Playing the Grünfeld: A Combative Repertoire 178483095X, 9781784830953

The Grünfeld Defence is well known to be one of Black's best and most challenging responses to 1.d4, and has long b

242 111 72MB

English Pages 504 [950] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Title Page
Symbols & Bibliography
Foreword by GM Nikita Petrov
Introduction
1 3.f3
2 Fianchetto System
3 Rare 4th & 5th Moves
4 Systems with e2-e3
5 Bg5 Systems
6 Bf4 Systems
7 Lines with Qa4†
8 Russian System
9 8.e5 & 8.Be2!
10 Rare Lines
11 5.Bd2
12 Exchange Variation
13 Be3 Systems
14 7.Nf3
15 7.Bc4
16 11.dxc5!
Variation Index
Recommend Papers

Playing the Grünfeld: A Combative Repertoire
 178483095X, 9781784830953

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Contents Title Page Symbols & Bibliography Foreword by GM Nikita Petrov Introduction Third Move Deviations 1 3.f3 2 Fianchetto System Sidelines 3 Rare 4th & 5th Moves 4 Systems with e2-e3 5 Bg5 Systems 6 Bf4 Systems 7 Lines with Qa4† 8 Russian System 9 8.e5 & 8.Be2! 4.cxd5 Nxd5 10 Rare Lines 11 5.Bd2 5.e4 Nxc3 6.bxc3 Bg7 12 Exchange Variation 13 Be3 Systems 14 7.Nf3 15 7.Bc4 16 11.dxc5! Variation Index

2

Table of Contents Title Page Symbols & Bibliography Foreword by GM Nikita Petrov Introduction 1 3.f3 2 Fianchetto System 3 Rare 4th & 5th Moves 4 Systems with e2-e3 5 Bg5 Systems 6 Bf4 Systems 7 Lines with Qa4† 8 Russian System 9 8.e5 & 8.Be2! 10 Rare Lines 11 5.Bd2 12 Exchange Variation 13 Be3 Systems 14 7.Nf3 15 7.Bc4 16 11.dxc5! Variation Index

3 5 7 9 25 63 136 166 216 302 375 411 461 544 593 623 689 747 834 898 940

950

A Combative Repertoire

Playing the Grünfeld By

Alexey Kovalchuk

Quality Chess http://www.qualitychess.co.uk

First edition 2020 by Quality Chess UK Ltd Copyright © 2020 Alexey Kovalchuk All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher. Paperback ISBN 978-1-78483-095-3 Hardcover ISBN 978-1-78483-096-0 All sales or enquiries should be directed to Quality Chess UK Ltd, Suite 247, Central Chambers, 11 Bothwell Street, Glasgow G2 6LY, United Kingdom Phone +44 141 204 2073 e-mail: [email protected] website: www.qualitychess.co.uk Distributed in North and South America by National Book Network Distributed in Rest of the World by Quality Chess UK Ltd through Sunrise Handicrafts, ul. Szarugi 59, 21-002 Marysin, Poland Typeset by Jacob Aagaard Proofreading by Colin McNab 3

Edited by Andrew Greet Cover design by Kallia Kleisarchaki and Adamsondesign.com Front cover photo by Kallia Kleisarchaki Photo page 5 by Dominic Cross Author photo on back cover by Mikhail Vikhrov

4

Key to symbols used ² ³ ± µ +– –+ = © „ ƒ ÷ ? ?? ! !! !? ?! ™ #

White is slightly better Black is slightly better White is better Black is better White has a decisive advantage Black has a decisive advantage equality with compensation with counterplay with an initiative unclear a weak move a blunder a good move an excellent move a move worth considering a move of doubtful value only move mate

Bibliography Avrukh: Grandmaster Repertoire 2A – King’s Indian and Grünfeld, Quality Chess 2018 Avrukh: Grandmaster Repertoire 8 – The Grünfeld Defence Volume One, Quality Chess 2011 Avrukh: Grandmaster Repertoire 9 – The Grünfeld Defence Volume Two, Quality Chess 2011 Dearing: Challenging the Grünfeld, Quality Chess 2005 Delchev: The Safest Grünfeld Reloaded, Chess Stars 2019 Hilton & Ippolito: Wojo’s Weapons Volume 3, Mongoose 2012 Kaufman: Sabotage the Grünfeld, New In Chess 2014 Kornev: A Practical White Repertoire with 1.d4 and 2.c4 Volume 2, Chess Stars 2013 Lemos: Opening Repertoire: The Fianchetto System: Weapons for White against the King’s Indian and Grünfeld, Everyman Chess 2014 Moskalenko: An Attacking Repertoire for White with 1.d4, New In Chess 2019 Pavlovic: Unknown Weapons in the Grünfeld, Thinkers 2016 Sakaev: How to Get the Edge Against the Gruenfeld, Chess Stars 2004 5

Sakaev: An Expert’s Guide to the 7.Bc4 Gruenfeld, Chess Stars 2006 Schandorff: Playing 1.d4 – The Indian Defences, Quality Chess 2012 Smith: e3 Poison, Quality Chess 2017 Svetushkin: The Ultimate Anti-Grünfeld: A Sämisch Repertoire, Chess Stars 2013 Electronic Resources ChessBase Mega/Correspondence Databases Roiz: Complete Repertoire for White after 1.d4 Nf6 2.c4 g6 3.g3, Modern Chess 2019 Marin: Positional Gruenfeld Repertoire (Parts 1, 2 & 3), Modern Chess 2017

6

Foreword by GM Nikita Petrov I first met Alexey at one of the countless junior tournaments in the south of Russia in which we competed. It was immediately obvious to me that he not only had a rare passion for chess, but was also a diligent worker. In fact, one may argue that Alexey’s scientific approach to chess is one of the reasons why he is not yet a grandmaster (although he will surely become one). However, his broad opening erudition and positional understanding certainly eclipse those of many strong GMs. The topic of the book you are holding in your hands does not really require an introduction. The Grünfeld Defence is not only an extremely popular opening at all levels of practical chess, but has also been analysed by many esteemed authors. Still, Alexey’s work is unique in the way it combines a stateof-the-art repertoire for Black with emphasis on typical Grünfeld middlegames. By guiding the reader through a number of carefully-chosen games in some of the most important variations, Alexey showcases a variety of key strategic concepts. Obviously, no serious research of such a topical opening can be possible without concrete analysis, and Alexey has presented innumerable novelties which readers may use to their benefit. I firmly believe that working on one’s openings yields the best results when studying theory is combined with developing a feel for the middlegame positions. This is arguably where the strongest point of Alexey’s work is. I am sure the reader will enjoy learning from Alexey, and that a lot of victories in the Grünfeld will be a well-earned reward.

Nikita Petrov Novorossiysk February 2020

7

8

Introduction

As you are almost certainly aware, the Grünfeld Defence arises after the opening moves 1.d4 Nf6 2.c4 g6 3.Nc3 d5 (or 3.Nf3 Bg7 4.Nc3 d5).

The Grünfeld is a top-class opening which has long featured in the repertoires of World Champions and other elite players. So how did it originate and what makes it such a tough opening for White to deal with? This introduction will answer those questions while offering a glimpse at what is to come in this book. Early History The first recorded occurrence of the above position dates back to 1855, when the Indian player Moheschunder Bannerjee played it against Scotland’s John Cochrane, in Calcutta. Bannerjee was unsuccessful and Cochrane, a strong attacking player whose 4.Nxf7 gambit against the Petroff still features in modern theory books, won by smothered mate in just 21 moves. Bannerjee is one of the reasons why 1.d4 Nf6 set-ups where Black refrains from occupying the centre with 1...d5 are collectively known as the Indian Defences. The database contains no more examples of our theme until 1922, when the following game took place.

9

Albert Becker – Ernst Grünfeld Vienna (4), 19.03.1922 1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.Nc3 d5 Legend has it Grünfeld was going to play a King’s Indian set-up with 4...d6, but he accidentally carried the pawn two squares ahead! Probably not true, but an amusing story nonetheless. 5.cxd5 Nxd5 6.e4 Nxc3 7.bxc3 c5 8.Be2 0-0 9.0-0 cxd4 10.cxd4 Nc6 11.Be3 Bg4 Following the appearance of Alekhine’s Defence in 1921, another new opening emerged with familiar ideas: having allowed White to occupy the centre with his pawns, Black attacks them using pieces and pawns. The following year was when Reti first started playing the opening that bears his name (1.Nf3 d5 2.c4), so the early 1920s was truly a golden era for the development of hypermodern opening systems.

12.d5 Ne5 13.Rb1 Nxf3† 14.gxf3?! Bh3 15.Rxb7? There was no good reason for White to weaken his kingside structure – and certainly not to give up the exchange. 15...Bxf1 16.Bxf1 Qd6 17.Bxa7 Be5 18.h3

10

18...Qa3? This squanders most of Black’s advantage. 18...Rfc8! would have activated Black’s last piece while preventing Bc4. The a-pawn will soon fall, and Black should have no trouble converting his material advantage. 19.Be3 Rfb8 20.Rxb8† Rxb8 21.Bc4 Qc3 22.Bb3 Qc7 23.Kg2 Bf4 24.Bd4 Be5 25.Be3 ½–½ Grünfeld played the same opening three more times in 1922 against strong opponents: he made a solid draw against Sämisch, outplayed and beat Kostic from a level position, and even came back from a worse position against Alekhine, exploiting some errors from the future World Champion to score an impressive victory. Subsequent Developments The idea of ceding the centre did not comply with the prevailing chess principles of the early 1920s, and Black’s opening was considered risky to say the least. Traditionalists attempted to refute it and in many games Black experienced some kind of problem in the opening, but the Grünfeld Defence was in no hurry to leave the foreground and move to the archive. Thanks to the efforts of Botvinnik, Smyslov, Simagin, Flohr and other leading players and theoreticians, the 1930s saw the Grünfeld enriched with many valuable strategic ideas. As with all openings, the process of discovery has continued over the decades, with new plans, resources and refinements being discovered for both sides. Why play the Grünfeld? 11

For almost a hundred years now, the Grünfeld has proven its resilience and vitality at the highest level of competition, with modern engine analysis confirming the absolute correctness and soundness of Black’s play. The lack of symmetry in the main variations predetermines a full-blooded battle, and the prospects of a sharp middlegame and/or a double-edged endgame has attracted such virtuosos as Stein, Korchnoi, Fischer and Kasparov to Black’s cause. The list of great players who have played the Grünfeld is practically endless; however, among modern grandmasters, Peter Svidler, Alexander Grischuk, Ian Nepomniachtchi and Maxime Vachier-Lagrave (MVL) are perhaps the most prominent of the elite players who specialize in it and whose games are worth studying. Although strategic principles are important, the Grünfeld gives rise to numerous sharp, forcing variations, many of which have been analysed through to the endgame. Lazy chess players will have a hard time in such a battleground, with either colour. However, the rewards are more than worth the effort: a well-prepared Grünfelder has excellent chances to seize the initiative and win convincingly against weaker players – and against stronger players, you can resist by virtue of the fact that you are playing high-quality, purposeful moves. White’s task of finding an opening advantage against the Grünfeld has become increasingly difficult over the years, and the fact that “the Anti-Grünfeld” (typically beginning with 1.Nf3 Nf6 2.c4 g6 3.Nc3) has become a recognised opening already says a lot. Some strong players have reverted to weird ideas including an early h2-h4 or g2-g4; but as we will see, a well-prepared Grünfelder should be delighted to encounter such moves. Strategy Although the Grünfeld Defence may give rise to different pawn structures, by far the most characteristic of them is shown below:

12

Having allowed the opponent to create a substantial pawn centre, Black will attack it with all his might. The dark-squared bishop eyes the d4- and c3-pawns, and ...c5 will increase the pressure. The b8knight often goes to c6, although this will depend on whether or not d4-d5 is a good reply. If White defends the d4-point with an early Nf3, Black will often pin it with ...Bg4. Other possibilities for Black involve ...Qa5 to attack c3, and possibly ...Rfd8 after castling. Obviously White has his own resources and there are various ways in which he can develop his pieces and support his centre. In the remainder of this introduction, I will offer a short preview of each of the sixteen chapters that comprise the repertoire I am advocating in this book. Chapter 1 1.d4 Nf6 2.c4 g6 is the starting point for the book. Before we arrive at the Grünfeld proper, the first chapter deals with the popular 3.f3, preparing e2-e4 without allowing the characteristic Grünfeld structure after ...d5 followed by a knight trade on c3.

My recommendation here may seem controversial as I prefer to avoid the ‘consistent’ 3...d5, after which 4.cxd5 Nxd5 5.e4 forces the knight to retreat to the somewhat passive b6-square. Some strong players continue to defend Black’s cause in that variation, but I like White’s chances. Therefore I have instead opted for 3...c5, which has attracted several of the world’s leading players. This usually transposes to a Modern Benoni with White having committed his pawn to f3, and the emerging positions promise a dynamic battle with mutual chances. Chapter 2 Next we will look at the Fianchetto System, which arises after 1.d4 Nf6 2.c4 g6 3.g3.

13

White hopes his bishop will exert a strong influence from g2, so I recommend blocking it with the reliable 3...c6 followed by ...d5, as played by various experts including Nepomniachtchi and MVL. Despite the possibility of a symmetrical structure, many of the variations are far from boring, and the chapter contains a number of sacrificial and other dynamic ideas. Chapter 3 Next we move on to the main Grünfeld tabiya of 1.d4 Nf6 2.c4 g6 3.Nc3 d5, starting with offbeat continuations such as 4.f3?! and 4.h4?!. Such moves are dubious and justifiably rare, but the recently fashionable 4.Nf3 Bg7 5.h4!? is more respectable.

14

Read on to see why I find 5...c6! to be the optimal reply. Chapter 4 Next we consider various set-ups involving an early e2-e3, the most common being 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.e3 Bg7 5.Nf3.

White can continue in various ways and Black will react accordingly. This chapter will also consider move orders where White delays the development of either knight, leading to different plans and nuances. Chapter 5 Next we will look at systems connected with an early Bg5. Beginning with 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Bg5, I recommend 4...Bg7!?, immediately offering a pawn sacrifice: 5.Bxf6 Bxf6 6.cxd5 (6.Nxd5?! Bg7 gives Black fabulous compensation) 6...c6!

15

As we will see, White can either hang on to his extra pawn or return it, but Black should be fine in all variations. The more common variant is 4.Nf3 Bg7 5.Bg5, when 5...Ne4! is our choice.

As we will see, White may retreat his bishop with 6.Bf4 or 6.Bh4, and it is important to know the differences between them and how Black should vary his response accordingly. Finally, 6.cxd5 Nxg5 7.Nxg5 e6 leads to a different type of battle where Black should be at least equal with the two bishops. Chapter 6 Next we will move on to lines with an early Bf4. Once again, White can play this move with or without 16

the knight on f3. First we will analyse 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Bf4 Bg7 5.e3 0-0, when a few moves are possible but 6.Rc1 Be6! will be our main line.

Black must be ready for all kinds of positions, as 7.cxd5, 7.c5 and 7.Qb3 are all valid options, but in each case I offer clear recommendations to show how Black gets a fine game. Next we will analyse the 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Bf4 variant. 5...0-0 is natural, when the main line of 6.e3 should be met energetically with:

6...c5! The main idea is 7.dxc5 Ne4! when the ensuing complications are by no means unfavourable to Black.

17

Chapter 7 This small chapter deals with a few quirky queen checks. After 1.d4 Nf6 2.c4 g6 3.Nc3 d5 White can try the immediate 4.Qa4†!?, although 4.Nf3 Bg7 5.Qa4† is the more common variant, known as the Flohr Variation. With 5...Bd7 6.Qb3 White tries to break Black’s harmony, having lured the bishop to what he hopes will prove an inferior square, but the opposite may prove true after:

6...dxc4 7.Qxc4 0-0 Depending on how White continues, Black may play a quick ...b5 or begin with ...Na6 and ...c5, with excellent counterplay. Chapter 8 Next we move on to the Russian System, one of White’s most challenging options against the Grünfeld. Usually it involves a knight on f3, although we will also examine lines after 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Qb3!? dxc4 5.Qxc4 Bg7 6.e4 0-0 where White leaves the royal knight at home and develops in other ways. The main variant arises after 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0.

18

From here we will examine 7.Bf4 followed by the main line of 7.e4, when various moves have been tried but I believe 7...a6! to be the best of them. A nice point is that 8.Bf4 can be met by 8...b5!, sacrificing the c7-pawn for the initiative. Chapter 9 The next chapter completes the repertoire against the Russian System by analysing White’s two most challenging options. 8.e5 is the more overtly aggressive of them, when 8...b5 9.Qb3 Nfd7 reaches another branching point where the most important moves are 10.h4!?, 10.Be3 and 10.e6. The reader should study every option carefully, as they each contain specific nuances. 8.Be2! is the latest trend and leads to a more positional game, with 8...b5 9.Qb3 c5! 10.dxc5 Be6! 11.Qc2 Nbd7 the critical line. Black will regain the c-pawn with a long manoeuvring game ahead; there are many nuances and challenges, but Black’s chances are no worse. Chapters 10 & 11 Both of these chapters discuss certain sidelines after 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5, which is why I have grouped them together.

19

Chapter 10 covers an assortment of variations, including 5.Nf3 Bg7 6.Bg5!?, which I call the Ulko Variation and which Shakhriyar Mamedyarov loves to play. However, after the standard 6...c5 7.Rc1, I will demonstrate that the ambitious 7...Nc6! is an excellent move. Another option is 5.g3, leading to a deferred version of the Fianchetto Variation. However, the presence of the knight on c3 makes life easier for Black, who continues 5...Bg7 6.Bg2 Nxc3 7.bxc3 c5 with convenient pressure against the enemy centre. The main subject of the chapter is 5.Na4!?, Nadanian’s system, which I recommend meeting with the active 5...e5!?. Chapter 11 continues the theme by analysing 5.Bd2, an important option which prepares to recapture on c3 with the bishop. After the natural 5...Bg7 6.e4, some Grünfelders try to avoid White’s scheme by retreating the knight, but I will show that the straightforward 6...Nxc3 7.Bxc3 0-0 gives Black fine play. Chapter 12 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 reaches a hugely important tabiya which forms the basis of the rest of the book.

20

This chapter shows how to meet such sidelines as 7.Ba3, 7.Bg5, 7.Qa4†!? and 7.Bb5†!?, which might easily confuse an unsuspecting victim. We will review each continuation, setting exact guidelines as to how Black should react. Chapter 13 The next chapter deals with systems involving the early development of White’s bishop to e3, starting with 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Be3 when White delays the development of the royal knight. Black obtains good play after 7...c5 8.Qd2 Qa5, with typical pressure against the enemy centre. 7.Nf3 c5 8.Be3 is the more popular and challenging variant, when 8...Qa5 is our move.

21

From here, both 9.Nd2 and 9.Qd2 0-0 10.Rc1 deserve close attention. In the latter case, I heartily recommend 10...Nd7!, a favourite continuation of Nepomniachtchi. Chapter 14 Next we cover the important position arising after 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Nf3 c5. After checking a bunch of sidelines, we will focus on 8.Rb1, a variant with a rich history.

After the standard 8...0-0 9.Be2 we reach an important juncture, where 9...Nc6! 10.d5 Ne5 is my recommendation. As with all the major variations, Black should come to the board well prepared, as the 22

price of a mistake may be high. However, the same is true for White, so your efforts will be rewarded. Chapter 15 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 is the final option which needs to be considered. White’s last move prepares to develop the knight to e2, rather than f3 where it invites a pin. 7...c5 8.Ne2

Here we play 8...Nc6, when 9.Be3 0-0 is normal. From this position, we will see how to handle White’s ideas involving delaying castling, before turning our attention to the most popular 10.0-0. This reaches another critical position, where 10...b6! is one of the most important theoretical advancements in recent Grünfeld history. Less than a decade ago this move was considered dubious, but then in 2011 Grischuk played a novelty against Aronian which caused a reassessment, and now it is the usual choice among players ‘in the know’ – which now includes you. The remaining part of the chapter deals with various ideas White may try, from the aggressive 11.h4!? to the solid 11.Qd2. Chapter 16 This final chapter concludes our coverage of 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.0-0 b6! by placing the critical 11.dxc5! under the microscope.

23

11...Qc7! is the essential reply, and if White exchanges on b6 then Black will obtain at least enough compensation for the pawn, due to his pressure along the queenside files. Instead, 12.Nd4! Ne5! 13.Nb5 Qb8 is the critical line, when 14.Bd5 and 14.Be2 both require careful study. *** I hope that readers will enjoy the book, and most of all enjoy playing this wonderful chess opening. The Grünfeld Defence is an indestructible shield against 1.d4; and no less important is the fact that Black will not merely sit in the trenches hoping for a draw, but will strive for an active, dynamic game with chances to seize the initiative. I wish good luck and many victories to you, dear reader! The author is grateful to Sergey and Alexander Osipov, Artem Solovyov, Alexander Shurov, Gleb Apryshko, Jacob Aagaard, Andrew Greet and Renier Castellanos for their invaluable assistance on this project. I dedicate this book to my son, Matvey Kovalchuk. Alexey Kovalchuk Torez, February 2020

24

A) 4.dxc5!? 18 B) 4.d5 d6 5.e4 e6 19 B1) 6.Ne2!? 19 B2) 6.Nc3 exd5 7.cxd5 Bg7 22 B21) 8.Be3 22 B22) 8.Nge2 26 B23) 8.Bg5 31

1.d4 Nf6 2.c4 g6 We begin our journey by investigating White’s alternatives to 3.Nc3, 3.Nf3 and 3.g3. By far the most important of them is: 3.f3 First seen in Nimzowitsch – Tartakower, Carlsbad 1929, this move was subsequently adopted by World Champion Alekhine against Bogoljubov. White is in no hurry to develop either of his knights, and wants to immediately occupy the centre with e2-e4. Black must decide between playing ...d5 anyway, enabling White to play cxd5 and e2-e4 when there is no knight trade on c3, and switching to some other approach such as a King’s Indian or Modern Benoni 25

set-up. White’s last move was recommended by Svetushkin in The Ultimate Anti-Grünfeld and Kaufman in Sabotage the Grünfeld. In modern practice, Aronian, Ding Liren, Vitiugov, Cheparinov and Nakamura are just a few of the strong players who use it. White’s other options do not require too much attention: 3.Bg5 can be met either by 3...Bg7, when White has nothing better than 4.Nc3 d5 transposing to Chapter 5, or by 3...Ne4 with a promising version of the 2...Ne4 Trompowsky. 3.d5 This prevents 3...d5 but is a violation of opening principles, as White moves a pawn for a second time while neglecting his development. Black gets a good game with any sensible continuation, but my favourite idea is: 3...b5! Avrukh recommends this move in his Grandmaster Repertoire 8. Having studied the position, I came to the conclusion that this continuation is the most promising, as Black obtains a much-improved Benko Gambit. 4.cxb5 a6 5.bxa6 Avrukh also mentions 5.b6N 5...c6 6.Qb3 Bb7 “with an interesting game.” I continued for a few more moves and found a remarkable idea: 7.dxc6 Nxc6 8.e3 Na5! 9.Qb4

9...Nc4!! A beautiful move: the knight cannot be taken either by the queen due to the ...Rc8 skewer, or by the bishop due to the hanging g2-pawn and rook on h1. After 10.Nc3 Nxb6³ Black enjoys a lead in development and the e3-pawn restricts the bishop on c1, so Black has a small but persistent advantage. 5...c6 Black is in no hurry to recapture on a6, but instead attacks White’s centre by using the ‘hook’ on d5. 26

Sam Shankland’s book Small Steps to Giant Improvement contains a lot of good advice about utilizing pawn hooks. 6.dxc6 6.Nc3 cxd5 7.Nxd5 Qa5† 8.Nc3 Bg7 9.Bd2 Qxa6!?N prepares to transfer the queen (Avrukh’s recommendation of 9...Bxa6 should also be fine but I prefer the text move). 10.e3 Qb7© Black has good compensation for the pawn, with pressure on the queenside while White’s dark-squared bishop is limited by the e3-pawn. 6...Nxc6 7.e3 7.Nc3 Bxa6 8.g3 Bg7 9.Bg2 0-0 10.Nh3 occurred in Dol – Van Son, Netherlands 1989, when 10...e6N 11.0-0 d5 12.Nf4 Qb8© would have given Black good Benko-style compensation for the pawn. 7...Bg7 8.Nf3 8.Nc3 0-0 9.Nf3 Qa5 10.Bd2 Bxa6© again gives Black compensation thanks to his play along the aand b-files and the strong bishop on g7. 8...0-0 9.Be2 Bxa6 10.Bxa6?! White should have preferred 10.0-0 Ne4© although Black has good pressure against White’s undeveloped queenside. 10...Qa5† 11.Nc3

11...Qxa6!N I believe this move offers Black more winning chances. The tempting 11...Ne4 should be met by 12.Bb7!N (Avrukh only considers 12.0-0?! Nxc3 13.bxc3 Rxa6³ as seen in Dzindzichashvili – Adorjan, Amsterdam 1978) 12...Nxc3 13.Qd2 Ne4 14.Qxa5 Rxa5 15.Bxc6 dxc6= when Black has compensation, though the c6-pawn is also weak. Black will most likely eat the queenside pawns, but during this time the c6-pawn should fall and, due to the limited material, the game is likely to end in a draw. 12.Qe2 Rfb8 13.Qxa6 Rxa6³ 27

Black has more than enough compensation for the pawn, due to his lead in development and powerful pressure along the a- and b-files. 3.Qc2!? This cunning move has occasionally been used by strong players as a surprise weapon. 3...Bg7 Black should be happy to switch to a King’s Indian defence, since White has decided much too early on the placement of the queen. 3...d5?! 4.cxd5 Nxd5 5.e4 Nb6 6.Nf3 would justify White’s 3rd move. However, 3...c5 is a good alternative. 4.e4 0-0 5.Nc3 d6 6.Be2 6.Nf3 Nc6 7.d5 Nb4 8.Qd1 e6 9.Be2 Qe7 10.Bg5 h6 11.Bxf6 Qxf6÷ occurred in Rotaru – Heydt, corr. 2015. Black is certainly not worse and has the long-term advantage of the bishop pair. 6...Nc6 7.Be3 e5 8.dxe5 8.d5 Nd4 9.Qd2 c5 10.dxc6 bxc6 11.Rd1 d5 12.exd5 cxd5 13.cxd5 Bf5µ occurred in Rumyantsev – Fritsche, corr. 2013; despite the extra pawn, White already had to think about saving the game. 8...dxe5 9.Rd1 Nd4 10.Bxd4 exd4 11.c5 Qe7 12.Rxd4 Qxc5µ White’s opening experiment had failed, as Black already enjoyed a lead in development as well as the bishop pair in Koneru – E. l’Ami, Wijk aan Zee 2006. 3...c5 This move usually leads to a Modern Benoni, with White having committed to the Sämisch System. In recent years, this continuation has been favoured by such Grünfeld aficionados as Nepomniachtchi, Vachier-Lagrave, Grischuk, Vidit and other strong grandmasters. The other big main line is 3...d5 4.cxd5 Nxd5 5.e4 when Black must retreat his knight to b6, having been deprived of the opportunity to exchange on c3. I think White has good chances to obtain some advantage here. We will analyse A) 4.dxc5!? followed by the normal B) 4.d5. A) 4.dxc5!? e6 5.Be3

28

5...b6! Black does not bother trying to regain the pawn and instead fights for the initiative. 6.cxb6 Another game continued: 6.Nc3 Ba6 7.cxb6 axb6 8.Bg5 (8.b3 Nc6 9.Bg5 h6 10.Bd2 Bg7 11.e4 d5³ gives Black excellent compensation for the pawn due to his lead in development and the more active location of his pieces) 8...h6 9.Bh4 Now in Caruana – Grischuk, Baku 2014, Black could have played:

9...Bxc4!N 10.Qd4 g5 11.Qxc4 gxh4 12.Qxh4 b5© Black has excellent compensation for the pawn, with ...b4 coming next. Note that 13.Nxb5? Qa5† 14.Nc3 Nd5µ only adds to White’s problems. 29

6...axb6 7.Bg5 7.Nc3 occurred in Bocharov – Swiercz, Moscow 2016, when the improvement 7...Ba6N transposes to the previous note. 7...h6 8.Bh4 This occurred in Shankland – Svidler, Hoogeveen (1) 2018. I recommend:

8...Nc6!N 9.Nc3 Bg7 10.e4 I also considered 10.e3 0-0 11.Nge2 Ba6 12.Nb5 d5 13.Nec3 Na5 14.cxd5 g5 15.Bg3 Nxd5ƒ when Black is clearly ahead in development and his initiative is unpleasant to deal with. 10...0-0 11.Nh3 d5 12.cxd5 exd5 13.Nf2 dxe4ƒ Black has more than enough compensation for a pawn. An illustrative line is: 14.Qxd8 Rxd8 15.Nfxe4 Nxe4 16.fxe4 g5 17.Bf2 Nb4 18.Rc1 Nxa2 19.Nxa2 Rxa2µ Black has a serious advantage. B) 4.d5 This is the most natural and common continuation. 4...d6 Black is also ready to fight for the centre. The d4-d5 advance has its pluses and minuses: White establishes a space advantage but Black’s dark-squared bishop will have a large operational space on 30

the h8-a1 diagonal. 5.e4 e6 Black gets ready to exchange on d5. White has two main modes of development: he can play B1) 6.Ne2!? intending to put the king’s knight on c3, or play more conventionally with B2) 6.Nc3. 6.Bg5 Bg7 leaves White with nothing better than 7.Nc3, transposing to variation B23 on page 31. Alternative knight developments do not combine well with the bishop on g5. For instance, 7.Qd2 exd5 8.cxd5 0-0 9.Ne2 b5 10.a4 b4 11.Ng3 Nbd7 12.Bh6 occurred in Klink – Colpe, Lueneburg 2017, and now Black could have seized the initiative with:

12...Nh5!N 13.Bxg7 (13.Nxh5 Qh4† is excellent for Black) 13...Nxg7 14.Bb5 f5³ Black has good play in the centre and his plans may include ...f4 and ...Ne5. B1) 6.Ne2!? This knight is often a problem figure in White’s set-up, so it heads for c3, allowing the other knight to go via d2 to c4. 6...Bg7 Black does not hurry to remove the tension in the centre, but continues the development of the kingside 7.Nec3 a6 Now Black wants to play ...exd5, and on cxd5 then ...b5. Therefore White’s next move is almost forced. 31

8.a4 Ignoring ...b5 does not lead to anything good for White. For instance: 8.Bg5 h6 9.Bh4 exd5 10.cxd5 b5 11.Qc2 g5 12.Bf2 Nbd7 13.a4 b4 14.Nd1 Ne5 15.Nd2 Nh5 16.Ne3 0-0 17.Bc4 Now in Laine – Le Page, corr. 2015, Black’s best continuation would have been 17...Nf4N 18.0-0 h5ƒ with the initiative on the kingside, which in the future could turn into an attack.

8...Nh5! A great move! Black not only threatens ...Qh4†, but also prepares to undermine the centre with ...f5. 9.Be3 White prepares Bf2 in case of a queen check. 9.Bd3? is obviously worse. 9...Qh4† 10.Ke2 occurred in Binas – Hostinsky, corr. 2016, when the simple 10...Nf4†N 11.Bxf4 Qxf4 12.Qd2 Qh4µ would have left White weak on the dark squares and with his king misplaced in the centre. 9...exd5 10.cxd5 f5! Immediately attacking the centre and preparing a possible ...f4 followed by ...Nd7-e5, forming a dark-square blockade. 11.exf5 White to some extent surrenders his control of the centre, but on the other hand avoids being smothered by ...f4. 11.Nd2 0-0 12.Bd3 Nd7 13.exf5? Re8! 14.Kf2 enabled Black to win with a not-too-difficult tactic: 32

14...Rxe3! 15.Kxe3 Bd4† 16.Ke2 Nf4† 17.Kf1 Nxd3–+ With an absolutely winning position, Goldbeck – Chatalbashev, Bamberg 2018. Another game continued: 11.Qd2 0-0 12.Be2 I also checked 12.g3N but White still cannot avoid an unpleasant blockade in the centre: 12...Nd7 13.Be2 f4! 14.gxf4 Qh4† 15.Bf2 Qxf4µ The e5- and f4-squares are completely controlled by Black. 12...f4 13.Bf2 b6 14.g4 Nf6 15.Bh4 We have been following Walczak – Nekhaev, corr. 2016. Here I prepared the improvement:

15...Nxe4!N 16.Bxd8 Nxd2 17.Nxd2 Rxd8 18.Nc4 Nd7 19.Nxd6 Be5 20.Nce4 Bxd6 21.Nxd6 Ne5 22.Nxc8 Raxc8³ 33

Black has a great knight on e5 against a bad white bishop, and is also structurally better, as the d5pawn is weak. White must therefore fight for equality and his path is not strewn with roses. 11...gxf5 12.Qd2 White constantly needs to think about the ...f4 advance. However, the problem with the text move is that the b1-knight loses its best square. As Tarrasch said: “If one piece is bad, the whole position is bad.” Trying to radically solve the problem with 12.f4 leads to a different set of problems after: 12...Qh4† 13.Bf2?! (This loses a pawn; 13.Kd2 would be the lesser evil although after 13...Nf6ƒ White has some problems due to his king in the centre) 13...Qe7† 14.Qe2 Nxf4 15.Qxe7† Kxe7 16.g3 Ng6µ White had no compensation for the pawn in Garriga Cazorla – Antoli Royo, Mislata 2014. 12...0-0 13.Be2

13...f4!N 13...Nd7?! 14.0-0?! f4! 15.Bf2 was the actual move order of the Nakamura – Bok game which we will return to below. However, this sequence gave White the opportunity to improve his situation with 14.Bg5!N 14...Qc7 15.Bh6 Ne5 16.Bxg7 Qxg7 17.0-0 f4÷ although even here Black is still not worse. The text move enables Black to seize the initiative while keeping his powerful dark-squared bishop. 14.Bf2 Nd7 15.0-0 Qg5 Connecting the queen to the attack. 15...Ne5!?N 16.Ne4 Kh8 17.Nbc3 Qe7 18.Kh1 Bf5ƒ also gives Black some initiative. 34

16.g4?! This overambitious move gives Black the key resource of a future piece sacrifice on g4 to open the kingside. The calmer 16.Kh1 would have been preferable, although 16...Ne5 17.Ne4 Qg6 18.Nbc3 Bf5³ still favours Black, as he can gradually develop his initiative on the kingside 16...Ne5 17.Kh1 Nf6 18.Rg1 Qg6 19.Bh4 Now Black has a choice of improvements.

19...Nfxg4!N This is the most convincing and flashiest continuation. The game continuation of 19...h5?! gave White the chance to improve his situation with 20.gxh5!N (rather than 20.g5? Nfg4! when White was in trouble in Nakamura – Bok, Isle of Man 2016) 20...Qxh5 21.Qxf4! Nfg4 22.fxg4 Rxf4 23.gxh5 Rxh4 24.Nd2 Kh8 when the position is unclear. 19...Kh8!?N is a promising alternative, showing that Black does not have to break through immediately to claim a good position. For instance: 20.Qe1 (20.Na3? enables Black to launch a crushing attack: 20...Nfxg4! 21.fxg4 f3 22.Bf1 f2 23.Rg2 Bxg4 24.Bxf2 Bf3–+) 20...Qf7 21.Bxf6 Qxf6 22.Nd2 Bd7³ Black is better because of the two bishops and the more active arrangement of his pieces. 20.fxg4 Kh8!! Amazingly, Black can follow up the piece sacrifice by developing his initiative slowly, as White can do nothing to oppose it. 35

21.Bd1 21.Na3? was covered under 19...Kh8!?N above. 21...Bd7 22.Na3 Rae8 23.Be1 Bxg4‚ The game is not yet over, but Black is the clear favourite due to his powerful attack. B2) 6.Nc3

This is the main move. White develops the knight to its most natural square before deciding what to do with his other pieces. 6...exd5 7.cxd5 Bg7 Now White has a choice. B21) 8.Be3 is a natural developing move which preserves the option of Nh3-f2 if Black allows it. Next we will consider B22) 8.Nge2 with the idea of Ng3, and finally the provocative B23) 8.Bg5. B21) 8.Be3 0-0 9.Qd2 White’s set-up resembles the Yugoslav Attack against the Dragon, and his kingside attacking potential should discourage Black from plans such as ...Ne8-c7 to prepare ...b5 (which can work well against other set-ups, as we will see in variation B22 below). Instead, Black will play in the centre and on the kingside with ideas such as ...Nd7-e5, ...Nh5 and ...f5, with typical counterplay. 9...a6 First we threaten ...b5, which White should prevent as usual.

36

10.a4 Nbd7! It used to be considered inaccurate to allow Nh3-f2, but it turns out that this set-up promises White less than was previously thought. 11.Nh3 This is the principled attempt to exploit Black’s last move. 11.h4?! Ne5 12.Bh6 saw White launch an unprepared attack in Sladek – Bardoux, Brno 2005. A good response is 12...Bxh6N 13.Qxh6 Qb6 14.Ra2 c4 when White’s attack is not really scary, for instance:

15.h5? Nxh5 16.g4 This typical attacking try is refuted by 16...Bxg4! 17.fxg4 Nxg4 18.Qd2 Ng3–+ when White’s position is falling apart. 11.g4?! Re8 12.h4 Ne5 13.h5 was another unprepared action in Rettig – Kobs, Wuerzburg 1987. I recommend the improvement:

37

13...b5!N 14.axb5 Bxg4! 15.fxg4 Nexg4 16.Bg5 axb5 Now it becomes clear why Black’s 13th move was needed: the open a-file falls into Black’s hands and helps to develop his attack. 17.Rxa8 Qxa8 18.hxg6 fxg6 19.Bxf6 Bxf6 20.Bg2 Qa1† 21.Nd1 Bd4 22.Ne2 Bf2† 23.Kf1 Rf8–+ Black has an absolutely won position. This and the previous line show the importance of developing properly before launching an attack. This positional rule is valid in most cases. 11.Bd3 runs into 11...Ne5 when White loses the bishop pair – and as the saying goes, “The worst bishop is better than the best knight.” 12.Nge2 Nxd3† 13.Qxd3 Nd7 14.f4 occurred in Othman – Mohd, Abu Dhabi 1995, when Black should have continued:

14...Nf6N 15.0-0 Bd7 16.a5 Bb5! 17.Qc2 Qe7µ White has a weak pawn on e4 while Black has the advantage of two bishops. It is not profitable for White to exchange on b5, as Black’s pawn structure 38

will be significantly improved. The most sensible alternative to the main line is: 11.Nge2 Ne5 12.Ng3 h5 13.Be2 13.Bg5 is an attempt to guard the h4-square but... 13...h4! It does not matter! 14.Bxh4 Nxe4 15.Bxd8 Nxd2 16.Bc7 In Lapiccirella – Vocaturo, Bratto 2006, the improvement 16...Nec4!N 17.Be2 Bd7 18.Bxd6 Nxf3† 19.gxf3 Nxd6µ would have been clearly better for Black because of the two bishops and the fact that White’s pawn structure is seriously weakened and will be vulnerable for a long time. 13...h4 14.Nf1 Re8 15.Bg5 Another game continued 15.h3 Nh5! 16.f4 Nd7 17.Bxh5 gxh5 18.g4 hxg4 19.hxg4 Nf6 20.Qg2 when, in Terry – R. Wells, England 2000, 20...Nxg4N 21.0-0-0 f5 22.e5 dxe5 23.d6 Bd7–+ would have been hopeless for White. 15...h3 16.g4 Bd7 17.Rg1 b5! 18.Rg3 Here I found an improvement on Black’s play from Van den Haak – Douglas, corr. 2006.

18...Qa5!N 19.Bh6 b4 20.Bxg7 Kxg7 21.Rxh3 I also considered 21.Nd1 Nexg4! 22.fxg4 Nxe4 23.Qf4 b3† 24.Nc3 Nxg3 25.Nxg3 Qb4µ when Black is doing great. 21...Nfxg4! 22.fxg4 bxc3 23.Qh6† Kf6 24.Qf4† Ke7 25.bxc3 Rab8 26.Qg5† f6 27.Rh7† Kd8 28.Qd2 Bxg4µ An amazing position! Black is clearly better, since his knight is excellently placed and his rooks apply pressure along their respective files. 11...Ne5 The knight is both perfectly located and hard to drive away. White would like to chase it off with f3f4 but then ...Ng4 will follow.

39

Playing for ...b5 is inferior: 11...Rb8?! 12.Nf2 Ne8 13.Be2 Nc7 14.h4! Unexpectedly, White can start attacking on the kingside. 14...b5 15.axb5 axb5 16.Bh6 b4 17.Ncd1 Bxh6 18.Qxh6 Nf6 19.h5‚ Black’s king is in real danger. 12.Nf2 Nh5! This idea is familiar idea to us: Black wants to play ...f5 and maybe even ...f4. 13.Be2 f5

14.exf5 This continuation has the disadvantage of weakening White’s control over the centre to some extent. However, White had to do something about the threat of ...f4. An aggressive alternative is: 14.f4!? Ng4 15.Bxg4?! White wants to keep the well-placed knight on f2, but trading off the bishop for a knight seems dubious. 15.Nxg4 fxg4 16.g3 Re8 17.Bd3 was played in Mikrut – Wachinger, Katowice 2017, when the natural 17...Bd7N 18.a5 b5 19.axb6 Qxb6 20.0-0 Qb4„ would have given Black plenty of play in the centre and on the queenside. 15...fxg4 16.g3 Bd7 17.a5 17.Qe2 b5 18.0-0 b4 19.Ncd1 Qe8 20.Qc2 occurred in Fus – Choroszej, Warsaw 2016, when Black missed the strong idea of 20...Nf6!Nµ intending ...Rf7-e7. 17...b5 18.axb6 Qxb6 19.0-0 Rfe8 20.Ra3 Rab8 21.Rc1 a5 22.Rc2 Qb4 23.Qc1 We have been following Ionescu – A. Vovk, Mamaia 2015. Here we can strengthen the play of the Ukrainian grandmaster by means of:

40

23...Bd4!N 24.Bxd4 cxd4 25.Nb1 Nf6 26.Nd2 a4 27.Nd3 Qb7 28.Rc7 Rec8! 29.Rxb7 Rxc1† 30.Nxc1 Rxb7 31.Nd3 Rc7µ Black is obviously better as he controls the open file and already threatens to invade on c2. 14...gxf5 Once again threatening ...f4. 15.Bh6 In principle, it is in White’s interests to trade off the dark-squared bishops – partly because Black’s bishop tends to be the best minor piece on the board, and also because ...f4 was such an unpleasant threat. White has also tried nudging the black queen first: 15.Bg5 Qe8 16.Bh6 This was the choice of Kamil Miton, a strong GM. Avoiding the bishop trade is riskier for White: 16.0-0?! Qg6 17.Kh1 Bd7 18.Bh4?! (18.f4 Ng4³) 18...Bh6 19.Qc2 Nf4 20.Bg3 Nxe2 21.Bxe5 Rae8 22.Nxe2 Rxe5 23.Nc3 b5–+ S. Silva – Kozlov, corr. 2015. In spite of material equality, White’s position is hopeless due to Black’s two bishops against two knights and the weakness of the d5-pawn. 16...Bxh6 16...f4!?N is a valid alternative; after 17.Bxg7 Nxg7 18.0-0 Qg6 19.Rae1 Nh5 Black is not worse and the position remains complex, with chances for both sides. 17.Qxh6 Qg6 18.Qxg6† In Miton – Jacko, Slovakia 2014, Black should have played:

41

18...hxg6N 19.g3 Nf6 The position is balanced. A sample continuation is: 20.f4 Neg4 21.Nfd1 Re8 22.h3 Ne3 23.Nxe3 Rxe3 24.Kf2 Re7 A plausible plan for Black is ...Bd7, ...Rb8 and ...b5. 15...f4!? This move has both pluses and minuses. On the positive side, White will be deprived of the possibility of f3-f4. On the other hand, an outpost on e4 is now available to him. The flexible 15...Qf6N 16.Bxg7 Nxg7 17.0-0 Bd7÷ is equally playable:

The position remains complicated but I believe Black’s chances are no worse than White’s. A logical 42

plan for the next few moves would be ...Kh8 and ...Rg8, followed by organizing an attack on the white monarch. 16.Bxg7 Nxg7 17.0-0 Qh4 18.Nfe4? We are following a game where White made this serious mistake, taking an important defender away from his kingside. White should have favoured the other knight: 18.Nce4N 18...Nf5 19.Ng4! This is an important resource that helps White. 19...Nxg4 20.fxg4 Ne3 21.Rxf4! Rxf4 22.Qxe3 Rf8 23.h3 Bd7 24.Nxd6 Qf6 25.Qxc5 Qxb2÷ White has sufficient compensation for the material but no advantage, and all three results are possible. 18...Nf5 The e3-square is inconvenient to protect. White does not want to retreat his knight to the passive d1square, so he must get ready for the knight to land on e3 at any moment. 19.Rf2 Bd7 Black should not hurry with ...Ne3 as it would not threaten anything yet. Instead he develops another piece and supports a possible ...b5. 20.g3 Qh6 21.Kh1 In Bocharov – Kovalev, Khanty-Mansiysk 2016, Black should have improved his position with:

21...Kh8!N In the game Black got carried away with 21...b5?; the position does not require such tough actions.

43

22.gxf4 Ng6 23.Bf1 Qxf4³ Black has a small but steady advantage due to his better pawn structure and more active pieces, especially the knights. B22) 8.Nge2

White directs his knight to g3 where he can have two future routes: Nh1-f2 or Nf1-e3/d2, depending on what Black does. 8...0-0 9.Ng3 a6 10.a4 The previous moves were quite logical: Black castled and threatened ...b5, while White continued the knight manoeuvre and prevented Black’s queenside expansion. Usually White meets ...a6 with a2-a4 automatically in these positions, but I found a game where a former World Champion ignored it. 10.Be2 b5 11.0-0 was seen in Topalov – Krysa, Gibraltar 2017, when I recommend improving with:

44

11...b4!N 12.Nb1 Nfd7 13.Nd2 h5 14.Nh1 a5 15.Nf2 Nb6³ White’s pieces are passive. 10...Nbd7 11.Be2 White continues developing his kingside and does not determine the position of the dark-squared bishop. Depending on the circumstances, it may be placed on g5 or e3. Less cunning is: 11.Bg5 11.Be3 h5 is the same thing with one less move played. 11...h6 12.Be3 h5

45

13.Be2 13.Bg5 Qe8! is a useful idea to know about, preparing ...Nh7. White has two main ideas: a) 14.Be2 Nh7 15.Bf4 occurred in Pereira – Yang Kaiqi, Kuala Lumpur 2008. White’s dark-squared bishop is rushing around, but still it cannot find shelter. Black could have emphasized this with 15...Ne5!N 16.0-0 h4 17.Nh1 g5! 18.Bd2 f5³ when his kingside initiative already seems quite threatening. b) 14.Qd2 Nh7 15.Bh6 is similar to the note to Black’s 11th move in the main line below. However, the difference is felt after 15...h4 16.Bxg7 Kxg7 when the f1-square is unavailable, so White had to settle for the clumsy 17.Nge2 in Cooke – V. Rajlich, Harkany 1999. Here I propose 17...Ne5N 18.Nf4 Qe7³ when the knight on f4 is less than impressive, and Black’s plans include playing on the queenside with ...Bd7 and ...b5, as well as undermining the centre with ...f5. 13...h4 14.Nf1 Rb8 I considered the following attempt to improve White’s play:

15.Nd2N 15.g4 was met convincingly by 15...Nh7 16.Nd2 Bd4! 17.Bh6 Re8 18.Nc4 Qf6 19.Bc1 b5 20.axb5 axb5–+ in Pitkaenen – Le Page, corr. 2010. White’s position is simply hopeless due to his lack of development. 15...Nh5! 16.Nc4 It is important that 16.f4 runs into 16...Bd4! 17.Bxd4 cxd4 18.Bxh5 dxc3 19.bxc3 gxh5 20.Qxh5 Kg7 21.Qg4† Kh6µ when White’s compensation for the piece is not enough. 16...Ne5 17.Nxe5 Bxe5 18.Qd2 f5³ Black begins to develop some initiative on the kingside. White has problems finding a decent plan, and will soon have to switch to defence. 11.Bf4!? was the interesting continuation of Andresen – Emonot, email 1995. I recommend a novelty: 11...Ne5N 12.Be2 Nh5! Not fearing any deterioration of the pawn structure due to the dynamic 46

possibilities of the position. 13.Nxh5 gxh5 14.0-0 f5 15.Qd2 Ng6÷ with an obscure game in which Black has good chances to develop an initiative on the kingside. The front h-pawn will march ahead to h3, while White must watch for a possible blockade in the centre by means of ...f4.

11...Ne8!? With the typical Benoni plan of ...Nc7, ...Rb8 and ...b5. The reader may be wondering why we don’t play 11...h5 here. The trick is revealed, showing why White waited so long to develop his dark-squared bishop: 12.Bg5! Qe8 13.Qd2 Nh7 14.Bh6 Rb8 15.Bxg7 Kxg7 16.Nf1 Qe7 17.Ne3² The royal knight is perfectly employed. 12.0-0 12.Be3 Rb8 leaves White nothing better than 13.0-0, transposing. 12.Bf4!? deserves further attention; I only found one game with this move, Boehme – Bochev, email 2014. I recommend 12...Bd4!?N with the possible continuation: 13.Qd2 f5 14.exf5 (14.h4 fxe4 15.Ngxe4 Ndf6 16.Bg5 Qa5÷) 14...gxf5 15.Bc4 Ne5÷ There arises a complex position with mutual chances. 12...Rb8 13.Be3 White finally determines the position of the bishop. 13.Kh1 was seen in Dziuba – Ftacnik, Czech Republic 2017. The problem of White’s last move is that the Nh1-f2 manoeuvre is no longer available, so 13...h5!N would have made a lot of sense: after 14.Re1 Ne5 15.Nf1 f5„ Black has good counterplay. A top-level game continued: 13.f4 Nc7 14.e5 Nothing essentially changes after 14.Kh1 b5 15.e5, Basso – Vedmediuc, Cutro 2019, when 15...f5!N 16.axb5 Nxb5 17.e6 Nb6 18.Nxb5 axb5 19.Bxb5 Bb7÷ reaches a double-edged position. 47

Here I found an interesting way to improve on Black’s play from Aronian – Nepomniachtchi, Wijk aan Zee 2017:

14...f5!N 15.exf6 15.e6!? Nf6 is not easy to assess. My main line continues 16.Rb1 b5 17.axb5 Nfe8 18.Bf3 axb5 19.Re1 Qe7 20.Be3 b4 21.Na4 Ba6³ when, in spite of the fact that White has a ‘nail’ on e6, Black’s position looks preferable due to the more harmonious arrangement of his pieces. 15...Qxf6 16.Kh1 Qd8 17.Nce4 Nf6= In this complex position, Black’s chances are not worse due to his strong bishop on g7 and the potential weakness of the pawn on d5. In a later game, Nepo faced a different move: 13.Re1!? Ding Liren – Nepomniachtchi, Abidjan 2019. There are various ways to handle the position, but my favourite is: 13...Bd4†N 14.Kh1 Qh4 Threatening ...Bf2. 15.Rf1 f5 16.exf5 gxf5 17.Bd3 17.f4 Ndf6„ 17...Ne5 18.Nge2 Nxd3 19.Qxd3 Be5 20.g3 Qf6 21.Bf4 Bd7 22.Rab1 b5 23.axb5 Bxb5 24.Nxb5 axb5= Black is fine. 13...Nc7 Now we are ready for ...b5, so White must decide whether to interfere with this plan or to ignore it and play in the centre.

48

14.f4 Now White has some aggressive ideas in the centre, so Black needs to start counterplay on the queenside. An important alternative is: 14.a5 b5 15.axb6 Rxb6 16.Qd2 This occurred in Leenhouts – De Boer, Dieren 2018. I recommend an unusual move:

16...Qe8!N 17.f4 I also checked 17.Rab1 f5 18.exf5 gxf5 19.Bh6 Bxh6 20.Qxh6 Qg6÷ when the queen connects to the defence in time and Black is not worse. 17.Rfe1 Rb4 18.Bf4 Qe7 19.Rab1 Re8÷ reaches a rather complicated position which requires practical verification. However, it seems to me that the ...Nc7-b5-d4 manoeuvre offers Black excellent prospects. 17...Rb4 18.Kh1 Nb5 19.Bd3 Qd8„ Again we reach a complex position where Black has his fair share of the play. White can also ignore ...b5 and look to develop his play in a different way: 14.Qd2 b5 15.Bh6 15.axb5 Nxb5 16.Nxb5 axb5 17.Ra7 h5 18.Rfa1 h4 19.Nf1 h3 20.g3 Qf6 21.b3 c4 22.Rb1 c3 23.Qd1 b4³ turned out well for Black in Sawicki – Merilo, corr. 2012, and he went on to win with the help of his protected passed pawn. 15.Bg5 f6 16.Be3 b4 17.Nd1 f5 18.Re1 fxe4 19.Nxe4 Nf6 20.Ndf2 Nxe4 21.Nxe4 Bf5 22.Bd3 occurred in Sanchez Ortega – Gerhards, corr. 2009, when 22...Bxe4N 23.fxe4 Qe7³ would be my suggestion. The text move was played in Aronian – Firouzja, Internet 2019. Here I like the improvement:

49

15...Bxh6!N 16.Qxh6 Qf6 17.Nd1 c4„ Followed by ...Nc5, with a good position. 14...b5 Black seizes space on the queenside and has a number of possible ways to develop his play in that area. 15.e5!? White clears the e4-square for either of his knights, and is ready to meet ...dxe5 with f4-f5. Each side must play carefully in such a sharp position, where the price of every tempo is extremely high. Another logical possibility is 15.axb5 Nxb5 16.Nxb5 axb5 17.Qd2 as seen in Mindlin – Baron, Jerusalem 2017. Here I suggest 17...Nb6!?N when the knight is ready to go to c4. Play may continue: 18.b4 (a real war begins after 18.e5!? Bb7 19.Ne4 dxe5 20.Bxc5 Bxd5! 21.Bxf8 Qxf8 22.Nc3 Be6 23.fxe5 Bxe5 24.Kh1 Nc4© when Black’s powerful bishops fully compensate for the small material investment) 18...Bxa1 19.Rxa1 Nd7÷ White has enough compensation for equality, but no more than that. 15...bxa4! A great solution! Keeping the pawn structure together is less important than opening the b-file for the rook, which may eat the pawn on b2 and/or go to b4 to fight for some central squares. 16.Nge4! The strongest move. White should keep the other knight on c3, as it may go to d5 in some lines.

50

16.Nce4 is inaccurate as the g3-knight is not so active. 16...dxe5 17.f5 occurred in Hesham – Kantans, Batumi (ol) 2018, when I recommend improving with:

17...Rxb2N 18.d6 Nb5 19.fxg6 hxg6 20.Bg5 Nc3 21.Bxd8 Nxd1 22.Bg5 f5 23.Rfxd1 fxe4³ White has some compensation for the material, but it’s obvious that he will have to fight for equality. 16.Nxa4 dxe5 17.f5 was seen in Knittel – Cole, corr. 2015, when natural and strong would have been:

17...Nf6!N 18.d6 Ncd5 19.Bg5 Rb4 20.fxg6 hxg6 21.Nc3 Nf4³ Once again White has some compensation for the pawn, but he will have to struggle for a draw. 16...dxe5 17.f5 It seems that White’s initiative has acquired a threatening character, but Black can defuse it with some precise moves. 51

17...gxf5! Avoiding the greedy 17...Rxb2? 18.d6 Nb5 19.fxg6 hxg6 20.Nd5+– when the significance of the d5square becomes apparent. 18.Rxf5

We have been following Vachier-Lagrave – Radjabov, Skopje 2015. Unlike the previous note, here Black could have obtained a more promising position by grabbing a pawn: 18...Rxb2!N This is stronger than the game continuation of 18...Nb6?!, when 19.Rg5! led to the unpleasant possibility of an exchange sacrifice on g7 to open up the black king. 19.Qe1 Kh8 20.Rh5 Rb4 21.Qg3 Ne8³ White has some initiative, but Black has three extra pawns. With careful play, he has decent chances to realize his extra material. The position remains challenging for both sides but I believe that White risks more than Black. B23) 8.Bg5

52

This is a cunning continuation, the point of which is to provoke ...h6, which will leave Black’s hpawn under attack in the future. 8...h6 Despite the above point, it’s best to play this move to avoid Qd2 followed by Bh6, when White exchanges the dark-squared bishops. 9.Be3 0-0 10.Qd2 a6 As we will see in the note below, the h6-pawn is indirectly protected. 11.a4 As usual, White prevents ...b5. He may or may not seize further queenside space with a4-a5. 11.Bxh6?! runs into a typical sequence for these positions: 11...Nxe4 12.Nxe4 Qh4† 13.g3 Qxh6 14.Qxh6 Bxh6 15.Nxd6 Nd7

53

16.Bh3 Nb6 17.Nxc8 Nxc8÷ Shishkov – Kantans, Liepaja 2017. In the resulting position, Black’s chances seem higher to me. Objectively the evaluation may not go beyond equilibrium, but only Black plays to win, because of the powerful dark-squared bishop and weakness of the isolated pawn on d5, which will be perfectly blockaded by a knight on d6. White’s extra pawn is practically irrelevant and Black will soon get active along the e-file and/or on the queenside with ...b5. 11...Nh7!? With this non-standard move, we defend against Bxh6 by preparing a check on h4, while also planning counterplay with ...f5. It would be a mistake to continue in the same manner as in variation B21 with 11...Nbd7?!, since after 12.Nh3 Ne5 13.Nf2² Black cannot play the key ...Nh5 move, due to the hanging h6-pawn. Because of this detail, White can claim a slight advantage. 12.Nge2 12.Bd3 is positionally risky, as Black can once again target the bishop. 12...Nd7 13.Nh3 Ne5 14.Nf2 Nxd3† 15.Nxd3 occurred in Ionescu – Ardeleanu, Bucharest 1993, and here I would recommend:

54

15...f5N 16.0-0 g5÷ With two strong bishops and kingside counterplay, Black is certainly not worse. 12.a5 was played at the highest level in Ding Liren – Nepomniachtchi, Sharjah 2017. I recommend a new idea:

12...Qe7!?N 13.Bc4 Nd7 14.Nge2 Ne5 15.Bb3 b5! 16.axb6 Rb8 17.0-0 Rxb6÷ The position is rather complicated but White has no opening advantage. 12...Nd7 13.Nf4 The knight is not ideally placed on f4, but White has nothing better. 13.h4 h5 does not improve White’s prospects: after 14.Nf4 Ne5 15.Be2 Qe7 16.a5 Bd7 17.g3 Rae8 18.0-0 f5„ Black is perfectly centralized. 55

13.Ng3 allows Black to target the knight in the usual manner. 13...h5 14.Be2 h4 15.Nf1 Re8 16.Bf2 occurred in Ponomariov – Gelfand, Beijing 2014, and now I suggest:

16...h3N 17.g4 Ne5÷ We reach a complex position where Black has ideas of ...Ng5 and/or ...f5 at the right time. 13...Ne5 14.Be2 Nf6 The knight returns to battle. The pseudo-active 14...f5? would be a clear mistake in view of 15.Ne6 Bxe6 16.dxe6± when White has achieved many pluses.

56

15.g4! The idea of this move is to vacate the g2-square for the knight, as well as seizing space. In general, White has a headache finding a good role for the knight on f4. Thus, after 15.b3 g5 16.Nd3, as seen in L.B. Hansen – Mortensen, Lyngby 1991, it is better to avoid exchanging White’s problem piece:

16...Ng6!N My analysis continues 17.Rc1 Re8 18.0-0 Nh5 19.Kh1 Bd7 20.f4 Nhxf4 21.Nxf4 Nxf4 22.Bxf4 gxf4 23.Qxf4 Qe7³ when Black has better prospects thanks to the two bishops, pressure against the e4-pawn and the possibility of playing ...b5 at a convenient time. It is worth considering 15.0-0 g5 16.Nh3 as played in Neukirchner – Sadzikowski, Heusenstamm 2018. Here I recommend:

57

16...Qe7N 17.f4 gxf4 18.Bxf4 Ng6 19.Nf2 (in the event of 19.Bxh6 Bxh3 20.gxh3 Nxe4 21.Nxe4 Qxe4 22.Bxg7 Kxg7³ Black enjoys the superior pawn structure) 19...Nxf4 20.Qxf4 Rb8 21.Ra3 Nh7 22.Ng4 Qg5³ Black is better because of the powerful dark-squared bishop. 15...Bd7 Black develops the last minor piece and supports a possible ...b5. 16.Ng2 h5 17.g5 Nh7 All this happened in Dao Thien Hai – Gdanski, Budapest 1996, but the critical continuation in this position has not yet been tested.

58

18.f4!N This aggressive move leads to great complications. 18.0-0?! I get the feeling that this seemingly logical move is inaccurate. 18...b5! Black captures the initiative. 19.Kh1 White gains nothing from 19.axb5 axb5 20.Rxa8 Qxa8, as it would be dangerous to grab a pawn: 21.Nxb5? Bxb5 22.Bxb5 Rb8 23.Qe2 Qa5 24.Bd3 Rxb2µ Black’s activity is extremely high. 19...Re8

20.Ra2 White does not have an effective plan and must try to find useful waiting moves. 20.f4?! brings White nothing but weaknesses: 20...Nc4 21.Bxc4 bxc4 22.Qc2 Qe7³ The e4-pawn is vulnerable, while Black has the bishop pair and the possibility of developing pressure along the b-file. I also checked 20.axb5 axb5 21.Rxa8 Qxa8 22.Bf4 (22.Bxb5?! Bxb5 23.Nxb5 Qa6 24.Qe2 Rb8 25.Nc3 Qxe2 26.Nxe2 Rxb2µ clearly favours Black) 22...Qb7 23.Nd1 Ra8³ when it is obvious that White should worry about equalizing. 20...Rb8 21.axb5 axb5 22.b3 Here it was worth paying attention to the development of Black’s initiative on the queenside.

59

22...b4!N 22...f6?! was an untimely blasting of the kingside. The game continued 23.gxf6 Qxf6 and now 24.Ra6?! Nf7 favoured Black in Dao Thien Hai – Gdanski, Budapest 1996, but 24.Ra7!N 24...Ra8 25.Rc7 would have been unclear; my main line from here ends in perpetual check some moves later. 23.Nd1 Qe7 24.Nf2 Ra8 25.h4 Ra3!ƒ Black has the better prospects. 18...Ng4 19.Bg1 Qe7 20.h3 Bxc3! 21.Qxc3 In this razor-sharp position, Black has a remarkable resource which should lead to a draw if both sides play correctly.

60

21...Nxg5!! The immediate 21...Qxe4 is less accurate due to 22.Kf1, when White has chances for an advantage. 22.fxg5 White should not be tempted to capture the other knight: 22.hxg4?! Nxe4 23.Qh3 Rae8 24.Be3

24...Nf6! White has many weak points to worry about. According to my analysis, best play continues 25.f5 Nxd5 26.Ra3 Qe5 27.gxh5 gxf5 28.0-0 Qxb2 29.Bh6 Qxe2 30.Qg3† Qg4 31.Qxd6 Qd4† 32.Rf2 Nf6 33.Bxf8 Rxf8 34.Rg3† Ng4 35.Qxd4 cxd4ƒ and Black keeps some initiative. 22...Qxe4 23.Qf3 23.Kf1 is now strongly met by 23...Ne5, when ...f6 is a powerful threat.

61

23...Qb4† 24.Qc3 Qe4= The line ends in a surprising repetition, which neither side can avoid. Conclusion 3.f3 is quite a poisonous system. We meet it with 3...c5, when 4.dxc5!? should be considered but 4.d5 d6 5.e4 e6 is deservedly the main line. After White builds his centre with e2-e4, he must decide how to develop his minor pieces. One idea is 6.Ne2 intending Nec3, when Black should avoid an immediate clarification of the central structure and instead go for the 8...Nh5! idea. The main line is 6.Nc3, and after 6...exd5 7.cxd5 Bg7 White has another choice. Against 8.Be3 Black should transfer the queen’s knight to e5, not fearing the Nh3-f2 manoeuvre due to the plan of ...Nh5 and ...f5, which promises Black a good game. Next we looked at 8.Nge2 intending Ng3 and Be2, keeping the dark-squared bishop flexible. In that case, Black should go for a different set-up with ...Ne8, ...Nc7, ...Rb8 and eventually ...b5. Complex positions may arise, especially after f3-f4 and e4e5, but after reading this chapter the reader will be sufficiently armed. Finally, 8.Bg5 should be met by the plan of ...h6 followed by ...Nh7 with an indirect defence of the h6-pawn, followed by ...Nd7-e5 with a promising game.

62

A) 5.Qa4!? 37 B) 5.Nf3 Bg7 6.0-0 0-0 39 B1) 7.Nc3?! 41 B2) 7.Nbd2 43 B3) 7.b3 dxc4! 8.bxc4 c5! 49 B31) 9.e3 49 B32) 9.Bb2 50 B4) 7.Qb3 54 C) 5.cxd5 cxd5 6.Nf3 Bg7 62 C1) 7.0-0 62 C2) 7.Nc3 Ne4 66 C21) 8.Qa4† 67 C22) 8.Qb3 68

63

1.d4 Nf6 2.c4 g6 3.g3 My database indicates that future World Champion Alexander Alekhine was the first to play this move played against Oscar Mueller in 1923. In the same year, it was also adopted by future World Champion Max Euwe, as well as the positional genius Akiba Rubinstein – against none other than Ernst Grünfeld himself. White hopes that his bishop on g2 will exert pressure on the long diagonal. Another important point is that White will be well placed to fight for the d5- and e4-squares. The Fianchetto System is generally favoured by positional players who prefer a solid game. As such, it was firmly included in the repertoires of Botvinnik and Karpov, while nowadays Carlsen, Ding Liren and Mamedyarov are among those who use it. White can also enter this system via a different move order such as 3.Nf3 Bg7 4.g3, in which case we respond in the same way as shown below, with an inevitable transposition. 3...c6 Black prepares ...d5 with a solid barrier against the fianchetto bishop. The main alternative is 3...Bg7 4.Bg2 d5, leading to a different type of game. 4.Bg2 4.d5 does not require serious attention – White must really fear the Grünfeld Defence to make such a move. A noteworthy example continued 4...Bg7 5.Bg2 Qa5† 6.Nc3 Qb4 7.Qd3 d6 8.Kf1?! cxd5 9.cxd5 Na6µ when White had serious coordination problems in Petrosian – Ree, Amsterdam 1973, which the former World Champion solved by offering his much-lower-rated opponent a draw on the next move. Obviously White can start with 4.Nf3 or 4.Nc3 but it makes no real difference, as 4...d5 will almost certainly transpose to one of our main lines after a subsequent Bg2. 4...d5 We will analyse three main options: A) 5.Qa4!?, B) 5.Nf3 and C) 5.cxd5. 5.Qb3 is nothing to fear. 5...Bg7 6.Nc3 0-0 7.cxd5 Grabbing a pawn is too risky. (White should settle for 7.Nf3, when 7...Qb6 8.0-0 transposes to variation B4) 7...cxd5 8.Nxd5 Nxd5 9.Bxd5

64

9...Nc6 10.Bxc6 bxc6 11.Nf3 Bh3 12.Ng5 Bg2 13.Rg1 Bd5 14.Qd3 Qb6 15.Nf3 Rfd8 16.Kf1 c5µ Kalinin – Gobet, Thessaloniki (ol) 1988. A) 5.Qa4!? This move was recommended by Avrukh in Grandmaster Repertoire 2A. He presented some nice ideas for White, but theory develops quickly and an attractive new idea has emerged for Black. 5...Bg7 6.cxd5 0-0! This idea was introduced by Vachier-Lagrave, sometime after Avrukh’s book was published. Black sacrifices a pawn, relying on his rapid development to develop his initiative. 7.dxc6 White can also decline the pawn sac and play aggressively with: 7.e4 cxd5 8.e5 As seen in Mamedyarov – Vachier-Lagrave, Saint Louis 2018. Here I propose an exciting new idea:

65

8...Ne4!N 9.f3 Bd7 10.Qa3 White appears to be winning a piece for no compensation, but Black has a spectacular resource. 10...Nd6!! 11.exd6 exd6 White is far behind in development and his king is stuck in the centre, while the d4-pawn is vulnerable. My analysis continues: 12.Ne2 Re8 13.Qd3 13.Nbc3 is an interesting alternative suggested by Jacob Aagaard. Black responds with 13...Nc6 14.0-0 b5! when my line continues: 15.Qxd6 b4 16.Rf2 bxc3 17.bxc3 Rxe2 18.Rxe2 Nxd4 19.Rd2 Qe8 20.Rxd4 Qe1† 21.Bf1 Bh3 22.Bb2 Qe3† 23.Kh1 Qxf3†= With perpetual check. 13...Qb6 14.Nbc3 Bxd4 15.Kd1 Bxc3! It is worth giving up the prized bishop to keep up the momentum of the attack. 16.Nxc3 Bc6 17.Qd2 a5© A fantastic variation! Black has two pawns for his piece, while White has difficulty coordinating his forces and his king remains stuck in the centre. Due to these factors, Black has full compensation. 7...Nxc6 This position has been tested a few times at the highest level.

66

8.Nf3 Nd5 9.e3 9.Nc3 Nb6 10.Qd1 was played at a high level in Edouard – Jones, Batumi (ol) 2018. I recommend improving with:

10...Be6!?N 11.0-0 Nxd4 12.Nxd4 Qxd4 13.Qxd4 Bxd4 14.Bh6 Rfd8 15.Rac1 (15.Bxb7 is not scary due to 15...Rab8 16.Bf3 Nc4ƒ) 15...Nd5 With roughly equal play. 9.0-0 Nb6 10.Qa3 was another interesting sequel, seen in Rodshtein – Alexeenko, Skopje 2019. Here I can offer another novelty:

67

10...Bg4!N 11.e3 e5 12.Nxe5 Nxe5 13.dxe5 Nc4 14.Qb4 Nxe5 15.Nc3 (15.Qxb7?! is too risky in view of 15...Rb8 16.Qxa7 Be2 17.Re1 Bf3ƒ when White is lagging in development and Black has a nasty initiative) 15...Nf3† 16.Kh1 Qd7 The game might end in a draw after17.e4 Bh3 18.Nd5 Nh4 19.gxh4 Bxg2† 20.Kxg2 Qg4† 21.Kh1 Qf3† 22.Kg1 Qg4†= 9...Nb6 10.Qd1 e5 11.0-0 11.dxe5N is worth considering. Play continues 11...Qxd1† 12.Kxd1 Rd8† 13.Ke2 Bf5 14.Rd1 Rxd1 15.Kxd1 Rd8† 16.Ke2 Bd3† 17.Ke1 Nc4© when it is obvious that Black has enough compensation for two pawns, and White should switch to defence. 11...Bg4 12.dxe5 Qxd1 13.Rxd1 Nxe5 14.Nbd2 All this happened in Vitiugov – Vachier-Lagrave, Karlsruhe/Baden-Baden 2018, and here I recommend another novelty:

68

14...Nbc4N 15.Nxe5 Nxe5 16.f3 Be6© White has problems developing his queenside, so Black has plenty of compensation for the pawn. B) 5.Nf3

White develops the kingside. Black could consider taking on c4 but I prefer straightforward development. 5...Bg7 6.0-0 Here or on the next move, White could transpose to one of the later variations by exchanging on d5. However, most players who intend to go for this plan will tend to do it on move 5 to avoid any ...dxc4 69

ideas. White gains no advantage from delaying castling, for instance: 6.b3 0-0 7.Bb2 dxc4 8.bxc4 c5 9.d5 White should prefer 9.0-0, transposing to variation B32 on page 50. 9...e6 10.dxe6 10.Nc3 exd5 11.Nxd5 Nxd5 12.Bxg7 Ne3! 13.fxe3 Kxg7 14.0-0 Qe7³ leaves White with a terrible pawn structure. 10.Nh4 exd5 11.cxd5 Nbd7 12.0-0 b5 13.Nd2 Re8³ gives Black a harmonious position with good chances to advance his queenside pawns. 10...Qxd1† 11.Kxd1 Bxe6 12.Ng5 Bc8 13.Nd2 h6 14.Ngf3 Nc6 15.Ne5 Nxe5 16.Bxe5 In Markus – Jelen, Otocec 2016, Black could have established some advantage with:

16...Ng4!N 17.Bxg7 Kxg7 18.Ke1 Rb8³ The position of the white king makes it hard for him to connect the rooks and get coordinated.

70

6...0-0 Some strong players have taken on c4, but I am happy to complete our kingside development and wait for White to commit to a plan. We will analyse four main options: B1) 7.Nc3?!, B2) 7.Nbd2, B3) 7.b3 and B4) 7.Qb3. The risky 7.Bf4 dxc4 8.Qc1 b5 9.Bh6 was played in Romanishin – Macieja, Koszalin 1997. I recommend 9...Qb6!?N 10.Nc3 Bb7 11.b3 cxb3 12.axb3 Nbd7³ when White has some compensation for the pawn, but it is not enough for equality. Some strong players have tried: 7.Qa4 White defends the c4-pawn and gives Black a chance to speak. Usually Black responds by moving one of his knights via d7 to b6, after which White retreats his queen (having already traded pawns on d5) and tries to play against the bad knight on b6. This is not much of an achievement though, as White loses time with his queen and the knight on b6 is not so bad anyway. 7...Nfd7 8.cxd5 8.Qc2!? is another idea, when 8...dxc4 is logical since the queen must move yet again to recapture the pawn: 9.Qxc4 e5 10.dxe5 Nxe5 11.Nxe5 Bxe5 12.Nd2 (12.Qc2 Na6 13.Bd2 Qe7 14.Bc3 Re8 15.Bxe5 Qxe5 also led to roughly equal play in Navara – Ragger, Germany 2014) 12...Be6 13.Qc2 Na6 14.Nf3 Bf6 15.Be3 Nb4 16.Qe4 Nd5÷ In Gleyzer – Cintins, corr. 2014, Black’s pieces were harmoniously placed. 8...Nb6 9.Qd1 cxd5 Something similar to this can arise if Black meets 5.Qa4!? with 5...Nfd7 (rather than the pawn sacrifice which I recommended earlier). This is a slightly improved version for Black, as White has already committed his knight to f3, thus losing out on the interesting options of developing the knight 71

via e2 (after e2-e3) or even h3, as Avrukh recommends. 10.Nc3 Nc6 11.Bf4 Black has no problems after 11.h3 Bf5 12.Bf4 Rc8 13.Rc1 as seen in Sedlak – Rodshtein, Valjevo 2012. I like the look of 13...f6!?N to prepare ...e5, as shown after 14.g4 Be6 15.e3 Bf7 16.Bg3 e5÷ when Black is not worse. A more recent high-level game continued 11.e3 Bf5 12.b3 Rc8 13.Bb2 Qd7 14.Qe2 Bh3 15.Rfd1 Bxg2 16.Kxg2 e6 17.Rac1 a6 18.h4 h5 19.Ba3 Rfd8 20.Bc5 Qc7 21.Ba3 Nd7 22.Na4 Qa5 23.Qe1 Qxe1 24.Nxe1 Bf8 25.Bxf8 Kxf8= with an equal endgame in Yu Yangyi – Ding Liren, Huaian 2017. 11...Bf5 12.Rc1 This occurred in Pigusov – Huzman, Sverdlovsk 1987. I recommend the following improvement:

12...h6!N Preparing to drive White’s bishop away with ...g5. 13.b3 13.h4 prevents our plan but weakens White’s kingside slightly. A good reply is 13...f6! 14.Nd2 Bg4³ when Black can think about ...e5, and the bad outlook of the g7-bishop is only temporary. 13...g5 14.Bd2 Rc8³ Black has the more pleasant position. B1) 7.Nc3?! This has been quite popular, which is surprising, considering that Black can already fight for the advantage by capturing the pawn. 7...dxc4 8.Ne5 72

It is surprisingly difficult for White to regain the pawn. White may also simply ignore the c4-pawn and aim for Catalan-style gambit play: 8.e4 b5 9.Re1 Bg4 10.Bf4 (10.h3 Bxf3 11.Bxf3 e5 12.d5 b4 13.Na4 cxd5 14.exd5 Nbd7µ left Black with a healthy extra pawn in Fricot – Wurschner, email 2001) This occurred in W. Schiller – Bubir, email 2001, and here I recommend the natural:

10...Nbd7N 11.h3 Bxf3 12.Bxf3 e5! This is an important resource in such positions. 13.dxe5 Ne8 14.e6 fxe6 15.e5 Rc8³ White has some compensation for the pawn but it is not enough for equality. I also checked: 8.a4 Na6! Eyeing the b4-square which was weakened by White’s last move. 9.Ne5 It does not look beneficial to capture space on the queenside: 9.a5 Nd5 10.Qa4 Ndb4 11.Na2 occurred in Sinanan – Young, corr. 2010, when 11...Nxa2N 12.Rxa2 Qd5 13.e3 Qb5 14.Bd2 Rd8 15.Bc3 Bf5µ would have left White with no compensation for the pawn. 9.h3 has also been tried but is too slow: 9...Nd5 10.e4 Ndb4 11.Be3 Nd3 12.Qd2 (12.b3 Be6µ) 12...e5µ In Mozny – Cernousek, Czech Republic 2014, Black not only had an extra pawn, but also a positional advantage. 9.e4 seizes control of the centre but weakens the d3-square, which Black is perfectly placed to exploit after his previous move9...Nb4 10.Ne5 Be6 11.Be3 Nd7 12.f4 Nxe5 13.dxe5 Qa5µ Black had a healthy extra pawn in Hardie – Spiller, Newport 2004. The text move is met by a thematic idea:

73

9...Ng4! 10.f4 In Heinatz – Hirn, Germany 1998, Black could have secured a clear advantage with: 10...Nxe5N 11.fxe5 Be6µ White has no compensation for the pawn.

8...Ng4! Black gets an excellent position with the same motif as in the last note above. 9.Nxg4 The pseudo-active 9.f4 was well met by 9...Nxe5 10.fxe5 Be6 11.e4 Na6µ in A. Fernandes – Vitor, Evora 2006. It is not clear what White has achieved, but Black maintains a strong position with the 74

knight heading for d3, and he has an extra pawn to boot. 9.Nf3 looks like a pretentious way to offer a repetition, but Black can play for more with 9...e5! 10.dxe5 Nxe5 11.Nxe5 Qxd1 12.Rxd1 Bxe5 13.Bh6 Re8 14.Ne4 Na6 15.Nd6 Bxd6 16.Rxd6 as seen in Dao Thien Hai – Dela Cruz, Manila 2008. Black has a few decent options here but 16...Bf5N³ would be my choice, when White’s dark-squared bishop is a useful asset, but it does not fully compensate for the missing pawn. 9...Bxg4 10.h3 Be6 11.e3 Qd7 12.Kh2 Na6 13.a3 Nc7 14.Ne4 We have been following Vierroth – Kl. Mueller, Bad Wildbad 1998. I recommend improving with:

14...Bd5N 15.Nc5 Qc8 16.e4 Be6 17.Be3 Rd8µ The well-placed knight on c5 cannot compensate for White’s missing pawn and the pressure against d4. B2) 7.Nbd2

75

White develops a piece and guards the c4-pawn, but the knight does not have many active prospects. 7...a5 A useful move. White must develop his dark-squared bishop at some point and b2-b3 is the most likely way to do it, so Black anticipates this and gets ready to chip away at White’s structure with ...a4. 8.b3 White usually plays this anyway, in spite of Black’s plan. We will consider some alternatives where White develops in a different way: 8.e3 Bf5 9.Qe2 is rather passive. 9...a4 10.h3 occurred in Kociscak – Flumbort, Austria 2018, and here I like:

76

10...e6N Fortifying the centre, since there is no need to fear for the light-squared bishop: 11.g4 (11.Nh4 Be4 leaves White nothing better than 12.Nhf3, and after 12...Re8 I slightly prefer Black) 11...Be4 12.Nxe4 Nxe4 The loss of the bishop pair is inconsequential, and Black is at least equal. He can choose between different plans both on the queenside and on the kingside, and the a4-pawn provides a useful space advantage. 8.Ne5 Nbd7 9.Ndf3 a4 10.cxd5 (10.Be3?! runs into 10...Nxe5 11.Nxe5 dxc4 12.Nxc4 Ng4³ when Black trades off the dark-squared bishop because the d4-pawn is weak) Now in Hegeler – Joachim, Germany 2000, Black could have fought for the advantage with:

10...Nxe5!N 11.Nxe5 Nxd5 12.e4 Nb6 13.Be3 f6 14.Nd3 Nc4³ Again White must allow his darksquared bishop to be traded for a knight, which should work to Black’s advantage in the long run. 77

8.Re1 Bf5 9.Qb3 (9.Nh4 Be6 leads nowhere for White: 10.c5 b6 11.cxb6 Qxb6 12.Nhf3 was seen in Vijayeendra – Uphoff, Regensburg 2018, when 12...Ne4!N 13.e3 f5 14.b3 Nd7³ would have given Black active prospects with chances to seize the initiative on either flank) This position was reached in Illescas Cordoba – Salgado Lopez, Linares 2014, and here I found a nice improvement:

9...b5!N 10.cxb5 a4 11.Qb4 cxb5 12.Qxb5 Ra5 13.Qb4 Nc6 14.Qa3 Qb6© Black has full compensation for the sacrificed pawn, and White already has to worry about the unpleasant threat of ...Nb4-c2.

8...a4 Further expanding the space. Now White must reckon with ...axb3 – although Black will often prefer to maintain the tension rather than hurry with this exchange. 78

9.Bb2 This is the most popular choice although I checked some alternatives. 9.c5 was seen in Goksel – T. Petrosian, Antalya 2004, and now I like 9...Ne4N, when 10.b4 takes us to the 9.b4 line below. 9.a3 occurred in Gross – Osterroth, corr. 1979, and now 9...Ne4!N is a good example of Black maintaining the tension on the queenside and causing difficulties for White. A logical continuation is 10.Bb2 Nxd2 11.Nxd2 dxc4 12.Nxc4 Be6³ when White is under some pressure on the queenside. 9.bxa4 damages White’s structure: 9...Qa5 10.Qb3 Qxa4 11.Bb2 dxc4 12.Nxc4 Be6 13.Nfd2 Nbd7 14.Rfc1 Nd5 15.Qxa4 Rxa4 16.a3 Rfa8³ In Haria – Kobo, Stockholm 2018, Black benefited from a more active positioning of the pieces as well as pressure along the a-file. 9.b4 Ne4 White already has problems connected with his lack of development. 10.c5 I also checked 10.e3 Bf5 11.Nxe4 Bxe4 12.Qe2 as played in Neese – Benne, RhedaWiedenbrueck 2001, when I recommend here 12...dxc4N 13.Qxc4 b5! 14.Qe2 a3³ would have been good for Black. The backward c-pawn is hard to attack and Black’s knight can head for c4, while the b4-pawn is a long-term weakness. 10...Nc3 11.Qe1 Nb5 12.e3 In Rogozenko – C. Hess, Bargteheide 2013, Black could have caused serious problems with:

12...Bf5!N 13.Kh1 Bd3 14.Rg1 Nd7µ Black’s pieces are much more active than their counterparts, and ...e5 is on the way. 79

9.Ba3 This is a more respectable alternative which has been tested at a high level. 9...Re8 10.e3 Bf5 11.Qe2 Nbd7 12.cxd5 Nxd5! An important detail. 13.e4 This is the critical move to check, but the ensuing complications are not too scary. 13.Nh4 has also been played a few times: 13...Be6 14.Qd3 (14.Ne4 Nc7 15.Nc5 Nxc5 16.Bxc5 axb3 17.axb3 Bxb3 18.Qb2 Bc4 19.Rxa8 Qxa8³ left White with insufficient compensation for the missing pawn in Vitiugov – Ding Liren, Shenzhen 2018) 14...Qb6 15.Rfc1 Nc7 16.Nc4 Qa6 17.Bb2 Red8÷ Black had a comfortable position in Loginov – Tasic, corr. 2012. 13...Nc3 14.Qd3 Not better is 14.Qe3 Nb5 15.exf5 Nxa3 16.fxg6 hxg6 17.Rac1 Nb5 18.Rfd1 Nf6 19.Nc4 Nd5 20.Qd2 axb3 21.axb3 e6³ when Black enjoyed a better pawn structure in Vachier-Lagrave – Ding Liren, Abidjan 2019. 14...Nb5 15.exf5 Nxa3÷ Black has a good position and he soon took over in a recent high-level game:

16.bxa4 Rxa4 17.Qb3 Ra7 18.Ng5 Rf8 19.Be4 Qb6 20.fxg6 hxg6 21.Qf3 Qxd4µ Black had a healthy extra pawn in Caruana – Vachier-Lagrave, Saint Louis 2019. 9...Bf5 The bishop moves to its most natural post where it influences the queenside as well as the e4-square. 10.e3 This move supports White’s centre but weakens the d3-point.

80

In the event of 10.bxa4 Qa5 11.Qb3 Qxa4 White will have to justify his weakened structure:

12.Rfc1 (Trying to keep the extra pawn only makes things worse: 12.Qxb7?! Nbd7 13.Qb3 Rfb8 14.Qxa4 Rxa4 15.Bc3 dxc4 16.Nh4 Nd5µ Horie – Aberg, corr. 2015. Despite the fact that Black has doubled, isolated pawns, he is clearly better due to the powerful knight on d5 and the weakness of the a2-pawn.) A draw was agreed here in Csom – Bronstein Wrexham 1995. Play might continue 12...dxc4N 13.Nxc4 Nbd7 14.Nfd2 Rfc8 15.e4 Be6÷ when Black is not worse at all, and I would prefer his position due to the pressure along the a-file. A few strong grandmasters have tried: 10.Ne5 Nbd7 11.Nxd7 11.h3 Qa5 12.g4 Be6 13.f4? was a blunder, after which the game cannot be saved. (I also checked 13.Rc1N 13...h5 14.Bc3 Qd8 15.g5 Ne8 16.Nd3 axb3 17.axb3 Bf5³ when Black enjoys the better pawn structure and White’s kingside has been weakened.) 13...a3 14.Bxa3 (14.Bc1 Qc3–+) 14...Qxa3 15.f5 gxf5 16.gxf5 Nxe5 17.dxe5 Qc5† 18.Rf2 Nd7 19.fxe6 fxe6 20.Nf3 Nxe5–+ Black had an extra pawn and a dominating position in Kunin – Sasikiran, Berlin 2015. 11...Qxd7 12.bxa4 Bh3 13.Qc2 This occurred in Kosic – Ragger, Neum 2011. I would suggest:

81

13...Bxg2N 14.Kxg2 h5!? Or 14...Ra7 immediately. 15.h3 Ra7³ Followed by ...Rfa8, with pressure along the a-file. White can also drive the bishop away, at the cost of misplacing his knight: 10.Nh4 Be6 11.Qc2 11.e4?! proves to be misguided after 11...dxe4 12.Nxe4 Nxe4 13.Bxe4 c5 14.Bxb7 Ra7 15.Bg2 as in Lai – Leenhouts, Maastricht 2017, when 15...cxd4N 16.Nf3 Nc6 17.Ng5 Bd7 18.Ne4 Qa5ƒ would have given Black a nasty initiative on the queenside. 11.e3 b5 12.Nhf3 Qb6 13.c5 Qa5 14.Rc1 occurred in Barwich – Leisebein, email 2008, when 14...Qc7N÷ would have given Black a good position with chances to break with ...e5 in the future. 11.Re1 has been tried by some strong players but it does not change much. 11...b5 12.e3 Qb6 13.c5 Qc7 14.b4 a3 15.Bc3 Ra4 16.Qc2 occurred in Vekelis – Hrubcik, corr. 2016, when a natural continuation is: 16...Na6N 17.Rab1 Qd7 18.Bf1 Ne4 19.Bd3 Nxd2 20.Qxd2 Nc7 21.f3 f6 Preparing ...e5 is a typical plan for such positions. It is dangerous for White to try to get active with 22.e4?! in view of 22...dxe4 23.fxe4 f5ƒ when White’s centre turns to ruins and Black seizes the initiative. 11...b5 12.Qd3 12.Bc3 was played in Gukesh – Gandhi, Sharjah 2019. White’s last move prepares Bb4 in some lines, but there is a risk that the bishop may be targeted by Black’s queenside pawns. I recommend 12...Qd7N with the possible continuation: 13.Rfc1 Rc8 14.e3 Na6 Suddenly White has to worry about ...b5-b4 and ...a3, and if 15.cxb5 cxb5 16.a3 Ne4!ƒ Black obtains a strong initiative. We have been following B. Jensen – Perkiomaki, corr. 2010. Here I recommend improving with: 82

12...Qb6N 13.c5 Qb7 14.f4 axb3 15.axb3 Rxa1 16.Rxa1 Na6³ It is difficult for White to find a constructive plan, and his pawn advances have left some holes which Black may be able to exploit in the long run.

10...Ne4 11.Nxe4 White does not want to endure the strong knight and simply exchanges it. 11.Qc1 defended against ...a3 in Maiorov – Valsecchi, Tortoreto 2015. 11...Qb6N is a decent reply, when 12.Nxe4 Bxe4 transposes to the 12.Qc1 line in the notes to the main line below. 11.bxa4 gives White nothing as usual. In Hera – Manik, Slovakia 2010, the most precise continuation would have been: 11...Nxd2N 12.Nxd2 Bd3 13.Re1 dxc4 14.Rc1 Qb6 15.Bc3 Qa6 16.a5 Rd8 17.Bb4 83

e5 18.dxe5 Qb5 19.Be7 Rd7 20.Bd6 Bxe5 21.Bxe5 Qxe5ƒ White’s pawns on a2 and a5 are weak. 11...Bxe4 12.Qe2 It is natural to develop the queen and connect the rooks. 12.Ne1 Bxg2 13.Nxg2 gives White nothing. In Kalme – Benko, New York 1958, Black could have fought for an edge with 13...Qa5!?N 14.Rb1 e6 15.Nf4 Nd7³ when White’s bishop is bad and Black’s queenside pawn ram gives him the freer game. 12.Ne5 is similarly unimpressive. In Salzmann – Willmann, email 2014, the simple 12...Bxg2N 13.Kxg2 e6 14.Nd3 Nd7 would have given Black slightly the more pleasant side of equality. It’s not logical to relieve the tension with 12.cxd5?!, as played in Pelts – Westerinen, Thessaloniki (ol) 1984, in view of 12...Qxd5!N when the black queen is strongly activated. 13.Ne1 Bxg2 14.Nxg2 Qb5 15.Nf4 Nd7³ Black can develop pressure along the a-file by doubling rooks, and White must also watch out for the idea of ...e5. 12.Qc1 was played in Barczay – Blees, Lillafured 1989, with White’s last move protecting against ...a3. 12...Qb6N is a good answer, and after 13.c5 Qc7 14.Ne1 Bxg2 15.Nxg2 Nd7³ Black has good chances, both in the centre with ...e5, and on the queenside thanks to the a4-pawn. White has a bad bishop on b2 and has difficulty finding a plan. Another possible scheme is: 12.Ba3 Re8 13.Bh3?! 13.Qe2 was seen in Ehlvest – Kantans, Harjumaa 2010, and here I recommend 13...e6N 14.Bh3 h5!? 15.Nd2 Bf5 16.Bxf5 exf5ƒ when Black has good control over the centre, as well as favourable tension on the queenside. The text move was played by a GM rated over 2600, but he overlooked a powerful resource for Black. 13...c5! 14.cxd5 The main tactical point is that 14.Bxc5? axb3 15.axb3? Rxa1 16.Qxa1 Bxf3–+ leaves Black with an extra piece. Now in Markowski – Bayram, Antalya 2004, Black’s most accurate continuation would have been: 14...Bxd5N 15.bxa4 Qa5 Black has an excellent game, since 16.Bxc5? is still impossible due to 16...b6 17.Ba3 Qxa4 and Black once again wins a piece. Finally, 12.Bh3 is another way to safeguard the light-squared bishop from exchange. A good reply is 12...axb3 13.axb3 Rxa1 14.Bxa1 as played in Zifroni – Manor, Tel Aviv 1998, when a good continuation is:

84

14...e6!N 15.Nd2 Bf5 16.g4 Bd3 17.Re1 Nd7³ Black has a harmonious position with a favourable pawn structure. The bishop looks unstable on d3 but it is quite safe, and Black’s plans may include taking over the open a-file. 12...Qb6 Threatening the b3-pawn. On occasion, the queen can switch to a6 to exert pressure against the c4pawn.

13.Nd2 13.Bc3? was a mistake in Nikolic – Ruck, Murska Sobota 2008, due to 13...axb3N 14.Qb2 dxc4 15.axb3 Rxa1 16.Rxa1 cxb3 17.Nd2 Bxg2 18.Kxg2 Nd7µ when White has no compensation for the 85

loss of material. 13.c5 captures space on the queenside, but weakens some light squares. In Ladva – Volodin, Tallinn 2016, it would have been good to play 13...Qa6N 14.Qd2 Nd7 15.Rfc1 Rfe8 16.Bc3 h5ƒ when Black has a certain initiative, and it’s hard for White to find an active plan. 13...Bxg2 14.Kxg2 Nd7 15.Rfc1 15.Qd3 was a slightly clumsy way to position the queen in Kociscak – Kreisl, Zillertal 2015. A good response would have been: 15...Rfc8N! 16.Rfc1 (16.c5 Qa6 17.Qxa6 Rxa6 18.b4 Nf6³ favours Black, who may transfer his knight via e8-c7-b5) 16...dxc4 17.bxc4 Qxb2 18.Rab1 Nc5! It is clear that White’s handling of the queen has failed. 19.dxc5 Rd8 20.Qc2 Qxc2 21.Rxc2 Ra7³ Black’s bishop is stronger than the knight, and the doubled c-pawns are more vulnerable than Black’s backward b-pawn. After the text move I found a slight improvement on Nikolic – N. Nguyen, Berlin 2015.

15...Rfc8!?N Ensuring that Black will be well placed to contest the c-file in the event of an exchange on d5. White is under slight pressure due to the tension on the queenside, while Black can always choose whether to change the structure with ...dxc4 or manoeuvre in some other way. B3) 7.b3 White defends the c4-pawn and prepares to develop the bishop on b2. The drawback of this move is that he has slightly weakened himself on the long diagonal. 7...dxc4! 86

Black wastes no time in beginning his counterattack in the centre. If Black goes for a solid set-up similar to the previous line, White will be able to develop his knight to the optimal c3-square rather than d2, with chances for a small plus. 8.bxc4 c5! White’s pawn structure has been weakened to some extent, so he relies on the activity of his pieces. There are two main possibilities: B31) 9.e3 and B32) 9.Bb2. B31) 9.e3 Nc6 10.Bb2 Qb6

11.Qb3 This is the usual continuation. I checked two other options: 11.Qe2 cxd4 12.exd4 Bg4 sees Black develop strong pressure against White’s centre: 13.d5 Na5 14.Na3 e6

87

15.h3 Bxf3 16.Bxf3 exd5 17.cxd5 Rae8 18.Qb5 Qxb5 19.Nxb5 Nc4 20.Ba3 Nxa3 21.Nxa3 Ne4³ In Semenov – Khanas, corr. 2016, Black was better because of the more active arrangement of his pieces, as well as the superior pawn structure. At the highest level White has tried: 11.Qc1 cxd4 Avrukh recommended this as a novelty in Grandmaster Repertoire 8, and it has proven a reliable choice for Black. 12.Nxd4 It is worse to go with hanging pawns: 12.exd4?! Bf5 13.Nc3 Rac8 14.Rd1 Rfd8³ In Markus – Tadic, Valjevo 2018, White’s c4- and d4-pawns were convenient targets. 12...Bd7 13.Nd2 Rac8 14.Bc3 Avrukh’s line continued 14.Rb1 Qa6 15.Bc3 Nxd4 16.exd4 b5 with a good game for Black. This seems like a reasonable assessment, although I would also like to point out that 15...b6!? 16.Rd1 Rfd8÷ is a good alternative. The position is rather complicated but Black has no weaknesses, whereas White will always have to be mindful of the isolated pawn on c4. 14...Nxd4 15.exd4 Bf5 16.Qa3 Bd3 17.Rfe1 Bxc4 18.Nxc4 Rxc4 19.Rab1 Qa6 20.Qxa6 bxa6 21.Bb4 Rxd4= In Mamedyarov – Radjabov, Shamkir 2018, the position was simplified and the rivals soon agreed to a draw. 11...Na5 12.Qxb6 axb6 13.Nbd2 Rd8 14.Rfc1 Bf5 15.Nh4 15.Nb3?! is well met by 15...Ne4³. 15...Nc6 16.Nb3 We have been following Blagojevic – Vega Gutierrez, Skopje 2016. Here I recommend a new idea: 88

16...Be4!?N 17.dxc5 17.f3 gives Black a pleasant choice between 17...g5!?, 17...Bh6!? and the calmer 17...Bd3 18.Rc3 Bf5, each of which enables him to press for an advantage. 17...Bxg2 18.Nxg2 bxc5 19.Nxc5 Ne4! 20.Bxg7 Nxc5 21.Bc3 Ra4© Black has full compensation for the sacrificed pawn. White’s c4- and a2-pawns are weak, and Black may look to cause trouble with various knight jumps. B32) 9.Bb2

This is the main line, developing as fast as possible. 89

9...Qb6 10.Qb3 This is the most natural and popular move. 10.Qc1 This has also been tested at a high level. 10...cxd4 11.Nxd4 Bd7 12.Nd2 White played too actively with 12.c5?! Qc7 13.Na3 Nc6 14.Ndb5 Qc8 15.Rd1 Bh3 16.Bh1 Qf5 17.Qc2 Qh5³ in Tran – Narayanan, Bhubaneswar 2016, when the pawn proved to be more vulnerable on c5 than it was on c4.

12...Nc6 13.c5 13.e3 leads back to the 11.Qc1 line in the notes to variation B31 above. 13.Nc2 Rac8 14.Rb1 Qc7 15.Bc3 b6 16.Qb2 Ne8 17.Ne3 Bxc3 18.Qxc3 Nf6 19.Rfc1 Qe5 20.Qxe5 Nxe5= gave Black a comfortable position in Donchenko – Wang Hao, Isle of Man 2018. 13.N2b3 Rac8 14.Rd1 Rfd8 15.Rb1 Qc7÷ occurred in So – Caruana, Leuven 2018. The position is roughly equal but I would prefer Black as he has no pawn weaknesses. 13...Qc7 14.Nc4 14.N2b3 was seen in Tate – Banzea, Edinburgh 2019, and now after 14...Nxd4N 15.Bxd4 e5 16.Bb2 Rfe8÷ Black has no problems, while White should take care not to become worse. 14.Nb5 Qa5 15.a4 occurred in Rohan – Kaliksteyn, Charlotte 2018, and here I suggest 15...Qb4!N 16.e3 Rfd8 17.Nd4 Rac8÷ when the c5-pawn may become weak in the future. The text move was played in Efimenko – Areshchenko, Lvov 2015. Here it looks natural to play:

90

14...Rad8N 15.Rb1 Nxd4 16.Bxd4 Bc6 17.Be5 Qc8 18.Bxc6 bxc6= Black does not have a shadow of a problem. 10...Ne4! A great move, opening the long diagonal for the g7-bishop. Black can either trade his knight once the b1-knight comes out, or retreat it to d6 to put pressure on the c4-pawn. 11.e3 The hasty 11.Ne5?! only worsened White’s position after 11...cxd4 12.Qxb6 axb6 13.Bxd4 Nd7 14.Bxe4 Nxe5³ in Cosma – Daulyte, Calimanesti 2014, when Black was attacking the c4-pawn as well as threatening ...Nf3†. Trading knights with 11.Nc3 makes life simpler for Black. 11...Nxc3 12.Bxc3 occurred in Calotescu – Gara, Kiskunhalas 1995, and here I recommend: 12...cxd4N 13.Qxb6 axb6 14.Bxd4 Nc6 15.Bxg7 Kxg7 16.Rfb1 Ra6÷ Despite Black’s doubled pawns, White has to be careful as his own pawns are even weaker. 11...Nc6 Black develops another piece while putting more and more pressure on the d4-pawn. 12.Ne5! It is important for White to activate the bishop on g2. The inaccurate 12.Rd1?! was played in Banusz – Seger, Germany 2014, when 12...Na5!N 13.Qxb6 axb6 14.Na3 Nd6³ would have reached a position where Black’s knights exert considerable pressure 91

on the c4-pawn. 12...Nd6 13.Nxc6 The most accurate. Now the power of the bishop g2 increases even more, as the c6-pawn will be a target to attack. I also checked 13.dxc5 Qxc5 14.Nd3?!, which nobody has played as yet, but may seem like a tempting try to sacrifice a pawn for compensation. (White should prefer 14.Nxc6 bxc6 which transposes to our main line below.) After 14...Qxc4 15.Bxg7 Kxg7 16.Rc1 Qxb3 17.axb3 White has some play for the pawn, but Black gets an excellent game by returning it as follows: 17...Bf5 18.Nc5 Rfd8! 19.Nxb7 Nxb7 20.Bxc6 Rac8 21.Na3 (An important tactical point is that 21.Rxa7?? loses to 21...Rxc6! 22.Rxc6 Rd1† 23.Kg2 Be4†) 21...Na5 22.Ba4 Rxc1† 23.Rxc1 Rd3³ White suffers from the weakness of the b3pawn as well as the first rank and the light squares on the kingside. Another not-so-challenging option is: 13.Nd2 cxd4 14.exd4 14.Qxb6?! axb6 15.exd4 Be6 16.Rfc1 was seen in Shimanov – Gorovykh, Irkutsk 2010, when 16...Bxe5N 17.dxe5 Nf5³ would have favoured Black, as his harmonious pieces and better pawn structure are more significant than White’s bishop pair. 14...Be6 15.Rfc1 Rfc8 16.Qxb6?! White should avoid the temptation of doubling Black’s pawns, as the open a-file will be more important. 16.c5 Bxb3 17.cxb6 Ba4 18.bxa7 Rxa7 was equal in Zhou Weiqi – Ding Liren, Danzhou 2010. White has enough activity on the h1-a8 diagonal to compensate for his inferior pawn structure, but not to claim any advantage. 16...axb6 17.f4 In Jakovenko – Svidler, Dagomys 2010, Black could have pressed with:

92

17...Ra4!N The a2- and c4-pawns are weak. The key line continues: 18.d5 Nxe5 19.fxe5 Bd7 20.exd6 Bxb2 21.dxe7 Black should not be seduced by the win exchange, but does best to play: 21...Bd4† 22.Kf1 Bc5³ Black retrieves the e7-pawn and remains with two powerful bishops in an open position. White’s weaknesses have not gone away, and he faces an unpleasant defence. 13...bxc6 14.dxc5 14.Qxb6?! remains untested, for good reason. After 14...axb6 15.Bxc6 Nxc4 16.Bxa8 Nxb2 17.Be4 cxd4 18.exd4 Bxd4 19.a4 Rd8 20.Ra3 Ba6 21.Re1 f5³ “Black has strong pressure” according to Avrukh. 14...Qxc5 15.Bxg7 Kxg7 16.Nd2 Be6

93

17.Rfc1 I also checked 17.Qa4N which has yet to be applied in practice. My line continues 17...Rfc8 18.Rfc1 Rab8 19.Nb3 Qe5 20.Qxa7 Bxc4 21.Na5 Ra8 22.Nxc6 Rxa7 23.Nxe5 Rxa2= with full liquidation of the queenside. Another option is: 17.Qc3† Kg8 18.Rfc1 18.Rac1 is hardly an improvement, and 18...Rab8 19.Nb3 Qxc4 gave Black easy equality in Sturt – Kleinman, Montreal 2012. 18...Rab8 19.Nb3 19.Rab1 was played in Fressinet – Van Wely, Bastia 2012, when 19...Rxb1N 20.Rxb1 Rc8 21.Rc1 f6 22.Nb3 Qxc4 23.Qa5 Qe2 24.Qxa7 Kf7 25.Qa3 Bd5 would have reached an approximately equal endgame. 19.Nf3!? was an interesting idea in Gavrijski – Koronowski, corr. 2012. I recommend 19...Qxc4N, capturing a pawn with the intention of returning it with interest, as seen after 20.Qxc4 Nxc4 21.Nd4 Ne5 22.Nxe6 fxe6 23.f4 Ng4 24.Rxc6 Rfd8 25.Rxe6 Kf7 26.Ra6 Rb2© when Black’s activity fully compensates for the pawn deficit. 19...Qxc4 20.Qa5 20.Qe5 Qb5 21.Rc5 Qb4 22.Rxc6 occurred in N. Pert – Griffiths, Sheffield 2011, when 22...Rfc8!N 23.Nd4 Qb2 24.Rd1 Rxc6 25.Nxc6 Qxe5 26.Nxe5 Rb2ƒ would have made things tricky for White. Another game continued 20.Qxc4 Bxc4 21.Na5 Bd5 22.Nxc6 Bxc6 23.Rxc6 Rfc8 24.Rac1 Rxc6 25.Bxc6= with an equal ending, Howell – Gelfand, Tallinn 2016. 20...Qb4 21.Qxb4 21.Qxa7 Rb7 22.Qa6 Rb6 23.Qa5 Bxb3 24.Qxb4= was another equal position in V. Ivanov – R. Gordon, corr. 2012. 94

21...Rxb4

22.Nd4 22.Nc5?! is inferior. In Lemos – Fabian, Buenos Aires 2011, Black should have continued 22...Bd5!N 23.Nd3 Rb7 24.f3 Rc8³ when White still needs to prove his compensation. 22...Bd5 23.Nxc6 Bxc6 24.Rxc6 Rc8 25.Rxc8† Nxc8 26.Rc1 Nd6 27.Rc7 Rb1† 28.Bf1 Rb7= The endgame was equal in Jianu – Shishkin, Baia Sprie 2012. 17...Rab8 18.Qd3! 18.Qc3† Kg8 leads back to the analysis above. The text move is slightly trickier, as White keeps the check in reserve while protecting the c4-pawn with his rook, thus freeing up the knight for active duty. 18...Rfc8 19.Nb3 Qb4 20.Qd4† f6! Accuracy to the end. Playing by analogy with the 17.Qc3† line with 20...Kg8?! does not work so well here in view of 21.c5 Nf5 22.Qe5 a5 23.Be4 a4 24.Bxf5 Bxf5 25.Nd4² when Whites was able to press with a good knight against a bad bishop in Ovalle – Krueger, corr. 2014. 21.Qxa7 21.c5 is not so scary now, as 21...Nf7 is a convenient reply. My analysis continues: 22.Qxb4 Rxb4 23.Nd4 Nd8 (23...Bd5 24.Bxd5 cxd5 25.Ne6† Kg8 26.Rab1 Rc4 27.Nf4 Ne5 28.Nxd5 R8xc5 29.Nxe7† Kf7 30.Rxc4 Nxc4 31.Rb7 Nd6 32.Rxa7 Nb5 33.Rb7 Nd6= is also acceptable)

95

24.a3 Ra4 25.Nxe6† Nxe6 26.Rc3 Kf7 27.f4 h5 28.Kf2 Rc7 29.Bf1 Nf8 30.Bh3 Ne6= With a level endgame. 21...Rb7 22.Qc5 Bxc4 23.Qxb4 Rxb4

24.Na5 Ra4 25.Nxc6 Kf8 White has temporarily won a pawn, but he cannot hold on to it because the a2-pawn is hopelessly weak. 26.Rc2 Rca8 27.Nb4 R8a5 28.Rb2 Bf7 With the idea of ...Nc4. 96

29.Nc6 Rxa2 30.Raxa2 Rxa2 31.Rxa2 Bxa2 The players agreed a draw in De Jong – Ljubicic, corr. 2016. B4) 7.Qb3

This was Rubinstein’s choice against Grünfeld in their 1923 game, and is quite a popular continuation among modern grandmasters including Mamedyarov, Vachier-Lagrave, Giri, Svidler, Yakovenko and Wang Hao. As for opening books, it was recommended by Burgess in A Cunning Chess Opening Repertoire for White. White’s idea is to defend the c4-pawn while putting pressure on b7 and d5, thus hindering the development of Black’s queenside pieces. 7...Qb6 We have no objection to a queen exchange. The c8-bishop is ready to go to f5 or e6, and in some lines the queen may go to a6. 8.Nc3 This is the natural developing move, which puts pressure on Black’s centre. Other continuations are worse, for instance: 8.Qxb6?! axb6 9.cxd5 Nxd5 Black’s queenside pawns are strong and the open a-file helps him. 10.e4?! 10.Nc3?! Nxc3 11.bxc3 increases Black’s structural advantage on the queenside, and 11...b5 12.Bf4 Be6 13.a3 Nd7µ was unpleasant for White in R. Thomas – Daugman, England 2014. The a3- and c3-pawns are weak and Black’s knight can easily make its way to c4. 97

White should play more modestly with 10.Bd2 Rd8 11.Nc3 Be6 12.a3 as in Kozma – Trifunovic, Marianske Lazne 1960, although 12...b5N³ still leaves Black with the better game. 10...Nb4 11.Na3 Be6 12.Bg5 Re8 13.Bd2 In Antic – Kleiser, Vienna 2011, Black should have continued:

13...N8a6N Defending the knight and threatening ...Bxa2. 14.b3 Red8 15.Bc3 b5 Black already has a decisive advantage. The pawns on d4 and a2 are weak, and the knight on a3 is terrible. 8.c5 This is not the most timely way to gain space. 8...Qxb3 9.axb3 Na6 10.Nc3 10.Bd2 Ne4 11.Bc3 Re8 12.b4 Nc7 13.Ne5 f6 14.Nd3 Nb5³ saw Black’s knights take up excellent positions in Suez Panama – Longa Yauca, Gibraltar 2016. 10...Nd7 11.Rd1 Strengthening the d4-pawn. 11.Ra4 is another way to protect the central pawn, which also stops the knight from coming to b4. After 11...e5 12.e3 Nc7 13.Bd2 e4 14.Ne1 Black had seized a lot of space in Ernst – l’Ami, Groningen 2003, and now 14...b6!N would have hit White’s pawn chain while also threatening to win the exchange with ...Ba6. White has nothing better than 15.f3 (15.cxb6 Nxb6 16.Ra1 Ba6–+ is even worse for him) 15...exf3 16.Nxf3 bxc5µ when Black enjoys a healthy extra pawn. 11.e4 dxe4 12.Nxe4 looked active in Teske – Balinov, Graz 1996, but 12...Nf6N 13.Nc3 Nb4³ would have given Black some advantage thanks to the great outpost on d5. 11.Bf4 Re8 12.e4 dxe4 13.Nxe4 Nf8 14.Ra4 was seen in the grandmaster duel Nikolic – 98

Marjanovic, Belgrade 1987. I recommend improving with 14...Be6N 15.Nfd2 Red8 16.Be3 Bd5³ when the pawns on b3 and d4 are weak; the latter in particular can be targeted by ...Ne6 and ...Nc7-b5. 11...e5 12.e3 Re8 13.Bd2 This occurred in Volodin – Duzhakov, St Petersburg 2018. I recommend a novelty:

13...e4N 14.Ne1 Nf6 15.Bf1 Nc7 16.b4 a6 17.h3 Bf8³ The position is closed and it will be hard for either side to break through. However, White’s pawn structure on the queenside is completely frozen while Black has some flexibility on the kingside, so he can certainly try for a win. 8...Rd8 Black defends the d5-pawn and takes aim at d4. 9.Rd1 White maintains the tension and asks Black how he intends to strengthen his position. White has tried broadly the same range of alternatives as on the previous move, none of which offer him anything special: 9.Qxb6?! axb6 10.cxd5 Nxd5 11.Nxd5 Rxd5 once again favours Black:

99

12.Rd1 (12.e3 occurred in Drygalov – Mischuk, Kolomna 2016, when 12...Rb5!N 13.Re1 Bf5 14.Bf1 Rb4 15.a3 Nd7³ would have given White problems with the development of the queenside pieces) 12...Rb5 13.e4 Be6 14.a3 Now in Rogozenko – Donchev, Debrecen 1992, the simple 14...Nd7N³ would have been promising for Black, as it is difficult for White to develop his queenside – the rook on b5 does an excellent job. 9.cxd5?! is another premature release of the central tension: 9...Qxb3 10.axb3 Nxd5 11.Bg5 Kf8

12.Rfc1 (12.Rfd1 was seen in Gormally – Collins, England 2015, when 12...Nb4N 13.Rd2 h6 14.Bf4 Be6³ would have targeted the weak pawn on b3) This position was reached in Thybo – Yankelevich, Stroebeck 2017. Here I recommend the simple 12...Nb4N 13.Bd2 N8a6³ when the b3-pawn is a longterm weakness and the knight on b4 is an excellent blockader. 100

A contest between two creative GMs continued: 9.Bd2 Be6 10.Na4 Qxb3 11.axb3 Nbd7 12.Rfc1 This was Rapport – Jones, Isle of Man 2018, and here I would recommend rerouting the light-squared bishop to a better square:

12...Bf5N 13.h3 Be4÷ Black has a completely reliable position. 9.Bf4 Some strong players have tried this move, but the bishop is not particularly well placed here. 9...dxc4 10.Qxc4 Be6 11.Qd3 Na6 12.Qc2 12.Qd2 c5 13.Be5 cxd4 14.Bxd4 Qa5 15.Rfd1 occurred in Maletin – Kapnisis, Plovdiv 2008, when 15...Nb4!N 16.Qf4 Nc6 17.Be3 Nd5 18.Nxd5 Bxd5³ would have left Black with the more active pieces, as Avrukh points out. 12.Be5 was played in Maletin – Vokarev, Nizhnij Tagil 2007, when 12...Bh6!?N would have been quite promising. After 13.Qc2 Nb4 14.Qb1 Nd7÷ Black will either trade his knight for the enemy bishop or damage White’s pawn structure if the bishop retreats to f4. 12...c5 13.Na4 Qb5 14.Nxc5 We have been following Alekseev – Radjabov, Astrakhan 2010. A natural improvement is:

101

14...Rac8N 15.Rfd1 Nxc5 16.dxc5 Nd5÷ Black enjoys full compensation for the pawn due to a number of factors: the pawns on c5 and b2 are vulnerable, the g7-bishop is strong and ...Nxf4 might be awkward. 9.Qa3!? This has been tried in some correspondence games. A logical continuation is: 9...dxc4 10.Qxe7 Re8 11.Qa3 11.Qd6?! Bf5 12.Qf4 was a strange manoeuvre in Lauer – Cuccumini, corr. 2013. I recommend 12...Nbd7N 13.Re1 Nd5 14.Qh4 Nxc3 15.bxc3 Qa5³ when Black has the upper hand in piece activity. 11...Na6

102

12.Qa4 I also checked 12.b3!?N 12...cxb3 13.axb3 (13.Qxb3?! Be6 14.Qd1 Bf5 15.e3 Ne4³) 13...Bf5 14.e3 Ne4 15.Na4 Qd8 16.Nh4 Bd7÷ when the position is complicated but White does not have a shadow of an advantage. 12...Qb4 13.Qxb4 Nxb4 14.Bd2 Be6 15.Rac1 Rad8 16.a3 Nbd5 17.Rfe1 Nxc3 18.bxc3 Ne4÷ This was Cruzado Duenas – Akdag, corr. 2015. I would slightly prefer Black’s position because of the strong knight on e4 and the potential weaknesses on c3 and a3. 9.c5 Once again, it is not profitable for White to gain space on the queenside, although many strong grandmasters have played this way, and it was also recommended by Burgess in the aforementioned book. 9...Qxb3 10.axb3 Compared to the 8.c5 line in the note to White’s previous move, Black has lost some flexibility as he has already committed his rook to d8. This is far from a useless move though; if White ever plays e2-e4, the rook will work optimally along the d-file.

10...Na6 11.Bf4 Nd7 12.Rfd1 f6! This is the most logical way of preparing ...e5. Burgess only considers 12...Re8 but, as explained above, the rook has the potential to be well placed on d8, so it is too early to spend another tempo moving it away. 13.e4 The slower 13.Bd2 Nf8 14.Ra4 Nc7 15.b4 e5 16.Rda1 a6³ gave Black the better game in Afloarei – Dreisch, corr. 2016. White lacks a good plan, while Black can develop his play in the centre and on the kingside. 13...dxe4 14.Nxe4 Nf8 15.Bf1 We have been following Li Bo – Lin Yi, Daqing 2018. Here Black can improve with: 103

15...Ne6N 16.Be3 Nec7 17.Bc4† Kf8³ Black has the better pawn structure and can improve his position by moving the a6-knight to b4 and the c7-knight to d5, with a great blockade. Finally, some high-level games have continued: 9.h3 Be6 10.c5 10.Ng5 was an interesting try in Aronian – Ivanchuk, Wijk aan Zee 2006, but 10...dxc4!N 11.Qa3 Nd5 leads to complications which favour Black. For instance: 12.Nxe6 fxe6 13.e4 Nb4 14.Na4 Qa5 15.Nc5 Qxa3 16.bxa3 Bxd4 17.Nxe6 Bxa1 18.Nxd8 Nd7 19.Nxb7 Nd3 20.Be3 Bb2³ 10...Qa6 11.Bf4 I also checked 11.Ng5N 11...Bc8 12.Bf4 h6 13.Nf3 Be6÷ when Black has a comfortable position, with ...Nbd7 and ...b6 on the agenda. 11...h6 12.a4 Here I found an improvement over Inarkiev – Dominguez Perez, Loo 2014.

104

12...Nbd7N 13.Qa3 13.a5? allows a simple tactic: 13...Nxc5 14.dxc5 d4µ 13...Ne4÷ Black is not worse at all.

9...Bf5 10.Ne1 White puts pressure on the d5-point and prepares to deploy the knight on d3. As usual, 10.Qxb6?! axb6 only helps Black: 11.cxd5 Nxd5 12.Nxd5 Rxd5 13.Bg5 Rb5 14.Rd2 Kf8 15.Bf4 Nd7³ gave Black some advantage in Grabarczyk – Engel, Hamburg 2016, due to the active rooks putting pressure on White’s queenside.

105

10.Ne5? is an unfortunate choice because of 10...Qxb3 11.axb3 Bc2, winning a pawn. 12.Rd2 Bxb3 13.c5 was played in Kilgus – Manik, Vienna 2015, when Black’s best would have been:

13...a5!N 14.Ra3 a4 15.Nxa4 Bxa4 16.b3 Ne4 17.Bxe4 dxe4 18.bxa4 Bxe5 19.dxe5 Rxd2 20.Bxd2 Nd7µ White suffers from pawn weaknesses on a4, c5 and e5. 10.Qa3?! is unsuccessful here. 10...dxc4 11.Nh4 Be6 12.Qxe7 Na6 13.e3 occurred in Schalkwijk – Cardoso Garcia, corr. 2011, when it would have been best to play:

13...Nd5N 14.Nxd5 cxd5 15.Nf3 Bg4³ Black’s pieces are harmoniously arranged, while White lags in development. As usual, 10.cxd5 Qxb3 11.axb3 Nxd5 leaves White with a somewhat damaged pawn structure. 106

12.Bg5 f6 13.Bd2 Nb4 14.e4 Bg4 15.Rac1 (I also checked 15.d5N 15...Nd7 16.h3 Bxf3 17.Bxf3 Nc2 18.Ra4 Nb6 19.Ra2 Nd4 20.Bg2 Nxb3 21.Be3 Nc4 22.Bxa7 Kf8÷ when Black is not worse) 15...N8a6 16.Be3 Prohaszka – Areshchenko, Abu Dhabi 2016. I recommend improving with:

16...Nc7N 17.Rd2 Bf8 18.Ne1 Be6 19.d5 cxd5 20.Ne2 Rd7 White’s compensation for the pawn is questionable. Some elite players have tried: 10.Bf4 Qxb3 11.axb3 Bc2 12.Rdc1 Bxb3 13.c5 White has enough compensation for the pawn, but Black still has a choice between two moves which lead to approximate equality.

13...Ne4!? 107

This was chosen by Grünfeld expert Nepomniachtchi. 13...Bc4 14.Bxb8 Rdxb8 15.Ne5 Bb5 16.Nxb5 cxb5 17.c6 e6 18.e3 bxc6 19.Nxc6 Rb7 20.Ra6 was equal in Shirov – Gelfand, Tromso (ol) 2014. White will pick up the pawn on a7, after which inevitable simplifications and the opposite-coloured bishops make a draw highly probable. 14.e3 Nd7 15.Nxe4 dxe4 16.Nd2 Bd5 17.Bxe4 Bxe4 18.Nxe4 f5 19.Ng5 e5 20.dxe5 Bxe5 21.Rd1 Bxf4 22.Ne6!? A good try, but the position remains equal all the same. 22...Bg5! 23.Nxd8 Rxd8 24.f4 Be7 25.Rxa7 Nxc5 26.Rxd8† Bxd8 27.b4 Ne6 28.Rxb7 Bc7= Svidler – Nepomniachtchi, Sochi 2016.

10...Be6 The best defence is an attack: Black overprotects the d5-pawn while threatening ...dxc4. 11.c5 Finally, White removes the tension from the centre. 11.Na4 Qxb3 12.axb3 Na6 13.Nd3 Bf5 14.Ndc5 Nxc5 15.Nxc5 b6 16.Nd3 a5= gave Black comfortable equality in Schulz – Ganiev, corr. 2014. 11...Qxb3 12.axb3 Na6 13.Ra4 The rook is well placed here, as it guards the d4-pawn while supporting a future advance of White’s b-pawn. 13.Bf4 was played in Cumming – Van Wieringen, corr. 2014. I recommend 13...h6N 14.h3 (14.h4 Nh5 15.Bc1 Bf5 also leads to balanced play) 14...g5 15.Be5 Nd7 16.Bxg7 Kxg7= with approximate equality. 13.f3 Nd7 14.Ra4?! was played by a strong GM in Wang Hao – Durarbayli, Shamkir 2014, when Black 108

missed a powerful idea:

14...Ndxc5!N 15.dxc5 d4 16.Rxa6 dxc3 17.Rxd8† Rxd8 18.Rxa7 cxb2 19.Bxb2 Bxb2 20.Nd3 Bd4† 21.Kf1 Bxb3 22.Rxb7 Bc4µ Black is obviously better because of the two powerful bishops, and in the case of 23.Rxe7? Rb8 24.Re4 Rb1† 25.Ne1 Rd1 26.f4 Be6 White must give up the exchange: 27.Rxd4 Rxd4 28.Bxc6 Rc4 Black has good winning chances. 13.h3!? was an interesting try in Bruzon Batista – Gauri, Philadelphia 2019. I suggest an improvement:

13...Bc8!?N The bishop retreats to prepare ...e5. 14.Nc2 14.Bf4 is no better in view of 14...Nd7 15.Nf3 f6 16.e4 dxe4 17.Nxe4 Nf8 18.Nc3 Ne6÷ when 109

the d4-pawn is weak. 14...Bf5 15.Ne3 Bd7 16.Nc2 16.g4 h5 17.f3 e5 18.dxe5 Ne8 19.Nc2 Bxe5÷ gives Black a good game. 16...h5 17.Bg5 Re8 18.Bxf6 Bxf6 19.b4 Nc7 20.b5 White has made some progress on the queenside but Black keeps a solid position and his bishop pair could prove useful in the long run. My line continues: 20...Reb8 21.f4 e6 22.bxc6 bxc6 23.b4 h4 24.g4 g5 25.e3 gxf4 26.exf4 Bg7= The position remains dynamically balanced. Black has an extra pawn island, but he is poised to open things up with ...f6 and ...e5, when his bishop pair will start to become more relevant. 13...Bf5 The bishop seems to be swinging on a pendulum but it is worth moving it again to prepare ...e5. 14.Bf4

14...Ne8!?N This certainly deserves attention, although the decision will depend on how you evaluate the tactical line below. The game continued: 14...Nd7 Preparing ...e5. If that was the end of the story, Black would have the better game, but White could have implemented an interesting tactical resource. 15.Nxd5!N 15.Bg5?! is a waste of time. In Maghsoodloo – Wang Hao, Chengdu 2017, the simplest reply would have been 15...f6 16.Bd2 Nc7 17.b4 e5³, reaching a typical situation where Black can look 110

to develop his initiative in the centre and on the kingside, while White’s play on the queenside is deadlocked. 15...cxd5 16.c6 Nac5

17.c7 17.cxb7!? Nxb7 18.Bxd5 is another interesting possibility, leading to a double-edged endgame. 17...Nxa4 18.cxd8=Q† Rxd8 19.bxa4 e6 20.h3 h5 21.f3 e5 22.Bg5 Re8 With approximate equality. 15.Rda1 Nac7 The d4-pawn still requires the defence of the rook on a4, so Black can afford to leave his a-pawn undefended for a moment before consolidating with ...a6. Here is a sample line: 16.h3 h5 17.e3 a6 18.Nf3 Ne6 19.Rb4 Nxf4 20.gxf4 Rab8= Black has the two bishops but they cannot display all their strength in a mostly blocked position, so the chances are approximately equal. C) 5.cxd5 cxd5

111

This is the most common continuation of all. White removes the central tension and, to some degree, expands the field of activity of his bishop on g2. 6.Nf3 6.Nc3 Bg7 leaves White with nothing better than 7.Nf3, transposing to variation C2 below. Other options such as 7.e3 followed by Nge2 exist of course, and are occasionally tested by players looking to avoid theory, but we don’t need to prepare specially for such ideas. 6...Bg7 We will analyse C1) 7.0-0 and C2) 7.Nc3. An immediate leap into the centre was also tested at a high level: 7.Ne5 Ne4 8.Nd2 Nd7 9.Nxd7 Qxd7 10.Nxe4 dxe4 11.Bxe4 Qxd4 12.Qxd4 Bxd4 13.0-0 Bh3 14.Rd1 (14.Bxb7 Bxf1 15.Bxa8 Bxe2 16.Bc6† Kd8 17.Bf4 e5 is also equal, and after 18.Bxe5 Bxe5 19.Re1 the players agreed a draw in Vorobiov – Areshchenko, Bad Wiessee 2016) 14...0-0-0 15.Bf4 Bxb2 16.Bxb7† Kxb7 17.Rdb1 Rd1† 18.Rxd1 Bxa1 19.Rxa1 Having turned the position into a lifeless desert, the players agreed a draw in Mamedyarov – Gelfand, Sochi 2014. C1) 7.0-0

112

White takes care of his kingside before making any decision about how to develop the queenside pieces. 7...0-0 8.Nc3 Trying to avoid this natural and logical continuation leads nowhere for White: 8.Bf4 Ne4 9.Nbd2 (9.Nc3 is covered under 9.Bf4 in the notes to the main line below) 9...Nc6 10.Be3 Bf5 11.Nb3 Qd7 12.Nh4 Bh3ƒ White already had to fight for equality in Schwefer – Degro, Germany 2010. Black has no problems after: 8.Ne5 Ne4 9.Nd2 9.e3 Nc6 10.Nxc6 bxc6 11.Nd2 (11.Qc2 is well met by 11...Qb6 12.Nc3 Nxc3 13.Qxc3 Re8³ followed by ...e5) was once played against me in an online game, when 11...Nd6!?N 12.Nb3 e5÷ would have been the best route to a comfortable position for Black. 9...Nd6 10.b3 In Krogius – Balashov, Leningrad 1971, it would have been logical to play:

113

10...Nd7N 11.Ndf3 Nf6 12.Nd3 Bf5 13.Nfe5 Qa5÷ Black is in no way worse, and is well placed to fight for the open c-file. White achieves nothing with: 8.Qb3 Ne4 9.Nc3 Nc6! Black is already pressing for the initiative. 10.e3 Neither 10.Nxd5?? Be6–+ nor 10.Qxd5?? Nxc3 11.Qxd8 Nxe2†–+ is an option for White. 10.Rd1 was well met by 10...Nxc3 11.Qxc3 Bg4 12.Be3 Bxf3 13.Bxf3 e6 14.Kg2 Qb6 15.Rac1 Rfc8ƒ when the vulnerable d4-pawn made things a bit awkward for White in Wieser – Jicman, Graz 2014. 10...Nxc3 11.bxc3 11.Qxc3 Bf5 12.Bd2 Qb6 13.Ne5 Bxe5 14.dxe5 Rfd8 15.Rfc1 Be4ƒ is also promising for Black. 11...Na5 12.Qb4 b6 13.Nd2 Ba6 14.Re1 Rc8³ The pressure along the c-file gave Black the better chances in Haffner – Riess, Hanerau 2015. 8...Ne4 This move equalizes without much trouble at all. 9.Nxe4 This continuation has the drawback that the d4-pawn is weakened even more, and the f3-knight will be forced to move again. Alternatives are also nothing special for White: 9.Qb3 leads back to the earlier note on 8.Qb3. 9.Bf4 Nxc3 10.bxc3 gives Black a favourable pawn structure. 10...Nc6 11.Rb1 b6 12.e4 dxe4 13.Ng5 114

was seen in Granberg – O. Larsen, Snekkersten 1978, when it would have been logical to continue:

13...Bb7N 14.Nxe4 Na5³ Black blockades the hanging pawns and will exert pressure on the queenside. Several grandmaster games have continued: 9.Ne5 Nxc3 10.bxc3 Nc6 11.Nxc6 11.f4 does not improve White’s position. 11...Na5 12.e4 dxe4 13.Bxe4 Be6 14.f5?! occurred in Arango Arenas – Florez, Cali 2016, when 14...Bxf5!N 15.Bxf5 Bxe5! 16.Qf3 Bg7 17.Bh3 b6³ would have left White with insufficient play for the lost pawn. 11.Nd3 b6 12.e4 dxe4 13.Bxe4 Bb7 14.Re1 was seen in N. Petrov – Durarbayli, Skopje 2015, when the simple 14...Qd7N 15.Qa4 Rfd8 16.Bg5 h6 17.Bf4 Rac8³ would have given Black a typical plus due to the weakness and lack of mobility of White’s hanging pawns. 11...bxc6 12.Qa4 This generally leads to simplifications and an early draw. 12.Ba3 Ba6 13.Re1 Re8 14.e4 e6 15.e5 is more ambitious, but after 15...Bf8 16.Bxf8 Rxf8 17.Qa4 Bb5 18.Qa3 Qc7 19.Bf1 Bxf1 20.Rxf1 c5ƒ Black could think about more than equality in Nyzhnyk – Piorun, Wroclaw 2011. 12...Qb6 13.Ba3 Qa6 14.Qxa6 Bxa6 15.Rfe1 Rfe8 16.e3 e6

115

17.Bf1 Bxf1 18.Kxf1 Bf8 19.Bxf8 Kxf8 20.Rab1 Reb8 21.Rb3 Rb6 22.Reb1 Rab8 23.Ke2 Ke7 24.Kd3 Kd6 25.c4 c5= The position remained absolutely equal in Wen Yang – Li Chao, China 2017. 9...dxe4 10.Ne5 The knight is well placed in the centre, although White will have to watch out for ...f6 ideas. 10.Ng5 gives White nothing after 10...Qxd4 11.Qxd4 Bxd4 12.Nxe4 Nc6, for instance13.Bh6 (13.Nc3 Rd8 14.Rd1 Be6 15.Be3 Bxe3 16.fxe3 occurred in Donner – Botvinnik, Wijk aan Zee 1969, when 16...Rxd1†N 17.Rxd1 Rc8³ would have been slightly better for the sixth World Champion) 13...Rd8 14.Rab1 Be6 15.b3 Rac8= Black was comfortably equal in Smejkal – Jansa, Smederevska Palanka 1982.

116

10...Qd5 The queen defends the e4-pawn and prepares ...Rd8 to attack the d4-pawn in some lines. 11.b3 White prepares to develop the bishop on b2 in order to support both the pawn and the knight. 11.Qa4? b5 12.Qb4 a5 13.Qxe7 Qxd4 14.Bf4 Ra7µ left White in a difficult situation in Schoebel – Zuse, Karlsruhe 1988. His knight on e5 is not yet lost due to tactical reasons, but its future is still extremely shaky. 11.f3?! Bxe5! 12.dxe5 Qxe5 13.fxe4 Be6 14.Bh6 Re8 15.Qd2 Nc6³ gave Black a positional advantage in Gines Abad – Fernandez Barrera, Almansa 2014, as the loss of his dark-squared bishop is less important than White’s doubled e-pawns, which are not only weak but also restrict the bishop on g2. 11.e3?! is too passive, and 11...Bxe5 12.dxe5 Qxe5 13.Qc2 Bf5 14.Bd2 Nd7 15.Rac1 Nf6 16.Bc3 Qe6³ left White with insufficient compensation for the pawn in Concio – H. Nguyen, Hanoi 2018. 11.Bf4 f6 12.Nd3 Nc6 gives White some coordination problems, and the further 13.Be3 Rd8 14.Nf4 Qf7 15.d5 e5 16.dxe6 Bxe6 17.Qa4 f5 18.Nxe6 Qxe6 19.Rab1 Nd4ƒ left White facing an unpleasant initiative in S. Parker – Jablonski, corr. 2011. Another possibility for the development of the bishop and the protection of the pawn lies in: 11.Be3 Nd7 12.Nxd7 12.Qc2 f5! is excellent for Black. 13.f4 Nb6 14.Rfd1 Be6 15.b3 a5 16.Bf2 Rfc8 17.Qb2 a4 18.Rab1 axb3 19.axb3 occurred in Wojtaszek – Miroshnichenko, Polanica Zdroj 2001, when 19...Qa5Nµ would have made it hard for White to defend against the positional plan of ...Nd5 combined with a queen exchange and invasion of the rooks. 117

12.Nd3 Nf6 13.Qa4 was seen in Granda Zuniga – Krasenkow, Madrid 1998, when the precise 13...Bf5!N 14.Rac1 Rfd8 15.Ne5 Nd7 16.Nxd7 Bxd7 17.Qc4 Bc6 18.Qxd5 Bxd5 19.b3 Rac8³ would have given Black the upper hand thanks to his more active bishops and the slight weakness of the d4-pawn. 12...Bxd7 13.f3 exf3 14.Bxf3 We have been following Skousen – Chittka, Helsingor 2018. I recommend:

14...Qa5!N 15.Qb3 Rad8 16.Rfd1 16.Qxb7?! Bc8 17.Qe4 Bf5 18.Qf4 leaves White uncoordinated, and after something like 18...a6 19.g4 e5 20.dxe5 Bxe5 21.Qg5 Be6³ Black’s initiative is worth more than White’s extra pawn. 16...b6 17.Bd2 Qa6ƒ The position is close to equal but I would take Black on account of his more compact pawn structure. 11...Nc6 12.Bb2 White continues with his plan. Trading pieces does not bring White any benefits: 12.Nxc6 Qxc6 13.Bb2 13.Be3 Bf5 14.Qc1?! (14.Rc1 seems better, although after 14...Qa6 15.Qd2 Rfc8÷ I would still choose Black due to the weak d4-pawn) The text move was seen in De Andres Gonalons – Nabavi, Basel 2006, when the improvement 14...Qa6!N 15.Qc4 Qd6 16.Rac1 Rac8 17.Qa4 a6³ would have left the d4-pawn weak and the e3-bishop passive. This position was reached in Hübner – Bacrot, Albert/Berlin 1998, and some other games. For some reason, no one has played the following logical continuation:

118

13...Qd5!?N 14.Rc1 Bf5 15.Rc5 Qd7 16.Qd2 Rfd8 17.e3 e5 18.Rd1 exd4 19.Bxd4 Bg4 20.Rdc1 Bxd4 21.exd4 Qxd4 22.Qf4 Rac8 White still has to prove his compensation for the lost pawn. 12...Bf5 13.Nxc6 bxc6! Now White has to reckon on the possibility of ...c5. The d4-pawn remains a target and Black can already play with virtually no risk. 14.Qc2 14.Rc1 is well met by 14...Rfd8 15.e3 c5!, for instance: 16.Qe2 cxd4 17.exd4 Rac8 18.Rxc8 Rxc8 19.Rd1

119

19...h5! 20.h3 Qd7 21.Kh2 a5 22.Ba3 a4 23.Bc5 e5 24.dxe5 Qe6 25.b4 Bxe5 26.Re1 h4ƒ Black skilfully built up a nasty initiative in Norman – Killer, corr. 2016. 14.e3 c5 15.Qe2 cxd4 16.Bxd4 was played in Cumming – Kuhne, corr. 2013. From a human point of view, the most natural continuation would be 16...Bxd4N 17.Rfd1 e5 18.g4 Be6 19.exd4 f5 20.dxe5 Qxe5 21.Qe3 and now the most energetic idea is:

21...Rae8! 22.gxf5 gxf5 23.Qxa7 Ra8 24.Qe3 Kh8© Black has full compensation. White is still not worse but he cannot hold on to his extra pawn, for instance: 25.Kh1 Rg8 26.a4 Rg6 27.Qd4 Qxd4 28.Rxd4 Bxb3= 14...Bxd4 15.Bxd4 Qxd4 16.e3 Qe5 17.Qxc6 120

Despite Black’s doubled e-pawns, the position is equal, since White’s bishop is passive and Black will be the first to take control of the queenside files. 17...Rfd8 18.Rad1 Rac8 19.Qa6 Qc5 20.Rxd8† Rxd8 21.Qa4 Rd2 22.Bxe4 Bxe4 23.Qxe4 Rxa2 24.Rd1 Qf5 25.Qxf5 The players agreed a draw in Kanep – Esipenko, Helsinki 2016. C2) 7.Nc3

White develops his knight and fights for the centre, postponing castling for the time being. 7...Ne4 It is useful to be able to trade off the c3-knight at the right moment. The text move also clears the long diagonal for the g7-bishop. We will analyse the attempt to disrupt Black’s play with C21) 8.Qa4†, followed by the main line of C22) 8.Qb3. 8.0-0 is another popular continuation, when 8...0-0 leads back to variation C1 above. 8.Bd2 is hardly dangerous: 8...Nc6 9.0-0 0-0 10.e3 Bg4 11.h3 Bxf3 12.Bxf3 Nxd2 13.Qxd2 e6 14.Rac1 occurred in Vukic – Djuric, Cetinje 1992, when the simple 14...Rc8N would have been completely equal. 8.Ng5 This is a strange idea and after: 8...Nxc3 9.bxc3 0-0 10.0-0 h6 White risks becoming worse, for instance: 121

11.Nh3 11.Nf3 Nc6 12.Re1 b6 13.Ne5 occurred in Alm – Rother, Germany 2015, when 13...Nxe5N 14.dxe5 Be6³ would have left Black with the better pawn structure. 11...Nc6 12.Nf4 e6 13.e4 dxe4 14.Bxe4 Bd7 15.d5 exd5 16.Bxd5 Ne5 17.Bxb7 In Aronian – Anand, Saint Louis 2017, Black missed the strongest continuation:

17...g5!N 18.Ne2 Rb8 19.Bg2 Bb5µ Black has more than enough compensation for the pawn due to the activity of his pieces; the bishop on b5 coordinates perfectly with the knight to control the key light squares in the centre. We are also not afraid of: 8.Nxe4 dxe4 9.Ne5 f6!? This is the move to make if you want a fighting game. 9...Nd7 is a reliable equalizer, when 10.Nxd7 Qxd7 transposes to the earlier note on 7.Ne5. 10.Qa4† This occurred in Gustafsson – Glek, Deizisau 2000. I recommend a new move:

122

10...Kf8N 11.Nc4 Bd7 12.Qc2 I also considered another queen retreat: 12.Qa3 Be6 13.b3 Nc6 14.Bb2 f5 15.Rd1 Bxc4 16.bxc4 Qa5† 17.Qxa5 Nxa5 18.c5 b6= White’s bishop pair is not felt, as Black’s knight will be excellently placed on c4. 12...f5 13.Be3 Be6 14.f3 Nc6 15.Rd1 exf3 16.Bxf3 Bd5 17.0-0 Rc8 18.Qa4 Bxf3 19.exf3 Qd5 20.Ne5 Bxe5 21.dxe5 Qe6= Black has no problems. C21) 8.Qa4† Nc6 9.Ne5 Nxc3 10.bxc3 0-0 Black is ready to sacrifice a pawn for the sake of activity. 11.Nxc6 bxc6 12.Qxc6 Black also has no problems if White declines the pawn: 12.0-0 Qb6 13.e4 Nothing is given by 13.Ba3 Qa6 14.Qxa6 Bxa6 15.Rfb1 Rfb8 16.Bxe7 Bxe2 17.Bd6 Rd8 18.Bc7 Rdc8, and a draw resulted in Cori Tello – Nepomniachtchi, Batumi (ol) 2018, and several other games. 13.Re1 does not change much. 13...Qa6 has proven a reliable reply, and 13...Ba6 also seems fine, inviting 14.e4 with a transposition to the 14.Re1 note below. 13...Ba6 14.Rd1 14.Re1 e6 15.Bf4 Bd3 16.Rad1 dxe4 17.Bxe4 Bxe4 18.Rxe4 Rfd8 19.Ree1 Qb5 20.Qa3 Bf8 21.Qb3 occurred in Cerrato – Rook, corr. 2012, when I would prefer 21...Rd7N 22.Rb1 Qa6 with equal chances. 14...Bb5 15.Qa3 e6 16.Bf4 Rfe8 17.Bd6 123

Now in Kalinin – Yloenen, corr. 2013, it would have been natural to play: 17...Bf8= Black does not have a shadow of a problem. 12...Qa5 13.Bd2 13.0-0 Be6 14.Bf4 Rac8 15.Qb7 Rxc3 16.e3 Bf6 17.Bb8 occurred in Persson – Vasilev, corr. 2016. I think the simplest route to an equal position would have been: 17...a6N 18.Ba7 Rc7 19.Bb6 Rxb7 20.Bxa5 Rc8= Black has no problems. We have been following Nakhbayeva – Lei Tingjie, Moscow 2015. I found a useful improvement:

13...Rb8!N Sacrificing a second pawn in order to gain even more activity. 14.Qxd5 Rb5 15.Qc4 15.Qf3 Bb7 16.e4 Rb2 17.Qe3 Ba6 18.Bf1 Rfb8 19.Bxa6 Qxa6 20.Rc1 Rxa2© offers Black excellent compensation. An illustrative line is:

124

21.Qe2 Qc8 22.0-0 Rbb2 23.Rfd1 Bh6 24.f4 a5 25.Qd3 Qh3! Threatening to take on f4. 26.Qe3 e5! 27.dxe5 Qd7µ Black wins the bishop. 15...Ba6 16.Qd3 Rb2 17.Qe3 e5 18.d5 e4 19.0-0 Qxd5 20.Rad1 Qc4© Black has regained one pawn while keeping enough of an initiative to compensate for the other. C22) 8.Qb3

White continues to exert pressure in the centre; the d5-pawn now finds itself under attack. 8...Nxc3 9.bxc3 This is the correct decision. Despite the fact that the pawn becomes backward and potentially weak, it 125

is more important for White to offer the d4-pawn secure protection while avoiding misplacing the queen. 9.Qxc3?! is clearly worse: 9...Nc6 10.0-0 0-0 11.Be3 (11.Bf4 Bg4 12.Rfd1 Rc8 13.Qb3 Bxf3 14.Bxf3 Nxd4µ was poor for White in Volodin – Vachier-Lagrave, Internet 2018) 11...Bg4 12.Qd2 Bxf3 13.Bxf3 e6 14.Rac1 Qb6 15.Rfd1 Rfc8 16.Kg2 In Bochev – Badolati, corr. 2014, Black could have fought for the advantage with:

16...Na5!N 17.b3 Rxc1 18.Rxc1 Nc6³ The d4-pawn requires attention and White’s bishops are not making their presence felt at all. 9...0-0 10.Nd2 This is the usual move and certainly a logical choice. The bishop on g2 is now working at full capacity, pressing down on the d5 pawn. White also prepares e2-e4, which will further expand the bishop’s sphere of influence. 10.0-0 Nc6 11.Nd2 e6 transposes to the main line below. 10.Ba3 has been tried at the elite level, with the idea of preventing ...e6. 10...Nc6 11.Nd2 Be6 12.0-0 (12.Qxb7?! is inadvisable due to 12...Qa5 13.Qxc6 Qxa3 14.0-0 Rfc8 15.Qb5 Rab8 16.Nb1 Rxb5 17.Nxa3 Rb2ƒ when White is under serious pressure) 12...Qd7 13.Rfe1 Rac8 14.Qb5 Rfd8 15.Rac1 b6 16.e3 Now in Wang Yue – Gelfand, Beijing 2013, Black could have aimed for more than equality by means of:

126

16...Rc7!N 17.Bf1 h5 18.h4 a6! 19.Qb3 Ra8ƒ With some initiative. 10.Ne5 is a more active knight jump but it seems inferior to the main line. 10...Nc6 11.f4 was seen in Danielian – Havlikova, Reykjavik 2015, and here I recommend a new move:

11...Be6!N 12.Qxb7 If White does not capture the pawn, Black will simply play ...Na5, ...Rc8 and ...f6, with unpleasant pressure along the c-file. 12...Qa5 13.Nxc6 (I also checked 13.Bd2 Bxe5 14.Qxc6 Rfc8 15.Qb7 Bd6 16.Qb1 Rab8 17.Qd3 Bf5 18.Qe3 Ba3µ when White is clearly suffering) 13...Qxc3† 14.Kf2 Qxa1 15.Qb3 Bxd4† 16.e3 Bf6 17.Ba3 d4 18.Rxa1 Bxb3 19.Nxe7† Kg7 20.e4 Be6 21.e5 Bxe7 22.Bxe7 Rab8 23.Bxf8† Kxf8µ White faces a difficult endgame, as Black’s rook is extremely active and the a2-pawn is weak.

127

In the 21st century, flank attacks with h2-h4 seem to be an option at almost every step, and this is no exception. 10.h4 has been tried, although Black can mostly ignore the kingside demonstration: 10...Nc6 11.h5 Na5 12.Qb4 Bf5 13.hxg6 hxg6 14.Nh4 Be6 15.Bf3 Re8 16.Ng2 Rc8ƒ Black’s accurate play gave him some initiative in Ruiz Jarabo Pelayo – Jedinger, corr. 2015.

An important point is that the natural 17.Nf4 is well met by 17...Bf5! 18.Nxd5 Nc6 19.Qb3 (the excessively greedy 19.Qxb7? is refuted by 19...e5 20.dxe5 Rxe5 21.Ne3 Re7 22.Qb3 Na5 23.Qd1 Rd7 24.Nd5 Be6–+ when White suffers material losses) 19...e5 20.dxe5 Nxe5 21.Bg2 Nd3† 22.Kf1 Nc5© Black has more than enough compensation for the pawn, since the bishop on g7 is extremely strong, the c3-pawn is weak and the white king is misplaced. 10...e6 11.e4 White saves a tempo by delaying castling and immediately creates tension in the centre. However Black is not obliged to defend the d5-pawn. 11...Nc6! An excellent resource, preparing ...Na5. 12.0-0 White should get castled at some point – especially if he is considering exchanging on d5, in which case a check along the e-file could prove inconvenient. 12.exd5 is certainly not dangerous: 12...Na5 13.Qb4 exd5 14.0-0 b6 15.Re1 Be6 16.Nf1 Rc8 17.Ne3 Qd7÷ Black had a comfortable position in Kuznetsov – Cerrato, corr. 2012. 12.Ba3 Re8 13.0-0 Na5 14.Qd1 b6 transposes to the 14.Ba3 Re8 line, as covered in the notes to the main line below. 128

A top-level game continued 12.e5 b6 13.0-0 Ba6 14.Re1 Rc8 15.Qd1 Re8 16.a4 Na5 17.Re3 and now Black’s play from Caruana – So, Paris 2017, can be improved by means of:

17...Re7!N 18.Ba3 Rec7 19.Rc1 h5 20.Bb2 Bf8³ Black is slightly better due to his queenside pressure and the long-term weakness of the c3-pawn. 12...Na5 Black drives the queen away from attacking the d5-square, while preparing employment for the lightsquared bishop after ...b6. 13.Qd1! I believe this is White’s most precise continuation, for a few reasons which will be discussed shortly. The first is a tactical point which will be highlighted below. 13.Qb4 has been tried, but the queen looks clumsy here: 13...b6 14.Re1?! (White should settle for 14.exd5 exd5, transposing to the Kuznetsov – Cerrato game under the 12.exd5 line above) Now in Castellanos Rodriguez – Colmenares, Cesenatico 2015, Black could have seized some advantage with:

129

14...dxe4!N 15.Nxe4 Bb7 16.Bf4 Bd5³ Black has achieved full control of the d5- and c4-squares, and White’s queenside pawns will be weak in the long run. 13.Qc2 looks natural but after 13...b6 14.Ba3 Re8 15.Rfe1, as played in Rabinovich – Botvinnik, Moscow 1935, we can improve upon the play of Mikhail Moiseyevich by means of:

15...dxe4!N 16.Nxe4 Bb7 Black has a comfortable position and will establish a typical blockade on the d5- and c4-squares. The key tactical point is that 17.Nd6? Bxg2 18.Nxe8? loses to 18...Bc6 19.Nxg7 Qd5–+ when White’s king is caught. If the same thing happened with the queen on d1 instead of c2, White would defend with f2-f3.

130

13...b6 14.e5! This is the most challenging continuation. White captures space and can later think about a pawn storm with f2-f4, g2-g4 and f4-f5. Meanwhile, Black’s bishop is restricted by the e5-pawn. White tried a plan of transferring the knight away from the d2-square with 14.Re1 Bb7 15.e5 Qd7 16.Nf1 Rfc8 17.Qf3 in Fries Nielsen – Bjork, Graz 1978. However, after 17...Rc7N with the idea of doubling rooks along the c-file and pressuring the c3-pawn, Black’s position looks preferable. White can also activate his bishop before making any commitment in the centre: 14.Ba3 Re8 15.e5 15.Re1 is the same as the Rabinovich – Botvinnik game quoted earlier, except that White’s queen is on d1 here. Black has a choice of playable lines: a) Despite the important change in the queen’s position, 15...dxe4!? 16.Nxe4 Bb7 is still playable, the justification being 17.Nd6 Bxg2 18.Nxe8 Bc6 19.Nxg7 Qd5 20.f3 Kxg7÷ when Black has enough compensation to claim equal chances. b) 15...Bb7 is more solid. 16.e5 Rc8 17.Rc1 occurred in Kaster – Schwab, Trier 1992, when the simple 17...Ba6N 18.Re3 Qd7= would have reached a typical situation for this variation, with roughly equal chances. 15...Ba6 16.Re1 Rc8 17.Re3 Qd7 18.h4 This position was reached in l’Ami – Smirin, Stockholm 2013. Here I recommend a pendulum manoeuvre:

131

18...Bh6!N 19.f4 Bf8 It was worth provoking the advance of White’s f-pawn. Now his queen will be unable to utilize the f4-square and his king is slightly more exposed, which could make a big difference if Black breaks through along the c-file at some point in the future. 20.Bxf8 Rxf8 21.h5 Qe7÷ Black is not worse, as his pieces occupy good squares and the pressure along the c-file distracts White from his kingside ambitions. 14...Ba6 15.Re1 Rc8 The plans of both sides are determined by the pawn chains: White will look to attack on the kingside and Black will find counterplay on the queenside. 16.Re3 Re8! The point of this mysterious-looking rook move will soon become clear. 17.h4

132

17...Re7! Black gets ready to double his rooks and burden White with pressure on the backward c-pawn. 18.h5 White plays aggressively on the kingside but will have to consign one of his minor pieces to the passive defence of the c3-pawn. 18.Ba3N 18...Rec7 19.Bb4 deserved attention, when the bishop defends the pawn from a more active square. Nevertheless, after 19...Bh6 20.Rf3 Qd7ƒ I still like Black’s chances. 18...Rec7 19.Bb2 19.Nb1!?N 19...Bh6 20.f4 Bf8 21.hxg6 hxg6 22.g4 f6 23.exf6 Qxf6÷ reaches a complex and unclear position, although it is hard to believe that Black can be doing badly when White’s queenside pieces look so ridiculous. Defending with the bishop looks natural. This is where we should improve on Black’s play from an elite-level game.

133

19...Qg5!?N The game continued with 19...Nc4?, which releases tension prematurely and enables White to develop an initiative on the kingside. 20.Nxc4 Bxc4 21.hxg6 hxg6 occurred in Caruana – Karjakin, Tashkent 2014, when 22.Rf3!N b5 23.Bc1 a5 24.Bd2ƒ would have put Black under some pressure. 19...Bh6!?N 20.f4 g5!? is a second possible improvement, when 21.Qg4 Kh8 22.f5 exf5 23.Qxf5 g4÷ leads to an extremely difficult and double-edged game. 20.hxg6 hxg6 21.a4 Bf8÷ Black has his fair share of the chances. Unlike the game continuation noted above, we have avoided premature exchanges so White’s pieces remain cramped on the queenside. Conclusion The Fianchetto System is an important positional option, to which Black must respond carefully in order not to fall into a bad position. I chose the solid set-up with 3...c6 4.Bg2 d5 in order to restrict the g2-bishop. I find this to be the most reliable continuation, but its solidity does not mean that Black plays only for a draw – we saw many variations in the chapter in which he was able to seize the initiative. In terms of specific variations, 5.Qa4!? was the first major line we considered. This modern move was recommended by Avrukh, but 5...Bg7 6.cxd5 0-0! is an attractive idea which offers Black fine compensation for the sacrificed pawn. We then considered lines after 5.Nf3 Bg7 6.0-0 0-0 where White avoids trading pawns on d5. 7.Nbd2

134

is certainly no threat: Black gets a fine position with 7...a5 followed by ...a4 and ...Bf5, with chances to seize the initiative. 7.b3 is a better move but 7...dxc4! 8.bxc4 c5! is a good answer. Black will obtain the better pawn structure but he must be mindful of the power of the g2-bishop; ultimately, the chances are equal. Finally, 7.Qb3 keeps more tension in the position. 7...Qb6 is a good answer, especially as any premature releasing of the tension with cxd5, Qxb6 or c4-c5 is likely to benefit Black. Our third and final branch dealt with the symmetrical position after 5.cxd5 cxd5 6.Nf3 Bg7. If White develops routinely with 7.0-0 0-0 8.Nc3, then 8...Ne4 solves all Black’s problems and even enables him to fight for the advantage in numerous lines. It is slightly more challenging to delay castling with 7.Nc3, although even then 7...Ne4 is a good answer. The main line continues 8.Qb3 Nxc3 9.bxc3 when White can try to stir up trouble with Nd2 and e2-e4, but Black keeps a full share of the chances.

135

A) 4.f3?! 75 B) 4.h4?! 76 C) 4.Nf3 Bg7 79 C1) 5.g3 dxc4 6.Qa4† Nfd7 7.Qxc4 Nb6 79 C11) 8.Qb3 80 C12) 8.Qd3 81 C2) 5.h4!? c6! 82 C21) 6.Qb3!? 83 C22) 6.Bg5!? 84 C23) 6.cxd5 85

1.d4 Nf6 2.c4 g6 3.Nc3 d5 This position is key, and over the course of the book we will examine every plausible continuation White may try. In this chapter we will examine a variety of sidelines. We will start with A) 4.f3?!, before moving on to B) 4.h4?! and finally some rare tries after C) 4.Nf3 Bg7.

136

4.g3 c6 leads to the previous chapter; Black could also investigate 4...dxc4!? with a possible transposition to variation C1. 4.Bd2 can be met by 4...Bg7, when White hardly has anything better than 5.cxd5 Nxd5 which transposes to Chapter 11. 4.e4? looks absurd, and indeed 4...dxe4 5.f3 exf3 6.Nxf3 Bg7 7.Be2 c5 8.d5 0-0 9.0-0 e6 10.Bg5 exd5 11.Nxd5 Be6µ leaves White with no compensation for the missing pawn. Another overaggressive idea is: 4.g4? dxc4 5.h3 White’s attack has ended without starting. 5.g5 Nd5 is no fun for him either. For instance, 6.Bg2 c6 7.Qa4 was played in De Nucci – Mekhitarian, Campinas 2010, when 7...b5!N 8.Nxb5 Nb6 9.Qa5 cxb5 10.Bxa8 Nxa8 11.Qxb5† Bd7 12.Qxc4 Bg7–+ would have refuted White’s play. His structure has been weakened and Black’s minor pieces will be too strong. 5...Bg7 6.e4 This occurred in Wehmeier – Sygulski, Germany 1995. I recommend striking at the enemy centre:

6...c5!N 7.dxc5 7.d5 is no better in view of 7...b5 8.g5 Nh5 9.Nxb5 Qa5† 10.Nc3 Bxc3† 11.bxc3 Qxc3† 12.Bd2 Qe5–+ when White’s position is simply hopeless, as he is threatened with ...Qe4 and ...c3. 7...Qxd1† 8.Kxd1 Be6 9.Be3 Na6µ Black has a clear lead in development and the c5-pawn is weak. A) 4.f3?!

137

This is not the worst of White’s weird options, but it still looks pretentious. 4...c5! Black strikes at the centre with good effect, since White’s last move weakened his king and deprived the knight of a good square. 5.cxd5 This is White’s best try. The alternative is 5.dxc5 d4 6.Nb5 e5 7.Bg5 Be7 8.Nd6† Bxd6 9.cxd6 as played in Caron – Duplessis, Quebec 2005. After analysing the position, I propose the following improvement:

9...h6!N 10.Bh4 Nc6 11.c5 g5 12.Bf2 Qa5† 13.Qd2 Qxc5 14.e3 Qxd6³ Despite White’s advantage of the two bishops, Black stands better due to his lead in development and extra space in the centre. 5...Nxd5 6.Na4!? This surprising move leads to an interesting game. 6.e4 Nxc3 7.bxc3 Bg7 gives Black a pleasant version of a normal Grünfeld. 8.Bb5† Bd7 9.Bxd7† Qxd7 10.Ne2 cxd4 11.cxd4 Nc6 12.Be3 0-0 13.0-0 occurred in Gluth – Speck, Germany 1991, when a logical continuation would have been:

138

13...Na5!N 14.Rc1 b5÷ There arises a typical Grünfeld middlegame where Black’s chances are at least equal. Obvious ideas include ...Rac8 and ...Nc4, and in the future Black may create a strong passed pawn on the queenside. 6...Bg7! A promising pawn sacrifice. 7.Nxc5 Nc6 8.Nb3 Nb6 9.e3 e5

10.Ne2 10.dxe5 Qxd1† 11.Kxd1 Nxe5 12.Bb5† Ke7 13.Bd2 Rd8 14.Rc1 Be6© offers Black plenty of 139

compensation for the pawn. White’s king is unhappily stuck in the centre and Black has a powerful bishop on g7, while his knights can target the c4- and d3-squares. 10...Be6 11.dxe5 Qxd1† 12.Kxd1 0-0-0† 13.Nbd4 Bxe5 14.Ke1 Nb4 15.Kf2 Nd3† 16.Kg1 Kb8 17.b3 Nxc1 18.Rxc1 Rhe8© We have been following Troia – Jodar Arias, corr. 2012. Black has full compensation for the pawn, with two strong bishops and possibilities such as ...Nd5 to create tension around the weak e3-point. Overall it is obvious that White risks more than Black when following this path. B) 4.h4?! This demonstrably erroneous move has been used by such grandmasters as Azmaiparashvili, Moradiabadi, Kruppa and Cebalo. Usually they make this move hoping to surprise, confuse and overwhelm the opponent. 4...c5! All according to the ‘rules’ of chess: when encountering a flank operation, we respond with a blow to the centre. 5.dxc5 5.cxd5 Nxd5 6.dxc5 is hardly an improvement in view of 6...Nxc3 7.Qxd8† Kxd8 8.bxc3 Bg7 9.Kd2 Nd7 10.Ba3 Nf6 11.f3 Nd5 12.Bb2 Kc7µ when White’s extra pawn is not felt at all: the c5pawn will not last much longer, while White suffers from a lag in development, an ugly pawn structure and weak dark squares. 5...d4 6.Nb5

140

6...e5! This is the correct way to defend the central pawn: there is no need to fear a check on d6. 6...Nc6?! is less convincing. True, after 7.e3 e5 8.exd4 exd4 9.Bf4 Bxc5 10.Nc7† Kf8 11.Nxa8 Bb4† 12.Bd2 Qe7† 13.Qe2 Bxd2† 14.Kxd2÷ Black has good compensation for the material, but it’s not at all obvious if he can claim an advantage, unlike our main line. 7.e3 White tries to undermine the centre; other moves are no better. 7.Qa4? is a waste of time: 7...Nc6 8.Bg5 Bxc5 9.0-0-0 (the greedy 9.Bxf6 Qxf6 10.Nc7† Kf8 11.Nxa8 Bb4† 12.Kd1 Qxf2 13.Nf3 e4–+ reaches a position where the only thing I could recommend for White would be to resign) 9...0-0 10.e3 a6 11.exd4 Bg4 12.f3 axb5 13.Qxb5 Bxd4 14.fxg4 Rxa2 White’s further resistance is meaningless, Obodchuk – Vuelban, Backi Petrovac 2017. White can, of course, immediately trade his clumsy knight for our black bishop with 7.Nd6† Bxd6 8.cxd6 Qxd6 9.e3 Nc6 10.Be2, as played in Dambrauskas – Nguyen, Novy Bor 2013. However, White is too far behind in development and his position is virtually hopeless after:

141

10...Bf5!N Threatening to advance the pawn to d3. 11.g4 Bd7 12.h5 0-0-0 13.Kf1 Be6–+ White’s pieces are devoid of harmony and Black should win. 7.Bg5 Bxc5 8.a3 (8.Bxf6? runs into 8...Qa5†! 9.Nc3 dxc3–+ when White can already resign) was played in A. Burnett – Biolek, Marianske Lazne 2015, when Black’s strongest continuation would have been:

8...Be7!N 9.e3 a6 10.exd4 axb5–+ White has no compensation for the lost piece. 7.b4 is an overoptimistic attempt to retain the extra pawn. 7...Nc6 8.Bg5 (8.a3 a6 9.Nd6† Bxd6 10.cxd6 Qxd6µ leaves White far behind in development) In Danada – Kukel, Banska Stiavnica 2012, Black could have gone for the brutal approach: 142

8...Nxb4!N The justification is seen after 9.Bxf6 Qxf6 10.Nc7† Kd8 11.Nxa8 Bf5 12.Rc1 Bxc5 13.a3 Na2 14.Ra1 Nc3 15.Qd2 Kc8–+ when Black will round up the knight on a8 to reach something close to material equality, while White’s position is in ruins. 7.Nf3 Developing the second knight is slightly more resilient than some of the moves mentioned above, but even here White’s position is not to be envied. 7...Bxc5 8.Nxe5 a6

9.Nd3 9.Na3 Bb4† 10.Bd2 Bxd2† 11.Qxd2 Ne4 12.Qb4 (or 12.Qf4 Qa5† 13.Kd1 Nxf2† 14.Qxf2 Qxe5–+ leaves Black with both a healthy extra pawn and a big positional advantage) 12...Qc7 13.Nd3 Nc6 143

14.Qa4 0-0 gives Black a decisive advantage, as White will have to make too many concessions to complete development and ensure his king’s safety. 9...axb5 10.Nxc5 bxc4 11.Bg5 We have been following Daus – Spiteri, corr. 2007. Here I recommend:

11...h6!N 12.Bxf6 12.Bd2 Qe7–+ 12...Qxf6 13.e3 d3–+ White has serious problems completing development, and will most likely have to give up a piece for nowhere near enough compensation. 7...Bxc5 8.exd4 exd4 9.Nf3 I found another improvement over the existing games.

144

9...Nc6!N 9...0-0 10.Be2 Nc6 has been played and is quite promising for Black, but the text move is more precise: Black brings another piece into play and saves time by delaying castling. 10.Bd3 10.Be2 enables Black to threaten ...d3 and thus win a tempo: 10...a6 11.Na3 Bf5 12.Bd3 Qe7† 13.Kf1 Bxd3† 14.Qxd3 Nb4 15.Qd1 0-0-0–+ With his king in such a position, White will not last long. 10...a6 11.Na3 Qe7† 12.Qe2 Nb4! 13.0-0 Qxe2 14.Bxe2 d3 15.Bd1 0-0µ The d3-pawn is extremely strong and White will have a hard time completing his development. C) 4.Nf3 Bg7

145

This is a popular position which can lead to some major theoretical lines, but in this section we will stick to White’s rare 5th moves, specifically C1) 5.g3 and C2) 5.h4!?. 5.c5? is pointless, and 5...b6 6.Qa4† Bd7 7.Qb3 bxc5 8.Nxd5 cxd4 9.Nxd4 0-0 10.Nxf6† Bxf6 11.Qc4 c5 12.Qxc5 Na6µ clearly favoured Black in Stapinski – Lahdenmaeki, corr. 2010: he has already completed development, while White’s king is stuck in the centre. 5.Ne5? is a waste of time. After 5...0-0 6.e3 c5 7.Be2 Be6 8.Bf3 cxd4 9.exd4 Nc6 10.0-0 dxc4 11.Nxc6 bxc6 12.Bxc6 Rb8µ Black was clearly better in Preuss – Schmidt, Germany 2001, due to his superior pawn structure and excellent rook on b8, which makes it hard for White to develop his darksquared bishop. Finally, 5.Qd3? is another move which violates the basic principles of chess. Weirdly, the database indicates that some players rated in the region of 2400-2500 played this way, but since most of those games continued 5...dxc4 6.Qxc4, the most likely explanation is that 5.Qb3 was actually played, but the incorrect (in every sense of the word) 5.Qd3 was accidentally inserted. In any case, this silly move can be met by 5...Bf5N when White has nothing better than 6.Qd1, after which 6...0-0 followed by ...c5 obviously leaves White facing an uphill battle to equalize. C1) 5.g3 If White unexpectedly reverts to a fianchetto we are not obliged to transpose to the previous chapter by playing ...c6, but can instead simply take the pawn. 5...dxc4 6.Qa4† 6.Bg2 can be met by 6...c6 when White will struggle to regain his pawn. 7.Ne5 Ng4 8.Nxg4 Bxg4 9.d5 0-0 10.0-0 Qc8 11.Qc2 was the continuation of Petr – Novotny, Czech Republic 2008, when 146

11...Bh3N 12.Bxh3 Qxh3 13.Rd1 e6 14.d6 Nd7µ would have left White with no real compensation for the pawn, as the d6-pawn is securely blocked and thus not dangerous. 6...Nfd7 7.Qxc4 Nb6 We will consider C11) 8.Qb3 and C12) 8.Qd3. C11) 8.Qb3 This seems slightly the less appealing of the two options, as Black is likely to win a tempo with ...Be6 at some point. 8...0-0 9.Bg2 9.Bf4 Be6 10.Qd1 Nc6 11.e3 was seen in Milojevic – Mastrapovic, Vrbas 2011, when it would have been good for Black to play:

11...Nb4!N 12.Rc1 (or 12.Bg2 Bc4µ) 12...c5! 13.a3 (13.dxc5 Qxd1† 14.Rxd1 Nc4µ offers Black more than enough compensation and White is under serious pressure) 13...Nc6 14.Bg2 (after 14.dxc5 Nd5³ the g7-bishop is just a monster) 14...cxd4 15.Nxd4 Nxd4 16.exd4 Nd5³ Black is better because the d4pawn is isolated and weak.

147

9...Nc6 10.e3 A more aggressive alternative is: 10.d5!? Na5! (Avrukh recommends 10...Bxc3† 11.Qxc3 Nxd5 but I am not so keen on the position after 12.Qc4 Qd6 13.0-0© when White has some compensation for the pawn, due to Black’s weak dark squares) 11.Qb4 Nac4 12.0-0 a5 13.Qc5 (I also considered 13.Qb3N 13...e6 14.dxe6 Bxe6 15.Qc2 Qe7ƒ when Black has a comfortable position and a small initiative) 13...Qd6 14.Qxd6 cxd6³ In Almond – Jo. Hodgson, Paignton 2005, Black had more active pieces, especially the strong knight on c4, as well as pressure against the d5-pawn. 10...Be6 11.Qd1 Qc8! 11...Bc4 was less convincing after 12.b3 Ba6 13.Ne4÷ in Baumbach – Biedermann, corr. 2014, as White obtains counterplay with his knight going to c5. 12.0-0 Rd8 13.Ng5 Bc4 14.Re1 h6 15.Bh3 Here I recommend improving upon Black’s play in Perisic – Plenkovic, Bosnjaci 2014:

148

15...f5!N 16.b3 Ba6 17.Nf3 e5³ White is behind in development and Black has created unpleasant pressure against the d4-point. C12) 8.Qd3

8...0-0 9.Bg2 Mobilizing the queenside pieces with 9.Bf4 Nc6 10.Rd1 does not help White, as he is still behind in development. 10...Bf5 11.Qd2 Bg4 12.Bg2 Bxf3 13.Bxf3 Nxd4 14.Bxb7 Nc4 15.Qc1 Rb8 16.b3 Rxb7 17.bxc4 c5µ was model play by Black in Meier – Carlsen, Karlsruhe/Baden-Baden 2017.

149

9...Nc6 10.Bf4 10.e3 fails to hold the centre together in view of 10...e5 11.0-0 exd4 12.exd4 Nxd4 13.Nxd4 c5µ as played in Suez Panama – Djuric, Fort de France 2012. Another game continued 10.0-0 Nxd4 11.Nxd4 Qxd4 12.Qxd4 Bxd4 13.a4 (13.Nb5 Be5 14.Bh6 proved ineffective after 14...Rd8 15.Rfd1 Bd7µ in Bancod – Ashwin, Subic Bay 2009) 13...a5 14.Nb5 Be5³ and White’s compensation for the pawn was clearly not enough in Othman – Frhat, Manama 2009. 10...Nxd4! This simple tactical trick offers Black an excellent game. 11.Nxd4 e5 12.Bxe5 After 12.Ndb5 Qxd3 13.exd3 exf4 14.Nxc7 Rb8 15.0-0 fxg3 16.hxg3 Rd8µ Black’s advantage was defined by the two bishops and the weak d3-pawn in Hübner – Topalov, Dortmund 1996. 12...Bxe5 13.Rd1 We have been following Massimini Gerbino – Krueger, corr. 2003. Black has more than one possible improvement, with the choice coming down to taste:

13...c5!?N Angling for a favourable endgame. Black can also keep the queens on with 13...Qe7!? 14.0-0 c6³, with a small but stable advantage due to the two bishops. Later he may be able to convert his queenside pawn majority into a valuable passed

150

pawn. 14.Nf3 Qxd3 15.Rxd3 Bg7³ Black has the better prospects thanks to the bishop pair. C2) 5.h4!?

This aggressive move prepares h4-h5. The ever-creative Alexander Morozevich was the first to play this move at a high level, and it was subsequently taken up by other leading players including Grischuk, Nakamura, Nepomniachtchi, Mamedyarov and even Garry Kasparov himself. Unlike the dubious 4.h4?! line which was dealt with earlier in the chapter, here White does not have to worry about a quick ...c5, since the knight on f3 helps to control the central squares. Obviously advancing the h-pawn at this early stage is not without drawbacks: White weakens his kingside and spends a tempo moving a pawn instead of developing a piece. 5...c6! This has taken over as the exclusive choice of elite Grünfeld players. Black strengthens the d5-point, preventing the f6-knight from being overloaded by the need to control both d5 and h5. Black also threatens to take on c4 and keep his extra pawn with ...b5. 5...c5? turns out badly after 6.dxc5 Qa5 7.cxd5 Nxd5 8.Qxd5 Bxc3† 9.Bd2 Bxd2† 10.Qxd2 Qxd2† 11.Nxd2± and Black was simply a pawn down in Fridman – Kantans, Riga 2014. 5...0-0?! is also not good in view of 6.cxd5 Nxd5 7.h5‚ with an easy attack. Theory is still developing in this variation, but it seems to me that C21) 6.Qb3!?, C22) 6.Bg5!? and C23) 6.cxd5 are the most important options, the last being White’s most popular and reliable 151

continuation. 6.h5? is too optimistic, and 6...Nxh5 7.cxd5 cxd5 8.Rxh5 gxh5 9.Bf4 Nc6 10.Qb3 0-0 11.0-0-0 Bf5µ left White with no real compensation for the sacrificed material in Gattegno – Sos Andreu, San Sebastian 2017. 6.e3 0-0 7.Bd2 was played in Radovanovic – Vuckovic, Kragujevac 2015. A good response would have been:

7...e6N 8.Be2 Nbd7 9.0-0 b6 10.Qc2 Bb7³ We have reached a version of the Schlechter Variation of the Slav Defence, but with White having played the strange h2-h4. Another way to gambit a pawn is: 6.Bf4!? dxc4 7.a4 White is insufficiently prepared for 7.e4?! and Black quickly seizes the advantage: 7...b5 8.a3 Bg4 9.Qd2 Bxf3 10.gxf3 Nh5 (10...Nbd7!?Nµ is a good alternative, leaving White with no compensation for the pawn) 11.Be3 e5 12.dxe5 Qxd2† 13.Kxd2 Bxe5 14.a4 b4 15.Nd1 Now in Antipov – Howell, Gibraltar 2018, Black should have brought the second knight into play: 15...Nd7!N 16.Rc1 0-0-0 17.Bxc4 Bf4µ Black has the more active pieces and securely controls the blockading squares on e5 and f4. 7.e3 Be6 8.Ng5 initiates a typical sequence for such positions, but here it leads to Black’s advantage: 8...Bd5 9.e4 h6 10.exd5 hxg5 11.dxc6 Nxc6 12.Bxg5 In Prestel – Shishkin, Untergrombach 2019, it would have been great for Black to play 12...Qa5!N 13.Bxc4 Ne4 14.Bd2 Nxd2 15.Qxd2 Nxd4 16.Rd1 0-0-0!µ when White’s king is stuck in the centre, and his kingside has been weakened by the earlier h2-h4. The text move was played in Mamedyarov – Grischuk, Beijing 2013. I recommend a novelty:

152

7...c5!N Attacking the centre in typical Grünfeld style. I think the following line is the best that White can do: 8.e3 cxd4 9.Qxd4 Qa5 10.Qxc4 Nc6 11.Qb5 a6 12.Qxa5 Nxa5 13.Be2 Bg4 14.Rd1 Rc8³ Black has the simple plan of plonking his knight on c4. White will not be able to tolerate its presence there, so Black will get the advantage of the bishop pair after its exchange. C21) 6.Qb3!? This move has rarely been encountered in practice but it nevertheless deserves close attention. I suggest gaining time by chasing White’s queen: 6...dxc4 7.Qxc4 b5! 8.Qb3 Maybe someone in the future will test 8.Qd3!?N 8...0-0 9.e4 Na6 10.Be2 c5 11.Nxb5 cxd4 12.Nbxd4 Qb6 13.e5 Ng4 14.Qb5 Qxb5 15.Nxb5 Nb4 when Black will win the pawn on e5, and White will have to think about equalizing. 8...Be6 9.Qc2 b4 10.Na4 Bf5 11.Qc4 This was Rawicz – Czajkowski, Czudec 2016, and here I found a promising pawn sacrifice:

153

11...0-0!N 12.Bf4 White should decline the offer, as 12.Qxb4?! Nbd7 13.e3 e5! 14.dxe5 Rb8 15.Qc4 Ng4ƒ renders his position close to hopeless, as it is hard to suggest anything against the plan of taking on e5 with a knight and then infiltrating to d3. 12...Nbd7 13.e3 Rc8 Black is all set to challenge the enemy pawn centre with ...c5. My analysis continues: 14.Be2 c5 15.dxc5 Qa5 16.Qa6 Qxa6 17.Bxa6 Nxc5! The exchange sacrifice is an important detail to justify Black’s play. 18.Bxc8 Nd3† 19.Ke2 Rxc8 20.Rhd1 Nxf4† 21.exf4

154

21...Bd7! 22.Rxd7 White should return his extra material, as 22.b3?! Nh5 only helps Black. 22...Nxd7 23.Rd1 Rc7÷ The computer evaluates this line with zeros, but in a practical game Black’s chances are somewhat higher due to the fact that White’s structure has been weakened. Black also has a bishop remaining, which makes it easier for him to conduct the endgame with play on both flanks. C22) 6.Bg5!?

White aggressively gambits a pawn. Even though it is not fully sound, it has been tried by some 155

strong players and Black needs to play accurately to prove his superiority. 6...dxc4 7.e4 This is the best attempt to justify White’s play. 7.e3 occurred in Pulvett Marin – Gorny, San Sebastian 2018, when 7...Be6!N 8.Ne5 Nd5µ followed by ...f6 would have made life unpleasant for White. 7.a4 Be6 8.e3 Nbd7 9.Nd2 Nd5 10.Bxc4 h6 11.Bxd5 Bxd5 12.Nxd5 hxg5 13.Nc3 gxh4 14.Qg4 e5µ left Black with a healthy extra pawn in Karthikeyan – Gaikwad, Aurangabad 2018. 7.Bxf6 exf6 8.e3 Be6 9.Nd2 enabled White to restore material equality in Javanbakht – Firouzja, Teheran 2015. However, Black could have obtained a positional advantage by simple means:

9...Nd7N 10.Bxc4 Bxc4 11.Nxc4 0-0 12.h5 f5³ With ideas such as ...Qg5 and ...f4 in the air, White has some problems to solve. 7...Be6! This is stronger than 7...b5, when 8.e5 Nd5 9.h5 was rather double-edged in Nepomniachtchi – Giri, Beijing 2013. It seems to me that best play would be 9...gxh5!?N 10.Rxh5 (10.a4?! Bg4³) 10...Bg4 11.Rh4 h5 12.a4 Nxc3 13.bxc3 Qd5 14.Be2 Nd7÷ when any result is possible. 8.e5 Nd5 9.h5 My improvement is:

156

9...Nxc3!N 9...Nd7 10.h6 Bf8 11.Ne4÷ led to an unclear situation where White was helped by the strong knight on e4 in Kasparov – Nepomniachtchi, Saint Louis (rapid) 2017. That is why we should exchange it off. 10.bxc3 Nd7 11.h6 Bf8 12.Be2 f6 13.Be3 Bd5µ Black will continue unwinding his tangled kingside pieces with ...e6, remaining with a safe extra pawn, while it is hard to suggest a good plan for White. C23) 6.cxd5 cxd5

This way White avoids any speculative sacrifices. 157

7.Bf4 The most logical move. Before playing e2-e3, White takes the bishop outside the pawn barrier. 7.Qb3 Nc6 8.Bf4 0-0 9.e3 transposes to 9.Qb3 in the note to White’s 9th move in the main line. 7.Bg5?! places the bishop on a worse square than in the main line: 7...Nc6 8.e3 (8.Qb3 does not pose the slightest problem. 8...0-0 9.Bxf6 Bxf6 10.e3 occurred in Natalicchio Escalante – Suez-Panama, Gibraltar 2009, and here I recommend the energetic 10...e5N 11.dxe5 Nxe5 12.Rd1 Nxf3† 13.gxf3 Be6³ when White suffers from a broken pawn structure and an unsecured king, while Black has two bishops.) 8...Nh5! Preparing to eliminate White’s dark-squared bishop. 9.Be2 h6 10.Bf4 Nxf4 11.exf4 In Shimanov – Izoria, Burlingame 2017, it would have been worthwhile to secure the kingside with:

11...h5!N 12.Ne5 0-0 The key point is that the erroneously aggressive 13.g4? runs into 13...Qb6! 14.gxh5 Qxb2 15.Rc1 Bf5 16.hxg6 Nxe5 17.dxe5 Rac8 18.gxf7† Rxf7 19.Qd2 Qxd2† 20.Kxd2 Bh6 21.Nxd5 Rd8–+ when White is left without a piece. 7...Nc6 8.e3 0-0

158

9.Ne5 Since the position is closed, White can justify moving an already-developed piece. The idea of the move is to prevent ...Bg4 and threaten to compromise Black’s structure by exchanging on c6. 9.Rc1 Bg4 10.Qb3?! (White should settle for 10.Be2, although after 10...e6 he has no trace of an advantage) 10...Bxf3 11.gxf3 occurred in Rozman – Haessel, Edmonton 2018, when the natural 11...Na5N 12.Qa3 Rc8 13.Be5 Nc6 14.Be2 e6 15.Bd6 Re8 16.Kf1 Qd7³ would have given Black slightly better chances due to his superior pawn structure. White can play for safety with: 9.Be2 Bg4 10.0-0 10.Ne5?! Bxe2 11.Qxe2 Rc8 12.Qb5 Nh5 13.Bh2 e6³ turned out well for Black in Khismatullin – Ponkratov, Khanty-Mansiysk 2014, since White’s position has been weakened by the h2-h4 move. 10...e6 11.Rc1 11.Bh2 occurred in Nesterov – Moiseenko, Batumi 2018, when 11...Ne7!N 12.Rc1 Nf5³ once again gives White problems with the h4-pawn. 11.Nd2 Bxe2 12.Nxe2 Nh5³ was a similar story in Maksimenko – Kolosowski, Gorzow Wielkopolski 2014. 11...Rc8 12.Qb3 Qe7= Melkumyan – Erdos, Germany 2013. Objectively, the position is equal; however, in a practical game I would choose Black’s side, as the pawn on h4 looks out of place. Another popular continuation is: 9.Qb3 Na5 10.Qa3 Some strong players have tried: 10.Qb4 Nc6 11.Qa3 (11.Qb3 Na5= repeats the position) 11...Bg4 12.Ne5 (or 12.Be2 Rc8 13.0-0 Ne4 14.Rfc1 e6³ when the weakness of the h4-pawn gives White a lot 159

of trouble) 12...Nxe5 13.Bxe5 Ne4 14.Bxg7 Kxg7 15.Bd3 Nxc3 16.bxc3 This was Al Sayed – Muhammad, Manila 2013, and now 16...Rc8N³ gives Black the more comfortable game thanks to his better pawn structure. 10...Bf5 11.Be2 11.Rc1!? deserves attention, although it has only been played in one game, Diermair – Hill, Lisbon 2018. I recommend 11...Rc8N 12.Be2 Nc4 13.Bxc4 dxc4 14.Qxa7 Nd5 15.0-0 Nxf4 16.exf4 e6© when Black has good compensation thanks to the two bishops, with the light-squared bishop being especially strong. 11...Ne4 12.Nxe4 The weaker 12.Rc1?! was played in Kvetny – Kreisl, Austria 2017, when 12...Nc4!N 13.Bxc4 dxc4 14.Nxe4 Bxe4 15.Rxc4 Bxf3 16.gxf3 Qd5ƒ would have given White problems due to his broken pawn structure. 12...Bxe4 13.0-0 Perhaps White should have kept the knight away from c4 by means of: 13.b3N 13...Nc6 14.0-0 e6 15.Ng5 Bf5 16.Bd6 (16.g4?! is well met by 16...e5 17.dxe5 Bc8 18.Rad1 h6 19.Nf3 Bxg4³ when the white king does not feel safe) 16...f6„ Black is not worse, as he has fine counter-chances on the kingside. 13...Nc4 14.Qc3 I also checked 14.Qb4N 14...Qb6 15.Qxe7 Bf6 16.Qd7 Qxb2 17.Rfe1 Rad8÷ when Black is certainly not worse, as his pieces are active. 14...Rc8 15.b3 Bxf3 16.Bxf3 Ne5 17.Qb4 In Al Sayed – Sarana, St Petersburg 2018, it would have been strong to play:

17...Nd3!N 18.Qxb7 Nxf4 19.exf4 e6 20.Qxa7 Qd6 21.g3 Ra8 22.Qc5 Qxc5 23.dxc5 Bxa1 24.Rxa1 Rfb8µ With a big advantage for Black.

160

9...Nxe5! Black offers a temporary pawn sacrifice. 10.Bxe5 This seems safest. 10.dxe5 is playable, though after 10...Ne4 White should decline the sacrifice: 11.Nxe4 (11.Qxd5?! looks too risky in view of 11...Nxc3 12.Qxd8 Rxd8 13.bxc3 Be6 14.Be2 Rac8³ when Black regains the pawn and keeps the upper hand due to his better pawn structure) 11...dxe4 12.Bc4 Qc7 13.Rc1 Qa5† 14.Qd2 Qxd2† 15.Kxd2 Bg4 The chances are about equal. 10...Ne4! Another strong move: Black should benefit from exchanging pieces. 11.Bxg7

11...Kxg7 Black is not afraid of doubled pawns, since the pawn on e4 will capture additional space and control some important squares. 11...Nxc3!? is a good alternative. 12.bxc3 Kxg7 13.Be2 was played in Ftacnik – Mista, Germany 2016, when 13...h5!N 14.0-0 Qd6 15.Rc1 Bd7 16.Qb3 Rab8 17.c4 dxc4 18.Qxc4 Rfc8ƒ would have given Black slightly the more pleasant position. In the future he can aim for ...b5 and ...a5, eventually creating a passed pawn. 12.Bd3 12.Nxe4 dxe4 13.Bc4 Bd7 14.0-0 Qb6 gives Black comfortable equality. 161

White could also try 12.h5!?N, when 12...Nxc3 13.bxc3 Qd6÷ reaches a complex position in which Black’s chances are not worse. 12...Nxc3 13.bxc3

13...h5! Depriving White of the opportunity for h4-h5. 14.Qb3 b6 15.0-0 Bg4!? This provocative move is not the only decent option, but it was vindicated by the game continuation. Black aims to provoke f2-f3 with a view to weakening White’s kingside. Black had some good alternatives. One example is 15...Rb8N 16.c4 Bb7 17.cxd5 Bxd5 18.Bc4 Bxc4 19.Qxc4 b5 mobilizing the queenside pawn majority, although the evaluation has not yet gone beyond equilibrium. 15...Qd6!?N is another option, leading to a complex strategic game. 16.f3?! White should have avoided weakening his dark squares, in which case the position would have remained about equal. 16...Be6 17.Qb2?! Another inaccuracy: White plays too passively.

162

It would have been better to try 17.Rfc1 Rc8 18.a4 (but not 18.c4?! dxc4 19.Bxc4 Bxc4 20.Rxc4 Rxc4 21.Qxc4 e6 22.g3 Qb8 23.Kg2 Rc8³ when Black captures the open file, with a small but steady advantage) 18...Bf5 19.Ba6 Rc7 20.Qb4 e6 21.g3 f6ƒ when the possibility of ...g5 still gives White something to think about. 17...Rc8 18.Rac1 Rc7

19.e4! Playing actively in the centre is White’s best chance. 19.Qb4 Bc8 20.a4 e6 21.g3 Bb7ƒ would be unpleasant for White, as he has many weak squares and his kingside looks shaky. 19...dxe4 20.fxe4 Bc4 21.Rcd1 21.Bxc4 Rxc4 22.e5 b5! is excellent for Black, since 23.Qxb5 Rxc3! 24.Rxc3 Qxd4†µ leaves him with a healthy extra pawn. Now Black’s play can be improved with:

163

21...Bxd3!N In the game Black preferred 21...e5 with the same general ideas. However, after 22.Bxc4 Rxc4 23.Qb5 Qc7, in El Gindy – Grischuk, Tbilisi 2017, White missed a chance to simplify with 24.Qxe5†! Qxe5 25.dxe5 Rxe4 26.Rd7 Rxe5 27.Rxa7 Re4 28.g3 Re3 29.Kg2 Rxc3 30.Rf2³ when he should be able to hold the pawn-down endgame without too much trouble. 22.Rxd3 e5 23.Qf2 Qe7³ In this heavy-pieces ending, Black’s chances are higher due to White’s numerous pawn weaknesses, namely h4, c3 and a2. Conclusion White’s weird 4th moves are justifiably rare in practice, since they generally entail a lot of risk on White’s side. For example, 4.g4? sees White trying to start a clearly unprepared attack, resulting in the loss of a pawn after 4...dxc4, while 4.f3?! and 4.h4?! both run into 4...c5! when Black obtains excellent central play. If White wishes to try something offbeat, he should at least start by developing a piece and supporting his centre with 4.Nf3 Bg7 before deciding what to do. 5.g3 is not a bad move, although Black obtains a good game with 5...dxc4 and he could even transpose to the previous chapter with 5...c6, so it is not clear what White hopes to achieve with this move order. 5.h4!? is the most important branch of the chapter. 5...c6! is the best reply, when White has to choose a plan. The most aggressive option is to sacrifice a pawn on c4, but such plans carry a lot of risk and Black should obtain the advantage with correct play. 6.cxd5 cxd5 is the main line, leading to an Exchange Slav structure where Black’s chances are not worse, and we saw many lines in which White 164

could be made to suffer for weakening his kingside with h2-h4.

165

A) 3.Nf3 Bg7 4.e3 91 B) 3.Nc3 d5 4.e3 Bg7 96 B1) 5.Qb3 97 B2) 5.Nf3 0-0 101 B21) 6.b4 103 B22) 6.cxd5 104 B23) 6.Be2 c5! 110 B231) 7.dxc5 111 B232) 7.0-0 112

1.d4 Nf6 2.c4 g6 This chapter will deal with systems involving an early e2-e3. We will consider A) 3.Nf3 Bg7 4.e3 and B) 3.Nc3 d5 4.e3 separately, as the game will take on a completely different character according to which knight White develops first. 3.e3 Bg7 does not change much, as White will almost certainly play one of the two knight moves in the 166

near future. A) 3.Nf3 Bg7 4.e3

4...0-0 For the moment, we continue to play flexibly. 5.Be2 White proceeds with a very specific move order, which is designed to delay the development of the b1-knight. GM Axel Smith recommends this system in his 2017 book e3 Poison. The logic of White’s approach is that 5...d5 can be met by 6.cxd5 Nd5 7.e4 when there is no ...Nxc3, and so the knight must retreat to the suboptimal b6-square. White has good chances for an opening advantage in this line, despite losing a tempo with e3-e4. This anti-Grünfeld continuation was played by the Montenegrin player Kosic versus Ivanisevic in 2005 but it was seemingly forgotten until 2008, when Shengelia applied it. At the elite level, Grischuk and Ivanchuk have tried it. Since White’s set-up is rather new, fresh positions can arise quite early in the opening, which is good for those who are striving for a creative, independent game. 5...b6!? This continuation has been chosen by such Grünfeld specialists as Nepomniachtchi, Vachier-Lagrave and Bok. Black will fianchetto the bishop and follow up with ...e6 and ...c5. 5...c5 is a good alternative, when White must either allow ...cxd4, with a likely transposition to the reversed Tarrasch as covered later in variation B232, or play 6.d5, reaching a version of the Benoni where the early e2-e3 is hardly White’s most threatening set-up. 167

6.0-0 Bb7 7.Nc3 By now, White is usually ready to develop his knight. I checked two other possibilities: 7.d5 was an attempt to block the b7-bishop in Tikkanen – Howell, Tromso (ol) 2014. I recommend improving with 7...Na6N 8.Nc3 Nc5 9.Qc2 e6„ when Black has started the fight against d5, and may consider ...c6 in the near future. The bishop on b7 is not at all bad, as it adds to the pressure on the d5point. 7.b4!? is unusual, but the plan of capturing space on the queenside deserves attention. 7...Ne4 8.Bb2 a5 9.b5 d6 10.Qc2 (I also checked 10.Nbd2N 10...Nxd2 11.Nxd2 e5 12.Bf3 Bxf3 13.Qxf3 Nd7 14.Rfd1 f5 with an approximately equal position) 10...Nd7 11.Nbd2 f5 12.a4 e6 13.Nb3 was seen in Hayrapetyan – Davtyan, Yerevan 2018, and here I recommend:

13...Qe7N 14.Nfd2 Nxd2 15.Qxd2 (15.Nxd2 is safer although it would be an admission that the earlier Nb3 move did not achieve its goal) 15...Qg5 16.f3 Bh6³ The weak e3-pawn is inconvenient to defend.

168

7...e6!? Black does not yet define his plans. He may yet play ...d5 or choose a more flexible plan involving ...c5. Smith does not considering this continuation in e3 Poison. 8.b3 It is logical to develop the bishop on b2. Again, the attempt to block the b7-bishop with 8.d5 is not dangerous: 8...Re8 9.e4 exd5 10.exd5 Na6 11.Bg5 h6 12.Be3 occurred in Danielian – Jumabayev, Yerevan 2018, and here I recommend putting pressure d5 by means of:

12...c6!N 13.Qd2 Kh7 14.Bd4 Nc5÷ Black targets the d5- and e4-points, with good play. 169

8.Qc2 c5 9.Rd1 Nc6 10.a3 cxd4 11.exd4 was seen in Manuri – Baltgailis, Quebec 1993, when it would have been logical to play on the c-file against the enemy queen:

11...Rc8N A nice possibility is 12.Bf4 Na5 13.Bd6 Nxc4! 14.Bxf8 Qxf8 15.Bxc4 Rxc4© when Black’s extra pawn and strong bishops fully compensate for the sacrificed exchange. Finally, I checked: 8.b4 The seizure of space gives Black the chance for a powerful counterstrike with ...c5. 8...d6 9.Qc2 9.a4 c5! 10.b5 Nbd7 11.Qc2?! occurred in Klarner – Moesche, Bavaria 1998. Black already had a pleasant position, and White’s last move makes it even more tempting to play along the c-file: 11...Rc8N 12.Ba3 Re8 13.dxc5 Nxc5³ The c4-pawn is a long-term weakness, whereas White is unable to mount a meaningful attack against d6. 9...Nbd7 9...c5!? is also interesting. 10.Rd1 a5 11.b5 Qe7 12.a4

170

12...e5 For fans of a more complex game, I can recommend 12...Rae8!?N 13.Bb2 Nh5 14.Rac1 f5„ with interesting prospects on the kingside. 13.dxe5 Nxe5 14.Bb2 Nxf3† 15.Bxf3 Bxf3 16.gxf3 Rae8 In Paravyan – Alekseenko, Khanty-Mansiysk 2018, Black was comfortably equal, with decent chances to take over the initiative in the future. 8...c5! 8...d5 has been employed by some strong players but I like the flexibility of the text move. 9.Bb2 9.Ba3 is playable but it seems a clumsier way to develop the bishop. 9...d6 10.Qc2 occurred in Pham – Danielian, Riyadh 2017, when 10...Na6!N would have been logical. Play may continue 11.Rac1 Qe7 12.Rfd1 Nb4 13.Qb1 Rfd8÷ when Black’s chances are not worse, and the purpose of the bishop on a3 remains a mystery. 9...Qe7 Again we choose a flexible continuation, delaying the decision on whether to advance the d-pawn one or two squares. 10.dxc5 White opens space for his bishop along the a1-h8 diagonal and will not have to reckon with the possibility of ...cxd4 in the future. 10.d5?! has not been tested, for good reason: after 10...Ne4 11.Rc1 Na6 12.a3 Rae8 13.dxe6 dxe6

171

14.Qc2 Nxc3 15.Bxc3 e5³ Black has a pleasant position and can expand with ...f5 and ...e4. 10.Re1 was tried in Granda Zuniga – Iturrizaga Bonelli, Mexico City 2016. Out of many playable continuations, I would suggest 10...d5!?N 11.dxc5 bxc5 12.cxd5 exd5 13.Rc1 Nbd7 14.Qc2 Rac8„ with a dynamic game, typical for the hanging-pawns structure. 10.Qc2 was played in Moeschinger – Klauser, Lenk 1989. I think this would have been a good moment to release the tension in the centre: 10...cxd4N 11.exd4 My idea is 11...Na6 12.a3 d5 13.cxd5 when Black has a powerful pawn sacrifice:

13...Nc7! 14.dxe6 Nxe6 15.Rfe1 Rad8 16.Rac1 Rfe8© Black has excellent compensation. The d4-pawn is weak, the bishop on b7 is strong and at any convenient time you can play ...Bxf3 and ...Nxd4. Another example continued 10.Rc1!? Rd8 11.Qc2 Nc6 12.dxc5 bxc5 13.Rfd1 and now in Schuette – Laqua, Germany 2017, it would have been logical to play: 13...Rab8N A plausible continuation is 14.Na4 Ne8 15.Bxg7 Kxg7 16.Qb2† Kg8 17.Ne1 d6÷ when the d6-pawn is easily defensible, while Black has a central majority and a lot of flexibility with his structure. Finally, 10.Qd2 was played in Meduna – Stocek, Ostrava 2016. Here I like 10...d5!?N 11.dxc5 bxc5 12.cxd5 exd5 when Black plays with hanging pawns, in a position full of dynamics. Play might continue: 13.Rfd1 Nbd7 14.Rac1 Bh6 15.Ba3 Rac8„ The c5- and d5-pawns are under tension, but Black will always have counterplay in such a dynamic position. 10...bxc5 11.Qd2 White gets ready to centralize the rooks. 11.Qc2 Rd8 12.Rfd1 was seen in Tomaszewski – Gdanski, Gdynia 1987. With the queen on c2 rather 172

than d2, I would avoid ...d5 in favour of the flexible 12...d6N 13.Ne1 Nc6 14.a3 Rab8÷ when Black is fine. 11.Nb5!? d5 12.Be5 was an interesting idea in Ovezdurdiyeva – Saduakassova, Tashkent 2017. I recommend a dynamic response:

12...Rd8!N 13.Nc7?! The critical move, but a mistake. (It is better to play 13.cxd5!? Bxd5 14.Qc1 Nc6 15.Bb2 Rac8 16.Rd1 Nb4÷ when Black has plenty of activity to compensate for the isolated c-pawn.) 13...Nc6 14.Nxa8 Nxe5 15.Nxe5 Ne4 16.f4 dxc4 17.Qe1 Nd2 18.Rf2 c3³ The knight on a8 will be destroyed at a convenient moment, and the c3-pawn is a serious threat to White. 11...Rd8 There is also nothing wrong with 11...d5N, which transposes to the earlier note on 10.Qd2. 12.Rfd1 d5 Another reasonable option is to adopt a flexible pawn structure with 12...d6!? 13.Rac1 Nc6 14.a3 Rab8÷ when Black has a comfortable position: the d6-pawn is well defended and the b3-pawn may become the object of attack. The text move offers Black dynamic play with hanging pawns. We will follow an instructive example for a few more moves.

173

13.cxd5?! This slight inaccuracy opens the e-file, making ...d4 more of a threat in some lines. 13.Qc2N looks more reliable, when 13...Nc6 14.Na4 Nb4 15.Qc1 dxc4 16.Bxc4 Rxd1† 17.Qxd1 Nbd5 18.Qc2 Nb6 reaches an approximately equal position. 13...exd5 14.Rac1 Bh6! The bishop exerts unpleasant pressure and threatens ...d4, taking advantage of the unfortunate position of White’s pieces on d2 and c1. 15.Na4 Ne4 16.Qc2 Another small inaccuracy: the queen strays further from the kingside, leaving the f2- and e3-points even weaker. 16.Qe1N was more resilient although 16...Nd7 17.Bd3 Rac8 18.Rc2 Qe6 19.h3 Nef6 offers Black excellent counterplay, with ideas such as 20.Qa5 d4! in the air, striking at the kingside. 16...Nd7 17.b4? This is a typical idea when playing against hanging pawns, but it’s a mistake. 17.Bd3 is better although it runs into an unpleasant sacrifice: 17...Nxf2! 18.Kxf2 Bxe3† 19.Kf1 d4 20.Qe2

174

20...Rac8! 21.b4 Bxf3 22.gxf3 Qg5 23.Qg2 Bxc1 24.Rxc1 Qe3ƒ Black has a nasty initiative. 17...d4! Increasing the power of both bishops. 18.bxc5?! 18.Nxc5! was objectively a better try, although 18...Ndxc5 19.bxc5 dxe3 20.c6 exf2† 21.Kf1 Bxc1 22.cxb7 Rxd1† 23.Qxd1 Re8 24.b8=Q Rxb8 25.Qxc1 Qb4 26.Qc2 Qxb2 27.Qxe4 Qxa2µ leaves Black with excellent winning chances. We have been following Taimanov – Psakhis, Moscow 1981, in which Black played 18...dxe3 and went on to win a good game, although at one point White could have made the outcome unclear. At this point, a much more powerful continuation would have been:

175

18...Nxf2!N 19.Kxf2 Bxe3† 20.Kf1 Ne5! 21.c6 Nxf3 22.Bxf3 Ba6† 23.Rd3 Re8–+ Black has a decisive attack. B) 3.Nc3 d5 4.e3

This modest-looking continuation was played in the distant past by such giants as Euwe, Alekhine, Flohr, Petrosian, and subsequently by Kramnik, Ivanchuk, Nakamura, Mamedyarov, Andreikin and others in the modern era. Usually, White plays this way in order to avoid long, forcing theoretical lines. The game usually acquires a calm, positional character, with White’s plans including moves like Nf3, Be2, 0-0, b2-b3 (or b2-b4) and Bb2. In chess clubs, such a construction is often called the turtle system.

176

4...Bg7 We will focus on two main options: B1) 5.Qb3 followed by B2) 5.Nf3. 5.b4 0-0 6.Nf3 transposes to variation B21. White can play on the queenside with: 5.Bd2 0-0 6.Rc1 6.Nf3 transposes to the 6.Bd2 line in the notes on page 102. 6...c5 7.dxc5 Na6 8.Nxd5?! 8.cxd5 is better although 8...Nxc5 is fine for Black. Trying to hang on to the d5-pawn only leads to trouble for White: 9.Bc4?! Bf5 10.Nge2 Bd3 11.Bb3 was seen in Cusi – Yermolinsky, Chicago 2003, when 11...Qb6!N 12.0-0 Rfd8 13.Re1 Nxb3 14.Qxb3 Qxb3 15.axb3 Nxd5³ would have led to an advantageous endgame for Black, thanks to his bishop pair and White’s doubled b-pawns. 8...Ne4 9.f3 Nxd2 10.Qxd2 Nxc5 The immediate 10...e6 has been played a few times but the text move is stronger. 11.Rd1 e6 12.Nc3 Bd7 13.Nge2 Qh4† 14.g3 Qxc4 15.Nd4 Qb4ƒ In Cirulis – Kappes, corr. 2010, Black had the two bishops and White suffered from some kingside weaknesses. A somewhat serious option is: 5.cxd5 Nxd5 6.Nxd5 6.Qb3 Nxc3 7.bxc3 0-0 8.Nf3 c5 transposes to the 7.Qb3 line in the notes to variation B22 on page 105. Another possibility is 6.Bc4 Nxc3 7.bxc3 c5 8.Ne2 (8.Nf3 again transposes to variation B22) 8...0-0 9.0-0 Qc7 10.Bb3 (I also checked 10.d5N 10...Nd7 11.e4 Ne5 12.Bf4 a6 13.a4 Bd7 with a roughly equal position) as played in Stelter – Sailer, Bavaria 2004. Here I recommend an improvement: 10...b6!N 11.e4 Nc6 12.Bf4 e5 13.Bg3 Ba6 14.dxe5 Rad8 15.Bd5 Nxe5 16.Qc2 c4„ Black intends to move his queen to escape the pin, followed by plonking his knight on d3; and if it is exchanged, then Black will enjoy the two bishops. 6...Qxd5 7.Ne2 The knight comes to c3 next. This has been tried by some strong players, and has occasionally been recommended as a convenient low-theory option against the Grünfeld. I like the following reply. 7...0-0 8.Nc3 Qd6 9.Be2 c5! 10.d5 10.Ne4N is not scary in view of 10...Qc6! 11.d5 Qb6 12.0-0 e6 13.Nc3 exd5 14.Nxd5 Qd8 15.e4 Nc6 16.Be3 b6 17.Qd2 Nd4÷ and Black’s chances are not worse. 10...e6 11.e4 This occurred in Sibilio – Safarli, Rogaska Slatina 2011, and now I recommend:

177

11...exd5N 12.exd5 Re8 13.0-0 Nd7 The passed d-pawn has the potential to be strong, but also a weakness. My illustrative line continues: 14.Be3 a6 15.a4 b6 16.Qd2 Be5 17.g3 Nf6 18.Rfe1 Bh3 19.Rad1 h5 20.Bf3 Bd7÷ The d-pawn is securely blocked, and Black can think about seizing the initiative with ...b5. B1) 5.Qb3 This move is infrequently met but is nevertheless quite logical, as the queen puts pressure on d5 and b7. Korchnoi played this way several times, and several other strong GMs have experimented with it. One of the advantages of this move is that White avoids heavy theory and often succeeds in reaching a position where the players have to think for themselves at an early stage. It also has some drawbacks of course: White commits the queen rather early and may have to waste time moving her again in the future. 5...e6 5...dxc4 invites White to develop his bishop with tempo and 5...c6 leads to a version of the Slav, so the text move is my preference. 6.Qa3 Preventing Black from castling short. White causes us no problems with: 6.Nf3 0-0 7.Bd2 Another game continued 7.Be2 b6 8.0-0 Bb7 9.Rd1 Nbd7 10.Bd2 dxc4 11.Bxc4 c5 12.dxc5 Nxc5 13.Qa3 and now in Karsa – P. Popovic, Biel 1981, the logical 13...Bxf3N 14.gxf3 Nfd7 15.f4 Qh4

178

16.Bf1 e5³ would have given Black good chances to develop an attack on the enemy monarch, due to its weakened pawn cover. 7...b6 GM Yusupov reached this position several times with White. It seems to me to be somewhat strange, since Black has no problems and may even be a little better. 8.Rc1 8.cxd5 exd5 is likely to transpose. For example, 9.Be2 Bb7 10.0-0 Nbd7 11.Rfd1 Re8 12.Rac1 c6 was the move order of the Burkart – Gustafsson game noted below. 8...Bb7 9.cxd5 exd5 10.Be2 Nbd7 11.0-0 c6 12.Rfd1 Re8 13.a4 13.Be1 Bf8 14.a3 Bd6 15.Qc2 Qe7 16.Rb1 Rac8³ was seen in Burkart – Gustafsson, Germany 2018, with Black’s chances being somewhat higher due to the more active arrangement of his pieces. 13...a5 14.Ne1 Nf8 15.Nd3 Ne6 16.Bf1 This position was reached in Yusupov – Kasimdzhanov, Germany 2014, and here I propose an interesting manoeuvre:

16...Ng4!N 17.h3 Nh6 18.Ne2 Nf5³ Black’s chances are somewhat higher as his pieces are more harmonious.

179

6...a5! Black deals with the situation on the a3-f8 diagonal by preparing the ...Na6-b4 manoeuvre, thereby putting White’s queen in an unfortunate position. 7.Nf3 7.Bd2 was played in Rodshtein – Walach, Katowice 2017, when 7...dxc4!N 8.Bxc4 c6! would have been an effective way to gain space on the queenside. For instance, 9.Qb3 b5 10.Be2 Nbd7 11.Nf3 0-0 12.0-0 Re8 13.Qc2 e5³ and Black’s chances are somewhat higher, since he can develop his play in the centre or on the queenside thanks to the advanced pawns. 7...c6 Black could also consider 7...dxc4!?N 8.Bxc4 c6 with similar ideas as in the previous note. White should be able to do better than transposing to it with an early Bd2, but Black’s scheme is still playable. 8.b3 White safeguards the c4-pawn and prepares to return his queen to ‘base camp’ with Qb2. At the same time, it is clear that the whole idea of White’s 5th move did not yield any benefits, and his queen manoeuvres have only resulted in a loss of time. One strong GM has made the serious mistake of 8.Bd2?, allowing 8...dxc4 when Black simply wins a pawn, as recapturing would cost White a piece. 9.Qa4 b5 10.Qc2 occurred in Zubov – Krasenkow, Warsaw 2009, when the most accurate continuation would have been:

180

10...Nbd7!N (the game continuation of 10...a4 gave Black a clear advantage, but there is no need to spend a tempo blocking the queenside) 11.a4 Ba6 12.axb5 cxb5 13.e4 b4 14.Nd1 Qc7–+ White has no compensation for the pawn. 8.c5?! avoids the loss of a pawn but gives Black easy play. 8...0-0 9.Be2 Nbd7 10.0-0 Re8 11.b4 b6 12.Rb1 Ba6 13.Bxa6 Rxa6 14.b5 cxb5 15.Rxb5 bxc5 16.dxc5 In Finegold – Troff, Saint Louis 2012, Black missed the opportunity to win a pawn outright with:

16...Qc7!N 17.Bd2 a4 18.Rfb1 Bf8µ The c5-pawn is doomed. 8.Qb3 looks too sophisticated. 8...0-0 9.Be2 was seen in Arnaudov – Givon, Albena 2013, when Black could have gone for the familiar idea of capturing space on the queenside: 9...dxc4N 10.Bxc4 b5 11.Be2 Nbd7 12.0-0 Qe7 13.Qc2 e5³ Black has achieved the optimal arrangement of his pieces and 181

has active prospects both in the centre and on the queenside. 8.cxd5 exd5 9.Bd3 Na6 10.0-0 Nb4 11.Bb1 0-0 12.Bd2 Re8 13.Ne5 occurred in Zajogin – Weber, Germany 2014, when Black missed the strongest plan:

13...c5!N 14.Qa4 Rb8 15.f4 Bd7 16.Qd1 Nc6 17.Ne2 Qb6µ The pressure against d4 and b2 will cost White a pawn. Finally, 8.Ne5 was tried in Ovsepyan – Schlichtmann, Bavaria 2016, when the natural and strongest reply would have been: 8...Nbd7!N 9.f4 c5! 10.Be2 (10.dxc5? Nxe5 11.fxe5 Ng4 gives Black a big advantage) 10...dxc4 11.Nxc4 0-0 12.dxc5 Nd5! 13.Nxd5 exd5 14.Nd6 b6 15.cxb6 Qxb6 16.Nxc8 Raxc8 17.0-0 Rfe8 18.Qd3 a4© It is difficult for White to deal with Black’s unpleasant initiative and develop his queenside without a material concession. It is worth noting that the greedy 19.Qxd5? leads to disaster after 19...a3 20.Qxd7 axb2 21.Bxb2 Bxb2 22.Rad1 Qxe3† 23.Rf2 Ba3 24.Qd3 Bc5–+ when White suffers decisive material losses. 8...Na6 Black continues with his plan. However, I want to emphasize that we will not be in a rush to put the knight on b4, but in most cases will go for some combination of ...Ne4, ...b6 and ...c5, acting in the centre. 9.Be2 White also finishes the development of the kingside. Other continuations do not bring any benefits: 9.Bd2 Ne4 10.Rd1 b6 11.Be2 occurred in Bitelmajer – Krysa, Vicente Lopez 2016, when it was possible to press on the centre with 11...c5!N 12.0-0 0-0 13.Be1 Bb7 14.cxd5 exd5 15.Nxe4 dxe4 16.dxc5 Qe7³. The e4-pawn squeezes White’s position, while the weakness of the d3-square and the 182

unfortunate placement of the bishop on e1 also count in Black’s favour. It seems strange to me that Korchnoi went for these positions. In one of the games he chose 9.c5 but soon became worse after 9...Nd7 10.Bxa6 Rxa6 11.Na4 0-0 (11...Qe7!?) 12.Bb2 Qe7 13.0-0 Re8 14.Rac1 in Korchnoi – Kurnosov, Chelyabinsk 2007. I like the following straightforward plan:

14...f6!N 15.Rfe1 e5µ Black has the advantage of two bishops, as well as excellent prospects for developing an initiative on the kingside, while White’s play has reached a dead end.

9...Ne4! An excellent move. It is not profitable for White to trade off the knight, as the e4-pawn will squeeze him and the clearing of the long diagonal will give Black the opportunity for ...c5, with powerful pressure against d4. 183

10.Qb2 10.Nxe4?! dxe4 turns out badly for White after 11.Nd2 Nb4 12.Qb2 c5 13.a3 Nc6 14.Nxe4 cxd4µ when he has big problems on the long diagonal. 10.Na4 Nb4 11.Qb2 c5 12.0-0 0-0 13.Rd1 occurred in Korchnoi – Krasenkow, Lvov 2000, when 13...Bd7!N 14.Nc3 a4! 15.Bd2 a3 16.Qc1 Nxd2 17.Qxd2 Qa5µ would have been unpleasant for White due to Black’s bishop pair and pressure in the centre. 10...0-0 11.0-0 b6 Black is ready to complete development with ...Bb7 and ...c5, while White has difficulties with the bishop on c1. 12.Rd1 White tried 12.a3 Bb7 13.Qc2 Nxc3 14.Qxc3 c5 15.Bb2 cxd4 16.exd4 in Ringoir – Molner, Internet 2014, when a strong manoeuvre would have been:

16...Nc5!N 17.a4 Ne4 18.Qd3 Re8 19.Rfd1 Ba6³ The black pieces are more active than their counterparts, and the d4- and c4-pawns should be excellent targets to attack. 12...Bb7 13.Bd2 White has to develop his queenside somehow. 13...Nxd2! A timely exchange. Black seems to be exchanging a powerful knight for a not-so-active bishop; but if he neglects to do so, the bishop will drop back to e1 and have chances to play a more active role in the 184

long term. Therefore it’s best to make the trade now and enjoy the advantage of two bishops. 14.Qxd2

So far Black had done everything right in Nakamura – Caruana, Thessaloniki 2013, but here his play could be improved. 14...Qc7!N Caruana preferred to reroute his knight with 14...Nb8, which is not a bad decision, and after 15.Rac1?! Nd7 he went on to build a decisive advantage before squandering the win in an endgame. However, Nakamura could have improved with 15.Ne5!N when it is not so easy to exploit Black’s bishop pair. Best play continues: 15...f6! (15...Nd7 16.Nxd7 Qxd7 17.Rac1 Rfd8 18.Na4 Qc7 19.c5 b5 20.Nb6 Ra7 21.f4ƒ gives White the easier game) 16.Nd3 Qe7 17.Na4 Nd7 18.c5 b5 19.Nb6 Ra7 20.f4 Qd8 21.Nxd7 Qxd7= Black’s bishops are unable to showcase their potential in such a blocked position, although he can still obtain adequate play by preparing ...e5. 15.Rac1 Rad8 16.Qb2 dxc4 17.bxc4 c5 It is worth giving White an outpost on b5 if it means improving the scope of both bishops. 18.Rb1 Bxf3 19.Nb5 Qb8 20.Bxf3

185

20...e5! 21.dxe5 Bxe5 22.Qc2 Bxh2† 23.Kh1 Be5 White certainly has some compensation and it will not be at all easy to convert Black’s extra pawn; nevertheless, it’s better to be a pawn up than a pawn down. B2) 5.Nf3

5...0-0 It makes sense to castle and keep the other pieces flexible until White reveals what he is doing. His main options are B21) 6.b4, B22) 6.cxd5 and B23) 6.Be2.

186

6.Qb3 e6 transposes to 6.Nf3 0-0 in the notes to variation B1 above. 6.h3 is sometimes played but after 6...c5 White hardly has anything better than 7.Be2, when 7...cxd4 8.exd4 Nc6 9.0-0 transposes to variation B232. 6.Bd3 is similar to variation B23 and I see no advantage of having the bishop on d3 instead of e2, so Black can follow the same recipe, the likely result being a transposition after 6...c5 7.0-0 cxd4 8.exd4 Nc6 (8...Bg4!? is a reasonable alternative with White’s bishop on a suboptimal square) 9.h3 dxc4 10.Bxc4. 6.b3?! is asking for trouble: 6...c5 7.Bb2 cxd4 8.Nxd4 e5 9.Nf3 d4! 10.exd4 exd4 11.Qxd4 (11.Nxd4 Qb6 12.Na4 Qa5† 13.Qd2 Re8† 14.Be2 Qxd2† 15.Kxd2 Ne4†ƒ also gives Black a dangerous initiative for the sacrificed pawn) 11...Qa5 12.0-0-0 Ng4 13.b4 Qxb4 14.Qxg7† Kxg7 15.Nd5† In Lauko – Janecek, Tabor 2010, Black missed the simple refutation of White’s play: 15...Qxb2†N 16.Kxb2 Nxf2 17.Rd2 Nxh1 18.Nc7 Na6 19.Nxa8 Bg4–+ The knight on a8 dies the death of the brave; and even if White could pick up the knight on h1 just as easily, Black would still have an extra pawn. A final alternative is: 6.Bd2 This is not such a bad move and it has been employed by some strong GMs, but it doesn’t cause us any problems. 6...e6!? 6...c5 is perfectly playable but I rather like the modified Queen’s Gambit Declined with the bishop on g7. 7.Rc1 7.Be2 dxc4 8.Bxc4 Nbd7 9.b4 occurred in Fridman – Bok, Saint Louis 2018, when 9...Qe7!?N seems to me like a good novelty. Play may continue 10.Bb3 e5 11.0-0 e4 12.Ng5 h6 13.Nh3 g5ƒ with a complex game, where the h3-knight is out of play and I like Black’s chances on the kingside. 7...Nbd7 8.cxd5 8.Bd3 dxc4 (8...c5!?N 9.0-0 b6 10.cxd5 exd5÷ is a valid alternative) 9.Bxc4 c5 10.0-0 b6 11.Qe2 a6N (the game actually continued 11...Bb7 but I would rather not allow 12.Ba6, even though Black is not worse here either) 12.a4 Bb7 13.Rfd1 Qe7 14.Be1 Rfd8 15.b3 (I think White should settle for equality with 15.d5N 15...exd5 16.Nxd5 Nxd5 17.Bxd5 Ne5 18.Bxb7 Qxb7=) 15...cxd4 16.exd4 Qd6 17.Ng5 h6 18.Nge4 This occurred most recently in Caruana – So, Wijk aan Zee 2020, and now 18...Qe7!?N would have enabled Black to fight for the advantage due to the isolated d4-pawn. 8...exd5 9.b4 c6 10.Be2 We have been following Novikov – Epishin, Tallinn 1986. The position resembles the Carlsbad lines in the Queen’s Gambit Declined, and White has begun his thematic minority attack on the queenside. I recommend meeting this plan as follows:

187

10...a6N 11.a4 Nb6!? 11...b5 is a good alternative: 12.a5 (otherwise ...Nb6 comes) 12...Ne4 13.0-0 Nd6 and Black takes aim at the c4-square. The text move has the following idea: 12.b5 The calm 12.0-0 can be met by 12...Ne4!, and if 13.b5 cxb5 14.axb5 a5 15.Nxe4 dxe4 16.Ne5 Be6ƒ Black has active pieces and a mobile passed pawn. 12...cxb5! 13.axb5 a5„ Black’s pawns have been split into three ‘islands’ but he is compensated by several active possibilities: his bishop will be well placed on e6 or f5; his knights can jump to e4 and/or c4; the king’s rook can go to c8; and the passed a-pawn may advance at an opportune moment. B21) 6.b4

188

Seizing space on the queenside is a natural idea, especially as it holds up Black’s counterplay with ...c5. 6...c6 7.Bb2 7.Be2?! (or 7.Bd3?!) gives Black easy play with 7...dxc4 8.Bxc4 b5 9.Bd3 a5 10.bxa5 and now in Sliwa – Plater, Krakow 1953, Black could have played 10...b4!N 11.Ne2 c5 12.Bb2 Rxa5³ with a somewhat better position due to the weakness of the a2-pawn. 7...Be6 Developing the bishop and asking White what he intends to do about the c4-pawn. 8.c5 This is most logical. 8.cxd5 is not completely stupid, but Black gets a comfortable game with either 8...Nxd5 or 8...cxd5!? with ideas of ...Nbd7-b6. On the other hand, if White tries to maintain the tension between the pawns on c4 and d5 then he risks becoming worse, as the following lines demonstrate: 8.Qb3?! a5! leads to big problems for White. 9.b5 is clearly the move he would like to play, but now in Stevic – Chelushkina, Pozarevac 2012, Black missed a chance to ram her opponent with:

189

9...a4!N 10.Qb4 (10.Nxa4 is refuted by 10...dxc4 11.Bxc4 cxb5 12.Qxb5 Ra5 13.Qb3 Bxc4 14.Qxc4 b5–+ and White loses a piece) 10...a3! 11.Bxa3 dxc4 12.Qxe7 Nd5 13.Qxd8 Rxd8 14.bxc6 Nxc3 15.c7 Rc8 16.cxb8=Q Rcxb8 17.Bd6 Rd8 18.Bc5 b6 19.Bxb6 Rdb8 20.Bc5 Rxa2–+ Despite White’s extra pawn, he is too far behind in development, and the penetration of Black’s rook to the first rank (not to mention the powerful passed pawn) will decide the game. Some grandmasters have dislodged the bishop: 8.Ng5 Bf5 9.h3 (9.Bd3?! Bxd3 10.Qxd3 Nbd7 11.Nf3 Nb6 12.cxd5 cxd5³ favoured Black thanks to the outpost on c4 in Sgarito – Grammatica, corr. 2013) Here I found a way to improve on Black’s play from Kveinys – Bartel, Germany 2008: 9...a5N 10.b5 a4! Ramming with the a-pawn again! 11.a3 (11.Nxa4? dxc4 gives Black a serious advantage) 11...Qa5 12.cxd5 Nxd5 13.Qd2 Nb6³ Black has a lead in development and his ideas include ...cxb5 and ...Rc8, as well as ...e5. Finally, 8.Nd2 allows Black to strike on the queenside in a different way: 8...a5! 9.b5 dxc4 10.bxc6 Nxc6 11.Bxc4 Bxc4 12.Nxc4 b5! 13.Ne5 (13.Nxb5? Qd5 is terrible for White) Now in Ibrahimli – Kantans, Poti 2016, Black could have obtained some advantage with 13...Qd6!N 14.Nd3 e5 15.a4 b4 16.Nb5 Qd7 17.Nxe5 Nxe5 18.dxe5 Qxd1† 19.Rxd1 Ng4³ when he regains the pawn with the better endgame, due to the strong protected passer. 8...Nbd7 9.h3 a5 10.a3 10.b5 was played in Strating – I. Sokolov, Amsterdam 2001. Once again, the best solution is ramming: 10...a4!N 11.Rc1 (11.Nxa4? is refuted by 11...Qa5† 12.Nc3 Nxc5–+ followed by ...Ne4, when White’s position is falling apart) 11...Qa5 12.bxc6 bxc6 13.Bd3 Bf5 14.0-0 a3 15.Ba1 Rfb8³ The bishop on a1 is shut out of play and all of Black’s pieces occupy good positions, so it is obvious that White must switch to defence.

190

10...Ne4 11.Bd3 Bf5 12.Qc2 axb4 13.axb4 In Malakhatko – Konguvel, Latschach 2013, Black’s best continuation would have been:

13...Nxc3N 14.Bxf5 gxf5 15.Bxc3 Rxa1† 16.Bxa1 e6 The doubled f-pawns are not really weak; nor is White’s queenside space advantage a problem, since a number of pieces have been exchanged. My illustrative line continues: 17.Bc3 Qc7 18.Qa2 Nf6 19.0-0 Ne4 20.Bd2 Rb8 21.Ra1 b6 22.Be1 bxc5 23.bxc5 h6 24.Qa5 Qc8= Each sides controls an open file. The c6-pawn has the potential to be weak; on the other hand, Black has an active knight and theoretically the better bishop, so overall the position remains balanced. B22) 6.cxd5 Nxd5

191

This continuation was used as White by such masters of the past as Rubinstein, Maroczy and Keres. More recently, it has been employed by GMs Eljanov, Aleksandrov, Pelletier and especially Pavel Tregubov who has played it quite a lot. In contrast to the main lines with e2-e4, the pawn on e3 offers useful support to the d4-point, but the drawback is that the c1-bishop remains passive. 7.Bc4 This is White’s most logical and natural continuation, developing a piece while attacking an enemy unit. 7.Be2 c5 transposes to 7.cxd5 Nxd5 in the note to variation B23 on page 110. The quiet 7.Bd2 does not pose much of a problem to us: 7...c5 8.Qb3 Nxc3 9.Bxc3 (9.bxc3 transposes to 9.Bd2?! in the notes to the 7.Qb3 line on the next page) 9...cxd4 10.Nxd4 e5 11.Nb5 Be6 12.Bc4 Bxc4 13.Qxc4 This occurred in Duong – Areshchenko, Kuala Lumpur 2002, and here I suggest: 13...Nc6N 14.Rd1 Qe7 15.0-0 Rfd8³ White’s knight will be thrown back by ...a6 and White’s coordination will be disrupted, which gives us reason to prefer Black’s chances. A final alternative is: 7.Qb3 Nxc3 8.bxc3 c5 If we compare this to our main line, it seems hard to believe that White can be better off having played Qb3 instead of the developing move Bc4, but we will check some lines anyway. 9.Ba3 This is the most active try. 9.Be2 transposes to the 7.cxd5 line in the note to variation B23 on page 110. 9.Bd2?! proved to be too passive after 9...Nc6 10.Be2 e5 11.dxc5 e4 12.Nd4 Qg5 13.g3 Qxc5µ in Keglevic – Krzyszton, corr. 1973, when White’s light squares and c3-pawn were weak. 192

9.Bd3 is also nothing special. One example continued 9...Nd7 10.Rb1 b6 11.0-0 Bb7 12.Ng5 when, in Veiga Rodriguez – Kappes, corr. 2013, 12...Rc8N 13.Re1 e5³ would have given Black strong pressure in the centre, while the purpose of the knight on g5 remains unclear. 9...cxd4 10.cxd4 Nc6 11.Be2 Qa5† 12.Nd2 Be6 13.Qb2 Black’s play from Kreiman – Shipov, Internet 2001, can be improved by means of:

13...Rad8!!N Remarkably, Black reveals his intention to sacrifice a piece, and there is nothing White can do about it. My analysis continues: 14.Rb1 Nxd4! 15.exd4 Bxd4 16.Qb4 Qe5 17.Rc1 Qg5! 18.g3 a5 19.Qb5 Rd5 20.Nf3 Rxb5 21.Nxg5 Rxg5 22.Bxe7 Rf5 23.Bxf8 Bxf2† 24.Kd2 Rd5† 25.Kc3 Kxf8µ Black has two pawns for the exchange, plus a mighty pair of bishops as well as some initiative against White’s exposed king. 7...Nxc3 7...Nb6 is quite playable but I consider the text move to be more combative. 8.bxc3 White has a well fortified centre, but in the future we can undermine it with ...c5 and ...e5. As mentioned earlier, the big disadvantage of White’s set-up is that the dark-squared bishop is passive, and a future e3-e4 will weaken the d4-point, as well as losing a tempo by comparison to the main lines where e2-e4 is played in one move. 8...c5 9.0-0 9.Bb2 can be met by 9...Qc7 or 9...Nc6, with a likely transposition to the 11.Bb2 line on the next page after a few moves.

193

9...Qc7 Threatening a discovered attack by ...cxd4. 10.Qe2 This is the usual move, protecting the bishop and supporting a possible e3-e4 push. 10.Qb3?! is hardly logical, as 10...Nc6 leaves White vulnerable to ...Na5 ideas. 11.Ng5?! e6 12.Be2 Na5 13.Qc2 cxd4 14.exd4 occurred in Alonso Centeno – Rojo Lopez, Valladolid 1991, when 14...e5!Nµ would have given Black a big advantage. 10.Qc2?! is another unsuccessful location for the queen. In Torrents Senal – Boix Moreno, La Pobla de Lillet 2017, it would have been logical to play 10...Bg4N 11.Nd2 Nc6 12.h3 Bf5 13.e4 Bd7 14.d5 Na5 15.Bd3 Rfe8³ when Black is ready to strike at the enemy centre with ...e6. 10.Qd3?! also leaves the queen less than optimally located. 10...Rd8 11.Bb3 was seen in Nemeth – Ronyai, Budapest 2006, when the typical developing move 11...b6N would have been good for Black. For instance: 12.Rd1 Nc6 13.Qc4 e6 14.e4 Bb7 15.Bg5 Rd7 16.dxc5 Rxd1† 17.Rxd1 Na5 18.Qd3 Qxc5µ Black is better due to weak pawns on c3 and e4. Moreover, Black has the option of ...Nxb3 at any moment, and if White preserves his bishop by retreating it then the knight can go to c4. 10.Bb3?! is another dubious decision, since the bishop will be hit by ...Na5 sooner or later. After 10...Nc6 11.Ba3 b6 12.Qe2 Na5 13.Bd5 Rb8 14.Nd2 Qd7 15.Be4 Qa4 16.Bb2 e5³ Black’s pieces were perfectly arranged in Simagin – Korchnoi, Tallinn 1965. Finally, retreating with 10.Be2 does not make much of an impression. 10...b6 11.a4 Nc6 12.Ba3 Rd8 13.Qc2 Na5 14.dxc5 bxc5 15.Rab1 occurred in Caselas Cabanas – Ruck, Santiago 1995, when I would suggest:

194

15...Bd7!N 16.Rfd1 Rab8 17.h3 h6³ Black’s bishops point menacingly towards the weak a4- and c3pawns.

10...Nc6 Black has a straightforward plan of completing development with ...Na5, ...b6, ...Bb7 and ...Rac8. 11.Ba3 White develops his bad bishop and attacks the c5-pawn. Other moves are less logical, for instance: 11.Bb2 Bg4 12.Rac1 e5 13.h3 Bxf3 14.Qxf3 cxd4 15.cxd4 exd4 16.exd4 Bxd4 17.Ba3 occurred in Weinrich – Chuchelov, Goch 1991, when 17...Rfd8!N 18.Bd5 Rac8 19.Rfd1 Bg7³ would have left White with questionable compensation for the pawn. 11.h3 Rd8 12.Bb2 Na5 13.Bd3 b6 14.Be4 Bb7 15.Bxb7 occurred in Pinto Santamaria – Moreno Uribarri, San Sebastian 2017, and now instead of misplacing his knight on b7, Black should have preferred 15...Qxb7N 16.e4 Rac8 17.d5 Qd7³ followed by ...Qa4 or ...e6. 11.a4 Na5 12.Ba2 was played in Cavatorta – Sbarra, Porto Mannu 2009, when the standard developing plan would have worked well: 12...b6N 13.e4 Bb7 14.Rd1 e6³ In my understanding, Black’s position is preferable in view of his powerful bishops and pressure along the c-file. 11.Rd1 This has been played by Tregubov, the leading specialist of this variation. 11...b6 12.Bb2 e6 13.e4 Bb7 14.Rac1 Na5 15.Bd3 Rac8 16.Qe3 Rfd8 17.h4 h6 18.e5 18.Nh2 was played in Partikas – Vetter, corr. 1994, when 18...b5!N 19.Ba3 cxd4 20.cxd4 Nc4 21.Bxc4 bxc4ƒ would have given Black a promising game, with the bishop pair and a passed pawn. 18...Qc6 19.Qe2?! 195

I also studied 19.h5N 19...g5 20.Ne1 cxd4 21.cxd4 Qa4 22.f4 gxf4 23.Qxf4 Qxa2 24.Rd2 Rxc1 25.Bxc1 Qd5 26.Rf2 Rd7 27.Nc2 Nb3 28.Rf3 Nxc1 29.Rg3 Kh8 30.Qxc1 Rc7 31.Qf4 Qd8 when Black has two bishops and an extra pawn, and White must seek salvation. We have been following Tregubov – Kulaots, Tallinn 2016. I recommend: 19...h5!N 20.Bb5 Qe4 21.Qxe4 Bxe4 22.Ng5 Bb7³ Black’s chances are preferable due to the outpost on d5; he can improve his position with the simple plan of ...cxd4, ...Bd5 and ...Nc4. 11...b6! Indirectly defending the c5-pawn. Black intends to develop according to the familiar scheme of ...Bb7, ...Na5 and centralizing the rooks. 12.Rac1 White anticipates an exchange on d4, in which case his rook will press along the open c-file. 12.dxc5 bxc5 13.Bxc5 is certainly not dangerous in view of 13...Ne5 14.Nxe5 Bxe5 15.Bd4 Bxh2† 16.Kh1 Bd6 with some initiative for Black. 12.Rab1!? deserves attention. In Herraiz Hidalgo – Vila Gazquez, Barcelona 2018, it would have been logical to play 12...Rd8N 13.Rfd1 Na5 14.Bd3 Bb7 15.e4 Rac8÷, reaching a typical Grünfeld situation where Black has pressure against the centre and along the c-file, and his chances are not worse. A different type of position occurs after: 12.Rfd1 Na5 13.Bd5!? Rb8 14.e4 An older game continued 14.Rac1 Rd8 15.dxc5 bxc5 16.Ng5 e6 17.Bf3 Rxd1† 18.Rxd1 Bd7 19.c4 Ba4 20.Rc1 Nc6 21.Ne4 Nb4³ with a more harmonious position for Black in Flesch – Simagin, Salgotarjan 1967. The knight is strong on b4 and the bishop on a4 prevents White’s rook from occupying the open d-file. 14...Rd8 15.Bb3 e6 16.Qe3 Bf8 17.Bb2 Bb7 18.h4 In Guichard – Aloma Vidal, La Massana 2015, this would have been the ideal moment to release the central tension:

196

18...cxd4!N 19.cxd4 After 19.Rxd4 Bg7µ the c3-pawn is weak and Black has ready-made pressure along the c-file. 19...Nxb3 20.axb3 Qc2µ Black enjoys the advantage of the bishop pair, and the e4-pawn is weak. 12...Rd8

13.Ng5?! White threatens f7 and leads his knight to e4 in the future, but the plan proves to be flawed. 13.Rfd1 e6 14.Ng5?! was the same kind of idea in Khan – Schenning, corr. 2013. 14...Bb7N looks like 197

a natural reply, when it is not clear what White is doing on the kingside and I prefer Black’s chances. A slightly earlier correspondence game continued: 13.Bd3 Bb7 14.Rfd1 e6 15.h4?! White should play more modestly with 15.h3, though after 15...Rac8÷ Black is not worse and it is not clear how White should strengthen his position. The text move was played in Kolanek – Huerga, corr. 2012. I am not impressed by White’s attacking plan, and propose the following improvement:

15...Rac8N 16.h5 Rd5! 17.hxg6 hxg6 White has spent a couple of tempos opening the h-file, but Black is the one who gets to use it! My illustrative line continues: 18.dxc5 bxc5 19.Ba6 Bxa6 20.Qxa6 Ne5 21.Nxe5 Rxe5 22.c4 Rh5 23.g3 Rb8 24.Kg2 Be5 25.Rh1 Rb6 26.Qa4 Qb7† 27.f3 Rd6 28.Qa5 Qb6 29.Qxb6 Rd2† 30.Kg1 Rxh1† 31.Kxh1 axb6–+ White cannot avoid material losses, as Black’s pieces are too active. 13...e6 14.h4 White would love to ram us on the kingside but this action will not bring him success, as his pieces are not suitably placed to support an attack. Transferring the knight to the centre also gives Black a wonderful game: 14.Ne4 Bb7 15.Ba6 Na5 16.Bxb7 Qxb7 17.Nd2 Qc6µ Reed – Grimault, Gorges 2017. White’s knight manoeuvre via f3-g5-e4d2 achieved nothing except losing time and enabling Black to strengthen his position. 14...h6 15.Ne4?! The knight continues its ill-conceived journey, spending precious time in the process.

198

It was better to admit the earlier mistake with 15.Nf3, although 15...Bb7 16.Rfd1 Rac8 17.Ba6 Na5³ is still pleasant for Black. The bishop on a3 is restricted by the c5-pawn, and White’s light squares may become weak after the exchange of those bishops. Nevertheless, this was the lesser evil and nothing terrible has happened. 15...Bb7 16.Bd3 f5! The first call to storm the bastions of the king! 17.Ng3?! 17.Nd2 was better although 17...Ne5! is a great move which yields a clear advantage:

18.dxe5 Qd7 19.Bc4 Qxd2 20.Bxe6† Kh7 21.Qxd2 Rxd2 22.Rcd1 Rad8 23.Rxd2 Rxd2µ Black will recapture the e5-pawn to reach an endgame with excellent winning chances, thanks to the active rook and superior pawn structure on the queenside. The text move avoids the trick noted above but leads to even greater problems. 17...h5! With ...Qe7 coming, the h4-pawn is doomed. We could end the opening analysis here or earlier, but it is worth including the rest of the game, as it did not last very long. 18.e4 This desperate attack is doomed to fail; but if you are going to die anyway, you may as well put on some music. 18...f4 19.e5 199

19.Nh1 looks ridiculous, and after 19...cxd4 20.cxd4 Qf7 21.d5 Nd4–+ Black will play ...f3 next, although he could also win by picking up the pawns on d5 and h4. The text move is a desperate attempt to stir up the water.

19...Nxe5!? Black plays safely and prevents any possible attacks. Eating the piece was also fine: 19...fxg3N 20.Bxg6 Ne7 21.Bc2 Qc6 22.fxg3 Rf8–+ and the g2-square must be constantly monitored, so White’s attack is over. 20.dxe5 fxg3 21.f4?! This loses immediately, but there was no real chance of salvation. 21.Bxg6 Qxe5 22.Qxe5 Bxe5 is a winning endgame for Black. For instance, 23.fxg3 Rd2 24.Rf2 Rxf2 25.Kxf2 Rf8† 26.Kg1 Bxg3 27.Rc2 Ba6 28.c4 Rf4–+ and White’s weak pawns on h4, c4 and a2 will fall like autumn foliage.

200

21...Qe7 22.Bxg6 Qxh4 23.Qxh5 Rd2! Perhaps this is the detail that White missed. 24.Rf3 24.Qxh4 loses to 24...Rxg2† 25.Kh1 Rh2† 26.Kg1 Rxh4 when there is no defence again ...Rh1 mate. The text move was played in Hobaica – Fusco, Pinamar 2008, but White resigned without waiting for 24...Qh2†. B23) 6.Be2 This has been White’s most popular choice, simply completing the development of the kingside before turning to the other flank. In the early days this move was played by Lasker and Tarrasch, and it was even used successfully in a World Championship game by Petrosian against Botvinnik in 1963. Modern-day grandmasters who have played this way include Mamedyarov, Fressinet, Tomashevsky and Milov. 6...c5! Immediately attacking the centre is the most logical and powerful continuation. The Grünfeld Defence effectively becomes a reversed Catalan where White’s extra tempo should be enough to secure equality, but not to fight for an advantage. White usually proceeds with either B231) 7.dxc5 or B232) 7.0-0. A third possibility is: 7.cxd5 Nxd5 8.Qb3!? 201

The more popular 8.0-0 gives Black time to create pressure against the d4-point: 8...Nc6 9.Nxd5 (9.Qb3?! Be6 10.Qxb7 cxd4 11.exd4 transposes to the 9.cxd5 Nxd5 10.Qb3?! line in the note to variation B232 below) 9...Qxd5 10.dxc5 Qxc5 11.e4 Rd8 12.Qa4 Now in Brichikova – Ilieva, Sunny Beach 2007, the logical 12...Nd4!N 13.Nxd4 Bxd4µ would have left Black with a powerful bishop on d4, and a general advantage in piece activity. 8...Nxc3 9.bxc3 Nc6 10.0-0 b6 11.Rd1 Qc7 12.e4 12.Ba3 Na5 13.Qa4 Bd7 14.Bb5 Bxb5 15.Qxb5 a6 16.Qe2 was played in Traudes – Schaefer, Germany 2011, and here I recommend 16...Qc6!N intending to deploy the queen on a4. For instance, 17.Bc1 e6 18.h3 Qa4³ when Black can plonk the knight on c4 any time he feels like it, and White’s bishop is passive. 12...Na5 13.Qa3 Another option is 13.Qb1 Bb7 14.Be3 as played in Thys – Hackenberg, corr. 2016. Now I like 14...cxd4N 15.cxd4 Rac8÷ with a typical situation for the Grünfeld Defence. I would choose Black’s side if given the choice, as his bishops exert meaningful pressure on White’s centre, and the c-file is in Black’s hands. 13...Bb7 14.Bd3 We have been following Kuhne – Z. Ivanovic, corr. 2015. Here I suggest:

14...e6N 15.Rb1 Rac8÷ With an unclear position, where Black’s chances are not worse. B231) 7.dxc5 dxc4

202

8.0-0 White can hardly benefit from delaying castling: 8.Qxd8 Rxd8 only helps Black’s development, and after 9.Bxc4 Nbd7 10.c6 bxc6 11.0-0 Nb6 12.Be2 Nfd5 13.Bd2 Nxc3 14.Bxc3 Bxc3 15.bxc3 c5³ White was under some pressure in Sermier – Ponizil, Prague 2016. 8.Bxc4 Qxd1† 9.Kxd1 can be played in the hope of utilizing the king in the centre in the endgame, but it is risky for White to play this way when there are still many pieces on the board. 9...Nbd7 10.c6 bxc6 11.Ke2 was played in S. Kasparov – Aronian, Minsk 1998, when Black missed a powerful idea:

11...Nd5!N 12.Nxd5?! Objectively White should try something else, but he is in deep trouble no matter 203

how he plays. 12...cxd5 13.Bxd5 Ba6† 14.Ke1 Rab8–+ White’s king is stuck in the centre and Black’s pieces are exceedingly active, ...Nc5-d3† being the principal threat. 8...Na6 This is the most flexible way of picking up the c5-pawn. 9.Qd4 This was Mamedyarov’s choice. Another option is 9.c6!? Qc7 10.Qa4 Nc5 11.Qxc4 Qxc6 12.b4 Nce4 13.Qxc6 bxc6 14.Bb2 a5 15.bxa5 Ba6 16.Bxa6 when the players agreed a draw in Acosta – Giesel, corr. 2017. Play would continue 16...Rxa6 followed by ...Rxa5, when the mutual weaknesses of the a2- and c6-pawns balance each other. 9...Qc7 10.Bxc4 10.Qxc4 Nxc5 also gives White no more than equality, for instance: 11.Nb5 (11.b4 Na6 12.Qxc7 Nxc7 13.Bd2 was seen in Miladinovic – Videnova, Belgrade 2013, when 13...Bd7N= followed by centralizing the rooks would have given Black a comfortable game) 11...Qb6 12.b4 Galyas – Hagen, Budapest 2013. Here I suggest:

12...a6!N 13.bxc5 Qxb5 14.Qxb5 axb5 15.Bb2 Ne4 16.Bxg7 Kxg7 17.Bxb5 Ra5 18.a4 Nxc5= With a balanced endgame. 10...Nxc5 11.Qh4 Black’s play from Mamedyarov – Giri, Beijing 2012, can be improved by means of:

204

11...Bf5N 12.b3 h6! Intending to drive the queen away. 13.Bb2 g5 14.Qd4 Rfd8 15.Qe5 Qxe5 16.Nxe5 e6 17.Rfd1 Nfd7= Black does not have a shadow of a problem. B232) 7.0-0 White simply castles, believing his foundations in the centre to be strong. 7...cxd4 8.exd4 This move gives rise to a reversed Tarrasch Defence, where White’s extra tempo at least gives him some equalizing chances... If White avoids the isolated queen’s pawn, Black also gets an easy game. 8.Nxd4 dxc4 9.Bxc4 a6 Intending ...b5 and ...Bb7. 10.a4 Qc7 11.Qe2 11.b3?! seriously loosens the long diagonal. 11...Rd8 12.Qc2 occurred in Zhigzhitzhapov – Angarov, Irkutsk 2017, when 12...Ng4!N 13.g3 Nc6 14.Ne4 Qb6µ would have left White with significant weaknesses on both sides of the board. 11...Rd8 12.f3 12.h3 was played in Tica – Shariyazdanov, Opatija 2013. I recommend improving with 12...e6N 13.b3 Nbd7 14.Ba3 Ne5³ when Black will gain the advantage of the bishop pair. 12...Nbd7 13.Ba2 Nb6 14.Rd1 205

We have been following an older game, Barcza – Korchnoi, Stockholm 1962. My improvement is:

14...e5!N With good play for Black. The critical line continues: 15.a5!? exd4 16.axb6 Qxb6 17.exd4 Be6 18.Bxe6 Qxe6 19.Qxe6 fxe6³ Black’s chances are higher, as the d4-pawn is weaker than the e6-pawn and the activity of the bishops on c1 and g7 is different.

8...Nc6 If you have played with White against the Tarrasch Defence, you should be able to handle Black’s position using exactly the same ideas, with just minor changes to take into account White’s extra tempo when the situation demands it. By the way, this position can also be reached via the Panov-Botvinnik Attack against the Caro-Kann, when Black opts for the ...g6 set-up and White goes for timid 206

development Nf3 and Be2 instead of the more ambitious main lines. 9.h3 Defending against ...Bg4. It would be too much to offer a fully comprehensive repertoire against the reversed Tarrasch Defence; but as usual, I will give a selection of examples to show White’s main plans and how Black should deal with them: 9.Bf4 Bg4 10.h3 (10.c5?! Ne4 is covered below in the note on 9.c5) 10...Bxf3 11.Bxf3 dxc4 12.d5 Nb4 turns out well for Black. 13.d6 e6 14.d7 Nd3 15.Bd6 occurred in Materniak – Slawinski, corr 1999, and now it looks great to play:

15...Nxd7N 16.Bxb7 Rb8! 17.Bxf8 Kxf8 18.Be4 Nxb2 19.Qd6† Kg8 20.Rac1 Be5 21.Qd2 Qa5 22.Qxd7 Bxc3µ Black is materially at least okay with two pawns and a knight for a rook, and the passed c-pawn will make White’s life miserable. 9.Re1 Bg4 10.cxd5 Nxd5 11.h3 Be6 12.Bf1 Rc8 13.Bg5 h6 14.Be3 Nxe3 15.fxe3 Such a transformation can often happen in these positions: White no longer has an isolated pawn, but 15...f5 16.Qa4 Bf7³ still leaves Black with a preferable position due to the two bishops, and in the future he can open the game with ...e5. 9.c5 Relieving the central tension in this way gives Black an additional hook on the queenside. 9...Ne4 10.Be3 10.Bf4?! Bg4 practically forces White to lose a tempo: 11.Be3 (11.Ne5? simply lost a pawn after 11...Nxe5 12.Bxg4 Nxg4 13.Qxg4 Bxd4 14.Nxe4 dxe4–+ in Djordjevic – Cabrilo, Pozarevac 2013) 11...b6 12.cxb6 Nxc3 13.bxc3 axb6 14.h3 Bxf3 15.Bxf3 Na5 16.Re1 e6 17.Be2 Qc7 18.Qc2 Rfc8³ Liu – Brattain, Dayton 2017. Black has a small but stable advantage due to his superior pawn structure and 207

pressure along the a- and c-files. 10...b6 11.cxb6 Nxc3 12.bxc3 axb6 13.Qd2 We have been following Brumen – Smirin, Rabac 2004, and here I recommend:

13...Na5!N 14.Bh6 Ba6 15.Bxg7 Kxg7 16.Bxa6 Rxa6³ Black’s superior structure and outpost on c4 give him some advantage. White’s other way of releasing the central tension is: 9.cxd5 Nxd5 White has tried various ideas from here, but in all cases he struggles to gain enough activity to offset the weakness of the isolated pawn.

10.Qb3?! 208

This is the most active and forcing continuation, but it does not work. White should settle for a different continuation, although none of the following alternatives inspires much confidence in his position: 10.h3 Be6 11.Ne4 Qc7 12.b3 Bf5 13.Ng3 Nc3 14.Qd2 Nxe2† 15.Nxe2 Rad8³ left Black with the bishop pair as well as the isolated pawn as a target in Ozturk – Peng Zhaoqin, Beijing (rapid) 2008. 10.Re1 Be6 11.Bg5 h6 12.Be3 Rc8 13.Qd2 Kh7!? (13...Nxe3N can also be played, with similar ideas as in the line below) 14.Rac1 Qa5³ gave Black active pieces while the d4-pawn remained a concern for White in Ravi – Hodgson, London 1989. 10.Bg5 h6 11.Be3 Be6 12.Qd2 occurred in Ozgibcev – Kryakvin, Sochi 2018, and here I recommend 12...Nxe3N 13.fxe3 Rc8 14.Rfd1 Qb6 15.Na4 Qc7 16.Nc5 Bf5 17.Rac1 Rfd8³ when Black enjoys the bishop pair and the better pawn structure. 10...Be6! 11.Qxb7 Nxd4 12.Nxd4 Bxd4 White has traded off his weak pawn but the price is too high, as Black dominates the centre.

13.Bh6 13.Rd1 Rb8 14.Qa6 Nb4 15.Qa4 Qb6µ also gave Black an obvious advantage in G. Gonzalez – Goldenberg, Mar del Plata 1961. The powerful bishop on d4 puts pressure on the f2-weakness as well as on White’s queenside. 13...Rb8 14.Qa6 Rxb2! 15.Nxd5 Qxd5 16.Bxf8 Kxf8 17.a4 Rb6 18.Qd3 Bxa1 19.Rxa1 Qxd3 20.Bxd3–+ Black had excellent chances to realize his extra pawn in Sepulveda Herrera – Lopez Silva, Santiago de Chile 2014. 9...dxc4 10.Bxc4 Na5 We drive the bishop away to enable ...Be6. Later, Black’s knight and/or bishop may utilize the c4square.

209

11.Be2 11.Bd3 is no better. 11...Be6 12.Re1 Nc6! (but not 12...Rc8?! 13.Rxe6! fxe6 14.Qe2 Kh8 15.Ng5© when White has a promising initiative for the exchange) 13.Bf1 (13.Rxe6? does not work here: 13...fxe6 14.Qe2 Nxd4–+ White has no compensation) 13...a6 14.Ng5 Bd5 15.Be3 e6 16.Rc1 Rc8 17.Bd3 Qd6 18.Qd2 Rfd8³ The players agreed a draw in Sanner – Monasterio, email 2004, but Black stands better. 11...Be6 Continuing the plan.

12.Bg5 This is White’s most logical choice. The bishop goes to a fighting position and prepares, on occasion, the exchange of dark-squared bishops by Qd2 and Bh6. 12.b3? covers the c4-square but destabilizes the knight on c3. A queenside fianchetto hardly ever combines well with an IQP on d4, and this is no exception. 12...Rc8 13.Bb2 Nd5 14.Qd2 Nxc3 15.Bxc3 Nc6 16.Rfd1 Bd5 17.Ne5 e6 18.Rac1 Qh4 19.Ng4 Rfd8µ occurred in Frieser – Rooze, Bad Sooden 2016, when Black had strong pressure against the d4-point while White had no targets for counterplay. 12.Ne5 has been the most popular continuation but I like Black’s chances after 12...Nd5. For example, 13.Nxd5 Bxd5 14.Be3 occurred in Hauksson – Hjertenes, Norway 2000, when 14...Qd6N 15.Bf3 Rfc8 16.b3 Nc6³ would have prepared further exchanges to exploit the weakness of the isolated pawn. Another thematic continuation is: 12.Re1 Rc8 13.Bg5 13.Ne5 Nd5 14.Bd2 Nc6 15.Nxc6 Rxc6 16.Bf3 Rc4 17.Qe2 Rxd4 18.Bxd5 Bxd5 19.Be3 Bc4 210

20.Qf3 Rd7µ left Black clearly better in Zhukova – Beshukov, Berlin 1995, thanks to the two bishops in an open position. 13.Ng5 Bc4 14.Bf3 is also not dangerous. In Rausis – Fominyh, Cairo 2001, Black should have played 14...Nd5!N intending 15.Nxd5 Bxd5 16.Bxd5 Qxd5 17.Rxe7 Nc6 18.Re1 Nxd4 19.Be3 Rfd8µ when Black’s pieces are beautifully centralized. 13...h6 14.Bh4 Nc4 15.Bxc4 Bxc4 16.Ne5 This position was reached in J. Horvath – Bernasek, Austria 2012. I propose expanding on the queenside with:

16...a6!?N 17.Rc1 Bd5 18.Qd2 b5³ White has no initiative to compensate for the isolated pawn, and any further piece exchanges will take Black closer to a favourable endgame. 12...Rc8 13.Rc1 13.Re1 was covered in the note above, via 12.Re1 Rc8 13.Bg5. 13...Nc4

211

14.Bxc4 I am not convinced by White’s decision to trade off his bishop, even though the knight on c4 was unpleasant for him. 14.b3 expels the knight at the expense of loosening White’s queenside. 14...Nd6 15.Re1 was seen in Peng Xiongjian – Sivuk, Sanya City 2017, and here I suggest expanding on the queenside: 15...a6N 16.Bd3 Re8 17.Qd2 b5 18.Bf4 Qb6³ Because of the isolated d4-pawn, Black’s chances are somewhat higher. 14...Bxc4 15.Re1 Re8 White will have to contend with Black’s bishop pair in addition to the isolated pawn. 16.Qa4

212

16...Qb6 Black defends the a7-pawn and attacks b2. Our ‘iron friend’ offers 16...a6!?N 17.Ne5 Be6 18.Qb4 h6 19.Bf4 Nd5 20.Nxd5 Qxd5³ as its preferred way for Black to keep some edge. 17.Qa3 Be6 18.Rcd1 I also checked 18.Be3N 18...Nd5 19.Nxd5 Bxd5 20.Ne5 when Black keeps the upper hand with the help of a remarkable idea:

20...Ra8!N Defending the a-pawn before neutralizing White’s active queen. 21.Rc3 Qd6 22.Qxd6 exd6 213

23.Nc4 Re6³ Black’s chances are higher thanks to the two bishops. We have been following Krainski – Maksimenko, Jastrzebia Gora 2015. The most logical continuation would be:

18...h6!N 18...Rcd8?! was the game continuation, when 19.b3!?N would have left Black with nothing special. 19.Bh4 Red8 20.Rd2 g5 21.Bg3 Bf8³ The black pieces are quite harmoniously arranged. The dark-squared bishop appears to be temporarily misplaced by the need to defend e7, but in the future Black can prepare an explosive ...e5 break, targeting the queen on a3. Conclusion The e2-e3 systems are certainly not the most ambitious way to meet the Grünfeld Defence, as White limits his central activity and restricts his dark-squared bishop. Nevertheless, we should not take White’s plans lightly. We began by examining 3.Nf3 Bg7 4.e3 0-0 5.Be2, which was recommended by Axel Smith in e3 Poison. By delaying the development of the knight to c3, White renders ...d5 less attractive, but 5...b6 6.0-0 Bb7 7.Nc3 e6!? followed by ...c5 gives Black a comfortable game with good prospects for counterplay. We then turned our attention to 3.Nc3 d5 4.e3 Bg7, when White has several options. 5.Qb3 e6 6.Qa3 interferes with Black’s castling, but 6...a5! is a good answer, with ...c6 to follow. If White is not careful, he can lose a pawn by allowing ...dxc4; and even if he defends with b2-b3, Black gets an 214

excellent game with ...Na6 followed by either ...Nb4 or ...b6 and ...c5, rendering the queen on a3 useless. Next we looked at simple development with 5.Nf3 0-0, when the path divides6.b4 c6 7.Bb2 is not a bad way to develop, but 7...Be6 is an adequate response. We then considered 6.cxd5 Nxd5 7.Bc4, when 7...Nxc3 8.bxc3 c5! gives Black a thematic Grünfeld set-up with a simple plan of ...b6, ...Bb7 and ...Na5, while White’s dark-squared bishop is stuck behind the e3-pawn. Finally, 6.Be2 c5! gives Black a healthy reversed Catalan, which could become a reversed Tarrasch Defence in the event that White allows ...cxd4 and a subsequent IQP structure. Despite the extra tempo compared with a normal Tarrasch, Black has an excellent game and the isolated pawn usually becomes a burden on the defence.

215

A) 4.Bg5 Bg7!? 119 A1) 5.e3 119 A2) 5.Bxf6 Bxf6 121 A21) 6.Nxd5?! 121 A22) 6.cxd5 c6! 121 A221) 7.dxc6 122 A222) 7.Rc1!? 123 A223) 7.e4 0-0 125 A2231) 8.Bc4?! 126 A2232) 8.Nf3 127 A2233) 8.e5 129 B) 4.Nf3 Bg7 5.Bg5 Ne4! 133 B1) 6.Bf4 134 B2) 6.Bh4 Nxc3 7.bxc3 dxc4! 142 B21) 8.e3 143 216

B22) 8.Qa4† 144 B3) 6.cxd5 Nxg5 7.Nxg5 e6 148 B31) 8.Qd2 150 B32) 8.Nf3 156

1.d4 Nf6 2.c4 g6 3.Nc3 d5 This chapter will deal with lines where White develops his bishop to g5 in the early stages of the game. There are two main variants: A) 4.Bg5 and B) 4.Nf3 Bg7 5.Bg5. A) 4.Bg5 This method of developing the bishop immediately threatens Bxf6 and Nxd5, in contrast to systems with Bf4, where White aims at the c7-point, as well as to the early e2-e3 lines, where White leaves the bishop blocked behind the pawn chain. According to my database, this move was first played by Alekhine at Vienna 1922, when he lost to none other than Ernst Grünfeld himself. Later in the 20th century, Euwe, Petrosian, Botvinnik and Taimanov all made their contributions to the system. Among modern players, Mamedyarov, Aronian, Carlsen and Tomashevsky are among its adherents. 4...Bg7!? 4...Ne4 has been by far the most popular choice but simply ignoring White’s threat has a certain appeal. The young Peter Svidler was the first to play this way at a high level, in 1991. It was subsequently taken up by Adorjan, Yandemirov and, in the 21st century, Kasparov, Grischuk, Aronian and Carlsen, which may serve as a certain stamp of quality.

217

We will examine the rather timid A1) 5.e3 followed by the more critical A2) 5.Bxf6. 5.Nf3 leads to variation B. A1) 5.e3 c5 6.Nf3 It is worth knowing that 6.dxc5 can be strongly met by 6...Ne4! 7.Nxe4 dxe4 8.Qxd8† Kxd8 9.0-0-0† Ke8 10.Ne2 as played in S. Ivanov – Baigildin, Satka 2011, when I recommend:

10...Bd7!N Anticipating the c5-c6 desperado before playing ...Na6. 11.Nd4 a5 12.f3 Na6³ Black is a little better thanks to his pawn structure. 218

6...cxd4 7.exd4 7.Nxd4 Ne4! 8.Nxe4 dxe4 resembles the previous note; here too, Black should obtain a favourable endgame. For instance, 9.Qd2 h6 10.Bh4 Nc6 11.Nxc6 Qxd2† 12.Kxd2 bxc6 13.Rb1 was played in La Rosa – Heiermann, corr. 2013, when Black could have fought for the advantage with:

13...a5!N 14.Be2 a4 15.Rhd1 Be6 16.Bg3 Ra7³ With pressure along the half-open b-file. 7...0-0! Black is not afraid to sacrifice a pawn for the sake of rapid development. 8.Bxf6 8.Be2 gives White a strange version of variation B232 of the previous chapter, where he has played Bg5 instead of castling. 8...Be6!?ƒ is one good reply. 8.cxd5 Nxd5 makes life easy for Black, for instance: 9.Be2 h6 10.Bh4?! (White should prefer 10.Be3 as in Menna Barreto – Perez, Buenos Aires 2017, although 10...Be6N 11.0-0 Nc6 12.Rc1 Rc8³ still gives Black a pleasant game playing against the IQP) 10...Nc6 11.0-0 Be6 12.Qd2 In De Wit – Faybish, Belgium 1997, Black could already have taken decisive action:

219

12...g5!N 13.Bg3 g4 14.Ne5 Nxc3 15.Qxc3 Nxd4–+ With a healthy extra pawn as well as a positional advantage. 8...exf6 9.cxd5 No better is 9.Nxd5 Nc6 10.Be2 f5 11.0-0 Be6 12.Qb3 Na5 13.Qb5 as in Kraiouchkine – Kiewra, Montreal 2013, when 13...a6 14.Qb4 Bxd5 15.cxd5 Qxd5 16.Rac1 Rfc8³ would have left the d4-pawn chronically weak. 9...Nd7 10.Be2 Nb6 11.Qb3

11...a5! 12.a4 220

It is logical to prevent ...a5-a4. 12.0-0 a4 13.Qb5 Re8 14.Rfe1 Bd7 15.Qd3 Ra5 16.Qd2 Nxd5 17.Bc4 Nb6 18.Bd3 f5³ was pleasant for Black in De Souza – Azevedo Pessoa, Brazil 2003. 12...Qd6 13.0-0 Rd8 14.Rfe1 We have been following Melnikov – Orlov, St Petersburg 2005. Here I suggest:

14...Qb4!N 15.Qa2 Bf8 16.h3 Bf5 17.Rac1 Rac8© Black has more than enough compensation for the pawn because of the two bishops and the active distribution of his pieces. A2) 5.Bxf6 Bxf6 By making this trade, White signifies his intention to accept the pawn sacrifice. Black, in turn, receives compensation in the form of a strong dark-squared bishop and pressure against d4. We will consider the dubious A21) 6.Nxd5?! followed by the main line of A22) 6.cxd5. A21) 6.Nxd5?! Bg7 7.e3

221

7...c5! Black is well placed to exert powerful pressure against d4. 8.Nf3 Nc6 9.Be2 White traded one weak pawn for another with 9.dxc5 Bxb2 10.Rb1 Bg7 11.Qd2 0-0 12.Bd3 in Sachdev – Kiran, New Delhi 2008. I recommend targeting the c5-pawn as follows:

12...Nb8!N 13.0-0 Na6 14.Nd4 Nxc5 15.Be2 e6 16.Qb4 b6 17.Nc6 Qe8 18.Nce7† Kh8 19.Nxc8 Rxc8³ The c4-pawn is isolated and weak, so White must switch to defence. 9...e6 10.Nc3 cxd4 11.exd4 Nxd4 12.Nxd4 Qxd4 13.Qb3 222

Exchanging queens does not alter the evaluation: 13.Qxd4 Bxd4 14.Rc1 Ke7 15.0-0 Bd7 16.Rfd1 Be5 17.Bf3 Rab8³ Black’s bishop pair gave him a lasting edge in Sinclair – Krivoshey, Dos Hermanas 2003. 13...0-0 14.Rd1 We have been following Burman – Grigutavicius, Skelleftea 1999. I suggest improving with: 14...Qc5N 15.Bf3 Rb8 16.0-0 b6 17.Nb5 a6 18.Nd6 Be5 19.Nxc8 Rfxc8³ Despite opposite-coloured bishops, White is a long way from securing a draw. The c4-pawn is weak and ...b5 is threatened. And even if White manages to stabilize his queenside, Black will still have the more potent pawn majority and will be able to press without any risk. A22) 6.cxd5

This is the correct capture. White destroys the central pawn and prepares to seize more space with e2e4. Unlike the previous variation, Black finds it harder to attack the d4-point, which tends to be the central focus of his counterplay in these positions. 6...c6! A great solution, continuing the assault on White’s centre. The first player may react with A221) 7.dxc6, A222) 7.Rc1!? or A223) 7.e4. Some strong GMs have tried returning the pawn with: 7.e3 cxd5 8.Bb5† Omitting this check does White no favours. For example, 8.Qb3 e6 9.Nf3 0-0 10.Be2 Nc6 11.0-0 Na5 12.Qc2 was played in Bermejo – Carames, Buenos Aires 1966, when 12...b6!N 13.b4 Nc4 223

14.Rfc1 Bb7 would have given Black a nasty initiative. 8...Kf8 Moving the king does not inconvenience Black in the least, as the g7-square is readily available. 9.Nge2 The evaluation is about the same after the other knight development: 9.Nf3 Kg7 10.0-0 Bg4 11.h3 Bxf3 12.Qxf3 e6 13.Rac1 a6 14.Ba4 b5 15.Bb3 Nd7 16.Rfd1 Rc8 17.Ne2 Qb6³ Despite the symmetrical pawn structure and opposite-coloured bishops, Black was able to exert unpleasant pressure in Fressinet – Vachier-Lagrave, Nancy 2012. 9...Kg7 10.Qb3 10.Nf4 e6 11.0-0 was played in Karasev – Svidler, St Petersburg 1995, when 11...Be7N 12.Rc1 Bd7 13.Be2 h5 14.Na4 b6 15.Nd3 Nc6³ would have given Black good prospects with the two bishops. 10...e6 11.0-0 a6 12.Bd3 Nc6 13.Rac1 In Nakamura – McShane, Internet 2004, it would have been good for Black to play:

13...Na5N 14.Qd1 Bd7 15.Nf4 Rc8³ Black not only has the two bishops: his knight can also go to c4 at any moment, and he has the option of capturing space on the queenside with ...b5. A221) 7.dxc6

224

Obviously this direct move is an important option to consider, and it has been tested at the highest level. 7...Bxd4 8.cxb7 8.Qb3? gives away White’s material advantage while needlessly exposing the queen to attacks by the enemy pieces. 8...Nxc6 9.e3 Bg7 10.Bb5 0-0 11.Nf3 Be6 12.Qa3 Qb6 13.Bxc6 bxc6 14.0-0 Rab8 15.Nd1 c5µ was clearly better for Black in Hoang Thi Nhu – Abdumalik, Sharjah 2014, thanks to her active pieces, better development and bishop pair. 8...Bxb7 9.e3 I checked two other continuations: 9.Nf3?! Bxc3† 10.bxc3 Qa5 11.Rb1 (11.Qb3 0-0 12.e3 transposes to the note to White’s 11th move in the main line below, where Black also stands slightly better) occurred in Moiseenko – Le Quang Liem, Khanty-Mansiysk 2013, when natural and strong would have been:

225

11...Qxc3†N 12.Nd2 Bc6 13.Qb3 Qa5³ Black is somewhat better due to the lagging development of White’s kingside and the pressure along the a8-h1 diagonal. 9.Qb3 Bxc3† 10.Qxc3!?N (10.bxc3 Qc7 11.Nf3 Nd7 12.e3 Bxf3 13.gxf3© gave Black excellent compensation for the minimal material deficit in Korniushin – Yandemirov, Novgorod 1997, due to White’s poor pawn structure) 10...0-0 Black’s lead in development offers at least enough value for the pawn, for instance: 11.Nf3 Nd7 12.Qa3 Qb6 13.g3 e5 14.Qb3 Qa5† 15.Qc3 Qxc3† 16.bxc3 e4 17.Nd4 e3 18.Rg1 exf2† 19.Kxf2 Nf6© Black has excellent compensation; the c3-pawn is weak and it is unlikely that White will be able to hold it. 9...Bxc3† 10.bxc3 Qa5 11.Qd4 This is White’s best try. 11.Qb3?! proved to be inaccurate after 11...0-0 12.Nf3 Bxf3 13.gxf3 Nd7 14.Qb4 Qf5 15.Be2 Rab8 16.Qe4 Qxe4 17.fxe4 Rb2 18.Rd1 Nc5µ in Orlinkov – Najer, Moscow 1996. Black will restore the material balance and White will be left with severely weak pawns. 11...0-0 12.Qb4 This position was reached in Ponomarenko – Amann, corr. 2003. I recommend improving with:

226

12...Nc6!N 13.Qxa5 Nxa5 14.Nf3 Rfc8 15.Rc1 Rab8 16.Be2 Bd5³ Black has strong pressure against the weak pawns on c3 and a2, and White faces a difficult defence. A222) 7.Rc1!? 0-0 7...cxd5? would be refuted by 8.Nxd5, of course. 8.dxc6 White can also switch plans with: 8.e4 Bg7 9.Nf3 9.dxc6N 9...Nxc6 10.d5 Nd4 11.Nf3 Nxf3† 12.Qxf3 Qb6© gives Black excellent compensation; the dark-squared bishop is especially strong. 9...cxd5 10.exd5 10.e5? was a strategic error in Battaglini – Deutsch, Petah Tiqwa 2011. If White wishes to play this pawn structure, he should do so without having committed his knight to f3 and rook to c1. Black could have obtained an excellent position with 10...Nc6N 11.h3 a6 12.Be2 e6 13.0-0 f6 14.exf6 Qxf6µ when the dark-squared bishop comes to life and the d4-pawn is extremely weak. 10...Nd7 11.Bd3 11.Bc4N 11...Nb6 12.Bb3 Bg4 13.h3 Bxf3 14.Qxf3 Bxd4 regains the pawn with a pleasant game for Black. 11...Nb6 12.Qb3 e6 13.0-0 I also checked 13.dxe6N 13...Bxe6 14.Qc2 Bg4 15.Be2 Bxf3 16.Bxf3 Re8† 17.Be2 Bxd4 when Black is fine. 13...Nxd5 14.Nxd5 Qxd5 15.Qxd5 exd5 16.Rc7 Rb8 17.Ne5 Be6= All this happened in Chow – Ahluwalia, England 2018. White has an active knight on e5, while 227

Black has the advantage of two bishops. White’s rook looks impressive on c7 but Black can easily exchange it or chase it away, so overall the chances can be estimated as equal.

8...Qxd4 9.Qxd4 Bxd4 10.cxb7 Bxb7 11.Nf3 Bf6 The position bears a certain resemblance to the previous variation. In return for the sacrificed pawn, Black has some initiative thanks to the bishop pair and the backwardness of White’s kingside development. 12.e3 12.g3?! looks inadvisable: 12...Nd7 13.Bg2 Rab8 14.Na4 (14.Rc2? proved to be a serious mistake after 14...Bxf3! 15.Bxf3 Rxb2! 16.Rxb2 Bxc3† 17.Rd2 Rb8 18.Kd1 Bxd2 19.Kxd2 Rb2† 20.Kc3 Rxa2µ when Black had a healthy extra pawn in Nezad – Kulaots, Gibraltar 2011) In Jumabayev – Ganguly, Guangzhou 2010, it would have been logical to continue:

228

14...Rfc8N 15.0-0 Ba6!ƒ With unpleasant pressure against the pawns on b2 and e2. 12...Na6 13.Bc4 A plausible alternative is 13.Be2N 13...Rac8 14.0-0 Rfd8 when Black’s powerful dark-squared bishop helps to keep the balance. 15.Rfd1 (or 15.a3 Nc5 16.Rfd1 Nb3 17.Rxd8† Rxd8 18.Rd1 Rc8© when it is hard for White to neutralize the pressure on the dark-squared diagonal) 15...Rxd1† 16.Bxd1 Bxf3 17.Bxf3 Nb4 18.Kf1 Nd3 19.Rc2 Nxb2 20.Rxb2 Rxc3 With a dead equal endgame. I also considered 13.Bb5!?N 13...Rab8 14.0-0 Rfd8 15.Rfd1 (not better is 15.Bc4 Rbc8 16.b3 Nb4 17.a3 Nc6 followed by ...Na5, when Black wins a pawn on the queenside anyway) 15...Rxd1† 16.Nxd1 Bxf3 17.Bxa6 Bxd1 18.Rxd1 Rxb2= Once again, Black regains the pawn and the draw is virtually certain. Returning to our main line, I propose a novelty to improve on Li Yankai – J. Garcia, Ho Chi Minh City 2017.

229

13...Nb4!N I like this provocative move. 14.b3 After 14.0-0 Rac8 15.b3 Rfd8 White will not be able to retain his extra pawn. For instance, 16.a3 Nc6 17.Ne4 Bb2 18.Rc2 Bxa3÷ and White must take care not to become worse against the two bishops. Another important line is: 14.a3 Rfc8! 15.Bxf7† (also playable is 15.b3 Nc6 16.0-0 Na5 17.Nd5 Bxd5 18.Bxd5 Rxc1 19.Rxc1 Rd8 20.Bc4 Bb2= when White must say goodbye to the extra pawn and the position is approximately equal) 15...Kxf7 16.axb4 Rc4 17.b5 Rb4© Black wins back one pawn and retains full compensation thanks to the bishop pair and active rook on b4. 14...Bxc3† 15.Rxc3 Nxa2 16.Rd3 Rad8= Black has comfortable equality. A223) 7.e4

230

With this move White avoids accelerating Black’s development and opening the game for the bishops, as occurred in the previous variation. Instead, he seizes the centre and prepares to gain more space with e4-e5. 7...0-0 Black evacuates the king from the centre in order to avoid an annoying check on b5, as would have occurred if he recaptured on d5 prematurely. We will consider A2231) 8.Bc4?!, A2232) 8.Nf3 and A2233) 8.e5. An obscure, but nonetheless interesting, alternative is: 8.h4!? I only found two examples of this move, but it deserves attention. 8...cxd5 9.e5 Bg7 10.h5 Nc6

231

11.h6! Seizing space and playing against Black’s bishop is more important than opening the h-file. 11.hxg6 was played in Hillarp Persson – Andreasen, Denmark 2016, when 11...fxg6!N would have activated Black’s rook, which is more important than keeping his pawn structure tidy. A logical continuation would be 12.Qd2 Bf5 13.Be2 Rc8 14.Rd1 Qd7 15.Bf3 Rcd8 16.Bxd5† e6 17.Bxc6 Qxc6 18.Nf3 Bg4„ when Black is at least equal. 11...Bh8 12.Qd2 f6 13.f4 Bf5 14.Nf3 We have been following Sammalvuo – Kantans, Panevezys 2017. White’s strategic goal is to keep the bishop on h8 locked out of the game, but it will not be easy for him to keep control as there are many pieces on the board and it is not clear where his king should go. Black should have strived to undermine the enemy centre with:

232

14...e6N 15.Be2 g5! 16.g3 Qb6 17.Rd1 Rae8÷ Reaching an unclear position in which Black’s chances do not seem worse at all. A2231) 8.Bc4?! This seems like a normal developing move, but it gives Black the chance to fight for the advantage. 8...cxd5 9.exd5!? This was the choice of Levon Aronian so we should pay attention to it – even though he was unsuccessful. 9.Nxd5 Bg7 enables Black to regain the pawn with ease: 10.Ne2 Nc6 11.0-0 (11.e5?! proved an overoptimistic choice after 11...Be6 12.Ne3 Bxc4 13.Nxc4 Qd5 14.Ne3 Qa5† 15.Qd2 Qxd2† 16.Kxd2 Rfd8 17.f4 Nxd4µ when White’s exposed king was a problem in B. Szabo – Lei Tingjie, Graz 2016) 11...e6 12.Ndc3 Nxd4 13.Nxd4 Bxd4

14.Qd3 Be5 15.Rad1 Qc7 16.g3 a6 17.Bb3 b5³ Struik – Siebrecht, Dieren 2007. Black’s chances are higher due to the two bishops; meanwhile Black has no weaknesses and it is not clear what White should do. 9...Qb6 10.Nge2 Qxb2 11.Rb1 Qa3 12.0-0 Nd7

233

White’s strategic goal seems to be to use the doubled pawns to smother Black in the centre, but Black has enough resources and his superior structure should tilt the balance in his favour. 13.Qd2 Jumping the knight to the centre also does not bring any dividends: 13.Ne4 Nb6 14.Rb3 Qa4 15.Bb5 Qa5 16.d6 Be6 17.Rb2 Now in Aaron – Ovetchkin, Mumbai 2014, Black could have simply grabbed the open file: 17...Rfc8N 18.Re1 Bd5µ Black’s pieces are much better placed, and White’s position is devoid of harmony. 13...Nb6 14.Bb3 Qa5 15.Ng3 Bd7 16.Nge4 We have been following Aronian – Carlsen, Moscow (blitz) 2009. I think the most accurate continuation would have been: 16...Bg7N 17.d6 exd6 18.Nxd6 Bc6³ The bishop pair and the weakness of the d4-pawn promise Black a great game, whereas White must suffer and play for a draw. A2232) 8.Nf3

234

This logical developing move has been played at a high level. 8...cxd5 White must make an important decision: to eat or not to eat the pawn on d5 – and in the case of the former, will it be his knight or pawn that does the dining? 9.exd5!? This ambitious choice reminds us of Aronian’s play in the previous variation; although once again, Black is well placed to deal with White’s plan. White can also decline the pawn and gain space: 9.e5 Bg7 In this case, the drawback of White’s 8th move is immediately felt, as White has to reckon on the unpleasant possibility of ...Bg4. 10.Qd2 White tried the aggressive 10.h4 in Turner – Navara, Sunningdale 2007, when 10...Bg4!N would have been best. For instance, 11.Be2 Nc6 12.Qd2 f6 13.exf6 exf6 14.0-0 Qd6³ and Black has two bishops, while the h2-h4 move has only weakened White’s kingside. White also gets nowhere with: 10.Qb3 Nc6 11.Rd1 (11.Bb5?! Bg4µ produces the unpleasant threat of ...Bxf3 when the d4-pawn will be vulnerable) In Nadirjanova – Abdusattorova, Tashkent 2012, the simple 11...e6N 12.Be2 f6 13.exf6 Rxf6 14.0-0 Rf4µ would have left Black with a strong pair of bishops and a weak pawn on d4 to attack. The text move was Nakamura’s choice but his play was rather unconvincing. 10...Nc6 11.Bb5?! Bg4 12.Ng1 f6 13.h3 We can improve on Black’s play from Nakamura – Kasparov, Saint Louis (rapid) 2017, by means of:

235

13...Bf5!N The bishop is more active here than on e6, where Kasparov placed it. 14.exf6 Bxf6 15.Nge2 Qd6 16.0-0 a6 17.Bxc6 bxc6 18.Na4 e5³ Black opens the game for his bishops and White must battle for equality. White’s objectively best move may well be: 9.Nxd5 This is unambitious and not difficult for Black to meet, but White should probably admit that he has failed to gain an opening advantage and play for equality. 9...Bg7 White has little chance of keeping his extra pawn, as d4 is too vulnerable. 10.Ne3 Qa5† 11.Qd2 Qxd2† 12.Kxd2 Rd8 13.Rd1 White returns his extra pawn, realizing that there is no advantage in their position. 13.Bd3?! proved an unfortunate choice after 13...Nc6! 14.d5 Nb4 15.Bc4 e6 16.Ke2 exd5 17.Nxd5 Be6 18.Rhd1 Rac8 19.Rac1 Re8µ when White was bound hand and foot in Johansen – Kengis, Adelaide 1990. 13...Bxd4 14.Nxd4 Rxd4†= The players agreed a draw here in Rosenberger – Rupprecht, Bavaria 2003. 9...Bg4 10.Be2 White achieves nothing special by overprotecting the d5-pawn: 10.Bc4 Nd7 11.0-0 (11.Qd2? Bxf3 12.gxf3 Nb6µ left White with a terrible pawn structure in Vasiukov – Dzindzichashvili, Tbilisi 1973, with his extra pawn not being felt in any way) 11...Bxf3 12.Qxf3 Bxd4 13.Qe4 Bc5 14.g3 Nf6 15.Qd3 a6 16.Rac1 Rc8³ Lauer – Meier, corr. 2013. The d5-pawn, which is isolated, blocks the bishop on c4, 236

and only Black can realistically play for a win. 10...Qb6 11.Qd2 Bxf3 12.Bxf3 Bxd4 An equally strong alternative is 12...Qxd4!?N 13.Qxd4 Bxd4 14.0-0 Nd7 15.Rad1 Bxc3 16.bxc3 Ne5 17.Be2 Rfc8ƒ with some initiative in the endgame. 13.0-0 Nd7 14.Rfd1 Bg7 15.d6 exd6 16.Qxd6 We have been following Zielinski – Piesik, Katowice 2018. Here it looks natural to play:

16...Ne5!N 17.Qxb6 Nxf3† 18.gxf3 axb6 19.Rd6 Ra6 20.Rad1 b5ƒ Black has the upper hand thanks to his active bishop, and can press in the endgame without any real risk. A2233) 8.e5 Bg7

237

White captures space in the centre and limits Black’s dark-squared bishop. 9.Bc4 This has been almost a universal choice, actively developing the bishop and defending the extra pawn on d5. 9.h4 cxd5 takes us back to the 8.h4!? line in the notes on page 125. 9.dxc6? Nxc6 only speeds up Black’s development and White quickly collapsed after 10.Nge2 Qb6 11.Qb3 Nb4 12.Na4 Qa5 13.Nec3 Be6 14.Bc4 Bxc4 15.Qxc4 Nc2† 16.Ke2 b5 17.Qd3 Nxa1–+ in Yuffa – Artemiev, Chelyabinsk 2018. Less active continuations do not pose any difficulties for Black. For instance, 9.Bd3 cxd5 10.Nge2 Nc6 11.0-0 occurred in Naotunna – Ozer, Sochi 2016, when a good continuation would have been: 11...e6N 12.f4 Bd7 13.Kh1 f6 14.Rc1 Rf7³ It is not clear how White should improve his position, and the advantage of the two bishops can gradually take effect. 9...b5! Black takes the opportunity to expand on the queenside with tempo. 10.Bb3 White decides to keep the bishop on the a2-g8 diagonal, where it keeps control over d5. On the other hand, the bishop becomes a target to be rammed by Black’s a-pawn. 10.Bd3?! This is an unfortunate choice, as White leaves the d4-pawn undefended. 238

10...b4 11.Nce2 In the event of 11.Na4 Qxd5 12.Nf3 Bg4 White must lose a tempo to defend the d4-pawn. 13.Be2 Nd7³ Black can follow up with ...c5 to break down White’s centre. 11...Qxd5 12.Qc2? White was losing a pawn whatever he did, but it would have been better to stop the black queen from returning to d5 after destroying the pawn on g2. The lesser evil was 12.Nf3 Bg4 13.Qc2 Bxf3 14.gxf3 Qxf3 15.Be4 Qh3³, but even here White has problems as his king will not feel safe. 12...Qxg2 13.Ng3 13.Be4 gives nothing due to 13...Bf5! 14.Bxg2 Bxc2 15.Rc1 Bd3µ with two bishops and an extra pawn. 13...Qd5 14.N1e2 Ba6 15.Be4 Qc4 16.Qd2 Rd8 17.Rc1 In Y. Vovk – Astaneh Lopez, London 2012, Black could have safely played:

239

17...Qxa2!N 18.Qxb4 Nd7 White is in trouble, for instance: 19.Rxc6 Bxe2 20.Nxe2 Rab8 21.Qc3 Qa1† 22.Qc1 Qxb2µ Black has the more active pieces as well as an extra pawn, so White will have to suffer in a dismal position. It is more reasonable to take an additional step back: 10.Be2 b4 11.Na4 Qxd5 12.Nf3 12.Bf3?! looks tempting but after 12...Qa5 13.b3 Rd8 it turns out that the d4- and e5-pawns are too weak. For instance, 14.Rc1 Qc7 15.Ne2 Bxe5µ and White’s centre has been destroyed. This position occurred in Torma – Nguyen, Budapest 2017. I think Black should have played:

240

12...Nd7N 13.0-0 13.Qb3 is well met by 13...Qxb3 14.axb3 c5! 15.Nxc5 Nxc5 16.dxc5 a5 17.c6 Be6÷ when Black is not worse. The pawns on e5, c6 and b3 are all weak, and Black should be able to pick up one of them to restore material equality – and with two bishops, he may be able to strive for the advantage. 13...c5! 14.dxc5 An interesting alternative is 14.Rc1!? cxd4 15.Bc4 Qa5 16.e6 Ne5 17.Nxe5 Qxe5 18.exf7† Kh8÷ when the pawn on f7 is annoying, but Black has two bishops and a strong passed pawn of his own. 14...Bb7 15.a3 b3 16.Qxd5 Bxd5 17.Rfd1 Bc6 18.Nc3 Nxc5 19.Nd5 Bxd5 20.Rxd5 Rac8ƒ Black has the easier play, since the pawn on b3 hampers White’s play on the queenside and the b2pawn may become an object of attack in the future.

10...a5! To paraphrase a well-known expression: after saying “b”, then you should say “a”! Black follows up the previous pawn advance by continuing his offensive action on the queenside. 11.a4 White will try to stop the advancement of the pawns, but his bishop becomes less mobile because of the a4-pawn. 11.h4!?N has never been played but one should always consider such ideas in positions with a fianchetto. I suggest 11...c5! when a sensible continuation is: 12.Nge2 a4 13.Bc2 a3 14.bxa3 Rxa3 15.0-0 Ba6 16.Ne4 cxd4 17.Qxd4 b4 18.Rfe1 Bxe2 19.Rxe2 Nc6 20.Qd1 Nxe5= With a comfortable position for Black. 11.Nf3N allows 11...a4 12.Bc2 cxd5 when White’s position is already starting to become dangerous, for instance:

241

13.Bd3 (13.Nxb5? is unplayable due to 13...Qa5† 14.Nc3 a3 15.Qd2 axb2 16.Rb1 Bg4–+; and 13.0-0 is well met by 13...Bg4 14.h3 Bxf3 15.Qxf3 Nc6 16.Rad1 e6³ when the d4-pawn is an eternal weakness, while Black’s queenside pawns give him a useful space advantage) 13...b4 14.Nxa4 Bg4 15.Nc5 Nc6 16.Be2 Qb6 17.0-0 Bxf3 18.Bxf3 Nxe5 19.dxe5 Qxc5 20.Qxd5 Qxd5 21.Bxd5 Rad8 22.Bb3 Bxe5ƒ Black should be able to pick up the b2-pawn, after which he will play for a win with zero risk. To interesting and beneficial complications leads: 11.a3 c5! Black has already sacrificed one pawn, and now he leaves two move en prise! His primary goal is to destroy the defence of e5 and activate his pieces, regardless of material losses. 12.Nxb5 12.Nf3 should be met by 12...Bg4 13.Nxb5 Nd7 14.a4 cxd4 15.h3 Bxf3 16.Qxf3 Nxe5„ when Black’s chances are not worse. Later, he may regroup with ...Ne5-d7-c5 followed by ...e6. I also considered 12.dxc5 Nd7! 13.c6 Nxe5 14.Nge2 e6 15.Nd4 Ba6„ when we reach a strange situation: White’s pawns look impressive but Black has wonderful counterplay against the white centre, and I would take his position if given the choice. 12...a4! 13.Bc4 13.Bxa4? Qxd5 gives White big problems. 13...Ba6 14.Ne2 Qa5† 15.Nec3 Nd7 16.f4 Nb6 17.Be2

242

17...Bxb5 18.Bxb5 Nxd5 19.Be2 Nxf4 20.0-0 Nxe2† 21.Nxe2 cxd4 Having been completely outplayed in the complications, White resigned in Simon – Scherbakov, corr. 2015. In a normal game White would surely fight on for a while, but in a correspondence game the decision to give up is understandable. 11...b4 12.Nce2 Natural and best. It may seem tempting for White to send the knight to the c5-outpost, but that would be a mistake: 12.Ne4? cxd5 13.Nc5 Nd7! The tactical point is seen after: 14.Bxd5? 14.Nd3 is preferable, though after 14...e6 15.Nf3 Bh6 16.0-0 Ba6µ Black’s advantage is significant. I also checked 14.Nxd7 Qxd7 15.Ne2 Ba6 16.0-0 Rac8µ when Black is much better because of the two bishops, and ...f6 will put further pressure on White.

243

14...Nxc5! 15.Bxa8 15.dxc5 is met by 15...Ba6 when White’s king will not feel safe. 15...Ba6 16.dxc5 Qxa8 17.Ne2 17.f3 Rd8 also gives Black a decisive attack, with ...Rd5 or ...Qd5 likely to follow. 17...Qxg2 18.Rf1 Bxe5–+ Black’s bishops will tear White’s position to shreds. 12...cxd5 13.h4 White does not intend to castle as his king feels safe enough in the centre. Instead, he embarks on an immediate attack, aiming to use the rook on h1. Calm development leads nowhere: 13.Nf3 Bg4 14.Qd3 e6 15.0-0 Nc6 16.h3 Bxf3 17.Qxf3 f6³ Black is better because the e5- and d4-pawns are weak. I also checked 13.Qd2 Ba6 14.Nf3 Nc6 15.0-0 Qd7 16.Rfc1 Nd8 17.Rc5 Bb7 18.Rac1 Ne6 19.R5c2 Rac8ƒ when Black’s bishops can show their strength, and ...f6 may come at any moment. Here I propose an improvement for Black:

244

13...h5!N In the game Black ignored his opponent’s plan with 13...Ba6, which is not bad either. After 14.h5 Nc6 15.hxg6 fxg6 in Pantsulaia – Gopal, Khanty-Mansiysk (ol) 2010, White should have caught up on development with 16.Rc1N 16...Rc8 17.f4, with approximately equal play. The text move may not guarantee Black an advantage, but it feels more logical to prevent White from opening the h-file. Here are some lines I analysed: 14.Qd2 14.Nf4 is not dangerous in view of 14...Ra7! 15.Nge2 (15.Nxd5? loses to 15...Rd7–+) 15...e6 16.Rc1 Ba6 17.Bc2 Rc7 with an initiative for Black. 14...Nc6 15.Rc1 Qb6 16.Nf3 Once again, White should avoid getting greedy: 16.Bxd5?! Bb7 17.Nf3 Rfd8 18.Rc5 e6 19.Rb5 Qa7 20.Be4 Nxd4 21.Rxb7 Nxf3† 22.Bxf3 Rxd2 23.Rxa7 Rxe2† 24.Bxe2 Rxa7 25.f4 Rc7 26.Kd2 f6 27.exf6 Bxf6ƒ White will lose a pawn, although the opposite-coloured bishops offer him reasonable drawing chances.

245

16...Rd8 17.Qf4 Bf5 18.Bc2 Bg4÷ Black is at least equal, and White needs to reckon with ideas such as ...f6 or simply ...Bxf3. B) 4.Nf3 Bg7 5.Bg5 Finally we come to the version with the knight on f3, which can also arise via 4.Bg5 Bg7 5.Nf3, as noted earlier.

5...Ne4! This continuation was used for the first time at a high level by Spielmann in 1934, and was subsequently employed by Keres, Botvinnik, Korchnoi, Uhlmann and others. Black counterattacks the 246

bishop on g5 and essentially gives White a three-way choice: he can retreat the bishop with B1) 6.Bf4 or B2) 6.Bh4, or allow it to be traded by playing B3) 6.cxd5. 6.Nxd5?? blunders a piece away: 6...Nxg5 7.Nxg5 e6–+ Another feeble attempt is: 6.e3? Nxg5 7.Nxg5 e5 8.Nf3 8.h4 exd4 9.exd4 occurred in Moiseenko – Novik, St Petersburg 2013, and here I recommend putting pressure on the d4-point: 9...Nc6!N 10.Qe2† Kf8 11.Nxd5 Nxd4 12.Qd3 Qe8† 13.Be2 Qxe2† 14.Qxe2 Nxe2 15.Kxe2 c6 16.Nc3 h6 17.Nge4 Ke7µ Black’s bishops are clearly stronger than the knight pair. 8...exd4 9.exd4 9.Nxd4 c5 10.Nf3 d4 11.exd4 cxd4 12.Nd5 0-0 13.Be2 b5! 14.0-0 bxc4 15.Bxc4 Be6 16.Qb3 Nc6µ was unpleasant for White in Carmo – Machycek, corr. 2013, due to the threat of ...Na5, plus the two powerful bishops of course. 9...0-0 10.cxd5 c6 11.dxc6 Nxc6 12.d5 We have been following Besir – Santhosh, corr. 2007. White can hardly hope to retain his extra pawn when he is so far behind in development, and Black could have kept some advantage with:

12...Nb4!N 13.Be2 Bxc3† 14.bxc3 Nxd5 15.Qd4 Re8 16.Rd1 Be6 17.0-0 Qc7µ The c3-pawn is weak and needs protection, and ...Nf4 could be troublesome. 6.Qc1?! is a strange move and 6...c5! strikes hard at the centre. 7.Bh6 (White’s position quickly turned to ruins after 7.Nxd5? Nc6 8.Qf4 Qa5† 9.b4 Nxb4–+ in A. Hernandez – So, Las Vegas 2015)

247

7...Bxd4! 8.Nxd4 cxd4 9.Nxd5 Nc6 10.f3 Qa5† 11.Bd2 Nxd2 12.Qxd2 Qxd2† 13.Kxd2 0-0 14.e3 Rd8ƒ Mastalerz – Troff, Cappelle-la-Grande 2016. Despite the fact that White has not lost any material yet, his position is clearly worse, as his king is unsafe and his knight will soon be driven away from d5. 6.h4!? is interesting, but not objectively such a great move in my opinion. After 6...Nxg5 7.hxg5 it is true that White has opened the h-file, but the Grünfeld bishop is an excellent defender. 7...dxc4 8.Qa4† (8.e3 occurred in Dragun – Sieciechowicz, Warsaw 2012, when 8...Be6N 9.Be2 c5 10.dxc5 Nd7 11.c6 bxc6 12.Nd4 Bd5³ would have left White with insufficient compensation for the pawn, and ...e5 is an unpleasant threat) 8...c6 9.Qxc4 Qb6 10.0-0-0 Be6 11.Qa4 This was Yuffa – Timofeev, Sochi 2018, and here I can recommend the bold continuation:

11...0-0!N 12.e3 Qc7 13.Qc2 Rc8 14.Kb1 b5ƒ Black has good prospects on the queenside while the 248

attack along the h-file is not so terrible: even if White destroys the h7-pawn, the black king will feel relatively safe thanks to the bishop on g7. B1) 6.Bf4 This move is somewhat less popular than the other bishop retreat covered in variation B2 below, but there have been plenty of strong players who have favoured it, valuing the increased control over the e5-square. Korchnoi was the first to play this way, against Uhlmann in 1960 in Buenos Aires. It was subsequently taken up by Simagin, Gligoric, Taimanov and Bronstein, as well as numerous modern grandmasters including Carlsen, Kramnik, So, Gelfand, Eljanov, Karjakin and others. 6...Nxc3 7.bxc3 c5! Black fights for the centre in the characteristic fashion for this opening. Unlike variation B2 below, I cannot recommend 7...dxc4 here. The difference is felt after 8.e3 b5? (8...Be6 is correct although White can still fight for the initiative) 9.a4 c6 10.Qb1! a6 11.axb5 cxb5 12.Qe4+– when Black loses a piece thanks to the bishop on f4. 8.e3 Approximately the same type of position is reached after: 8.cxd5 Qxd5 9.e3 cxd4 10.cxd4 Nc6 11.Qd2 11.Be2 Qa5† 12.Qd2 Qxd2† 13.Kxd2 0-0 leads straight to our main line below. 11.Qa4?! Bd7 12.Bc4 was an inferior try in Fieandt – T. Ernst, Jyvaskyla 2001, when 12...Nxd4N 13.Bxd5 Nxf3† 14.Bxf3 Bxa4 15.Rb1 Rc8ƒ would have yielded some initiative for Black. Finally, 11.Rc1 was seen in Galacki – Kusmierek, corr. 2014, when 11...0-0N 12.Qd2 Bf5 13.Be2 Qa5 14.0-0 Qxd2 15.Nxd2 Rfd8= would have been the simplest route to equality.

249

11...0-0 12.Rc1 Bf5 13.Rc5 13.Be2 transposes to the 11.Rc1 line above. We have no problems after 13.Bc4 Qa5 14.Ke2 when, in Shalamberidze – Evdokimov, Cesme 2015, the natural 14...Qxd2†N 15.Nxd2 e5 16.Bxe5 Bxe5 17.dxe5 Nxe5ƒ would have given Black some initiative, since the d3-square is weakened and Black’s queenside pawn majority may later become a distant passed pawn. 13...Qe4 14.Ng5 Qb1† 15.Rc1 This occurred in Villarreal – Connelly, corr. 2014, and here I like: 15...Qb4N 16.e4 Qxd2† 17.Bxd2 Bd7 18.d5 Nd4 19.Rc7 Rfd8 20.Bd3 e6³ Black’s pieces occupy optimal places, while White’s lack coordination.

250

8...0-0 9.cxd5 This is the usual choice: White wants to develop his other bishop without losing a tempo after ...dxc4. 9.Bd3 dxc4 10.Bxc4 will be covered below in the note on 9.Be2. 9.Qb3?! looks active but is in fact dubious, as the queen may be hit by ...Nc6-a5. 9...cxd4 10.exd4 (White is not helped by 10.cxd4 Qa5† 11.Nd2 dxc4 12.Bxc4 Nd7 13.Rb1 e5 when the players agreed a draw in Wegner – Reinke, Mecklenburg VP 2001, but Black is better after 14.dxe5 Nxe5 15.Bxe5 Bxe5 16.Qb5 Qxb5 17.Rxb5 Bg7³ due to his two bishops in an open position plus a useful pawn majority on the queenside) 10...dxc4 11.Bxc4 Nc6 12.Be2 b6 13.0-0 Bb7 14.Rfe1 Rc8 15.Rac1 In Musitani – Gromov, corr. 2006, it would have been logical to play:

15...e6N 16.h3 Qe7 17.Qa4 a6ƒ Black is somewhat better, since White has more pawn islands and the c3-d4 pawn duo is more of a weakness than a strength. Finally, a contest between elite players continued: 9.Be2 dxc4 10.Bxc4 Nd7 11.0-0 Nb6 12.Bb3 (12.Bd3?! is worse in view of 12...Nd5 13.Rc1 Nxf4 14.exf4 Qd6³) 12...c4 13.Bc2 Nd5 14.Qd2 Nxf4 15.exf4 This was Eljanov – Caruana, Shamkir 2016, and here I recommend:

251

15...b5N 16.a4 Bb7 17.Ne5 f6 18.Ng4 b4 19.cxb4 e5 20.fxe5 fxe5 21.d5 Bxd5= Black has the bishop pair but their prospects are somewhat limited, so the chances are roughly equal. 9...Qxd5 10.Be2 White has a reliable centre but he has yet to castle, which gives Black time to provoke a queen exchange and then develop his queenside pieces with comfort. 10.Bd3?! is worse, as 10...Nc6 11.0-0 e5! targets both of White’s bishops. 12.dxe5 Nxe5 13.Be2 Qe6 14.Rc1 occurred in Babychuk – Karakas, corr. 2015, when it would have been best for Black to avoid the knight trade with:

14...Nc6!N 15.Ng5 Qe7 16.Qd6 h6 17.Qxe7 Nxe7 18.Ne4 Nd5 19.Bd6 Rd8 20.Bxc5 Bf5 21.Nd6 252

Nxc3 22.Bf3 Bd3 23.Rfe1 Nxa2ƒ It is obvious that Black is playing to win. Black also has an easy game after: 10.Qb3 Be6 11.Qxd5 11.Rb1?! is even worse. 11...cxd4 12.Nxd4 Qxb3 13.axb3 Bd5 14.Bg5 was seen in Xu Yinglun – Gopal, Doha 2014, when it would have been best for Black to play: 14...Rc8!N 15.c4 Be4 16.Rc1 h6 17.Bh4 e5 18.Nb5 Bf8 19.f3 a6 20.Nd6 Bxd6 21.fxe4 Bb4† 22.Kf2 Nd7µ Despite the fact that White has two bishops, they cannot show all their power. The light-squared bishop is restricted by the c4pawn, while Black also has a better pawn structure: White’s e-pawns are doubled and e4 and b3 are weak, especially with ...Nc5 coming in the near future. 11...Bxd5

12.Bd3 12.Be2 cxd4 13.cxd4 Rc8 14.Kd2 Nd7 15.Rhc1 Nf6 16.Bc7 Bf8 17.Ke1 e6 18.a4 Ne4µ was obviously better for Black in Aseev – Svidler, St Petersburg 1998. Black threatens ...Bb4 and his pieces are perfectly located in the centre. 12...Nc6 13.Rb1 13.e4? Be6 14.d5 does not work due to 14...c4! 15.Bxc4 Bxc3† 16.Ke2 Bxa1 17.Rxa1 Na5 18.Bd3 Bg4µ. 13...b6 14.Ke2 cxd4 15.cxd4 Bxa2 16.Rb2 Bd5³ White had to fight for a draw a pawn down in Koziel – Prado, corr. 2009. 10...cxd4 11.cxd4 According to the rules of a positional game, White captures towards the centre and avoids creating an extra pawn island. 11.exd4?! is illogical. 11...Nc6 12.0-0 b6 13.Qd2 e5 14.Nxe5 Nxe5 15.Bxe5 Bxe5 16.Bf3 was seen in 253

Chojnacki – Krainski, Poznan 2006, and here I can recommend:

16...Qd6N 17.Bxa8 Bxh2† 18.Kh1 Ba6 19.Be4 Bf4 20.Qc2 Bxf1 21.Rxf1 Bg5³ Despite the oppositecoloured bishops and level material, White faces an unpleasant defence. Black has a better pawn structure and a safer king, and the presence of heavy pieces means that White is a long way from reaching the kind of endgame that would guarantee him half a point.

11...Qa5†! Trading queens will help Black. 12.Qd2 White cannot afford to avoid the exchange, as 12.Nd2? e5! gives him big problems in the centre: 254

13.Bxe5 (no better is 13.dxe5 Nc6 14.Qb3 Rd8 15.Rd1 as in Zakharevich – Polovodin, Podolsk 1992, when 15...Be6!N 16.Qxb7 Rac8 17.Bg5 Rxd2 18.Rxd2 Nxe5 19.Be7 Nc6–+ would have put White in a hopeless position, with his king stuck in the centre and the a5-e1 diagonal ruining his life) 13...Bxe5 14.dxe5 Rd8 15.Qc2 Bf5 16.Qb2 In Sarana – Chirila, Internet 2019, Black missed the strongest continuation:

16...Na6!N 17.Rd1 Rac8 18.e4 Bxe4 19.e6 Bxg2 20.exf7† Kxf7 21.Rg1 Bd5–+ White’s position is terrible: his king is stuck in the centre and his knight remains in a nasty pin, so Black should win. 12...Qxd2† 13.Kxd2 13.Nxd2?! Nc6 14.0-0 e5! gives Black easy play: 15.dxe5 Nxe5 16.Rab1 b6

255

17.Nf3 (17.Rfc1 is pointless: 17...Bf5 18.e4 Be6 19.Nc4 Nc6 20.a4 Nd4 21.Bf1 Rfd8 22.f3 Rac8 23.Kf2 Rc5³ occurred in Holm Pedersen – F. Olafsson, Varna 1962, when Black was taking over the cfile and his knight was dominating in the centre) 17...Nc6 18.Rbc1 Bb7 19.Rfd1 Rfd8 20.Bg5 Rxd1† 21.Rxd1 h6 22.Bh4 Nb4 23.Rd7 Bc6 24.Rc7 Be8 25.Nd4 Be5 26.Re7 Bd6 27.Rb7 Nxa2ƒ White needs to worry about equality. 13...Nc6 Black threatens ...Rd8 followed by ...e5. Depending on how White plays, he may even carry out this break as a temporary pawn sacrifice. Alternatively, he can continue mobilizing his queenside pieces. 14.Rhc1 White intends to create pressure along the b- and c-files, essentially asking Black the question of how he will continue to develop his queenside. The following alternatives are not scary: After 14.Bb5 Be6 Black need not fear the damage to his pawn structure, as the two bishops will more than make up for it: 15.Bxc6 bxc6 16.Rhc1 Rac8 17.Rc5 Rfd8 18.a4 f6 19.Rb1 Bf5 20.Rb7 e5 21.Bg3 Bf8 22.Ra5 c5–+ Gallego Alcaraz – Espinosa Veloz, Santa Clara 2017. Despite the fact that White’s rooks are very active, his position is hopeless, since his centre is collapsing and his king falls under an attack. Korchnoi opted for the mysterious 14.Rhb1 b6 before immediately moving his rook again. 15.Rc1 Bb7 16.Rc3 Rfd8 17.Ke1 This was Korchnoi – Uhlmann, Buenos Aires 1960, and now I like the idea of transferring the bishop to the f8-a3 diagonal:

17...Bf8!?N 18.Kf1 Rdc8 19.Rac1 e6 20.e4 Nb4!? This is a way to keep the position complex and continue the fight. (20...Bb4 21.Rb3 Bf8 gives White nothing better than repeating with 22.Rbc3=) 21.Rxc8 Rxc8 22.Rxc8 Bxc8 23.a4 f6= The position remains balanced but plenty of play remains.

256

14.Rhd1 occurred in Gorczyca – Bednarska, Ustron 2001, when thematic development with 14...Rd8N 15.Ke1 Bf5 16.a3 Rac8= would have given Black a comfortable position, similar to the main line below. 14.Rac1?! Rd8 15.Rc5 is somewhat risky. In Krainski – Toma, Rowy 2002, Black missed a nice idea:

15...e5!N 16.Nxe5 Nxe5 17.Bxe5 b6 18.Bxg7 (18.Bc7 bxc5 19.Bxd8 cxd4 20.Bc7 dxe3† 21.fxe3 Bb7ƒ gives Black the upper hand due to her better pawn structure) 18...bxc5 19.Bf6 Rd7 20.Rc1 cxd4 21.Bg4 dxe3† 22.Kxe3 Rd6 23.Be7 Rd5 24.Bf3 Bb7 25.Bxd5 Bxd5 26.a3 Bxg2³ Black emerges with an extra pawn, although White has decent drawing chances due to the opposite-coloured bishops. 14.Rab1 This move has been chosen by Portisch and Korchnoi, but the rook could be hit by ...Bf5. 14...Rd8 Black starts with the standard developing move, threatening ...e5. 15.Bd3 15.Bb5 Bf5 16.Bxc6 occurred in Portisch – Timman, Tilburg 1980, when Black took on b1 and agreed a draw after further simplifications. Instead he could have fought for the advantage with 16...bxc6!N, not fearing the damage to the pawn structure, as the bishop pair more than makes up for it. It is also worth noting that the isolated c-pawn can quickly be mobilized with ...c5 in order to attack White’s pawn centre. For instance: 17.Rbc1 Rac8 18.Rhd1 c5 19.Ke1 Be4 20.Rxc5 Rxc5 21.dxc5 Bc3† 22.Ke2 Rxd1 23.Kxd1 Bb4 24.Nd4 Bxc5³ Black has the classic advantage of two bishops in an open position, and White must fight for a draw. 15...e5! 16.Nxe5 Nxe5 17.Bxe5 Bxe5 18.dxe5 In Pavlidis – Zak, Herceg Novi 2005, Black’s easiest option would have been to head for a rook endgame with:

257

18...Bf5N 19.Rb3 Rd5 20.Ke2 Bxd3† 21.Rxd3 Rxe5= The endgame is equal and a draw is likely.

14...Rd8 Black threatens ...e5, so White is forced to spend a tempo safeguarding his king. 15.Ke1 White can block the file with 15.Bd3 but then 15...e5! leads to exactly the same type of equal endgame that Black could have reached in Pavlidis – Zak, as noted just above: 16.Nxe5 Bxe5 17.Bxe5 Nxe5 18.dxe5 Bf5 19.Rc3 Rd5 20.Ke2 Bxd3† 21.Rxd3 Rxe5=

258

15.Bb5 is not dangerous. In Makhno – Myshkovskiy, St Petersburg 2017, Black could have solved all his problems with a thematic idea: 15...e5!N 16.Bxc6 exf4 17.Be4 fxe3† 18.fxe3 Bf8 19.Rab1 Rb8 20.Ke2 Bd6 Black is comfortably equal and can try to play for a win with the help of the bishop pair. 15.Rc5 prevents ...e5 but, aside from that one purpose, the rook is not well placed. Black reacted well with 15...Be6 16.a4 Bd5 17.Ke1 e6 18.Rcc1 Rdc8 19.Bd6 Na5 20.Bb5 Nb3 21.Rxc8† Rxc8 22.Rb1 Na5 23.Bb4 Be4 24.Rd1 Nb3ƒ in Dreev – Artemiev, Moscow 2014. In retrospect, White’s 15th move wasted time and Black was able to take over the c-file, causing his opponent some problems.

15...Bf5! Preventing Rab1. From here the bishop can go to e4, where it will influence many parts of the board. The game now enters a phase of slow manoeuvring. 16.a3 Preventing ...Nb4-d5 is always useful. Here are some other examples: Black gets an excellent position after 16.Nh4?! Be4 17.Bf3 f5 18.Bxe4 fxe4 19.Rab1 Rd7 20.Rc5 as in Bator – Rytshagov, Stockholm 1990. Here Black missed a powerful idea:

259

20...a5!N Intending to mobilize the knight. 21.Kf1 (21.a3?! prevents ...Nb4 but leads to even bigger problems: 21...Bf8! 22.Rcb5 e5 23.Bxe5 Bxa3 24.f3 Be7 25.g3 a4–+ and the a-pawn decides the outcome of the battle) 21...Nb4 22.f3 b6 23.Rc4 b5 24.Rc5 Nd3 25.Rcxb5 Nxf4 26.exf4 Rxd4µ White’s knight would be inferior to Black’s bishop in an open position at the best of times, and this is all the more true when it stands on such a square as h4. Doubling the rooks along the c-file yields nothing special for White. 16.Rc4 Be4 17.Rac1 occurred in Spassov – Tseshkovsky, Gmunden 2007, and here I like the idea of:

17...Bd5!?N 18.R4c2 Bf8 Improving the dark-squared bishop, which is currently being obstructed by the d4-pawn. A sample continuation is: 19.Ne5!? Bxg2 20.f3 Nxe5 21.Bxe5 Bh3 22.Rc7 Rd7 23.a4 Rad8 24.Rb1 b6 25.Rbc1 f6© White has compensation for the sacrificed pawn due to the activity of his 260

pieces, but it is sufficient only for equality. Many grandmasters have played: 16.h3 h5 17.Nh4 Bc8 17...Be4!?N is an interesting option which has not yet been applied in practice. An important feature of this move is that it avoids a quick repetition, as can happen after the main move. Play may continue 18.Bf3 Bxf3 19.Nxf3 Nb4 20.Kd2 Nd5 21.Bg3 b5= when Black is in perfect order: he can transfer his knight via b6-c4 or fight for the c-file with ...Rac8. 18.Bc4 18.Nf3 Bf5 19.Nh4 repeats the position – but if you want to continue the game, you still have the possibility of 19...Be4!? as I have indicated. With the text move, White continues the struggle and strives for the advantage, but Black can instigate not unprofitable complications...

18...Nxd4! 19.Bc7 19.exd4 Rxd4 20.Nxg6 Re4† 21.Be3 Bxa1 22.Bd5 Rxe3† 23.fxe3 Bf6 can only be dangerous for White. 19...Rd7 20.exd4 Rxd4 21.Nxg6 Re4† 22.Kf1 Bxa1 23.Bd5 Bf5 24.Bxe4 Bxe4 25.Nxe7† Kf8 26.Rxa1 Bd3† 27.Kg1 Kxe7= The endgame is equal and a draw ensued in Ivanchuk – Caruana, Wijk aan Zee 2015. 16...Rac8 Black brings his last piece into play and it is not clear what White should do next. We will follow another Caruana game for a few more moves. 17.h3 17.Bb5 has been played but the bishop will soon be driven back. 17...Be4 18.Kf1 a6 19.Be2 Na5 261

20.Nd2 Bc6 21.Ke1 occurred in Arencibia Rodriguez – Bruzon Batista, Havana 1999, when Black could have reverted to his thematic central break:

21...e5!?N 22.Bxe5 Bxe5 23.dxe5 Rxd2 24.Kxd2 Nb3† 25.Ke1 Nxa1 26.Rxa1 Bxg2 Black has a comfortable position. Objectively the endgame is equal, but it is not impossible to imagine Black mobilizing his queenside majority while White struggles to advance his split kingside pawns in an effective way. 17...h5 18.Bc4 e6 19.Bf1 It is hard to guess exactly what White was hoping to achieve, but his position remains okay. 19.Bg5!?N leads to more energetic play: 19...f6 20.e4! Na5! 21.Ba2 Rxc1† 22.Rxc1 Bxe4 23.Bxe6† Kh8 24.Ne5! The position has become extremely lively, but the forcing line 24...Re8 25.Bf7 Rf8 26.Bxg6 Bxg2 27.Bd2 Nb3 28.Nf7† Rxf7 29.Bxf7 Nxc1 30.Bxc1 Bxh3 31.Bxh5 Bf8= results in a completely equal position all the same. 19...a6 20.Bg5?! White goes a step too far in provoking the black pawns forward. 20.Be2N would maintain the equilibrium. A possible continuation is 20...Bf6 21.Bg5 Bxg5 22.Nxg5 f6 23.Nf3 Kg7 24.Nd2 e5 25.d5 Rxd5 26.e4 Rxd2 27.Kxd2 Bxe4© when Black has full compensation for the sacrificed exchange. 20...f6 21.Bh4 Here Black could have fought for the advantage with:

262

21...g5!N The game continued 21...Na5?! 22.Nd2 Bc2, which looks strong until you see 23.Bc4!. Perhaps Caruana ‘yawned’ at this resource, and the game was later drawn in Karjakin – Caruana, Skopje 2015. 22.Bg3 h4 23.Bh2 b5 Black’s prospects are improved by the considerable amount of space he has gained. 24.Nd2 24.a4 is an important move to consider, but Black has a strong answer: 24...Nb4! 25.Bc7 Rd7 26.Ba5 Rxc1† 27.Rxc1 Nd3† 28.Bxd3 Bxd3 29.Rc8† Kf7 30.axb5 axb5³ Black has a lasting advantage thanks to the two bishops and the passed b-pawn. 24...e5! 25.dxe5 Nxe5 26.Bxe5 fxe5 27.e4 Bd7³ Black keeps the better chances in the endgame and White will have to defend accurately to reach a draw. B2) 6.Bh4

263

This move first appeared in 1960, but it took another ten years before it was played at a high level, by the Russian international master Orest Averkin, against Krogius. For a long time, this option remained in the shadows, but the Dutch international master Bert Enklaar started playing it regularly in 1972, and little by little it became more popular. Benko, Beliavsky, Ftacnik and Ivkov took it up; and in the 21st century, Carlsen, Nakamura, Ipatov and Wang Yue have all played it. Compared with the previous variation, White’s bishop exerts no influence over the e5-square or the queenside. Instead, the bishop eyes the e7-pawn and prevents it from moving. White generally does not aim for a big advantage and is content with a small positional pull. It tends to appeal to players who like to avoid long, forcing theoretical lines. 6...Nxc3 7.bxc3 dxc4! Using the fact that the bishop is on h4 rather than f4, Black eats a pawn. This move was first played by GM Robatsch against Enklaar in 1972, and it was subsequently used by such players as Kasparov and Vachier-Lagrave, along with most other Grünfeld specialists. It is worth exploring the pawn sacrifice B21) 8.e3 followed by the main line of B22) 8.Qa4†. 8.e4 is ambitious but not dangerous. 8...b5 9.a4 c6 10.Be2 Nd7 11.0-0 0-0 12.Qc2 Nb6 13.Ne5 Bxe5 14.dxe5 Nxa4 15.Bxc4 Nb6 16.Ba2 was seen in Jurka – Jeske, Brno 1986, when 16...a5!N³ would have emphasized Black’s advantage. White’s compensation for the pawn is insufficient, and he will have to battle for salvation. Another questionable sideline is: 8.Qb1?! Dimakiling – Ni Hua, Manila 2007. A logical novelty is:

264

8...c5!N 9.Qb5† This is the critical move to check. White should prefer 9.e3 although 9...cxd4 10.exd4 0-0 11.Bxc4 Qc7³ gives Black an easy position edge, as the pawns on c3 and d4 are weak. 9...Nd7 10.e3 a6 11.Qxc4 b5 12.Qd5 Suddenly Black can launch a decisive attack. 12...Qa5! 13.Qxa8 Qxc3† 14.Ke2 0-0 15.Rd1 Nb6 Despite being a rook up, White is defenceless. For instance: 16.Qc6 Na4 17.Nd2 Bg4† 18.f3 Bh6! 19.Qe4 Bf5 The end is near. 20.Qe5 Nb2 21.Re1 f6 22.Qxe7 cxd4 23.f4 dxe3 24.Qxe3 Nd3 25.Kf3 Bxf4–+ B21) 8.e3 b5 9.a4 c6 It is important to understand this position in order to appreciate why Black’s 7th move was a good choice. 10.Be2 White gets nowhere with 10.axb5 cxb5 11.Ne5 Bb7 12.Rb1 Qd5, since the following tactical sequence is disastrous for him: 13.Bxc4 bxc4 14.Qa4† Bc6 15.Rxb8† Rxb8 16.Nxc6 Qb5–+ Diaconescu – Palmo, Finland 1977. 10.Nd2 a6 11.Be2 is also no problem: after 11...0-0 12.Bf3 Ra7 13.0-0 Qc7 14.Bg3 e5µ White had no real compensation for the pawn in Murdzia – Kasimdzhanov, Germany 2013. 10.Qb1 was tried in a super-GM game but it’s not a great move. 10...a6 11.Be2 (a key point is that 265

11.axb5 cxb5 12.Qe4 can be met by 12...Ra7µ when, due to the placement of the bishop on h4 rather than f4, the b8-knight does not hang and Black remains with an extra pawn) 11...Nd7 12.0-0 Nb6 13.Ne5 Bxe5 14.dxe5 Nxa4 15.Bf3 In Nakamura – Aronian, Saint Louis 2014, the most accurate continuation would have been:

15...Qc7!N 16.Qb4 c5 17.Qb1 Ra7 18.Bg3 Be6µ Any compensation for the sacrificed pawns is not visible, so Black has a big advantage. 10...Nd7 11.Nd2 Nb6 12.Bf3 Now in Valov – Kulaots, Minsk 2015, a strong continuation would have been:

12...Nxa4!N 13.Bxc6† Bd7 14.Bxa8 Qxa8© 266

Black has more than enough compensation for the exchange. B22) 8.Qa4†

The above analysis makes it clear that White should regain the pawn immediately, but the drawback is that he must spend time moving his queen about. 8...c6 The point of this move is to enable the queen to come to a5 next. She will be ideally placed on that square, threatening the c3-pawn. As we will see in the main line below, Black will be able to increase the pressure on that pawn by manoeuvring his knight. 9.Qxc4 Qa5 10.e3 10.Nd2?! prepares to drive the active queen away, but it gives Black the opportunity for 10...e5!. Then 11.Nb3 Qc7 12.e3 0-0 13.Be2 Bf5 14.0-0 Nd7 15.Bf3 occurred in H. Hansen – Borkenhagen, email 2010, when 15...Be6N³ would have favoured Black; he has a free game and the knight on b3 is not well placed. 10.Qb3 seems a strange choice, as after 10...Be6 White cannot take on b7 as the c3-pawn needs protection. Instead he played 11.Qb2 in Semcesen – Y. Vovk, Lvov 2012, when Black should have immediately struck at the centre:

267

11...c5!N 12.e3 cxd4 13.Nxd4 0-0 14.Bd3 Bd5 15.0-0 e5÷ Black is at least equal due to his better pawn structure. Finally, the ambitious 10.e4 Be6 11.Qd3 Nd7 12.Nd2 b5 13.Be2 was played in Ipatov – Timofeev, Plovdiv 2012. I recommend improving with:

13...b4!N 14.Rc1 bxc3 15.Rxc3 c5 16.Ra3 Qb4 17.d5 The key detail is 17...c4! 18.Qxc4 Qxa3 19.dxe6 Nb6 20.Qb5† Kf8 21.0-0 Qd6³ when White’s compensation for the exchange is insufficient. 10...Na6! The knight is heading for c5, taking advantage of the vulnerable c3-pawn and rook on a1, in order to advance to e4 or a4 to put further pressure on c3. White must also reckon with the possibility of ...Be6, 268

harassing the queen, which remains bound by the need to protect the c3-pawn. 11.Qb3 White retreats in order to meet ...Be6 with Qb2. The simplistic 11.Be2?! allows Black to gain too much time against the queen: 11...Be6 12.Qd3 Nc5 13.Qc2 Bf5 14.Qb2 Nd3† 15.Bxd3 Bxd3 16.Qb4 Qxb4 17.cxb4 a5 18.Kd2 Be4µ Marzo Pie – Erenberg, Riga 2018. Black holds all the aces, with two strong bishops and unpleasant pressure on the queenside. 11.Nd2 This is a slight improvement on the above line, but it still gives Black an easy game. 11...Be6 12.Qd3 Nc5 13.Qc2 Bf5

14.Nc4 This seems to be the least of the evils. 14.Qb2?! Nd3† 15.Bxd3 Bxd3 favours Black due to the two bishops. After 16.Nb3 Qh5 17.Bg3 b6 18.f3 0-0 19.Kf2 e5µ the light-squared one cutting through the a6-f1 diagonal was especially strong in Can – Nedev, Porto Carras 2011. 14.Nb3!? was not such a bad move in Harmon Vellotti – Troff, Chicago 2015. Still, after 14...Nxb3N 15.Qxb3 e5 16.Bc4 0-0 17.Be7 b5 18.Be2 Be6 19.Bb4 Bxb3 20.Bxa5 Bd5 21.0-0 Rfe8ƒ Black has some initiative. He threatens ...exd4 and ...Bxd4; the bishop on d5 conveniently protects the weak pawn on c6; and Black may also consider the position plan of ...exd4 followed by ...f5-f4. 14...Qa6 15.Qd1 b5 16.Nd2 Here I found a useful improvement:

269

16...Na4!N For once, 16...Nd3†?! 17.Bxd3 Bxd3 is not so effective. 18.Nb3 Bc4 occurred in Michenka – Tilicheev, Ostrava 2012, when 19.Nd2!N would have equalized. 17.Rc1 0-0 18.Be2 c5 19.0-0 Rfe8³ Black is ready for further undermining of the centre with ...e5, and the invasion of White’s camp by means of ...Nb2-d3 may also be on the cards.

11...Nc5! Black needs to improve the knight before it is too late. The text move virtually forces a queen exchange and White’s centre will be somewhat weakened.

270

11...Be6 12.Qb2 Nc5?! gets nowhere after 13.Rc1! when c3 is securely defended and the b7-pawn will fall. 12.Qb4 Qxb4 13.cxb4 After skipping to the c-file, the pawn traveller returns to its original file and forces the black knight to continue its journey. 13...Ne4 Why does the knight come to e4 when there is no longer a pawn to attack on c3? The knight is actually heading for d6, where it has two roles: it may go to f5 to hunt down White’s bishop, or it can stay controlling the b5-square so that ...a5 will carry real force. 14.Bd3 14.Rc1?! is inaccurate, since the rook is likely to be needed on the a-file once ...a5 is played. 14...Nd6 15.Bg3 Preparing to trade off the knight is the lesser evil for White. (15.Be2?! a5 16.a3 axb4 17.axb4 Bf5 18.0-0 Ra2 19.Rfe1 Rb2µ allowed Black to gobble the b4-pawn in Meduna – Gavrikov, Prague 1988) 15...a5 16.Bxd6 exd6 17.b5 Bd7 18.bxc6 bxc6 19.Bd3 Ke7 20.0-0 Rhb8³ Netusil – Petr, Czech Republic 2015. Black is better because of the two bishops, even though material is level and Black has an extra pawn island. Black’s centralized king may also benefit him in the endgame. 14...Nd6 Black continues with his plan. 15.Rb1!? White gets ready to answer ...a4 with b4-b5. 15.0-0?! Connecting the rooks is not a bad idea, but it does not help White to move his king so far from the centre and queenside in the endgame. 15...a5! Black follows his usual plan and his position is already more promising. 16.a3 Be6 17.Rfc1 After 17.Nd2 f5 18.h3 the rivals agreed to a draw in P. Nguyen – Bernal Varela, corr. 2011, but Black could have kept trying with 18...axb4N 19.axb4 Kd7 20.Bg3 b5³ with ideas of ...Rhb8 and ...Ra4 to make headway on the queenside. 17...Kd7 Black keeps his king well centralized, avoiding White’s mistake from two moves ago. With his rooks connected, he can turn his attention to the queenside. 18.Ne5† White can try to sacrifice a pawn, but in my opinion it is doubtful: 18.b5?! Nxb5! 19.Rab1 Rhc8

271

20.a4 Nd6 21.Rb6 f6³ White has some compensation due to the pressure along the queenside files, but he is still fighting for equality. If White acts calmly and is not in a hurry to create counterplay, then Black captures the a-file. For example, 18.Bg3 Rhc8 19.h3 Ra7! 20.Nd2 axb4 21.axb4 Rca8³ when possession of the open file provides Black with a lasting advantage. 18...Bxe5! Black gives up his bishop, which was blocked by the d4-pawn in any case. White’s bishop pair will not be so relevant, since the dark-squared one is inactive. 19.dxe5 Nf5 Black went on convert his advantage in an opposite-coloured-bishop endgame in Netusil – Konopka, Czech Republic 2004. Some correspondence players have improved on the above game with 15.Ke2!?, keeping the king centralized. Play continues 15...a5 16.a3 Be6 17.Rhc1 Bd5 when an important resource is:

18.b5! Nxb5 19.e4 Be6 20.Rab1 Bd7 21.a4! White plays with a lot of energy, but it is only enough for equality. 21...Nxd4† 22.Nxd4 Bxd4 23.Rxb7 f6 24.f3 Bc8 25.Rb3 Kd7 26.Bf2 Bxf2 27.Kxf2 Ra7= Munoz Moreno – Glazman, corr. 2012. The position is greatly simplified and 28.Rbc3 can be met by 28...Ba6 when Black does not have a shadow of a problem. 15...Be6 16.a4 Bd5 17.Ke2 f6 18.Bg3 In Cvetnic – Krecak, corr. 2009, it would have been good for Black to play:

272

18...e5!N 19.dxe5 fxe5 20.Nxe5 Ne4! 21.Bxe4 Bxe4 22.Rbg1 0-0 23.f3 Bc2 Black has a lot of activity to compensate for the missing pawn. My illustrative line continues: 24.Nd7 24.a5 Rad8 25.Rc1 Ba4© offers Black lasting compensation with the two bishops, and I would prefer his position if given the choice. 24...Rfe8 25.Nc5 Bh6 26.Bf2 a5 27.Rc1

27...Bxa4! 28.Nxa4 axb4 Despite being a knight for a pawn up, White must play with a lot of precision to stop the queenside 273

passers. Here is the rest of my analysis: 29.Nb2 Ra2 30.Rc2 b3 31.Rd2 Bg7 32.Rb1 c5 33.e4 Rc8 34.Be1 b5 35.Nd3 c4 36.Nb4 Rca8 37.Nxa2 bxa2 38.Rbd1 a1=Q 39.Rxa1 Bxa1

40.Rd5 Ra2† 41.Kf1 c3 42.Rxb5 c2 43.Rb8† Kf7 44.Rc8 Be5 45.Bd2 Bxh2 46.Ke2 Bf4 47.Kd3 Ra3† 48.Ke2 Ra2= After all that, the game must end in a draw. B3) 6.cxd5

This move has been played in many more games than the two previous options combined. The 274

Spanish player Rieber Arnal was the first to employ it, against Spielmann in 1934. Later it was taken up by Lasker, Petrosian, Smyslov, Simagin and Taimanov. In modern times, Kasparov, Nakamura, Seirawan, Vitiugov and Wojtaszek have played it. White’s idea is to reach a Carlsbad structure, as usually associated with the Queen’s Gambit Declined, in the hope that Black’s bishop will not be well placed on g7. The disadvantage is that Black gains the advantage of two bishops, which could become more relevant later in the game. A complex positional battle lies ahead, where each side makes its own plans. 6...Nxg5 7.Nxg5 e6 Regaining the pawn and reaching the aforementioned pawn structure. White can either go for some quirky queen play with B31) 8.Qd2 or recentralize his knight with B32) 8.Nf3. It would be dangerous for White to go pawn-grabbing: 8.Qa4†?! c6 9.dxc6 Nxc6 10.Nf3 Bd7 Now White has to spend additional time moving his queen to safety. 11.Qd1 11.0-0-0 is a creative approach but White is merely trading one set of problems for another, as now his king becomes the target: 11...0-0 12.Qa3 b5! 13.Nxb5 Rb8 14.e4 Qb6 15.Nd6 Nb4 16.e5 Bc6 17.Bc4 Rfd8 18.Rhe1 Bxf3 19.gxf3 Bf8µ In Lindsalu – Trubetskoi, corr. 2009, Black had the better pawn structure and good attacking chances. 11...Qb6 Black simply regains the pawn and gets the better game. 12.Qd2 12.e3 did not solve White’s problems after 12...Qxb2 13.Qc1 Qxc1† 14.Rxc1 Ke7 15.Be2 Rac8 16.Kd2 Rhd8 17.Rb1 e5 18.d5 Na5 19.Rhd1 f5 20.h4 Rc5 21.e4 b6µ in Stock – Cross, corr. 2012. Black has pressure along the c-file as well as the two bishops. 12...Nxd4! 13.0-0-0 13.Nxd4 Qxd4 14.Qxd4 Bxd4 15.Rc1 Ke7 16.e3 Bg7µ gave Black a significant endgame advantage in Araz – Savchenko, Riyadh 2017. The text move is a better try but Black has everything figured out:

275

13...Bc6! 14.Nxd4 Rd8 15.e3 e5 16.Qe1 exd4 17.exd4† Kf8 18.d5 Re8 19.Qd2 Bd7 Despite being temporarily a pawn down with an apparently misplaced king, Black soon takes control. 20.Kb1 Rc8 21.Rc1 Bd4 22.Bd3 Bxf2 23.Qh6† Kg8 24.Ne4 Bd4 25.b3 Bg7 26.Rxc8† Bxc8 27.Qc1 Bf5µ White had serious problems in Sitar – Kaan, corr. 2013, due to being weak on the long diagonal, with ...Qd4 an especially nasty threat. In several games White has tried a different knight retreat: 8.Nh3 It may seem tempting to put the knight on f4, but in reality it will be rather poorly placed there. 8...exd5 9.Nf4 c6 10.e3 0-0 11.Be2 11.Bd3 does not change much. In Schlapp – Schnitzspan, Germany 2013, it would have been good to play 11...Re8N 12.0-0 Bf8 13.Qd2 Nd7 14.Rfe1 Nb6 15.Rac1 Bd7³ when White’s counterplay is not visible, and I value Black’s chances due to his bishop pair and potential to develop pressure in the future. 11...a5 12.0-0 We have been following Truskavetsky – Zinchenko, Alushta 2005. Here I propose a thematic knight manoeuvre for this structure:

276

12...Nd7N 13.Qc2 Nb6! Depending on circumstances, the knight may jump to c4, especially after b2-b4, or go via c8 to d6. 14.Rac1 Playing for a minority attack is thematic for this structure, but it does not work here: 14.Rab1 Re8 15.Nd3 (15.b4? leads to even greater difficulties after 15...Nc4 16.Bxc4 dxc4 17.b5 Bf5–+) 15...Bf5µ White’s pieces are disharmonious. 14...Re8 15.Bd3 Bf8 16.Rfe1 Bd6 17.Nfe2 Kg7³ Black’s chances remain higher thanks to the two bishops. B31) 8.Qd2

277

This move is associated with the idea of depriving Black of the right to castle. 8...exd5 9.Qe3† Without this check, White’s previous move would make little sense. Still, it will be easy for Black to castle artificially, while White will have to spend time moving his queen again. 9...Kf8 10.Qf4 At least White gets to threaten mate in one. One-time threats do not pose real problems though, and the queen may be attacked in the future. 10.h4 h6 11.Nf3 Bf6 12.Qf4 transposes to our main line below. 10.Nf3 c6 11.g3 was played in Shabalov – Hellers, New York 1993, and here it looks promising to play:

11...h5!N 12.Bg2 Bh6 13.Qd3 Bf5 14.Qd1 Qb6 15.Qb3 Bg7 16.0-0 Nd7³ Black has a slightly better game with the two bishops. White loses a lot of time with 10.Qd2 Bf6 11.Nf3 Kg7 12.e3 as in Rajkovic – Sahovic, Vrsac 1979. Black can play calmly with:

278

12...c6N 13.Bd3 Be7 14.0-0 Re8 15.Rab1 a5 16.h3 Be6 17.Rfe1 Nd7³ Black’s chances are somewhat higher: a natural plan is to attack on the kingside with ...Bd6, ...g5, ...h5 and so on. At a high level, White has tried: 10.g3 Bf6 11.h4 h6 12.Nf3 Kg7 13.Qf4 13.Bg2 occurred in Roj – Pribyl, Czechoslovakia 1991. Black could have obtained a great game with 13...c6N 14.0-0 Re8 15.Qd2 Bf5 16.Rfe1 Nd7³ when he is ready to advance his kingside pawns and attack. 13...c6 14.Bg2 Na6 15.0-0-0 White’s king also fails to find a safe haven with 15.0-0, as played in Negulescu – Peptan, Romania 1992. Black can already start attacking with 15...g5!N, and after 16.Qd2 gxh4 17.e4 Nc7 18.Rae1 Rg8 19.exd5 Nxd5‚ he will create threats on the g-file in addition to having an extra pawn. We have been following Nakamura – Giri, Monte Carlo 2011. I think Black should play:

279

15...h5!N Preventing White’s g-pawn from advancing, before turning out attention to the queenside. Play could continue: 16.Ne5 Nc7 17.e4 dxe4 18.Bxe4 Ne6 19.Qe3 Qb6µ Black enjoys the usual advantage of two bishops, and the d4-pawn is weak. 10...Bf6 Preventing mate on f7 while preparing an excellent parking space for the king – while also gaining a tempo against the knight. 11.h4 This is the natural choice: White defends his knight and prepares a future attack with h4-h5. Still, Black’s bishop pair remains the dominant trump card. White’s attack is not dangerous as most of his pieces will be unable to support it. The knight on g5 looks active, but it only interferes. The alternative is: 11.Nf3 Kg7 12.e4?! White tries to play energetically but opening the position for the bishops is a bad idea. 12.h4 h5 13.g3 leads back to the 10.g3 line noted above. 12.g4 h6 13.0-0-0 Be6 14.h4 transposes to 14.g4 in the notes to the main line below. 12.e3 is not so bad. However, in Volkov – Aronian, Omsk 1996, Black would still have had the better chances after 12...h5!?N, for instance: 13.h4 c6 14.Bd3 Nd7 15.0-0 Nf8 16.Rac1 a6 17.a3 Bg4 18.Ne5 Be6 19.Na4 Nd7 20.Nxd7 Bxd7 21.Nc5 Bc8 22.g3 g5‚ The bishop pair and kingside attack make a dangerous cocktail. 12...dxe4 13.Nxe4 In Perz – Spiekermann, email 1999, it would have been best to continue: 280

13...Bf5!N 14.0-0-0 Exchanging pieces does not solve White’s problems, as the d-pawn will be weak in the endgame: 14.Nxf6 Qxf6 15.Qe5 (White can’t be too greedy, as 15.Qxc7?! Re8†‚ leads to a dangerous attack) 15...Nc6 16.Qxf6† Kxf6 17.0-0-0 Rhd8µ White must switch to a miserable defence of the d4-pawn. 14...Bxe4 15.Qxe4 Nc6³ The isolated d-pawn is a long-term weakness, and further exchanges could result in endgame difficulties for White.

11...h6! Excellent timing. It would be a mistake to play 11...Kg7?! due to 12.e4! dxe4 13.0-0-0 h6 14.Ngxe4 Bf5 and now in 281

Szczepkowska – Kulon, Poznan 2015, White could have obtained a dangerous attack with: 15.g4!N 15...Bxe4 16.Nxe4‚ 12.Nf3 The solid approach is best. Sometimes White initiates complications with 12.Nxd5!? but he comes off worse after: 12...Bxg5 13.Qe5 Rg8 14.hxg5 Nc6 15.Qe4 Bf5 16.Qf3 Nxd4 17.Qa3† Kg7 18.gxh6† Kh7 19.Ne3 We have been following Adler – Akesson, Vasteras 2016, where White was a pawn up and he eventually prevailed. However, he is behind in development and his pawn structure is weak, and Black could have established some advantage with accurate play:

19...c5!N 20.Rc1 (20.Qxc5? is impossible due to 20...Rc8 21.Qxa7 Qd6 22.Qa3 Rc5 23.g3 Rd8 24.f4 Qb6 25.Kf2 Nc2 when Black has a decisive attack and will also win material) 20...Qg5 21.g3 Rae8 22.Bg2 b6 23.Kf1 Re7 24.Rh4 Rge8³ White’s position is unpleasant: his extra pawn is not felt, but the same cannot be said of his unsafe king. 12...Kg7 Finally, the king finds a safe haven. White will try in the future to launch an attack by advancing his g-pawn, but such ideas should not be dangerous without the support of a dark-squared bishop. Black has plenty of resources, and will generally aim to complete the development of the queenside and, if possible, exchange queens, to remove any attacking threat while seeking to exploit the bishop pair in the endgame. 13.0-0-0 Since White plans to attack on the kingside, it is logical to park his king on the opposite flank.

282

13.g3 leads back to the 10.g3 line noted above. 13.g4 Be6 14.0-0-0 transposes to 14.g4 in the notes to the main line below. 13.e3 Be6 14.Bd3 c6 15.0-0-0 Nd7 transposes to the main line below. The aggressive 13.e4!? is not a bad idea. White gets no advantage, but nor is he worse: 13...dxe4 14.Nxe4 Bf5 15.0-0-0 This was Rodriguez Perez – Chicote Franco, Valladolid 1992, and here I suggest:

15...Nc6N 16.Ng3 Qd5! 17.Nxf5† gxf5 18.Bd3 Nxd4 19.Nxd4 Qxd4 20.Qxd4 Bxd4 21.Bxf5 Rad8= There arises an ending with opposite-coloured bishops, which is approximately equal. 13...Be6 14.e3 14.e4?! is too risky, as 14...dxe4 15.Nxe4 Bxa2 is a safe pawn grab. White tried 16.g4 Nd7 17.Bd3 in Schroll – Epishin, Vienna 1991, when it would have been good for Black to reroute his dark-squared bishop:

283

17...Be7N 18.g5 h5 19.Ne5 Bd6³ White’s compensation for the pawn is insufficient. I also checked: 14.g4 Nd7 15.Bg2 (15.e3 leads to 15.g4 in the notes to the main line below) 15...Be7 16.g5 In Manea – Ocnarescu, Amara 2007, Black could have gone on the offensive with:

16...Bd6!N 17.Qd2 (if 17.Ne5 c6³ Black protects the d5-pawn and threatens to increase the pressure on the pinned knight) 17...Nb6 18.Ne5 Qe7 19.a3 Rac8 20.gxh6† Rxh6 21.h5 Rch8 22.hxg6 Rxh1 23.Bxh1 fxg6 24.Rg1 Bxe5 25.dxe5 c6³ White does not have much of an attack, but the e5-pawn is weak and ...Nc4 could cause trouble.

284

14...Nd7 Having completed development, Black can proceed solidly with ...c6 or more actively with ...c5, according to circumstances. 15.Bd3 This has been the most popular move, although White can try a few different options and move orders without affecting the character of the play a great deal. Here are some examples: 15.Ne5 occurred in S. Foisor – Khotenashvili, Khanty-Mansiysk (ol) 2010, when 15...Rc8!?N would have been a good way to prepare active play on the queenside: 16.Kb1 h5 17.Be2 c5„ The slow 15.Be2 was played in Krause – Budzyn, Neumuenster 2013, when 15...Rc8N would also have been good. A sample continuation is 16.Kb1 c5 17.g4 cxd4 18.Nxd4 Be5 19.Nxe6† fxe6 20.Qb4 Bxc3 21.bxc3 Qb6µ when the c3-pawn is weak and White must fight for a draw in a bad position. 15.g4 initiates an immediate attack which is hardly dangerous. 15...Be7 16.Qg3 occurred in Troff – Defibaugh, Arlington 2011, and here I like a different method of playing on the queenside:

285

16...a5!N An instructive moment may occur after 17.Ne5, when the exact 17...h5! pinpoints the h4pawn as weakness. My analysis continues 18.Nxd7 Bxd7 19.gxh5 Rxh5 20.Be2 Bxh4 21.Qg2 Rh8 22.Qxd5 Bc6 23.Qxd8 Raxd8µ when White’s attack is over and Black’s two bishops are monstrously powerful. 15.Qg3 seems illogical, as White rules out g2-g4, and 15...c5! starts an attack against the white king. 16.Bd3 was played in Drnovsek – Jelen, Otocec 2009, when the natural 16...Rc8N would have connected the rook to the attack. Play may continue: 17.Kb1 Re8 18.h5 g5

Black has a good position, and the tempting 19.Nb5? proves to be a mistake after: 19...c4 20.Bc2 Qa5 21.Nd6 c3 22.Nxe8† Rxe8 23.Rd3 cxb2 24.Rb3 Rc8‚ Black has a strong attack. 15.Kb1 286

Finally, this prophylactic move is a natural choice which discourages ...c5 ideas to some degree, but there are other scenarios in which taking the king to the flank is less than helpful for White – specifically, after a queen exchange. We will follow an instructive example. 15...c6 16.Bd3 a5!? An interesting move, hinting at the possibility of gaining queenside space with ...a4, while keeping the thematic ...Qb8 in reserve. Trading queens immediately is perfectly fine: 16...Qb8 17.Ne2 Qxf4 18.Nxf4 and now in Fistek – W. Schmidt, Poznan 2007, it would have been good to continue 18...Bg4!N 19.Rc1 h5 20.Nd2 Nb6 21.Nb3 Be7 22.Nc5 Rab8³ when Black risks nothing, and any opening of the position will bring his bishops to life.

17.Qg3?! An instructive error: White intends Ne2-f4 but Black can force the queens off before that happens, and in the meantime the g2-g4 advance is blocked. 17.g4!? was more logical, though I still prefer Black’s chances after 17...Qb8 18.Qxb8 Raxb8 19.Rdg1 Be7 20.Ne2 Bd6 21.Nf4 Rhf8 22.Bc2 Nb6³ due to his bishop pair, and after 23.Nxe6†?! fxe6 Black’s rooks will exert crushing power along the f-file. Another idea was to improve the position of the knight: 17.Ne2!? h5 (17...Qb8 is slightly less strong here; nevertheless, after 18.Qxb8 Raxb8 19.h5 g5 20.Ng3 a4 21.Nf5† Bxf5 22.Bxf5 Nb6ƒ Black still retains some initiative, with ideas including ...Nc4 and ...b5) 18.Ng5 Nf8 19.Qg3 Bd7 20.Nf4 a4‚ With an attack on the white king. 17...Qb8! Now White has zero attacking chances and he will have to defend a worse endgame against the bishop pair. 18.Ne2 18.Qxb8 Raxb8 19.Ne2 Bg4³ doesn’t change matters: White has no counterplay and Black can gradually strengthen his position, though it’s a long way from winning. 287

18...Qxg3 19.Nxg3 Now Black begins to rearrange his pieces: the dark-squared bishop belongs on d6 and the knight will go to f6, then onward to e4. 19...Be7 20.Rc1 Bd6 21.Ne2 Nf6 22.Nf4 Bd7 The bishop is not doing much right now but it could prove valuable later, so we should protect it from being exchanged. 23.Rc2 Rhe8 24.Ne2 Ne4 25.Nc3 Black could have significantly increased his advantage with:

25...Nxc3†!N 25...f5 allowed 26.Na4! when White prevented Black’s queenside pawns from advancing, although Black still kept the better chances in Kekelidze – Khalifman, Bad Wiessee 1998. 26.Rxc3 26.bxc3 b5 is also no fun for White. 26...b6! The queenside pawns are set in motion. 27.Rhc1 c5 28.dxc5 bxc5 White’s best bet is to sacrifice the exchange to prevent further domination. 29.Rxc5 Bxc5 30.Rxc5µ White has a bit of compensation but Black’s prospects are clearly higher. 15...c6 16.g4 16.Qg3?! only loses time. After 16...Qb8 17.Qxb8 Raxb8 the players agreed a draw in Smejkal – Jansa, Czech Republic 1999, but Black could have been more ambitious. A possible continuation is 18.h5 g5 19.Ne2 Bd8 20.Ng3 Kf6³ when the power of the two bishops may eventually be felt. 16...Qb8 17.Rdg1 288

17.Qxb8 Raxb8 18.Rdg1 Be7 leads to the same. 17...Be7 18.Qxb8 Raxb8 19.Ne2 19.Ne1 Nf6 20.f3 occurred in Kovalenko – Tukhaev, Alushta 2009, and here I would recommend:

20...h5!N 21.Bf5!? Bd7 22.Bxd7 Nxd7 23.g5 Nb6³ The e3-pawn may become weak, especially with Black’s knight heading for c4. Moreover, the pawns located on the dark squares g5 and h4 may become targets for Black’s bishop in the future. 19...Bd6 20.Nf4 Rbe8 21.Bc2 We have been following P. Wang – Roberts, Cappelle-la-Grande 2008. Here I suggest:

289

21...Nb6 22.Nxe6† fxe6= Black has a comfortable position, a sensible plan being ...Rhf8 and ...e5, with a good game. B32) 8.Nf3 exd5

Petrosian, Averbakh and Seirawan all loved to play this way. White does not try to threaten mate or do anything fancy on the kingside. Instead he solidifies his centre and will usually aim for a minority attack on the queenside. For the time being, Black’s bishop pair is not really felt, as the g7-bishop hits a solid barrier – but as usual, the bishops have a lot of potential to become powerful as the game goes on. 9.e3 White intends to complete the development of his kingside before launching a classic minority attack. 9.e4? makes no sense whatsoever, and 9...dxe4 10.Nxe4 0-0 11.Be2 Nc6 12.d5 Nd4 13.0-0 Nxe2† 14.Qxe2 Re8 15.Qc2 Bf5 16.Rfe1 Qxd5–+ left Black with two bishops and an extra pawn in Patterson – Eilmes, email 1998. 9.Qb3?! achieves nothing, other than blocking White’s b-pawn. 9...c6 10.e3 0-0 11.Be2 Nd7 12.0-0 Nf6 13.Rac1 was seen in Milanovic – Vadasz, Kecskemet 1990, when a logical continuation would have been:

290

13...Qe7N 14.h3 Bf5 15.Rfe1 Rfe8³ Black’s pieces occupy harmonious positions and it is not clear what White’s plan should be in the absence of a b4-b5 advance. White has also tried fianchettoing the bishop but it seems strange to aim it at the d5-pawn. 9.g3 0-0 10.Bg2 c6 11.0-0 Nd7 12.b4 a6 13.a4 Nb6 14.e3 occurred in Arkell – Allen, Southampton 1986, when it would have been logical and natural to play:

14...Bf5N 15.a5 (15.b5 is not dangerous in view of 15...Nc4 16.bxc6 bxc6 17.Nd2 Nb2 18.Qc1 Nd3ƒ when the knight disrupts the harmony of the white pieces, and ...Qa5 will put additional pressure on the queenside) 15...Nc4 16.Nd2 Nd6³ With the queenside closed, Black has the better prospects on the kingside.

291

9.b4 White can launch his minority attack immediately, although in some scenarios it may facilitate counterplay with ...a5. 9...0-0 10.e3 Qd6 11.Qb3 Also possible is: 11.a3 c6 12.Be2 a5 13.b5 c5 14.0-0 Nd7 15.Ra2?! (another game continued 15.Rc1 c4 16.a4 Nb6 17.e4 dxe4 18.Nxe4 and now in Tenikashvili – R. Golubev, Sochi 2018, 18...Qb4!N 19.Nc5 Bf5µ would have left Black clearly better due to the bishop pair and White’s weak pawns on d4 and a4) This was Arkell – Lputian, Cappelle-la-Grande 1991, and here Black can improve with 15...Rd8!N, overprotecting the d5-pawn. After 16.dxc5 Nxc5 17.Nd4 Ne6 18.Qd2 Nxd4 19.exd4 Be6µ Black’s chances are much higher due to the two bishops and the weakness of the d4- and a3-pawns. 11...Be6 12.Be2 c6 13.0-0 Nd7 14.Rac1 Rfc8 15.Na4 b6 16.Ne1 In Mogila – Khusnutdinov, Astana 2010, it would have been an ideal time to strike at the b4-pawn:

16...a5!N 17.a3 No better is 17.b5 c5 18.Qa3 Qf8 19.Nf3 Rab8µ when Black’s position is much more promising. 17...axb4 18.axb4 Rc7 19.Nc3 b5!µ Black’s ideas include ...Nd7-b6-c4 and ...Rca7, capturing the a-file completely, while the b4-pawn remains a long-term weakness. 9...0-0 10.Be2 It is sensible to get castled before starting a fight on the queenside. Neither 10.Rc1 nor 10.b4 has much independent significance: Black replies 10...c6 and the play will almost certainly transpose somewhere below, since White will have to finish developing his kingside at some point. 10.Bd3 This has been significantly less popular than going to e2. Each move has some minor pros and cons 292

compared to the other, but the general plans for both sides will remain about the same. 10...c6 11.0-0 Qe7 12.Rb1 12.a3 has been played a few times but it seems a slightly odd choice, since 12...Nd7N 13.b4 a6 leaves White needing to spend additional time overprotecting the b4-pawn before he can play a3-a4 to threaten b4-b5. 14.Qb3 Nb6 15.a4 Be6 16.Nd2 Rfb8 White still cannot play b4-b5, as ...axb5 followed by ...c5 would give Black a huge advantage. The engine suggests 17.a5 but then 17...Nd7 18.Na4 Re8 19.Nc5 Nf6³ enables Black to prepare a kingside attack at his leisure, since White froze the queenside. 12...Be6 13.b4 a5 14.a3 axb4 15.axb4 In Dorn – Guennigmann, Rotenburg 2009, it would have been worth playing:

15...b5!N Intending to transfer the knight to c4.

293

10...c6 Black’s plan involves: ...Qe7, ...Be6 and ...Nd7-b6, along with ...a6 if White threatens b4-b5 along the way. White’s thematic minority attack will not be dangerous, since a2-a4 and b4-b5 can be met with one of two good plans. One option will be ...axb5, and after axb5 Black can strike at the centre ...c5!. The other is ...cxb5, and after axb5 then ...a5! with a strong passed a-pawn, perfectly supported by the knight on b6. 11.0-0 11.h3 seems an odd choice. 11...Qe7 12.0-0 Nd7 13.Qc2 occurred in Sogorin – Sabbatini, corr. 2011, when 13...Re8N 14.Bd3 Nb6 15.Rfe1 Be6³ would have given Black the more enjoyable game. 11.b4 a6 12.0-0 Qe7 13.Qb3 was seen in Gscheidmeyer – Hawranke, Germany 1980. I recommend:

13...Rd8!N Overprotecting the d5-pawn to enable a future ...c5. An illustrative line is: 14.a4 Be6 15.Rfc1 Nd7 16.b5 axb5 17.axb5 c5! 18.Nxd5 Bxd5 19.Qxd5 Nb6 20.Qb3 cxd4³ 11...Qe7 From here, the queen monitors the b4-square and supports future operations on the kingside and in the centre. For instance, a plan with ...f5-f4 may prove dangerous, after suitable preparation. 12.a3 White prepares to build up on the queenside but his plan is slow, and Black has more than one good way to deal with it. 12.Rb1 would be the normal move in the Queen’s Gambit Declined where this structure occurs, but here it is not helpful for White to play b4-b5 too quickly. After 12...Be6N 13.b4 Black even does well to provoke it with 13...a5! 14.b5 c5 15.Rc1 Nd7„ when Black has excellent play in the centre, and 294

White must switch to a defensive game. 12.Rc1 Nd7 13.Na4 Nf6 14.a3 was played in Miandro – Viviani, Porto Sant Elpidio 2016, and here I can recommend a thematic knight manoeuvre for such positions:

14...Ne4!N 15.Nc5 Nd6³ White has no real prospects on the queenside, while Black can gradually develop his kingside play. A classic game from more than 80 years ago continued: 12.Ne1 Nd7 13.Rc1 Nb6 14.Nd3 Re8 15.Nc5 Here I can offer a better solution than the move played by Paul Keres:

15...Rb8!N Clever prevention. 295

15...Nc4 16.Bxc4 dxc4 occurred in Vaitonis – Keres, Tartu 1938, and now instead of 17.Ne2? b6µ, which gave Black an obvious advantage, White could have played 17.Qa4!N with a double-edged game, since 17...b6? would be refuted by 18.Qxc6. That is why we move the rook away. 16.Bd3 White should probably try something else, but Black has a pleasant game in any case, so I just offer this line for illustrative purposes. 16...Nc4 17.Bxc4 dxc4µ With the rook on b8, Black has clearly better chances thanks to the bishop pair. The c4-pawn may be doubled but it is not weak; it captures space and controls the d3-square, which could make a useful outpost in the future. Unlike in the Keres game noted above, it would be unfortunate for White to play: 18.Qa4? b6 19.N5e4 b5 20.Qc2 b4–+ White loses a piece. 12...Nd7 13.b4 a6! It is useful to take control over the b5-square before putting the knight on b6. 14.Qb3 14.Rc1 occurred in H. Schmidt – Claus, Germany 1994, when it would have been good to follow our standard plan: 14...Nb6N 15.Re1 Nc4 16.Bxc4 dxc4 17.Nd2 Be6 18.Qe2 b5 19.Nde4 f5 20.Nc5 Bf7µ Black has two bishops along with the easy plan of ...a5, opening the queenside. 14.Na4N is an untested but quite logical move to prevent ...Nb6. We can meet it with 14...Qd8, preparing to send the knight in a different direction without allowing an invasion on b6. A possible continuation is 15.Nc5 Nf6 16.a4, when Black can transform the position with: 16...a5! 17.b5 b6 18.Nb3

18...Ne4! 19.Rc1 (Black’s idea is that 19.bxc6? allows 19...Nc3 20.Qd2 Nxe2† 21.Qxe2 Ba6–+ when 296

Black wins the exchange and the game, as White will not be able to hold on to the c6-pawn) 19...cxb5 20.Bxb5 Bd7 21.Qd3 Nd6³ The a4-pawn is weak and Black’s knight may go to c4 at any moment, while the knight on b3 is not happy at all. 14...Nb6 15.a4 Be6 Black develops the last of his minor pieces and is ready to put his knight on c4, which explains White’s next move. 16.Nd2 As usual, 16.b5 is not scary. 16...axb5 17.axb5 c5 18.Qb2 was the continuation of Arzamastsev – Druzhinina, Sukhumi 2006. Black has several promising continuations but my favourite is:

18...Rxa1N 19.Rxa1 Rc8 20.Rc1 g5!µ The central tension already favours Black, so it is logical to expand on the kingside. She threatens to drive away a key defender of the d4-point, and White’s position is clearly under a lot of stress.

297

16...f5! This move is excellently timed, just as yet another white piece moved away from the kingside. Not only does the white king have few defenders, but the f2-e3-d4 pawn chain is also vulnerable to attack by ...f4. 17.g3 Rae8 18.Rac1 f4! Sacrificing a pawn to open the bishop’s path towards h3, in addition to opening files for the rooks. Black’s attack practically plays itself. Although the game continuation works perfectly, a slower build-up with 18...g5 would also have worked well. For instance, 19.a5 Nc8 20.Bh5 Bf7 21.Bxf7† Qxf7 22.Nf3 Qh5 23.Kg2 f4‚ and White faces some tough questions on the kingside. 19.gxf4?! White gets his priorities all wrong. There is no question that 19.exf4 Bxd4 20.a5 Nc8 21.Nf3 Bf6µ would have left him clearly worse because of the pawn structure and Black’s strong pair of bishops – but at least White’s king would not have been in immediate danger. 19...g5! Hooking on the pawn to open additional roads to the kingside. 20.f5 Now White comes under attack. Black has serious pressure on the f-file. 20.fxg5 gives Black an advantage after any sensible move, but the strongest is 20...Bxd4! 21.a5 Nd7 298

22.exd4 Qxg5† 23.Kh1 Qxd2–+ when White’s pawn structure is terrible and his king remains too exposed. 20...Rxf5 21.Kh1 White tries to organize his defences. The attempt to find counterplay with 21.b5 yields nothing: 21...axb5 22.axb5 c5! 23.Bg4 (23.dxc5? d4– + wins a piece) 23...c4 24.Qd1 Rf6 25.Kh1 Ref8 26.Bxe6† Qxe6 27.Qe2 h6µ The protected passed pawn on c4 is a long-term trump, while Black keeps his attacking pressure along the f-file. 21...Ref8 22.Nd1 Kh8 23.Rg1 Bg8 Black’s advantage is stable and he can afford to spend a few tempos reorganizing his pieces. 24.Bf1 Bf7 25.b5 cxb5 26.axb5 Black’s play up to now has been virtually perfect, and here he could have posed his opponent the greatest difficulties with:

26...Nc4!N It is worth mentioning that the game continuation of 26...a5 was also pretty good, and Black went on to win convincingly in Dittmann – Milev, Kienbaum 1958. 27.Qc2 27.Nxc4 dxc4 28.Bxc4 Rxb5 29.Qa2 Rc8–+ wins material, as White’s poor bishop is pinned in two different directions. 27...Bg6 28.Nxc4 299

28.Bxc4 dxc4 29.Qxc4 is hardly an improvement, as either capture on b5 gives Black a serious advantage due to his superior coordination, strong bishops and safer king.

28...Rxf2! 29.Qxf2 Rxf2 30.Nxf2 dxc4 31.Rxc4 Bf8µ Materially White has a tiny advantage of two rooks versus a queen, but Black is the clear favourite. The queen has many targets to attack, and she is supported by two powerful bishops. By combining threats on the kingside with advancing the distant passed a-pawn, Black has good chances to break the defence. Conclusion White’s systems involving an early Bg5 typically lead to a long positional struggle, so it’s no wonder that they were used by such players as Botvinnik and Petrosian. We started by considering the immediate 4.Bg5, when we have no hesitation in sacrificing a pawn with 4...Bg7!?. Following 5.Bxf6 Bxf6 6.cxd5 c6! we get a strong initiative if White trades on c6, so he usually opts for 7.e4 0-0 followed by returning the extra pawn in some way. The resulting positions require some study but Black’s chances are in no way worse. We then analysed the more popular variant with 4.Nf3 Bg7 5.Bg5, when 5...Ne4! leads to a further split. 6.Bf4 has been favoured by several top players and deserves to be more popular, although we saw that 6...Nxc3 7.bxc3 c5! generally leads to an endgame where Black has no problems. We then moved on to 6.Bh4 Nxc3 7.bxc3, when it is important to remember that the change in the placement of White’s bishop makes 7...dxc4! a good move. 6.cxd5 is the most popular continuation but it is strategically rather risky, as 6...Nxg5 7.Nxg5 e6 leads to a stable central structure where Black has the long-term advantage of the bishop pair. 8.Qd2 exd5 9.Qe3† Kf8 is no great problem, and we saw that Black often manages to force a queen exchange with 300

a timely ...Qb8, steering towards a favourable endgame where the bishops will make their presence felt. Finally we analysed the more popular 8.Nf3 which seems rather toothless: White’s main plan in the Carlsbad structure is the minority attack with b4-b5, but we saw that Black is well equipped to meet it. If White’s queenside play comes to a standstill, Black will be well placed to seize the initiative on the kingside, as we saw in the final variation.

301

A) 4.Bf4 Bg7 5.e3 0-0 164 A1) 6.Qb3 166 A2) 6.cxd5 168 A3) 6.Rc1 Be6! 170 A31) 7.cxd5 171 A32) 7.c5 Game 1 173 A33) 7.Qb3 c5! 8.Qxb7 Qb6 9.Qxb6 axb6 180 A331) 10.dxc5 182 A332) 10.Nf3 184 B) 4.Nf3 Bg7 5.Bf4 0-0 187 B1) 6.Rc1 187 B2) 6.e3 c5! 189 B21) 7.Qb3 190 B22) 7.dxc5 Ne4! 192 B221) 8.Be5 193 302

B222) 8.Rc1 194

1.d4 Nf6 2.c4 g6 3.Nc3 d5 This chapter deals with lines in which White develops his bishop to f4 on move 4 or 5. This move takes aim at the c7-pawn, and the queenside in general, while developing the bishop outside the pawn chain in preparation for e2-e3. Just as in the previous chapter, we will examine two versions of White’s plan: the immediate A) 4.Bf4 and B) 4.Nf3 Bg7 5.Bf4. In both cases, we will look to strike at the enemy centre with a timely ...c5. A) 4.Bf4 This move was first employed by Sämisch, against Ernst Grünfeld in 1922. Over the years it has been tested by numerous top players, including world champions Capablanca, Alekhine, Euwe, Botvinnik, Petrosian, Karpov, Kramnik and Anand. 4...Bg7 5.e3 5.Nf3 will be covered in variation B. The current section will deal with lines where White avoids that knight move, or at least postpones it for long enough to take the game into independent channels. As you might expect, if White avoids supporting the d4-point with either of those two main moves, then he risks having his centre blown away by a quick ...c5, as the following lines demonstrate: 5.Qb3?! c5! White is not ready to go wandering with his queen, and his centre starts collapsing like a house of cards. 6.cxd5?! The lesser evil would be 6.Nxd5N 6...Nxd5 7.cxd5 cxd4 8.Nf3 0-0 9.Rd1 Qa5† 10.Bd2 Qb6³, as given by Avrukh. However, Black still enjoys the better game due to the non-development of White’s kingside. 6...cxd4 7.Qa4† 7.Nb5? 0-0 8.Nc7 Nh5 9.Bg3 occurred in De Groot – Hebels, Netherlands 1990, when 9...Nxg3N 10.Qxg3 Na6 11.Nxa8 Qa5† 12.Kd1 Bf5–+ is the winning line pointed out by Avrukh. Black will pick up the knight on a8 and pulverize White’s hopelessly exposed king. 7...Nfd7 8.Nb5 Na6 9.Rc1 0-0 10.Nc7 Ndc5 11.Rxc5 Nxc5 12.Qa3 In Flear – Bejaoui, Tunis 2000, Black’s most convincing continuation would have been:

303

12...b6!N 13.Nxa8 Qxd5 14.b4 Trying to free the knight does not help: 14.Nc7 Qf5 15.b4 Qxf4 16.bxc5 Qxc7–+ White is material down in addition to having a wretched position. 14...Qf5 15.bxc5 Qxf4 16.Nf3 Qf5 17.Qc1 Bb7–+ White is too far behind in development and Black’s initiative is only increasing. I also checked another lunge of the queen: 5.Qa4†?! Although Karpov and Riazantsev have played this way, it is risky. 5...Bd7 6.Qb3 This position was reached in Karpov – Morozevich, Moscow (blitz) 2008, and several other games. Most games have continued 6...Nc6 with adequate play for Black, but there is an even stronger move which is yet to be tested:

304

6...Nh5!N 7.Qxb7 7.Bd2 c5! is great for Black, for instance: 8.dxc5 d4 9.Nd5 e6 10.g4 exd5 11.gxh5 dxc4 12.Qxc4 00µ White’s extra pawn does not come close to compensating for the serious lag in his development. 7...Nc6 8.Bxc7 Qc8 9.Qxc8† Rxc8 Black has temporarily sacrificed two pawns, but White is far behind in development and his centre is crumbling. Best play continues: 10.Be5 Nxe5 11.dxe5 Bxe5 12.e3 Bxc3† 13.bxc3 dxc4³ Material is equal and Black still enjoys a lead in development. He has various active ideas, including ...Nf6-d5 to target the c3-pawn, or simply taking over the open b-file. 5.Rc1?! This move also fails to pay sufficient attention to the centre, thus enabling Black to seize the initiative with: 5...Nh5! 6.Bg5 6.Be3!? dxc4 7.Qa4† occurred in I. Sokolov – Smirin, St Petersburg 2010, when 7...c6!N 8.Qxc4 Nd7 9.Nf3 0-0 would have given Black the more pleasant game; he has a few ways of developing counterplay in the centre, and White still has to decide how to develop his kingside. 6.Bd2 c5 7.e3 cxd4 8.exd4 dxc4 9.d5 a6 10.Bxc4 0-0 11.Nf3 b5 12.Bb3 Bb7 13.Be3 Nd7³ was also promising for Black in Chazalette – Boeken, corr. 2012; his ideas include ...Rc8 and ...Nb6-c4. 6...h6 7.Bh4 c5

305

8.Nxd5 This was Korchnoi’s choice, but it does not solve all of White’s problems. 8.dxc5? is asking for even more trouble, as demonstrated after 8...d4 9.Nd5 g5 10.e4 e6 11.b4 exd5 12.exd5 Qe7† 13.Kd2 Nf6 14.Bg3 Ne4† 15.Kc2 a5–+ in Li Ruofan – Khotenashvili, China 2018. Materially White is just about okay with three pawns versus a piece, but her position is hopeless due to the weak king and general disharmony of her pieces. 8.e3 cxd4 9.exd4 dxc4 10.Nf3 0-0 11.Bxc4 Bg4 12.0-0 Nc6 13.d5 Nd4 14.Re1 Bxf3 15.gxf3 Qc7 16.Bf1 Qf4µ was also depressing for White in Dzagnidze – Svidler, Gibraltar 2012, due to her terrible pawn structure and Black’s mighty knight on d4. 8...Nc6 9.e3 cxd4 10.exd4 10.Be2!? was a reasonable attempt to complicate the game in Kakageldyev – Odeev, Tashkent 2017, but the precise reaction 10...g5!N 11.Bxh5 e6 12.Nf4 gxh4 13.Nf3 Qa5† 14.Nd2 0-0 15.0-0 Rd8µ would have left Black with the bishop pair and more active pieces. 10...Be6 11.Be2 11.Nf3 was played in Taboada – Klimakovs, email 2008, when 11...g5!N would have been strong, for instance: 12.Nd2 Nxd4 13.Qxh5 Rc8 14.Ne3 gxh4 15.Qxh4 Qa5µ Black has a huge initiative in return for a tiny material investment. 11...Bxd5 12.cxd5 Qxd5 13.Bxh5 In Korchnoi – Vaganian, Reggio Emilia 1987, Black would have done well to insert an intermediate check before recapturing on h5:

306

13...Qa5†!N 14.Kf1 gxh5 15.Nf3 0-0µ Black is clearly better, thanks to his powerful bishop and the weak pawns on a2 and d4. 5...0-0 5...c5 is a serious alternative, and the recommendation of Avrukh. Play may continue 6.dxc5 Qa5 7.Rc1 dxc4 8.Bxc4 0-0 9.Nf3 Qxc5 10.Bb3 Nc6 11.0-0 Qa5 12.h3÷ when, objectively speaking, Black is not doing badly, but personally I don’t like this position very much. White has excellent bishops and the queen on a5 may become a target. All this is, of course, rather abstract reasoning, but still it seems to me like the kind of position where Black would not have to do too much wrong to allow White an easy initiative. Castling is more flexible: depending on White’s response, we may either follow up with ...c5 or look for counterplay in some other way. We will analyse three main options: A1) 6.Qb3, A2) 6.cxd5 and A3) 6.Rc1. 6.Nf3 converts to variation B2 on page 189. 6.Bd3 c5 gives Black easy play. 7.Nge2 White avoids a possible ...Bg4 pin which could occur if his knight was on f3. (I also checked 7.dxc5 as played in Beukema – Leenhouts, Netherlands 2018, when 7...Be6!N 8.cxd5 Nxd5 9.Nge2 Nd7 10.Nxd5 Bxd5 11.0-0 Nxc5 would have given Black a clear initiative) 7...Nc6 8.0-0 Now in Thomassin – Himaras, Athens 2005, a logical continuation would have been:

307

8...Nh5!N 9.cxd5 cxd4 10.dxc6 dxc3 11.Nxc3 Nxf4 12.exf4 bxc6³ Black enjoys the advantage of the two bishops. A1) 6.Qb3

White tries to fight for the initiative by attacking d5, but it is risky to try such things when the kingside pieces are undeveloped. 6...c5! Black strikes against d4.

308

7.cxd5 This has been White’s usual choice. A long time ago, Capablanca and Reshevsky preferred a different continuation: 7.dxc5 Ne4! 8.cxd5 Qa5 9.Nge2 Nxc5 10.Qd1 10.Qc4 Nba6 11.Nd4 Bd7 12.Rb1 led to an eventual victory for White in Capablanca – Flohr, Netherlands 1938, but the result may have been different if Black had found the energetic 12...e5!N 13.dxe6 Nxe6 when his lead in development more than makes up for the sacrificed pawn. My illustrative line continues 14.Nb3 Qb6 15.Nd5 Qd8 16.Rd1 Rc8 17.Qe4 Bc6 18.Bxa6 bxa6 19.0-0 Re8 20.Qf3 Nxf4 21.exf4 a5 22.h3 Ba8³ when Black’s initiative outweighs White’s extra pawn, as the two bishops are colossally strong in such an open position. The text move looks safer, and was played in Reshevsky – Kashdan, USA 1942. My improvement is:

10...Nba6!N 11.Nd4 11.a3 can be ignored by 11...Bd7!, since 12.b4? runs into 12...Nxb4 13.axb4 Qxb4 14.Rc1 Rac8–+ and Black regains the piece, as the pin is too strong. 11...Ne4 12.Qb3 White had no choice but to move the queen again, so it is clear that his strategy has been unsuccessful. 12...Bd7 13.Bxa6 Nxc3 14.0-0 Nxd5 15.Qxb7 Nxf4 16.Qxd7 Qxa6 17.exf4 Rfd8 18.Qc6 Qb6 19.Qxb6 axb6 20.Nb3 Ra4µ Black is still a pawn down for the moment, but White’s pawns on a2, b2 and f4 are all weak. Moreover, Black’s rooks are more active than their counterparts and his bishop is stronger than White’s knight in the open position. 7...cxd4 8.exd4 This position has occasionally arisen via the Panov-Botvinnik Attack against the Caro-Kann – but in 309

that opening, White usually avoids the early Bf4 in favour of the much-more-useful Be2, intending Bf3 to overprotect the d5-pawn. Black’s life is significantly easier with the bishop on f4, especially after the following method of generating counterplay.

8...a5! 9.Nf3 a4! 10.Nxa4 10.Qc4 enables Black to continue the energetic play with 10...b5! 11.Qxb5 (or 11.Nxb5 Nxd5 12.Bd2 Bb7 13.Be2 Nc6 14.0-0 Qb6© also gives Black ample play for the pawn) 11...Ba6 12.Qb4 Bxf1 13.Kxf1 Nxd5© when he has more than adequate compensation, due to White’s misplaced king and Black’s future pressure against the pawns on b2 and d4. 10...Nxd5 11.Bd2 Nc6 12.Bc4 Na5 13.Bxa5 Qxa5† 14.Nc3 Nxc3 15.bxc3 In Mikhalchishin – Bagirov, Frunze 1979, Black could have developed his initiative with:

310

15...Bg4!N 16.Be2 Be6!? 16...Bxf3 17.Bxf3 Bxd4³ offers Black a risk-free advantage; the opposite-coloured bishops offer White some drawing chances, but he will have to suffer for a long time. 17.Qb2 Qa3 18.Qxa3 Rxa3 19.c4 Rd8 20.0-0 Bxd4 21.Nxd4 Rxd4 22.Rfc1 Rd2 23.Bf3 Raxa2 24.Rxa2 Rxa2 25.Bxb7 Bxc4 With an extra pawn and same-coloured bishops along with a pair of rooks, Black has decent winning chances. A2) 6.cxd5 Nxd5

The central exchanges remove some tension from the position, although both sides still need to approach the position with care. 7.Nxd5 Qxd5 8.Bxc7 Na6 Black outpaces his opponent in development, which compensates for the minimal material deficit. 9.Bxa6! White should return the extra pawn in order to catch up with his development. Trying to retain the extra pawn would be too risky, as shown after: 9.Bg3?! e5! 10.Nf3 White has nothing better. 10.dxe5? leads to a hopeless position after 10...Qa5†! 11.Qd2 Nb4–+ when White must defend against ...Nc2†, and with moves like ...Rd8, ...Be6 and ...Rac8 on the agenda, White’s position will soon collapse. 311

10.Bxe5 Bxe5 11.dxe5 is a slight improvement on the above line, but 11...Qa5† 12.Qd2 Nb4 13.Nf3 Bf5 14.Nd4 Rac8 15.a3 Nc2† 16.Nxc2 Qxd2† 17.Kxd2 Rxc2† 18.Ke1 Rxb2µ still left White facing an extremely difficult defensive task in Francovig – Spiegel, corr. 2009. 10.Bxa6 was played in P. Varga – Demkovich, Budapest 1997, when 10...bxa6!N would have maintained Black’s initiative. For instance, 11.Nf3 exd4 12.0-0 d3 13.Qb3 Rd8µ when White has returned the extra pawn in order to complete development, but Black remains clearly better thanks to the strong passed pawn and the two powerful bishops. 10...exd4 11.Nxd4 In Colom Andres – Exposito Amaro, Barcelona 2011, it would have been good for Black to play:

11...Rd8!N Simply bringing another piece into play. A brief illustrative line is: 12.h4!? Qa5† 13.Qd2 Qxd2† 14.Kxd2 Nb4³ Black has more than enough compensation for the pawn, due to the poor placement of White’s king.

312

9...Qxg2 10.Qf3 Qxf3 11.Nxf3 bxa6 White has spoiled Black’s pawn structure on the queenside, but Black has the bishop pair to compensate for it. 12.Rc1 Another game continued 12.Rg1 Be6 13.b3 Rfc8 14.Rc1 a5 15.Ke2 a4 16.bxa4 Bxa2 17.a5 Be6 18.Rc5 Bf8 19.Rgc1 Bg4= when Black’s bishop pair was just as important as White’s well-placed rooks in Frijling – Thierry, corr. 2009. 12...Bb7 13.Ke2 a5 14.Rhd1 14.Bxa5 leads to an immediate repetition: 14...Ba6† 15.Kd2 Bb7 16.Ke2= 14...a4 15.Rc5 Rfc8 16.Rdc1 e6 17.a3 After 17.Bd6 Rd8 18.Bc7 Rd5 19.Ne5 the players agreed a draw in Turov – Ftacnik, Pardubice 1998. The position is certainly equal, although Black could try playing on with the two bishops. 17...Bf8 18.R5c3 f6 19.Ne1 Kf7 20.Nd3 We have been following Grigorov – Debray, France 2007. White has active pieces and Black must keep an eye on the doubled a-pawns, but his bishop pair remains useful. My suggested improvement is:

313

20...h5!?N The position remains balanced. You certainly don’t need to know any more than this, but I added a few lines for illustrative purposes. 21.h4 It is dangerous for White to go after the a4-pawn: 21.Rc4?! Ba6! 22.Rxa4? (White should swallow his pride and retreat with 22.R4c3, when no real harm has been done) 22...Bb5 23.Ra5 a6 White’s rook is trapped, and after 24.a4 Bxd3† 25.Kxd3 Bb4 26.Rac5 Bxc5 27.Rxc5 Ke7µ he does not have enough compensation for the exchange. 21...Ba6 22.Kd2 Bb5 Securing the queenside pawns. 23.Ba5 Rxc3 24.Rxc3 Black can no longer play ...Rc8 since his bishop moved to b5, so it looks like the rook’s arrival on c7 will cause problems. However, Black can create counterplay with:

314

24...g5! 25.hxg5 fxg5 26.Ne5† Kg8 A logical conclusion would be: 27.Bb4 Bxb4 28.axb4 Rf8 29.f3 g4 30.fxg4 Rf2† 31.Kc1 Rf1† 32.Kd2 Rf2†= With perpetual check. A3) 6.Rc1

White anticipates a possible ...c5 and at the same time takes aim at the c7-pawn. He intends to develop the kingside later, once the queenside situation has been clarified.

315

6...Be6! Black bets on his fast development. Korchnoi was the first to implement this active move against Borisenko in 1958, but it remained on the margins of theory until the 21st century, during which its popularity took off. Black’s last move poses a threat to the c4-pawn, so White usually responds with A31) 7.cxd5, A32) 7.c5 or the most popular A33) 7.Qb3. White has also tried sacrificing the pawn: 7.Nf3 dxc4 8.Ng5 Bd5 Korchnoi preferred 8...Nd5 but there is no reason to allow White to take on e6. The text move leads to a typical sequence: 9.e4 h6 10.exd5 hxg5 11.Bxg5 Nxd5 12.Bxc4 12.Qf3?! Nb6 13.Qxb7 Qxd4 14.Be2 Qe5 15.Be3 N8d7 16.Rd1 occurred in Murdzia – Florstedt, Germany 2017, when 16...Rfd8!N 17.Qc6 Nf6³ would have left White with a degree of compensation for the sacrificed pawn, but not enough for equality. Black intends ...Nfd5 when life could quickly become unpleasant for White. 12...Nb6 13.Bb3 Qxd4 14.Qxd4 Bxd4 15.Nb5 Bxb2

16.Rc2!? This seems more precise than 16.Rxc7 Nc6 17.Rxb7 as played in l’Ami – Bok, Internet 2016, when 17...Na5!N would have been an excellent temporary pawn sacrifice. Play continues 18.Rxe7 Nxb3 19.axb3 a6 20.Nc7 Rac8 21.0-0 Ba3 22.b4 Bxb4 23.Re4 a5ƒ when the remote passed pawn makes the endgame tricky for White. 16...Be5 17.Nxc7 Bxc7 18.Rxc7 Nc6 19.Rxb7 e6 20.Rc7 Nd4 21.Bd1 21.Kd2 Nxb3† 22.axb3 Nd5= is also balanced: the pawns on a7 and b3 will be destroyed, and there will be too little material left for either side to create serious chances. 316

21...Nd5 22.Rc4 e5 23.h4 Rfc8 24.Rxc8† Rxc8 25.Rh3 Rb8 26.Bb3 Nf4 27.Bxf4 exf4 28.Rd3 Nxb3 29.axb3 Kg7 30.Rd7 Rxb3 31.Rxa7 Rb1†= Accurate play on both sides resulted in complete equality in Dreev – Sarana, Yaroslavl 2019. A31) 7.cxd5 Nxd5 8.Nxd5 Qxd5

White opens the c-file and forces Black to think about the hanging c7-pawn. On the other hand, Black’s pieces are active and White’s queenside pawns also need protection. 9.b3!? This is the most ambitious option. White’s idea is that in certain lines when ...c5 is played, he will be able to respond with dxc5 without losing the b2-pawn. 9.Bxc7? is poor in view of 9...Qxa2 10.Qd2 Rc8 11.Ne2 Nc6 12.Bg3 Na5µ when White lags in development and faces the unpleasant threats of ...Nb3 and ...Nc4. 9.Qb3? is also weak. In Bajc – Pozun, Otocec 2017, the simple 9...Qa5†!N 10.Qc3 Qxa2–+ would have given Black an extra pawn, since taking on c7 would either cost White his b-pawn or allow a horrible pin with ...Rc8. 9.a3 looks natural but it allows Black to equalize without much difficulty: 9...Qa5† 10.Qd2 Qxd2† 11.Kxd2 c6 12.Bc4 Bxc4 13.Rxc4 Nd7 14.Nf3

317

14...c5! The key move, leading to simplifications. 15.dxc5 Bxb2 16.a4 Ba3 17.Rb1 Nxc5 18.Nd4 Ne4† 19.Ke2 Rfc8 20.Rxc8† Rxc8 21.Rxb7 a5 22.Rc7 Nc3† 23.Kd3 Rxc7 24.Bxc7 Nxa4 25.Bxa5= Beinoras – Kulaots, Tallinn 2015. 9...Na6! This multipurpose move defends the c7-pawn and supports a possible ...c5 break, while also preparing a possible raid on White’s queenside with ...Nb4. I am not so happy with Black’s position after 9...Qa5† 10.Qd2 Qxd2† 11.Kxd2 when, compared to the 9.a3 line above, it is harder for Black to carry out ...c5. It’s still too early to talk about White’s advantage, but I would like to avoid it anyway. 10.Nf3! It is time for White to catch up with the development of his kingside. 10.Be2?! allows 10...Qa5† 11.Qd2 Qxd2† 12.Kxd2 when White’s king was misplaced in Oye Stromberg – Hjorth, Ostersund 2018, and Black could have caused problems with:

318

12...Nb4!N 13.a3 (13.Rxc7? runs into 13...Nd5 14.Rxb7 Nxf4 15.exf4 Bd5 16.Bf3 Bxb7 17.Bxb7 Rad8µ when White clearly does not have enough for the exchange) 13...Nd5 14.Bxc7 Rfc8 15.Bg3 Rxc1 16.Kxc1 Rc8† 17.Kd2 Bf5 18.Bc4 Nb6 19.Ne2 Nxc4† 20.bxc4 Rxc4 21.Rc1 Be6³ Black has the advantage of the bishop pair, with chances to create a distant passed pawn on the queenside. 10.Qd2?! This move prevents the check on a5 but does nothing to aid White’s kingside development, so Black can strike at the centre and seize the initiative. 10...c5! Well played! White’s seemingly fortified centre begins to collapse. 11.Nf3 An important point is that 11.dxc5? is refuted by 11...Qxd2† 12.Kxd2 Rfd8† 13.Ke1 Nb4–+ with a decisive attack. Please also note that 11.Bxa6? is impossible due to 11...Qxg2. 11...cxd4 12.Bc4? White commits a serious error, losing a pawn for no compensation. The lesser evil was 12.exd4N 12...Rac8 13.Bc4 Qe4† 14.Qe3 Qxe3† 15.Bxe3 Nb4 16.0-0 Bf5³ when Black has an excellent game due to his active pieces and the isolated d-pawn, but White still has reasonable chances to save the game. 12...Qe4 13.0-0

319

13...Bxc4!N 13...dxe3 gave Black a safe extra pawn in Matsenko – Stankovic, Chelyabinsk 2017, but the text move is even more powerful. 14.Rxc4 e5 15.Bg3 b5 16.Rcc1 Bh6µ The pressure on the e3-pawn makes White’s position extremely unpleasant. 10...Qa5† Now Black cannot play 10...c5? because of 11.Bxa6! and Black doesn’t have ...Qxg2. 11.Qd2 Nb4 12.Be2 b6! Preparing ...c5. 13.Bxc7 In the event of 13.0-0 c5 14.Ra1 Rac8 15.a3 Nc6 16.b4 Qa4÷ Black’s chances are not worse: there is pressure on the d4-pawn, especially with ...Rfd8 coming next. 13...Qa3 14.0-0 Nxa2 15.Ra1 Qxb3 16.Ba6 Qc3 This simple tactical solution forces a queen exchange while keeping an extra pawn. White will have enough activity to maintain the balance, but nothing more. 17.Qxc3 Nxc3 18.Bb7 18.Rfc1 is met by 18...Na2 when the rook cannot stay on the c-file due to the ...Nb4 fork. Therefore it should move to b1 or d1, when the game is likely to end in a repetition after 19...Nc3. 18...Rae8 19.Ba6 320

The position is equal no matter what – but if Black wishes to try for a full point, my suggestion would be:

19...Nd5!?N 19...Bf5 kept things balanced and a draw ensued in Zhak – Ham, corr. 2010. If you are content with a draw, then 19...Ra8N= would be a simple way to repeat the position. The text move forces the play and creates an imbalance. 20.Rfc1 Nb4 21.Bb7 a5 22.Bxb6 Rb8 23.Rc7 Bd5! 24.Bxa5 Bxb7 25.Bxb4 Bxf3 26.Bxe7 Bd5 27.Bxf8 Bxf8= White has a rook and two pawns for two bishops, which is enough for equality. At the same time, the unbalanced material offers some scope for either side to outplay the opponent. A32) 7.c5 Due to the difficulty of fitting a complete Grünfeld repertoire into a single volume, complete illustrative games are few and far between. However, Grischuk’s play in the following encounter was too accurate and instructive to ignore.

GAME 1

Boris Gelfand – Alexander Grischuk Moscow 2014 321

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Bf4 Bg7 5.e3 0-0 6.Rc1 Be6! 7.c5

White seizes space on the queenside and may follow up with b2-b4 and a general advance on that flank. Boris Gelfand loves to play with a space advantage so it is no surprise that he chose this option. 7...c6 Black starts by solidifying his pawn chain. His subsequent plans may involve queenside play with ...b6 and/or a central break with ...e5. His light-squared bishop tends to be something of a problem piece in this structure, so he will generally look to exchange it if given the chance – not only for its counterpart, but also a knight, especially after an early Nf3. 8.Bd3 White develops the bishop in such a way as to discourage ...Bf5 while preventing a possible ...Ne4. Here are some other possibilities: 8.Be2 has the slight drawback of failing to control the e4-square, allowing 8...Nbd7 9.h3 Ne4. Then 10.Nf3 was played in Socko – Piorun, Warsaw 2011, when it would have been good to play on the queenside with:

322

10...b6!N 11.Na4 (11.b4 allows the typical trick 11...bxc5 12.bxc5 Ndxc5! when White should be able to regain the pawn with equality, but he would generally prefer to preserve his pawn chain) 11...b5 12.Nc3 a5 13.Bd3 Bf5 14.Ne2 a4 15.0-0 f6÷ Black has good prospects, with ...e5 on the agenda. 8.b4?! looks natural but is actually an instructive mistake. In Ivanchuk – Grischuk, Elancourt 2013, Black could have emphasized the drawback of this move with:

8...b5!N Fixing the b4-pawn in preparation for ...a5. 9.Bd3 (9.a4?! runs into the strong response 9...a5! 10.axb5 axb4 11.Nb1 cxb5 12.Bd3 Nc6 13.Bxb5 Bd7 14.Ne2 Qa5 15.Bd3 b3†µ when Black’s passed pawn is dangerous) 9...a5 10.a3 Nh5 11.Bg5 axb4 12.axb4 Na6 13.Rb1 Qe8 14.Bh4 Bd7 15.Nge2 e5³ White has been forced to defend the b-pawn, and Black is now ready to develop his initiative on the kingside. 323

8.h3 is sensible, but Black can follow the same recipe as in our main line: 8...Bc8! 9.Nf3 Nfd7 10.e4!? (10.Bd3 transposes to the main line below; the text move gives White’s move order an independent character but is not really an improvement) 10...dxe4 11.Nxe4 Nf6 12.Nc3 Be6 13.Qd2 In S. Zhigalko – Deac, Calimanesti Caciulata 2016, there was no reason for Black not to complete his development with:

13...Nbd7N 14.Be2 Re8 (14...b6!?÷ also deserves attention and leads to interesting play) 15.0-0 Nd5 16.Bg3 Nxc3 17.Qxc3 a5÷ Black has the d5-square and the d4-pawn is weak, so his chances are not worse. A final alternative is: 8.Nf3 Bg4 As has already been said, Black’s task becomes easier when he can exchange his light-squared bishop. 9.Be2 9.Qb3 was played in A. Mastrovasilis – Erdogdu, Aix-les-Bains 2011, when Black could have initiated interesting complications with: 9...Nh5!N (the game continuation of 9...b6 is also not bad) 10.Qxb7 Nxf4 11.Qxa8 Qd7 12.Nb5! cxb5 13.c6 Qc7 14.Qb7 Qa5† 15.Nd2 Na6 16.h3 Bc8 17.Qxa7 Qb4ƒ The position remains extremely complicated but it is Black who owns the initiative. White’s king is less than secure, his queenside pawns are weak and the f4-knight cannot be taken. 9...Nbd7 10.0-0 Re8 11.b4 Bxf3 12.Bxf3 e5 Black makes his main breakthrough and seizes the initiative. 13.dxe5 Nxe5 14.Be2 This was Cebalo – Sedlak, Djakovo 2011, and here I can recommend a small restructuring:

324

14...Nfd7!N 15.Qc2 Qe7 16.Rfd1 a5 17.a3 axb4 18.axb4 Ra3³ White’s bishops are unable to show their strength; Black’s central dominance and control of the open a-file are far more relevant, with ...Rxc3 the immediate threat.

8...Bc8! The bishop was hindering Black’s main plan, namely the ...e5 break, hence the need to move it. The c8-square has the advantage of preparing ...b6 and ...Ba6 as a possible plan to exchange the ‘bad’ bishop. The fact that Black has spent time returning the bishop to its starting position does not matter since the position is closed. 9.h3 325

This is the most logical continuation, preventing ...Bg4 before developing the knight. Consider White’s other options: 9.Nf3 was played in Simantsev – Drygalov, Chelyabinsk 2017. We have already discussed the benefits of exchanging the light-squared bishop – and even having burned another tempo on ...Bc8, the advice still holds true: 9...Bg4!N 10.0-0 Nbd7 11.h3 Bxf3 12.Qxf3 Re8 13.Bh2 e5³ Black captures the initiative. The peculiar 9.a3 was played in Hanssen – Verlinde, corr. 2015, when 9...Nbd7N would have been natural and strong. For instance, 10.Nf3 (or 10.h3 Nh5! 11.Bh2 e5³) 10...Nh5! 11.Bg5 Qe8³ followed by ...e5, seizing the initiative. 9.b4 Re8 10.Nf3 Bg4 11.Be2 Nbd7 12.0-0 Bxf3 13.Bxf3 e5 is another example of Black successfully carrying out his plan. 14.dxe5 Nxe5 15.Be2 was the continuation of Rajesh – Matinian, Sitges 2017, and here I suggest regrouping with:

15...Nfd7!N 16.Re1 Nf8 17.h3 Ne6 18.Bg3 Qe7³ Black has perfectly centralized pieces, and can think about ...a5 and/or ...d4 in the future. 9.Bg5 has been tested at GM level but it should not pose problems. 9...b6 10.b4 a5 11.b5 bxc5 12.dxc5 Nbd7 13.bxc6 Nxc5 14.Bxf6 Bxf6 15.Be2 Bxc3† 16.Rxc3 Ne4 17.Rc2 was seen in Duda – Bok, Doha 2015, when Black could have activated the queen:

326

17...Qd6!N 18.Nf3 Qb4† 19.Nd2 Nc3µ Black will exchange on e2 next, leaving White with a problem king. Finally, 9.Bg3 is a strange kind of waiting move which was tested at the elite level. 9...b6 10.b4 Nh5 11.Bh4 White’s bishop manoeuvre does not seem logical, and Black reacted well with 11...a5! 12.cxb6 axb4 13.Na4 Nd7 14.Rxc6 Bb7 15.Rc1 in Wojtaszek – Nepomniachtchi, Wijk aan Zee 2017. Here it would have been best to destroy the pawn:

15...Nxb6!N 16.Nc5 Bc6 17.Rc2 Qd6³ Black is ahead in development and threatens to open the centre with ...e5, and he may also look to exert pressure along the a-file. 9...Nfd7! 327

Another excellent move, preparing ...e5. 10.Nf3 I also considered 10.Bg5, in which case we can switch to queenside play: 10...b6! 11.cxb6 axb6 12.Nge2 Nf6 13.0-0 Bb7÷ Black intends ...c5 and ...Ne4 with an active game. 10...e5! Black carries out his plan, regardless of the fact that White seems to have the e5-square under control. 11.dxe5 Black has nothing to fear from 11.Bxe5 Nxe5 12.Nxe5 (12.dxe5? Nd7 favours Black) 12...Bxe5 13.dxe5 Nd7 14.0-0 Nxe5 15.Ne2 b6 16.b4 a5 17.b5 bxc5 18.Rxc5 Qb6÷ when the c5-rook, the b5pawn and the d3-bishop are all under fire, so White must take care not to become worse. 11...Nxc5 12.Bb1 Another GM game continued 12.0-0 Nbd7 13.Bb1 when, in Socko – Baron, Riga 2015, it would have been good to continue:

13...Qb6N 14.Qc2 Re8 15.Rfd1 (15.Rfe1 Nxe5 16.Nxe5 Bxe5 17.Nxd5 cxd5 18.Bxe5 Rxe5 19.Qxc5 Qxb2 20.a4 gives White enough compensation for equality, but nothing more) 15...Ne6÷ Black’s chances are not worse. For example, after 16.Bg3 Nc7 17.a3 Nxe5 18.Nxe5 Bxe5 19.Na4 Qa6 20.Bxe5 Rxe5 21.Nc5 Qa5 White is under pressure to prove the correctness of his pawn sacrifice.

328

12...Nbd7 13.b4!? This bold move was Gelfand’s choice. However, the pawn may prove to be too exposed on this square, and Black’s knight will be perfectly located on e6. At first I thought this move was a mistake but it is actually okay, as long as White follows it up correctly. The more restrained 13.0-0N transposes to Socko – Baron above. 13...Ne6 14.0-0 White has to part with his bishop, since 14.Bg3 runs into 14...Ng5! 15.Nxg5 Qxg5 16.h4 Qe7 17.f4 Qxb4³ when the drawback of White’s 13th move becomes clear. 14...Nxf4 15.exf4 Nb6 Grischuk makes room for the c8-bishop while activating the knight, which can go to c4 when the time is right.

329

16.Qd4?! 16.b5!N would have organized some much-needed counterplay, while also making sense of the earlier b2-b4 move. The position is dynamically balanced and there are all kinds of possibilities. Here is one line I analysed: 16...Re8 (16...Bh6!? is another idea, since g2-g3 is no longer possible) 17.Re1 Bd7 18.f5! gxf5 19.bxc6 bxc6 20.Bd3 Kh8 21.Qc2 f6 22.e6 Rxe6 23.Rxe6 Bxe6 24.Ne2 Nc4© Black is temporarily two pawns up with the bishop pair, but White has enough compensation for equality. It was a mistake of White to place the queen on the diagonal of the ‘Grünfeld bishop’, and Black’s next move must have been an easy decision. 16...f6! 17.b5 Nc4 Even though Black keeps an advantage after the game continuation, 17...fxe5!N would have been slightly more accurate, as it should reach the same position as the game, but without allowing White the improvement noted at move 20 below. Play continues 18.Nxe5 (18.fxe5? is impossible here due to 18...Rxf3 19.gxf3 Qg5† 20.Kh1 Bxh3 21.Rg1 Qh5 and White can resign) 18...Nc4 19.bxc6 (19.Bd3 Nxe5 20.fxe5 Qb6µ; 19.Rfe1 Nxe5 20.fxe5 Bh6 21.Rcd1 Qe7 22.bxc6 bxc6µ) 19...bxc6 20.Bd3 when we rejoin the game. 18.bxc6 bxc6 19.Bd3 fxe5 White has a choice of two continuations, and it is a difficult task to determine the lesser of the evils.

330

20.Nxe5?! Gelfand must have believed that this move gave more chances for salvation, but 20.fxe5! would have been better. Play continues 20...Rxf3 21.Bxc4 Kh8! 22.Ne2 Bxh3 23.Bxd5 Qxd5 24.Qxd5 cxd5 25.gxf3 Bxf1 26.Kxf1 Bxe5 27.f4³ when White is a pawn down but he should be able to hold, thanks to his active rook and knight. 20...Bxe5! This leads to a situation where White will be a pawn down with no counterplay. The alternative is less convincing: 20...Nxe5 21.fxe5 Qb6 (21...Qe7 leads to equality after 22.f4 Bd7 23.Na4 g5 24.Qc5 Rae8 25.Qxa7 gxf4 26.Nb6 Bf5 27.Qxe7 Rxe7 28.Bxf5 Rxf5 29.Rxc6 Bxe5=) 22.Ne2 Bd7 23.f4 g5 24.g3 Bxh3 25.Rf2 Bd7÷ Black has an extra pawn and two bishops, but White’s connected, mobile passed pawns mean that any result is possible. 21.fxe5 Nd2 22.Rfd1 Qg5 23.Qe3 Nf3† 24.Kh1 Qxe3 25.fxe3 Nxe5

331

26.e4? The plan of attacking Black’s centre with e3-e4 is correct, but White should have prepared it with 26.Bc2! in anticipation of ...d4. Play continues 26...Kg7 27.e4 when 27...d4! is nevertheless the right move: after 28.Rxd4 Be6 29.Bb3 Bxb3 30.axb3 Rfb8 31.Rb1 Kf6³ material is equal, but White faces a difficult defence because Black’s king and knight are more active than their counterparts, while the b3and e4-pawns are weak. 26...d4 27.Na4 Rb8! Grischuk is not too concerned with preserving his extra pawn: instead he activates his pieces and gets ready to invade the enemy camp. 28.Bc4† Kg7 29.Rxd4 29.Nc5 does not improve the situation in view of 29...Rd8 30.Rd2 Kf6 31.Kh2 Rb4µ when Black's extra pawn has serious weight.

332

29...Rb4! Pinning the bishop while conveniently gaining a tempo by attacking the knight. 30.Nc5 Rf2! White’s pieces lack harmony, so Black can start attacking the king. The immediate threat is ...Bxh3, since recapturing will allow ...Nf3 with a mating net. 31.a3 Untying the bishop. The alternative was 31.Nd3 Rd2! 32.Nxb4 Rxd4 33.Ba6 Rxe4 34.Bxc8 Rxb4µ when Black should convert his extra pawn with good technique. 31...Rbb2 32.Na4 Gelfand must have been counting on this resource, but Black has a powerful answer.

333

32...Bxh3! Black ‘sacrifices’ the exchange but picks up two pawns for it, while also exposing White’s king. 33.Nxb2 33.gxh3? loses trivially: 33...Rh2† 34.Kg1 Nf3† 35.Kf1 Rbf2# 33...Bxg2† 34.Kh2 Rxb2 35.Kg3 g5 36.Rcd1 h5 The passed pawns inexorably march forward. For White, certain death awaits, but Gelfand continues to resist stubbornly. 37.R4d2 Rxd2 38.Rxd2 Bxe4 39.Re2 h4† 40.Kf2 Ng4† 41.Kg1 Nf6 A much easier win was 41...h3! 42.Rxe4 h2† 43.Kg2 h1=Q† 44.Kxh1 Nf2† 45.Kg2 Nxe4–+ followed by a routine conversion of Black’s two-pawn advantage. 42.Be6 Kg6 43.Re1 Bf3 44.Re5 g4 45.Bf5† Kh6 46.Bd3 Grabbing the a7-pawn does not save White: 46.Ra5 g3 47.Rxa7 Nd5 48.Rh7† Kg5 49.Bh3 Nf4 50.Bf1 Bd5 51.a4 c5 52.a5 c4 53.a6 c3 54.a7 h3 55.Bxh3 c2 56.Rc7 Nxh3†–+

334

46...Bd5 47.Bf1 g3 48.Bh3 Ne4 49.Be6 Nf2 50.Bxd5 cxd5 51.Kg2 d4 Black has three passed pawns, supported and secured by the knight. White has no counterplay and can only resist by cutting off the black king for as long as possible. 52.Rd5 d3 53.a4 a5 54.Kf3 Gelfand rejects 54.Rxa5 in view of 54...d2 55.Rd5 d1=Q 56.Rxd1 Nxd1 57.a5 Nc3 58.a6 Nb5 when Black halts the a-pawn and White should resign. 54...Kg6 55.Rd8 Kf7 56.Rd4 Kf8 56...Ke6 would have reached the position from move 61 in the game. It is possible that Grischuk was in time trouble and opted to play a few waiting moves before committing to the winning line. 57.Rd5 Ke8 58.Rd4 Ke7 59.Rd5 Kf6 60.Rd6† Kf5 61.Rd4 Ke6 62.Rd8 Ke5 63.Rd7 h3 64.Kxg3 Ne4† 65.Kf3 65.Kxh3 d2 wins: Black threatens to cut off the rook with ...Nd6 and the only way to prevent it, namely 66.Rd3, allows a fork on f2. 65...h2 66.Kg2 d2 0–1 A33) 7.Qb3

335

This has been White’s most popular choice, aggressively attacking the b7-pawn. 7...c5! A great move! Black leaves the pawn to its fate and energetically fights for the initiative. 8.Qxb7 White plays in a principled manner, hopes to neutralize Black’s initiative and retain an extra pawn. An unimpressive alternative is: 8.Nf3 If White does not wish to take on b7, then why did he put his queen on b3 in the first place? 8...Nc6 9.dxc5 9.Qxb7? does not bring any benefits: on the contrary, after 9...cxd4 10.exd4 Na5 11.Qa6 dxc4 12.Be2 Nd5 13.Nxd5 Bxd5 14.Bd2 Nc6µ Black was obviously better in Petrosyan – Vorontsov, Moscow 2014, due to his active pieces and the weakness of the d4-pawn.

336

9...Ne4! This position has been tested in several correspondence games. Although there are complications in store, Black’s chances are higher. 10.Qxb7 10.cxd5 Bxc3†! 11.bxc3 Bxd5 gives Black some initiative. 10.Rd1 Qa5 11.cxd5 Nxc5 12.Qc4 Na4! leads to complications which favour Black: 13.Bc7 b6 14.Qxc6 Nxb2 15.dxe6 Bxc3† 16.Rd2 f6 17.Bb5 Bxd2† 18.Nxd2 Rac8 19.h3 a6 20.Be2 Rfd8 21.0-0 Rxd2 22.Qb7 Re8 23.Bxb6 Qxa2µ White did not have enough for the exchange in Ballow – Haugen, corr. 2016. 10...Qa5 11.Qxc6 Nxc3 12.bxc3 Bxc3† 13.Kd1 Rad8 14.Nd4 dxc4 15.Bc7 Qb4 16.Bxd8 Rxd8 17.Rxc3 Qxc3 18.Qc7 Rxd4† 19.exd4 Qxd4† 20.Kc1 Qa1† 21.Kd2 c3† 22.Kd3 Bf5† 23.Ke3 c2 24.Bd3 Qxh1 25.Qb8† Kg7 26.Qb2† f6 27.Qxc2 Bc8 28.g3 We have been following Rumyantsev – Denisov, corr. 2016. Surprisingly, the sequence from moves 10 to 28 was almost forced! White’s king cannot be caught immediately, but it will remain in a precarious position for some time. I think Black should have continued:

337

28...Bd7!Nµ Preventing c5-c6. The game goes on but White has a difficult task, as shown by the following sample line: 29.Be4? This tempting move loses. 29...Qe1† 30.Kf3 Bh3! 31.c6 f5 32.Bd3 e5 33.Be2 Bg4† 34.Ke3 f4† 35.gxf4 exf4† 36.Kxf4 Bxe2–+ Winning a piece and the game. You also need to be prepared for complications after: 8.dxc5 This was seen in Neuman – Erdos, Deizisau 2013, among other games. Black has usually continued with 8...Na6, which is not a bad move, but I prefer a novelty:

338

8...Ne4!N We saw the same kind of idea in the note above. Here the moves Nf3 and ...Nc6 have not been played, but Black’s plan still works well. 9.Qxb7 9.Nxe4 dxe4 10.Qxb7? is refuted by 10...Qa5† 11.b4 Qa3–+ when White’s king will be caught. 9.cxd5 must be met by 9...Bxc3†! 10.bxc3 Qxd5 11.Qxd5 Bxd5 12.f3 Nxc5 13.a3 Nbd7 14.e4 Bb3³ when White’s bishop pair is less important than Black’s superior pawn structure and well-placed pieces. 9...Nxc3 10.bxc3 10.Qxa8? loses to 10...Nxa2–+ with ideas of ...Qa5† and ...Bxb2, when White’s king falls under a decisive attack. 10...Qa5 11.Ne2 Once again, it would be too dangerous to play 11.Qxa8? Bxc3† 12.Rxc3 Qxc3† 13.Ke2 Qxc4† 14.Ke1 Qc1† 15.Ke2 Nd7 16.Qxa7 d4–+ with a decisive attack on the white king. 11...Nd7 12.Qb5 Qxa2ƒ Black has a lot of activity, while White still has to develop the kingside and secure his king. 8...Qb6 The only move. Despite the impending queen exchange, Black will have a nice initiative to compensate for the minimal material shortage. 9.Qxb6 White has no choice. The greedy 9.Qxa8? was played in Szuper – Shishkin, Krakow 2012, when the natural 9...Nc6!N 10.Qxf8† Kxf8 would have been winning for Black. For instance, 11.cxd5 Nxd5 12.Nxd5 Bxd5 339

13.Rxc5 Qxb2–+ when White’s king is in peril and he will have to shed material. 9...axb6 Black’s play is based on his central pressure along with the open a-file. We will analyse A331) 10.dxc5 and A332) 10.Nf3. 10.cxd5 Nxd5 11.Nxd5 Bxd5 12.dxc5 bxc5 transposes to the 11.cxd5 line in the notes to variation A331 below. The passive 10.b3 was played in Bourgeois – Krysa, Buenos Aires 2013, when Black could have developed great activity with: 10...Nc6!N 11.Nf3 Nb4 12.a4 dxc4 13.bxc4 Bf5 14.Be2 Ne4µ White’s pawns on a4 and d4 are under serious pressure. A331) 10.dxc5 bxc5

11.a3 I checked two other moves: 11.Be5? is a strange choice, as Black will be able to develop the b8-knight with gain of tempo. 11...dxc4 12.Nh3?! was the continuation of Hammer – Le Quang Liem, Cap d’Agde (rapid) 2010, when Black’s best would have been: 12...Nbd7!N (though slightly less accurate, the game continuation of 12...Nc6 was also somewhat favourable for Black, who went on to win) 13.Bg3 Nd5 14.Bxc4 Nxc3 15.Bxe6 Nxa2 16.Bxa2 Rxa2 17.0-0 Rxb2–+ There is no reason why Black should not be able to convert his extra pawn. It would be risky for White to play: 11.cxd5 Nxd5 12.Nxd5 Bxd5 13.Rxc5 e6 340

Black is temporarily two pawns down but he has more than enough compensation, due to his active pieces aiming at the queenside, along with White’s exposed king. 14.Ne2 Rxa2 15.Nc3 Ra1† 16.Ke2 Nd7 17.Rc7 Nb6 18.f3 I also checked 18.Nxd5 Nxd5 19.Rc2 Rd8 20.Bg3 h6 21.e4 (21.f3 allows 21...Bxb2!, since 22.Rxb2? Nc3†–+ will cost White his rook) 21...Nb4 22.Rd2 Rxd2† 23.Kxd2 Rb1 24.Ke3 Nc6µ when White continues to have a hard time developing his kingside pieces. 18...Bc4† 19.Kf2 This position was reached in Manukyan – Melkumyan, Yerevan 2014. I can recommend a long, forcing continuation:

19...Bxc3!N 20.bxc3 Rfa8 21.Bh6 R8a2† 22.Kg3 Bxf1 23.Rb7 Rxg2† 24.Kh3 Rg1† 25.Kh4 Ra4† 26.e4 Rxe4†! If it were not for this resource in conjunction with Black’s next move, he would be dead lost. 27.fxe4 g5†! 28.Bxg5 Rxh1 29.Rxb6 Rxh2† 30.Kg3 Rg2† 31.Kf4 Kg7³ It is difficult to convert the extra pawn in such a position due to the small amount of material and opposite-coloured bishops, but Black can certainly try. 11...dxc4 12.Nd1 This position has been tested several times at a high level. It seems to me that White will do well to hold a draw. 12...Ra4 13.Ne2 White can invite a draw by repetition with 13.Nc3 Ra7 14.Nd1, as seen in Batchuluun – Zhou Jianchao, Al Ain 2015. However, Black is not obliged to defend the c4-pawn, but can instead utilize the slight improvement of his rook’s placement to continue the game with:

341

14...c3!N 15.Nxc3 Nh5 16.Nge2 Rb7 17.Nd1 Bb3© Black has excellent compensation for the pawn, and White’s position looks more difficult to handle, especially over the board. 13...Nc6 14.Nec3 Ra7 15.Nb5 Rb7 16.Nc7 In Batchuluun – Jumabayev, Incheon 2013, Black’s strongest continuation would have been:

16...Bf5!N 17.Na6 Nd7 18.Rxc4 Na5 19.Nxc5 Nxc4 20.Nxb7 Nxb2 21.Nxb2 Bc3† 22.Kd1 Bxb2ƒ White can hardly hope to keep his extra pawn, so he faces an unpleasant defensive task due to his unsecured king and undeveloped state of his kingside, as well as the knight on b7 which is stuck a long way from the other pieces.

342

A332) 10.Nf3

White seeks to finish development as quickly as possible, while avoiding central exchanges which generally only help to increase Black’s activity. 10...dxc4 Black does not capture the pawn for materialistic reasons, but rather to free the d5-square for his knight, or in some cases even the bishop. 11.dxc5 A harmless alternative is: 11.Ng5 In Jelfs – Shafi, Paisley 2018, Black could have obtained the upper hand with: 11...cxd4N 12.exd4 Bd5 13.Bxb8 White needs to get his move order right in order to minimize his disadvantage. On 13.Nxd5?! Nxd5 14.Bxb8 Black unexpectedly does not have to recapture on b8 but instead plays 14...b5! 15.Bg3 Rxa2 16.Be2 Rxb2µ when the d4-pawn is weak and the connected passers pose a serious threat. 13...Rfxb8 14.Nxd5 Nxd5 15.Bxc4 e6 16.Nf3 Nf4 17.g3 b5 18.Bf1 Nd5 19.a3 b4³ White has an extra pawn for the moment, but Black has more than enough compensation with serious pressure on the queenside. It seems quite likely that White’s three queenside pawns will be eliminated, resulting in an endgame where Black has four pawns versus three on the kingside, with at least some winning chances, depending on which other pieces remain on the board. 11...bxc5

343

12.Ng5 This move deflects the unpleasant idea of ...Nd5, but the drawback is that the knight will be left misplaced on g5. White’s other idea is to organize pressure on the c4-pawn: 12.Nd2 Nd5 13.Bxc4 (White erred with 13.Nxd5?! Bxd5 14.Bxc4 Bxg2 15.Rg1 Bc6 16.Bb3 Na6µ in M. Brown – Kudrin, Las Vegas 2012, when the threats of ...Nb4 and ...Bxb2, caused White’s position to collapse) 13...Bxc3 14.bxc3 Nxf4 15.exf4 Bxc4 16.Nxc4 Rxa2

17.g3 Nd7 18.0-0 Nf6 19.Ra1 Rfa8 20.Rxa2 Rxa2= The simplifications resulted in a level endgame in Batchuluun – Rodshtein, Tromso (ol) 2014.

344

12...Bd5

13.Rd1 I checked two other options: 13.e4?! This seemingly active move only creates an additional weakness, since the pawn has broken away from its base and will need to be protected. 13...Bb7 14.f3 14.Bxc4? is worse: 14...h6 15.Nf3 (after 15.Nxf7 Rxf7 16.0-0 Nc6 17.Bxf7† Kxf7 18.Rfd1 c4µ Black’s minor pieces were stronger than the rook and two pawns in Dehaybe – Pepermans, email 2013) 15...Nxe4 16.0-0 Nxc3 17.bxc3 Bxf3 18.gxf3 Ra4 19.Bxb8 Rxc4–+ Gurbanov – Givon, Beersheba 2014. White has defended the e4-pawn and wants to exploit the weakness of the doubled c-pawns. Black, on the other hand, has considerable dynamic capabilities. For example, he can plonk a knight on d4 and/or attack the b2-pawn, and can also consider ...Nh5 and ...Bh6 with the threat of ...f6. 14...Ba6 Black protects the pawn, not only for materialistic reasons, but also because it controls the d3- and b3-squares, thus holding the b2-pawn in place. It was worth moving the bishop anyway, as it was biting on granite in the form of the e4-f3-g2 pawn chain. 15.Be3 The bishop steps away from a possible ...Nh5 strike and attacks the c5-pawn. 15.Nh3?! looks cowardly. 15...Nc6 16.Nf2 Nd7 17.Nfd1 was played in Barseghyan – Li Di, Moscow 2015, when 17...Bd4!N 18.Nd5 Ra7 19.N5e3 Nde5 20.Nxc4 Bxc4 21.Bxc4 Nxc4 22.Rxc4 Rxa2µ would have left White badly uncoordinated, with his rook on h1 still not having entered the game. 15.Na4 Nbd7 16.Bxc4 Bxc4 17.Rxc4 Nh5 18.g3 h6 19.Nh3 g5 20.Nxc5 was played in 345

Khusnutdinov – Nepomniachtchi, Doha 2016, when Black could have won with: 20...Nb6!N 21.Rb4 gxf4 22.Rxb6 Rxa2–+ The threats include ...Ra1†, ...Bd4 and ...fxg3. White cannot castle because of the check on d4, and his position is completely devoid of harmony. 15...Nc6! “Activity first!” is Black’s motto, disregarding material. 16.Bxc5 Rfb8 Black is actively placed, whereas White is far behind in development and his pieces are disharmoniously located. 17.Ba3 Bh6 Pinning the knight and forcing White to weaken his kingside. Another interesting game continued: 17...Nd7!? 18.Nd1 (White can stop the knights from using the e5-square with 18.f4, but 18...Nb4 19.Nf3 Nc5 20.Ne5 Ncd3† 21.Bxd3 Bxe5 22.fxe5 Nxd3† 23.Kd2 Nxc1 24.Kxc1 Rb7µ gives Black excellent chances to realize his advantage) 18...Nce5 19.Bxe7 Nd3† 20.Bxd3 cxd3 21.Bd6 Rc8 22.Bf4 Rc2µ Despite being two pawns up, White’s position is not to be envied, as he has yet to solve the problem of his king and the h1-rook. 18.h4 Nh5 19.Rh3? Under pressure, White commits an error which loses a piece. 19.Kf2 was more stubborn although 19...f6 20.g4 Nf4 21.Nh3 Nxh3† 22.Bxh3 Bxc1 23.Rxc1 Nb4µ leaves Black the exchange up with excellent winning chances. 19...f6 20.g3 fxg5 21.hxg5 Bxg5 22.f4 In Morozov – Sutovsky, Heraklio 2017, Black’s most convincing route to victory would have been:

22...Bh6!N Prosaically keeping the extra piece. 22...Bxf4?! was the game continuation which still favoured Black, but there was no real need to give back material. 23.Nd5 Bc8 24.Rh2 e6 25.Bxc4 Nd4 26.Ne7† Kf7 27.Nxc8 Rxc8 28.Rh1 Bf8–+ 346

Black should win with his extra piece. White’s best idea is actually to give up playing for an advantage and instead look for a draw by simplifying the position: 13.Nxd5! Nxd5 14.Bxc4 Nxf4 15.exf4 Bxb2 16.Rc2 Bd4 17.0-0 Nc6= Juhasz – Rohan, Al Ain 2013. Objectively, the position is equal, although Black’s passed c-pawn and White’s doubled f-pawns offer Black some hope of gaining an advantage. 13...e6 White can proceed in various ways, but none of them offers a complete solution to his problems. 14.Bd6 I considered the untested 14.e4N 14...Bc6 15.f3, when 15...Nfd7! 16.Bxc4 Nb6 17.Bb3 N8d7 18.Ke2 Bxc3 19.bxc3 c4 20.Bc2 Rxa2ƒ gives Black some initiative due to the active rook on a2. 14.a3 This was played in Eilmes – Dulany, corr. 2013, and here I advise:

14...Nc6!N Based on the following idea: 15.e4 h6 16.exd5 exd5 Black gets a powerful fist consisting of the three central pawns, as well as an open file leading to White’s king. 17.Nf3 Rfe8† 18.Kd2 Perhaps White’s best chance is to return the extra piece, although in that case it is obvious that Black has won the opening battle: 18.Be3 d4 19.Bxc4 dxe3 20.fxe3 Ng4ƒ The e3-pawn is weak, and Black may consider ...Bxc3 followed by ...Rxa3 at any moment. 18...d4 19.Nb5 Ne4† 20.Kc1 d3ƒ 347

Black has a serious initiative for the piece. My illustrative line continues: 21.Bg3 Na5 22.Nc7 Bxb2† 23.Kxb2 Reb8† 24.Ka2 Nc3† 25.Ka1 Nb3† 26.Kb2 Nxd1† 27.Kb1 Rxa3–+ White can only surrender. We have been following Kohlweyer – Bobras, Germany 2011. My improvement is:

14...Rd8!N 15.Bxc5 Rc8 16.Bd4 16.Bd6 proves unsuccessful after 16...h6 17.Nf3 Ne4ƒ when the a- and b-pawns are vulnerable. 16...h6 17.Nf3 Nc6 18.Ne5 Nxd4 19.exd4 Ne4ƒ Black has more than enough compensation for a pawn, thanks to the two bishops and pressure against White’s queenside pawns. B) 4.Nf3 Bg7 5.Bf4 The Bf4 system with the inclusion of Nf3 is more popular than without. White’s plans do not entail a sharp game on the queenside, as we saw in the previous variations, but a simple development of the kingside pieces. White’s set-up was first employed by the Spanish master Rey Ardid in 1922, before more famous players such as Flohr, Bogoljubov, Capablanca, Petrosian and Korchnoi took it up. In more recent times, Kramnik, Ivanchuk, Dreev and Bareev have all played it. As usual, Black’s plans should include central activity with ...c5, and perhaps a timely ...Ne4 to open the long diagonal and put pressure on White’s centre and queenside. 5...0-0 White can either accelerate his queenside development with B1) 6.Rc1 or play the more natural B2) 348

6.e3. B1) 6.Rc1 This move is quite likely to transpose elsewhere, but it may also have its own significance. 6...Be6! Black follows the same formula as in the earlier variation A3. 7.Ng5?! White violates a well-known principle by moving the same piece repeatedly in the opening. Nevertheless, this has been played by some strong GMs and must be checked. 7.e3 transposes to 7.Nf3 in the notes to variation A3 on page 170; and 7.c5 c6 8.e3 leads back to the 8.Nf3 line in the notes to variation A32 on page 175. 7.cxd5 Nxd5 8.Nxd5 Qxd5 is rarely played, but is likely to transpose to variation A31 on page 171, as White will surely play e2-e3 at some point. 7.Qb3 c5 8.Qxb7 Qb6 9.Qxb6 axb6 10.dxc5 (10.e3 leads back to the recently analysed variation A332) 10...bxc5 11.a3 Nbd7 12.Ng5 was played in Chatalbashev – Sadzikowski, Gallipoli 2017, when Black missed a strong idea:

12...d4!N 13.Nxe6 fxe6 14.Nb5 e5 15.Bg3 e4© White’s kingside is completely undeveloped and Black’s central pawns have captured a lot of space, all of which adds up to more than enough compensation for the small material deficit.

349

7...c5! 8.dxc5 8.Nxe6 fxe6 supports Black’s centre and gives him an open f-file to go with his other trumps. 9.e3 (9.dxc5? d4 10.Na4 Ne4 11.e3 Qa5† 12.Ke2 Nc6–+ was even worse in Lomakin – D. Kovalev, Mukachevo 2012; White’s king can hardly expect to survive on such a square) 9...Nh5! 10.Bxb8 Rxb8 11.Be2 cxd4 12.exd4 dxc4 13.Bxc4 Bxd4µ Black’s pieces are more active and the f2-point is weak, Faghirnavaz – Darini, Tabriz 2013. 8...d4 9.Nb5 Nc6 10.Nc7 Bf5 11.Nxa8 e5 12.Bd2 e4 13.e3 For the moment Black is a rook down, but it’s obvious that White is playing a dangerous game. Notice that Black is in no immediate hurry to pick up the knight on a8: he can do so at any time, but for the moment it is more important to maintain the momentum of his initiative. We will continue following a top-level encounter. 13...h6 14.Nh3 Bxh3 15.gxh3 Ne5 16.Bg2 Nd3† 17.Kf1 Nxc1 18.Qxc1 d3 19.Bc3 In Grischuk – Caruana, Moscow 2012, Black could have secured a decisive advantage with:

350

19...Re8N 20.Qd2 Qxa8 21.b4 Qc8–+ White’s pieces are devoid of harmony and his extra pawn is relatively meaningless. Black still has some work remaining, but his advantage should be enough to decide the game. His next move is likely to be ...Qe6, targeting c4. B2) 6.e3

White prepares to develop his bishop and castle. So far, his centre is well fortified and his darksquared bishop sits happily outside the pawn chain. 6...c5! 351

Black, in accordance with the main strategy of the Grünfeld Defence, strikes at the enemy centre, having in mind ideas such as ...Qa5 and ...Ne4. We will consider B21) 7.Qb3 followed by the main line of B22) 7.dxc5, after first checking some minor alternatives: 7.Bd3?! cxd4 8.exd4 Bg4 9.0-0 Nc6³ threatens the d4-pawn, and White already has to resort to a concession such as 10.Be2. 7.Be2?! has been a popular choice but White already risks becoming worse. 7...cxd4 8.exd4 (8.Nxd4 dxc4 9.Bxc4 Qa5 10.Nb3 Qb4 was played in Kovac – Svoboda, Slovakia 2000, when 11.Qe2N would have been White’s best try, although 11...Nh5 12.0-0 Nxf4 13.exf4 Nc6³ still gives Black an enduring advantage due to the two bishops) 8...Nc6 9.h3 (9.0-0 transposes to 9.Bf4 in the notes to variation B232 of Chapter 4 on page 113) 9...dxc4 10.Bxc4 a6 11.a4 Bf5 12.0-0 Rc8 13.d5 Nb4 14.Qb3 a5µ White’s opening had clearly been a failure in Chandran – Kruger, corr. 2014. The isolated d5-pawn blocks the bishop on c4, and Black has ideas of ...Bc2 or ...Nd7-c5. 7.Rc1?! is just bad. 7...cxd4 8.Nxd4 (8.exd4 Nc6 9.Be2 dxc4 10.Bxc4 Bg4 11.d5 Bxf3 12.Qxf3 Nd4 13.Qd1 Nd7 14.Be3 [or 14.0-0 Rc8 15.Bb3 Nc5³ and Black’s knights are very active] 14...Nf5 15.0-0 Nxe3 16.fxe3 Rc8µ Black’s superior pawn structure gave him the upper hand in Persoons – Mirumian, Brasschaat 2012)

8...Ne4! 9.cxd5 Nxc3 10.bxc3 Qxd5 11.Qb3 Qc5 12.Qc2 e5 13.Nb3 In Tebbs – Evans, England 1998, Black could have targeted the queenside pawns: 13...Qa3! 14.Bg3 Be6 15.Be2 a5µ White’s pawns are weak and ...a4 is threatened. 7.h3 can hardly cause problems. Out of several good options, I suggest: 7...Be6 8.dxc5 (8.cxd5?! Nxd5 352

9.Be5 was played in McGuinness – Baxter, Wetherby 2016, when the simple 9...Bxe5N 10.Nxe5 Qa5 11.Qd2 cxd4 12.exd4 Rd8µ would have left the d4-pawn weak) This was S.J. Silva – A.C. Cardoso, Joao Pessoa 2016, and here I recommend another novelty:

8...Ne4N Activating the g7-bishop. 9.Nxe4 dxe4 10.Qxd8 Rxd8 11.Nd4 Na6 12.Be2 Nxc5³ Black is slightly better due to more active pieces. Finally, 7.cxd5 Nxd5 only helps Black to get active, for instance: 8.Be5 (8.Nxd5 Qxd5 9.Be5 Nc6 10.Bxg7 Kxg7 11.dxc5 Qxc5 12.a3 was seen in Verastegui Noguera – Henao, Bogota 1992, when 12...Bg4!N 13.Be2 Rfd8 14.Qc1 Qxc1† 15.Rxc1 Rac8 16.h3 Bxf3 17.Bxf3 Nd4 18.Rxc8 Nxf3† 19.gxf3 Rxc8³ would have left Black with a risk-free advantage in the endgame, thanks to White’s weakened kingside structure) 8...Bxe5 9.Nxe5 Nxc3 10.bxc3 Nd7 11.Nf3 Qa5 12.Qd2 We have been following Vlcek – Vrba, corr. 2010, and here I like:

353

12...Rd8N 13.Bd3 b6 14.0-0 Bb7³ Intending ...e5, with easy play for Black. B21) 7.Qb3 The queen lunge looks active and it prepares Rd1, but Black gets a good game by continuing to play against the enemy centre. 7...cxd4 8.Nxd4 It is hardly a good time for White to take on an IQP: 8.exd4? dxc4 9.Bxc4 Nc6 10.0-0 Na5 11.Qb4 Nxc4 12.Qxc4 Be6µ Black had the bishop pair as well as the better pawn structure in Loeber – Jap Tjoen San, Haarlem 1998. 8...Nc6 9.Nxc6 Exchanging knights is logical, since Black was threatening both ...Nxd4 and ...Na5. 9.Rd1?! was played in D’Costa – Lupulescu, Chalkidiki 2000, when Black missed a strong counter:

354

9...Nh5!N 10.Bg5 h6 11.Bh4 Nxd4 12.exd4 Nf4 13.cxd5 Re8ƒ It is difficult for White to complete the development of his kingside. 9.cxd5 is well met by 9...Na5! 10.Qa3 Nxd5. For instance, 11.Rd1 Nxc3 12.bxc3 Bg4 13.f3 Bd7 14.Nb3 Nxb3 15.Qxb3 was played in Kolpak – Gilbert, corr. 2010, when Black could have emphasized his advantage with:

15...Qe8!Nµ Unpinning the bishop while threatening ...Ba4. White lags in development and his queenside pawns are obviously weak. 9...bxc6 Curiously, in this line White seems to be aiming for Grünfeld-style play against Black’s pawn centre, 355

but his pieces are not well placed for the job. 10.Be2 This move looks obvious and it has been almost universally played, but I believe it is a slight inaccuracy. 10.Rd1?! is certainly not an improvement. In Lehotzky – Kitts, Lansing 1989, Black could have seized a big advantage with: 10...Nh5!N 11.Bg5 h6 12.Bh4 g5 13.Bg3 Nxg3 14.hxg3 Bg4

15.Be2 Bxe2 16.Kxe2 Rb8 17.Qc2 Qc8–+ White has an unsecured king and the bishop is stronger than the knight. Moreover, there is an unpleasant threat of ...Qa6. 10.cxd5?! was another mistake. In Rozwadowski – Zagorsek, Graz 2013, Black should have opted for 10...Nxd5!N when his pieces develop a lot of activity:

356

11.Bg3 Nxc3 12.bxc3 Qa5 13.Rc1 Be6 14.Bc4 Bxc4 15.Qxc4 Qb5 16.Qxb5 cxb5µ Black is better for a few reasons: the c3-pawn is weak and there is a great contrast between the bishops: Black’s aims at the weak pawn, while White’s merely shoots into thin air. 10.Be5! seems like the best attempt to nullify the pressure, although only one game has reached this position so far. In Kladiva – Hevesi, Fuzesabony 2004, Black could still have fought for the advantage with:

10...Be6!N 11.Be2 dxc4 12.Bxc4 Qa5 13.Bxe6 Qxe5 14.Bc4 Rab8 15.Qc2 Rb4 16.Bb3 Rfb8³ Due to the pressure on the b-file and the threat of ...c5-c4, White still has some problems to solve. 10...Nh5! 357

Black is not afraid to ruin his pawn structure. 11.Bxh5 gxh5 The h5-pawn is not particularly weak. Moreover, Black has the bishop pair and he threatens to win a piece with ...e5. 12.Bg5 This was Gurri Mancera – Bieszk, email 2011, and here I recommend an improvement:

12...dxc4!N 13.Qxc4 No better is 13.Qc2 Qa5 14.Bxe7 Re8 15.Bh4 Bf5 16.Qc1 Bd3–+ when Black’s bishops are extremely active and White’s king is stuck in the centre. 13...Qb6 14.Qb3 Qc5 15.Bf4 Ba6µ White’s position is extremely unpleasant due to his unsafe king and Black’s powerful bishops, while Black’s fractured pawn structure has little impact on the game. B22) 7.dxc5

358

This is the most logical and popular reply. White immediately eliminates the pawn, not wishing to allow ...cxd4. On the other hand, now the bishop on g7 begins to work at full capacity. 7...Ne4! Regaining the pawn can wait – Black’s priority is to activate his pieces. Now the bishop presses unpleasantly on the long diagonal, and ...Qa5 may follow. White’s main options are B221) 8.Be5 and B222) 8.Rc1. Other continuations are clearly worse, for instance: 8.Qb3?! Nxc3 9.bxc3 occurred in Prielozny – Bukovsky, Slovakia 2009, and here I can recommend:

359

9...Nd7!N 10.cxd5 Nxc5 11.Qc4 Bf5! 12.Be5 (12.Qxc5? Rc8–+) 12...Rc8 13.Qd4 f6 14.Bg3 Ne4 15.Nd2 Qa5 16.Bc4 Nxd2 17.Kxd2 Rfd8µ Despite the fact that White has an extra pawn, his position is not to be envied, as his king is in the centre and Black’s rooks exert serious pressure. Another mistake would be 8.Nxe4?! dxe4 9.Qxd8 Rxd8 10.Nd4 e5 11.Bg5 f6 when White had two pieces under fire in Ljubic – Colard, email 2001. True, White can keep the game going with 12.Nb5 fxg5 13.Nc7, but after 13...Na6 14.Nxa8 Be6µ Black will pick up the trapped knight with an obvious advantage. B221) 8.Be5 This has been tried by some strong players but it leads to the deterioration of White’s pawn structure. 8...Bxe5 9.Nxe5 Nxc3 10.bxc3 Qa5 11.Qd4 The alternative is: 11.Qb3 Qc7 12.Nf3 dxc4 13.Bxc4 Nd7 14.c6 Qxc6 15.Rd1 15.0-0 Nc5 16.Qa3 Qc7 17.Nd4 occurred in Docx – Gordon, Gibraltar 2013, when the natural 17...b6Nƒ followed by ...Bb7 would have left Black with an edge due to his better pawn structure. 15...Nb6 16.Rd4 In Pascua – So, Tagaytay City 2010, it would have been best for Black to play:

16...Nxc4N 17.Rxc4 Be6! 18.Rxc6 Bxb3 19.Rc7 Bxa2 20.Rxe7 Bd5µ The bishop is stronger than the knight in this open position, and the passed a-pawn will be tough for White to handle.

360

11...Be6! Kaufman gives 11...f6 as leading to equality and this is certainly not a bad option, but I prefer the text move. 12.cxd5 Bxd5 13.h4 13.e4 Be6 14.Bc4 Rd8 15.Qe3 was played in De Jong – Bartos, email 1998, when an improvement is:

15...Nd7!N Based on the following point: 16.Bxe6 Nxe5 17.Bd5 e6 18.Qg3 exd5 19.Qxe5 Re8–+ White never managed to secure his king.

361

13...Nc6 14.Nxc6 Bxc6 15.Qb4 Qc7 16.h5 White’s attack is not dangerous. In Marotz – Tajti, email 2007, Black could have continued:

16...Rad8N 17.hxg6 fxg6 18.Qh4 e6ƒ White’s king is not safe and his extra pawn does not make its presence felt at all. B222) 8.Rc1

Bringing the rook into play is the most logical continuation. Its only drawback is that the a2-pawn loses a defender. 8...Nd7! 362

Despite the above remark about the a2-pawn, it is too soon to play 8...Qa5?! because 9.Be5!² neutralizes the g7-bishop. The text move controls the e5-square, thus preparing ...Qa5. The d5-pawn is left to its fate, as there are more important matters than counting pawns. This move was even played by Vladimir Kramnik in one of his rare outings on the black side of the Grünfeld, which is a particular mark of quality. 9.cxd5 While Black develops his initiative, White agrees to ‘suffer’ in the hope of consolidating and eventually prevailing with his extra material. Other continuations are clearly worse, for instance: 9.Be2?! was seen in Ivana Maria – Padmini, Jalgaon 2012, when Black’s play can be improved:

9...Nxc3N 10.bxc3 dxc4 11.c6 Nb6! 12.Qxd8 Rxd8 13.c7 Rd5 14.Nd2 Rc5 15.Ne4 Rc6µ The c7-pawn looks dangerous but its days are numbered, as Black has surrounded her with ‘love and care’ and now threatens ...e5. 9.Qxd5?! Nxc3 10.bxc3 Qa5 White’s tripled pawns, like the Tower of Babel, will soon be crumbling. 11.Qd2 (after 11.Nd2 Nxc5 12.Nb3 Nxb3 13.axb3 Bxc3† 14.Ke2 Bg4†–+ the white king succumbed to the attack in Gundavaa – Wei Yi, Tagaytay City 2013) 11...Nxc5 12.Be5 Bxe5 13.Nxe5 In Ivana Maria – Schut, Athens 2012, Black’s most convincing continuation would have been:

363

13...Bf5!N 14.Qb2 Qc7 15.Nf3 Rad8 16.Nd4 Be4µ It is difficult for White to complete the development of the kingside, and the extra pawn is not felt at all.

9...Qa5 Black ramps up the pressure on c3. Despite the fact that White temporarily has two extra pawns, it becomes obvious that he will not be able to keep the extra material. 10.Nd4 White blocks the g7-bishop and supports a possible c5-c6 in some lines. I checked several other ideas:

364

10.Bd3? is poor. 10...Nxc3 11.Qd2 (11.bxc3 Bxc3† 12.Kf1 Nxc5 13.Bc4 b5 14.Bb3 b4–+ left White facing the threats of ...Ba6†, ...Nd3, and ...Rfd8 in Jepsen – Klifoth, Helsingor 2013) 11...Qxa2 12.c6 bxc6 13.dxc6 Nc5–+ White’s position was already hopeless in Malekian – Tari, Lund 2015. 10.Bc4?! is an unfortunate choice because the bishop can be hit: 10...Nxc3 11.bxc3 Bxc3† 12.Ke2 b5! 13.Qd3 Now in Zheliandinov – Novak, Bled 2000, it would have been best to play:

13...b4!N With the unpleasant threat of ...Nxc5 followed by ...Ba6. White has no choice but to give up the exchange with: 14.Rxc3 (14.c6 Nc5 15.Qc2 Ba6–+ is hopeless for White, who will suffer material losses due to the misplaced king and the weakness of the d3-square) 14...bxc3 15.Ne5 Qxc5 16.Rc1 Nxe5 17.Bxe5 Bd7–+ White’s compensation for the exchange is clearly insufficient. Also doubtful is: 10.Nd2?! Nxc3 11.bxc3 Nxc5 12.Qc2 Defending the c3-pawn. (It’s amazing how quickly a GM with a rating close to 2600 got a losing position after 12.e4?! Bxc3 13.f3 Bd7 14.Kf2 in Belous – Chirila, Saint Louis 2017, when 14...Bd4†!N 15.Kg3 Qxa2 16.Bc4 Qa5–+ would have left White a pawn down, with little hope of stopping the triumphant march of Black’s queenside passers.) 12...Bd7 13.Nc4 Qd8 14.Be2 Bf5 15.Qd1 Be4 16.d6 This position was reached in Glotz – Schirmer, email 2006. There arises a logical question: “Why not pick up a kingside pawn, with gain of tempo, while stopping White from castling?”

365

16...Bxg2!N 17.Rg1 Bh3 18.dxe7 Qxd1† 19.Kxd1 Rfe8 20.Bd6 Rac8 21.Nd2 Bd7µ White has a terrible pawn structure and the e7-pawn will be eaten after ...Bf6. Thus, Black will restore the material balance and put pressure on the weak pawns on a2, c3, and h2. White may also protect the c3-point: 10.Qc2 Nxc3 11.bxc3 Nxc5 Black picks up one pawn and threatens ...Bf5.

12.e4 12.Nd4 covers the f5-square, but 12...e5! 13.dxe6 Bxe6 14.Nxe6 Nxe6 15.Be2 Nxf4 16.exf4 Rac8ƒ was a good response in Malar – Roques, corr. 2015. Objectively the position is equal, but in a practical game Black could try to play for a win, as he will have some chances to create a passed pawn on the 366

queenside after winning the c3-pawn. 12...Bd7 13.Bd2 Rac8 Black has obvious compensation for the pawn, due to his better development and pressure against the weak pawns on c3 and e4. 14.c4 Another good example continued: 14.Be2 Qa4! 15.Qxa4 Bxa4 16.e5 Ne4 17.Rb1 Nxd2 18.Kxd2 Rfd8 19.c4 b5ƒ Rumyantsev – Vetter, corr. 2012. White cannot hold the extra pawn and Black’s initiative is unpleasant, since the bishops are working at full power and White’s king is not well placed. 14...Qb6 15.Be3 Bb2! 16.Rb1 Bc3† 17.Bd2 Ba4 18.Qc1 Bxd2† 19.Nxd2 Qa5 20.f3 b5ƒ All this happened in Littke – Dessaules, corr. 2013. White’s kingside remains undeveloped and Black’s initiative is becoming unpleasant: the immediate threat is ...bxc4, and if Bxc4 then ...Nd3† will be painful. One other continuation has been tested at a high level: 10.Be2 Nxc3 11.bxc3 Bxc3† 12.Kf1 Nxc5 A critical position has arisen, where White has tried a few different ideas. 13.h4!? The most aggressive continuation, intending to connect the rook on h1 to a kingside attack. Other moves are no better, for instance: 13.Qc2 Na4 14.e4 Bd7 15.Qb3 Rac8 16.h4 Bg7 17.Rxc8 Rxc8 18.Qxb7 Nc5 19.Qb1 occurred in Hlavacek – A. Davies, corr. 2015, when 19...Qa4!N 20.Nd2 Bc3 21.Bd1 Qb5† 22.Be2 Qxb1† 23.Nxb1 Nxe4µ would have left Black’s pieces more active, while the pawns on d5 and a2 are weak. 13.Be5 Bxe5 14.Nxe5 was seen in Golod – Charochkina, Yerevan 2014, when the precise 14...Qa3!N 15.Qc2 b6 16.Nc4 Qb4³ would have been advantageous for Black, since White’s king is unsuccessful and the d5- and a2-pawns are weak. 13...Bd7 14.Kg1 Rac8 15.h5 I can now recommend an improvement.

367

15...Bf6!N 15...Ne4!? is another decent move which led to an eventual victory for Black in Alekseenko – Wei Yi, Moscow 2016. The text move is slightly more accurate: Black removes the bishop from the sights of the white rook and threatens ...Ne4-c3. My line continues: 16.hxg6 hxg6 17.Ne5 Ba4 18.Qe1 Qxe1† 19.Rxe1 g5! White must misplace his bishop, as placing it on g3 would invite ...Ne4. 20.Bh2 Kg7³ The pawns on d5 and a2 are weak, and the rook on h1 is shut out of play. 10...Nxc3 The knight was a good defender, so Black eliminates it to expose the pawn weaknesses on the queen’s flank. 11.bxc3 Other continuations are worse, for example: 11.Rxc3? places the rook in a nasty pin. 11...Nxc5 12.Qc2 occurred in Tremblay – Abreu Delgado, New York 2018, when Black overlooked a killing blow:

368

12...e5!N 13.dxe6 Nxe6–+ White suffers material losses. 11.Qd2?! also cannot be recommended for White. 11...Qxc5 12.bxc3 Qxd5 13.c4 Qc5 14.Be2 e5 15.Nb3 Qe7 16.Bg3 was seen in Bustamante Orta – Espinosa Veloz, Oaxtepec 2016. Black already has the better pawn structure, and he could have increased his advantage by gaining space with his pawns:

16...h5!N 17.h3 a5 18.Qc3 a4 19.Nd2 e4 20.Qc2 Nc5µ White suffers from a cramped position and a lack of good squares for his pieces.

369

11...Qxa2!? White’s pawns begin to drop like autumn foliage, and he needs to play carefully so as not to get into a worse position. 12.Nb3! This is the least popular of the three options White has tested, but I believe it is his only chance to equalize. 12.c6?! This move looks tempting and it has been the choice of several strong players, but it is a mistake. 12...Nc5 13.Be2 13.c4? e5! 14.dxe6 was played in Ftacnik – Bok, Netherlands 2017, when 14...Qa5†!N would have been killing: 15.Ke2 (or 15.Qd2 Qxd2† 16.Kxd2 fxe6 17.Bg3 e5 18.cxb7 Bxb7µ followed by ...Nb3†) 15...fxe6 16.Bd6 Rd8 17.Be7 e5 18.Bxd8 Bg4† 19.f3 Rxd8 20.fxg4 exd4 21.cxb7 Qa3!–+ With a decisive attack. 13...Qxd5 14.Bf3 Qc4 15.Be2 Qa2! 15...Qd5 led to an early draw in Ivanchuk – Gelfand, London 2013, but Black can play for more. 16.Ra1 16.Bg5?! proved a worse choice after 16...e5 17.Be7 exd4 18.Bxc5 dxe3 19.0-0 Re8 20.cxb7 Bxb7 21.Bxe3 a5 22.Ra1 Qd5µ in Simantsev – Shishkin, Police 2013. Each side has a passed pawn, but the one on a5 is dangerous whereas the one on c3 is weak. 16...Qb2 17.0-0 e5 18.Rb1 Qxc3 19.cxb7 Bxb7 20.Nb5 Qa5³ In Hjorth – Bolda, corr. 2014, White had some activity, but still faced an uphill struggle against Black’s extra pawn. 12.Qb3?! 370

This move preserves White’s extra pawn but it is too ambitious. 12...Qxb3 13.Nxb3 a5! Intending to drive the defensive knight away and gobble the c5-pawn, all while advancing the passer towards its promotion square. 14.c6 White should keep his extra pawn, even though it is not so dangerous. Other moves are worse, for instance: 14.Kd2? a4 15.Nd4 Nxc5 forces White to worry about ...Bxd4 and ...Nb3†, among other things. This was well illustrated after 16.Bc4 e5 17.dxe6 Bxe6 18.Bxe6 fxe6 19.Bg3 Bxd4 20.exd4 Nb3† 21.Kc2 Nxc1 22.Rxc1 Ra6–+ when White had no compensation for the exchange in Rauk – Sirosh, Tallinn 2016. 14.Bc4?! a4 15.Nd4 Nxc5 restores material parity while leaving Black with a powerful passed pawn. 16.Ke2 Bd7 17.Bg5 Rfe8 18.f3 was seen in Gemy Vargas – Barbosa, Asuncion 2012, when 18...e6!N would have been perfectly timed. After the further 19.dxe6 Nxe6 20.Nxe6 Bxe6 21.Bxe6 Rxe6–+ the simplifications have led to an even greater exposure of White’s weaknesses: the c3-pawn is vulnerable and Black’s a-pawn should seal his victory. 14...Nb6! Kramnik chose 14...Nf6 and went on to draw with Fridman at Dortmund 2012. The text move also attacks the d5-pawn but avoids blocking the g7-bishop, while in some lines the knight may be able to utilize the c4-square. 15.e4 Defending the d5-pawn, but Black continues the aggression. 15.cxb7? Bxb7 gives Black too much activity: after 16.e4 a4 the a-pawn will run wild. 15.c4? is another unfortunate choice which exposes the long diagonal. 15...bxc6 16.Bc7 a4 17.Nd4 Nd7 18.dxc6 Nc5 19.Bb6 Bxd4 20.exd4 Nb3 21.Rc3 a3 22.Bd3 a2 23.Ke2 a1=Q 24.Rxa1 Nxa1–+ The knight on a1 is stuck, but Black still has an extra rook and he should be able to convert his advantage with careful play. White also failed to solve his problems with 15.d6 e5 16.d7 Bxd7 17.cxd7 exf4 18.Nc5 a4 19.Bb5 a3 20.Ke2 Nd5µ in Van Schyndel – Wolfrum, email 2010. The a3-pawn is much more dangerous than the one on d7, and ...Nxc3† is threatened. 15...bxc6

371

16.Bc7 White must hurry to eliminate the main threat, namely Black’s passed pawn. The inferior 16.dxc6?! a4 17.Na1 was played in Spalir – Alsina Leal, Pardubice 2016, when it would have been advantageous for Black to secure the c4-outpost: 17...Be6! 18.Bb5 Nc4 19.0-0 Ra5 20.Bxc4 Bxc4 21.Rfd1 Rc5 22.c7 e5 23.Bg5 f6 24.Be3 Rxc7–+ The position has stabilized, and Black enjoys the bishop pair plus a mighty passed pawn, which should amount to a decisive advantage. 16...Na4 17.Bxa5 cxd5 18.exd5 It may look as if White has managed to defend himself and remain with an extra pawn, but Black’s next move shatters the illusion. 18...Nc5! Destroying White’s dark-squared bishop, along with the d5-pawn and the overall harmony of his position. 19.Nxc5 Rxa5 20.Ne4 Rxd5µ Black has two great bishops and the c3-pawn is weak. White was unable to avoid a loss of material in Swayams – Melkumyan, Riga 2014.

372

12...Qa3 13.e4 White could also consider 13.Be2!?N 13...Nxc5 14.Nxc5 Qxc5 15.0-0 a5 16.e4 a4 17.Be3 Qa5 18.c4 a3ƒ when our iron friend gives 0.00, but in a practical game I would prefer Black’s chances with the far-advanced a-pawn. 13...Nxc5 14.Nxc5 Qxc5 15.Qd2 e6 16.Be3 We have been following Huerta Navarro – Lerida Macareno, corr. 2016. Here I recommend:

16...Qa5!?N 17.Be2 Rd8 18.c4 This looks safer than 18.Bf3 Bd7 19.0-0 Rdc8÷ when Black is not worse: he has created some pressure on the c3-pawn, and in the future he can mobilize the a-pawn. 373

18...Qxd2† 19.Kxd2 a5 20.Rc2 b5 21.cxb5 exd5 22.Bg5 Re8 23.exd5 Re5 24.Rhc1 Rxd5† 25.Ke1 Bf5 26.Rc7 Bb2 27.Rd1 Rxd1† 28.Kxd1 Be5= The position remains balanced, although the mutual passed pawns require both sides to take care. My illustrative line continues:

29.Re7 Bd6 30.Rb7 Bc8 31.Rb6 Bc5 32.Rc6 Bxf2 33.b6 Rb8 34.Bf4 Bd7 35.Bxb8 Bxc6 36.Bc7 Bxg2 37.Ba6 a4 38.b7 Bxb7 39.Bxb7 Bc5= With three pawns against a bishop and little material remaining, Black is in no danger and a draw is highly likely. Conclusion This chapter has dealt with two distinct variants of the Bf4 set-up, the first of which saw him delay the development of the royal knight with 4.Bf4 Bg7 5.e3 0-0, followed by some kind of action on the queenside. The main line continues 6.Rc1 Be6! with three further branches. 7.cxd5 Nxd5 8.Nxd5 Qxd5 enables Black to activate the queen, which offers him good counterplay. 7.c5 leads to a complex strategic battle where we have the model game Gelfand – Grischuk to guide us. Finally, in answer to 7.Qb3 Black must act boldly in the centre with 7...c5!, sacrificing a pawn in return for a lasting initiative, even after a queen exchange. The second part of the chapter dealt with 4.Nf3 Bg7 5.Bf4 0-0, when 6.e3 c5! once again offers Black thematic counterplay. 7.dxc5 is the critical reply, when 7...Ne4! increases Black’s dynamic activity. White will temporarily be two pawns up, but Black’s active pieces will enable him to restore the material balance sooner or later. In many cases, Black ends up with a strong passed a-pawn which tends to make endgames more dangerous for White. 374

A) 4.Qa4†!? Bd7 5.Qb3 dxc4 6.Qxc4 Bg7 202 A1) 7.Bf4!? 203 A2) 7.e4 205 B) 4.Nf3 Bg7 5.Qa4† Bd7 6.Qb3 dxc4 7.Qxc4 0-0 208 B1) 8.g3!? 209 B2) 8.e4 210 B3) 8.Bf4 214

1.d4 Nf6 2.c4 g6 3.Nc3 d5 This chapter will deal with White’s attempts to disrupt us with an early queen check. Once again, we will consider two slightly different versions of the concept: A) 4.Qa4†!? and B) 4.Nf3 Bg7 5.Qa4†. In both cases, the play can be compared with the much more popular Russian System (4.Nf3 Bg7 5.Qb3) which will be featured in the next two chapters. A) 4.Qa4†!? White delivers the check in an attempt to break Black’s harmony. This move was introduced by the

375

Argentinean player Renato Sanguinetti against Bahamonde, at Buenos Aires in 1946. Subsequently, Korchnoi, Aleksandrov, Mamedyarov, Vallejo Pons and Ivan Sokolov have all tried it. 4...Bd7 5.Qb3 Attacking the unprotected pawn on b7, as well as d5. Occasionally White has tried: 5.Qd1 Inviting 5...Bc8 with a repetition, as has occurred a few times when the players felt like taking a day off. If you prefer to continue the fight, my suggestion would be: 5...c6 6.e3 The untested 6.cxd5N 6...cxd5 7.Bf4 was suggested by GM Nikita Petrov, reaching an Exchange Slav where Black would not normally hurry to play ...g6 and ...Bd7 – but since these two moves have been awarded ‘for free’, the outlook is not bad at all. For instance: 7...Qb6 8.Qd2 Nc6 9.e3 Rc8 10.Bd3 (10.Nf3?! Ne4ƒ) 10...Nb4 11.Bb1 Black has a choice here, with 11...Nh5 being the solid continuation. (11...Ne4!? is a riskier and more dynamic solution, which interested readers may wish to analyse for themselves.) 12.Be5 (12.Bg5 h6 13.Bh4 g5 14.Bg3 Nxg3 15.hxg3 e6÷ gives Black good long-term prospects with the two bishops) 12...f6 13.Bg3 e6 14.Nf3 Nc6÷ Black has a good position. 6...Bg7 7.Nf3 0-0 Again, we have a version of a normal opening – this time a Schlechter Slav – where Black has been given two extra tempos. The ...Bd7 move is not particularly useful (nor harmful), but castling definitely helps! 8.Qb3 8.Be2 dxc4 9.Bxc4 Bg4 gives Black a Schlechter Slav with a full extra tempo. (9...c5 10.0-0 cxd4 11.exd4 Nc6÷ is also decent, intending ...Na5 and ...Rc8, with similar play as in variation B232 of Chapter 4.) 8...Qb6 9.Qxb6 axb6 10.cxd5 Nxd5 11.Nxd5 cxd5 12.Bd2 Rc8 13.Bd3 Nc6 14.Bc3 f6 15.0-0 e6÷ Black had no opening difficulties in Keinanen – Erdos, Batumi 2019. 5...dxc4 6.Qxc4 6.Qxb7? has never been played, with good reason: 6...Nc6 7.Bf4

376

7...e5! 8.Bxe5 Rb8 9.Qxc7 Qxc7 10.Bxc7 Rxb2–+ In return for a mere pawn, Black has a decisive initiative: the primary threat is ...Bb4, and White is too far behind in development to offer much resistance. 6...Bg7 White hopes that the bishop on d7 will obstruct Black’s standard plans of development, but it also provides us with some additional opportunities for counterplay with a timely ...b5, as we will see. White can either attack c7 with A1) 7.Bf4!? or seize the centre with A2) 7.e4. 7.Nf3 converts to variation B. 7.g3 was tested once at GM level but has not been repeated. 7...Bc6 8.Nf3 Bxf3 9.exf3 occurred in I. Sokolov – S. Ernst, Boxtel 2011. White has the bishop pair but they are unable to show all their power, and White’s pawn structure is not so healthy. I recommend:

377

9...e6!?N (9...0-0N is also fine and is quite likely to transpose) 10.Bg2 (10.Qb5† Nc6 11.Qxb7 Nxd4 12.Bb5† Kf8 is somewhat better for Black, although you can castle before playing ...e6 if you prefer to avoid this option altogether) 10...0-0 11.0-0 Nc6 12.Rd1 Ne7 13.Qb4 Rb8³ White’s pawn structure has been damaged and Black has firm control over the blockading square on d5, so his chances are somewhat preferable. A1) 7.Bf4!?

7...Na6 Black defends the c7-pawn and prepares to challenge White’s centre with ...c5, which will be the response to practically any move that White might try. 378

8.Nf3 8.e4 occurred in Papadopoulou – Pina Sierra, Oropesa del Mar 2000, when it was logical to strike at the centre with: 8...c5!N 9.d5 (9.dxc5?! is well met by 9...Rc8 since after 10.b4? Nxe4–+ White loses material; alternatively, after 9.Nf3 Qa5 10.Bd2 cxd4 11.Qxd4 Nb4ƒ White’s queen is not well placed and Black gains a lot of activity) 9...0-0 10.Be2

10...b5! It turns out that the d7-square is not such a bad place for the bishop. 11.Nxb5 Nxe4! 12.Qxe4 Bxb2 13.Rd1 Bxb5µ Black has an extra pawn as well as a nasty initiative. 8.Qb3 looks too passive, and 8...c5! 9.d5 b5! once again justifies the location of the bishop on d7. Then 10.Nxb5 Qa5† 11.Nc3 c4! 12.Qc2 Nxd5 13.Bd2 Nab4 14.Qc1 was the continuation of Guimard – Sanguineti, Buenos Aires 1968. Black has a dominating position and he won the game anyway, but the most convincing and beautiful route to victory would have been:

379

14...Ba4!N 15.Nxd5 c3!!–+ Breaking White’s defence. 8.Rd1 is also met by 8...c5!, when White may have trouble justifying the decision to place the rook on d1 so early. 9.e3 Qa5 10.Qb3 was seen in Julve – Lucas, France 2003, when it would have been logical to finish development:

10...0-0N 11.Nf3 cxd4 12.exd4 Rac8 It is difficult for White to complete the development of his kingside. For example, 13.Be2? would be hit by 13...Nb4! 14.0-0 Be6 15.Qa4 Qxa4 16.Nxa4 Bxa2µ when Black has an extra pawn as well as the better structure. 8...c5! In light of the above variations, this move should come as no surprise. 380

9.e3 9.dxc5 Rc8 is excellent for Black, since 10.b4 Be6 11.Qb5† Qd7 leaves White unable to keep his queenside together. 9...0-0 10.Be2 An instructive error is 10.d5? Qa5! 11.Nd2 as played in Harazinska – Krainski, Lazy 2014, when Black missed a good opportunity:

11...Nb4!N 12.Nb3 Qb6 13.Rc1 Rac8 Black threatens ...e6 when White’s centre falls apart, and 14.a3 loses to 14...Nbxd5! 15.Nxd5 Nxd5 16.Qxd5 Be6–+ when losing a pawn will be the least of White’s worries. 10.Qb3 is not so terrible although after 10...cxd4 11.exd4 Bc6 12.Be2 Nc7 13.0-0 Ncd5 14.Be5 Qb6³ there was no need to agree a draw in M. Gurevich – Akesson, Antalya 2004, as Black’s position is somewhat better due to the isolated d4-pawn and the active disposition of the pieces.

381

10...cxd4 11.Qxd4 Qa5 12.0-0 Rfd8 13.Bxa6 bxa6! 13...Qxa6 is perfectly playable for Black but the text move is more ambitious and simply a better move. Black keeps her queen active and prevents White’s queen from shifting to b4. 14.Rfd1 Be6 15.b4 Rxd4 16.bxa5 Rc4³ White faced a difficult endgame in Tica – Videnova, Malinska 2013, due to Black’s bishop pair and the active rook on c4. A2) 7.e4

White seizes space in the centre. Indeed, the e4- and d4-pawns look menacing, but Black need not 382

fear them. Rather, he will simply complete development as quickly as possible and then look to counterattack. 7...0-0 8.e5 White takes even more space, while hoping to shut the g7-bishop out of the game. On the other hand, the d5-square is weakened and the d4-pawn may become weak in the future. It should be noted that White is far behind in development, making it quite risky to make so many pawn moves early in the opening. Naturally I checked some other possibilities as well: 8.Nf3 transposes to variation B2 on page 210. 8.a4N prevents any ...b5 ideas but White spends a valuable tempo on a pawn move while behind in development. Black gets an excellent game after 8...Na6 9.Nf3 c5 10.Be2 cxd4 11.Qxd4 Bc6³ when he remains ahead in development, while the e4-pawn is weak. 8.Be2 does not change much. In Sormani – Chrestani, Brazil 1993, Black missed a powerful idea:

8...b5!N 9.Nxb5 Nxe4 The justification of Black’s play is revealed after: 10.Nxc7? (White should try something else although he is clearly worse in any case, for instance: 10.Nf3 Nd6 11.Nxd6 cxd6 12.00 Nc6 13.Be3 a5µ and Black will target the b2- and d4-pawns; or 10.Bf3 Bc6 11.Be3 Qd7 12.Nc3 Nd6 13.Qd3 Nf5µ when Black is ahead in development and has pressure against the weak d4-pawn) 10...Nc6 11.Nxa8 Qa5† 12.Kf1 Nxd4 Despite being a rook up, White is in serious trouble, the principal threats being ...Bb5 and ...Rc8. A sample line continues 13.Qa6 Qf5 14.Be3 Nc2 15.Qd3 Bb5!–+ when White must lose material while Black keeps his attack. 8...Be6! Black responds to the threat by making a threat of his own. This small finesse secures him an outpost 383

on d5. 9.Qb5 Attacking the unprotected pawn on b7 is the only challenging idea. 9.exf6? trades the queen for three pieces, but Black gets two extra pawns included in the bargain: 9...Bxc4 10.fxg7 Kxg7 11.Bxc4 Qxd4 12.Be2 Nc6 13.Nf3 Qb6 14.0-0 Rad8

15.Rb1 (or 15.Bb5 Nd4 16.Nxd4 Qxd4 17.Re1 c6 18.Ba4 Rd7µ when Black’s material advantage should prevail) 15...e5 16.Re1 This occurred in Lev – Huzman, Israel 1999, and now I agree with Avrukh’s suggestion of 16...f6!N 17.b4 Nd4 18.Bd1 Qa6µ. White has no serious counterplay, and Black’s queen and two pawns should eventually prove too much for the three minor pieces. 9...Nd5 10.Qxb7

384

10...Nb6! An important multifunctional move. Apart from defending the a8-rook, Black threatens to trap the queen with ...c6 followed by ...Bc8, while simultaneously attacking the d4-pawn. The inaccurate 10...Nxc3?! 11.bxc3 Bd5 12.Qb5÷ left Black with a certain amount of compensation for the pawn in Korchnoi – Kulaots, Puhajarve 2011, but the text move is much stronger and leads to Black’s advantage. 11.Qe4 Evacuating the queen from the danger zone and defending the d4-pawn, although the queen remains a target for Black’s pieces. 11.Nf3? is of course impossible due to 11...c6 12.d5 Bc8, winning the queen. 11...Bf5 12.Qh4 Nc6 Black has completed his development, whereas White has moved his queen seven times in the first twelve moves, while most of his other pieces remain in their starting positions. The main threats are ...Nxd4 and ...Nb4. 13.Bg5!? White is in trouble no matter what he does, so this seems like a reasonable attempt to confuse matters. 13.Be3 Nb4 might lead to a crushing attack after: 14.0-0-0?! (14.Rc1 would postpone the end although 14...Nc2† 15.Rxc2 Bxc2µ leaves White with clearly insufficient compensation for the exchange)

385

14...Rb8 15.Nf3

15...Na4! 16.Nxa4 Qd5 17.Nc3 Nxa2†–+ With a decisive attack. After the text move Black does not desperately need an improvement, but I will present a novelty all the same.

13...Nxd4!N This is the strongest and most stylish winning line. There is nothing wrong with the game continuation of 13...Qxd4, which represents a slower but more straightforward way to realize Black’s advantage. 14.Qxd4 (14.Nf3 Qxh4 15.Bxh4 Nxe5 16.Nxe5 Bxe5µ leaves Black with a healthy extra pawn) 14...Nxd4 15.0-0-0 Ne6 16.Be3 Bxe5–+ Chekhov – W. Schmidt, Polanica Zdroj 1981. 386

14.0-0-0 Bxe5 15.Bxe7 Qe8 16.Bxf8

16...Na4! This is the only winning move, so Black can certainly be forgiven for choosing the more prosaic winning line in the aforementioned game. 17.Nge2 Nxe2† 18.Bxe2 Nxc3 19.bxc3 Qxf8–+ Black’s bishops drill the queenside with decisive effect. The general conclusion is clear: if White follows the early Qa4† by hurrying to occupy the centre with pawns and neglects to develop his knight to f3, he is liable to reach a dangerous position where he lags far behind in development, while his ‘active’ queen presents a target which enables Black’s pieces to gain additional time. B) 4.Nf3 Bg7 5.Qa4†

387

The check on a4 is more logical in conjunction with 4.Nf3, compared to the previous variant where the check occurred on move 4. This way, White does not forget about the development of his kingside and defends the d4-pawn for the future. Flohr was the first to play this sequence, against Boleslavsky at Kiev 1944. The Czech grandmaster used it in several more games, so the variation bears his name. It was subsequently taken up by Petrosian, Ragozin, Gennady Kuzmin and Uhlmann. Among modern GMs, Nakamura, Wang Hao, Grischuk and Vitiugov have played it, although the elite players seem to have abandoned it in the past few years, as practice shows that White takes more risks than Black. 5...Bd7 6.Qb3 dxc4 7.Qxc4 7.Qxb7? is a bad idea: 7...Nc6 8.Bf4 Rb8 9.Qxc7 Qxc7 10.Bxc7 Rxb2 11.0-0-0 Rb7 12.Bg3 Bf5µ reached a situation with level material where White’s king was far from safe in Trmal – Hollis, Leningrad 1960. 7...0-0 We have reached the main branching point for the Flohr Variation. It is worth analysing the rare B1) 8.g3!?, followed by the ambitious B2) 8.e4 and finally the main line of B3) 8.Bf4. The modest 8.e3 gives Black easy play with 8...Na6 9.Be2 c5! 10.0-0 cxd4 11.exd4 when, in Reshevsky – Benko, New York 1960, it would have been good to play:

388

11...Qb6!N 12.Be3 Rfc8 13.Qb3 Qxb3 14.axb3 Nb4³ The isolated d-pawn and doubled b-pawns do not decorate White’s position. 8.Bg5 was played in Kreizberg – Aharon, Israel 2019, and a few other games. I recommend improving with:

8...b5!N 9.Nxb5 (9.Qb3 is hit by the unpleasant 9...c5! 10.dxc5 Qa5µ with threats of ...b4 and ...Ne4) 9...Ne4 10.Bf4 (the greedy 10.Nxc7 leads to pain and suffering for White after 10...Nc6 11.Nxa8 Nxg5 12.Nxg5 Qa5† 13.Qc3 Qxg5ƒ when the d4-pawn and the knight on a8 are both under fire) 10...c5! 11.e3 cxd4 12.Bd3 Qa5† 13.b4 Qxb5 14.Qxb5 Bxb5 15.Bxe4 Bc6 16.Bd3 dxe3µ Black emerges from the tactical skirmish with an extra pawn.

389

B1) 8.g3!? This rare move is worth checking, but Black easily get a comfortable game. 8...Be6 9.Qa4 Nd5 10.Bg2 Nc6 The position resembles a variation of the Fianchetto System, which arises after 3.g3 Bg7 4.Bg2 d5 5.cxd5 Nxd5 6.Nf3 Nb6 7.Nc3 0-0 8.0-0 Nc6 9.e3. This does not form part of our repertoire (3...c6 followed by ...d5 being our choice in Chapter 2) but two extra tempos should more than make up for the change of set-up. 11.0-0 11.e3?! Nb6 12.Qd1 reaches a standard position, except that normally White would have castled while Black’s bishop would still be on c8. 12...Qc8 13.0-0 Rd8 14.Ng5 Bc4 15.Re1 was seen in Perisic – Plenkovic, Bosnjaci 2014, when the optimal time arose for Black to strike in the centre:

15...e5!N 16.d5 e4 17.Bxe4 Bxc3 18.bxc3 Bxd5 19.Qf3 Ne5 20.Qf4 Bxe4 21.Qxe5 Bc6 22.Qf4 Qf5!µ Black’s bishop is much stronger than its counterpart, his pawn structure is better and he controls the only open file. 11.Ng5?! is a questionable attempt to get active at the expense of a pawn. 11...Nxd4 12.Nxe6 was seen in Bogdanovich – Pavlov, Konya 2018, when it would have been best to play:

390

12...fxe6!N 13.Bd2 c5µ White’s compensation for the pawn is clearly not enough. Black has perfectly placed his knights in the centre, and the doubled e-pawns control some important squares as well.

11...Nb6 12.Qd1 White also has problems after: 12.Qc2 Nxd4 13.Nxd4 Qxd4 14.Bxb7 Rab8 15.Bg2 Nc4 16.Rb1 (16.Rd1?! is worse due to 16...Na3! 17.Qd2 Qxd2 18.Rxd2 Nc4 19.Rc2 Nxb2 20.Bxb2 Rxb2 21.Rxb2 Bxc3µ when Black regained the exchange while keeping an extra pawn in Dantas – Tsuboi, Brazil 1998.) 16...Rfd8 17.Rd1 Qc5 18.b4 Rxd1† 19.Qxd1 Qd6 20.Qxd6 Nxd6 21.Nd5 Kf8 22.Be3 c6 23.Nf4 Bxa2 24.Rc1 Rxb4 25.Bxa7 We have been following Wolkenstein – G. Mueller, corr. 2007, and here 25...Bh6!?N³ would be my suggestion. Perhaps White can make a draw, helped by the small amount of material remaining, but defending such a position is hardly a dream outcome from the 391

opening. 12...Nxd4 13.Nxd4 Qxd4 14.Qxd4 Bxd4 15.Nb5 Another game continued 15.Bxb7 Rab8 16.Ba6? (16.Bc6N is a better try although 16...Nc4ƒ still gives Black useful pressure against the b2-pawn) when, in Solmajer – Tomazini, Otocec 2016, Black could have maximized his advantage with:

16...Nd5!N 17.Nxd5 Bxd5 18.Rd1 c5 19.e3 Bf3µ Black has a clear advantage in piece activity. 15...Be5 16.Bxb7?! White should prefer 16.f4N 16...c6 17.fxe5 cxb5 18.Bxb7 Rab8 when White’s two bishops offer some compensation for the doubled e-pawns, although I still slightly prefer Black. 16...Rab8 17.Bf3 Nc8! 18.Nc3 18.a4 a6 19.Nc3 Nd6 is similar.

392

18...Nd6!N 18...Rd8 gave Black slightly the more pleasant game in Pirtskhalava – Kasparian, Tbilisi 1949, but the text move is more accurate. 19.Na4 It is hard to suggest another way of organizing White’s queenside, since 19.Nd5 Bxd5 20.Bxd5 Bxb2 wins a pawn. 19...Rfc8µ Intending ...c5, with clearly better prospects for Black. B2) 8.e4

393

This ambitious continuation is risky for White, as his beautiful pawn centre immediately comes under fire. 8...b5! We already saw several examples of this move in the non-Nf3 part of the chapter, and it is an important resource here as well. 9.Nxb5 This is the usual continuation and for good reason, as any attempts to keep White’s pawn centre intact are doomed to fail. 9.Qd3 has only been tried in one game. 9...b4 10.Nb1?! (10.Nd1N is preferable although 10...c5 11.d5 Re8 12.Ne3 e6 13.d6 Bc6 14.e5 Nfd7 15.Nc4 Nb6 16.Bg5 Qc8 17.Nfd2 N8d7³ still favours Black; the d6-pawn is securely blockaded and the e5-pawn is weak, while Black is significantly ahead in development) In Mohota – Grover, New Delhi 2013, Black could have obtained a big advantage with:

394

10...c5!N 11.dxc5 Qc7 12.Be2 a5 13.0-0 Na6 14.Be3 Ng4 15.Bd4 Bb5 16.Qxb5 Bxd4 17.Nbd2 Bxb2 18.Rab1 Bc3µ Black’s last move threatens ...Bxd2, and the weak c5-pawn is another problem for White. Black also does well after: 9.Qb3 c5! 10.dxc5 It is more dangerous to play 10.e5 Ng4 11.Bxb5?! (11.dxc5 Na6 transposes to the line below) 11...cxd4 12.Nxd4 Bxb5 13.Ndxb5 a6 14.Na3 Qd4 15.Qc2 Nc6 16.Qe2 Qxe5 17.Qxe5 Ngxe5 18.0-0 Nd3µ when White’s strong centre was a distant memory and Black was dominating in Hübner – Kasparov, Brussels 1986. 10...Na6 11.e5 Ng4 12.Bxb5 12.h3?! Nxe5 13.Nxe5 Bxe5 14.Be3 occurred in Ubilava – Kengis, Kiev 1984, when 14...Qa5!N would have been strongest. For instance: 15.Be2? (15.0-0-0 is the lesser evil although 15...Bc6µ still gives Black excellent attacking chances) 15...Be6 16.Qc2 b4 17.Ne4 b3† 18.Qd2 Nb4 19.0-0 bxa2–+ 12...Nxc5 13.Qc4 Bxb5 14.Qxc5 Ba6 15.h3 Rc8 16.Qd5 16.Qxa7? proved to be too greedy after 16...Qa5 17.hxg4 Rxc3 18.Bd2 Qb5 19.Qxa6 Qxa6 20.Bxc3 f6–+ in Ciciotti – Wegelin, corr. 2012. Materially White is still okay, but his exposed king will be his undoing. Now in Lubczynski – W. Schmidt, Poznan 2004, Black missed a spectacular resource:

395

16...Nxe5!!N 17.Nxe5 Qb6 Black threatens to regain the piece with ...e6 or ...Bb7. The knight cannot retreat from e5 due to ...Rxc3, so it is difficult for White to complete development. 18.Bd2 The main alternative is: 18.Nd7 Qc7! 19.Nb5! A nice resource, but after 19...Bxb5 20.Qxb5 Rfd8 21.Be3 Rxd7µ Black remains clearly better: his pieces are more active and the b2-pawn is weak. 18...e6 19.Qe4 f5 20.Qe3 White has no choice but to return the extra piece. 20.Qf4? Rfd8 should be winning for Black. One illustrative line is21.0-0-0 Rd4 22.Qe3 f4 23.Qe1 Rb8 24.b3 Qc5 25.Nf3 Qa3† 26.Kb1 Rxb3† 27.axb3 Bd3# 20...Qxe3† 21.Bxe3 Bxe5µ Black keeps the advantage of two bishops in an open position, along with unpleasant pressure along the c-file.

396

9...Nxe4 Black’s scheme involves sacrificing the c7-pawn in order to seize the initiative. 10.Qxc7 White will be under some pressure no matter what, so he may as well capture the pawn while he has the chance. 10.Nxc7? is too greedy. 10...Nc6 11.Nxa8 Qa5† 12.Bd2 Nxd2 13.Nxd2 Nxd4 14.Nc7 occurred in Almeida Quintana – Herrera, Santa Clara 2002. White has an extra rook, but one has to take into account his lack of development, vulnerable king and the fact that the c7-knight cannot easily escape. Black has a choice of strong continuations but the most accurate is:

397

14...Rb8!N 15.Bd3 Bc6 16.0-0 Qxd2–+ White has castled but his king is anything but safe. For instance, 17.Rfd1 Qg5 18.Bf1 Qh4 19.Nd5 Rxb2 20.Nxe7† Kf8 21.Nxc6 Nf3† with mate to follow. White’s other option is to decline the sacrifice with: 10.Bd3 Nd6 11.Nxd6 cxd6 Black improves his structure while the d4-pawn remains isolated and weak, and the pressure along the b-file will also be unpleasant. A good illustrative example continued: 12.0-0 12.Be4 Be6 13.Qa4 was played in Zhou Jianchao – Ganguly, Subic Bay 2009, when Black retreated his bishop to d7 and a draw by repetition ensued. Instead, 13...d5!N 14.Bd3 Qb6 15.0-0 Bd7 16.Qb3 Nc6³ would have left White with difficulties concerning the weak d4-pawn and the pressure along the b=file (or the a-file, in the event that White exchanges on b6). 12...Nc6 13.d5 Ne5 14.Nxe5 Bxe5 15.Qh4?! This move looks too ambitious, although White was slightly worse no matter how he played. 15...Rb8 16.Rb1 Bf6 17.Bg5 Bxg5 18.Qxg5

398

18...e5! 19.Qxd8 Rfxd8 20.b3 Bb5 21.Rbd1 f5 22.f3 Rdc8µ Black had excellent winning chances in Boensch – Mohr, Berlin 1990, due to his superior pawn structure and control over the c-file.

10...Nc6 Black has no qualms about a queen exchange, as his remaining pieces are extremely active. 11.Bd3 White does not hurry to trade queens, which would only activate an enemy rook, but instead develops a piece while attacking Black’s knight.

399

Black has easy play after 11.Qxd8 Rfxd8 12.Bd3 (or 12.Nc3?! Nb4µ when the threat of ...Nc2† is unpleasant) as seen in M. Richter – Seger, Travemuende 2002, when Black’s play can be improved with:

12...Bf5!N 13.g4 Ng5 14.gxf5 Nxf3† 15.Kf1 Rac8ƒ Black’s pieces are fully mobilized and it is obvious that the d4-pawn will fall. White will face a difficult battle for a draw, due to his disharmonious pieces and multitude of pawn islands. 11...Nb4! An excellent, dynamic move to develop Black’s initiative. 12.Bxe4 Bxb5 Now White must decide whether or not to exchange queens. 13.Qxd8! This way, he minimizes his disadvantage. An instructive error is: 13.Qc3? Such a move is understandable, as White avoids activating an enemy rook by trading on d8, while gaining a tempo against the knight and generally hoping to protect his weaknesses with the help of the strongest piece. However, White’s king remains stuck in the centre and the presence of queens leads to dangers which outweigh the aforementioned benefits. 13...a5! Black’s strategy is based only on moving forwards. 13...Qd6!? is also promising but it is more natural to secure the knight with a pawn, thus keeping the queen available for other duties. 14.a3 400

White was correct to avoid 14.Bxa8? Qxa8 15.Be3 (or 15.a3 Rc8 16.Qxc8† Qxc8 17.axb4 Qc4–+ with mate in a maximum of four more moves) 15...Rc8 16.Qb3 Bc4 17.Qd1 Ba6 18.Rc1 Nd3† 19.Kd2 Rxc1 when he can resign. In Rasmussen – Schandorff, Denmark 2010, Black could have won with a natural continuation:

14...Rc8!N 14...Qd7?! was the inaccurate continuation of the game, although White was still under pressure and did not manage to solve the considerable practical problems in front of him. 15.Qd2 f5 16.Bb1 Qc7 17.Bxf5 Rxf5 18.axb4 Qc4 19.bxa5 Bh6 20.Qd1 Qe6† 21.Be3 Bxe3–+ White’s king is caught. 13...Raxd8 14.Bd2 f5 15.Bxf5 Rxf5 16.Bxb4 Black took the d4-pawn in Giorgadze – Khalifman, Germany 1998, and one other game. My improvement is:

401

16...e5!N A winning attempt. 16...Bc6!?N looks like a safe route to simplifications and a likely draw. 17.dxe5 Bxe5 18.Rd1 Bd3 19.Bc3 Bf4 The point is that White is forced to sacrifice the exchange in order to avoid bigger problems. 20.Rxd3 Rxd3 21.Ke2 Rd8³ White has a knight and two pawns for a rook, so materially he is not doing badly at all. However, Black’s pieces are more active and the knight may become overstretched in a future endgame against a rook with pawns on both sides, so Black can keep playing for a win. B3) 8.Bf4

402

This is the main line, which is understandable when you consider the effectiveness of the ...b5 strike in the previous variation. 8...Na6! Just as in variation A1, Black uses the knight to defend the c7-pawn while preparing counterplay with ...c5. It should also be noted that Black has already developed all four of his minor pieces and castled, so he is well prepared to attack White’s centre. 9.e4 White seizes the centre and gives himself the option of e4-e5. Black needs to play actively so as not to fall under the clamp. Fortunately, we are in the convenient position of being able to play ...c5 against virtually any move White may try, as the following lines demonstrate: 9.e3 c5! takes us back to variation A1 – see page 204 for the continuation. 9.Rd1 White makes a prophylactic move in anticipation of ...c5, but we can play it anyway. 9...c5! 10.dxc5 10.d5? occurred in Zilberman – Kulkarni, Arinsal 2009, when Black missed a powerful sequence: 10...b5!N 11.Nxb5 Qa5† 12.Nc3 Rab8!! Having already sacrificed a pawn, Black calmly gives up the exchange in order to penetrate along the b-file. 13.Bxb8 Rxb8 14.Rd2 Rb4 15.Qd3 Ne4! 16.Nxe4 Rxb2 Despite being a rook up, White is helpless, for instance: 17.e3 c4 18.Qxc4 Rb1† 19.Ke2 Bb5–+ 10...Qa5 11.Nd4 No better is 11.e4 Rac8 12.e5 Rxc5 13.exf6? Rxc4 14.fxg7 Re4† 15.Be3 Rd8 when White had no compensation for the squandered material in Ermenkov – Orev, Sofia 1973. Here I propose an improvement over Remlinger – Yankovsky, Los Angeles 2012. 403

11...b5!N Opening lines on the queenside and luring White’s knight to an unfortunate square. 12.Ndxb5 Rfc8 13.e3 Rxc5 14.Qa4 Qxa4 15.Nxa4 Bxb5 16.Nxc5 Bxf1 17.Kxf1 Nxc5µ Two knights do not always fare well against a rook and pawn – but on this occasion, Black’s superior development and coordination allow us to rate his chances as significantly better.

9...c5! Again we see a timely blow to the centre. Now the black queen can join the action via a5 and the queen’s rook is ready to go to c8. 10.e5 404

This has been the most popular choice, although it is risky to start such an attack when behind in development. Moreover, the d5-square is weakened. 10.d5?! does nothing to ease White’s position. 10...b5! is thematic and powerful, and after 11.Nxb5 as played in Demchenko – Gabrielian, Armavir 2010, I can recommend:

11...Nb4!N 12.a4 Nxe4! 13.Be2 (13.Qxe4? is impossible due to 13...Bf5–+ followed by a fork on c2) 13...a6 14.Nc7 Ra7 15.0-0 Bf5 16.g4 Nd6 17.Bxd6 exd6 18.Nxa6 Nxa6 19.gxf5 Bxb2 20.Rab1 Bf6µ Black has a vastly better pawn structure. White can also consider: 10.dxc5!? This rare move was played in Mihajlov – Van Haastert, Novi Sad 2016, and a few other games. White immediately relieves the tension. I recommend: 10...Qc8!?N 10...Be6N is also possible, when 11.Qb5 Bd7 12.Qxb7 is covered in the next chapter in Game 2 – see page 228 for the continuation from here (with one less move having been played). The only drawback of that line is that it might lead to a quick draw if White moves his queen between c4 and b5, so the text move is more suitable for those looking for a fighting game. 11.Rd1 11.e5 creates additional weakness. My analysis continues 11...Be6 12.Qb5 Nc7 13.Qa4 Nfd5 14.Nxd5 Nxd5 15.Rc1 Nxf4 16.Qxf4 Bxa2 17.Bc4 Bxc4 18.Rxc4 Rd8 19.0-0 Rd5³ when White’s position looks overextended due to the pawns on c5 and e5, while Black also has the stronger minor piece. 11...Be6 12.Qd4 Nd7 13.e5 Ndxc5³ Black enjoys a lead in development and is poised to gain more time with ...Rd8, knocking the queen off its active post.

405

10...Be6! An excellent multifunctional move, improving the bishop and fighting for the d5-square. 11.exf6!? White elects to trade his queen for three minor pieces, which makes for an interesting game, although Black gets an extra pawn as part of the bargain, which tilts the balance in his favour. 11.Qa4 is safer for White but also does nothing to challenge Black. 11...Nd5 12.Nxd5 occurred in Vila Gazquez – Barbero Sendic, Spain 2012. Black recaptured with the bishop, which is certainly not bad, but I slightly prefer:

12...Qxd5!N 13.Bxa6 bxa6 14.dxc5 Qxc5 15.0-0 Qb5ƒ Black has the easier game thanks to his pair of bishops. 11...Bxc4 12.fxg7 Kxg7 13.Bxc4 cxd4

406

14.Be5† White has two other possible ways to pick up the d-pawn: 14.0-0-0 Qa5 15.Rxd4 Rfd8 16.Re1 Rxd4 17.Be5† Kf8 18.Bxd4 occurred in Renette – Hoffmann, Belgium 2012, and here I like:

18...Nb4!N 19.a3 Nc6 20.Be3 Kg7³ White’s three pieces are not quite the equal of Black’s queen and pawn. 14.Rd1 Qa5 15.Be5† f6 16.Bxd4 e5 17.Be3 Rfd8 18.0-0 Rxd1 19.Rxd1 Rd8 20.Rxd8 (I also considered 20.Rc1N 20...Qb4 21.b3 Nc7 22.Ne1 b6ƒ when White has a difficult defensive task in front of him) 20...Qxd8 21.Bxa7 White won a pawn back in Kantor – Dolgener, Budapest 2011, but Black could 407

have forced a positional concession with:

21...Qa5!N 22.Be3 Qb4 23.Bb3 Nc5 24.Bxc5 Qxc5³ Trading the knight for the bishop clearly helps Black. Next he will advance his kingside pawns, and White will have to defend well to draw. 14...f6 15.Bxd4 e5 16.Be3

16...Nb4! The knight has done its job at the edge of the board, so now it heads for c2 in order to make a favourable trade. 17.0-0 Nc2 18.Rad1 Qc7 19.Bb3 Nxe3 20.fxe3 Rac8 21.Rf2 408

Another good example continued 21.Kh1 Rfd8 22.Nd5 Qc5 23.e4 a5 24.Rfe1 Qb5 25.Nc3 and now in Kaidanov – Robson, Saint Louis 2012, Black could have underlined his advantage with:

25...Qa6!Nµ Intending ...b5. After the text move, I see no reason not to deploy Black’s ‘lazy’ rook on the open file.

21...Rfd8!N 21...Qc5?! looks somewhat artificial, although Black remained slightly better in Galliamova – Lagno, Rostov-on-Don 2011. 22.Nd5 Qc6 23.e4 a5 24.g3 b5µ 409

Black’s plan includes the gradual advancement of her queenside pawns, and rook exchanges will also help to expose weaknesses for the queen to exploit. Conclusion The Qa4† lines are similar to the Russian System of the next two chapters, but with the difference that Black’s bishop stands on d7. White’s general idea is to lure the bishop away from b7, but its presence on d7 has its upside as well, as it supports the ...b5 move. We started by analysing 4.Qa4†!? Bd7 5.Qb3 dxc4 6.Qxc4 Bg7, followed by either 7.Bf4!? or 7.e4. In both cases, Black has a lot of counterplay and White suffers from slow development, especially on the kingside. White tends to do better with 4.Nf3 Bg7 5.Qa4† Bd7 6.Qb3, although 6...dxc4 7.Qxc4 0-0 still offers Black fine prospects. 8.g3!? leads to a version of the Fianchetto System with two extra tempos for Black, while 8.e4 b5! sees Black sacrifice a pawn to gain a powerful initiative. 8.Bf4 seems like White’s best bet, although 8...Na6 followed by ...c5 gives Black an easy game, with good chances to fight for the advantage. Since Black has good chances to seize the initiative in all these variations, we may conclude that the plan of Qa4† followed by Qb3 is not dangerous, but rather a degraded version of the Russian System, which we will consider in the next two chapters.

410

A) 4.Qb3!? dxc4 5.Qxc4 Bg7 6.e4 0-0 220 A1) 7.Bf4 220 A2) 7.Be2!? 221 B) 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 224 B1) 7.Bf4 Na6 8.e4 c5! 225 B11) 9.Rd1!? 226 B12) 9.dxc5 Game 2 227 B2) 7.e4 a6! 233 B21) 8.Qa4!? 236 B22) 8.Bf4 b5! 9.Qxc7 Qxc7 10.Bxc7 Bb7 238 B221) 11.e5 241 B222) 11.Bd3 243

1.d4 Nf6 2.c4 g6 3.Nc3 d5 In this and the next chapter, we will deal with White’s early queen sortie to b3 (without the 411

preliminary check on a4), which puts pressure on the d5-pawn in order to provoke an exchange on c4. Thereafter, he will capture the centre with e2-e4, so Black needs to act enterprisingly, energetically and dynamically in order not to fall under the positional clamp. White’s scheme is known as the Russian System, due to the influence of the various Soviet players who developed it over the years. Just as in previous chapters, we will start by analysing the immediate A) 4.Qb3!? before moving on to B) 4.Nf3 Bg7 5.Qb3. The latter is the Russian System proper, and coverage of it will run into the next chapter as well. However, the non-Nf3 version is so closely related that it makes sense to group the two variants together. A) 4.Qb3!? According to my database, Botvinnik was the first to play this move, against Levenfish at Leningrad 1933. The 4.Nf3 Bg7 5.Qb3 version has already been played a year earlier, by Ragozin against Romanovsky, also in Leningrad in 1932. Ragozin was Mikhail Moiseyevich’s second around that time, so the two of them may have analysed White’s system together. The text move was subsequently taken up by Levenfish, Euwe, Reshevsky, Makogonov, Flohr and Petrosian. It continues to be used by many strong grandmasters, including Carlsen, Nakamura, Topalov, Kramnik, Svidler, Yakovenko and Ivanchuk. 4...dxc4 5.Qxc4 Bg7 6.e4 White captures the centre, but at a price: his development lags behind and the exposed position of his queen will enable Black to gain additional time. A double-edged struggle lies ahead, and it remains to be seen whose trumps will prove stronger. 6.Nf3 leads straight to variation B. 6.Bf4 Na6 should also transpose in short order: 7.e4 0-0 is variation A1 below, while 7.Nf3 0-0 is variation B1. 6...0-0 We will look at A1) 7.Bf4 followed by the most popular A2) 7.Be2!?. 7.Nf3 leads to variation B2. 7.e5 has been played in a bunch of games and Black retreated his knight in all of them, but a better move is 7...Be6!N, transposing to variation A2 of the previous chapter, with one less move played. A1) 7.Bf4 White develops the bishop and attacks the c7-pawn. 412

7...Na6 A logical continuation that combines development with supporting the thematic ...c5 break. 8.Rd1 8.Nf3 transposes to variation B1. The text move is a reasonable attempt to obstruct Black’s counterplay, using the spare tempo gained by delaying Nf3. However, Black can obtain good play with: 8...c6! Preparing the queen’s sortie to a5, and only then playing ...c5 to attack the centre. 9.Nf3 Qa5 10.Qd3 In Ushenina – Harika, Dilijan 2013, it would have been logical to continue with:

10...c5!N 11.d5 11.Qb5 transposes to the note on 10.Qb5 in the later variation B11 on page 226, with one extra move having been played. 11...Nb4 12.Qb1 Nxe4! Initiating complications which are by no means unfavourable for Black. 13.Qxe4 Bf5 14.Qc4 e5! 15.Be3 Nc2† 16.Ke2 Nd4† 17.Ke1 b5ƒ The analysis could be extended, but it is obvious that Black has good compensation for the donated piece. White cannot castle, she lags in development and she will have difficulties bringing her kingside 413

pieces into battle. A2) 7.Be2!?

With this sly move, White does not hurry to develop the royal knight, but instead gives himself the option of f2-f3 to bolster the e4-pawn. Tigran Petrosian successfully introduced this move against Botvinnik in the 15th game of their 1963 World Championship match. More recently, Carlsen, Topalov, Kramnik and Yakovenko have all played it. 7...b6! The point of White’s move order is well revealed in the event that Black tries to follow the recipe that we will use against the main Russian System: 7...a6 8.Bf4 b5 9.Qxc7 Qxc7 10.Bxc7 Bb7 11.f3! Even though the engine assesses the position as close to equal and some strong players have defended Black’s cause, his compensation for the pawn is not so obvious to me. The text move is rare but quite promising. Black supports a future ...c5 break, while also preparing to deploy the bishop on a6 or b7 – or sometimes both squares in quick succession. White has several possible moves and there are not many games on the database, so you can expect a fresh, dynamic position. 8.Bf4 I will take this as the main line, as it’s a natural move and has been White’s most popular choice so far. 8.e5 This aggressive, space-gaining move is somewhat risky, since the weakened d5-square becomes an important factor. Nevertheless, it was the choice of some strong GMs, which obliges us to study it 414

closely. 8...Be6! We already encountered this theme in the previous chapter. 9.exf6? White sacrifices the queen for three pieces but Black gets two extra pawns as part of the deal, which renders White’s compensation inadequate. 9.Qa4 is more reliable although 9...Nd5 10.Nf3 c5! 11.dxc5 Nd7 still gives Black fine prospects. 12.Qh4 Nxc5 13.0-0 Nxc3 14.bxc3 Qc7 15.Ng5 h6 16.Nxe6 Nxe6 17.Bxh6 Qxe5 18.Bc4 Qf6 19.Qxf6 exf6 20.Bxg7 Kxg7³ left Black with the slightly better endgame in Bluvshtein – Ganguly, Edmonton 2009, due to White’s split queenside pawns. 9...Bxc4 10.fxg7 Kxg7 11.Bxc4 Qxd4 We encountered similar situations in the previous chapter. Once again, Black’s plan will be to exchange as many rooks as possible and gradually advance his pawns, especially on the kingside. 12.Bd5 c6 13.Nge2 Qf6 14.Be4 Nd7 15.Be3 Nc5!? Personally I would prefer 15...Rad8N 16.0-0 c5 17.Rad1 Ne5µ with further exchanges to follow. The text move practically forces White to trade his dark-squared bishop for a knight, but Black’s queenside structure is damaged. On the other hand, he still has two extra pawns and may be able to utilize the open b-file at some point. 16.Bxc5 bxc5 17.Bd3 In Lalith – Gupta, New Delhi 2014, Black advanced his a-pawn, but it would have been better to prioritize the centre by means of:

17...Rad8!N 18.Bc4 e5 19.0-0 Rfe8 20.Rac1 e4 21.Rc2 Qe5µ Intending ...e3 and the inevitable exchange of rooks. Another move which deserves attention is: 8.Be3!? Ng4 9.Bxg4 Bxg4 10.Nge2 415

White has lost the bishop pair but gained some time and strengthened his centre – but we can still attack it.

10...c5! 11.dxc5 Nd7! 12.cxb6 Nxb6 In return for the sacrificed pawn, Black gets active play for all his pieces. 13.Qa6 In the event of 13.Qb3 Rc8 14.Bxb6 Qxb6 15.Qxb6 axb6 16.f3 Be6 17.Nf4 b5© Black has full compensation for the pawn. His pieces are well mobilized and he threatens ...b4 followed by ...Rc2. 13...Bxe2!? If Black is content with a draw, then 13...Bc8 14.Qb5 Bd7 15.Qa6= is an immediate repetition. The text move is an option to play for a win from an equal position. 14.Qxe2 Na4! This move is an important detail, without which Black would stand worse. 15.Nxa4 Or 15.Rc1 Nxc3 16.bxc3 Qa5 17.0-0 Bxc3 18.Rfd1 e6= with an equal position. 15...Qa5† 16.Bd2 Qxa4 17.e5 Qc6 18.0-0 Qc2 19.Rae1 Qxb2 20.Bg5 Qxe2 21.Rxe2 e6= The endgame was approximately equal in Ferre Perez – Wanders, corr. 2017. Another logical move is: 8.Nf3!? Developing naturally. 8...c5! Black should immediately strike at the centre, provoking complications. 9.dxc5 Ba6 10.Qa4 Qc8!? Offering a pawn sacrifice in return for open lines. 11.Be3 416

11.cxb6!?N 11...axb6 leads to a different type of game. A sample line is 12.0-0 Bxe2 13.Qxa8 Bxf1 14.Kxf1 Qc4† 15.Kg1 Nc6 16.Qa4 Qxa4 17.Nxa4 Nxe4 18.Nxb6 Rb8 19.Nc4 Rb4 20.Nfd2 Nxd2 21.Nxd2 Bxb2 22.Bxb2 Rxb2 with an equal endgame. 11...bxc5 So far, this position has only been tested in correspondence games, and Black has done fine.

12.Bxa6 White did not achieve anything after 12.Bc4 Bxc4 13.Qxc4 Ng4 14.Nd5 Nxe3 15.Nxe7† Kh8 16.fxe3 Qb7÷ in Pepermans – Perez, email 2013; Black is no worse in view of the weakness of White’s b-pawn and doubled e-pawns. 12.Rc1 Bxe2 13.Kxe2 Nbd7 14.Rhd1 Nb6 15.Qb5 Rb8 16.b3 Qe6 17.Kf1 Nxe4 18.Nxe4 Qxe4 19.Qxc5 e6 20.Kg1 Rbc8 kept a balanced position in Kokoryukin – Bernal Varela, email 2011. 12...Qxa6 13.Qxa6 Nxa6 14.Ke2 Ng4 15.Bf4 f5!÷ We have been following Molzahn – Rach, email 2014. Despite the splitting of Black’s queenside pawns, his position is in good order, due to his active possibilities on the kingside.

417

8...Ba6 9.Qa4 White had better avoid 9.Qxc7? Qxd4 10.Rd1 Qb4 11.Rd2 as played in Potpara – Firat, Batumi 2010, when Black could have established a serious advantage with:

11...Ne8!N 12.Qd8 Bxc3 13.bxc3 Qxe4 14.Qd5 Qxd5 15.Rxd5 Bb7 16.Bf3 Nf6 17.Rd2 Bxf3 18.Nxf3 Nc6µ With a safe extra pawn, as well as the better structure on the queenside. 9...Bb7! Having driven White’s queen to a slightly worse position, Black targets the e4-pawn. 10.Bf3 418

10.f3 c5 11.d5 occurred in H. Johnson – De Oliveira, email 2014, and here I recommend:

11...Nh5N 12.Be3 e6! 13.dxe6 fxe6 Black’s activity is more important than the slight damage to his pawn structure. For instance, 14.Rd1 Qe7 15.Nh3 Nc6³ and Black is somewhat better due to his control over the d4-outpost. 10...Qd7! Another strong move, posing White some problems. 11.Qxd7 Nfxd7 12.e5 Bxf3 13.Nxf3 So far, Black had done everything right in Babula – Ftacnik, Czech Republic 2010, but on the next move he played too passively with 13...c6. A clear improvement is:

419

13...c5!N The critical continuation is: 14.d5 This must have been the move that Ftacnik was afraid of. Anything else gives Black easy play with ...Nc6, ...Nc5-e6 and so on. 14...Na6 15.0-0-0 c4³ Black will transfer one of his knights to d3, where it will be extremely powerful. We may also note that Black has a pawn majority on the queenside, which can be mobilized in the near future. B) 4.Nf3 Bg7 5.Qb3 The Russian System is most often encountered with precisely this move order. As mentioned earlier, Vyacheslav Ragozin was the first to play this way, against Romanovsky in Leningrad 1932. Subsequently, the young Botvinnik incorporated it into his repertoire. 5...dxc4 6.Qxc4 0-0 We will analyse B1) 7.Bf4 followed by B2) 7.e4. The latter move is the main line by far, and most of its coverage will be found in the next chapter. B1) 7.Bf4 Na6 8.e4 White has nothing better. For instance, the timid 8.e3 is simply met by 8...c5 9.dxc5 Qa5 when Black is 420

already at least equal. 8.Rd1 gives Black a choice: 8...c6 invites 9.e4 with an immediate transposition to variation A1. Alternatively, 8...Be6!? is a logical attempt to exploit White’s move order. 9.Qb5 (9.d5?! Nxd5 10.Nxd5 c6 clearly favours Black) 9...Nd5 10.Be5 c6 11.Qa4 occurred in Hawellek – Kreuzholz, Germany 2016, when 11...Bxe5N (or 11...f6!?N 12.Bg3 Qb6) 12.Nxe5 c5 would have given White some problems. Seizing the centre is White’s most logical plan and we must consider it carefully, as we no longer have the option of ...a6 and ...b5, as recommended in the main lines that follow.

8...c5! Black takes decisive action in the centre and gets ready to move the queen to a5. We will consider two main replies: B11) 9.Rd1!? and B12) 9.dxc5. 9.e5 Be6 takes us back to variation B3 of the previous chapter, with one less move played – see page 216 for the continuation. 9.d5 is an interesting option to gain space. 9...e6 10.Be2 (10.d6 Nb4 seems more dangerous for White than for Black) 10...exd5 11.exd5 occurred in Khotenashvili – Lagno, Huaian 2016, and several other games. I recommend a new and dynamic move, leading to intense complications:

421

11...b5!N 12.Nxb5 (12.Qxb5?! seems worse in view of 12...Nb4 13.0-0 Nfxd5 14.Rfd1 Bd7 15.Qc4 Nxf4 16.Qxf4 Nc2 17.Rac1 Nd4ƒ when Black’s chances are preferable due to the bishop pair) 12...Re8 13.Nd6

13...Nb4! Starting a war on the board. 14.Nxe8 Qxe8 15.0-0 Nfxd5 16.Rae1 Ba6 17.Qxc5 Bxe2 18.Bd6 Qb5 19.Rxe2 Qxe2 20.a3 a5 21.Re1 Qd3 22.axb4 Nxb4= The storm has subsided and a roughly equal position has arisen. B11) 9.Rd1!? This seems a reasonable move although it has only been played once at the elite level. 422

9...Qa5!? I like this rare move more than 9...cxd4, as was played in Vitiugov – Caruana, Skopje 2015. 10.d5!? This is White’s most ambitious choice so I present it as the main line. 10.Rd2!? was played in Pischikov – Priasmoro, Pardubice 2014, and just one other game so far. It should lead to a complicated game with mutual chances. I recommend 10...Rd8N as an improvement, and after 11.d5 Black has a choice: 11...Bd7 This move leads to equality. (For fans of a sharper game, I can recommend 11...Ne8!? 12.h4 Nd6 13.Bxd6 exd6 14.h5 Nc7 15.hxg6 hxg6 16.Qd3 b5÷ with double-edged play.) 12.Be2 b5 13.Nxb5 Nxe4 14.Qxe4 Bxb5 15.0-0 Bxe2 16.Qxe2 Nc7

17.Rfd1 Rd7 18.Ne5 Bxe5 19.Bxe5 Rad8 20.d6 exd6 21.Rxd6 Rxd6 22.Rxd6 Re8 23.h3 Nb5 24.Rd7 Nd4 25.Qe4 Nc6 26.Qxc6 Qe1† 27.Kh2 Qxe5†= The ending is equal. 10.e5!?N has yet to be tested in practice but it also leads to interesting and complex play: 10...Nd7 11.Be2 Nb6 12.Qb5 (or 12.Qb3 Bg4 13.Qb5 Qxb5 14.Nxb5 Rfc8÷ and Black has no problems) 12...Qxb5 13.Nxb5 Be6÷ Black is fine and in a practical game I would favour his position, as he has slightly more active pieces and the weakness of the e5-pawn may make itself felt. For example, after the plausible 14.dxc5?! Nxc5 15.Nc7 Rac8 16.Nxe6 Nxe6 17.Be3 Rc2³ White’s bishop pair is outweighed by the penetration of Black’s rook to the second rank, so White is forced to switch to defence. 10.Qb5 This move has been played a few times but Black has nothing to fear from the queen exchange. 10...Qxb5 11.Bxb5 cxd4 12.Nxd4 Nc5 13.e5 423

This is more logical than 13.f3?! as played in Raveggi – Paganelli, Italy 1987, at which point 13...a6!N 14.Be2 Nfxe4! 15.fxe4 e5 16.Be3 exd4 17.Bxd4 Bxd4 18.Rxd4 Be6³ would have favoured Black due to his better pawn structure. 13...Ng4 14.Bg5 In Lados – Swol, Mikolajki 1991, Black should have continued:

14...Bd7!N 15.Bxd7 Nxd7 16.e6 Nb6 17.h3 Nf6 18.exf7† Kxf7= Black’s isolated e-pawn is easy to defend, and the slight structural weakness is balanced out by Black’s activity, with his rooks heading for the c- and d-files and his knights looking to cause trouble for White on the queenside. 10...Bd7 Black prepares a timely ...b5. This move reminds us of one of the conclusions from the previous chapter: that provoking ...Bd7 before heading for this type of position is not particularly helpful for White. 11.Rd2!? This sensible prophylactic move has only occurred in one game. 11.Nd2 is recommended by my ‘iron friend’ and was played in J.P. Santos – J. Feller, Lucerne (ol) 1982. My suggestion is 11...b5!N 12.Nxb5 Nxe4 13.Qxe4 Bxb5 14.Bxb5 Qxb5 15.Nc4 Bd4 16.b3 Nb4÷ with a complex and unclear game in which Black’s chances are no worse. 11.Bd2!? leads to interesting complications. In Tillmann – Kais, Moehnesee 2010, the strongest move would have been 11...Nb4N, with the following point: 12.Qb3

424

12...Nxe4! 13.Nxe4 Qxa2 14.Qc4 (14.Qxa2?? Nc2† 15.Ke2 Bb5† 16.Qc4 Bxc4#) 14...Qxb2 15.Bc3 Bxc3† 16.Qxc3 Qxc3† 17.Nxc3 Rfd8÷ Black has three pawns for the piece, and his chances are no worse in this complicated queenless position. 11...e6 12.Be2 exd5 13.exd5 In Studzinski – Kuma, Poland 1990, it would have been natural and good to continue:

13...Rfe8!N With a complicated game. Note that White can immediately lose by making a seemingly natural move:

425

14.0-0? b5! 15.Nxb5 Re4–+ Winning a piece. B12) 9.dxc5

GAME 2

Alexander Rustemov – Emil Sutovsky Polanica Zdroj 1999 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.Bf4 Na6 8.e4 c5 9.dxc5

This natural move has been White’s most popular choice, but trying to keep an extra pawn is risky for him, as we will see. 9...Be6 Attacking the queen with tempo; the c5-pawn will be destroyed later. 10.Qb5 Bd7! Black offers the b7-pawn in order to accelerate his development. 11.Qxb7 426

White’s only ambitious approach is to take the pawn, but it will require him to play defensively while the black pieces develop more activity. 11.Qc4 offers to repeat the position. This position has already been encountered in the previous chapter on page 216, where 11...Qc8!?N would be our choice if an early draw is not satisfactory. 11...Nxc5 12.Qb4 Ne6 The knight is perfectly located here, where it eyes the f4- and d4-squares. In addition, it conveniently gains a tempo against the bishop, giving Black more time to develop his initiative.

13.Be5 White is trying to position the bishop actively, but it will soon be thrown back. Some other games have continued: 13.Bg3 a5! A typical move to drive White’s queen away from the defence of the e4-pawn. 14.Qa3 No better is 14.Qb3 a4 15.Qa3 Bc6 16.Nd2 when, in Smithers – De Waard, corr. 2001, Black missed an extremely strong queen manoeuvre: 16...Qd7!N 17.Rd1 Qb7–+ The pressure on the e4-pawn is already decisive. 14...Bc6 15.Rd1 Qb6 White now faces the question of what to do about the vulnerable e4-pawn.

427

16.Bd3 16.e5 Nh5 does not help White, especially after the pseudo-active. 17.Nd5? Bxd5 18.Rxd5 as played in Sitnik – Balenovic, Sisak 2000, at which point 18...Nhf4!N 19.Rd1 Rfd8–+ would have made it practically impossible for White to finish developing and hide his king. 16...Nc5 17.e5 Nh5 18.Be2 Nxg3 19.hxg3 e6 20.0-0 Nd7 21.Qd6 Rad8 22.Qd4 This occurred in Ftacnik – Ilincic, Prague 1989, and now I agree with Avrukh’s proposed improvement: 22...Qb8!Nµ Picking up the e5-pawn, after which Black will have clearly better chances thanks to the two bishops. 13...a5! Just as in the previous note, the white queen is forced to retreat to a less favourable post. 14.Qa3 No better is: 14.Qb3 a4 15.Qb4 Bc6 16.Bc4 (16.Be2 is well met by 16...Qa5! 17.Qxa5 Rxa5 18.0-0 Nxe4 19.Bxg7 Kxg7³ when Black is somewhat better due to his active pieces and the weakness on b2) 16...Nxe4 17.Nxe4 Bxe4 18.Bxg7 Nxg7 19.Bxf7† It may look as though Black has missed a simple trick, but 19...Rxf7 20.Qxe4 Qa5† 21.Kf1 Rb8© left him with more than enough compensation for the small material shortage in Machycek – Van Oosterom, corr. 2007, due to the unpromising location of White’s king.

428

14...Qb6! From here the queen eyes the targets on b2 and f2, as well as the important b4- and d4-squares. Thus, White has to reckon with unpleasant ideas such as ...Ng4 and ...Qb4. Black has more than enough play for a pawn and White must make a difficult choice between several unpleasant options. Emil Sutovsky has done especially well from this position: in addition to our main illustrative game, he subsequently claimed two other victims. 15.Bc4 15.Be2? was a surprising mistake in F. Berkes – A. Zhigalko, Legnica 2013, when both of these extremely strong grandmasters overlooked 15...Ng4!N 16.Bg3 Nf4!µ with a near-decisive advantage for Black. 15.h3 addressed the ...Ng4 threat in Schroer – Mikhalevski, Internet (blitz) 2000, but does not advance White’s development. I recommend 15...Rac8!N, activating the rook and preventing White’s bishop from coming to c4. Play continues 16.Be2 Bc6 17.0-0 Nxe4 18.Nxe4 Bxe4³ when Black’s chances are somewhat more promising due to the weak pawn on b2. 15.Rd1 was seen in Yakovich – Maletin, Loo 2014, when 15...Ng4!N would have been thematic. After 16.Bg3 Qb4 17.Qxb4 axb4 18.Nd5 we have a powerful sacrificial opportunity:

429

18...Bxb2! 19.Nxe7† Kg7 20.Rxd7 Bc3† 21.Kd1 Nc5 22.Rc7 Nxe4µ Black has superb compensation for the piece thanks to his ultra-active pieces and the unfortunate position of the white king, despite the fact that there are no queens on the board. To a certain simplification of the position leads: 15.Nd5 Nxd5 16.exd5 Nd4 17.Bxd4 Bxd4 18.Nxd4 Qxd4 19.Be2 Rab8 Even though White has exchanged a lot of pieces and remains a pawn up for the moment, he is far from being equal due to his retarded development and weak pawns on b2 and d5. 20.0-0 Rxb2 21.Rad1 21.Qxe7 Ba4 22.Ba6 (22.Bf3? Re8 23.Qc7 Bb5 24.Rfd1 Qxf2† 25.Kh1 Rxa2 would be even worse for White) 22...Qxd5µ is mentioned by Avrukh; the a2-pawn is weak and White’s defence is not easy. 21...Qb4 22.Qxb4 Rxb4 23.Rb1 23.Bf3 fails to solve White’s problems after 23...Rfb8 24.Rfe1 Kf8 25.d6 exd6 26.Rxd6 Be6 27.a3 Rb3µ when a long and difficult defence awaits him. 23...Rfb8 24.Bd3 This was Mir Mahmoud – T. Gara, Istanbul (ol) 2012, and here I suggest a simple improvement of Black’s king:

430

24...Kf8Nµ Black enjoys the better structure and the d5-pawn is quite likely to be destroyed in the future. Sutovsky’s second ‘customer’ in this line opted for: 15.Bxf6 Eliminating the threat of ...Ng4, but trading the bishop for the knight is an obvious concession. 15...exf6 Black has excellent compensation; White has tried various continuations but none of them solves his problems. 16.Bc4 16.Nd5 Qc5 17.Qxc5 Nxc5 18.0-0-0 Be6 19.Nd2 f5µ occurred in Novoa – Jelen, email 2002. Black’s bishops exert unpleasant pressure on the queenside and White’s extra pawn does nothing to alleviate his suffering. 16.Rd1 Bc6 17.Bc4 f5! activates both bishops, and after the further 18.Nd5 Bxd5 19.exd5 Bxb2 20.Qb3 Qxb3 21.axb3 Bc3† 22.Kf1 Nc5µ White had a depressing position in Mittelman – Sutovsky, Ramat Aviv 2000. Despite the opposite-coloured bishops and White’s theoretically better pawn structure, Black has several advantages: the b3-pawn is weak, White’s bishop is restricted by its own pawns and it will take White a few more moves to get the h1-rook into the game. 16...f5 17.exf5 Nd4 18.0-0 This position was reached in Elizarov – Kovalchuk, Rostov-on-Don 2019, and here I should have played:

431

18...Nxf3†!N 19.gxf3 Bxf5 There is no need to fear: 20.Nd5 Qd4 21.Ne7† Kh8 22.Nxf5 gxf5–+ With a decisive attack on the white king. Returning to the game, White tries to catch up on development but for complete happiness he will have to castle. Before that can happen, it is Black’s move and he begins to create unpleasant threats.

15...Ng4! Threatening the bishop and the f2-pawn, so White’s answer is forced.

432

16.Bg3 Qb4! Exploiting the unsuccessful position of the white queen, Black continues developing his initiative in the endgame. 17.Qxb4 axb4 18.Nd5 Bxb2 19.Rb1? What looks like a natural and logical move actually turns out to be a serious inaccuracy. Another game continued: 19.Rd1 Ba4 20.Rb1?! This move saves material but does not solve White’s problems. (He should have tried 20.Rd2!N 20...Bc3 21.Nxe7† Kg7 22.0-0 Bxd2 23.Nxd2÷ as given by Avrukh. White has reasonable compensation for the exchange, but no more than equality.) 20...Bc3† 21.Ke2 Nc5 22.e5 We have been following Johansson – Pasko, corr. 2001. In this difficult, tangled position, Black could have pressed with:

22...e6!N 23.Nxb4 Ne4 24.Rhc1 Rfc8 25.Bd3 Nxg3† 26.hxg3 h5³ Black will pick up the e5-pawn to equalize the material count, and the a2-pawn will remain weak. The limited material on the board will make it difficult to get more than a draw, but Black is the one who will be trying. 19.0-0! is perhaps the best way to sacrifice the exchange: 19...Bxa1 20.Nxe7† Kg7 21.Rxa1 Nc5 22.h3 Nf6 23.Be5 Be6

433

24.Nd5 Bxd5 25.exd5÷ In Meissen – Morozov, corr. 2017, White’s extra pawn and bishop pair yielded compensation for the exchange, but only enough for equality. 19...Bc3†! Forcing White to make a difficult choice: he must either lose the right to castle or give Black a dangerous passed pawn. 20.Ke2 20.Nxc3?! bxc3 21.0-0 was played in Bacrot – Sutovsky, Albert (6) 2001. Quite why such a strong player as Bacrot not only followed a line in which Sutovsky enjoyed previous success but also deviated with an inferior continuation for White remains a mystery. In any case, Sutovsky moved his queen’s rook and won in only eight more moves, but it would have been slightly more accurate to play 21...Rfc8!N 22.Bd5 c2 23.Rbc1 Ra3µ as given by Avrukh, when the c2-pawn is a big pain for White. 20...Nc5 The knight takes aim at the e4-point. 21.Nb6 White tries to reduce Black’s attacking firepower by exchanging his knight for a bishop. The greedy 21.Nxe7†? Kg7 22.e5 Ne4 23.h3 allows a beautiful combination:

434

23...Rxa2†! 24.Bxa2 Bb5† 25.Kd1 Ngxf2† 26.Bxf2 Nxf2† 27.Kc1 Nxh1–+ Black emerges with an extra pawn and his attack continues. I also checked 21.h3 Nf6 22.Nxe7† Kg7 23.e5 Nfe4 24.Nd5 Be6 25.Bh4 Rae8µ when White’s position is not to be envied: the e5-pawn will soon be destroyed and his exposed king remains a problem.

21...Be6! Black sacrifices the exchange in order to launch an attack on the white king. 22.Bxe6 fxe6!? 435

Opening the f-file and taking control of the d5-square. 22...Nxe6N was also good: 23.Nxa8 Rxa8 24.Rhd1 Rxa2† 25.Kf1 Nf6 26.e5 Ne4µ Black has fantastic compensation for the exchange due to his active pieces and the strong passed pawn. 23.Nxa8 Rxa8µ Such positions are always unpleasant to defend, and it’s not surprising that White commits a further error. 24.h3? Driving away the active knight may seem logical but it leads to even greater problems. 24.Nd2 was the lesser evil, although 24...Bxd2 25.h3 Nxe4 26.hxg4 Nxg3† 27.fxg3 Bc3µ still gives Black excellent chances to realize his advantage. 24...Rxa2† 25.Kf1 Nf6 26.e5 Nd5 The knight from g4 has moved to an even better outpost in the centre, and his position is winning.

27.Kg1 b3 28.Kh2 b2 29.Rbd1 Ne4 30.h4 h6 31.Rd3 Kf7 32.Kh3? A final blunder brings the game to a close. 32...Nxf2† Faced with the loss of a rook, White resigned. 0–1 B2) 7.e4 436

Finally we arrive at the main tabiya of the Russian System. 7...a6! With this move, Black prepares to expand on the queenside with ...b5, capturing space and knocking the queen away from its active post, before hitting the centre with ...c5. Alekhine was the first to play this move, against Euwe in Game 12 of their World Championship match in 1935. Even though Black lost the game, it is quite remarkable that the move was abandoned for nearly thirty years thereafter. It remained obscure until the Hungarian players Barcza, Ribli and Barczay all played it in their national championship of 1969, after which it became known as the Hungarian Variation. It went on to become the main line and it remains Black’s most popular choice to this day, with Kasparov, Grischuk, Anand, Svidler, Giri and Nepomniachtchi among its adherents. In the remainder of this chapter, we will focus on two significant sidelines: B21) 8.Qa4!? and B22) 8.Bf4. We will deal with White’s two most popular and challenging options, namely 8.e5 and 8.Be2!, in the next chapter. 8.Qb3 has played in hundreds of games but after 8...b5 (8...Nc6!? is an extra possibility against White’s chosen move order) we are almost certain to transpose. For instance, 9.a4?! is the 8.a4?! line on page 235, while 9.e5 and 9.Be2 lead respectively to variations A and B2 of the next chapter. 8.Bd3?! is a clumsy choice, as the bishop only interferes with White’s play. 8...b5 White already has to resort to a strange-looking move:

437

9.Qb4N (9.Qb3?! c5µ exposed the shortcoming of White’s 8th move in Patterson – Clancey, corr. 2014; due to the location of the bishop on d3, White must lose a pawn without any compensation) 9...Bb7 10.Be3 Ng4 11.Rd1 Nxe3 12.fxe3 e5µ Attacking the centre and expanding the influence of the bishop on g7 clearly favours Black. 8.Be3?! is premature: 8...b5 9.Qb3 (9.Qd3 Bb7 10.a3 Ng4 also targets the dark-squared bishop; 11.Bd2 was played in Bultman – Dedina, corr. 2012, when 11...e5!N 12.Nxe5 Nxe5 13.dxe5 Nd7 14.Rd1 Nxe5³ would have given Black the better game thanks to his faster development and more active pieces) 9...Bb7 White lacks a convenient way to defend the e4-pawn. For instance:

10.d5 (10.e5 Ng4 11.Be2 c5 12.Rd1 Nxe3 13.fxe3 occurred in Savola – Michielsen, Jyvaskyla 2009, when 13...c4!N 14.Qc2 Bh6µ would have given Black the better pawn structure and two powerful 438

bishops) 10...c6 11.Rd1 cxd5 12.Nxd5 Nxd5 13.exd5 Nd7 White’s problems are far from over, as the d5-pawn is weak: 14.Be2 Nf6 15.d6 e6 16.0-0 In T. Ward – Steele, East Lansing 1995, Black should have continued 16...Nd5!Nµ when the d6-pawn will be destroyed sooner or later. 8.h3?! looks like a waste of time. 8...b5 9.Qd3 (9.Qb3 Bb7 10.e5 Nfd7 gives Black an improved version of variation A of the next chapter, as h2-h3 is unnecessary) 9...Bb7 10.a3 c5 11.dxc5 In Pena – Herrera, Medellin 2016, it would have been natural to play:

11...Nbd7!N 12.Be3 Qc7 13.Rc1 Nxc5 14.Qc2 Ne6µ White is lagging behind in development and he has problems with the weak e4-pawn. 8.a3 is passive and 8...b5 9.Qd3 Bb7 10.Be2 c5 gives Black an easy game, for instance:

439

11.dxc5 (no better is 11.d5 c4 12.Qc2 as played in Proba – Leichter, Germany 2012, when 12...Qd7!?N 13.0-0 e6 14.dxe6 Qxe6µ would have left the e4-pawn weak) 11...Qxd3 12.Bxd3 Nbd7 13.Be3 Now in Tuovinen – Holmsten, Helsinki 1998, the natural 13...Rac8!µ would have enabled Black to win back the c5-pawn sooner or later, and the e4-pawn will remain weak. 8.Bg5 is another strange choice: 8...b5 9.Qb3 Bb7

10.d5?! This move only worsens White’s position. (The untested 10.e5N is a better try although Black can counter it with a powerful piece sacrifice: 10...c5! 11.exf6 exf6 12.Bf4 cxd4 13.0-0-0 Nc6 14.Kb1 Rc8© The position can be analysed further, but it is obvious that Black has excellent compensation, with two extra pawns and the opportunity to develop the initiative on the queenside by means of ...Qb6, ...f5 and ...Na5-c4.) 10...c6 11.Rd1 cxd5 12.exd5 In Auer – Pielmeier, Germany 1997, it would have been logical to continue:

440

12...Nbd7N 13.Be2 Qa5µ The d5-pawn is weak and White is behind in development. Finally, another clear mistake would be: 8.a4?! b5! Utilizing the nuance that White cannot take on b5 on pain of suffering material losses. 9.Qb3 9.Qc5 has seldom been played. 9...Bb7 10.axb5 Nbd7 11.Qb4 axb5 12.Rxa8 Qxa8 13.Bxb5 Bxe4 14.Nxe4 Qxe4† 15.Be2 c5 16.Qd2 cxd4 17.Qxd4 Rc8 18.Be3 was the logical continuation of Doren – Mezera, corr. 2016, and here I can recommend 18...Qc2!N 19.Qd2 Qxd2† 20.Bxd2 Rc2 when White should abandon his queenside pawn: 21.0-0 Rxb2³ White has some drawing chances due to his bishop pair and the limited amount of material, but Black can obviously be happy with the outcome of the opening. 9...c5! 10.dxc5 Another example continued 10.e5 Ng4 11.h3 cxd4 12.hxg4 Be6 13.Qc2 dxc3 14.axb5 Qd5! 15.bxc3 axb5 16.Rxa8 Qxa8 17.Bxb5 Bxg4µ when Black was clearly better in Muzyka – Tazelaar, corr. 2013. The weak e5-pawn is likely to be destroyed and the c3-pawn may also become vulnerable, whereas Black’s structure is much more solid. 10...Be6 11.Qa3 White fell into even worse trouble with 11.Qc2?! b4 12.Nd1 b3 13.Qb1 Qa5† 14.Bd2 Qxc5 15.Bc3 Nc6 16.Ne3 Ng4 17.Bxg7 Nxe3 18.fxe3 Qxe3†–+ when his position was already hopeless in Booth – Chow, Parramatta 2008.

11...b4! 12.Qxb4 Nc6 13.Qb6 13.Qa3 was played against me by a talented junior, and after 13...Rb8 14.Ng5 Nb4–+ White’s position was hopeless due to the unstoppable ...Nc2† in Simonyan – Kovalchuk, Rostov 2019. 13...Qc8 14.Bf4 Nd7 15.Qc7 Nb4 16.c6 16.Rc1?! Qxc7 17.Bxc7 Nxc5 18.Rd1 Bb3–+ occurred in Mesko – Mrdja, Sibenik 2010. White was 441

too far behind in development, with ...Nc2† on the agenda again. The text move is more stubborn but Black remains on top after: 16...Bxc3†! 17.bxc3 Nc2† 18.Kd2 Nxa1 19.cxd7 Bxd7 20.Bc4 Qxc7 21.Bxc7 Rac8 22.Rxa1 Rxc7µ Veiga Rodriguez – Savio, corr. 2012. White only has one pawn for the exchange and no other compensation, so Black should win with good technique. B21) 8.Qa4!? This is the only credible way of preventing ...b5. However, moving the queen again has its own drawbacks. 8...c5! Obviously 8...b5? 9.Bxb5± is the point of White’s last move. 9.dxc5 Bd7 Even though ...b5 was prevented, Black still manages to develop while gaining time against the white queen. 10.Qa3 This seems like the best try, as the queen defends the c5-pawn and avoids further harassment by the enemy pieces. 10.Qc2 poses no problems after 10...Qa5 11.Bd2 Qxc5 12.Be2 and now in Farago – Bezemer, Triesen 2013, it would have been good to activate the light-squared bishop by means of:

12...Bc6N 13.Bd3 Nbd7 14.0-0 Ne5 15.Nxe5 Qxe5³ Black’s position is somewhat freer and easier to 442

play. 10.Qc4 Be6 11.Qb4 Nc6! 12.Qa4 (12.Qxb7? Nxe4! 13.Nxe4 Bd5 gives White major problems) was a peculiar sequence in Glukhovtsev – Karakas, corr. 2015. White loses additional time with his queen, hoping to prove that Black’s pieces have been lured to unfavourable squares. I think Black should bring yet another piece into play, while targeting the c5-pawn:

12...Rc8!N 13.Be2 Nd7 14.Be3 Bxc3†! It is worth spoiling the enemy pawn structure, as White is unable to exploit the weakening of the dark squares. 15.bxc3 Qa5 16.Qxa5 Nxa5 17.c6 Nxc6 18.Ng5 Nd8³ White’s bishop pair is outweighed by his weak queenside structure and the numerous outposts available to Black’s minor pieces, including e5, c5 and c4. A contest between strong GMs continued 10.Qb3 Bc6 11.e5 when, in Sethuraman – Kovalenko, Voronezh 2013, it would have been most logical to play:

443

11...Nfd7!N Attacking both of the vulnerable pawns. 12.Be3 Nxe5 13.Nxe5 Bxe5³ The vulnerability of the g2-pawn makes it difficult for White to develop his light-squared bishop. 10.Qb4 a5 11.Qc4 (11.Qxb7?! Bc6 only helps Black) 11...a4 12.Be2 occurred in Rook – Yloenen, corr. 2012, and now Black’s play can be improved with:

12...Nc6!N 13.0-0 Be6 14.Qb5 Qc8 15.Nxa4 (15.Rd1 a3³) 15...Ra5 16.Nb6 Rxb5 17.Nxc8 Rxc5 18.Nb6 Nxe4³ This position can be assessed as somewhat better for Black due to the active rook and strong bishop on g7, both of which exert pressure against White’s queenside. 10...Bc6 11.Nd2 Defending the e4-pawn. 444

Returning the extra pawn with 11.Be2 Nxe4 gets White nowhere. 12.Nxe4 Bxe4 13.0-0 Nc6 14.Be3 occurred in Gyimesi – Sedlak, Hungary 2010, and here I recommend: 14...Qc7N 15.Rad1 Rad8³ It’s not clear what White should be doing. Even if he exchanges all the rooks, Black will keep some initiative due to his more active minor pieces. After the text move I found a useful improvement:

11...e6!N The purpose of this move will soon be clarified. 11...a5?! allows White to consolidate with: 12.f3! Nfd7 (12...Na6?! is the move Black would like to play, but after 13.Nc4 Nb4 14.Qb3± his manoeuvre has not achieved anything) 13.Nc4 Bd4 14.Be3 Bxe3 15.Nxe3 e6 Black was somewhat short of equality in Wang Hao – Grischuk, London 2012. 12.Be2 12.f3?! is much less attractive here, since 12...Nh5! immediately threatens a nasty check on h4. After the further 13.g3 f5 14.Bc4 (or 14.Be2 f4ƒ with unpleasant problems for White) 14...Nd7 15.Bxe6† Kh8 16.0-0 Ne5 Black has a serious initiative for a pawn, for instance: 17.Kg2 f4 18.g4 Qg5‚ White is in trouble on the kingside. 12...a5 13.0-0 13.f3 Nh5 14.g3 f5 is not quite as scary as the note above, as Black has spent a tempo on ...a5, but he has a pleasant initiative all the same. 13...Na6 445

Black develops his knight in the ideal way, without being thwarted by the Nc4/Qb3 set-up as shown in the note to Black’s 11th move above.

14.Rd1 Nb4 15.Qb3 Qe7 16.Qc4 Nd7³ Black regains the c5-pawn and his active pieces will give him the upper hand. B22) 8.Bf4

First played by Euwe against Alekhine in their 1935 World Championship match, this active move is a lot less popular than the two main lines covered in the next chapter. Nevertheless, it is a serious option which regularly appears in modern grandmaster praxis. White develops his bishop and attacks the c7-pawn. 446

8...b5! Black sacrifices the c7-pawn to speed up his development. Despite the impending queen exchange, his activity will offer at least enough compensation for the minimal deficit. 9.Qxc7 There is no turning back. The whole point of White’s last move was to destroy the c7-pawn, so it would seem strange to pass up the opportunity. Consider the alternative queen moves: 9.Qb3?! Bb7 10.e5 Nd5 11.Bd2 is a clear loss of time for White. 11...c5 12.dxc5 Nxc3 13.Qxc3 Nd7 14.Rc1 Bxf3 15.Qxf3 Nxe5 16.Qe4 f5 17.Qc2 was seen in Quinteros – Moreno Carnero, Buenos Aires 1997, when Black’s strongest continuation would have been:

17...Nc6!N 18.Bc3 Qd5 19.Qb3 Bxc3† 20.Rxc3 Qxb3 21.Rxb3 e5µ Black is far ahead in development and his knight has an excellent outpost on d4. In the time it takes White to safeguard his king and bring the h1-rook into play, Black will create unpleasant threats. 9.Qd3?! Bb7 10.a3 c5! also causes problems for White:

447

11.dxc5 (no better is 11.d5 as played in Yu Kaifeng – Xu Junwei, Qinhuangdao 2017, when 11...c4!Nµ would have been consistent) 11...Nbd7 12.c6 Bxc6 13.Ne5 Nxe5 14.Bxe5 Now in Delmont – Dragojlovic, La Fere 2007, 14...Qa5!Nµ would have underlined Black’s advantage; he is well ahead in development and is ready to centralize the rooks. A final inferior, time-wasting option is: 9.Qc5?! Bb7 10.Bd3 Other moves are even worse, for instance: 10.d5? Nbd7 11.Qa3 c5 12.dxc6 Bxc6 13.Nd2 Qb6 14.Be3 Qb7 15.Qxe7 b4 16.Nd1 Rfe8 17.Qd6 Nxe4–+ and White’s position was collapsing in Szalanczy – Szeberenyi, Balatonbereny 1997. 10.Bxc7? Qc8 leaves White unable to justify his greed, for instance: 11.Be5 Qxc5 12.dxc5 Nbd7 13.Bd4 b4 14.e5 bxc3 15.exf6 cxb2 16.Bxb2 Bxf6 17.Bxf6 exf6 18.Rc1 Rfe8† 19.Be2 Rac8 20.c6 Bxc6 21.Kd2 Rcd8–+ In Ungureanu – Ghinda, Bucharest 1974, Black’s extra pawn was incidental, the unfortunate position of White’s king in the centre being the deciding factor. We have been following Castro Perez – Stankovic, La Roda 2018. Black has several good continuations but the strongest and most spectacular is:

448

10...e5!!N The game continuation of 10...Nbd7 was also no fun for White. 11.Nxe5 11.Bxe5? is even worse in view of 11...Nfd7! 12.Qxc7 Nxe5–+ and Black wins a piece. 11.dxe5!? is well met by: 11...Nh5 12.Be3 Nd7! (12...Qxd3 picks up the loose bishop but allows 13.Rd1 Qc4 [13...Qc2?? 14.0-0 leaves Black’s queen desperately short of squares] 13.Qxc4 bxc4 14.g4 when White regains the piece with unclear consequences) 13.Qa3 Nxe5 14.Nxe5 Bxe5 15.Rd1 Qf6 16.0-0 Nf4µ Black has extremely active pieces, the knight on f4 being especially dangerous. If White exchanges it, Black will have the bishop pair and White’s dark squares will be weak. 11...Re8 12.Nf3 Nc6 13.0-0 Nd7 14.Qd5 Rb8

15.Qg5 Qxg5 16.Nxg5 Nxd4 17.Bxc7 Rbc8 18.Bd6 Nc5 19.Bxc5 Rxc5 20.Nh3 Ne6µ 449

White remains a pawn up for now, but Black’s superior activity and coordination make him the clear favourite. 9...Qxc7 10.Bxc7 Bb7 The e4-pawn is under heavy fire; not only from the pieces already attacking it, but also from the possibility of ...b4, driving away the defensive knight.

White has two serious options: B221) 11.e5 and B222) 11.Bd3. Other moves are worse, for instance: White can hardly hope for success with 11.Bxb8?! Raxb8 12.e5 as played in Reyes Cortes – Garcia Gonzalvez, corr. 2013. The most precise continuation is:

12...Nh5!N 13.Rd1 Rfd8µ The knight can go to f4 at any moment, and Black also has ideas of ...b4 and 450

...f6, so his compensation for the pawn is more than sufficient. 11.d5?! b4 12.Na4 Nxe4 13.Rd1 occurred in Porper – Vadasz, Borsodtavho 1991, when Black could have simply played:

13...Nd7!N Presumably Black was afraid of 14.Ba5, but after 14...Rfc8 15.Bxb4 Nef6! he targets the d5-pawn and White’s king will remain stuck in the centre. Play may continue 16.Nc3 Nb6 17.d6 exd6 18.Bxd6 Nfd5µ when White’s position is extremely unpleasant; he will have to suffer far too much for the sake of a mere pawn. Finally, simply abandoning the e4-pawn makes Black’s task too easy. 11.Be2?! Nxe4 12.0-0 Rc8 13.Ba5 Nc6 14.Bb6 occurred in Henkel – Rossberg, Verden 2016, and here I recommend:

451

14...Nf6!Nµ White is clearly worse because of the isolated d4-pawn; moreover, Black threatens the unpleasant ...Nd7, underlining the unfortunate placement of the bishop on b6. B221) 11.e5 The drawback of this move is that the d5-square is weakened while the b7-bishop becomes extremely strong. 11...Nd5 12.Nxd5 Some other games have continued: 12.Ba5 Nf4! Preventing White from developing the kingside. 13.0-0-0 13.a4 Nc6 14.Bb6 occurred in Raijmaekers – Petkov, email 2003, when 14...Nb4!Nµ would have made White’s position difficult to defend. 13.Rd1 Bxf3! 14.gxf3 Nc6 15.Bb6 f6 16.e6 f5 17.a4 b4 18.Ne2 Nd5 19.Bc5 Rf6 20.h4 Rxe6ƒ gave Black a nasty initiative in Hilgert – Pollina, email 2000. White’s king is stuck in the centre and he has a bad pawn structure. White’s current illusion of wellbeing is quickly shattered by:

13...Nh3! 14.Re1 14.Rd2? runs into 14...Bh6 of course. 14...Nxf2 15.Rg1 Nc6 16.Bb6 Nb4 17.Kb1 In Lautenschlager – Reussner, corr. 1990, Black could have safely played:

452

17...Nbd3!N 18.Re3 18.Bxd3 Nxd3 19.Rd1 Nf4µ gives Black an easy advantage, with two bishops and more than one weakness to attack. 18...Nf4 19.Re1 Rfc8 20.a3 Nd5 21.Nxd5 Bxd5 22.Bc5 Rc7 White’s position remains shaky and his coordination leaves a lot to be desired. 12...Bxd5

13.Be2 13.Bd3 Nc6 14.Bb6 Rab8 15.Bc5 does not enable White to keep the extra pawn because of 15...Bxf3 16.gxf3 Nxd4! 17.Bxd4 Rfd8 18.Bxg6 and now in Darazs – Cseke, Hungary 2009, the easiest 453

continuation would have been:

18...Rxd4N (18...hxg6?! only led to equality in the game, but 18...fxg6!?N is another possible way to play for a win) 19.Be4 Bxe5 20.Ke2 Rbd8³ The opposite-coloured bishops should enable White to draw with precise defence, but Black will be able to press without taking the slightest risk. 13...Rc8 14.Ba5 The bishop aims for c3, where it will defend the d4-pawn. 14.Bb6 Rc2 15.b3 Nc6 16.Bd1 Rc3 17.0-0 Rb8 18.Bc5 was seen in Dreev – Neverov, Uzhgorod 1987. Black’s play can be improved by means of:

18...e6!N 19.Be2 Bxf3 20.Bxf3 Nxe5 21.Be4 Nd3 22.Ba7 Rbc8 23.g3 Nb4³ Black is better because of the weak d4-pawn along with his control over the open file. 454

14...Nc6 15.Bc3 Rab8 16.a3 a5 17.Rd1 Even worse is: 17.Rc1?! Bxf3 18.Bxf3 Nxd4 19.Bg4 Rc4 20.0-0 Nb3 21.Be2 Rc5 22.Rcd1 b4 23.axb4 axb4 24.Bd4 Rc2 25.Rfe1 Nxd4 26.Rxd4 Rxb2–+ Black went on to convert his extra pawn in Gohil – Luecke, Muenster 1990. 17...b4 18.axb4 axb4 19.Bd2

19...Bxf3 20.Bxf3 Nxd4 21.Bg4 Rc4µ All this happened in Antoshin – Kalinitschew, Moscow 1982. Black is clearly better due to the weak e5-pawn and active pieces, with the knight and the c4-rook combining to create unpleasant threats. B222) 11.Bd3

455

This has been played in fewer games than the previous variation but it has scored more positively for White. He defends the pawn while developing a piece and preparing to castle. 11...b4 Naturally Black takes the opportunity to regain the e4-pawn and leave the d4-pawn isolated. 12.Na4 White plays contrary to Tarrasch’s precepts, according to which the knight on the edge of the board was a disgrace. For a player in the 21st century, such dogmas have long since ceased to be relevant; indeed, the text move has been by far the most popular choice, with a view to moving the knight to b6 or c5. The passive 12.Ne2 gives Black more than one good option: 12...Nxe4 (there is also 12...Bxe4!? 13.Bxe4 Nxe4 14.Rc1 as played in Guillem Sanchez – Espell, Almoradi 1998, when 14...a5N 15.0-0 Nf6³ gives Black the more pleasant position) 13.0-0 Nf6 14.Ne5 This position was reached in Bourgoin – Szczepankiewicz, email 2012, and once again I like the preventative 14...a5N³, keeping the bishop out of a5 and preparing ...Nd5. Black’s position is somewhat better thanks to the isolated d4pawn. 12...Bxe4 This move leads to the exchange of bishops, which, in conjunction with the rule that “when playing against an isolated pawn, it is beneficial to exchange as many minor pieces as possible,” makes perfect sense. At the same time, avoiding simplifications with 12...Nxe4!? also looks interesting, especially as it was the choice of Grünfeld expert Peter Svidler, among other strong players. 13.0-0 Nf6 14.Rac1 (14.Ne5 456

Nc6 15.Nc5 Rfc8 16.Nxc6 Bxc6 17.Be5 Bb5 18.Rfd1 Bxd3 19.Nxd3 Nd5 20.Bxg7 Kxg7 21.Rac1 a5³ left Black with a nagging edge because of the isolated pawn in Dyachkov – Svidler, Maikop 1998) 14...Nbd7 15.Ne5 Rfc8 16.Bc4 e6 17.Ba5 Rab8 18.b3 Bh6÷ A complex position had arisen in Almiron – Fleetwood, corr. 2015. Black’s chances are not worse, yet it is too early to talk of an advantage playing against the isolated pawn, since White’s pieces are active enough. 13.Bxe4 The untimely centralization of the king with 13.Ke2?! leads only to the loss of the d4-pawn. In Shestakov – Kanter, Gornozavodsk 2009, Black missed the opportunity to destroy the key defender with: 13...Bxf3†!N 14.Kxf3 Nc6 15.Bb6 Nd5 16.Be4 Nxb6 17.Bxc6 Nxa4 18.Bxa4 Bxd4µ With an extra pawn and a superbly centralized bishop. 13...Nxe4 14.Rc1 14.0-0 Nf6 15.Rac1 Nbd7 transposes to the next note. 14...Nd7 15.Ke2! The king may play an active role in the endgame and it is not too exposed in the centre. 15.0-0 This slightly inaccurate move was played in Donner – Ree, Amsterdam 1971. Although it would be an exaggeration to suggest that White is already worse, his position becomes slightly trickier than in the main line. 15...Nef6N This novelty immediately transposes to a couple of other games. 16.Rc4! White’s best idea is to double along the c-file while gaining a tempo against the b4-pawn. Although Black has a fine position, it is not so easy for him to exploit the weakness of the d4-pawn. White’s other options are inferior: 16.Ne5 Nxe5 17.dxe5 Nd5 18.f4 was seen in Dyachkov – Petukhov, Alushta 2003, when 18...f6!N 19.exf6 Bxf6 20.Rfd1 Nxc7 21.Rxc7 Rad8 22.Rcd7 Rxd7 23.Rxd7 g5ƒ would have offered Black some initiative in the endgame. His bishop works on both flanks and must be stronger than the knight, especially when it joins forces with the black rook. 16.Nc5 Rfc8 17.Nxd7 Nxd7 18.Rc6 a5³ also gives Black the upper hand due to the long-term weakness of the isolated pawn.

457

16...Nd5 17.Rfc1 a5 It is difficult for Black to exploit his superior pawn structure, and he should avoid trading off the bishop on c7 as this would activate White’s rooks. Instead, he should look to strengthen his position in small ways. 18.g3 Rfc8 19.b3 Ra7 20.Be5 Rxc4 This move gives White the opportunity to improve his pawn structure at the expense of his control over the c-file. The alternative is 20...Rca8!?N 21.Rc8† Bf8 22.R1c6 Rxc8 23.Rxc8 f6÷ when White’s activity is sufficient for equality. 21.bxc4 White could also have considered 21.Rxc4!?N 21...f6 22.Bf4 Kf7 23.Be3 with an approximately equal position. 21...Nxe5 Here White should have preferred:

458

22.Nxe5!N 22.dxe5? Bh6 23.Rc2 Rc7 24.Kf1 Nc3 25.Nxc3 Rxc4 26.Ke2 bxc3 27.Kd3 Ra4 left White a pawn down with no immediate way of regaining it. Even though his position was still tenable with precise defence, the practical problems proved too great in Kozul – Volokitin, Austria 2014. 22...Bxe5 23.dxe5 Nc7 24.f4 Ne6 25.Kf2 Kf8= The two kings come to the centre and the endgame is equal.

15...Nef6 16.Ne5 16.Rc4N is playable, when the position resembles the previous note. Having the king in the centre may seem to benefit White, although the downside is that he will have to watch out for the possibility of a knight landing on f4 with check. Black is certainly not worse in the ensuing position, so White

459

sensibly decided to simplify in the main game which we are following. 16...Nxe5 17.dxe5 Nd5 18.Rhd1 Nxc7 19.Rxc7 Bxe5 20.Rxe7 Rfe8 21.Rdd7 Bf6 22.Rxe8† Rxe8† 23.Kf3 All this happened in Vidit – Nepomniachtchi, Berlin (rapid) 2015, and here I would suggest: 23...Rc8N= A draw looks highly likely. Conclusion The Russian System is an important option which Black should be well prepared for, otherwise he risks falling under a positional grip. White has an advantage in space and relies on his powerful centre, but Black has a significant lead in development and will look to target White’s queen and his central pawns to create counterplay. We started with the version where White refrains from developing the royal knight: 4.Qb3!? dxc4 5.Qxc4 Bg7 6.e4 0-0 when there are two main branches. The first is 7.Bf4 attacking the c7-pawn, when 7...Na6 is a convenient reply to prepare ...c5. It is useful to know that 8.Rd1 can be met by a small modification of Black’s plan with 8...c6! followed by ...Qa5 and ...c5 with thematic play. The second option is the tricky 7.Be2!? when I recommend the rare but logical 7...b6!, with ideas to bring the bishop to a6 and/or b7, while also supporting the thematic ...c5. We then moved on to the more mainstream Russian System where White first develops the royal knight: 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 with another fork in the road. 7.Bf4 Na6 8.e4 c5! gives Black good chances, with Sutovsky’s play in Game 2 an excellent demonstration of his resources. 7.e4 introduces the main line of the Russian System, when 7...a6!, the Hungarian Variation, is our choice. Then 8.Qa4!? is well met by 8...c5! 9.dxc5 Bd7, while 8.Bf4 b5! 9.Qxc7 Qxc7 10.Bxc7 Bb7 also gives Black good prospects. In all these lines, Black gives up a pawn and sometimes additional material in the short term, but his dynamic initiative always provides at least enough counterplay. Having dealt with those lines, we are now ready to move on to White’s two most popular options of 8.e5 and 8.Be2!.

460

A) 8.e5 b5 9.Qb3 Nfd7 247 A1) 10.Be2!? 249 A2) 10.Ng5!? 251 A3) 10.h4!? Game 3 253 A4) 10.Be3 c5 11.e6! cxd4! 258 A41) 12.exf7† 261 A42) 12.Nxd4 261 A5) 10.e6 fxe6 264 A51) 11.Qxe6† 266 A52) 11.Be3 Nf6 269 A521) 12.h4 270 A522) 12.a4! 271 B) 8.Be2! b5 276 B1) 9.Qd3 276 B2) 9.Qb3 c5! 10.dxc5 Be6! 11.Qc2 Nbd7 277 461

B21) 12.c6!? 279 B22) 12.Be3 Game 4 281

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.e4 a6! Having got acquainted with the Russian System in the previous chapter, we are ready to tackle the two most challenging continuations. A) 8.e5 seizes space in the centre and has been White’s most popular choice overall, while the currently trendy B) 8.Be2! is perhaps the most critical move of all. A) 8.e5 White captures space and drives the knight away from f6. In the event that it retreats to d7, White will have the option of e5-e6, destroying the integrity of our pawn structure. The play may become sharp and somewhat forcing, so both sides should be well prepared. The text move was first played by Portisch against Adorjan in Budapest in 1970. It was subsequently taken up by Smyslov, Kozul, Atalik and Gulko, as well as at World Championship level by Karpov, who used it twice against Kamsky in the FIDE World Championship final at Elista 1996. Nowadays it is not so popular at the elite level, as the current verdict is that Black obtains plenty of counterplay. 8...b5 Unlike some of the earlier variations involving e4-e5, 8...Be6?! is not a serious option here in view of 9.Qb4 when White wins material for inadequate compensation. 9.Qb3 This is the only credible continuation, as White should keep the d5-square under control to prevent Black’s knight from moving there. Besides, it is useful to prevent ...b4, as well as to support a possible e5-e6. Other queen moves are hardly ever played, for good reason: 9.Qe2?! Nd5 10.Ne4 was seen in Warter – Hillenbrand, Heusweiler 2007, and here I recommend:

462

10...Nc6!N 11.Be3 Ndb4 12.a3 Bf5! 13.axb4 Bxe4 14.Qd2 f6ƒ Black has completed his development and is already putting White’s centre under pressure, so it is obvious that the opening has been a success for him. 9.Qc5?! loses time: after the further 9...Nfd7 10.Qd5 Nb6 11.Qxd8 Rxd8 the exchange of queens favours Black, who can exert pressure on the weak d4-pawn. 12.Bg5 Nc6 13.Rd1 Now in Garber – Lehiste, Estonia 2000, Black should have made a final developing move before striking at the centre:

13...Be6!N 14.b3 f6 15.exf6 exf6 16.Be3 Bf8 17.Be2 Bb4 18.Rc1 Nd5 19.Bd2 Ba3 20.Rd1 Ncb4ƒ Black has a lot of activity and the pawns on a2 and d4 are weak, all of which makes the position unpleasant for White. 9.Qd3?! Nd5 10.Nxd5 Qxd5 11.Be3 occurred in Stimpel – Ryba, Bad Wiessee 1997, when it would 463

have been natural and strong to continue with:

11...Rd8!N Preparing ...c5 with a lot of force, and if 12.Rc1 then 12...c5! comes anyway: 13.Be2 (13.Rxc5?! Qxa2 14.Be2 Qxb2 15.0-0 Bf5µ leaves White without any visible compensation for the pawn) 13...Qxa2 14.Qc3 Be6 15.Qxc5 Nd7ƒ The b2- and d4-pawns are weak, and Black’s pieces are more active.

9...Nfd7 Other knight moves have been tried, but the text is the main line and the most logical. The knight supports the undermining of White’s centre with ...c5, and may also be relocated via b6 to d5. 9...Be6?! is again inadvisable due to 10.Qa3 Nd5 11.Bxb5² when Black’s compensation for the lost pawn is insufficient. 464

An important tabiya has been reached, in which White has tried several continuations. We will pay particular attention to the interesting sidelines A1) 10.Be2!? and A2) 10.Ng5!?, followed by the three most popular options: A3) 10.h4!?, A4) 10.Be3 and A5) 10.e6. 10.Bf4?! c5 puts White’s centre under serious pressure. 10.a4?! White lingers with his development in the hope of making gains on the queenside. 10...c5 11.axb5 (another game continued 11.e6 fxe6 12.axb5 cxd4 13.Ne4 Nb6 14.Bd3 Qd5µ when Black enjoyed an extra pawn and central domination in Beliavsky – Fluvia Poyatos, Barcelona 2007) 11...cxd4 12.Nxd4 Bb7 13.Be3 (after 13.Bf4 Nxe5 14.Rd1 Qc7 15.Be2 Nd3† 16.Rxd3 Qxf4³ Black’s powerful bishops drilled both of the long diagonals in Aalderink – Gretencord, email 2009) 13...Nxe5 14.Rd1 Qa5 15.h3 This occurred in Cruz – Bissmann, corr. 2010, and here I can recommend:

15...axb5N 16.Qxb5 Qc7³ The pressure from the strong bishop on b7 makes it difficult for White to finish the development of his kingside. A1) 10.Be2!? This calm move leads to interesting play, and several strong GMs have experimented with it. 10...c5 11.e6 Trying to spoil Black’s pawn structure is thematic, but it strengthens the g7-bishop. 11.dxc5 Nxc5 leads to a quieter game. Play continues 12.Qd5 Qxd5 13.Nxd5 and now 13...Nbd7! is a nice detail, temporarily sacrificing the e7-pawn: 14.Be3 (14.Nxe7† Kh8 15.Nxc8 Raxc8 leaves White unable to save the e5-pawn: 16.Bf4 Rfe8 17.0-0 Nxe5= and Black is fine) 14...Bb7 15.Rd1 Rfd8 16.Nxe7† Kf8 17.Nd5 Ne6 18.Ng5 Bxe5 19.0-0 Nd4 20.Bxd4 Bxd5 21.b3= The position is equal and 465

further exchanges are inevitable, Roubaud – Machycek, corr. 2012. 11...cxd4! 11...fxe6 has been chosen by some strong players but I prefer to offer a piece sacrifice for the initiative.

12.exf7† Accepting the offer is dangerous: 12.exd7 Bxd7 13.Ne4 Be6 14.Qc2 Bd5 15.0-0 Nd7 16.a4 Rc8 17.Qb1 bxa4 18.Re1 occurred in Peigney – Bittner, corr. 2011, and here I like the idea of mobilizing Black’s central pawns:

18...f5!N 19.Ng3 e5 20.Bxa6 Ra8ƒ With two powerful central pawns for the piece, it is obvious that 466

Black has full compensation and White will have to defend well to contain his initiative. 12...Kh8 13.Ne4 Bb7 14.Nfg5 14.Qd3!? led to an interesting struggle in a correspondence game: 14...Nc6 15.Nfg5 Nf6 16.Qh3 Bc8 17.Qh4 h6 18.0-0 d3 (18...e5!?N is another option which could lead to any result over the board, though the machine just calls it equal) 19.Bf3 Nd4 20.Ng3 Nxf3† 21.Nxf3 Bb7 22.Bxh6 Bxf3 23.gxf3 Rxf7 24.Be3† Kg8 25.Ne4 Qd5 26.Ng5 Rff8 27.Rfd1 Rac8= Foulds – Karakas, corr. 2015. The position is complex and each side has its own weaknesses, but the position remained dynamically balanced throughout. 14...Bxe4 Normally Black would prefer not to trade such a bishop for a knight, but on this occasion preventing the threats of Qh3 and Ne6 is the top priority. 15.Nxe4 Ne5 16.Qh3 A riskier alternative is: 16.Ng5 d3 17.Bf3 GM Sergey Volkov, an expert in the Russian System, played this way quite recently. 17.Ne6 had been seen previously: 17...Qd6 18.Nxf8 dxe2 19.Kxe2 (After the mistaken 19.Ne6? Nbc6 20.f4 Nd3† 21.Kxe2 White went on to draw in Flumbort – Bagi, Budapest 2012, but 21...Nc5!N 22.Rd1 Qxe6† 23.Qxe6 Nxe6–+ would have refuted his play. Black’s two minor pieces are stronger than the rook and the pawn, especially as f7 and b2 are weak and Black has a firm grip on the d4outpost.) 19...Bxf8 20.Rd1 Qc6 21.Kf1 Nbd7³ All this happened in Moskvichev – Rawlings, corr. 2014. Black’s advantage is not as great as in the other example noted; nevertheless, his minor pieces are stronger than White’s rook and pawn. 17...Nxf3† 18.Nxf3 Nc6 19.0-0 e5 20.Re1 Na5 21.Qe6 Qc8 22.Qe7 Perhaps White should settle for 22.Rxe5N 22...Nc4 23.Qxc8 Raxc8 24.Rd5 Rxf7 25.Rxd3 Nxb2 26.Bxb2 Bxb2³ when he should make a draw with accurate defence, although Black’s superior minor piece and potential to make a passed pawn on the queenside still give him reason to strive for victory. In Volkov – Alsina Leal, Linares 2019, Black went back with his queen to d8 to force a draw. Instead he could have played for a win by means of:

467

22...Nc4!N 23.b3 e4 24.Qxe4 Bxa1 25.Bh6 Nd6 26.Qxd3 Rxf7 27.Rxa1 Nf5 28.Bg5 Rd7³ White has some compensation for the exchange, but not enough to claim equal chances.

16...Qd7 17.Qxd7 Nbxd7 Black is poised to destroy the f7-pawn, so White needs to generate some activity to avoid simply being a pawn down. 18.h4 d3 19.Bf1 Rxf7 20.h5 Kg8 21.hxg6 hxg6 22.a4 bxa4 23.f4 A previous game continued 23.Rxa4 Nb6 24.Ra2 Rc8 25.Kd1 Nd5 26.Rxa6 Nf6 27.Nxf6† exf6 28.Rh3 and now in Catt – Yloenen, corr. 2012, Black could have played for a win with:

468

28...Rd7!N 29.Bd2 Ng4 30.Rf3 Rcd8³ Black is somewhat better due to the strong passer on d3, which renders the f1-bishop passive, as well as the active knight on g4, which currently threatens a nasty fork on h2. 23...Nc4 24.Rxa4 Ndb6 25.Ra2 Rc8 26.Bxd3 Ne5 27.Ke2 Nxd3 28.Kxd3 Nd5 29.Ra4 Rc6 30.Ng5 Rf8ƒ Despite the limited material remaining on the board, Black kept some initiative in Vaitonis – Stone, email 2016, due to White’s wandering king and the weak pawns on f4 and b2. A2) 10.Ng5!?

469

This is another not-so-mainstream approach which leads to interesting positions. The American GM Conrad Holt has played it numerous times, and a few elite players including Carlsen have used it as a one-off weapon. White threatens e5-e6 with additional punch, while also introducing tactical motifs such as Nxf7 and Ne6 in certain scenarios. 10...Nc6 Attacking the central pawn, which is no longer defended by White’s knight. 11.Be3 Black has nothing to fear from the following attempts by White to force the play: 11.Nxf7?! Rxf7 12.e6 wins the exchange but Black gets more than enough compensation for it: 12...Nxd4 13.exf7† Kf8 14.Qd1

14...Ne5! Black’s initiative is extremely strong. (14...Nc5 was playable but less convincing in Nakamura – Gelfand, London [rapid] 2013) 15.Bf4 Bb7 16.Bxe5 Bxe5 17.Qd3 c5 18.0-0-0 Qc7 19.Kb1 Rd8 20.Qe3 c4 21.f3 Kxf7µ Nenneman – Jansa Girona, corr. 2016. Black has a bishop and a pawn for a rook, and his mighty knight on d4 in conjunction with the mobile pawn majority on the queenside promises superb play for the minimal material investment. 11.Ne6!? This is slightly better than the above line, but Black is still in good shape after: 11...Na5! 12.Nxd8 Nxb3 13.axb3 Rxd8 14.Nxb5 14.Bxb5!? Bb7 15.Bc4 Bxg2 16.Rg1 Bb7 17.Be3 Nb6 18.Ke2 c5 19.Rgd1 cxd4 20.Bxd4 Nxc4 21.bxc4 was seen in Husemann – Meier, email 2013, when 21...g5!?N÷ would have prevented f2-f4 and left Black with good chances due to the bishop pair. 14...c6 15.Nc3 c5 16.Bg5 cxd4 17.Nd5 Bb7 18.Nc7 Rac8 19.Bxe7 Bxe5 20.Bxd8 Rxd8 21.Nxa6 d3 470

22.Nb4 We have been following Vayser – Glazman, corr. 2010, and here it looks natural to play: 22...Bxb2N 23.Ra7 Nc5 24.Rxb7 d2† 25.Kd1 Nxb7 26.Bc4 Nc5÷ Material is equal and the game should end in a draw, but White has to be more careful due to the threat posed by the far-advanced passed pawn. 11...Na5 Forcing White to decide where to put his queen. Depending on his choice, Black may follow up either with ...c5 or with ...Nb6 followed by landing a knight on c4. 12.Qd1 12.Qb4?! is inaccurate in view of 12...c5 13.dxc5 Nxe5 14.Rd1 Bd7 15.Qh4 h6 16.Nf3 as seen in Mihajlov – Pruijssers, Fagernes 2017, at which point I suggest an improvement:

16...Ng4!N 17.Bc1 e5 18.Be2 Qxh4 19.Nxh4 Nf6 20.Be3 Rfd8 21.Nf3 Nc6 22.0-0 Be6µ Black has a great game due to his control over the important central squares d5 and d4. 12.Qc2!? is playable although Black has no reason to complain after 12...c5. A good example continued: 13.f4 cxd4 14.Bxd4 h6 15.Nf3 Bb7 16.Rd1 Qc7 17.Qf2 b4 18.Na4 Rad8 19.Bd3 b3 20.a3 g5 21.0-0 gxf4 22.Rc1 Qb8 23.e6 fxe6 24.Bxg7 Kxg7 25.Rfe1 Rf6 26.Kh1 e5 27.Qh4 This occurred in Bjuhr – Haugen, corr. 2015, and here I would opt for 27...Qd6N÷ when the position is extremely tangled. White has no advantage though he seems to have adequate compensation for the two sacrificed pawns, for reasons which can be more readily understood by computers than by humans.

471

12...Nb6 Black takes the sting out of e5-e6 and prepares to plonk a knight on c4. We will follow another correspondence game which featured logical play on both sides: 13.h4 h6 14.Nf3 Nac4 15.h5 Nxe3 16.fxe3 g5 17.Bd3 Bg4 18.Qc2 c5 19.dxc5 Nd7 20.Rd1 Qb8

21.Nd5 Nxe5 22.Nxe7† Kh8 23.0-0 Bxf3 24.gxf3 Nxd3 25.Rxd3 Qg3† 26.Qg2 Qc7 27.Nf5 Qxc5 28.Nxg7 Kxg7÷ Bjuhr – Mertens, corr. 2016. Black is not worse at all and he might even search for more than equality in a practical game, due to having only two pawn islands to White’s three.

472

A3) 10.h4!?

GAME 3 Michael Zeihser – Alan Rawlings Correspondence 2013 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.e4 a6! 8.e5 b5 9.Qb3 Nfd7 10.h4!?

This interesting move was first played by Lputian, against Romanishin at Vilnius 1980. It has subsequently been taken up by such strong GMs as Anand, Ivanchuk, Topalov, Kasimdzhanov and Radjabov. White postpones further development of his pieces in order to hack away at the kingside with h4-h5, connecting the rook on h1 to the attack. 10...c5! Black should apply the classic rule: “In response to flank activity, you must counterattack in the centre!” Of course, attacking White’s pawn centre is the main goal of the Grünfeld anyway, so the decision is an easy one to make. With that being said, some strong players have preferred 10...Nb6, but Black’s results with that move have not been encouraging. 11.e6 White should seek to spoil the integrity of the black pawn structure. 11.h5? is a bit rash: after 11...cxd4 12.hxg6 hxg6 13.Ne4 Nxe5 White has opened the h-file but his 473

pieces are not suitably mobilized for an attack. 14.Nfg5 Nbc6 15.Qg3 f6 16.f4 fxg5 17.fxe5 Nxe5–+ left White three pawns down and unable to create serious threats in Rudd – Ahluwalia, Jersey 2018. 11.Qd5?! sets up a queen exchange, an odd idea in conjunction with the aggressive h2-h4. In Lorenc – Zezulkin, Polanica Zdroj 1999, it would have been natural to play 11...Ra7!N 12.dxc5 b4 13.Na4 Nxe5 14.Qxd8 Nxf3† 15.gxf3 Rxd8µ when material is equal and Black is obviously better, because almost all of the enemy pawns are weak. 11.Be3?! was also inconsistent in C. Hartman – M. Bergstrom, Stockholm 1982; it is too late for White to try and keep his centre together. 11...Nc6!N was a logical reply, developing another piece and increasing the pressure. Play may continue 12.dxc5 Ndxe5 13.Nxe5 Bxe5µ when Black enjoys a lead in development and control over the d4-square. 11...c4! Seizing space and forcing the queen to retreat to a less active position. 12.Qd1 Retreating all the way, as going to c2 may enable Black to gain time with a future ...Bf5. 12.exf7† Rxf7 13.Qd1 Nb6 reaches our main line below. 12.Qc2?! Nb6 13.exf7† Rxf7 14.Ng5 Rf8 15.Be3 occurred in Golubka – Von Bahr, Vaxjo 2015, when it would have been logical to continue with:

15...Nc6!N 16.h5 Bf5 17.Qd2 h6 18.Nf3 g5µ Black is fully mobilized and continues to exert pressure on the d4-pawn.

474

12...Nb6! Rather than having a relatively useless extra pawn on e6, Black speeds up his development and seeks activity.

13.exf7† Rxf7 14.h5 It is logical to advance the pawn again, otherwise White’s 10th move will prove to be a loss of time and weakening. Now the question becomes whether White’s attack on the h-file is dangerous or not; will he have enough resources to develop the initiative on the kingside? The pseudo-active 14.Ng5?! removes the defence of the d4-pawn, and the knight causes no great trouble for Black. 14...Rf8 15.Be3 Nc6 16.h5 Nxd4 17.Rc1 was seen in Bass – Gustafsson, Eksjo 1983, when Black could have secured a clear advantage with:

475

17...e5!N 18.Nxh7?! Tempting but wrong. (However, after 18.hxg6 hxg6µ Black can continue ...Bf5 and White’s compensation for the lost pawn is not visible.) 18...Rf4!!µ 19.hxg6 (after 19.Bxf4 exf4 20.hxg6 Bf5–+ the white king is too exposed) 19...Bf5 20.Qh5 Qd6 Black will capture the g6-pawn next, with a large advantage. Let’s see what happens when White abandons the h4-h5 operation and reverts to normal development. 14.Be2 Nc6 15.Be3 occurred in Pataki – Muhari, Paks 1998, and here I suggest:

15...Bf5N 16.0-0 Rc8³ Black has an active game and the isolated d4-pawn is a long-term weakness. It is not easy to suggest a plan for White: having avoided h4-h5, it’s not quite clear what he should do next. 14.Be3 Nc6 15.Ne5?! (White should prefer 15.h5, transposing to 15.Be3 in the notes to the main line below) was a dubious operation in Ehlvest – Sutovsky, Koszalin 1998. In the game, Black exchanged with the knight and got a comfortable position, but he could have punished White’s play more severely with:

476

15...Bxe5!N 16.dxe5 Qc7 17.f4 Be6 18.Be2 Rd8 19.Qc1 Nd4µ White’s attack is not dangerous and the knight on d4 is extremely unpleasant for him. 14...Nc6 While White loads up for an attack along the h-file, Black activates his pieces and aims at the d4pawn. The kingside attack should not be too dangerous, as White’s pieces are not well placed to generate mating threats. 15.hxg6 15.Be3 is occasionally played, but after 15...Nd5 White has nothing better than transposing by exchanging on g6 in the near future. 15...hxg6 16.Be3 White should protect the central pawn while developing a piece. 16.Be2? is a serious inaccuracy. 16...Bf5!? (simply taking the pawn with 16...Nxd4 is also advantageous for Black) 17.Be3 Nb4 18.Rc1 e5 19.a3 (19.dxe5 Rd7 20.Bd2 Nc6µ gives Black tremendous play for a pawn, with strong pressure along the d-file) Now in Kroeze – Van Haastert, Wijk aan Zee 1998, the strongest continuation was:

477

19...Rd7!N 20.axb4 exd4 21.Bg5 Qe8–+ Black will easily regain the piece, and his much more active pieces give him a decisive advantage.

16...Nd5! The knight takes up a proud position in the centre and forces White to worry about a possible exchange on e3. In the event of a knight trade on d5, the black queen will take up an active post in the centre. 17.a4 It is useful for White to chip away at the queenside. 17.Bh6N 17...Bf6 doesn’t change anything: Black can play exactly as in the main line but with the a478

file closed. Simplifying with 17.Nxd5 Qxd5 18.b3 cxb3 19.Qxb3 Be6 20.Qxd5 Bxd5 does not fully solve White’s problems, for instance: 21.Bd3 Bxf3 22.gxf3 Nxd4 23.Be4 Rd8 24.Rd1 Rd6ƒ White was able to hold a draw in Plomp – Jensen, corr. 2012, helped by the well-known strength of two bishops in an open position. However, in a practical game Black would have reasonable winning chances: he has an extra pawn, a strong knight in the centre and the potential to create a passed pawn on the queenside. 17.Be2 Nxe3 18.fxe3 was seen in Nucci – Lahdenmaeki, corr. 2016, and here it would seem logical to focus on White’s weakened dark squares:

18...Qc7N 19.Ne4 Bf5 20.Neg5 Qg3† 21.Kf1 Rff8ƒ Black has the advantage of two bishops, while the white king can hardly feel completely safe. 17.Qd2 Bf5 18.Rc1 occurred in Keuter – Lohmann, corr. 2005, when Black missed a fine opportunity to blow up the centre. 18...e5!N Black has superb offensive capabilities, as shown after: 19.Bg5

479

19...exd4!! (19...Qd7 also leaves White in trouble but the queen sacrifice is stronger and more beautiful) 20.Bxd8 dxc3 21.Qxd5 cxb2 The combination of the b2-pawn and the threats to White’s king proves too much. 22.Bg5 Re8† 23.Be2 Nb4 24.Qd1 Nd3† 25.Kf1 Nxc1–+ White cannot stop the coming of the new queen. 17...Rb8 Defending the b5-pawn. The a-file is conceded, but it doesn’t matter as there are no invasion points. The b8-rook may support a future advance of Black’s queenside pawns, should the position justify it. 18.axb5 18.Be2 Nxe3 19.fxe3 Qc7 reaches a position which I would not want to play with White. The further 20.axb5 axb5 21.Qd2 Bf5 22.Kf1 Na5 23.Nd5 Qd7 24.Nb6 Rxb6 25.Qxa5 Re6 26.Ra3 Rf8³ saw Black keep some edge thanks to his two bishops and more active pieces in Haznedaroglu – Jedinger, corr. 2014. 18...axb5 19.Bh6 The exchange of dark-squared bishops would be useful for White, but we can avoid it easily enough. 19.Be2 Nxe3 20.fxe3 Qc7 transposes to the previous note. 19...Bf6 20.Ne5 White searches for active counterplay. A simple move like 20.Be2?! would be worse, as 20...Rh7 21.Rh2 Bf5 22.Kf1 e6 23.Nxd5 Qxd5µ leaves White with a clear weakness on d4 which is inconvenient to defend. Moreover, his position is

480

devoid of harmony and he has no counterplay.

20...Bxe5 Black naturally cannot tolerate such a knight, but trading off the bishop works well again. 21.dxe5 Ndb4 Black has some excellent outposts for his knights, and White’s advantage of two bishops does not matter because they cannot show their strength. 22.Qxd8† Nxd8 23.Rd1 23.0-0-0 does not solve White’s problems: 23...Ndc6 24.Be3 (24.f3 Bf5 25.Bg5 Rb7 26.f4 occurred in Vertes – Laffranchise, corr. 2012, when 26...Rf8!Nµ would have improved Black’s rook, preparing a possible invasion along the a-file, while the constant possibilities of ...Bd3 and ...Nd3† are just as unpleasant) 24...Bf5 25.b3 In Kazoks – Kudryavtsev, corr. 2014, it would have excellent to play:

481

25...Bd3!N 26.f4 Bxf1 27.Rhxf1 Rbf8µ White must contend with a variety of unpleasant threats, including ...Nxe5, ...cxb3, ...Nd3† and ...Na5. 23...Bf5 24.Nxb5?! White tries to make a draw with the help of a tactical operation, but he falls short of his goal. Perhaps 24.Be3 was the lesser evil, although defending the position a pawn down after 24...Nc2† 25.Ke2 Nc6 26.f3 Nxe5³ could not have been much fun for him in Stoyanov – Gorokhovsky, corr. 2013. 24...Nc2† 25.Kd2 25.Ke2? immediately leads to a hopeless position after 25...Bd3† 26.Rxd3 cxd3† 27.Kxd3 Rxb5 28.Kxc2 Rxf2†–+ when White is material down while his king remains under attack. 25...Rxb5 26.Bxc4 Rxb2 27.f3 Na3† 28.Ke1 Nc6 29.Bxf7† Kxf7 When making his 24th move, White must have been counting on reaching this position and holding a draw. The limited amount of material certainly makes it hard for Black to realize his advantage, but in the correspondence game he eventually managed it. 30.Bc1 Rb3 31.Kf2 Nc4 32.Kg3 N4xe5 33.Bg5 Bc2 34.Rde1 Rb5 35.Bd2 Nd4 36.Bc3 Nec6 37.Bxd4?! With hindsight, it would have been wiser to avoid this exchange. More stubborn would have been 37.Rh7† Kg8 38.Rhh1 Rc5 39.Ba1 e5µ when the bishop on a1 may still come in handy in defence. 482

37...Nxd4 38.Rh8 Bb3 39.Rb1 Rc5 Black’s general scheme is to mobilize his pawns and eventually create a passed e-pawn. He should avoid trading off his rook, as it will be needed to put more pressure on White’s position.

40.Rd8 Nf5† 41.Kh2 Bd5 42.Rbb8 Kf6 43.Rbc8 Ra5 44.Rc2 g5 45.Rc7 e6 46.Rf8† Ke5 47.Rc2 Ne3 48.Re2 Kd4 49.Kg3 Kd3 Black pieces have become as active as possible, and the time has come to prepare ...g5-g4 to clear the way for the e-pawn. 50.Rb2 Ra4 51.Rg8 g4! A timely pawn sacrifice to set the e-pawn in motion. White has no choice but to take it, otherwise Black simply plays ...gxf3 and wins the f-pawn within a few moves. 52.fxg4 e5 53.Rg5 e4 54.Re5 Ra8 55.Rf2 Rd8 56.Re7 Nc4 57.Ra2 e3 White is powerless to stop the passed pawn. 58.g5 Nd6 59.Ra3† Kd2 60.Kg4 e2 61.Rae3 Bc4 62.g6 Nc8 63.R7e4 Bd3 64.Re5 Rd5 65.Re6 Rd6 White resigned in view of 66.g7 Rxe6 67.Rxe6 Bh7 68.Kg5 Bg8 69.Re8 e1=Q 70.Rxe1 Kxe1, when there is nothing he can do to stop Black from walking his king back to gobble the g-pawns, followed by mating with bishop and knight. 0–1 A4) 10.Be3

483

This calm move was introduced by Portisch against Ribli, at Budapest 1971. Many other strong players have used it, including Karpov, Bareev, Beliavsky, Kozul and Piket, although so far it has not attracted any of the current world elite. White’s idea is easy enough to understand: he develops a piece and supports his centre. 10...c5 Attacking White’s centre is the standard plan and requires no further explanation. 11.e6! White tries to spoil Black’s pawn structure. This is a worthy idea in its own right, and it is especially logical when you consider that ...cxd4 was threatened, after which the e5-pawn could easily have been destroyed. Other continuations are clearly worse: 11.Be2?! is a strange pawn sacrifice. 11...cxd4 12.Nxd4 Nxe5 13.0-0 (13.Rd1 leads to the 11.Rd1?! line below) In Morley – Hryniw, corr. 2006, Black could have maximized his advantage with:

484

13...Ng4!N 14.Bxg4 Bxg4µ With a strong pair of bishops, in addition to the extra pawn. 11.Rd1?! is another dubious way to sacrifice a central pawn: 11...cxd4 12.Nxd4 Nxe5 13.Be2 Qc7 14.Nd5 Qb7 15.0-0 e6 16.Nb4 Qe7 17.Rc1 Bb7 18.Rfd1 a5 19.Nd3 Nc4 It is surprising that this position has been reached in several correspondence games. 20.Nc5 (20.Qxb5 regained the pawn in Deforel – Odeev, corr. 2015, but 20...Nxe3N 21.fxe3 Bh6µ would have left him with big problems due to the weakness of the e3-pawn and the dark squares generally; alternatively, 20.Nxb5 Nxe3 21.fxe3 Na6 22.Nf4 occurred in Kund – Pascual Perez, email 2010, when 22...Be5N would have been excellent for Black thanks to his two bishops and better pawn structure) 20...Bd5 White is already clearly worse, and the further mistake 21.Qxb5? Nxe3 22.fxe3 Qg5–+ left him with insurmountable problems in Chomicki – Lloyd, corr. 2015. 11.Qd5 Ra7 12.dxc5 was the innocuous continuation of Galliamova – Qin Kanying, Jakarta 1993. Black’s play can be improved with:

485

12...b4!N 13.Na4 Nxe5 14.Qxd8 Nxf3† 15.gxf3 Rxd8³ White’s pawn structure is unsuccessful. 11.dxc5?! Nxe5 12.Nxe5 Bxe5 13.Rd1 is a misguided attempt to play actively; after 13...Qa5 14.Bd2 Be6 15.Nd5 Qd8 16.Bh6 the arising complications prove favourable to Black after:

16...Nc6! 17.Nf6† (17.Bxf8 Qa5† 18.Ke2 Rxf8 leaves White’s king horribly misplaced, which is much more relevant that his extra exchange: after 19.g3 b4 20.Qc4 Bxb2 21.Bg2 Rb8 22.Kf1 Rb5–+ Black was ready to destroy the c5-pawn and White’s king was still unable to find refuge in Balutescu – Maffei, corr. 2010) 17...exf6 18.Rxd8 Rfxd8 19.Qa3 In Moraes – Falconi, corr. 2008, Black had more than enough compensation for the queen. My suggestion would be:

486

19...Bd5!N Eyeing the g2-pawn; after 20.f3 b4 21.Qa4 Nd4 22.Be3 Bc6 23.Qa5 Nc2† 24.Kf2 f5–+ Black must win.

11...cxd4! This is the most accurate move. With the bishop on e3 under attack, Black does not have to fear the reply of exf7† followed by Ng5, which would have been problematic in the analogous situation after 10.h4. Now the centre is cleared of pawns, the d7-bishop has greater freedom of action and the knights can jump into the centre, attacking the white pieces. 11...c4 worked well in the previous variation but is less convincing here: 12.exf7† Rxf7 13.Qd1 Nb6 14.Ne5 Rf8 15.a4 Bb7 16.axb5 axb5 17.Rxa8 Bxa8 18.Nxb5 N8d7 19.Nxc4 Kh8 20.f3÷ Black still had to prove his compensation for the two sacrificed pawns in Francisco – Garcia, email 2004. 487

We will analyse two main options: A41) 12.exf7† and A42) 12.Nxd4. 12.Bxd4 Bxd4 13.Nxd4 Nc5 14.exf7† Rxf7 transposes to variation A41 below. Accepting the piece sacrifice is too risky: 12.exd7?! Qxd7 13.Rd1 (13.0-0-0? proved even worse after 13...e5 14.Kb1 Qa7 15.Bxd4 exd4 16.Ne4 Be6–+ when White was a pawn down with a worse position in Har Zvi – Greenfeld, Israel 1998) 13...e5 14.Ne2 Bb7 15.Nxe5 White had no space for his pieces, so he returns the piece in an attempt to break Black’s grip. 15...Bxe5 16.Nxd4 In Arakelov – Zakhartsov, Voronezh 2001, Black could have increased his advantage with:

16...Bd5!N 17.Qb4 Nc6 18.Nxc6 Qxc6 It is hard to offer White decent advice, since after 19.f3?! Bxf3–+ his position crumbles. A41) 12.exf7† Rxf7

488

13.Bxd4!? This move leads to simplifications, although Black should retain a minimal advantage. 13.Nxd4 looks like White’s best bet, when 13...Nc5 leads to variation A42 below. 13.Ng5? is an important move to check, but 13...e6! refutes it: 14.Qxe6 (if 14.Nxf7 Kxf7–+ then White will still lose a piece thanks to the fork, and it is obvious that two minor pieces will be stronger than the rook; 14.Nxe6 Qe7 15.0-0-0 dxe3–+ left White with no compensation for the lost piece in Mester – Feher, Hungary 1997) 14...Ne5 15.Nxf7 Nxf7 16.Qd5

16...dxc3! 17.Qxa8 cxb2 18.Rb1 Bc3† 19.Ke2 Bf5–+ White suffers material losses and his king is in grave danger to boot. 489

13...Bxd4 14.Nxd4 Nc5 15.Qd5 Qxd5 16.Nxd5 e5 17.Rc1 17.Nf3?! proved worse after: 17...Bb7 18.Rd1 Kg7 19.Be2 e4 20.Ne5 Rf5 21.Nb6 Rxe5 22.Nxa8 Nd3† 23.Bxd3 exd3† 24.Kf1 Bxa8 25.Rxd3 Rc5µ Jensen – Sorensen, corr. 2011. Black’s two minor pieces are stronger than White’s rook and pawn; moreover, the white king is most unfortunately placed. 17...Nbd7 18.Nc7 Ra7 19.Nde6 Nxe6 20.Nxe6 Nb6 21.Rc6 Rf6 Black had another interesting opportunity to develop an initiative against White’s king: 21...Bxe6!?N 22.Rxe6 Na4 23.b3 Nc3 24.a3 Rac7 25.Rxe5 Rfd7 26.f3 Kg7 27.h4 Kf6 28.Re8 Rd1† 29.Kf2 Nd5ƒ Black has more than enough compensation for the pawn, as White’s kingside is paralysed. 22.Rxb6 In Marez – Van Tricht, corr. 2011, Black offered to trade rooks, but I prefer: 22...Bxe6!?N 23.f3 Kf7³ It should be a draw, but White still has to be careful. A42) 12.Nxd4

I think this is White’s best bet. The knight defends the e6-pawn and prevents the b8-knight from entering the game. 12...Nc5 13.exf7† Rxf7 14.Qd5 White should take the opportunity to force a queen exchange, otherwise his inferior development and centralized king would put him at risk in the middlegame.

490

14.Qb4?! Qd6! 15.Nc2 Nd3† 16.Bxd3 Qxd3 17.Nd4 Bg4 18.Nde2 Bxe2 19.Nxe2 Nc6 20.Qb3 Qxb3 21.axb3 Bxb2µ left Black with a healthy extra pawn in Marcel – Vega Chirino, Havana 1999. Just as bad is: 14.Qd1?! Bb7 15.Rc1 15.Ndxb5 was played in a ‘freestyle’ game, Bednarski – ‘SuperSongoku’, Internet (rapid) 2006, when 15...Nbd7!N would have accelerated Black’s development while preparing to place the knight on e5. Play may continue 16.Nd4 Ne5 17.Qc2 Bxg2! 18.Bxg2 Ned3† 19.Kf1 Bxd4 20.Bd5 e6 21.Bxd4 exd5‚ when White’s king is not comfortable, to say the least. 15...Qd6 16.Nf3 Bxf3 17.gxf3 Nc6 18.Qxd6 exd6 19.Rd1 We have been following Epishin – Svidler, Groningen 1997, and here Black’s play can be improved by:

19...Ne5!N 20.Rxd6 Nxf3† 21.Ke2 Re8 Black has a strong initiative. My analysis continues: 22.Bg2 Nd4† 23.Kd1 Nf5 24.Bxc5 Nxd6 25.Bd5 Bxc3 26.bxc3 Ne4 27.Bxf7† Kxf7 28.Bd4 Rc8 29.Kc2 Nxc3 30.Bxc3 b4 31.Kb3 Rxc3† 32.Kxb4 Rf3µ Despite the material equality and limited resources remaining on the board, White’s situation is extremely difficult: his pawns are weak and Black’s king will approach the pawn on h2. 14...Qxd5 15.Nxd5 Nbd7 It is not easy for White to bring his light-squared bishop out, due to the possibility of ...Bb7 when the g2-pawn may be lost. 16.Nc7!? White lures Black’s rook to the not-so-good a7-square, while hoping to cause some annoyance with further knight hops. 491

It does not help White to go backwards: 16.Nc3?! Ne5 17.Rd1 Ng4 18.Nc2 Nxe3 19.Nxe3 Bb7 20.Be2 Now in Trmota – Vodicka, Czech Republic 1999, it would have been perfect to take control of the d5-square before applying pressure along the c-file:

20...e6!N 21.0-0 Rc8µ Black’s pieces are excellent and his bishops are especially powerful on the long diagonals. 16.Rc1?! is also not great: 16...Bb7 17.Nb4 (17.b4 Bxd5 18.bxc5 occurred in Van Dijk – Petkov, email 2005, when 18...Nf6!N 19.a4 bxa4 20.Bc4 a3µ would have made life difficult for White) 17...Rc8 18.f3 This was Plagnaud – Amann, email 2009, and here I recommend:

18...a5!N 19.Nbc2 b4 The crucial point is that it is bad for White to win the exchange: 20.Bc4? (20.b3 492

is the lesser evil although 20...Nb6µ followed by ...Nd5 leaves Black in full control) 20...Nb6 21.Bxf7† Kxf7 22.0-0 Nd3 23.Rb1 Nd5 24.Bf2 Ba6–+ Despite being the exchange up, White’s position in untenable: Black’s pieces are extremely active: the knight dominates on d3 and the bishops cut like knives. 16.b4 Bb7 17.Nc7 was an aggressive attempt in Savic – Cukic, Yugoslavia 1999. The natural reply is:

17...Rc8N 18.Ncxb5 Na4! 19.Na3 Ndb6© All of Black’s pieces are mobilized, and it is difficult for White to develop the kingside and secure his king. White’s extra pawn is not felt at all, and ...Nd5 is coming; in short, Black has more than enough compensation. Finally, 16.Rd1!? is not a bad move although Black does fine in the following forcing line: 16...Bb7 17.Nc7 Rc8 18.Nde6 Bxb2 19.Bxc5 Nxc5 20.Nxc5 Rf5 21.N7xa6 Bxa6 22.Nxa6 Bc3† 23.Ke2 Rf6 24.Ke3 (24.f4 does not change matters: 24...Rxa6 25.Kf3 Rc5 26.Rb1 Ra5 27.Bd3= and neither side had a chance to gain the advantage in Johnston – Laffranchise, corr. 2012) 24...Re6† 25.Kf3 Rxa6 26.Bxb5 Rxa2= Chekmasov – Dulany, corr. 2013. In the final position a draw is practically inevitable, as long as White takes a bit of care to look after his ‘active’ king.

493

16...Ra7 17.Ne8!? White needs to play quite inventively to avoid falling into serious difficulties. 17.Nce6? The knight may look beautiful on e6, but it does not justify giving up a pawn. 17...Nxe6 18.Nxe6 Bxb2 19.Rb1?! 19.Rd1! was more stubborn although Black kept the upper hand after 19...Bc3† 20.Ke2 Ra8 21.Ng5 Rf6 22.Ne4 Rc6 23.Rc1 b4 24.Nxc3 bxc3 25.Bd4 c2 26.Kd2 Nf6 27.Bxf6 Rxf6 28.Bd3 Bf5 29.Bxf5 Rxf5 30.Rhf1 Raf8 31.f3 Rd8† 32.Kxc2 Ra5µ in Uecker – V. Ivanov, corr. 2009. Material is equal for now, but Black’s rooks are extremely active and the a2-pawn is weak, so White has a difficult defence ahead. 19...Bc3† 20.Kd1 Ra8 21.a4? This leaves the knight on e6 as cannon fodder. White should have tried 21.Nc7, although 21...Rb8 22.Rc1 Be5µ keeps Black as the clear favourite due to his extra pawn. Now in Siegel – Herrera, Havana 1998, Black could best have capitalized on his advantage with:

494

21...Nf6!N 22.Nc7 Bg4† 23.Kc2 Rc8 24.Kxc3 e5 25.axb5 Rfxc7† 26.Kd2 axb5 27.f3 Nd5–+ Black’s activity is too much, and the white king falls under attack. 17...Bxd4 18.Bxd4 Rf8 19.Rc1 The pseudo-active 19.b4?! leads only to the loss of a pawn after 19...Rf4 20.Bxc5 Nxc5 21.bxc5 Bd7 22.Bd3 Bxe8 23.Ke2 Rc7 24.Rhc1 Rg4 25.Kf3 Rg5 26.c6 Rxc6µ as seen in Balta – Sherwood, corr. 2016. 19...Rxe8 20.Bxc5 Nxc5 21.Rxc5 Rd8 22.Be2 We have been following Zalcik – Ainutdinov, corr. 2012. I recommend:

495

22...Be6N 23.b3 Rd4ƒ Black keeps a small initiative, and White still needs to work to secure full equality. A5) 10.e6

This move, which forces the damaging of Black’s pawn structure, can be regarded as the main line of the 8.e5 variation. Like some of the earlier variations, Portisch was also the first to play this one – against Adorjan in Amsterdam in 1971. It has subsequently been used by several leading players, including Karpov, Kasparov and Kramnik just among the K’s. It has also been a popular battleground for correspondence players. 10...fxe6 We will consider A51) 11.Qxe6† and A52) 11.Be3, the latter being the more popular and challenging option. 11.h4?! is unjustified. 11...Nf6 12.Be3 Nc6 13.0-0-0 Rb8 14.Kb1 occurred in Jenkinson – Mercader Martinez, corr. 2014, when the strongest continuation would have been:

496

14...Qe8!N 15.h5 b4 16.Na4 Na5 17.Qc2 b3‚ With a strong attack. 11.a4?! aims for aggressive action on the queenside but 11...c5! counterattacks in the centre, according to the classical principle. An instructive example continued: 12.Qxe6† Kh8 13.Qe4 Ra7 14.Be3 Bb7 15.d5 b4 16.Ne2

16...Rf5! 17.Rd1 Qg8 18.Qb1 White’s play on the queenside clearly did not bring him any benefits in Berko – Ambirk, corr. 2012, and 18...Bxd5N would have left White a pawn down in a clearly worse position. 11.Be2 is a bit too timid. 11...Nf6 12.a4 b4! 13.Qxb4 Nc6 14.Qc5 Qd6 15.0-0 Rb8 16.Rb1 was seen in Sieciechowicz – Piorun, Chotowa 2009. Both sides have certain weaknesses but Black is better placed 497

overall, and this could have been emphasized with:

16...Nd5!N 17.Ne4 (17.Qxd6 cxd6 improves Black’s pawn structure, and after 18.Nxd5 exd5 19.Be3 Rb4µ the pawns on d4, b2 and a4 are weak, while the rook on b4 is superb) 17...Qxc5 18.dxc5 Rb4µ Black wins a pawn for no counterplay, and White must switch to defence. 11.Ng5!? leads to interesting complications which are by no means unfavourable for Black: 11...Bxd4! 12.Nxe6 Bxf2† 13.Ke2

13...Nc5! 14.Nxc5† e6 The upshot of the tactical skirmish is that Black has sacrificed a piece to leave White’s king marooned in the centre. After the further 15.N3e4 Bxc5 16.Nxc5 Nc6 17.Be3 Nd4† the opponents agreed to a draw in Paredes – Yloenen, corr. 2012, but Black might have done well to continue: 18.Bxd4 Qxd4 19.Qc3 Qf2† 20.Kd1 e5 21.Nd3 Rd8 22.Kc1 Qf8! The possibility of a check 498

on h6 prevents the white king from feeling safe. My analysis continues:

23.Be2 Bf5 24.Rd1 Rd4 25.Nxe5 Qh6† 26.Rd2 Rad8 27.Nf3 Rxd2 28.Nxd2 Qxh2 29.Qe3 Qxg2© Black has picked up a third pawn for the piece; meanwhile, the white king remains insecure and the a1rook cannot enter the game. A51) 11.Qxe6† Kh8

The simple recapture on e6 has been tested at World Championship level. 12.Qe4! White gains time against the rook on a8 in order to transfer the queen to h4 for attacking purposes. 499

Other moves are worse, for instance: 12.Be3?! occurred in May – Schwab, Trier 1992, when Black missed a simple opportunity to exploit White’s exposed queen: 12...Ne5N 13.Qd5 Qxd5 14.Nxd5 Nxf3† 15.gxf3 Nc6 16.Bg2 Nxd4µ With the superior pawn structure and strong knight on d4, Black has much the better game. It is also not safe to play 12.Be2?! as in E. Atalik – Zakhartsov, Kazan 2013, when 12...Nf6!N 13.Qb3 Nc6 14.Be3 Qd6! would have prepared to gain additional time against the white queen. Play may continue 15.Rc1 Be6 16.Qc2 Nb4 17.Qd2 Nxa2 18.Nxa2 Bxa2 19.0-0 Bd5µ when Black has collected a pawn for no visible compensation. 12.Ng5?! looks tempting but is too greedy. 12...Nf6 13.Nf7† Rxf7 14.Qxf7 Nc6 15.Be3 occurred in Gagunashvili – Predojevic, Calvia (ol) 2004, and here I can recommend a new idea:

15...e5!N 16.dxe5 Nxe5 17.Qb3 Bb7© Black has more than enough compensation. It is difficult for White to develop his kingside because the bishop on b7 unpleasantly puts pressure on g2, while ...Neg4 is threatened. White’s extra exchange is of no benefit whatsoever. 12.Qd5!? is the best of the minor alternatives, though it is hardly scary for Black. 12...Nb6 13.Qxd8 Rxd8 14.Bf4 Ra7 15.0-0-0 Nc6 16.Be3 reached a position with mutual weaknesses in Slivko – Lloyd, corr. 2014. A logical continuation would be:

500

16...Be6N 17.Kb1 Nb4 18.a3 Bf5† 19.Kc1 N4d5 20.Nxd5 Nxd5÷ Black is not worse: he has plenty of activity for his pieces and is poised to take on e3, gaining the advantage of two bishops. 12...Nb6 13.Qh4 13.Be2 Bf5 14.Qh4 Nc6 15.Bh6 transposes to our main line below; this was the actual move order of the stem game between Portisch and Adorjan from 1971. 13...Nc6 We have reached the key position of the 11.Qxe6† variation, where White must make a choice. 14.Bh6! This natural move is strongest. 14.Be3? is well met by 14...e5 15.Ng5 h6 16.0-0-0 and now in Quiroga – Ripari, Buenos Aires 1998, Black could have developed a strong attack with:

501

16...b4!N 17.dxe5 Qe8 18.Nce4 Nxe5 19.Kb1 Bf5 20.Be2 Na4–+ White faces insurmountable problems: the threats include ...Nxb2, ...c5-c4 and even the simple ...Kg8. 14.Bd3? It’s amazing that this ‘improvement’ over the main line was prepared and played by Karpov in his FIDE World Championship match against Kamsky. 14...Rxf3! Karpov must have either overlooked or underestimated this move in his preparation. 15.gxf3 Nxd4 16.Be4 I recommend a significant improvement:

16...Qf8!N 502

16...Bf5 gave Black plenty of compensation in Karpov – Kamsky, Elista (5) 1996, but White was able to escape with a draw. The text move is stronger; Black hurries to put additional pressure on f3. 17.Bf4 17.Bxa8? Nxa8 18.Qf4 Nxf3† 19.Ke2 Nd4† 20.Ke1 Nc2† 21.Kf1 Qxf4 22.Bxf4 Nxa1–+ leaves White in a hopeless situation, with a puny rook versus two minor pieces and two pawns. 17.Qg3 is not ridiculous, though 17...Bf5 18.Be3 Rd8 still leaves Black with more than enough play for the exchange. 17...Nf5 18.Qg5 Bf6 19.Qg1 Nh4! White’s position is devoid of harmony and he finds it difficult to protect his king. A possible continuation is: 20.Bxa8 Bxc3†! 21.bxc3 Qxf4 22.Qg3 Nxa8 23.Qxf4 Ng2† 24.Kd2 Nxf4–+ We reach an unusual material balance, with White’s two rooks versus Black’s three minor pieces plus extra pawn. The latter should prevail, as the knights have plenty of well-supported outposts and White has numerous pawn weaknesses which Black can target once he gets coordinated. 14...Bf5 15.Be2 15.Ng5? would be devastating if it were not for 15...Qxd4, winning a pawn and refuting the attack. 15.Bxg7† Kxg7 16.Be2 occurred in Ostenstad – Sigfusson, Reykjavik 1988, when Black has a choice of improvements: 16...e5N is the simplest option, transposing to 16.Bxg7† Kxg7 in the note below; 16...e6!?N is another good option available against White’s chosen move order.

15...e5! 16.Ng5 There is nothing to fear from 16.Bxg7† Kxg7 17.Qxd8 Raxd8 18.dxe5 Rfe8 when Black regains the pawn with a slight advantage in piece activity.

503

16.Bg5 is also not scary: 16...Qe8 17.dxe5 b4 18.Ne4 Bxe4 19.Qxe4 Nxe5 20.Nxe5 Qxe5 21.Qxe5 Bxe5 22.Rb1 Rae8ƒ Gerbich – Martynov, corr. 2014. Due to the fact that the white king did not have time to go to a safe place, Black has the initiative. 16...Bxh6 17.Qxh6 Qe7 18.dxe5 Nd4! 19.Rd1 c5 20.Rd2 All of this happened in Portisch – Adorjan, Amsterdam 1971 – the original game in which 10.e6 was introduced. I like the following new idea:

20...Rae8!?N The game continuation of 20...b4 21.Nd1 Nd5 also gave Black adequate counterplay, and he went on to win a good game. 21.f4 Qg7 22.Qxg7† Kxg7 Black will either regain his pawn or provoke some other kind of concession, for instance: 23.Nf3 I also considered 23.g3 b4 24.Nd1 h6 25.Nf3 Nxe2 26.Kxe2 Nc4 27.Rd5 Be4 28.Rxc5 Nxe5 29.Nxe5 Bxh1 when White has some compensation for the lost exchange, but he is fighting for equality rather than being better. 23...Nxe2 24.Rxe2 Bg4 25.Kf2 Rxf4ƒ Black has the initiative due to his more flexible pawn structure. A52) 11.Be3

504

This is the main continuation, strengthening the d4-point and preventing the thematic ...c5. Most of the strongest players have favoured this way of playing. White keeps in reserve the possibility of taking on e6, either with the queen or with the knight (following Ng5). 11...Nf6 The knight on d7 stood poorly so Black returns it to a more influential square, while incidentally defending e6. His further plans include ...Qd6, ...Nc6-b4 and finally ...Nbd5, getting a comfortable game. We will analyse the overaggressive A521) 12.h4 followed by the main line of A522) 12.a4!. 12.0-0-0?! is too risky, and 12...Qd6 13.Bd3 Nc6 14.Ne4 Nxe4 15.Bxe4 Bb7 16.Qc3 Nb4 17.Bb1 c5– + saw White fall under a decisive attack in Shabalov – Van Wely, Liepaja 2004. 12.Ne5?! gives nothing; in Tessedik – Kosa, Hungary 2012, the simple 12...Qd6!N 13.Be2 c5µ would have hit White’s centre while emphasizing the instability of the knight on e5. The creative 12.g3 Nc6 13.Bg2 was tested at a high level in Salgado Lopez – McShane, Germany 2019. White’s bishop development weakened the c4-square, which Black could have exploited with:

505

13...Na5!N 14.Qc2 Bb7 15.0-0 Nc4³ Black keeps an extra pawn and can prepare ...e5 in the future. 12.Rd1 Qd6 13.Ng5 Nc6 14.Be2 h6 15.Nf3 is a peculiar idea from correspondence play. 15...g5 16.a3 Bb7 17.Qc2 b4 18.Na4 bxa3 19.Nc5 axb2 20.Qxb2 Bc8 21.0-0 Rb8÷ Black had two extra pawns in Kopeikin – Yloenen, corr. 2012. It is hard to realize his material advantage because of his dodgy structure, but I would still take the extra pawns in a practical game. Another correspondence game continued 12.Be2 Nc6 13.0-0 Qd6 14.Rac1 when we come to an instructive moment:

14...Na5! (14...Nb4?! is thematic but it does not work so well here due to 15.Ng5ƒ) 15.Qc2 Bb7 16.Rfd1 Rad8 17.Nd2 Now in Swahnberg – Kuosa, corr. 2007, it would have been good to return the 506

knight:

17...Nc6!N The knight did its job of disrupting White’s plans, so now it returns to carry out the thematic regrouping. 18.Bf3 Nb4 19.Qb3 Nbd5 Black has a good position. A natural continuation is 20.Nde4 Nxe4 21.Nxe4 Qb6 22.Nc5 Nxe3 23.fxe3 Bxf3 24.gxf3 Rd5ƒ when Black can look to develop his initiative on the kingside, possibly with the help of a future ...e5 break. One should also not forget that Black still has an extra pawn. A521) 12.h4

This move is certainly ambitious but I think it’s too optimistic.

507

12...Nc6 13.Rc1 White has tried a few other moves: 13.Bd3 Qd6 14.Rc1?! Bb7 15.h5 Nxh5 16.Rxh5 gxh5 17.Ng5 was the unconvincing continuation of Belous – Codenotti, Kirishi 2012. Black has various routes to a big advantage but the best is:

17...Kh8!N 18.Qd1 Qh2 19.Be4 Bh6–+ Black is the exchange up with the initiative to boot. 13.Rd1 Qd6 14.Bd3 Nb4 15.Ne4 Nxd3† 16.Qxd3 Qd5 17.Nc5 occurred in Kopeikin – Kain, corr. 2012, and here it looks interesting to capture space on the queenside:

17...a5!?N The justification is that 18.Qxb5 allows 18...Ne4!ƒ with a useful initiative for Black; the white king does not feel completely safe. 508

13.0-0-0 is a somewhat risky continuation, as White provokes an attack. A good example continued: 13...Rb8 14.Kb1 b4 15.Na4 Na5 16.Qc2 b3 17.axb3 Nxb3 18.Bc4 Na5 19.Be2 Nd5 20.Nc5 Qe8 21.h5 e5! 22.hxg6 Bf5! 23.Bd3 (23.gxh7†? Kh8 24.Bd3 loses to 24...Nb4) 23...Qxg6 (23...Nb4?? is impossible here due to 24.Bxf5!+–) 24.dxe5 Bxd3 25.Nxd3 c5 26.Bxc5 Rfc8ƒ In Cardelli – Van Leeuwen, corr. 2015, White had to reckon on the unpleasant drilling of the enemy rooks along the queenside files. The text move is a rare but quite natural try, which has been played only once, in Heltzel – Henrichs, Maastricht 2011. I found an interesting new idea:

13...Nh5!?N With the simple idea of going to f4. The game continued with the thematic 13...Qd6, which is perfectly sensible and offers Black at least equal chances. 14.Be2 14.d5 is an obvious move to check, but 14...Na5 15.Qa3 exd5! 16.Qxa5 d4 17.Rd1 c5 18.Qxd8 Rxd8 19.Ne2 Bb7µ offers Black more than enough compensation for his piece, due to White’s poor development and the central pawns which threaten to steamroll through White’s position. 14...Nf4 15.Bxf4 Rxf4 16.Nxb5 axb5 17.Rxc6 Bb7 18.Qxe6† Kh8 19.Qe3 Rf8 20.Rc2 Rxa2 21.h5 Bxf3 22.Bxf3 Ra1† 23.Rc1 Bxd4 24.Qh6 Rxc1† 25.Qxc1 Qd6µ Black has an extra pawn and the white king does not feel safe.

509

A522) 12.a4! Utilizing the positioning of his queen, White tries to destroy the pawn structure and connect the a1-rook to the game. 12...b4! I like this move more than the similarly popular 12...bxa4. Black does not want to open the a-file for the enemy rook, so he gives up his material advantage, gaining time and intending to gain active play with his rook coming to the b-file and his knight heading for the b4-outpost. The fact that this move was chosen by such players as Gelfand, Svidler and Kamsky may serve as a certain quality mark. 13.Qxb4 Qd6!? The more popular 13...Nc6 may easily transpose, but there is still a slight possibility that the knight may be better on d7. 14.Qa3 Obviously White avoids trading on d6 and improving Black’s pawn structure. He is still open to a queen trade, but on his own terms. Other retreats suffer from the drawback of leaving the queen vulnerable to further harassment: 14.Qb3?! Nc6 15.Be2 Rb8 16.Qc2 sees the white queen rushing around the board, yet still not finding a safe haven. In Golubka – Kozhuharov, Sunny Beach 2015, the best continuation would have been:

16...Bb7N 17.0-0 Nb4 18.Qb1 Nbd5µ Black has pressure along the b-file and also threatens ...Nxe3, followed by taking advantage of the two bishops. 14.Qc4 is the move which might tempt us to develop the knight a different way: 14...Nbd7!? (there is 510

nothing much wrong with 14...Nc6, but it seems logical to threaten ...Nb6 to gain further time) 15.a5 Bb7 16.Be2 Bd5 17.Qd3

17...Bxf3! Forcing White to spoil his pawn structure. 18.gxf3 (White has no real choice, since 18.Bxf3? Ne5 19.Qe2 Nxf3† 20.Qxf3 Nd5 21.Qe2 Nxe3 22.fxe3 Rab8µ leaves his king stuck in the centre, while the bishop is stronger than the knight) 18...Nd5ƒ Jimenez Molina – Sanchez Rodenas, corr. 2016. White’s bishops cannot show their strength and his king lacks a safe haven.

14...Nc6 This position has been tested in games between strong grandmasters as well as in correspondence play, and the verdict is that Black has no problems. 15.Bc4 511

White takes aim at the e6-pawn while preparing to castle. 15.Be2 Nb4 16.Rc1 Bb7 17.h3 occurred in Moiseenko – Gelfand, Tromso 2013, when Black could have caused problems with:

17...Nh5!N 18.0-0 Nf4 19.Bxf4 Rxf4³ With the advantage of two bishops, plus pressure on the weak d4-pawn. Some correspondence games continued with: 15.Rc1 Bb7 16.h3 Defending against a possible ...Ng4. Alternatives are no better: 16.Ng5 Nd8 17.Bc4 Bxg2 18.Nxe6 Nxe6 19.Qxd6 exd6 20.Bxe6† Kh8 21.Rg1 Bb7³ favoured Black in Burger – Ham, corr. 2007, due to his better pawn structure and the unreliable position of the white king. 16.Be2 Nb4 17.0-0 Ng4! threatens ...Rxf3 and mate on h2. After 18.h3 Nxe3 19.fxe3 Rab8³ White suffered from weakened dark squares while Black’s bishops were excellent in Mesquita – Ham, corr. 2009. 16...Nb4 17.Bc4 Bxf3 18.gxf3 Kh8 19.a5 Rab8³ Galavics – Tazelaar, corr. 2015. Both sides have pawn weaknesses but White suffers from a lack of a safe haven for his king. His bishop pair does not matter, as the black knights have a number of good outposts. 15...Nb4 The knight is well placed here, threatening a fork on c2 and possibly rerouting to d5 later. 16.Qb3 Guarding c2 and taking aim at e6.

512

Another continuation is: 16.Rc1 Bb7 17.a5 (17.Bd2 Bxf3 18.gxf3 Kh8 19.Ne2 a5 20.Bxb4 axb4 21.Qb3 c6 22.Bxe6 occurred in Pavlicek – Husemann, email 2013, when 22...Nh5Nµ would have given White big problems due to his compromised pawn structure on the kingside) In Bercys – Sarkar, Vancouver 2012, it would have been a good time to go on the offensive with:

17...Ng4!N 18.Na2 Nxe3 19.fxe3 Nd5 20.Qxd6 exd6µ With the bishop pair and the better pawn structure. 16.0-0? This may look like a novice-level blunder but it was the choice of a GM with a rating of 2551. White sacrifices the exchange but does not get enough compensation for it. 16...Nc2 17.Qa2 17.Qxd6 was objectively the lesser evil, but improving Black’s pawn structure was clearly not part of White’s plan. After 17...exd6 18.Rac1 Nxe3 19.fxe3 Bh6 20.Rfe1 Rb8³ Black has a pleasant advantage thanks to the two bishops. 17...Nxa1 18.Rxa1 Kh8 19.Ne5 In Holt – Troff, Dallas 2016, it would have been good for Black to play:

513

19...Bb7!N 20.Bxe6 Rad8 Black has well-developed pieces and a material advantage. The key point is that White is unable to win back the exchange: 21.Nf7†? Rxf7 22.Bxf7 Ng4 White loses by force, for instance: 23.g3 Qf6 24.Qc4 Qf3 25.d5 Rf8 26.Be6 Nxe3 27.fxe3 Qxe3† 28.Kh1 Rf2–+ With devastation on the kingside.

16...Bb7! A great solution! A true Grünfelder should strive for an active, dynamic game, without fear of sacrificing material. Here the decision is easy to justify, as the bishop will exchange on f3 to compromise White’s pawn structure on the kingside, which will easily compensate for the lack of a 514

pawn. 17.Bxe6† 17.0-0 Bxf3 18.Bxe6† Kh8 19.gxf3 is a simple transposition to our main line. 17...Kh8 18.0-0 White cannot avoid the damage to his structure, so he should just finish his development. 18.Ng5? loses to 18...Bxg2 19.Rg1 Rab8–+ when Black has decisive threats of ...Nd3† and ...Qxh2. 18.Bc4? is also bad in view of 18...Rab8 19.Qd1 Ng4! with a huge initiative. 18...Bxf3 19.gxf3 Rab8 20.Bc4 The bishop returns to base to help protect White’s weaknesses. An interesting alternative is: 20.Ne4!? Nxe4 21.fxe4 Bxd4 22.Bxd4† Qxd4

23.Rad1 Qxe4 24.Rfe1 Qf4÷ Mesquita – Schoch, corr. 2016. White has sacrificed a pawn in order to activate the rooks; it should be enough for him to maintain the balance, but no more. 20...Qc6 21.Be2 21.d5?! proved worse after 21...Qd7 22.Qd1 Qh3 23.Re1 Qh4 24.Bf1 Rbd8³ in Butov – Coplin, corr. 2014. White’s d-pawn is doomed and he will remain with doubled f-pawns and an exposed king. 21...Nfd5 The knight is heading for f4. White’s extra pawn is not felt, and his compromised structure on the 515

kingside is the most relevant feature of the position. 22.Rad1 It would be much worse to swap knights: 22.Nxd5? Nxd5 The exchange brings no benefit to White, but Black’s rook becomes more active without the knight blocking it. 23.Qc4 Qd7

24.Ra2? An amazing error for a correspondence game, where it is expected that players will analyse with the help of a computer. More resilient would have been: 24.Rfe1! Rb4! (24...c5? 25.dxc5 Rb4 does not work here due to 26.c6!, and if 26...Qh3? 27.Qxd5 Rh4 28.f4 Qxh2† 29.Kf1+– the queen on d5 connects to the defence of the king, leaving Black hopelessly behind on material) 25.Qxa6 e5µ Black has good chances to succeed on the kingside. 24...e5 This is good enough but 24...c5!N would have been even more efficient, for instance: 25.dxc5 (or 25.Qxc5 Rf5 26.Qa3 Rh5 27.f4 Rh3‚ threatening ...Nxf4 with a decisive attack) 25...Rb4 26.Qd3 (26.c6 is refuted by 26...Qh3 27.Qxd5 Rh4–+ and White loses because of the f1-rook) 26...Rh4‚ Followed by ...Qh3 when the end is nigh. 25.dxe5? Apparently, White simply did not want to play and thus allowed a quick finish. 25.Rd1 would have lasted longer although 25...Nxe3 26.fxe3 Qh3 27.Rf1 exd4 28.exd4 Rf5‚ should still lead to the same result. 25...Rb4 26.Qxa6 Qh3 27.f4 Nxe3 28.fxe3 Qxe3† 29.Kg2 Rbxf4 30.Ra3 Rf2† 0–1 Husemann – Schenning, email 2012. 516

22...Qd7 23.Qc4 c6 24.Kh1 Nf4 25.Rd2 25.Rg1N does not change matters after 25...Nbd5 26.Rd2 a5ƒ when White’s extra pawn is absolutely not felt, and he will have to struggle to equalize. We have been following Beres – Procopi, corr. 2014. I think Black should have played:

25...Nbd5N Improving the knight and intending to improve the position with ...Qh3, or perhaps simply ...a5. My analysis continues: 26.Qxa6 Ra8 27.Qc4 Qh3 28.Rg1 Nxe3 29.fxe3 Nxe2 30.Qxe2 Rxf3 31.Nd1 Raf8 32.a5 Qh4ƒ White’s extra pawn will not be promoting anytime soon, while Black’s attacking ideas include ...e5 and ...Qe4. Objectively Black is no worse, and in practice White’s vulnerable king makes his position trickier to handle. B) 8.Be2!

517

This quiet developing move has taken over as the main line of the Russian System. It was recommended by Lars Schandorff in Playing 1.d4 – The Indian Defences and is the choice of most top players nowadays. Sosonko was the first to play it, against Stean in Islington in 1972. The RussianDutch GM went on to play it several more times, so I would suggest calling it the Sosonko Variation. Over the years, it has been used by numerous top players including Kasparov, Anand, Grischuk, Carlsen, Giri, Aronian, Karjakin and Gelfand. 8...b5 Black will have to play energetically and enterprisingly in order to avoid being caught in a vice-like crush. The text move is an obvious choice to drive the queen away. Black will follow up with the typical ...c5 break to gain counterplay against White’s centre. We will analyse B1) 9.Qd3 followed by the more popular B2) 9.Qb3. B1) 9.Qd3 This move enables Black to equalize relatively effortlessly. 9...c5! 10.dxc5 10.d5?! is too ambitious in view of 10...e6 11.Bg5 h6 12.Bxf6 Bxf6 13.e5 Bg7 14.d6 Bb7 15.0-0 Nd7µ as seen in De Santis – Sarno, Rome 1990. The d6-pawn is firmly blocked and the e5-pawn is a serious weakness. 10...Qxd3 11.Bxd3 Bb7 Black has sacrificed a pawn for the initiative, with speedy development and pressure on the e4- and c5-pawns. 518

12.Be3 Ng4 13.Rc1 Nd7 14.Nd1?! This is the most ambitious attempt to play for a win, but White risks too much. 14.Nd5? was the choice of Volkov, an expert in the Russian System, but this is also a mistake: 14...Bxd5 15.exd5 Nxe3 16.fxe3 Bxb2 17.Rc2 Ba3 18.c6 Nc5µ Volkov – Kovchan, Moscow 2012. White’s pawns are not as impressive as they may appear; the d5-pawn is actually an easy target for Black to attack. 14.b4! is White’s best bet, but it leads to simplifications and a likely draw: 14...a5 15.Nxb5 axb4 16.c6

16...Bxc6! 17.Rxc6 Nde5 18.Nxe5 Nxe5 19.Nc7 Nxc6 20.Nxa8 Rxa8 21.Bc4 Ra3 22.Ke2 Bd4 23.Bxd4 Nxd4† 24.Kd2 b3= In Satici – Cottegnie, corr. 2014, the rivals agreed to a draw, since further exchanges are inevitable. 519

14...Rfc8 15.b4 Nde5 16.Nxe5 Nxe5 17.Bb1 White keeps his extra pawn for the moment, and he has not had to give up the bishop pair or make any other obvious concession. Nevertheless, Black’s chances are superior: his knight can go to c4 at any moment and the queenside will be blasted open with ...a5. 17...Bc6 18.Nc3 e6 19.Ne2 a5 20.bxa5 Rxa5 We are following a correspondence game, in which Black went on to take control of the queenside with skilful manoeuvring:

21.Nd4 Nc4 22.Ke2 Be8 23.Nb3 Ra4 24.Bf4 Bc6 25.f3 f5 26.Rhe1 Bb2 27.Rcd1 Bc3 28.Rg1 Nb2 29.Rc1 Bg7 30.Be3 Nc4 31.Bd2 fxe4 32.fxe4 Na3µ 520

By now, in Nedoma – Yloenen, corr. 2012, White was clearly about to lose a pawn, either on a2 or e4. Either way, Black keeps the initiative and White will have to fight for a draw. B2) 9.Qb3

This is the more accurate retreat, avoiding the queen trade which occurred in the previous line. 9...c5! Once again, Black temporarily sacrifices a pawn in order to break White’s strong centre and get rid of the backward c7-pawn. 10.dxc5 White should eliminate the brave infantryman immediately, otherwise he risks becoming worse. 10.e5? is too risky: 10...Be6 11.Qd1 (11.exf6?! is bad as usual: 11...Bxb3 12.fxg7 Kxg7 13.axb3 cxd4 14.Nxb5 Nc6 15.Na3 occurred in Kainz – Seyb, Radebeul 2016, when 15...e5!N–+ would have given Black a decisive advantage due to the mighty central pawns) 11...Nd5 12.Nxd5 Qxd5 13.0-0 Nc6 14.dxc5 Qxc5 15.Be3 Qd5 16.Ng5 Qxe5 17.Nxe6 Qxe6µ White was a pawn down for insufficient compensation in Beles – Luptak, Slovakia 2000. 10...Be6! 10...Bb7 is a popular alternative, with pressure on the e4-pawn, but I am not so keen on Black’s position after 11.e5 Nfd7 12.Be3. I prefer the text move, as Black develops quickly and attacks the c5-pawn.

521

11.Qc2 Nbd7 The rare B21) 12.c6!? deserves attention, but B22) 12.Be3 is the main line by far. 12.Ng5 is not scary in view of 12...Nxc5 13.Be3 Rc8 14.Bxc5 Rxc5 15.Nxe6 fxe6 when Black’s activity counts for more than his doubled e-pawns:

16.Qb3 Re5! 17.0-0 (17.f4? Nxe4–+ would be disastrous for White) 17...Nxe4 18.Nxe4 Rxe4 Despite the opposite-coloured bishops and Black’s weakened structure, the extra pawn still counts for something, and White will have to suffer to make a draw. The game continued 19.Bf3 Rc4 20.Be2 Rc6 21.a4 Qd5 22.Qxd5 exd5 23.axb5 axb5 24.Rad1 e6µ when Black increased his advantage and went on to win in Lalith – Gupta, Bhubaneswar 2009. 12.b4 is hardly ever played, as it leads to an early liquidation of the queenside. 12...a5 13.Rb1 axb4 14.Rxb4 Nxc5 15.0-0 Qc8 16.Nd4 occurred in Mamedyarov – Svidler, Astrakhan 2010, when the most straightforward solution would have been:

522

16...Na4!N 17.Ncxb5 Qxc2 18.Nxc2 Bxa2 The position is essentially equal and likely to end in a draw due to the limited material, although White has to be slightly more careful in order to avoid becoming worse, as Black’s pieces are active and better coordinated. 12.0-0!? is a rare move which has been tested just once at the elite level. 12...Nxc5 13.Rd1 In Wojtaszek – Grischuk, Riga (rapid) 2013, it would have been good to play:

13...Qb8!?N Intending to activate the king’s rook on c8. A sample line is 14.Nd4 Rc8 15.Nxe6 Nxe6 16.a3 Ne8 17.Be3 Nd6÷ when White’s two bishops cannot show their strength, while Black’s knights have excellent prospects, with fine squares such as c4 and f4 available.

523

B21) 12.c6!?

This move is rarely played but it forces the play and deserves a close look. 12...Nb8 13.Ng5 White tries to cast doubt on the placement of the bishop on e6. 13.Nxb5? looks tempting but is unsound. 13...axb5 14.c7 Qc8 15.cxb8=Q Qxc2 16.Qxb5 Now in ‘papasi’ – Kovalchuk, Internet (blitz) 2020, I should have played: 16...Rxa2!N 17.Nd4 Qc8 (the aggressive 17...Rxa1!? 18.Nxc2 Rxc1† 19.Bd1 Rd8 20.Ne3 Bh6 21.Ke2 Nxe4 22.Qa5 Rd4 23.Re1 Bd7ƒ is promising but hardly necessary) 18.Rb1 Nxe4 19.Nxe6 Bc3† 20.Kf1 Qxe6 21.Be3 Nd6 22.Qd3 Bxb2µ White is a pawn down and his king is badly located. 13.a4? would be a mistake: 13...b4 14.Nd5 Nxd5 15.exd5 Qxd5 16.c7 Nc6 17.Be3 (17.0-0 Bf5µ forces a queen trade and Black will easily pick up the c7-pawn; nevertheless, this may still have given White more saving chances than the game continuation) 17...b3 18.Qc1 Qa5† 19.Kf1 Qxc7–+ White was not only a pawn down, but had also lost the right to castle in Eames – Alsina Leal, Hastings 2012. 13.0-0?! is another inferior move. After 13...Nxc6 14.Nxb5 I found a strong novelty:

524

14...Nxe4!N (14...Nb4 offered Black comfortable equality in Hernandez Carmenates – Mekhitarian, Sao Paulo 2009, but the text move aims for more) 15.Qxe4 Bd5 16.Qe3 axb5 17.Bxb5 Qc7 Black has more than enough play for a pawn. My analysis continues 18.a4 Bxf3 19.Qxf3 Nd4 20.Qd3 Rfd8 21.Qc4 Qxc4 22.Bxc4 Rdc8 23.Bd3 Nb3 24.Rb1 Rxa4³ when Black will win the b2-pawn and White will have to struggle for a draw. 13...Nxc6 14.Nxe6 fxe6 15.Nxb5 White should grab a pawn while he has the chance. He comes under pressure after: 15.Be3?! Nd4 16.Qd1 16.Bxd4 Qxd4 17.Qb3? occurred in Blasco Coll – Valdes Romero, Barcelona 2017, when Black missed a killing blow: 17...Ng4!!N White has nothing better than 18.Rf1 Nxf2 19.Qxe6† Kh8 20.Qd5 Qe3 21.Qd2 Nd3† 22.Qxd3 Rxf1† 23.Kxf1 Rf8† 24.Ke1 Qg1† 25.Kd2 Qxa1–+ when Black is the exchange up with an ongoing attack against White’s exposed king. 16...Nxe2 17.Qxe2 Here we can improve on Black’s play from Buhmann – Le Quang Liem, Khanty-Mansiysk (ol) 2010:

525

17...Qa5N 18.0-0 b4 19.Nd1 Qb5 20.Re1 Nxe4 21.Bh6 Nc5 22.Bxg7 Kxg7³ Black’s extra pawn is of some value, despite being doubled and isolated. 15...Rc8 16.Nc3 Nd4 17.Qd3 Nd7 18.0-0 Nc5 Black has a lot of activity and White will not be able to keep his extra pawn for much longer. 19.Qh3!? This seems like the most ambitious way to play. 19.Qd1 Nxe2† 20.Qxe2 Qd3 21.Re1 Qxe2 22.Rxe2 Nd3 is fine for Black, for instance:

526

23.Kf1 a5 24.f3 a4 25.a3 Nxb2 26.Bxb2 Bxc3 27.Rc2 Bxb2 28.Rxb2 Rc3= With a drawn endgame, M. Urban – Mamrukov, corr. 2014. 19...Na4 20.Bg4 Nxc3 21.bxc3 Rxc3 22.Be3 Qd6 23.Rac1 We have been following Oreopoulos – Van Unen, corr. 2013. I like the following active continuation:

23...Nc2!N 24.Bxe6† Kh8 25.Rcd1 Nxe3 26.Rxd6 Nxf1 27.Rd7 Rxh3 28.gxh3 Nxh2 29.Kxh2 Rxf2†= The complications have subsided, and another equal endgame has arisen. B22) 12.Be3

GAME 4 Aart van de Peut – Arild Haugen Correspondence 2016 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.e4 a6! 8.Be2! b5 9.Qb3 c5! 10.dxc5 Be6! 11.Qc2 Nbd7 12.Be3

527

This multifunctional move is the strongest and most popular choice. White develops the bishop and guards the c5-pawn, while also taking control of the d4-point. 12...Rc8 Rolling up the rook to win back the c5-pawn 13.Rd1 White is in no hurry to part with the c5-pawn. The rook on d1 stands well, putting pressure on the central squares and, thanks to the X-ray along the d-file, Black cannot immediately win back the pawn. Other continuations lead to an initiative for Black: 13.Ng5 Nxc5 transposes to the 12.Ng5 line in the note on page 278 above. 13.e5? merely adds another weakness, and after 13...Ng4 14.b4 Nxe3 15.fxe3 Qc7µ Black picks up the e5-pawn with an obvious advantage. 13.b4? is a bad idea, allowing the obvious 13...Nxe4!. White was obviously hoping that 14.Nd4 would turn out well for him, but Black can initiate a tactical storm: 14...Ndxc5! 15.Rd1 (15.bxc5? Bxd4–+ is simply winning) 15...Na4 16.Nxe6 Bxc3† 17.Kf1 Nd2† 18.Bxd2 Bxd2 19.Qxd2 Qxd2 20.Rxd2 fxe6µ In Dahm – Willms, Germany 1990, Black had an extra pawn (albeit doubled), but a more significant feature of the position was White’s misplaced king, which in turn prevents the rook on h1 from entering the game. 13.0-0 Nxc5 14.Rfd1 causes no problems. I propose an improvement over Morefield – Rohan, Linthicum 2018:

528

14...Qe8!N Taking control of the a4-square, so that White’s knight cannot go there in response to ...b4. My analysis continues: 15.e5 Ng4 16.Qd2 b4 17.Bxc5 Rxc5 18.Ne4 Rd5 19.Qxb4 Nxe5 20.Nc5 a5! Luring the queen to a bad square. 21.Qxa5 Qc6 22.Rac1

22...Nc4! A handsome tactical trick. 23.Nxe6 Rxa5 24.Nxf8 Bxf8 25.Bxc4 e6³ Black has decent chances to prove the superiority of his queen over White’s rook, knight and pawn. The most significant of the rare alternatives is: 13.c6 Rxc6 14.Nd4 This exchanging operation has been employed by several strong GMs, but I think it only helps Black. 14...Rd6 15.Nxe6 Rxe6 16.f3 16.0-0 was tested at a high level in A. Onischuk – Rodshtein, Khanty-Mansiysk (ol) 2010. I 529

recommend 16...Qb8!N, eyeing both the c-file and the kingside dark squares. Play may continue 17.Rac1 Rc8 18.Qb1 h5 19.Rfd1 Rec6 20.Nd5 Rxc1 21.Nxf6† Nxf6 22.Bxc1 Qe5 23.f3 h4 24.Be3 Nh5³ when the knight will go to f4, practically forcing White to trade his dark-squared bishop for it, after which the weakness of the dark squares will be felt. 16...Qb8 17.0-0 Rc8 18.Qb3 Nc5 19.Qb4 Na4! 20.Rad1 Black’s idea is that 20.Nxa4 is met by 20...Nd5! 21.exd5 Rxe3 22.Qd2 Qa7 23.Kh1 bxa4³ when White’s dark squares are weak. 20...Nxc3 21.bxc3 Nh5ƒ White had some problems in Zimmermann – Tiemann, corr. 2014. The c3-pawn needs to be defended, and ...Nf4 must also be reckoned with. 13...b4! Driving the knight away is more accurate than: 13...Qa5?! 14.a3 b4 15.axb4 Qxb4 16.c6 Rxc6 17.Nd4 Rd6 18.Nxe6 Rxd1† 19.Bxd1 fxe6 20.0-0ƒ White had good prospects in Khanas – Morozov, corr. 2019, due to the two bishops and superior pawn structure. 14.Nd5 This is by far the most popular sequel. The knight is strong in the centre, so Black is practically forced to exchange it. The alternative is also playable: 14.Na4!? Qa5 15.0-0 15.Nd2? is a clear mistake due to the tactical blow 15...Nxe4!. White continued with 16.Rc1 and went on to salvage a draw in Sturt – Bagi, Budapest 2015, but 16...Ndxc5!N 17.Nxc5 Nxc5 18.Bxc5 Bf5 19.Qc4 Bxb2 20.Qxa6 Rxc5 21.Rxc5 Qxc5–+ would have given Black a decisive advantage, thanks to his extra pawn along with two bishops in an open position. 15.b3 is an attempt to stabilize White’s position but it is rather slow. 15...Ng4 16.Bd4 Bxd4 17.Nxd4 Nxc5 18.Nxc5 Rxc5 19.Qd2 Bc8 20.0-0 Qc7 21.Bxg4 Bxg4 22.f3 In Latzke – Smolny, Stuttgart 2015, Black should have continued: 22...Bd7!N 23.Kh1 (23.Qxb4 e5 24.Ne2 Rc2 gives Black more than enough activity for the pawn, which will soon be recovered) 23...a5³ Black’s bishop has the potential to outwork the knight in an endgame with pawns on both flanks, and the open c-file is in Black’s hands. 15...Nxe4 16.Qxe4 16.c6 was played in the high-level game A. Onischuk – Svidler, Foros 2008. Here you can improve on the play of the renowned Grünfeld specialist with 16...Nef6!N when Black has good prospects. My illustrative line continues 17.b3 Bd5 18.Nd4 e6 19.Qd3 Nb8 20.Rfe1 Be4 21.Qc4 Rfe8 22.Qc5 Qxc5 23.Nxc5 e5 24.Nf3 Bxc6 25.Nxa6 Nd5 26.Nxb8 Nxe3 27.fxe3 Rxb8³ when Black has two strong bishops while the pawns on a2 and e3 are weak. 16...Qxa4 17.Ng5 Nf6 18.Qh4 Bf5 19.Qc4 h6 20.Nf3 Be6 21.Qxa6 Qxa6 22.Bxa6 Ra8 23.Bb7 Rxa2 24.Bd4?! White should have settled for equality with 24.c6 Nd5 25.Nd4 Rxb2 26.Nxe6 fxe6 27.Rc1 Nc7 28.Rb1 Rc2 29.Rxb4 Nd5 30.Re4 Nxe3 31.fxe3 Rd8 32.Rxe6 Bf6 33.c7 Rxc7= when a draw is almost 530

guaranteed. The text move is more ambitious, but White is in more danger than Black.

24...Rb8 25.Bc6 Ne8 26.Rd2 Nc7 27.h3 Rd8 28.Rfd1 This occurred in Bredenhof – Doren, corr. 2012, and here I suggest: 28...h5!?Nƒ With ideas such as ...b3 and ...Bh6 on the agenda, Black owns the initiative. 14...Bxd5 The powerful knight should be destroyed immediately. 15.exd5

531

15...Nxc5! Utilizing simple tactics to regain the pawn. White can still put his faith in the bishop pair but Black’s knights will have enough good central squares available, while the isolated d-pawn could become a target in the future – although this pawn also exerts a cramping effect and provides White with a potential outpost on c6, so it is too early to know if it will prove an asset or a liability. 16.0-0 This natural move connects the rooks and provides security to the white king. White’s rare attempts to deviate have not gone well: 16.Qd2? was played in L. Andersen – Hagen, Ballerup 2011, when Black missed the refutation: 16...Nfe4!N 17.Qc2 (White has nothing better than going back with the queen, as 17.Qxb4 Rb8 18.Qc4 Qa5† is terrible for him) 17...Qa5 18.0-0 Qxa2µ White has no compensation for the lost pawn. 16.Bc4?! was another dubious try in Forsberg – Olofsson Dolk, Sweden 2015. White sets himself up to be pinned along the c-file, and Black could have exploited the error with:

532

16...Qa5!N 17.0-0 Na4 18.d6 exd6 19.Rxd6 Nd5 20.Qb3 Nxe3 21.fxe3 Nc5µ The pawns on e3, b2 and a2 are all weak. 16.Bxc5N is harmless, but it is still slightly surprising that nobody has tried it in practice yet. Black regains the piece with 16...Qa5 17.Bxb4 Qxb4† 18.Qd2 and here I would recommend: 18...Qb6!? The fighting move. (18...Qe4 leads to simplifications, as seen after 19.Qd3 Qxd3 20.Bxd3 Nxd5 21.Bxa6 Rc6 22.Bb5 Rc5 23.Ba4 Bxb2= with a level endgame) 19.0-0 Rfd8 20.Qd4 Qxd4 21.Nxd4 Rxd5ƒ Taking advantage of the fact that it is impossible to take on a6 because of ...Rcd8, Black avoids excessive simplifications and forces White to play with a certain amount of accuracy to secure a draw. 16...a5 Black takes the a-pawn out of the enemy bishop’s line of vision, while securing the b4-pawn. The knight on c5 remains indirectly protected due to the would-be pin along the c-file. After a series of natural and in some cases forced moves, we have reached a serious branching point where White has a few options of roughly equal merit. 17.Rfe1 Since White has no great way of forcing the play, he settles for a slight improvement of his rook, hoping for future pressure along the e-file. 17.Bxc5?! is pointless due to 17...Nd7, when 18.Bxe7? only compounds White’s difficulties: 18...Rxc2 19.Bxd8 Rxd8 20.Rd2 Rxb2 21.Rxb2 Bxb2µ The d5-pawn is weak, and Black has excellent chances to promote a pawn on the queenside. 17.Bc4!? This move is recommended by Lars Schandorff in Playing 1.d4 – The Indian Defences. 17...Nce4 533

Schandorff mainly focused on 17...Qd6, as played in Aronian – Grischuk, Kazan 2011, but I prefer the text move. 18.Qe2 The Danish GM evaluated this position favourably for White, but subsequent games and analysis seem to indicate that Black is fine.

18...Qc7 19.Bb3 I was surprised to learn that a strong GM opted for 19.Rc1?!, missing the powerful 19...Nc3! 20.bxc3 Qxc4 21.Qxc4 Rxc4 22.cxb4 Rxc1 23.Rxc1 axb4 24.Rc5 Rd8 25.d6 h6 26.g3 Rxd6µ when Black was able to convert his extra pawn in Potkin – Areshchenko, Yerevan 2014. 19...Ng4 20.Bc1 Another example continued 20.Bd4 Ng5 21.g3 Nxf3† 22.Qxf3 h5 23.h3 Nh6 24.d6 exd6 25.Qf4 Nf5 26.Bxg7 Kxg7 27.g4 hxg4 28.hxg4 Nh4 29.Rxd6 f6 30.Rfd1 Rcd8 31.Kf1 Qe7 32.Qg3 g5 33.Qd3 Rxd6 34.Qxd6 Re8= with simplifications and a subsequent draw in T. Williams – Roques, corr. 2014. 20...Nc5 21.Bc4 Rfd8 22.h3 It seems natural to expel the knight from g4. Another game continued: 22.Bg5 Bf6 23.Bxf6 exf6 24.Rfe1 f5 25.h3 Nf6 26.Ne5 Rd6 27.Qf3 (I also checked 27.Nc6N 27...Kg7 28.Qe5 Re8 29.Qd4 Rxe1† 30.Rxe1 Ne6 31.Qe5 Nc5 32.Qd4 Ne6= when the game may end in a repetition) 27...Re8 28.Nc6 Rxe1† 29.Rxe1 Nce4 30.Rd1 Nd7 31.Qe2 Kg7 32.f3 Nef6 33.Qe7 Ng8 34.Qe3 Ngf6= The threefold repetition was the logical conclusion in Reinhart – Mi. Petrov, corr. 2016. 22...Ne5 23.Bb5 This position occurred in Lesko – Gilbert, corr. 2014, and here I can recommend:

534

23...Nb7!?N Sending the knight on a journey. 24.Rfe1 Nxf3† 25.Qxf3 Nd6 26.Bc6 Nf5 27.Bf4 Qb6 28.g4!? Nd4 29.Rxd4 Qxd4 30.Rxe7 Qf6 White has enough compensation to maintain the balance, but nothing more. For instance: 31.Qe4 I also considered 31.d6 g5 32.Bg3 Qxf3 33.Bxf3 Rc1† 34.Kg2 Bf8 35.Rc7 Rb1 36.d7 Rxb2 37.Be5 Rd2 38.Bc6 Bg7 39.Bg3 Bf8 40.Rc8 R2xd7 41.Bxd7 Rxd7 42.Ra8 Kg7 43.Rxa5 Be7= with an equal endgame. 31...Bf8 32.Ra7 Bd6 33.Bxd6 Qxd6 34.Rxa5 Qf6 35.Qxb4 Rxc6 36.dxc6 Rd1† 37.Kg2 Qxc6† 38.Kg3 Qc7† 39.Kg2 Qc6†= With perpetual check. Several high-level games have continued: 17.Bb5!? Nce4

535

18.Bc6 There is also 18.Qd3!? Nd6 19.Bc6 Nf5 20.Bf4 Qb6 21.Ne5 Nh5 22.Be3 Qc7 23.Nd7 Nxe3 24.Qxe3 Rfd8 when the rivals agreed a draw in Felkel – Schulz, email 2016, in view of 25.Qxe7 Bxb2 26.Qe2 Bc3 with an equal position. 18...Qc7 19.Rc1 Nd6 20.Qe2 Nf5 21.Bc5 21.g3 21...Nxe3 22.Qxe3 Rfd8 23.Rc4 Ne8 24.Rc2 Nd6÷ was fine for Black in Laube – Mi. Petrov, corr. 2014. The knight on d6 is an excellent blocker, but it is extremely difficult for either side to make progress. The present position was reached in Wang Hao – Giri, Biel 2012. My novelty is:

21...Nd7N Eliminating one of the enemy bishops, with the possible continuation: 536

22.Rfd1 Nxc5 23.Rxc5 Qa7 24.Rc2 Rc7÷ Black is not worse. 17...Ncd7! The knight is heading for b6, in order to attack the pawn on d5. 18.Qb1 White still hopes to claim an advantage. Other moves are not particularly ambitious: 18.Qd2N can be met by 18...Ne4 19.Qd3 Nec5 20.Qb1?! (it is better to repeat the position with 20.Qd2=) 20...Nb6ƒ when there are problems with the defence of the d5-pawn. Another untested move is 18.Qa4N 18...Nb6 19.Qxa5 Nbxd5 20.Qa4 e6÷ when Black has a comfortable position; White’s bishops do not count for much and the a2-pawn is weak. 18.Qb3N 18...Ng4! gives Black an improved version of the next note to the main line below. The difference is felt after 19.Bf4 Qb6 20.Bg3 Nh6 when White’s queen is worse on b3 than b1:

21.Nd2 (there is no sense in 21.Bd3?! because 21...Nc5 would be awkward for White) 21...Nf5÷ Black has no problems and I find his position slightly preferable due to the active knights. 18.Bc4 Nb6 19.Bxb6 Qxb6 20.Qe2 occurred in Husemann – Ress, email 2012, when it looks best to transfer the knight to blockading duty:

537

20...Ne8N 21.Bb3 Nd6 22.Ne5 Rc7 23.Nc6 Bf6÷ Black has a comfortable position resembling the main game, where White is less safe than he may appear to be. 18...Nb6 But not 18...Ng4?! 19.Bf4 Qb6 20.Bg3 Nh6 when the exact 21.Bd3!ƒ prevents the knight from going to f5. 19.Ba6 Ra8 20.Qd3

20...Qb8! By preventing Bb7, Black threatens the d5-pawn. 538

Of course you cannot play 20...Nbxd5? due to 21.Bb7 Nxe3 22.Qxe3+– and White wins the exchange. At first I looked at 20...Qd6 but was not happy after: 21.Bb7 Rad8 22.Qb5 Nbxd5 23.Bc6 Rb8 24.Rxd5 Nxd5 25.Qxd5 Qxd5 26.Bxd5 Bxb2 For some reason, the rivals agreed a draw here in Aguiar Garcia – Novak, corr. 2016, but White could have tried to win without any risk. For instance, after 27.Bf4 Rbd8 28.Bb3 e6 29.Kf1² Black’s pawns are not going anywhere, and White can gradually strengthen his position. 21.Bxb6 White must relinquish the advantage of two bishops, otherwise he will lose the d5-pawn. 21...Qxb6 22.Bc4 Ra7

23.Ne5 White is still hoping to squeeze an advantage out of the position. With hindsight, however, he would have done better to simplify and make a draw. 23.d6 exd6 24.Qxd6 Qxd6 25.Rxd6 leads to simplifications, and after 25...Rc8 26.b3 Ne8 27.Rd3 Bf8 28.Red1 Kg7 29.Ne5 Nf6 30.f3 Re8 31.Nd7 Nxd7 32.Rxd7 the players agreed a draw in Brzezinski – Visigalli, corr. 2016. 23...Rc7 24.b3 24.d6 was White’s last chance to simplify the position: 24...exd6 25.Qxd6 Qxd6 26.Rxd6

539

26...Ng4! (26...Re8 leads to more straightforward equality after 27.Bb5 Ree7 28.Rd8† Bf8 29.Ba4 Rc5 30.Nd3 Rxe1† 31.Nxe1 Rc1 32.Rd1=) 27.Nxg4 Rxc4 28.Ne3 Rd4 29.Rxd4 Bxd4 30.Nc4 Rc8 31.b3 a4ƒ The computer just gives zeros, but in a practical game Black could still hope to make use of having a bishop against a knight, especially if he can target the pawn on b3. 24...Ne8! The knight moves to the ideal d6-square, from which it can go to f5 and d4. The manoeuvre also makes room for the bishop to go to f6, where it defends the backward e7-pawn. 25.Nc6 Bf6 26.Qg3 Nd6 27.Qf4 Kg7 28.Bf1 Rfc8 29.Rd3?! White continues to drift towards bigger problems. 29.Bd3! would have stopped the knight from going to f5. After 29...Qc5! 30.Ba6 Ra8 31.Bc4 a4ƒ Black retains some initiative, but White’s chances are better than in the game. 29...h5! 29...Nb5? allows White to carry out his plan with 30.Rh3 h5 31.g4. Black can avoid disaster with 31...g5 32.Qd2 Nd4 33.gxh5 Nxc6 34.dxc6 Qxc6 35.h6† Kh8 with a roughly equal position, but there is no point in allowing the attack in the first place. The text move prevents the Rh3 plan while usefully capturing space. 30.h3 30.Rh3?! is now senseless due to 30...Qc5! 31.Rd1 e6 32.Rhd3 Nf5µ when the knight on c6 no longer feels secure.

540

30.g4?! is also useless in view o30...hxg4 31.Qxg4 Rh8µ when White has merely weakened his own kingside.

30...h4! Seizing more space while preventing White from opening the kingside in a beneficial way. 31.Rf3 Nf5! Practically all of Black’s moves deserve exclamation marks. He is eager to exchange knights and queens to put White in a surprisingly poisonous endgame. 32.Bd3 Nd4 33.Nxd4 Qxd4 34.Qxd4 Bxd4 35.Bc4 Bf6 It would seem that Black does not threaten anything and a draw should be a natural result, but this is just an illusion. The bishop must stand on c4 to prevent the black rooks from invading, but this gives Black the chance to create a deadly passed pawn with the help of a timely exchange sacrifice.

541

36.Re4 a4 37.Rfe3 Bg5 38.Re2 Bf6 39.R2e3 Rxc4! The apparent wellbeing of White’s position turns out to be a hoax. 40.Rxc4? White fails to offer much resistance. The best chance was to keep a pair of rooks on: 40.bxc4! b3 41.axb3 a3 42.Re2 Bb2 43.Re1 a2 44.c5 a1=Q 45.Rxa1 Bxa1 46.b4 Kf8 47.Kf1 (not 47.Rxh4? Rd8 48.c6 Rxd5 49.c7 Rd1† 50.Kh2 Be5†–+) 47...g5µ I believe Black should be able to win with perfect play, but White can still pose technical difficulties.

542

40...Rxc4 41.bxc4 b3 42.axb3 a3 43.Re2 Bb2 44.Rxe7 44.c5 also does not save White: 44...a2 45.Re1 Kf8 46.Kf1 a1=Q 47.Rxa1 Bxa1 48.Ke2 Bd4 49.b4 Bc3 50.b5 Bd4 51.c6 Ke8 52.f3 Bb6 53.Kd3 f5 54.Kc4 Kd8–+ The king walks to d6 and White loses all his pawns on the queenside. 44...a2 45.Ra7 a1=Q† 46.Rxa1 Bxa1 47.Kf1 Bd4 48.Ke2 Kf6 49.f3 Ke5 50.Kd3 Bc5 51.Kc3 f5 52.b4 Be7 With zugzwang upon him, White resigned. This was an extremely instructive and enlightening game, showing how Black can play for a win in the thematic structure with an isolated pawn on d5, when not much appeared to be happening. 0–1 Conclusion The Russian System is an important option against the Grünfeld, which is both popular and theoretically challenging. This chapter picked up on the tabiya after 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.e4 a6!, when we analysed White’s two most important options. 8.e5 is an aggressive move which leads to difficult, somewhat irrational positions which demand careful preparation and accurate handling by both players. After reading this chapter you should be happy to play these positions with Black, as many of the variations give him good chances to take over the initiative. 8.Be2! is a more patient approach which has taken over as the main line at high levels nowadays. 8...b5 is our reply, when 9.Qb3 c5! 10.dxc5 Be6! 11.Qc2 Nbd7 is the critical continuation. We ended the chapter with an instructive game in which White invested his hopes in the bishop pair, but Black’s accurate play demonstrated that the isolated d-pawn had its drawbacks, which were exploited with skilful knight manoeuvres and an appreciation of Black’s chances to break through on the queenside in the endgame.

543

A) 5.Nf3 Bg7 291 A1) 6.Qa4†!? 292 A2) 6.Bg5!? 294 B) 5.g3 Bg7 6.Bg2 Nxc3 7.bxc3 c5 298 B1) 8.e3 299 B2) 8.Nf3 Nc6 301 B21) 9.e3 302 B22) 9.Be3 304 C) 5.Na4!? e5!? 307 C1) 6.e4!? 309 C2) 6.dxe5 Nc6 311 C21) 7.f4 312 C22) 7.a3 313

544

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 Exchanging on d5 is the critical test of the Grünfeld, especially when White follows up with e2-e4. However, we are not quite at that stage yet, as there are a number of other options he may try. In this chapter we will examine A) 5.Nf3, B) 5.g3 and C) 5.Na4!?. 5.Bd2 is an important sideline which will be analysed in the next chapter. The main line is of course 5.e4, and it will be covered in Chapters 12-16. 5.Qb3 is not dangerous. 5...Nxc3 6.bxc3 c5 7.e3 (7.Nf3 Bg7 8.e3 transposes to 7.Qb3 Nxc3 8.bxc3 in the notes to variation B22 of Chapter 4) 7...Bg7 8.Bb5† Nc6 9.Nf3 (9.d5? would be a mistake due to 9...a6 10.dxc6 axb5 11.cxb7 Bxb7 12.Qxb5† Qd7–+ with a double hit on g2 and c3) This position was reached in Farnik – Heitmann, Adelaide 2003, when it would have been good for Black to play:

9...Be6!N 10.Bxc6† (10.Qc2 Rc8 targets White’s queen; and a move like 10.Qb1 can hardly make White happy) 10...bxc6 11.Qc2 cxd4 12.cxd4 Rc8 13.0-0 c5ƒ Black has the more pleasant game with the two bishops. A) 5.Nf3 With this move, White avoids committing to a particular central formation for the moment. 5...Bg7

545

We will analyse A1) 6.Qa4†!? and A2) 6.Bg5!?. 6.g3 Nxc3 7.bxc3 c5 8.Bg2 transposes to variation B2 on page 301. 6.e4 is still possible of course, and the position after 6...Nxc3 7.bxc3 will be covered in Chapters 13 and 14. 6.Bd2 transposes to variation A of the next chapter on page 317. The only other possibility worth mentioning is 6.Qb3!? Nxc3 7.bxc3 c5 8.e3 0-0 9.Ba3!? (the more popular 9.Be2 has been covered on page 110; see the 7.cxd5 line in the notes to variation B23 of Chapter 4). This position was reached in Bruzon Batista – Sutovsky, Pamplona 2004, and numerous other games. I found a promising new idea:

546

9...Be6!N 10.Bc4 (10.Qxb7? is too risky in view of 10...Qa5 11.Bb2 cxd4 12.Nxd4 Nd7 13.Qb4 Qc7 14.Nxe6 fxe6 15.Qc4 Qb6µ when White is way behind in development) 10...Bxc4 11.Qxc4 Nd7 12.00 Qa5÷ Black has a comfortable position and is all set to fight for the c-file with ...Rac8. A1) 6.Qa4†!?

This move is interesting and has scored decently, but it seems to me that White’s queen is not located in the best way. 6...Nc6! 6...Bd7 is a popular choice which was recommended by Avrukh, but I believe the text move is 547

stronger. 7.e3 Securing the d4-pawn. 7.Nxd5 Qxd5 8.e3 0-0 is a simple transposition to our main line. Other continuations are riskier, for instance: 7.Qb3?! Be6! emphasizes the unfortunate position of White’s queen. 8.Qd1 (8.Qxb7? Ndb4 is no good for White) This position was reached in Santiago – Lima, Maceio 2011, when Black could have prepared a powerful blow to the centre by means of:

8...Bg4!N 9.e3 e5ƒ With an obvious initiative. 7.Ne5? is tempting at first glance but the illusion soon fades. 7...Bxe5 8.dxe5 Nxc3 9.bxc3 Qd5 10.Bf4 In Pipia – Kolosowski, Batumi 2010, Black had the luxury of choosing between several promising continuations, the strongest being:

548

10...Bf5!N White will have a hard time completing his development. For instance: 11.Rd1 Qc5 12.Rc1 Be6 13.e3 0-0-0 14.Be2 g5! 15.Bxg5 Rhg8 16.Bf4 Rxg2 17.Bf3 Rg6 18.Bxc6 bxc6µ Despite the damage inflicted on Black’s queenside structure, his position is clearly better. White’s king is stuck in the centre and the light squares are extremely weak, while the bishop on f4 is severely restricted by its own pawns. 7.e4?! This move has been tried by some strong players and has scored quite well, but it enables Black to pick up a central pawn. 7...Nb6 8.Qc2 Nxd4 9.Nxd4 Bxd4 10.Bh6 Bd7 11.Rd1 e5 12.Nb5 12.Ne2? allows 12...Bxf2†! 13.Kxf2 Qh4† 14.g3 Qxh6 15.Qc5 and now in Bazeev – Sargsyan, Internet (rapid) 2019, the simple 15...0-0-0N 16.Bg2 f6–+ would have left White with a hopeless position, with no compensation for Black’s two extra pawns. 12...Bxb5 13.Bxb5† c6 14.Bg7 Qd6! 15.Bxh8 Qb4† 16.Qd2 Qxb5 Having won a pawn earlier, Black has sacrificed an exchange for excellent compensation, as White cannot castle and his pieces are not well coordinated. A good example continued: 17.b3 f5 18.f4 fxe4 19.Bxe5 Bc5 20.Qc2 e3 21.Qe2 Qb4† 22.Kf1 a5 23.h4 Nd5 24.g3 Qe4 25.Kg1 00-0© Black maintained at least enough compensation in Bittner – Schakel, corr. 2014, thanks to his perfectly centralized pieces and the dangerous passed pawn. In a practical game, the whole line looks much more dangerous for White than for Black. 7...0-0 Having castled and broken the pin, Black is ready to strike in the centre with ...e5, so White needs to do something to justify the placement of his queen.

549

8.Nxd5 8.Be2?! was played in a high-level blitz game: 8...e5 9.Nxd5 Qxd5 10.dxe5 Nxe5 11.Nxe5? (11.0-0 is better, although 11...Bd7 12.Qd1 Qxd1 13.Rxd1 Nxf3† 14.Bxf3 Rad8 15.Kf1 b6µ still leaves White with difficulties developing his queenside) In Nakamura – Vachier-Lagrave, Internet (blitz) 2016, both players overlooked a tactical detail:

11...Qxg2!N 12.Bf3 Qh3 13.Bg4 Qh4 14.Bd7 Qxa4 15.Bxa4 Bxe5–+ With a healthy extra pawn. 8.Bb5 was tried in a contest of strong GMs, Wang Hao – M. Andersen, Abu Dhabi 2018. I think 8...Nxc3N 9.bxc3 Bd7 looks pleasant for Black, for instance: 10.Qb3 (10.e4 allows 10...a6 11.Bxc6 Bxc6 12.Qc2 e6³ when Black enjoys the advantage of two bishops) 10...Na5 11.Qa4 c6 12.Be2 c5 13.Qa3 Rc8 14.0-0 Qc7÷ The bishop on c1 is passive, and Black has comfortable play. 8...Qxd5 9.Bc4 9.Be2?! e5 transposes to the Nakamura – MVL game noted above. The text move was played in Gajewski – Swiercz, Warsaw 2012. I propose an improvement:

550

9...Qf5!N 10.0-0 a6 11.Be2 Be6 12.Bd2 Bd5÷ Black is in no way worse, and will soon put White under pressure with ...e5. A2) 6.Bg5!?

White brings the bishop to an active position and does not yet determine his policy in the centre. The Belgian IM Polaczek was the first to play this way at a high level, in 1990. The Russian IM Jaroslav Ulko subsequently helped to popularize it, so I would like to call it the Ulko Variation. To this day, this line has only been tested in a relatively small number of games, and its theory continues to evolve. Mamedyarov, Van Wely and Rychagov are among the strong players who have incorporated it into their repertoires. 551

6...c5 Black carries out the standard hit on the centre, putting the d4-point under additional voltage. 7.Rc1 This is the standard move, preparing Nxd5 and Rxc5 in some lines. 7.e4?! Nxc3 8.bxc3 transposes to the 8.Nf3?! line in the notes to variation B of Chapter 12 on page 341. 7.Nxd5?! Qxd5 8.Rc1 Qxa2 9.Rxc5 was a questionable experiment by White in Mamedyarov – Melkumyan, Doha (rapid) 2016. Black should have continued:

9...Bg4!N 10.h3 Bxf3 11.exf3 0-0 12.Bc4 Qxb2 13.0-0 Nc6³ White’s two bishops do not compensate for the loss of a pawn; moreover, the d4-pawn is isolated and weak. I also checked: 7.e3 cxd4 8.exd4 0-0 A few strong players have continued this way with White, but it does not require serious analysis. Black has an even more pleasant version of the reversed Tarrasch Defence positions from Chapter 4, as White’s bishop has gone to g5 slightly prematurely. 9.Be2 9.Bc4?! proves an unfortunate choice after 9...Nxc3 10.bxc3 Qc7 11.Qb3 Nc6 when the threat of ...Na5 practically forces White to waste a tempo retreating his bishop to e2. 9.Qb3 is another possible continuation: 9...Nxc3 10.bxc3 Nc6 11.Be2 (11.Rd1 was played in Maghsoodloo – Pilvaieh, Rasht 2017, when 11...Qa5!?N 12.Bd2 Qc7 13.Be2 Be6 14.Qa3 Bd5³ would have given Black good prospects against the hanging pawns, with ...Na5 on the way) 11...Qc7 12.0-0 Be6 13.Qa3 Bd5 14.Rfd1 Rfe8 15.Rab1 This occurred in Skytte – C. Pedersen, Denmark 2010, and 552

now 15...a6N³ would have prepared ...b5 followed by ...Na5-c4, with a clamp on the light squares and pressure along the c-file. 9...Nc6 10.0-0 h6 11.Be3 Be6 12.Qd2 Kh7 13.Ne4 We have been following Ozgibcev – Novik, Sochi 2005. Black’s play can be improved with:

13...Qb6!N 14.Nc5 Bf5 15.Rac1 Rad8µ The d4-pawn is isolated and weak. White has no counterplay and must worry about the possibility of ...e5. 7...Nc6! This is the maximalist option, leading to a complicated game where Black has good chances to fight for the advantage. If you are content with a solid, reliable game, you can happily play: 7...Nxc3 8.bxc3 0-0 Having relieved the central tension, Black no longer has to worry about Nxd5 followed by Rxc5. 9.e3 White strengthens the d4-point. 9.e4 is a rare choice via this move order, but it immediately transposes to variation B of Chapter 12; see the continuation of Game 5 on page 342. 9.Qd2!? has been tested by a number of strong players. I propose a novelty: 9...Bf5!?N Simply developing the bishop to an active position. (9...Bg4 10.e3 Bxf3 11.gxf3 Nc6 12.Bg2 Qd5 13.Bh4 e5 was seen in Hochstrasser – Sutovsky, Zurich 2009, when 14.c4!N would have been critical. Avrukh continues with 14...Qd6 15.dxe5 Qxe5 16.Qd5 Qb2 17.0-0 Qxa2 18.Qxc5 a5 and concludes that the apawn gives Black good counterplay. However, in my opinion, after 19.Rb1² White keeps the better chances. It’s not so easy for Black to promote his a-pawn, while the b7-pawn is weak and the g2-bishop will begin to work unpleasantly after f3-f4.) 10.e3 Nd7 Black has a perfectly playable position. A 553

sample continuation is 11.Bd3 Bxd3 12.Qxd3 Re8 13.0-0 h6 14.Bf4 e5÷ when White will have to break up his pawn centre and it is obvious that he obtained no advantage from the opening. 9...Be6 Black develops the bishop to an active position while targeting the a2-pawn, which explains White’s next move. 10.Qd2 This seems like the most reliable choice. 10.Qa4 may lead to some difficulties. 10...Nc6 11.Be2?! occurred in Banikas – Eljanov, Plovdiv 2003, and here I agree with Avrukh’s recommendation and analysis: 11...Qd5!N 12.c4 Qe4 13.dxc5 (13.d5?! is bad because of 13...Bxd5µ when it is clear that the queen is misplaced on a4) 13...h6 14.Bf4 g5 15.Bg3 g4 16.Nh4 Rad8 17.0-0 Rd2ƒ Black has a strong initiative. 10...Nc6 11.Be2 Rc8 12.0-0 h6 13.Bf4 cxd4 14.cxd4 Qa5! We have reached a typical scenario for the Grünfeld, resembling variation B1 of Chapter 5. White has an extra central pawn, while Black has a queenside majority which may eventually yield an outside passed pawn. Assuming that the level of piece activity is approximately balanced, such positions tend to be quite comfortable for Black, as is the case here. We will follow an instructive game for a few more moves. 15.Bc4?! This method of trading pieces leads to problems, as Black gets the opportunity for a quick ...e5. 15.a3N was better, although 15...Qxd2 16.Nxd2 Bd5 17.Rc5 e6 18.Rfc1 g5 19.Bg3 f5 20.Be5 Bxe5 21.dxe5 Rcd8ƒ still gives Black the easier play in the endgame. 15...Bxc4 16.Rxc4 Qxd2 17.Nxd2 e5 18.dxe5 Nxe5 With the long diagonal cleared for Black’s bishop, the plan of creating a distant passed pawn on the queenside becomes more dangerous. 19.Ra4 White hopes that his active rook will restrict Black’s queenside ambitions. 19.Bxe5 does not equalize: 19...Bxe5 20.Rfc1 Rxc4 21.Rxc4 Rd8 22.Rc2 Rd7³ Black has the stronger minor piece and an easy plan of creating a passed pawn. 19...a6

554

20.Rb1?! Attacking the b7-pawn is nothing more than a strike into thin air, since White will not be able to capture it anyway. 20.Bxe5 Bxe5 21.g3 was better, although 21...Rc6 22.Rc4 Rfc8 23.Rfc1 Rxc4 24.Rxc4 Rxc4 25.Nxc4 Bc7³ still leaves White facing a difficult defence. 20...Nd3! 21.Bg3? Of course 21.Rxb7?? allows 21...Nc5–+ with a deadly fork. 21.h3 is a lesser evil, although 21...Rfe8 22.Nf3 Re6 23.Bg3 b5 24.Ra3 Nc1 still leaves White under considerable pressure. We have been following Babu – Gupta, Hyderabad 2005. Black could have exploited his opponent’s latest error with:

555

21...Nc1!N 22.Kh1 22.Rxb7? Rfd8 wins easily. 22...Rfd8 23.Nf3 Rd1† 24.Ng1 Rf1 25.h3 Rd8 26.Bh2 Rdd1 27.g4 Bf6µ Black is dominating and White will surely suffer a loss of material before long. 8.Nxd5 Qxd5 9.Rxc5 Qxa2

10.e3 White sacrifices a pawn to fight for the initiative. Other moves are no better: 10.Qd2 h6 11.Bh4 was the choice of an expert in the 6.Bg5 variation, in Ulko – Yandemirov, Moscow 1998. In the game Black went for interesting complications by advancing with ...g5-g4, but the simple 11...0-0N seems like an improvement. For instance: 12.e3

556

12...e5! 13.Nxe5 Nxe5 14.dxe5 Be6 15.Bd3 b6 16.Rc2 Rfc8 17.f4 Qa1† 18.Qd1 Qxd1† 19.Kxd1 Bb3µ Black wins an exchange. 10.Qc1 0-0 11.e3 was seen in Nikolaidis – Livshits, Moscow 2003, and some other games. On each occasion, Black overlooked a strong resource:

11...Na5!N 12.Nd2 e5! 13.dxe5 Be6 14.Be2 b6 15.Rc2 Bxe5µ Black’s pieces are clearly better placed. 10...Qxb2 11.Bc4 This active developing move is natural enough, although two other options have been tested at GM level:

557

11.Bd3 h6 12.Bh4 e5! is strong, since capturing on e5 would allow a queen fork on b4. 13.0-0 exd4 14.Qa4 Qb4 15.Qxb4 Nxb4 16.Bb5† Kf8 17.Nxd4 Bxd4 18.exd4 Kg7 19.Re1 b6 20.Rc4 was played in Van Wely – Fier, Germany 2015, when 20...a5Nµ would have left White with insufficient compensation for the missing pawn, especially since Black has connected passers on the queenside. 11.Be2 h6 12.Bh4 e5 13.0-0 exd4 14.exd4 occurred in Grandelius – Salem, Istanbul (ol) 2012. Despite Black’s uncastled king, it was quite safe for him to gobble a second pawn:

14...Nxd4!N 15.Nxd4 Qxd4 16.Qxd4 (16.Bb5† Kf8 gets nowhere for White) 16...Bxd4 17.Rd1 g5 18.Rxd4 gxh4³ White’s active pieces do not provide full compensation for the queenside passers, and he faces a long struggle for a draw. 11...Qb4† 12.Nd2 0-0 13.0-0 We have been following Mamedyarov – Vachier-Lagrave, Doha (blitz) 2016. The following move has not yet been tested in practice:

558

13...h6!N I considered 13...e5N but 14.Ne4! leads to complications: White is not only eyeing the f6-square, but also planning Qa1 (or Qc1) with threats against the black queen. 14.Rb5 14.Bf4?! e5 15.Rb5 Qe7 puts White in trouble. 14.Bh4 e5 15.Rb5 (15.Ne4?! achieves nothing here due to 15...Bf5) 15...Qa3 also favours Black. The text move seems best. My analysis continues: 14...Qd6 15.Bf4 Qd7 16.Qc2 Kh7 17.Nf3 b6 18.Rh5!? This move seems trickiest, but Black has enough resources to deal with the attack. 18...e6 19.Ng5† Kg8 20.Nxe6 fxe6 21.Qxg6 Ne7 22.Qg4 Rxf4! 23.Qxf4 Qe8 24.Qe4 Qc6 25.Qxc6 Nxc6³ We reach an endgame where material is equal, but Black is clearly the one who is trying to win. B) 5.g3

559

White intends to fianchetto his bishop and exert pressure on the long diagonal. This move was first tested by the American player Jacob Bernstein, against Alekhine, at Carlsbad 1923. It was subsequently used by Capablanca, Tolush, Bogoljubov and Taimanov; and in the modern era, Aronian, Jobava, Giri and Mamedyarov have all tried it. White’s set-up is not connected with a dense body of opening theory, so it may appeal to players who want to obtain a fresh position early in the game. On the other hand, when we compare it to the more popular variant arising after 3.g3 d5 (rather than 3...c6, as recommended in Chapter 2) 4.cxd5 Nxd5, the early Nc3 has made Black’s life easier. 5...Bg7 6.Bg2 6.e4?! would be inconsistent. 6...Nxc3 7.bxc3 c5 8.Be3 Nc6 9.e5 Qa5 10.Qd2 Bf5 11.Bg2 was seen in Pontoppidan – Juntunen, corr. 1997, when Black could have obtained a nasty initiative with: 11...cxd4N 12.cxd4 Nb4 13.Kf1 Rc8 14.Nf3 0-0ƒ White clearly has problems, due to his king position and the difficulty of developing the h1-rook. 6...Nxc3 7.bxc3 c5 Black challenges the enemy centre in the usual fashion. White now faces a choice between B1) 8.e3 and B2) 8.Nf3. As we will see, each option has certain pros and cons. 8.Bb2?! commits the bishop for no reason, and 8...Qa5 9.Nf3 cxd4 10.Nxd4 0-0 11.0-0 e5 12.Nb3 Qc7 13.Qc2 Bf5 14.e4 Be6³ gave Black the upper hand due to his superior pawn structure in Altanoch – Downham, corr. 2001. B1) 8.e3 This move is usually associated with the development of the knight on e2, from where it defends the

560

c3-pawn and may later go to f4. 8...Qa5 9.Bd2 This is the most popular sequel, enabling White to recapture with cxd4 in the event that Black exchanges. In some scenarios, White may attack the queen with a timely c3-c4. 9.Ne2!? This move has also been tested at the highest level. White hopes that his piece activity will outweigh his pawn weaknesses after the following exchange. 9...cxd4 10.exd4 0-0 11.0-0 Nc6 12.Qb3 Rd8 13.d5 Ne5 14.Bg5 Another high-level game continued 14.Nd4 Bd7 15.Bg5 when, in Mamedyarov – Grischuk, Almaty 2008, Black could have played 15...Bf6!?N 16.Bxf6 exf6 when White’s d- and c-pawns may become weak. It is worth adding that 17.f4?! only helps Black after 17...Ng4 18.Rfe1 Rac8 19.Rac1 f5³ when White has an airy kingside in addition to his structural weaknesses. 14...Re8 15.Nd4 Bd7 16.Rab1 b6 17.Bf4 Rac8 18.Bxe5 Bxe5 19.Nc6 Qxc3 20.Qxc3 Bxc3 21.Nxa7 Ra8 22.Nc6 Rxa2 23.Rxb6 After a logical sequence, I can recommend a modest improvement for Black:

23...Bf6N÷ Keeping as much tension as possible in the position. The position should certainly be a draw with best play, but Black can try to pose problems using his bishop pair and active rook on the 2nd rank. (Instead, after the simplifying 23...Rb2 the players agreed a draw in K. Georgiev – Kramnik, Belgrade 1997.) 9...0-0 10.Ne2 White places the knight in such a way as to avoid blocking the g2-bishop. Later, the knight may go to f4 and perhaps d5. It is also worth noting that ...Qa6 is one of Black’s recurring ideas in this line, so a 561

further point of the knight on e2 is that the queen on a6 will not prevent castling. 10.Qb3 Qa6 11.Rb1 fails to cause any problems. In Inkin – Rozhnev, Gornozavodsk 2010, Black should have replied with:

11...Nd7! Black can follow up with any of ...Nb6, ...e5, or simply ...Rb8 followed by ...b6. Note that 12.Bxb7?! Bxb7 13.Qxb7 Qxa2 14.Qb3 Qa6µ only adds to White’s problems: he is weak on the light squares and unable to castle, and the passed a-pawn has great potential. 10.a4 has been played, although such a move hardly seems like it should be White’s top priority. 10...Nc6 11.Ne2 Rd8 12.0-0 Be6 13.Rb1 was seen in Issakainen – Sawatzki, corr. 1995, when Black could have thematically taken control of the a6-f1 diagonal:

562

13...Qa6!N 14.Re1 Bc4µ White has no counterplay and his centre is about to come under heavy fire after ...e5. White occasionally tries 10.Nf3 but, as you might guess, 10...Qa6! is an excellent reply. 11.Qe2 Qxe2† 12.Kxe2 In Heeke – Loewe, Ditzingen 2008, Black could have optimally developed his pieces with:

12...Nc6N 13.Rhd1 Be6 14.Kf1 Rac8³ Black has the more harmonious position and can choose between two tempting plans: a central strike with ...e5, and queenside play with ...Na5, ...cxd4 and ...Nc4. Finally, 10.Nh3 has been played by Gunina a couple of times. For instance, 10...Qa6 11.Qe2 Qxe2† 12.Kxe2 occurred in Gunina – Sutovsky, Gibraltar 2016. Black could have caused his opponent some problems with:

563

12...Nc6!?N Intending fast central play. For instance: 13.Rab1 e5! 14.dxc5 Be6 15.Rxb7 Bc4† 16.Ke1 Na5 17.Rb4 Rac8 18.c6 Bd3³ Black will pick up the c6-pawn and enjoy more than enough compensation for White’s remaining extra pawn: the d3-bishop cuts through the light squares, and the d2-bishop and h3-knight are passive. 10...Qa6! Again we see this multifunctional move. The queen eyes the light squares and defends the b7-pawn in order to release the bishop from c8. 11.0-0 Bg4 12.Re1 It is understandable that White avoided 12.f3?!N 12...Be6 13.Re1 Nc6 14.f4 Rfd8µ when her centre is under pressure and her light squares are weak. In Zatonskih – Almasi, Bastia 2013, Black’s most accurate continuation would have been:

564

12...Nc6!N Instead, the game continued with 12...Rd8, when 13.d5!N would have prevented the knight from going to c6, as well as to d7, which would cause the light-squared bishop to be trapped. Therefore Black would have to settle for something like 13...Bc8!? 14.e4 Nd7 15.Be3 b6 16.a4 Ne5÷ with a double-edged position. White can get rid of the a-pawn, but on the other hand the black knight can go to c4. 13.h3 Obviously 13.d5? would be a positional blunder here due to 13...Ne5 when the knight can go to any of f3, d3 or c4 with powerful effect. 13...Bd7 14.Nf4 Rac8 15.Bf1 c4 16.Bg2 e5ƒ Black has the more active game and the bishop on d2 is less than impressive. B2) 8.Nf3 Nc6

565

White has two main ways of reinforcing his centre: B21) 9.e3 and B22) 9.Be3. 9.Bb2?! This move places the bishop on a passive square. Moreover, Black can cause concrete problems by attacking it immediately. 9...Qb6! 10.Qb3 10.Rb1?! Bf5 11.dxc5 Qxc5 12.Rc1 Qb6µ gives Black the better pawn structure as well as a lead in development. 10.Qd2 0-0 11.0-0 Rd8 12.Rfd1 was played in Wichmann – Grosse Honebrink, Dresden 1993, when the time was ripe for action in the centre: 12...e5! 13.d5 e4 14.Ng5 Na5 15.Qc2 f5 16.Rab1 Nc4 17.Bc1 Qf6µ Black has an obvious advantage in piece activity and the c3-pawn is weak. 10...Be6 Developing with tempo. 11.Qa3N This is White’s best defensive try. 11.Qxb6?! axb6 gives Black an open a-file to add to his other trumps. 12.a3 0-0 13.0-0 Bd5 14.Rfd1 Rfd8 15.e3 was seen in Kupchik – Torre Repetto, New York 1925, when 15...e5!Nµ would have maximized Black’s advantage. 11...cxd4 12.Nxd4 12.cxd4 0-0 13.0-0 Nb4ƒ gives Black a lot of activity. 12...Bc4 13.Nxc6 bxc6 14.Qa4 Bb5 Black keeps a certain initiative and White will have to play accurately to equalize. White can also sacrifice the d4-pawn, hoping to gain some initiative. 9.0-0 cxd4 10.Nxd4 Nxd4 11.cxd4 Qxd4 12.Rb1 0-0 This position has been tested many times at grandmaster level. White can regain the pawn easily 566

enough, but he has little chance of an advantage, as the following examples demonstrate:

13.Be3 13.Qxd4 Bxd4 14.Bh6 Rd8 15.Bxb7 gives White nothing: 15...Rb8 16.Bxc8 Rdxc8 17.Rxb8 Rxb8 18.Rd1 Bc5 19.Rd7 Rc8 20.a4 a6 21.Kg2 f6 resulted in an absolutely equal endgame in Kamsky – Bacrot, Jermuk 2009. 13...Qc4 14.Qd2 14.Bd5 does not change much: 14...Qa6 15.Qb3 Qxe2 16.Bc5 Be6 17.Bxe6 Qxe6 18.Qxe6 fxe6 19.Rxb7 Rf5 20.Rc7 Rd5 21.Rb1 Be5 22.Rc6 a5 23.Rb7 Rb8 24.Rxb8† Bxb8 25.Bxe7 Kf7 26.Bc5 Rd1† 27.Kg2 Rc1 28.Rb6 Rxc5 29.Rxb8 reached a drawn rook endgame in Nakhbayeva – Lei Tingjie, Chengdu 2015. 14...Be6 15.Rfc1 15.Rxb7 also leads to simplifications. 15...Rad8 16.Qa5 Qxe2 17.Rxa7 was seen in Krivonosovs – Sykora, Spain 2004, when 17...h5N 18.Rxe7 Qxa2 19.Qxa2 Bxa2= would have been dead equal. 15...Rad8 16.Qxd8 Rxd8 17.Rxc4 Bxc4 18.Bxa7 We have been following Shankland – So, Saint Louis 2018. The simplest continuation seems to be:

567

18...Bxa2N 19.Rxb7 Rd1† 20.Bf1 Bf6= With easy equality. B21) 9.e3

This move secures the d4-pawn but is rather slow, and the c1-bishop can hardly be happy. 9...0-0 10.0-0 Qa5 11.Qb3 Black also obtains an excellent position after other moves: 11.Bd2 Qa6 12.Re1 occurred in Durda – Svana, Uzhgorod 2007, and now 12...Bg4!N would be my 568

choice (the game continuation of 12...Be6 was at least equal for Black, but it looks beneficial to provoke a slight loosening of White’s kingside): 13.h3 (13.Bf1 Qa3! is promising for Black, an important tactical point being that 14.h3 runs into 14...Bxf3 15.Qxf3 cxd4 when Black wins a pawn thanks to the pin along the 3rd rank) 13...Be6 14.Bf1 Bc4 15.Bxc4 Qxc4÷ White has the more difficult position to play, especially with ...e5 coming up. 11.Bb2 has been played by some strong GMs but Black has little to fear: 11...Rd8 12.Nd2 (Black seized the initiative after 12.Qe2 e5 13.Nxe5?! Nxe5 14.dxe5 Bxe5 15.c4 Rd2 16.Qxd2 Qxd2 17.Bxe5 in Kotsur – Bologan, Moscow 2004, at which point 17...Bf5!N 18.Bxb7 Re8 19.Bf4 Qe2 20.Bd5 Be4 would have given him every chance of earning the full point) 12...Be6 13.Ne4 cxd4 14.cxd4 In Kachur – Yermolinsky, Aktyubinsk 1985, it would have been good for Black to complete development with:

14...Rac8N My illustrative line continues: 15.Re1 (this is not forced, but it makes a certain amount of sense to pre-empt ...Bc4) 15...Bd5 16.Qb1 Qc7 17.Rc1 b6 18.Ng5 Bxg2 19.Kxg2 Qd6³ Black has active pieces along with the characteristically favourable pawn structure; his plans involve ...e5 and the gradual exploitation of his queenside majority. 11...b6 12.Nd2 The greedy 12.Ne5?! backfires after 12...Nxe5 13.Bxa8 Be6 14.Qc2 cxd4 15.exd4 Rxa8 16.dxe5 Qxe5© when Black has tremendous compensation for the exchange, with an extra pawn, a better structure and light-square holes on White’s kingside. 12.Bd2 looks rather passive, and after 12...Qa6 13.Rfc1 Be6 14.Bf1 Qb7 15.Qa4 Na5 16.Ne1 Black clearly had a comfortable position in Maslanka – Jasnikowski, Gdynia 1987. Among several promising continuations, my favourite is:

569

16...Bd7!?N 17.Qa3 Bf5µ Black is clearly better, since 18.dxc5 is well met by 18...Rad8 19.cxb6 axb6 20.e4 Qxe4 21.Bf4 Nc4 when Black remains in control. 12...Bd7 13.Bb2 Rac8 14.Rfd1 Rfd8 Black has a fine position which soon became a serious advantage in the following contest of grandmasters: 15.Bf1 cxd4 16.Nc4 Qh5 17.cxd4 Be6 18.Qa3 Bxc4 19.Bxc4 Ne5 20.Bd3 In Arutinian – Kacheishvili, Tbilisi 2007, Black missed a great opportunity:

20...Nf3†!N 21.Kg2 Ng5 570

With decisive threats on the kingside. B22) 9.Be3

This seems like the best way to strengthen the d4-point, as White keeps his bishop reasonably active. On the other hand, the bishop may get hit by a ...Na5-c4 manoeuvre. 9...0-0 10.0-0 cxd4 Black has tried several other moves but the immediate central exchange seems reliable enough. 11.Nxd4 This is the most challenging continuation: White activates his knight and clears the long diagonal for his bishop, at the expense of his pawn structure. 11.cxd4 This recapture is less popular and objectively rather harmless. 11...Be6 Black intends ...Bd5 to neutralize the bishop on g2, followed by queenside play involving ...Na5 and ...Rc8. 12.Qa4 White’s queen looks active but Black may be able to kick it away with ...a6 and ...b5 in the future. Another logical sequence is 12.Qd2 Rc8 13.Rfd1 as played in Rejfir – Florian, Prague 1955, when 13...Qa5!N 14.Qxa5 (no better is 14.Qb2 Bd5³) 14...Nxa5ƒ would have given Black some initiative due to his pressure on the queenside; his knight can jump to c4, attacking the e3-bishop. 12.Rb1 b6 also offers Black a great game. For instance, 13.Qa4 Rc8 14.Rfd1 Na5 15.Ne5 Bd5 16.Rbc1 Bxg2 17.Rxc8 Qxc8 18.Kxg2 was played in Antic – Erdogdu, Petrovac 2015, when 571

18...Bxe5!N 19.dxe5 Nc4 20.Qxa7 Qc6† 21.Kg1 Nxe3 22.fxe3 Qc2 23.Rf1 Qxe2 24.Qxe7 Qxa2³ would have given Black some winning chances: he has a passed pawn on the queenside, while White’s structure is broken and his king is somewhat exposed. 12...Bd5 13.Rab1 Na5 14.Bf4 14.Ne5 Bxg2 15.Kxg2 Qd5† 16.Kg1 b6ƒ gives Black a harmonious position while White is weak in the centre. The text move was played in Muelas Cerezuela – Celaya Tapiz, Benidorm 2003. Black advanced his b-pawn one square and went on to win, but he could have played more ambitiously with:

14...a6!N 15.Rfc1 b5 16.Qa3 Ra7!³ The last move not only eliminates the threat of Bc7, but also prepares ...Nc4 and ...Qa8 with an excellent position. 11...Na5 Naturally, Black avoids the knight trade and aims for the c4-outpost. We have reached a critical position where White has tried several different ideas. 12.Rb1 Occupying the open file is certainly a logical choice. 12.Qc1 Nc4 13.Rd1 is hardly scary. 13...Nxe3 14.Qxe3 Qc7 gave Black a pleasant game in Elski – Krause, corr. 2014, although White went on to equalize by advancing his c-pawn and forcing simplifications. 12.Qd3 is a sensible move, controlling the important c4-square. 12...a6 13.c4 (13.Rac1 Qc7 14.c4 is another possible move order which has also been played) 13...Qc7 14.Rac1 Rd8 15.Rfd1 We have been following Bai Jinshi – Zhao Jun, Hangzhou 2018. Black’s simplest equalizing tool is: 15...Bxd4N 572

16.Bxd4 Nc6 17.Qe3 Nxd4 18.Rxd4 Rxd4 19.Qxd4 Bd7 The position is clearly equal. A sample continuation is 20.Rb1 Be6 21.Rxb7 Rd8 22.Rxc7 Rxd4 23.Bd5 Bxd5 24.cxd5 Rxd5 25.Rxe7 Ra5 with a drawn rook endgame. 12.Nb5 has been played in several games, presumably with the idea of attacking the a7-pawn, but after 12...Bd7! it turns out that the knight jump was useless. 13.Rb1 (White daren’t play 13.Nxa7? due to 13...Nc4 14.Bd4 e5 15.Bc5 Qc7 16.Bxf8 Bxf8 17.Bd5 Nb6 18.Bb3 Rxa7–+ when Black’s minor pieces were much too strong for the rook in Siebrecht – Cossin, Calvi 2012) Now in Dragomarezkij – Eliseev, Moscow 2013, Black could have fought for the advantage with:

13...a6!N 14.Na3 Bf5 15.Bb6 Qxd1 16.Rbxd1 Bxc3 17.Nb1 Nc4 18.Nxc3 Nxb6 19.Bxb7 Ra7 20.Bf3 Rc8 21.Nd5 Nxd5 22.Bxd5 Rc2ƒ White is under some pressure due to the active rook on the 2nd rank. A tricky continuation is: 12.Nb3!? Qc7 12...Nc4!?N 13.Bd4 is an untested pathway which also looks about equal. 13.Nxa5 A logical alternative is 13.Qd5!? Nxb3 14.Qxb3 when, in Krivonosov – Vaitonis, Plateliai 2017, Black could have solved all his problems with a temporary pawn sacrifice: 14...Be6N 15.Qxb7 Qxb7 16.Bxb7 Rab8 17.Ba6 Bxc3 18.Rac1 Be5 19.a4 Rb2 20.Bb5 a6 21.Bxa6 Bd7 22.f4 Bd6 23.Bd4 Rb4 24.Rc4 Rxa4 25.Rxa4 Bxa4= With total equality. 13...Qxa5 14.Bd4?! This seemingly natural move leads to difficulties. White should prefer 14.Qb3!? Rb8 15.Rfd1 Bg4 16.Bf3 Bxf3 17.exf3 as in Krivonosov – Kulaots, Riga 2014, when 17...Qxc3N 18.Bxa7 Qxb3 19.axb3 Bxa1 20.Bxb8 Rxb8 21.Rxa1 Rc8= would have been the easiest route to an equal endgame. 14...Rd8 15.Qb3 In Roobol – Szekely, Haarlem 1999, Black could have caused some problems with: 573

15...Rb8!N 16.Rfd1 Be6 17.Qb2 Qa4³ White needs to play quite accurately so as not to worsen his position. For example: 18.Bxg7 Kxg7 19.Rxd8 Rxd8 20.Bxb7 20.Qxb7? is impossible in view of 20...Rd1† 21.Rxd1 Qxd1† 22.Bf1 Bh3 and Black wins. 20...Rb8µ The pin along the b-file is unpleasant. 12.Qa4 Finally, White may actively develop his queen while connecting the rooks. 12...Bd7 13.Qb4 Rc8 14.Rfd1 14.Rad1?! seems worse. 14...Rc4 15.Qb1 Qe8 16.Bxb7 e5 17.Nf3 Bf5 18.Qb2 was played in Ftacnik – Bacrot, Germany 2010. Out of many possible moves, I prefer 18...Qb8!?N 19.Qb5 Qxb7 20.Qxa5 h6 21.Qxa7 Qxa7 22.Bxa7 Ra8 23.Be3 Rxc3ƒ when Black will quickly pick up the a2-pawn to restore material equality. He will then enjoy a risk-free advantage due to his bishop pair and overall activity level, although White should be able to hold with careful defence. 14...Rc4 15.Qb2 15.Qa3?! allows Black to attack the queen with good effect: 15...Ra4 16.Qc1 (even worse is 16.Qc5?! b6 17.Qd5 Qe8 18.f4 e6 19.Qf3 as played in Moran – G. O’Connell, Dublin 2011, when 19...Qe7Nµ would have left White with no harmony among his pieces, as well as a poor structure) 16...Qc7 17.Nb3 b6 Black enjoyed an obvious advantage in Nezamyatin – Kuryshkin, Sochi 2017, due to White’s queenside weaknesses. The text move is more logical. In Spraggett – Yermolinsky, Las Vegas 1999, Black had many possible moves, but my suggestion would be:

574

15...a6!?N 16.Rac1 Ra4 I think White should take the opportunity to simplify with: 17.Nf5! gxf5 18.Qd2 Nc4 19.Qxd7 Qxd7 20.Rxd7 Nxe3 21.fxe3 Rxa2 22.Bxb7 Rxe2 23.Rxe7 a5= With a roughly equal position. 12...Bg4!? I think it is worth provoking the following pawn move. 13.h3 White can also ignore the bishop with 13.Qc2!?N intending Qe4. My analysis continues 13...a6 14.Rfd1 (14.Qe4 Bc8 15.Rfd1 Qc7 16.Qh4 Bf6 17.Qh6 Bg7 18.Qh4 Bf6= is a possible route to a repetition) 14...Qc7 15.Qe4 Bc8 16.Qh4 Rb8 17.Nf3 b5 18.Bh6 Bb7÷ with an unclear position, where White’s kingside attack does not seem too dangerous and the c3-pawn remains a target. 13...Bd7 14.Nb3 Nc4 15.Bd4 We have been following Nogueiras Santiago – Bruzon Batista, Santa Clara 2000. Now Black’s play can be improved by means of:

575

15...Rc8!N An interesting idea: Black sacrifices a pawn in order to mobilize his pieces as quickly as possible. 16.Nc5 16.Bxb7 Rc7 17.Bg2 Qc8 18.Bxg7 Kxg7 19.Kh2 e5© offers Black excellent compensation for the minimal material investment: his pieces are located perfectly and it is not easy for White to defend his weakness on c3. 16...Bc6 17.Nxb7 Qc7 18.Bxc6 Qxc6 19.Qd3 Nb6 20.Na5 Qe6 21.Bxg7 Kxg7 22.Qd4† f6 23.c4 Rfd8 24.Qg4 f5 25.Qh4 Rd2÷ Black is poised to regain the sacrificed pawn, and White must be careful not to become worse. C) 5.Na4!?

576

This variation is rather young, having been introduced in 1996 by its inventor, the Armenian IM Nadanian. Many players followed suit, due to the unexplored nature of the emerging positions, which offered a lot of space for creativity. Even today there are plenty of uncharted paths in this variation: as you will see below, I found novelties for Black as early as move 6 in some lines. Vachier-Lagrave, Riazantsev, Lysyj, Turov and Bu Xiangzhi have all played the Nadanian Variation. The main idea of course is to play e2-e4 while depriving Black of the opportunity to exchange his knight on c3. However, Black has more than one way to reach a good position against it. 5...e5!? Since White spent a tempo moving his knight away from the centre, Black is perfectly justified in offering a pawn sacrifice. Avrukh, Kaufman, Delchev and Agrest all recommend the most popular 5...Bg7, which is also quite playable. A tempting alternative is: 5...Nf6!? Preventing e2-e4 and leaving White with a knight on the edge of the board. 6.Nf3 6.Nc3 can be met by 6...Bg7 (if Black wants a draw, then 6...Nd5 repeats the position) 7.e4 0-0 8.Nf3 c5 9.d5 e6 10.Be2 exd5 11.exd5 Bf5 12.0-0 Ne4 13.Qb3 Qb6 14.Na4 Qxb3 15.axb3 Nd7 16.Nh4 Rfe8 17.Nxf5 gxf5 18.Bb5 Rad8÷ when Black was not worse in Narciso Dublan – Ivanchuk, Barcelona 2005; White has weak pawns on b2 and d5, while the knight on a4 looks awkward. 6...Bg7 7.Bf4 The most natural and popular continuation, but I also looked at other possibilities. 7.g3 does not create any problems for Black. 7...Nc6 8.Nc3 was seen in Atia – Holmsten, Elista (ol) 1998, and this would have been a good time to strike at the centre: 8...e5!N 9.dxe5 Qxd1† 10.Nxd1 Ng4 11.Bg2 Ngxe5= Black does not have a shadow of a problem. 577

There are also no difficulties after the calm 7.e3 0-0 8.Be2 b6 9.0-0 Bb7 10.Qb3 when, in Gmuer – Ham, corr. 2001, I would have favoured the transfer of the knight to d6 in preparation for ...c5: 10...Ne4!?N 11.Rd1 Nd7 12.Nc3 Nd6 13.Bd2 c5= Black has a comfortable position. 7...0-0 8.e3 Nbd7 9.Rc1 9.Be2 allows 9...c5! when White soon became worse after 10.Nxc5 Nxc5 11.dxc5 Qa5† 12.Qd2 Qxd2† 13.Nxd2 Nd5 14.Bg5 h6 15.Bh4 Bxb2 16.Rb1 Bc3µ in Loyfenfeld – Mende, Germany 2011. White’s king is stuck in the centre and the c5-pawn is weak. 9...c6 10.Nc3 Trying to preserve the bishop with 10.h3 allows 10...Qa5† 11.Nc3 Nd5 12.Bg3 Nxc3 13.bxc3 Qxa2³ banking a pawn. Here I can recommend an improvement on Lou Yiping – Gao Rui, China 2013.

10...Nh5!N 11.Bg5 h6 12.Bh4 g5 13.Bg3 13.Nd2 is well met by 13...e5!, for instance: 14.Qxh5 exd4 15.Bxg5 hxg5 16.exd4 g4 17.Be2 Nf6 18.Qc5 Re8 19.0-0 Nd5 20.Rfe1 Nf4 21.Bc4 Rxe1† 22.Rxe1 Bxd4 23.Qe7 Be6 24.Qxd8† Rxd8 25.Bxe6 Nxe6³ Material is equal and Black has the better chances in the endgame thanks to his active pieces. 13...Nxg3 14.hxg3 c5 15.Bd3 g4 16.Nh4 cxd4 17.exd4 Nf6³ Black may not be able to keep the advantage of two bishops due to the fact that Nf5 is likely to be played soon; nevertheless, he still has good prospects playing against the isolated d-pawn.

578

We will analyse C1) 6.e4!? and the main line of C2) 6.dxe5, after first checking a couple of unimpressive alternatives: 6.g3?! was played in Rodrigues da Silva – Vasconcelos, Joao Pessoa 2010, when 6...Nb6!N would have been strong, for instance: 7.Nxb6 axb6 8.dxe5 Qxd1† 9.Kxd1 Nc6 10.Bh3 (10.f4? turns out even worse after 10...Be6µ) 10...Bxh3 11.Nxh3 Nxe5 12.Kc2 Nc4³ Black has an easy game, with obvious pressure on the queenside after the bishop comes to g7. 6.a3 White spent a tempo defending against ...Bb4† in Lazarev – Miron, Avoine 2013. Once again, Black could have caused problems with the thematic knight move: 6...Nb6!N 7.Nxb6 axb6 8.dxe5 White may as well take the pawn. 8.d5 Bf5 9.f3 e4 10.Qd4 Rg8 11.Qe5† Qe7 12.Qxe7† Bxe7 13.g4 Bh4† 14.Kd2 Bd7 15.fxe4 Na6 16.Kc2 Bxg4³ leaves Black with the better pawn structure and superior development of the pieces. 8...Qxd1† 9.Kxd1 Be6 Black has more than enough compensation, as the following lines demonstrate. 10.f4 10.Nf3 Bg7 11.Bf4 Nc6 12.e3 0-0-0† 13.Ke1 f6!³ presents White with a dilemma: if he exchanges on f6, Black’s dark-squared bishop becomes a monster; and if he ignores the pawn, Black will restore material equality while keeping some initiative thanks to his superior development. 10...Nd7 11.Be3 I also considered 11.Nf3 Nc5 12.Nd2 g5 13.e3 gxf4 14.exf4 0-0-0© when White has developmental problems and his extra pawn is not felt at all. 11...f6 12.exf6 Nxf6 13.Nf3 Bb3† 14.Ke1 Ng4 15.Bd4 Bb4† 16.Bc3 Bxc3† 17.bxc3 0-0© White has problems with his kingside development and his queenside pawns are weak, so Black has 579

tremendous play for a mere pawn. C1) 6.e4!? This rare move deserves attention, although Black has more than one good way of meeting it. 6...Nb6 This has been the usual choice. There are two good alternatives: 6...Bb4† 7.Bd2 Bxd2† 8.Qxd2 Nf6 9.Nf3 occurred in Xiong – Bai Jinshi, Internet (rapid) 2019, when 9...0-0!N 10.dxe5 Nxe4 11.Qe3 Qd5 12.Be2 Qa5† 13.Nc3 Nxc3 14.bxc3 Bg4 15.0-0 Nc6= would have given Black a comfortable position; the e5-pawn is weak. Black can also consider a novelty: 6...Nf6!?N 7.Nf3 exd4 8.e5 Ne4 9.Bd3 9.Qxd4?! Qxd4 10.Nxd4 Bb4† 11.Ke2 Bd7µ leaves White in a clearly worse position. 9...Bb4† 10.Kf1 Bf5

11.Qe2 11.g4? is suicidal in view of 11...Nd2†–+ when White will suffer from catastrophic pawn weaknesses. 11.Qb3 cannot be recommended: 11...Nc6 12.Bh6 Qd7 13.a3 Nc5 14.Nxc5 Bxc5 15.Bxf5 gxf5 16.Ng5 Na5 17.Qd3 Be7 18.Re1 Qd5 19.Qxf5 Rg8 20.Qxh7 0-0-0ƒ White’s extra pawn is hardly relevant, while his king is badly situated and the h1-rook cannot enter the game. 11...Qd7 12.Bh6 White should not allow Black to evacuate the king. 12.a3?! Be7 just improves the bishop’s position. 13.Bh6 Nd2† 14.Qxd2 Bxd3† 15.Qxd3 Qxa4 580

16.Nxd4 (16.e6?! is not dangerous here because of 16...f6 17.Nxd4 Na6 18.Rc1 0-0-0 19.Rc4 Qa5µ when White has a problem with his king and with the bringing the h1-rook into the game) 16...Na6 17.Qb5† Qxb5† 18.Nxb5 c6 19.Nc3 0-0-0³ Black benefits from better development. b) 12.e6!? is more interesting: 12...Qxe6 13.g4 Bxg4 14.Qxe4 Bh3† 15.Ke2 Nc6 16.a3 Be7 17.Qxe6 Bxe6 18.Re1 Bd5÷ We reach a difficult position to assess. Black has three pawns for a piece and his forces are well centralized, while White will have to work to establish the harmony of his pieces.

12...Nd2† 13.Nxd2 Bxd3 14.Qxd3 Qxa4 15.e6! An important resource that allows White to equalize the game. 15...Qa6 16.exf7† Kxf7 17.Nc4 Re8 18.Qxd4 Bf8 19.Qd5† Qe6 20.Qxe6† Rxe6 21.Bxf8 Kxf8= With a roughly equal ending. 7.Nxb6 This move was played by GM Lysyj, an expert in the Nadanian Variation. However, it liberates the a8-rook and does not contribute to the improvement of White’s position. 7.d5?! cannot be recommended. In Luehrig – Yamato, email 2000, Black could have caused problems with:

581

7...Bb4†!N 8.Nc3 c6 9.Qb3 Bxc3† 10.Qxc3 0-0 11.Bh6 Re8 12.d6 Qxd6³ Despite Black’s weak dark squares, White cannot derive significant benefits from them and his compensation for the pawn is insufficient. 7.Be3!? I think White should prefer this noncommittal developing move, which was played in Trujillo – Narings, Vlissingen 2005. For Black I recommend: 7...Bb4†!?N Another decent option is 7...exd4N 8.Qxd4 Qxd4 9.Bxd4 Rg8 10.Nc3 Nc6 11.Bf6 Be7 12.Bxe7 Kxe7 13.0-0-0 Be6= when Black does not have a shadow of a problem. 8.Nc3 exd4 9.Qxd4 Qxd4 10.Bxd4 0-0 11.0-0-0

582

11...Nc6 12.Bf6 Nd7 13.Bh4 Bxc3 14.bxc3 Nb6 15.Nf3 Be6 16.Rd2 Rae8 17.Bb5 Nc4 18.Re2 Nd6 19.Bxc6 bxc6 There arises a complex endgame, which is double-edged but objectively about equal. 7...axb6 8.d5 We have been following Lysyj – Sasikiran, Moscow 2007. I think Black should have played:

8...f5!N 9.Bd3 9.exf5?! runs into 9...Bb4† 10.Bd2 Qxd5! 11.Bxb4 Qe4†µ when Black regains the piece with a better position. 9...fxe4 10.Bxe4 Bb4† 11.Bd2 Qh4 12.Bc2 Bg4 13.Ne2 Bd6ƒ Black has good attacking possibilities. C2) 6.dxe5

583

Eating the pawn and inviting Black to prove his compensation is White’s most natural and popular continuation. 6...Nc6 Developing a piece and attacking the e5-pawn. White’s main options are C21) 7.f4, protecting his extra pawn, and C22) 7.a3, preventing a check on b4. 7.Nf3!? was played at a high level in Lputian – Shirov, Montecatini Terme 2000, when 7...Bb4†N 8.Bd2 Qe7 9.a3 Bxd2† 10.Qxd2 Nb6 11.Nxb6 axb6 12.Qf4 0-0 13.e3 Ra5 14.Be2 Nxe5= would have given Black a comfortable position with no problems. 7.Bd2 does not make much sense, and 7...Bf5 already threatens ...Ndb4. 8.a3 was seen in Swapnil – Gupta, Hyderabad 2013, and here it seems logical to recapture the pawn: 8...Nxe5N 9.Nf3 Qf6 10.Qb3 0-0-0 11.Nxe5 Qxe5³ Black is well ahead in development and is ready to attack the white king. 7.e4 leads only to problems for White: 7...Ndb4 8.Qxd8† (8.Bb5? caused White’s king to come under deadly fire after 8...Qxd1† 9.Kxd1 Bg4† 10.f3 0-0-0†–+ in Pozsonyi – Telek, Nyiregyhaza 2004) 8...Nxd8 9.Kd2 Here I recommend an improvement over Barkman – Kudrin, Parsippany 2006:

584

9...Ne6!N 10.a3 Nc6 11.Nf3 Bg7 12.Bb5 Bd7 13.Nc3 Nxe5 14.Bxd7† Nxd7³ White faces some problems due to the weakened light squares on the queenside. C21) 7.f4

With this principled move, White attempts to preserve his material advantage. 7...Bf5 Threatening ...Ndb4. I can also recommend an idea suggested by my friend, FIDE master Gleb Apryshko: 7...Bg4!?N 585

With the idea to prevent e2-e4. 8.Nf3 8.a3? is a clear mistake here due to 8...b5 9.Nc3 Nxc3 10.Qxd8† Rxd8 11.bxc3 Na5 12.Nf3 Nb3 13.Rb1 Nxc1 14.Rxc1 Bxa3µ when Black has the bishop pair and the better structure, plus a passed apawn which could become extremely powerful later. 8...Ndb4 9.Qxd8† 9.a3?! is unsuccessful, as 9...Qxd1† 10.Kxd1 0-0-0† 11.Bd2 Nd5 gives Black a promising initiative. My analysis continues: 12.Ke1 Bh6 13.e3 f6 14.exf6 Rhe8 15.f7 Re7 16.Kf2 Nxf4 17.exf4 Bxf3 18.Be3 Bd5 19.Rd1 Rxf7 20.Rxd5 Rxd5 21.Bc4 Rdd7 22.Bxf7 Rxf7ƒ The f4-pawn is under pressure and White still needs to work hard to achieve equality. 9...Rxd8

10.Kf2 Nc2 11.Rb1 Bf5 12.Nc3 Bc5† 13.e3 Na3 14.Ra1 14.bxa3? Bxb1 15.Nxb1 Rd1–+ wins a piece back, leaving Black the exchange up with a winning position. 14...Nc2= The game ends in a surprising draw by repetition. 8.a3 This position was reached in Grego – Grammatica, corr. 2011. I propose a novelty: 8...g5!N Putting pressure on White’s pawn chain. The open g-file may prove surprisingly useful, as we will see. Unlike the position from the previous note with the bishop on g4, 8...b5? would be a disastrous choice here due to 9.e4! Bxe4 10.Bxb5+–. 586

9.e3 gxf4 10.exf4 Nb6 11.Nxb6 White gains nothing by avoiding a queen trade: 11.Bd2?! Rg8 12.Nf3 Be7 White is in big trouble, for instance: 13.Nxb6 (13.g3 prevents one threat but allows 13...Nd4!, and after 14.Rc1 Nxf3† 15.Qxf3 Nxa4 16.Bb5† Bd7 17.Qxb7 Bxb5 18.Qxb5† Qd7–+ White has no compensation for the donated piece)

13...Bh4†! 14.g3 Rxg3!–+ White’s position is falling apart. Capturing the rook loses trivially, and 15.Nxh4 Qxh4 16.hxg3 Qxg3† 17.Ke2 Bg4# is another nice mate. 11...Qxd1† 12.Kxd1 axb6 White still has an extra pawn but he is far behind in development, and his king remains in danger despite the exchange of queens. 13.Be3 Nb4 14.Bd2! I believe this is White’s only move to avoid being worse. In the event of 14.Rc1 0-0-0† 15.Bd2 Nd3 16.Bxd3 Rxd3 17.Nf3 Rg8© Black’s two bishops and active rooks more than compensate for the minimal material investment. 14...Bc2† 15.Kc1 Be4 16.Nf3 Nc2 17.Ra2 h6 18.b3 But not 18.Be2?! Rg8 19.g3 Nd4 20.Nxd4 Bxh1³ when White does not have enough compensation for the exchange.

587

18...Nb4 19.Bxb4 Bxb4 20.Kb2 Bc5 21.b4 Be3 22.Bc4 Bxf4 23.Re1 Bxf3 24.gxf3 c6= The resulting endgame is equal. C22) 7.a3 White prevents ...Bb4†, while acknowledging that defending the e5-pawn would lead to certain inconveniences. 7...Nxe5 Black regains the pawn while allowing himself to be deprived of castling rights. This should not matter too much, as the queens will be exchanged and White’s pieces are not active enough to harm the king in the centre. 8.e4 Nb6 9.Qxd8† Kxd8 10.Bg5†?! This move looks so tempting and it has been the choice of some top players, but it only helps Black’s development. White should prefer: 10.Nxb6!? axb6 11.Bd2 As played in Nadanian – Timar, email 2002. Here I recommend:

588

11...Bc5N 12.b4 Bd4 13.0-0-0 Ke7 14.Bf4 Black can now choose the type of game to suit his playing style and risk appetite. 14...Ba1!? 14...Nc6 looks like a reliable route to an equal endgame: 15.Nf3 Rxa3 16.Rxd4 Rxf3 17.Bxc7 Nxd4 18.gxf3 Bg4! 19.fxg4 Rc8 20.Kd2 Rxc7 21.f4 Ne6 22.f5 Nd4 23.Bd3 Kf6 24.fxg6 fxg6 25.Ke3 Ke5 26.Rf1 Ne6= White cannot strengthen his position and the game should end in a draw. The text move leads to a more obscure game.

15.Kc2 Rxa3 16.Rxa1 Rxa1 17.Bxe5 Rxf1 18.Bxh8 Rxf2† 19.Kd3 Rf1 20.Ke2 Rc1 21.Bd4 Bd7 22.h4 h5 23.Rh2 Rc4 24.Nf3 Rxb4÷ Any result is possible, but Black has three pawns for the sacrificed piece and objectively he is fine.

589

10...Be7 11.0-0-0† 11.Nxb6N 11...axb6 12.0-0-0† Ke8 13.Bf4 f6 transposes to the note to White’s 12th move below. 11...Ke8 12.Bxe7 This is the consistent sequel to White’s 10th move. Nevertheless, the bishop exchange simplifies Black’s development and leaves him with no opening problems, to put it mildly. Another option is 12.Bf4, although to play this would be to admit that White’s 10th move was unsuccessful. In Bu Xiangzhi – Sasikiran, Internet 2005, Black could have fought for the advantage with:

590

12...f6!N 13.Nxb6 (13.Nc3 Kf7 14.Nf3 Nxf3 15.gxf3 c6³) 13...axb6 14.Be2 (14.Bg3 c6 does not really change anything) 14...Be6 15.Nf3 Nxf3 16.Bxf3 c6³ Black intends ...Kf7 and ...b5-b4, with pressure on the queenside. 12...Kxe7 13.Nc5 This was MVL’s choice. The French super-GM evidently did not wish to open the a-file for Black’s rook, after which it might cause some annoyance by going to a4, for instance. I also checked 13.Nxb6 axb6 14.Bd3 (14.h3 Be6 15.f4 Nc6 16.Nf3 Ra4³ shows how the open file can be used as a highway for the black rook) 14...Rd8 15.Bc2 Rxd1† 16.Kxd1 Be6³ when the endgame is slightly better for Black because of his active pieces. 13...Nbd7 14.Nxd7 In Vachier-Lagrave – Caruana, Shamkir 2015, Black could have fought for the advantage by choosing the right recapture:

14...Bxd7!N 14...Nxd7 gave Black a comfortable position in the game, but Black gains more from developing his bishop. 15.Be2 15.Nf3 Nxf3 16.gxf3 Rad8³ gives Black a risk-free advantage due to his better pawn structure. 15...Bc6 16.f3 Rad8 17.Nh3 Ba4 18.Rxd8 Rxd8 19.Nf2 c5³ Black controls the open file and White will have to work for a draw.

591

Conclusion This chapter has dealt with a number of somewhat rare systems which can arise after 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5. We started with 5.Nf3 Bg7 6.Bg5!?, which I refer to as the Ulko Variation after its main developer. After the further 6...c5 7.Rc1 I chose 7...Nc6! as my main recommendation, as I believe it leads to a slight advantage for Black with best play, although 7...Nxc3 is a worthy alternative which offers Black a reliable position without complications. We then considered White’s deferred kingside fianchetto with 5.g3. Though it is playable, Black is helped by the availability of the simple plan of ...Nxc3 followed by ...c5, with typical pressure against the enemy pawn centre. Black obtains a comfortable game and can often seize the initiative, especially if he can execute a well-timed ...e5. Finally we looked at the rare but tricky system of Nadanian, beginning with 5.Na4!?. I recommend meeting it with the active 5...e5!?, when the main line leads to a pleasant endgame for Black. For those who strive for a more complex game, I included some coverage of 5...Nf6!?, avoiding the early simplifications.

592

A) 6.Nf3 317 B) 6.e4 Nxc3 7.Bxc3 0-0 320 B1) 8.Bc4 322 B2) 8.Rc1 324 B3) 8.Nf3 326 B4) 8.Qd2 c5! 9.d5 Bxc3 10.bxc3 Qd6! 327 B41) 11.Nf3 329 B42) 11.f4 330

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.Bd2 According to my database, the Soviet player Solomon Kogan was the first to play this move against Yudovich in 1938 in Moscow. However, the move was not popular for a long time and remained in the shadows of theory until the late 80s, when it was developed by GM Viktor Moskalenko and started to attract more followers. Over the years, it has been played by such grandees of chess as Anand, Kramnik, Nakamura, Mamedyarov, Ivanchuk, Shirov, Tomashevsky, Fedoseev and others. The idea behind this move is to recapture with the bishop after ...Nxc3, to neutralize the g7-bishop’s pressure on the long diagonal. It also has some drawbacks though: White lingers with the development 593

of the kingside, and his king may remain in the centre for a long time. This gives us reason to play energetically, often sacrificing a pawn or more for the initiative. 5...Bg7 We will analyse two main options: A) 6.Nf3 and B) 6.e4. A harmless alternative is: 6.e3 0-0 7.Bc4?! Taimanov played this move, but it enables Black to take the initiative. 7.Nf3 is White’s best bet, leading to the innocuous 7.Bd2 line in the note to variation B22 of Chapter 4 on page 104. 7.Rc1?! fails to fight for the centre, and 7...e5 8.dxe5 Bxe5 9.Nf3 Nxc3 10.Bxc3 Qxd1† 11.Kxd1 Rd8† 12.Nd2 Nc6 13.Bb5 Bxc3 14.Rxc3 Bd7 15.Bxc6 Bxc6³ favoured Black in Dempsey – Collins, Greystones 2001. The bishop is stronger than the knight, and White’s king does not feel safe in the centre. 7...Nb6 8.Bb3 c5 9.dxc5 N6d7 10.Na4 Na6 11.Bc3 Ndxc5 12.Bxg7 Kxg7 13.Nxc5 Nxc5 14.Qxd8 Rxd8 15.Bc2 b6 16.Ne2 Bb7µ White’s opening strategy had clearly failed in Taimanov – Korchnoi, Budva 1967. Black is ahead in development, with more active pieces and possession of the open d-file. A) 6.Nf3

Gligoric, Petrosian and Najdorf have played this way with White. More recently, Fridman, Zubov and Epishin have followed the same path, along with several other strong GMs. White’s set-up looks modest but it deserves close attention.

594

6...c5 Castling has been more popular, but I prefer the idea of attacking White’s centre immediately and narrowing his possibilities. 7.Rc1 This is the best and most popular move. Consider the alternatives: 7.dxc5?! leads to an ugly pawn structure for White: 7...Nxc3 8.Bxc3 Bxc3† 9.bxc3 Qc7 (9...Qxd1† was played successfully by Lagno against Cmilyte in a blitz game in 2013, but the text move is stronger) 10.e3 In Bue – Jansen, Gausdal 2008, 10...Nd7N³ would have left White with no compensation for his inferior pawn structure. 7.Nxd5!? Qxd5 8.Rc1 is more interesting: 8...Qxa2 9.Rxc5

9...Na6! (I prefer this to 9...Nd7, which was played in Chirila – Sarkar, Dallas 2014) 10.Rc2 0-0 11.e4 Qe6 12.Bd3 Qb6 13.h3 Bd7 14.Rc1 Rfc8 15.0-0 Rxc1 16.Qxc1 Rc8= Black was doing fine in Schmalz – Arnold, corr. 2016. 7.e4?! This has been played in quite a few games, but White may regret attempting to play actively. 7...Nxc3 8.Bxc3 cxd4 9.Nxd4 0-0 10.Bc4 Other continuations are worse still: 10.Be2?! is too passive. 10...Qb6 11.Nb3 occurred in Miton – Kot, Jaroslawiec 1995, when 11...Nc6!N 12.Bxg7 Kxg7 13.0-0 Rd8µ would have made it hard for the white queen to find convenient squares, while the knight on b3 is clearly misplaced. 10.Qd2?! is also unimpressive, and 10...Qb6 11.Rd1 Nc6 12.Nxc6 Qxc6 13.Bd3 Bxc3 14.bxc3 Be6 15.0-0 Rfd8µ left White with no compensation for his weakened pawn structure in Happel – Leeuwenburgh, Haarlem 1995. 595

10...Qc7 11.Qe2 Nc6 12.Rd1 Now I can offer an improvement on Scepanovic – Pajzer, Sombor 2008:

12...Na5!N 13.Bd3 Rd8 14.Nb5 Qb6 15.Bxg7 Kxg7 16.Nc3 Be6 17.0-0 Nc6³ Black is slightly better thanks to his control over the d4-point. 7.Qa4†?! Such a queen check can occur in several lines of the Grünfeld. White is trying to disrupt Black’s development but in this instance the check has the opposite effect. In Obolenskikh – I. Kuzmin, Angarsk 2015, Black could have punished his opponent’s last move with:

7...Bd7!N White has a choice of retreat squares for the queen. Whichever one he chooses, Black obtains a 596

strong initiative with the possibility of beautiful winning variations: 8.Qa3 8.Qb3 Nxc3 9.Bxc3 cxd4 10.Nxd4 0-0 11.Qxb7 Bxd4 12.Bxd4 Nc6 13.Bc3 Rb8 14.Qa6 Nb4 15.Bxb4 Rxb4© gives Black excellent compensation for minimal material damage. The natural-looking 16.Qa3? allows an immediate win with: 16...Qb6 17.b3 Rf4! Beautiful geometry. 18.e3 Qf6 19.Rd1 Qc3† 20.Rd2 Rxf2–+ 8...Nb4 9.Rc1 0-0 10.dxc5 a5 11.Na4 Bxa4 12.Qxa4 Nd7 13.Qb3 White fails to solve his problems with 13.a3 due to 13...Bxb2! 14.axb4 axb4 15.Qxb4 Bxc1 16.Bxc1 Ra1 with a decisive advantage.

13...b6! 14.cxb6 Qxb6 15.a3 Bxb2!! Killing White. 16.Rb1 16.axb4 Bxc1 17.Bxc1 Qxb4†–+ leads to a winning endgame for Black. 16...Bxa3 17.e3 Nc5 18.Qc3 Rac8 19.Bc4

597

19...Bb2!! A furiously beautiful move. 20.Rxb2 Ncd3†–+ White falls apart.

7...Nxc3 8.Bxc3 cxd4 9.Nxd4 9.Bxd4 has the idea to trade off the strong g7-bishop, but 9...0-0 10.Bxg7 Kxg7 11.Qxd8 Rxd8 12.e3 Be6 13.a3 Nc6³ left Black with a pleasant development advantage in Lukin – Yandemirov, Podolsk 1990. 9...0-0 10.e3 Bd7 11.Be2 Nc6 598

We have reached a critical position, which has been tested in several GM games. 12.Nb3 Most strong players prefer this move. 12.Nb5 fails to cause Black any inconvenience. 12...Be6 13.Qxd8 Raxd8 14.a3 was played in Fridman – Mikhalevski, Belgium 2009, and here I suggest:

14...Rd7N 15.0-0 a6 16.Bxg7 Kxg7 17.Nc3 Rd2³ The active rook is rather unpleasant for White. 12.0-0 Nxd4 13.Bxd4 Be6 (13...Bc6!? is also possible) 14.Bxg7 occurred in Laketic – Vucinic, Subotica 2019, when 14...Kxg7N 15.Qa4 Qb6 would have been equal. 12.Bf3 was played in Cvek – Langner, Czech Republic 2011, plus a few other games, all involving players rated over 2400. Surprisingly, nobody took the opportunity to force White’s king to a suboptimal position:

599

12...Nxd4!N 13.Bxd4 Qa5† 14.Qd2 Qxd2† 15.Kxd2 Rfd8 16.Bxg7 Kxg7ƒ White has some problems connected with his exposed king. Finally, mass simplifications do not help White: 12.Nxc6 Bxc3† 13.Rxc3 Bxc6 14.Qxd8 Rfxd8 15.Bf3 Bxf3 16.gxf3 Rac8 17.Rxc8 Rxc8 18.Kd2 This was Miton – Macieja, Lubniewice 2002, and now 18...Kg7N³ would have given Black winning chances with no risk. The position resembles an old game of Rubinstein in which Black’s king made its way to h3 in the pawn endgame. 12...Be6 13.0-0 13.Qc2?! Bf5 14.e4 Bxc3† 15.bxc3 Be6 16.0-0 Qb6 17.Rb1 Rfd8³ left White with an inferior pawn structure in Manolache – Banzea, Bucharest 2019. 13...Bxc3 14.Rxc3 Qb6 15.a3 This position was reached in Li Di – Laurent Paoli, Vandoeuvre 2019. I recommend:

600

15...Rac8N 16.Qc2 Bxb3 17.Qxb3 Qxb3 18.Rxb3 Na5 19.Rb4 Rfd8 20.Ra4 Nc6 21.Bf3 a6= White has a bishop against a knight but Black’s rooks are more active, so the endgame is equal. B) 6.e4

This natural move is the usual way for White to follow 5.Bd2. 6...Nxc3 6...Nb6 has been slightly more popular but I prefer the text move.

601

7.Bxc3 0-0 We have reached an important branching point, where B1) 8.Bc4, B2) 8.Rc1, B3) 8.Nf3 and B4) 8.Qd2 all deserve attention. 8.f4?! costs White too much time. 8...Nc6 9.d5 Bxc3† 10.bxc3 Na5 11.Be2 c6 12.c4 was played in Sserwaniko – Frobisha, Kampala 2018, when 12...Qc7N 13.Qc1 Qb6 14.Nf3 Qc5 15.Rb1 Bg4³ would have made life difficult for White. His centre may be strong, but he cannot castle and the c4-pawn is weak, while Black may also be able to undermine the centre with ...f5. 8.h3? prepares Nf3 without allowing a pin, but is painfully slow, and 8...c5 9.d5 e6 puts White under serious pressure already, for instance: 10.Nf3 (Or 10.Bc4 Bxc3† 11.bxc3 b5! 12.Bxb5 Qa5 13.Qb3 exd5 14.exd5 Nd7 15.Be2 Rb8 16.Qc2 Nb6 17.Rb1 Re8 18.Kf1 Bf5 19.Bd3 Bxd3† 20.Qxd3 Qxa2–+ left White in a mess, with disharmonious pieces and an unsafe king in Ballicora – Vandivier, USA 1998.) 10...exd5 11.exd5 Nd7 12.Bxg7 Kxg7 13.Qd2 Re8† 14.Be2 Nf6 15.Rd1 In De Francesco – Baldauf, Austria 2015, Black missed the strongest continuation:

15...Ne4N 16.Qc2 Qa5† 17.Kf1 Nf6µ White’s king is awkwardly placed and the d5-pawn is weak. 8.Be2 This move is a sly attempt to develop the kingside without allowing the ...Bg4 pin. 8...c5 9.d5 e6 Again we should open the centre and isolate the d5-pawn. 10.Nf3 exd5 11.exd5 Nd7 12.0-0 Nb6 13.Bxg7 Kxg7 14.d6 14.Qc2 Qxd5 15.Rfd1 Qf5 16.Qc3† Qf6 17.Qxc5 Be6³ gave Black a risk-free edge in Zueger – Schlick, Germany 2006.

602

14...Qf6 15.Rc1 15.Qd2 Rd8 16.Rfd1 occurred in Teodorescu – Ocnarescu, Buzias 2002, when 16...Be6N³ would have resulted in the d6-pawn being more of a weakness than a strength. 15...Rd8 16.Rxc5 Rxd6 17.Qb3 Be6³ Schweda – Groszpeter, Austria 2011. Despite the simplifications and symmetrical pawn structure, Black’s pieces are better mobilized and White is under some pressure 8.h4!? This aggressive move has been played by Shirov, among other strong players. 8...Nc6! I like this method of putting pressure on White’s centre. 8...c5 is a typical move but it looks worse here in view of 9.d5 Bxc3† 10.bxc3 e6 11.d6ƒ when Black had some problems in Shirov – Maletin, Loo 2014. 9.d5 9.Ne2 protected the pawn in Chirila – Troff, Internet 2017, when 9...h5!N would have halted White’s kingside plans. Play may continue 10.d5 Ne5 11.Nf4 c6 12.Be2 cxd5 13.Nxd5 Nc6 14.Rc1 e6 15.Bxg7 Kxg7 16.Nc3 Nd4³ when Black’s possession of the d4-outpost gives him a small but lasting advantage. 9...Bxc3† 10.bxc3 Ne5 11.Qd2?! White should have preferred 11.h5!?N 11...Qd6 12.Be2 c6 13.Nf3 cxd5 14.hxg6 Nxg6 15.Qd2 Qf4 16.exd5 Qxd2† 17.Nxd2 Nf4 with a roughly equal game. 11...e6 12.h5 In Kuzmanovic – Mitrovic, Vrbas 2015, Black could have punished his opponent’s previous inaccuracy with a most natural move:

603

12...exd5!N 13.hxg6 fxg6 14.Qxd5† Qxd5 15.exd5 Bf5µ Black has a healthy lead in development as well as the better pawn structure. B1) 8.Bc4 White develops the bishop to the active position, but the drawback is that the bishop may be targeted by ...Nd7-b6 or ...c6 followed by ...b5. 8...Nd7 This is the most flexible continuation. The knight might go to b6 or f6 in the near future, while Black can also postpone the decision with ...c6 followed by a general advance of the queenside pawns. 9.Ne2 White can delay the development of this knight with 9.Qd2, as played in Kuzmanovic – Maksimovic, Belgrade 2018. In that case, I would propose the multipurpose 9...a5!?N, intending 10.Nf3 Nb6 11.Be2 (11.Bb3 would only invite 11...a4) 11...Na4 12.0-0 Nxc3 13.bxc3 b6÷ when Black’s chances are certainly not worse. The bishop pair will be useful, especially once Black opens the centre with ...c5 or ...e5. 9.Nf3 This simple developing move has the drawback of allowing Black to pin the knight. In Borda – Csapo, Budapest 2008, Black moved his knight to f6, but it would have been better to go the other way:

604

9...Nb6!N To my students, I say that in most cases a knight on b6 is not worth much in this pawn structure, but the current position is an exception. Here the knight has real value on b6, because if White answers ...c5 with d4-d5, then his pawn centre will be destroyed by means of ...f5. 10.Be2 c5 11.0-0 The above idea is shown after 11.d5 Bxc3†! 12.bxc3 f5 when White’s centre collapses. He can avoid a loss of material with 13.exf5 Bxf5 14.c4, but 14...e6³ leaves Black with the better game: he enjoys a lead in development and White will have to accept an isolated pawn, one way or the other. 11.dxc5 is also not great in view of 11...Qxd1† 12.Rxd1 Bxc3† 13.bxc3 Na4³ when Black enjoys the better pawn structure. 11...Bg4 12.d5 Bxc3 13.bxc3 e6 14.dxe6 Bxe6÷ I prefer Black’s long-term chances in this double-edged position, due to White’s weak queenside structure. 9...c6! This is anything but a standard solution in the Grünfeld Defence. Usually the c-pawn advances two squares to attack White’s centre, but here we are planning ...b5, ...a5, ...Bb7, ...b4 and only then ...c6c5. 10.0-0 10.f4!? was seen in Bocharov – Matinian, Khanty-Mansiysk 2013, when it would have been logical to continue the space-gaining plan on the queenside: 10...b5N 11.Bd3 a5 12.0-0 Qb6÷ A complex game lies ahead but I like such positions for Black, whose plans include ...Bb7 and ...c5. 10.Qc2 This was played in Szczechowicz – Chmielewski, Sopot 1997. I offer a strong, though unsurprising, 605

novelty: 10...b5!N 11.Bd3 11.Bb3 a5 12.a3 Ba6 13.0-0 Qb6ƒ sees Black capturing more space on the queenside; moreover, it may be possible for him to undermine the centre with a timely ...e5. 11...e5 12.Bb4 12.0-0 Qb6ƒ is promising for Black. The text move is more critical but it enables Black to start a storm.

12...exd4! 13.Bxf8 Bxf8 The key line continues: 14.Qxc6? 14.Nxd4? Nc5–+ leads to a decisive material gain. 14.0-0 is objectively the best try, but 14...Ne5 15.f4 Nxd3 16.Qxd3 Bb7 17.a3 c5ƒ offers Black fabulous compensation for the exchange. The text move is critical as it gains more material, but Black can launch a decisive attack. 14...Nc5! 15.Qxa8 Nxd3† 16.Kd2 Nxf2‚ White’s king is doomed, for instance: 17.Qd5 Bh6† 18.Ke1 Nd3† 19.Kf1 19.Kd1 Qa5–+ changes nothing. 19...Qf6†–+ The end is nigh. Returning to our main line, Black once again has a simple improvement available.

606

10...b5!N 10...Nf6?! was a mistake in Huber – Tompa, Rieden 1996. It was not worth determining the position of the knight at this stage. Now it blocks the g7-bishop and has no prospects itself, except for exerting slight pressure on e4 which is easily dealt with: 11.Qc2!N (rather than 11.Ng3?! when the knight is consigned to defensive duties) 11...b5 12.Bd3 a5 13.Bd2 Bb7 14.Bg5² Black’s pieces are passive, his counterplay is delayed and the c6-pawn is backward. 11.Bb3 a5 12.Rc1 a4 13.Bc2 e5÷ There arises a position with mutual chances, where Black’s plans include ...Bb7 and ...c5. B2) 8.Rc1 White activates his rook and prepares to recapture on c3 with it, as occurs shortly. 8...c5 Black needs to put pressure on the centre at some point, and now is an opportunity. 9.d5 9.Nf3?! cannot be recommended: 9...cxd4 10.Nxd4?! (10.Bxd4N is the lesser of the evils, though 10...Nc6 11.Bxg7 Kxg7 12.Qxd8 Rxd8 13.h3 Nd4 14.Bc4 Nxf3† 15.gxf3 Bd7³ still leaves White struggling due to his compromised structure) 10...Qb6 11.Nb3 In Zhuravlev – Sadsnev, USSR 1971, it would have been natural and strong to continue:

607

11...Rd8N 12.Qc2 Nc6 13.Bxg7 Kxg7 14.Bc4 Nb4 15.Qb1 Bd7 16.0-0 Rac8 17.a3 Nc6µ White is under serious pressure on the queenside. 9.dxc5 is playable although it gives White nothing special: 9...Qxd1† 10.Kxd1 Rd8† 11.Ke1 Bh6 12.Ra1 This was l’Ami – Bok, Amsterdam 2015, and here I suggest:

12...Be6N (Avrukh proposes 12...Bd7N intending ...Bc6, which is certainly possible, but I prefer the text move) 13.Nf3 Rc8 14.Nd4 Bd7 15.Rd1 Rxc5= Black has no problems. 9...Bxc3† 10.Rxc3 e6 Disregarding the loose pawn on c5, Black hurries to undermine the centre as quickly as possible, hoping to exploit White’s slow development and uncastled king. 608

11.Bc4 White develops solidly. Nobody has grabbed the pawn with 11.Rxc5?!, for good reason: 11...Qh4!? (11...Nd7 is also good) 12.Qc2 Nd7 13.Nf3 Qf6 14.Qc3 Qd8 15.Ra5 b6³ White’s rook is misplaced and he stands worse. 11.d6?! This aggressive more is more dangerous for White than for Black. 11...e5! Black isolates the d6-pawn, frees his bishop and takes control of the central dark squares. 12.Nf3 12.Rxc5 Nd7 13.Rd5 Qb6 14.Nf3 Qxb2 15.Qb3 Qxb3 16.axb3 f6 17.Bc4 Kg7 18.Nd2 Rd8³ leaves the d6-pawn weak rather than strong. In the long run, Black can create a passed pawn on the queenside. 12...Bg4 13.Rxc5 Bxf3 14.gxf3 Nc6 15.Bh3 Avrukh mentions 15.d7 Qe7 16.Rd5 Nd4 17.Bh3 f5© and concludes that Black has good compensation, and I completely agree with him. Now Black can improve on Moiseenko – Rodshtein, Rijeka 2010, by means of:

15...Nd4!N This novelty was offered by Avrukh in Grandmaster Repertoire 9. 16.Rd5 16.Rxe5 Qxd6 17.Rd5 Qb4†ƒ is mentioned by Avrukh; White is under unpleasant pressure. 16...Qh4 17.Bg2 In the event of 17.Rxe5? Nc6–+ White loses a piece. 17...Ne6 18.Rxe5 Qf6 19.Rxe6 fxe6µ Despite the fact that White is materially okay, with a bishop and two pawns against a rook, the 609

position obviously favours Black, since the d6-pawn is weak and White’s kingside pawn structure has been compromised. 11...exd5 12.Bxd5 Nd7 13.Ne2 13.Nf3 Nf6 14.Rxc5 Qe7 gives Black no problems, as Avrukh points out. 13...Nf6 14.0-0 Nxd5 15.exd5 b6 16.Re3 Ba6 17.Re1 Bxe2 18.R1xe2 Qd6 19.h4 Rad8 20.Rd2 Rfe8= We have been following Shirov – Sutovsky, Kallithea 2008. A roughly equal ending has arisen, and the rivals soon agreed to a draw. B3) 8.Nf3

Without further ado, White develops the knight. The drawback of this continuation is that is invites Black’s obvious reply: 8...Bg4 Pinning the knight and thus increasing the pressure on the d4-point. 9.Be2 9.Qd2 Bxf3 10.gxf3 e5! gives Black plenty of activity. An accurately played game continued: 11.dxe5 Qh4 12.Bc4 Nc6 13.e6 fxe6 14.Bxe6† Kh8 15.Bxg7† Kxg7 16.Qc3† Qf6 17.0-0-0 Kh8 18.h4 Qxc3† 19.bxc3 Rxf3 20.h5 g5= Rydholm – Martin Sanchez, corr. 2016. Despite White’s scattered pawn structure, he has enough dynamic resources to maintain the balance. 9...e6!

610

Before hitting the d4-pawn with ...Nc6, Black anticipates the d4-d5 advance. 10.Qd2 White delays castling in order to free up the d1-square for the rook. 10.0-0 is met by 10...Nc6 when the threat of ...Bxf3 forces White to compromise his centre: 11.d5 (11.e5 was played in Morozov – Bouguern, email 2013, and here I propose 11...Bf5!?N 12.Re1 Be4, transferring the bishop to d5, with a complex game) Now in Gaertner – Sutter, Switzerland 1996, it would have been good to make the following minor-piece trades:

11...Bxc3!N 12.bxc3 Bxf3 13.Bxf3 exd5 14.exd5 Ne5 15.Be2 Qd6÷ Black is left with a strong knight versus an enemy bishop which is restricted by the d5-pawn. White’s position is objectively still okay, but I would definitely prefer Black. 10...Nc6 11.Rd1 Qe7 12.0-0 Rfd8 Once again, the pressure against the d-pawn compels White to advance it.

611

13.d5 Bxc3 14.bxc3 Bxf3 15.Bxf3 Ne5 15...exd5 16.exd5 Ne5 17.Be2 is another possible move order. 16.Be2 exd5 17.exd5 Qc5 18.Qd4 Nd7 Just as in the note to White’s 10th move above, Black has the easier game due to his better minor piece. 19.Bf3 Qxd4 20.Rxd4 Ne5 21.Be2 Nd7 22.Bf3 Nc5 23.a4 Rd7 24.Rc4 b6

25.g3? A serious mistake. 612

It was worth getting rid of a weak pawn with 25.a5, although the position remains more pleasant for Black because the bishop is restricted by the d5-pawn. 25...a5! Fixing the weakness on a4. 26.d6 Rad8 27.dxc7 Rxc7 White faced a difficult defence in Parkes Navea – Tortora, corr. 2008. B4) 8.Qd2 First played by IM Lalev in 1991, this move has since become established as the main line, having been played by such players as Le Quang Liem, Mamedyarov and Moiseenko. White plays flexibly, avoiding any commitment of the kingside minor pieces until Black reveals his intentions. 8...c5! Black immediately puts pressure on White’s centre. 9.d5 9.dxc5 gives White nothing. 9...Qxd2† 10.Kxd2 Rd8† 11.Kc2 was seen in Dreev – Roiz, Dagomys 2009, and here I recommend:

11...Be6!?N (Avrukh suggests 11...Bd7 with the idea of ...Bc6, but I prefer the text move) 12.Nf3 Rc8 13.Bxg7 Kxg7 14.Kd2 Rxc5= Black has no problem with this endgame.

613

9...Bxc3 10.bxc3 It is less logical to take with the queen: 10.Qxc3?! This removes a defender of the d5-point. 10...e6 11.Bc4 Surrendering the centre with 11.dxe6 Bxe6 only helps Black, for instance: 12.Nf3 Qe7 13.Be2 Nc6 14.0-0 Rfd8 15.Rfc1 b6³ In Solozhenkin – Holmsten, Finland 2005, Black owned the d4-outpost as well as the more useful pawn majority. 11.Rd1 exd5 12.Rxd5 Qe7 13.Bd3 was seen in Flear – Mista, Hastings 2016. It is clear that the rook on d5 is misplaced, and 13...b6N 14.Nf3 Nc6 15.0-0 Nb4 16.Re5 Be6µ would have left White’s pieces in obvious disharmony.

11...b5! With this dynamic strike, Black fights for the centre. 12.Bxb5 exd5 13.exd5 The practically suicidal 13.Qxc5? Na6 14.Qc3 was played in Rietze – Engel, Hamburg 2016, when 14...Rb8N–+ would have given Black a decisive advantage, since White is so far behind in development. 13...Qxd5 14.Nf3 Qe4† 15.Kf1 This was Kalavannan – Mis, Riga 2018. In this position, I recommend: 15...Nd7N 16.Re1 Qb7³ Black’s slightly inferior queenside structure is outweighed by his lead in development and pressure along the b-file.

614

10...Qd6! An important move. Black intends to undermine the enemy centre by means of ...f5 and/or ...e6, but first he should establish a blockade on the dark squares. White has two main options: B41) 11.Nf3 and B42) 11.f4. 11.Bc4?! only exposes the bishop, and 11...Nd7 12.f4 Nb6 13.Bb3 c4 14.Bc2 f5 15.Ne2 fxe4 16.Bxe4 Bf5µ gave Black a clear advantage in Gupta – Li Chao, Wijk aan Zee 2010, due to the weakness of the d5-pawn. 11.h4!? With this adventurous move, White tries to launch an attack. Objectively it is not fully correct, but it leads to interesting positions where both sides will have to be careful. 11...f5! Black was planning this move anyway, and it also coincides with the well-known principle of meeting a flank attack with a counterblow in the centre. 12.h5 fxe4 13.hxg6 Qxg6 14.Ne2 Nd7 15.Rh6 Qf7 16.Qg5† Kh8 17.0-0-0 In this sharp position, Black should maximize his piece activity by means of:

615

17...Rg8!N Improving on Ivanchuk – Dominguez Perez, Beijing (rapid) 2014, in which Black captured on f2. 18.Qh4 18.Qe3 is well met by 18...Nf6 19.f3 b6µ followed by ...Bb7 or ...Ba6, when White’s attack has been repelled and his d-pawn is weak. 18...Ne5! 19.Qxe4 Ng4 20.Rh4 Nxf2 21.Rf4 Nxe4 22.Rxf7 Bg4

23.Re1 23.Rxe7? Rae8–+ leaves White in a desperate situation due to Black’s active pieces and the horrible pin on the knight. 23...Bh5 24.Rxe7 Rae8 25.Rxe8 Rxe8 26.Nf4 Nxc3µ Black has every chance of realizing his extra pawn. 616

B41) 11.Nf3 This simple developing move enables Black to obtain a comfortable game. 11...Bg4 12.Ng5!? This is White’s most interesting attempt. 12.Ng1 invites a repetition, and indeed Black does not seem to have anything better than 12...Bc8. 12.Be2 Nd7 13.0-0 was played in Carter – Johnston, Colonie 2016. In the game Black continued with the thematic plan of ...e6 and ...exd5, which is enough to reach a satisfactory position. However, I prefer the idea of playing on the dark squares with:

13...Bxf3!?N 14.Bxf3 b5 15.Qe3 Rab8 16.Rfe1 e5= An interesting position has arisen where White has a protected passed pawn but it is well blockaded, and Black can look to create a passed pawn on the queenside in the future. Transferring his knight to the ideal d6-square is another attractive plan.

617

12...f6 The next few moves are practically forced, and result in an unusual pawn structure. 13.h3 fxg5 14.hxg4 Qe5 15.f3 This seems the most logical way for White to secure his pawns. 15.Qe3 Nd7 16.0-0-0?! is too optimistic. 16...a6 17.f3 b5 18.Kc2 Rab8 19.Be2 occurred in Stachowiak – Sieciechowicz, Legnica 2015, when it would have been good to play:

19...a5N‚ Advancing the queenside pawns would be a good plan for Black anyway – and it is especially unpleasant for White with his king located on that flank.

618

15...Nd7 16.a4 Kg7 17.Kd1 This position arose in Ruefenacht – Nataf, corr. 2014. Black has more than one decent continuation but I think the most human decision would be:

17...a6!?N Preparing a queenside advance. My illustrative line continues: 18.Kc2 h6 19.Qe3 Rfb8 20.Be2 b5 21.axb5 axb5 22.Rxa8 Rxa8 23.Bxb5 Qc7 24.Rb1 Ne5 25.c4? This seemingly tempting move leads to difficulties. White should settle for 25.Be2 Ra2† 26.Rb2 Ra1 27.Rb1 Ra2†= with a draw. 25...Ra2† 26.Rb2 Qa5 27.Qc3 Qxc3† 28.Kxc3 Rxb2 29.Kxb2 h5 Despite White’s extra pawn, the endgame is dangerous for him. For example:

619

30.gxh5 gxh5 31.Kc3 h4 32.Kd2 Nxf3†!–+ White’s position is hopeless. B42) 11.f4

This tricky move increases White’s central control and rules out a quick ...f5, which can now be met by e4-e5. The drawback is that White falls further behind in development, so Black should not hesitate to open lines. 11...e6 12.Bc4 b5! This pawn sacrifice enables Black to amplify the pressure on White’s centre. 620

13.Bxb5 exd5 14.Bc4 An interesting tactical trick, but White will not be able to preserve his extra pawn due to his lack of development. The greedy 14.Qxd5? leads to big problems for White after: 14...Qb6! 15.Be2 (after 15.Qxa8 Bb7–+ the queen is trapped) 15...Bb7ƒ Despite White’s extra pawn, he is under unpleasant pressure and will have to suffer for a long time. The simple 14.exd5 is not so bad although White can hardly hope to gain an advantage with it: 14...Nd7 15.Bxd7 Bxd7 16.Nf3 Bb5 17.Kf2 Rad8 18.Rad1 Rfe8 19.Rhe1

19...f6 20.a3 Rxe1 21.Rxe1 Qxd5 22.Qxd5† Rxd5= Gelfand – Dominguez Perez, Wijk aan Zee 2014. The position is about equal, although Black has some practical chances to cause problems using his rook and bishop, which tend to work together better than a rook and knight.

621

14...Bb7 15.Nf3 In the event of 15.Bxd5 Bxd5 16.Qxd5 Qxd5 17.exd5 Re8† White can hardly hope to retain his extra pawn. For instance, 18.Kf2 Nd7 19.Nf3 Nf6 20.Rhd1 Ne4† 21.Kf1 Nxc3 22.Rd3 Na4 with equality. 15...Rd8 16.Bxd5 Bxd5 17.Qxd5 Qxd5 18.exd5 Rxd5= After a stormy start to the game, the players were approaching a drawn harbour in Tomashevsky – Nepomniachtchi, Satka 2018. Conclusion 5.Bd2 is a rather popular set-up which continues to find a following among strong grandmasters. The ensuing positions are quite complex and require accurate handling from the second player. However, when armed with the knowledge from this chapter, the reader should not have opening problems. Black’s game is usually based on the undermining of White’s centre by means of ...c5; however, there are certain positions where he should favour the unusual plan of ...c6, ...b5 and ...a5, delaying ...c5 until space has been gained on the queenside. One way or another, Black will usually look to play on the dark squares, especially when an early Nf3 is played, allowing ...Bg4.

622

A) 7.Ba3 Nd7! 8.Nf3 c5! 9.Qb3! 0-0 333 A1) 10.Bd3 337 A2) 10.Be2 338 B) 7.Bg5 Game 5 341 C) 7.Qa4†!? Qd7! 350 C1) 8.Bb5!? 351 C2) 8.Qa3 353 C3) 8.Qb3 354 D) 7.Bb5†!? c6! 8.Ba4 0-0 9.Ne2 b5! 10.Bb3 a5 356 D1) 11.a4 359 D2) 11.0-0 Game 6 360

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Finally we will begin our investigation of White’s most popular and logical move. With this pawn 623

advance, he captures the centre and hopes to control the game with it. 5...Nxc3 6.bxc3 Bg7 Black develops the Grünfeld bishop to its characteristic home, and intends to attack White’s centre with ...c5. In this chapter, we will consider a number of significant sidelines, the main ones being A) 7.Ba3, B) 7.Bg5, C) 7.Qa4†!? and D) 7.Bb5†!?. The most popular options of 7.Be3, 7.Nf3 and 7.Bc4 will be covered in subsequent chapters. A) 7.Ba3 The idea of this move is to inhibit ...c5 and exert pressure along the a3-f8 diagonal. This move may be relatively rare but it has been known for a long time, having first been played by Grandmaster Levenfish against Rosenthal in Moscow in 1924. Capablanca subsequently tried it in 1926, but not many other strong players followed suit and the variation remained in the shadows for a long time. In modern times, Le Quang Liem, Milov and Granda Zuniga are among those who have played this way, although the move is mostly used as a surprise weapon rather than for its objective strength. 7...Nd7! Black insists on carrying out the thematic ...c5 blow, and White finds himself under pressure to justify committing his bishop to a3 so early. 8.Nf3 This seems most natural; White needs to develop his kingside somehow. 8.Rc1 c5 9.d5 Qa5 10.Qb3 0-0 11.Nf3 was seen in Garcia Gonzales – Lechtynsky, Cienfuegos 1985, when Black could have obtained good counterplay with:

624

11...Rb8!N 12.Be2 Nf6 13.Nd2 (13.Bd3 e6 is even worse for White; for instance, 14.0-0 exd5 15.exd5 Nxd5 16.Qxd5 Qxa3 17.Ng5 b5–+ when Black is a pawn up for no compensation, and he has two powerful bishops to boot) 13...e6 14.Nc4 (I also checked 14.d6 Bh6 15.e5 Bxd2† 16.Kxd2 Nd7 17.f4 b5 18.Bf3 c4 19.Qb2 Bb7µ when the d6-pawn is reliably blocked, and Black can break through on the queenside with ...a5 and ...b4) 14...Qc7 15.d6 Qc6 16.0-0 Nd7µ White’s pawns on e4 and d6 are weak, and he must also worry about ...b5. White can also try a hybrid of the Ba3 set-up with the big main line of Chapter 15: 8.Bc4 c5 9.Ne2 9.Qb3 0-0 10.Nf3 cxd4 transposes to the 9.Bc4?! line in the notes to the main line below. Avrukh does not mention the text move but I think it is worth considering. 9...0-0 10.0-0 Qc7 11.Rc1 a6 12.f4 I decided to take this aggressive continuation as the main line. A calmer way to play is 12.Qd2 Qa5 13.Qb2 Rb8 14.Bd3 b5 15.Bxc5 Nxc5 16.dxc5 as seen in Kovacs – Milic, Vienna 1952, when 16...Rd8N 17.Rcd1 Bg4 18.f3 Bd7µ followed by ...Qc7 would have won back the pawn, leaving Black clearly better thanks to his two bishops and better pawn structure. 12.e5 weakens the d5-square. 12...b5 13.Bd5 was played in Schlickenrieder – Blank, Hofheim 1995, when 13...Bb7N 14.Bxb7 Qxb7 would have left Black with plans including centralization of the rooks, ...cxd4 and ...Nb6-d5. White may try the disruptive 15.e6!?, but 15...fxe6 16.dxc5 Ne5 17.Nd4 Nc4 18.c6 Qc8 19.Bxe7 Rf7 20.Bg5 e5 21.Nb3 Qxc6³ leaves Black a little better thanks to his more active pieces; and later he can put pressure on the c3-pawn. 12...b5 13.Bd3 Here we can improve Black’s play from Golubov – Lesovoi, St Petersburg 2017, by means of:

625

13...Qa5N 14.Bxc5 No better is 14.Bb2 Qxa2 15.Qd2 Qa5µ when White lacks any compensation for the pawn. 14...Nxc5 15.dxc5 Qc7 16.Kh1 Qxc5µ Black is obviously better due to his bishop pair and superior pawn structure. 8.Qb3 This move usually transposes to one of the main variations. 8...c5 9.Rd1 Only this option has an independent meaning. 9.Nf3 is the main line below. 9.Bc4 0-0 10.Nf3 cxd4 transposes to the 9.Bc4?! line in the notes to the main line below. 9...0-0 10.Be2 Qa5 11.Nf3

626

11...Nf6! 12.Bxc5 12.e5? is hardly playable: 12...Be6 13.Qb2 Nd5 14.Bxc5 Nxc3 15.Qd2 b6 16.Rc1 bxc5 17.dxc5 Rab8 18.Bd3 occurred in Bouget – Rakotomaharo, Paris 2019, when 18...Rfd8!N 19.0-0 Qxa2 20.Rc2 Rxd3 21.Rxa2 Rxd2 22.Rxd2 Bd5–+ would have given Black a decisive advantage. 12.Bd3 Be6 also leads to problems for White after 13.Qb2 (or 13.d5 Bd7 14.Rc1 Rac8 15.Qb2 Nh5 16.g3 b5–+ intending ...b4) 13...b6 14.0-0 Bg4µ as given by Avrukh. The bishop on a3 has no purpose and White’s centre is under fire. 12.Nd2 cxd4 13.Bb4 Qg5! (Avrukh gives 13...Qe5 but the text move is stronger) 14.cxd4 Qxg2 15.Bf3 Qh3 16.Bxe7 Nxe4 17.Bxf8 Nxd2 18.Rxd2 Bxf8© gives Black more than enough compensation for the exchange. 12...Nxe4 13.Bb4 Avrukh also considers: 13.Bxe7 Re8 14.Bb4 Qc7 15.c4 a5 16.Ba3 a4 17.Qb2 Bg4 18.0-0 Ng5ƒ With a strong initiative for Black. We have been following De Virgilio – Molo, corr. 2008. This position was analysed by Avrukh and I completely agree with him that Black should continue:

627

13...Qb6!N 14.Nd2 14.c4 a5 15.Ba3 Qxb3 16.axb3 a4 17.bxa4 Rxa4 18.Ra1 Bf6µ does not help White. 14...Be6 15.Qa3 Nd6 16.0-0 Qc7³ Avrukh points out that Black will follow up with ...a5, and his position is in perfect order. I would add that Black benefits from a compact pawn structure, whereas the c3- and d4-pawns are long-term weaknesses.

8...c5! Although the bishop on a3 shoots at c5, it has failed to prevent Black from carrying out the thematic challenge to White’s pawn centre. Capturing on c5 is out of the question for White due to the hanging c3-pawn; but even if it were defended, taking on c5 would lead to serious structural damage. Thus, the supposed pressure exerted by the bishop along the a3-f8 diagonal proves to be a mere fiction. Note that 628

White has to do something about the threat of 9...cxd4, intending to meet 10.cxd4 with 10...Qa5†. 9.Qb3! The queen moves to its optimal position and makes room for the rook to go to d1. 9.Bd3? cxd4µ was dismal for White in Wedel – Schneider, corr. 1990, since 10.cxd4?? Qa5†–+ picks up the dark-squared bishop. 9.Rb1?! was seen in Lacunza – Hoehne, email 2013, when 9...Qa5!N 10.Qb3 0-0 11.Be2 Nf6 12.Bxc5 Nxe4 13.Bb4 Qc7 14.c4 Bg4³ would have been excellent for Black. 9.Qd2?! Qa5 emphasizes the unfortunate position of the a3-bishop. Following 10.Bb2 0-0 11.Be2 Nb6 12.0-0 Black had achieved a convenient arrangement of his pieces in Gering – Frotscher, Blankenburg 2008, and he could have secured a significant advantage with:

12...cxd4!N 13.Nxd4 (13.cxd4 Qxd2 14.Nxd2 Na4 gives Black a big advantage) 13...Rd8µ The weak c3-pawn will cause White a lot of pain. Another doubtful continuation is: 9.Bc4?! cxd4 10.Qb3 At least this move enables White to save his pawn structure. 10.0-0? was seen in Hvenekilde – J. Fries-Nielsen, Aalborg 1979, when Black could have safely eaten the pawn: 10...dxc3!N 11.Bxf7† Kxf7 12.Ng5† Ke8 13.Ne6 Qa5 14.Nxg7† Kf7 15.Qb3† Kxg7 16.Bxe7 Nf6 17.Bb4 Qb6 18.Qxc3 Be6 White has nothing like enough compensation for the piece. 10.Nxd4 leads to an unpleasant game for White due to his weak structure. 10...0-0 11.0-0 Qc7 12.Qb3 Nc5 13.Qb4 occurred in Kladiva – Vadasz, Balatonbereny 1995, when Black could have picked up the central pawn: 13...Nxe4N 14.Rfe1 Bf5 The tactical point is that 15.g4? is refuted by 629

15...a5 16.Qb3 Qf4 17.gxf5 Qg4† 18.Kf1 Nd2 mate! 10...0-0 11.cxd4 Nb6 12.0-0 White cannot save the two bishops with 12.Be2 Be6 13.Qb4 Rc8 14.0-0, as played in Akobian – Xiong, Saint Louis 2018, in view of Svidler’s recommendation: 14...Nc4!N 15.Rac1 Nxa3 16.Qxa3 Rxc1 17.Rxc1 Qb6³ Black’s bishop pair gives him the more pleasant game. 12...Nxc4 13.Qxc4 Bg4 A critical position has arisen, where White must find the best way to minimize his disadvantage.

14.Ne5 14.Rad1?! was a clear case of “the wrong rook” in I. Ivanov – Lonoff, Chicago 1990. I recommend 14...Qd7N 15.d5 Bxf3 16.gxf3 b5 17.Qe2 a5 18.Bb2 Bxb2 19.Qxb2 b4µ when Black benefits from the better pawn structure and safer king, as well as the possibility to create a passed pawn on the queenside in the future. 14.Rac1 placed the rook more actively in Bruno – Dvirnyy, Rome 2013. Still, after 14...Re8N 15.Rfd1 Qd7³ Black has the more pleasant game: he may or may not trade in his advantage of two bishops by exchanging on f3 and weakening White’s kingside structure. 14...Be6 15.Qb4 Qb6 16.Qxb6 axb6 17.Bxe7 Rfe8 18.Bb4 Bxa2³ We have been following Laczay – Laffont, email 2013. Avrukh calls the position equal, but I think Black has practical winning chances due to the advantage of two bishops.

630

9...0-0 We will consider two main options: A1) 10.Bd3 and A2) 10.Be2. 10.Bc4?! cxd4 takes us back to the 9.Bc4?! line in the notes above. 10.Rd1 Qa5 11.Be2 reaches the note on 8.Qb3 on page 334. 10.Rc1 was played in Frost – Gibbons, corr. 2005. 10...Rb8!?N seems a good reply, after which I analysed an interesting illustrative line: 11.Be2 Nf6 12.e5 Nd5 13.Bxc5 Nf4! 14.Bxa7 Be6 15.Bc4 Bxc4 16.Qxc4 b5 17.Qc6 Ra8 18.Bb6 Qb8 19.Ra1 Ra6 20.Qe4 Qc8 21.Qxf4 Qxc3† 22.Ke2 Qc4† 23.Ke1 Rxb6ƒ Black has a dangerous initiative; White’s extra pawn is clearly less relevant than the fact that his king is stuck in the centre. A1) 10.Bd3 Unlike variation A2 below, White guards the e4-pawn in anticipation of ...Nf6. On the other hand, the placement of his queen and bishop encourages Black to expand on the queenside and threaten ...c4. 10...Rb8! 11.0-0 11.Rd1 b5 12.dxc5? was played by a strong GM but he was in for a rude awakening:

631

12...b4! Black already has a decisive advantage. After the further 13.Bxb4 (13.cxb4 is no better in view of 13...Nxc5 followed by ...Nxd3† with ...Ba6 to follow, when Black is winning regardless of whether or not White exchanges queens) 13...a5 14.c6 axb4 15.cxd7 Bxc3† 16.Ke2 Bxd7 17.Bc4 Qc7 18.Ng5 e6 19.Rd3 Ra8–+ White’s position was hopeless in Khenkin – Maletin, Tashkent 2013. 11...b5 12.Qc2 12.Be2 Bb7 13.Qc2 was tried in Gomez Fontal – Herrera, Santa Clara 2001, when 13...a6N 14.d5 Qb6 15.Bc1 e6µ would have given Black a superb position. 12...a6 13.Bc1 Improving the position of the bishop. White has a couple of roughly equivalent alternatives: 13.e5N has not been tested but is worth considering. A good response is 13...b4! 14.cxb4 cxb4 15.Bc1 Nb6 16.Be3 Nd5³ when Black benefits from the outpost on d5. 13.Rad1 saw White preparing for central play in Thys – Radhey Derling, corr. 2012. 13...Qc7N is a logical reply to take the queen off the same file as the rook, and after something like 14.Rfe1 Bb7 15.Bc1 Rbc8 16.Qb2 Rfd8 17.Bg5 Nb6³ Black has harmoniously arranged his forces. 13.Rfd1!? Qc7 14.e5 Bb7 15.Ng5 reached an interesting situation in Szenczy – Laczay, email 2008. I propose an improvement:

632

15...e6!N The point is that the tempting 16.Ne4? has a flaw. (White should prefer 16.Qe2 although 16...Qa5 17.Bb2 Bd5 18.Ne4 Rfd8ƒ gives Black the more pleasant game) The refutation is 16...Rfc8 17.Nd6 cxd4! 18.Nxc8 Rxc8–+ when White faces inevitable material losses. 13...Bb7 14.Bg5 Nb6 Black activates the knight, as the prospect of White taking on c5 is not scary. 15.Rac1 We have been following Kazmierski – Iula, email 2013. This would have been a good moment to relieve the central tension in a somewhat unusual way:

633

15...c4!N 16.Be2 Qc7 17.Bh4 Rbe8 18.Rfe1 Qc6ƒ Black’s pieces are excellently placed, whereas White does not have the harmony needed to advance his central pawns in an effective way. A2) 10.Be2

10...cxd4 Black’s biggest problem is choosing between a number of promising alternatives. 10...Qc7 is recommended by Avrukh in Grandmaster Repertoire 9, and leads to a promising game for Black. 10...Rb8!? is another idea which deserves close attention, and is approved by our “iron friend”. 11.0-0 (11.e5 occurred in Hirneise – Naumann, Germany 2012, when 11...cxd4 12.cxd4 Nb6 13.0-0 Be6 14.Qb4 Nd5 15.Qd2 Qd7ƒ would have enabled Black to derive a long-term benefit from the d5outpost) 11...b5 12.Rad1 Qc7 13.e5 In this position from Noble – Goffin, corr. 2005, Black had a strong continuation:

634

13...Bb7!N 14.e6 c4 15.exf7† (15.Qb4 fxe6 16.Qxe7 Rbe8 17.Qd6 Qc8³ reaches essentially the same type of position) 15...Rxf7 16.Qb1 Qc6³ Black has an active game with a battery along the a8-h1 diagonal as well as good prospects along the semi-open f-file. 11.cxd4 Nf6 Black takes aim at the e4-pawn, practically forcing it to advance and thus securing the desired outpost on d5. 12.e5 Nd5 13.0-0 b6 14.Ng5!? This move prepares a restructuring of White’s pieces: the bishop goes to f3 and the knight will go to e4, before the rooks choose their positions. Another game continued 14.Bc4 Bb7 15.Rac1 Bh6 16.Rc2 a6 17.Bc1 and now in Yehuda – C. Wolff, email 2010, a natural and promising continuation would have been:

635

17...Bxc1N 18.Rfxc1 b5 19.Bxd5 Qxd5ƒ White’s control over the c-file is less important than Black’s superior minor piece and the long-term mobility of the queenside pawn majority compared to White’s rigid central formation.

14...h6 15.Bf3 Bb7 16.Ne4 Qd7 17.Rab1 This move is not a mistake exactly, although White’s plan is not completely clear. 17.Rac1 Rfc8 18.g3 occurred in Noble – Lahdenmaeki, email 2007, when 18...Rxc1N 19.Rxc1 Rd8 would have been essentially equal, although I would slightly prefer Black’s side. 17.Rfe1 Rac8 18.Rad1 is arguably the most logical continuation, centralizing the rooks. In Atalik – Lin Weiguo, Beijing 1997, it would have been logical to play: 636

18...Rc7N Preparing to double on the open file. A sample continuation is 19.h3 Rfc8 20.Bg4 e6 21.Bd6 Rc6 22.Bf3 Nc3 23.Nc5 Rxc5 24.Bxb7 Qxb7 25.dxc5 Nxd1 26.Rxd1 Qc6 27.cxb6 axb6 28.Rb1 Qe4 29.Qxb6 Bxe5= when the equilibrium is maintained. 17...Rab8 Black overprotects the bishop, enabling him to consider moves such as ...Nc7 or ...Nf4 without allowing a troublesome Nc5 trick. We will follow an instructive game for a few more moves. 18.Nd6 This move appears tempting but it only leads to a simplified position where White is left with a weak, isolated pawn on d4. Nevertheless, White is still not objectively worse. 18.Qb5? would be a mistake in view of 18...Qxb5 19.Rxb5 Rfc8µ when Black takes over the open file. 18.Rfe1 was a sensible way to maintain the tension. Then 18...Nc7 19.Qe3 Rfd8 20.Rbd1 Ba8 reaches a typical situation where I marginally prefer Black because of the useful outpost on d5, but the computer just gives zeros. 18...exd6 19.Bxd5 Bxd5 20.Qxd5 Rfd8 21.exd6 White correctly avoids exchanging bishops, as after 21.Bxd6? Bf8 22.Rfe1 Rbc8 23.Qf3 Bxd6 24.exd6 Qxd6µ the weakness of the d4-pawn is much more noticeable. 21.Qxd6? is also bad in view of 21...Qa4 22.Qb4 Rxd4 23.Qxa4 Rxa4 24.Bd6 Re8µ when the pawns on a2 and e5 are weak. 21.f4 is not completely satisfactory due to 21...dxe5 22.Qxd7 Rxd7 23.dxe5 Rd2ƒ when Black’s active 637

rook gives him some winning chances in the endgame.

21...Bf8 22.Rfe1 Rbc8 23.Qf3? White makes his first serious error of the game; there was absolutely no need to remove the queen from d5, where it stood perfectly. 23.g3 Bxd6 24.Rbc1 Rxc1 25.Bxc1 Bb4 26.Qxd7 Rxd7 27.Re8† Kg7 28.Rc8 Be7 29.Be3 Bf6 30.Rc3= would have been more accurate. By utilizing a few simple tactics to keep the d4-pawn safe, White simplifies to a safe endgame where his king will quickly get to d3. 23...Bxd6 24.Bb2 24.Bxd6 Qxd6 25.Rbd1 Rc4µ gives White problems with the weak d-pawn. So far, Black had done everything right in Gagunashvili – Mchedlishvili, Tbilisi 2015, but here he failed to maximize his advantage.

638

24...Rc2!N 24...Qf5?! saw Black exchange queens too early. The text move leads to a clearly improved version where Black’s rook has already taken up a great position on c2. 25.Qd3 Qf5 26.Qxf5 gxf5 27.g3 Rdc8µ The doubled f-pawns are hard to attack, while Black’s active rooks and the weakness of the a2- and d4-pawns gives him excellent chances in the endgame. B) 7.Bg5

GAME 5 Reinhard Wegelin – Hans-Joachim Quednau Email 2013 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bg5

639

White develops the bishop actively, with one possible plan being Qd2 and Bh6 to remove the Grünfeld bishop. The bishop also pins the e7-pawn, while tying the black queen to its defence. On the other hand, unlike the Be3 lines examined in the next chapter, the bishop does not support the d4-pawn. The present variation is rather young, having first been tested at a high level by Yuri Kruppa, against Ganguly in Dubai 2004. The Ukrainian GM played it in several more games, so it would not be unreasonable to call 7.Bg5 the Kruppa Variation. It has subsequently been used by such grandmasters as Navara, Nakamura, Ponomariov, Korobov, Milov, Krasenkow and Nyback, and is generally a good choice for those who like to avoid mainstream theory and reach an original position relatively early. 7...c5 Black challenges the centre in the usual way. 8.Rc1! Already White needs an ‘only move’ to avoid falling into difficulties. The text move indirectly defends the d4-pawn by tactical means, while also defending the c3-pawn and thus preparing d4-d5, gaining space and avoiding an accident on the d4-square. Other continuations are clearly worse, for instance: 8.Nf3?! Nc6 gives White nothing better than 9.Bb5, which is covered under the 8.Bb5†?! line below. 8.Qd2? is senseless, and 8...cxd4 9.cxd4 Qxd4 10.Qxd4 Bxd4 11.Rc1 Nc6 12.Bb5 Bd7 13.Ne2 Bb6µ left White with no compensation for the missing pawn in Dresher – LeCours, corr. 2013. 8.Bb5†?! has been tried in a few correspondence games but the bishop only interferes with White’s play: 8...Nc6 9.Ne2 (another game continued 9.Nf3 cxd4 10.Nxd4 0-0 11.Bxc6 bxc6 12.Nxc6 Bxc3† 13.Ke2 Qe8 14.Nxe7† Kh8 15.Rc1 Ba6†µ when White’s king was in a deplorable situation in Heilala 640

– Lahdenmaeki, corr. 2010) 9...0-0 10.Be3 White walks headfirst into a simple tactical refutation:

10...Nxd4! 11.cxd4 Qa5† 12.Qd2 Qxb5 13.0-0 cxd4 14.Nxd4 Qa4–+ In M. Brown – A. Johansen, corr. 2013, Black enjoyed a healthy extra pawn as well as two powerful bishops in an open position.

8...0-0 Black should evacuate his king from the centre before undertaking anything active. The tactical point of White’s last move is revealed after 8...cxd4 9.cxd4 Bxd4?? 10.Rxc8! Qxc8 11.Qxd4 when White’s bishops will easily outwork Black’s rook and pawn. 9.Nf3 White develops the knight and defends the d4-pawn, at the cost of allowing a ...Bg4 pin. Two other 641

continuations have been tried: 9.Ne2 seems a strange choice. 9...h6 10.Be3 Qa5 11.Qd2 Nc6 12.d5 Ne5 13.Nf4 occurred in Unander – Maatman, Vlissingen 2018, when Black missed a nice opportunity:

13...f5!N 14.exf5 Bxf5 15.Be2 Rad8 16.c4 Qxd2† 17.Bxd2 Rd6 18.0-0 Ra6³ Black is well placed to attack the weak pawns on a2 and c4. White can also relieve the central tension immediately: 9.d5 Qd6 The most common continuation is 10.Nf3, when 10...Bg4 leads straight to our main line below. However, White also has some independent possibilities:

642

10.Be2!? 10.Bd3 e6 11.c4 exd5 12.exd5 was played in Kokeza – Nagy, Senta 2013, when Black missed a powerful resource: 12...b5!N 13.Nf3 bxc4 14.Bxc4 Nd7 15.0-0 Nb6„ White already has to worry about maintaining the balance due to the pressure on the d5-pawn. 10.Qd2 Nd7 11.Nf3 e6 12.dxe6 Qxe6 13.Bd3 Nf6 14.Qe2 Re8 15.Nd2 occurred in S. Ernst – Van Kampen, Wijk aan Zee 2013, when a natural and strong continuation would be: 15...Bd7N 16.0-0 Bc6 17.f3 Qd7³ Black’s pieces are well positioned and his ideas include ...Rad8 and ...Nd5. The text move is a cunning attempt to prepare Nf3 without allowing the knight to be pinned to the queen, but Black can exploit the slightly timid nature of the bishop move by striking at the centre: 10...f5! 11.exf5 Bxf5 12.Nf3 In A. Rasmussen – M. Andersen, Helsingor 2019, Black could have continued the central theme with:

12...e6!N 13.Bc4 b5! 14.Bxb5 exd5 15.0-0 Nd7÷ Black’s pieces are beautifully placed and he certainly does not stand worse. 9...Bg4! Black increases the pressure on the d4-pawn, virtually forcing White to advance it. On occasion, we can exchange the bishop for the enemy knight in order to better control the central dark squares. 10.d5 Capturing space while safeguarding the central pawn is natural and logical. 10.Be2? simply led to the loss of a pawn after 10...cxd4 11.cxd4 Bxf3 12.Bxf3 Bxd4 13.0-0 Nc6µ in Bjorksten – Sarkar, Las Vegas 2015. 10.Be3 loses a tempo with the bishop and thus is hardly a way to fight for the advantage. 10...Qa5 643

11.Qd2 This position was reached in Manush – Abhishek, New Delhi 2007. The position is similar to variation B3 of the next chapter on page 380, but here Black has been given the move ...Bg4 for free! It’s not enough for Black to claim an advantage, but it certainly does his position no harm. A good continuation is:

11...Nc6!?N 12.d5 Rad8 13.Qb2 Bxf3 14.gxf3 Nd4 15.Bg2 c4 16.f4 Nb5 17.e5 f6 18.Qb4 Qxb4 19.cxb4 fxe5 20.fxe5 Bxe5 21.Rxc4 Nd6÷ Black has a good position. 10...Qd6! The queen has a reputation for being a bad blockader, but here we see an exception to the rule. In this instance, it is useful to take control over the e5- and f4-squares, in preparation for Black’s central idea: a blow in the centre with ...f5. The immediate 10...f5?! is well met by 11.Qb3! Kh8 12.Nd2ƒ when White seizes the initiative, as demonstrated in a number of games. 11.Be2 White continues developing while avoiding any damage to his kingside structure, so it is hardly surprising that this is the most common move by far. On occasion, White may also have the interesting tactical possibility of e4-e5. 11.Qd2?! allows White’s kingside to be weakened: 11...Bxf3 12.gxf3 Nd7 13.Be2 f5! 14.0-0 f4 Black threatens to trap the bishop with ...h6, so White has to hit the self-destruct button for his pawn centre: 15.e5 Nxe5 16.Bxf4 Nxf3† 17.Bxf3 Qxf4 18.Qxf4 Rxf4µ In Llupa – Stella, Biella 2015, Black had an extra pawn as well as the more active pieces, especially the rook on f4. A rare continuation is: 11.h3 Bxf3 12.Qxf3 It makes no sense to spoil the pawn structure with 12.gxf3? Nd7 13.Qd2 as played in Koller – 644

Heigerer, Neumarkt am Wallersee 2016, when the thematic 13...f5!N would have given Black an even better version of the previous note: 14.exf5 Rxf5 15.Be2 Raf8 16.Be3 c4–+ White’s position is riddled with pawn weaknesses and Black should be winning with accurate play. 12...Nd7 13.Bb5 I also checked the more relaxed 13.Be2 f5 14.g3 e6 15.c4 Ne5 16.Qg2 exd5 17.exd5 Nf7 18.Be3 Be5 19.0-0 f4ƒ when White has a hard time dealing with Black’s initiative. 13...Ne5 14.Qe3 c4 15.0-0 a6 16.Ba4 Nd3 17.Rb1 b5 18.Bc2 Here we can improve on Black’s play from Altanoch – Buker, corr. 2012, with a highly thematic move:

18...f5!N Black has great counterplay. My main line continues: 19.f4 fxe4 20.Qxe4 Rad8 21.Bxd3 cxd3 22.Qxd3 Qxd5 23.Qxd5† Rxd5 24.Bxe7 Rc8³ Black will restore material equality by picking up the pawn on c3. In the resulting endgame, he has some winning chances due to his superior activity and chances to create a passed pawn on the queenside. Black also has no problems after: 11.Qb3 Bxf3 12.gxf3 b6 White’s two bishops are less significant than his spoiled pawn structure. 13.Be3 White spends a tempo relocating his exposed bishop to a safer home. The overoptimistic 13.f4? was played in Sieciechowicz – Shishkin, Baia Sprie 2010, when Black missed a clear refutation of his opponent’s mistake: 13...h6!N 14.e5 Qc7 15.Bh4 Bxe5! After this combination, White’s position falls apart. 16.fxe5 Qxe5† 17.Be2 Qe4!–+ Black regains the piece with a decisive advantage. An interesting correspondence game continued: 13.h4!? Nd7 14.Bb5 Nf6 15.Kf1 Nh5 16.Re1 Rac8 17.c4 Rc7 18.Qc2 Be5 It is worth paying attention to how Black established control over the f4-square. 645

19.Bh6 Rd8 20.a4 Bf4 21.Qc3 e5 22.Bxf4 Nxf4³ Black’s unassailable knight was clearly stronger than the enemy bishop in Nyvlt – Gburek, corr. 2012. 13...f5! 14.h4 Nd7 15.h5 Ne5 16.Be2 f4 Again Black has a powerful clamp on the dark squares, although White may try the witty countermeasure:

17.Bd4!? cxd4 18.cxd4 Rfc8! You should not cling to the extra material, but rather activate your pieces. 19.dxe5 Bxe5 20.Rc2 Rxc2 21.Qxc2 Qc5 22.Qxc5 bxc5 23.hxg6 hxg6³ The resulting endgame was not so simple in Neves – Stone, corr. 2012. Despite the equal material and opposite-coloured bishops, the presence of rooks along with certain positional advantages enable Black to press for a win without risk. Such endgames were masterfully played by Anatoly Karpov.

646

11...Nd7! It is important to establish full control over the e5-square before doing anything rash. 11...f5? was a mistake in Pitterson – Elsness, Khanty-Mansiysk (ol) 2010, in view of the following rebuttal: 12.e5!N 12...Bxe5 13.Nxe5 Qxe5 14.f3 f4 15.h4 Bf5 16.Qd2 e6 17.d6± White threatens to push his d-pawn even further, as well as simply taking on f4. 12.0-0 This is the usual continuation, simply finishing development. 12.Nd2?! is well met by 12...Bxe2 13.Qxe2 Rfe8 when problems are in store for White. A good example continued: 14.Qb5 e6 15.Nc4 Qc7 16.d6 Qc8 17.a4 b6 18.a5 a6 19.Qb1 b5 20.Nb6 Nxb6 21.axb6 Qd7 22.Qd3 Reb8 23.0-0 Be5 24.Be7 Rxb6 25.Rfd1 a5µ Jenkinson – Fenwick, corr. 2008. White has no real compensation for the pawn, as the passer on d6 is firmly blocked. A few high-level games have continued: 12.h3 Bxf3 13.Bxf3 b5 14.0-0 (14.c4 was played in Nyback – Svidler, Khanty-Mansiysk 2009, when 14...bxc4!N 15.Rxc4 Bd4 16.0-0 Ne5 17.Ra4 f5„ would have brought about the collapse of White’s pawn centre, forcing him to switch to defence) 14...c4 15.Qd2 Nc5 16.Qe3 In Jorgensen – P.H. Nielsen, Horsens 2013, it would have been logical to plunge the knight into the heart of White’s position:

16...Nd3!N 17.Rb1 a6 18.Rfd1 Bxc3 19.Be2 Rfc8 20.Bxd3 cxd3 21.Qxd3 b4³ The transformation has resulted in a slight edge for Black, due to the possibility of creating a distant passed pawn on the queenside. Finally, strengthening the centre with 12.c4!? is not a bad option, although Black has a few reasonable ways to play against it. 12...e6 (12...f5!? 13.exf5 Bxf3!? 14.Bxf3 Bd4 is a dynamic option, but after 15.Bh4 Ne5 16.Bg3 Nxf3† 17.gxf3 Qf6 18.fxg6 hxg6 19.0-0 Qxf3 20.Qc2 Rf5 21.Rce1 the rivals agreed a draw in a balanced position in Koistinen – Muukkonen, corr. 2014) 13.0-0 exd5 14.exd5 Rfe8 647

15.Re1 Bxf3 16.Bxf3 Bd4 17.Qb3 Ne5 18.Be2 f6 19.Bf4 Re7 20.Bf1 Rae8 21.Re2 g5 22.Bg3 f5 23.Bxe5 Bxe5= The position remained equal in Avotins – Cerrato, corr. 2014.

12...f5! Black must undermine the enemy centre before he suffocates from a lack of space. 13.Nh4! White needs to produce this precise move to avoid falling into a worse position. I encountered 13.Bh4?! Bxf3 14.Bxf3 in ‘Witik’ – ‘Kovalchuk94’, Internet (blitz) 2019, when I missed an opportunity to win material:

648

14...Bh6!N The primary threat is ...g5, trapping the bishop. White has nothing better than 15.exf5 Bxc1 16.Qxc1 Ne5 17.Be2 gxf5µ when his compensation for the exchange is clearly insufficient. 13.Nd2?! also cannot be recommended for White. 13...Bxe2 14.Qxe2 was seen in Reis – Pawelzik, Bavaria 2016, when Black could have advantageously forced matters with:

14...h6!N 15.Nc4 Qa6 16.Bxe7 Ne5! 17.Bxf8 Rxf8 18.exf5 Rxf5 19.Rfe1 Qxc4 20.Qxc4 Nxc4µ Black’s excellent minor pieces are clearly stronger than White’s rook and pawn, especially as the c3and d5-pawns are weak. 13.Be3?! allows Black to establish a typical bind on the central dark squares: 13...f4 14.Bd2 Bxf3 15.Bxf3 a6 16.Rb1 b5 17.a4 Rab8 18.axb5 axb5

649

19.Bg4 Ne5 20.Be6† Kh8 21.Ra1 Nc4µ White’s position was unpleasant with the bishops unable to show their strength in M. Rudolf – Laghetti, email 2011. Finally, a simple exchange on f5 gives Black a lot of activity: 13.exf5?! Rxf5 14.Bh4 (14.Be3 was played in Cheng – A. Smirnov, Melbourne 2013, when 14...Ne5!N 15.Nxe5 Bxe2 16.Qxe2 Bxe5 17.f4 Bg7 18.c4 e6 19.dxe6 Qxe6³ would have favoured Black: his pieces are well centralized and ...Re8 will come next) Now in I. Rivera – Videnova, Tromso (ol) 2014, it would have been good to continue:

14...Ne5!N 15.Bg3 Bxf3 16.gxf3 Raf8 17.c4 Bh6 18.Rc3 Bf4³ White is under some pressure, as her kingside structure is broken and the light-squared bishop is passive. 13...Bxe2 14.Qxe2 fxe4 This is the only move to have been tested from the present position. The following alternative looks equally playable: 14...f4!?N 15.g3 fxg3 16.hxg3 e6 17.Rfd1

650

17...exd5 Trading the central pawns is the simplest way to maintain the balance. Players who yearn for a more complicated game may prefer 17...Rae8!? 18.Bf4 Qe7 19.d6 Qf6÷ when White’s passed pawn is securely blocked and the position is rather challenging for both sides. 18.Rxd5 Qe6 19.Rcd1 Nb6 20.Rd6 Qc4 21.Qxc4† Nxc4 22.Rd7 b5 23.Rb7 Nb6 24.Be3 Rfb8 25.Rxb8† Rxb8 26.Bxc5 Bxc3= The endgame is equal, although there are enough imbalances to ensure that any result would still be possible over the board. 15.Qxe4 Nf6 16.Bxf6 White has to exchange his bishop so as not to lose the pawn on d5. 16...Bxf6 17.Nf3 b5 Now White has to reckon with the possibility of further queenside advances, leading to the eventual creation of a distant passed pawn.

651

18.c4 This move is rather double-edged: White removes his pawn from the firing line of the bishop, but the scope of the bishop increases and the possibility of a passed a- or b-pawn becomes more real. I also considered 18.a4N 18...a6 19.Rfd1 Rfc8 20.axb5 axb5 21.h4 b4 22.cxb4 cxb4 23.Rxc8† Rxc8 24.Rb1 Bc3ƒ when the passed pawn is certainly an asset, and White needs to play precisely to avoid becoming worse. 18...b4 Black now has an obvious plan of ...a5-a4 before deciding how to advance further. 19.Rce1 a5 20.g3 a4 21.Qc2 Qd7 Black anticipates a possible Re6 attack, and prepares to transfer the queen to b7 to support ...b3. Another tempting continuation is: 21...Rfb8!? 22.Re6 Other moves lead to problems for White: 22.Re3?! Qd7 23.Ne5 Qf5µ and White will be hard pressed to stop the queenside pawns. 22.Nd2?! Qd8 23.Re6 Qf8! 24.Kg2 Bg7 25.Rfe1 Qf5ƒ is also difficult for White: if 26.Qxf5 gxf5 27.Rxe7 b3 28.axb3 a3–+ the a-pawn cannot be stopped. 22...Qc7 23.Nd2 Rb6 24.Rfe1 Rxe6 25.Rxe6 Qb7 26.Ne4 Bd4 27.Ng5 Bf6 28.Ne4 Bd4 29.Ng5 Bf6= Black must settle for a repetition, as 29...b3? allows 30.Rxg6† with a decisive attack.

652

22.Ne5! White must create counterplay before the queenside pawns continue their march. The text move prepares any of Nc6, Nxg6 or Ng4. 22.Re6 is less accurate in view of 22...Qb7 23.Rb1 Ra6 24.Re3 Qc8 25.Ne5 e6ƒ when White has some problems. 22...Qb7 Black continues his plan and threatens ...b3. 23.Nc6? Planting the knight on c6 appears tempting but is actually a serious mistake which allows the b-pawn to advance further. White defended better in an earlier correspondence game: 23.Re3!? Bxe5 I also checked 23...Ra6N 24.Rfe1 Bxe5 (but not 24...Rfa8? due to 25.Ng4‚ with danger on the kingside) 25.Rxe5 b3 26.axb3 axb3 27.Qc3 Qb4 28.Qxb4 cxb4 29.Rxe7 Ra2 30.f4 Rc2 31.d6 b2 32.d7 Rc1 33.Re8 b1=Q 34.Rxf8† Kxf8 35.d8=Q† Kf7 36.Qd5† Kg7 37.Qe5† Kh6 38.Qg5† Kg7= and the game ends with a perpetual check. 24.Rxe5 b3 25.axb3 axb3 26.Qc3 Ra2 27.Qe3 Rfxf2 28.Rxf2 b2 29.Rxe7 b1=Q† 30.Rf1 Qxf1† 31.Kxf1 In Koegler – Canovas Pardomingo, corr. 2009, the rivals agreed to a draw. The finish would have been: 31...Qb1† 32.Qe1 Qd3† 33.Kg1 Qd4† 34.Qe3 Qd1† 35.Qe1 Qd4†= With perpetual check.

653

23...b3! 24.Qb1 24.axb3 axb3 also leaves White in trouble. A sample continuation is 25.Qe4 b2! and now if 26.Nxe7† Qxe7 27.Qxe7 Bxe7 28.Rxe7 Black wins in beautiful style:

28...Rxf2!!–+

24...e6!! A great move, sacrificing a pawn to open the e-file. The deeper point is that, once Black has traded a pair of rooks and advanced his pawn to b2, the absence of a pawn on e7 will make it easier for Black to get his queen to f5, breaking White’s blockade on b1. 25.Rxe6 Rae8 26.Rfe1 654

This way White at least keeps control over the e-file, but the b2-pawn remains the deciding factor. The alternative is 26.Rxe8 Rxe8 27.axb3 Qxb3 28.Qc1 a3 29.d6 Qd3 30.Rd1 Qf3 31.d7 Rf8 32.d8=Q Bxd8 33.Nxd8 Qxf2† 34.Kh1 a2 when the passed pawn proves to be stronger than a knight:

35.Ne6 Qf3† 36.Kg1 Rb8–+ White is powerless against the terrible threat of ...Qxd1† followed by ...Rb1, or simply ...Rb1 immediately. 26...Rxe6 27.Rxe6 b2 28.f4 a3 29.Kg2 White has no counterplay and the superiority of the bishop over the knight is evident. White has an extra pawn and a protected passer on d5, but the pawn cannot advance as it must stay defending the knight. 29.Ne5 is a natural try to reactivate the knight, but it does not save the game: 29...Qb4 30.Nd7 Bd4† 31.Kg2 Qxc4 32.Nxf8 Qxd5† 33.Kh3 Kxf8 The queenside pawns decide matters, for instance:

655

34.Re4 c4 35.Qe1 c3 36.Re8† Kg7 37.Qe7† Kh6–+ White has no meaningful threats against the king, while the queenside pawns remain unstoppable. 29...Qf7! The queen is heading for f5, as discussed previously. 30.Qe4 Bd8! Black continues his plan in the most accurate way, maintaining control of the e7-square in order to prevent a knight fork. In the event of 30...Bc3? 31.h4 Kh8 32.Qd3µ White still has chances to resist. 31.Qb1 31.Nxd8 allows the blunt 31...Qf5! 32.Qxf5 gxf5 33.Rb6 Rxd8 when White has no counterplay against the b-pawn. For instance, 34.Kf2 Re8 35.d6 Rd8 36.Ke3 Rxd6–+ and White can resign. 31...Qf5 32.Re1

656

32...Kg7 Black has time to improve his king. 32...Qxb1 33.Rxb1 Re8 also wins after accurate play. 33.Qxf5 gxf5 34.Rb1 Re8 35.d6 The passed pawn is White’s last hope, but Black is well placed to deal with it.

35...Re6! 36.Na7 36.Nxd8 Rxd6 37.Nb7 Rd2† 38.Kf3 Rc2–+ leaves White with no answer against ...Rc1.

657

36...Rxd6 37.Nb5 Rd2† 38.Kh3 Rd3 39.Nxa3 Equally hopeless is 39.Re1 Ba5 40.Rb1 Bb4 41.Kh4 Kg6 42.h3 Rd2–+ followed by ...Rc2-c1. 39...Rxa3 40.Rxb2 Ra4 41.Rd2 Be7 0–1 C) 7.Qa4†!?

This is an interesting continuation, whose goal is to break the harmony of the black pieces. The queen on a4 is placed flexibly, freeing the d1-square for the rook, while being ready to drop back to a3 or b3 if attacked. A positive feature of this variation is that it delays Black’s attacking plans against the d4point. The move was first played by Mikhail Gurevich, against Ivan Sokolov at Reggio Emilia 1988. It was soon tried by some other strong players, including Kasparov – a clear indication of quality. Among modern GMs, Carlsen, Nakamura, Morozevich and Moiseenko are among those who have tried it. 7...Qd7! I consider 7...c6 and 7...Nc6 to be unsatisfactory; and although 7...Bd7 and 7...Nd7 have been tested by some leading players, the text move is my favourite. The queen may appear clumsy on d7, but this is an illusion. If White exchanges queens, Black’s position will be perfectly acceptable; and if the queen retreats, Black can develop his queenside harmoniously with ...b6, ...Bb7, ...c5 and ...Nc6. We will focus on three main options: C1) 8.Bb5!?, C2) 8.Qa3 and C3) 8.Qb3. Trading queens leads to a different type of game, where Black should have no great problems: 8.Qxd7† Nxd7 9.Nf3 9.Be3!? 0-0 10.Rc1 (also in the event of 10.Bc4, Marchand – Gaborit, Pau 1989, I like 10...c6!?N 11.Nf3 b5 12.Be2 Nc5„) was played in Riazantsev – Mamedyarov, Sochi 2017, and here I propose an 658

interesting novelty: 10...c6!?N Black wants to expand with ...b5 and capture the c4-square. 11.Nf3 b5 12.Bd3 Nb6 13.h4 Be6 14.Rc2 h5 15.0-0 Bc4÷ Black has no problems.

9...b6! Preparing ...Bb7, with pressure on the centre. 10.Bd3 10.Be3 Bb7 11.e5 e6 12.h4 f6 13.exf6 Nxf6 14.Ne5 0-0÷ reached an interesting situation in Shengelia – Erdos, Austria 2013. Each side has its pawn weaknesses but Black has little to fear: in the future he can play ...Nd5 and ...c5, with an active game. The text move pre-emptively defends e4 but allows Black to get active in a different way: 10...Nc5!? 11.Bc2 Ba6 12.Be3 Nd3† 13.Kd2 Nb2 14.Kc1 Nc4 15.Ba4† Kf8 16.Bc6 In Khenkin – Jaracz, Andorra 2001, the logical continuation would have been: 16...Rd8÷ With a double-edged game. Each side has certain difficulties with the location of the kings. Black has a strong knight on c4, and White has a fine bishop on c6, which prevents counterplay with ...c5 – although Black can remove the blockade with ...Na5 whenever he wishes. C1) 8.Bb5!? This move has been tried numerous times at a high level. White is determined to lure our pawn to c6. 8...c6 9.Be2 Another continuation deserving attention is: 9.Bd3!? 0-0 10.Ne2 10.Nf3 c5 11.Qxd7 occurred in Mamedyarov – Vachier-Lagrave, Saint Louis 2018. Usually in such positions it is preferable to take back with the knight, and 11...Nxd7N seems natural here too. Play may 659

continue 12.Be3 b6 13.0-0 Bb7 14.Rac1 Nf6 15.e5 Nd5 16.Bd2 Nc7 17.Bg5 Rfe8 18.Rfd1 Rac8ƒ when Black has good pressure against the enemy centre. 10...c5 11.Qxd7 Nxd7 12.Be3 Rd8 13.Rd1 This position was reached in Anton Guijarro – Gabuzyan, Martuni 2015. I propose the following novelty:

13...cxd4N 14.cxd4 e5!? 15.Bc4!? I also considered 15.d5 Bf8 16.Nc3 Nc5 17.Bc2 Bd7÷ when the d5-pawn is securely blocked and Black has good prospects on the queenside. 15...exd4 16.Nxd4 Ne5 17.Bd5 Ng4 18.Bg5 Nf6 19.f3 Rd7 20.Bxf6 Bxf6 21.Ke2 Rb8 22.Rd3 Rc7 23.Nb5 Rc2† 24.Rd2 Rxd2† 25.Kxd2 Be6 26.Bxe6 fxe6 27.Ke2 a6 28.Nd6 b5= Objectively the position remains equal, but in a practical game White would still have to be slightly careful, as Black’s bishop can support the movement of the queenside pawns.

660

9...0-0 10.Nf3 10.Qa3 enables White to hinder ...c5, but not for long. 10...b6 11.Nf3 c5 12.0-0 cxd4 13.cxd4 Bb7 14.Qe3 Nc6 15.Bb2 Rac8 16.Rab1 e6 17.Rfd1 Na5 18.d5 occurred in Bartholomew – Ding Liren, Internet (rapid) 2019, and here I suggest:

18...Bxb2N 19.Rxb2 exd5 20.exd5 Rc5÷ White has to worry about defending the d5-pawn. 10...c5 11.Qxd7 There are no problems after 11.Bb5 Qc7 12.0-0, as in Gonzalez Velez – Fluvia Frigola, Montcada 1999, when I would suggest: 12...a6N 13.Be2 cxd4 14.cxd4 Qc6 15.Qxc6 Nxc6= Black has a fine position. 661

11...Nxd7 12.Be3 We have been following S. Singh – Khusenkhojaev, New Delhi 2019. Black can interpret the position in various ways but the one I like the most is:

12...Rd8N 13.0-0 Nf6 14.e5 Nd5= As usual in such positions, the knight is excellent on d5 and Black has no problems. C2) 8.Qa3 This move is not wildly different from variation C3 below. 8...0-0 9.Nf3 b6 10.Bd3?! This has been the most popular choice but the bishop is slightly vulnerable on this square. 10.Bg5?! is also not too impressive. In Szeberenyi – l’Ami, Hengelo 2002 10...c5N would have been natural and strong. For instance, 11.Rd1 Qg4 12.e5 Nc6 13.h3 Qe4† 14.Be2 cxd4 15.cxd4 Bb7 16.Qb2 Rfd8 17.Be3 Rac8 18.0-0 Na5³ when Black has the better pawn structure and all his pieces are active. 10.Be3 Bb7 11.Bd3 seems like a slight improvement although Black has little to fear. 11...c5 12.0-0 e6 13.Rfd1 Rc8 14.Rab1 was seen in Balokas – Kovchan, Paleochora 2015, and here I recommend:

662

14...cxd4N 15.cxd4 Nc6÷ There arises a typical Grünfeld middlegame of dynamic equilibrium. A sample continuation is 16.d5 exd5 17.exd5 Ne5 18.Nxe5 Bxe5 19.Bb5 Qf5÷ when Black has a rich game: the c-line is in his hands and the d5-pawn may become weak, so White must play accurately to avoid being worse. 10...c5! 11.0-0 Nc6 Thanks to White’s 10th move, Black was able to generate intense pressure against d4. 12.Bb5 12.Bg5?! cxd4 13.Nxd4 occurred in Lalith – Gupta, New Delhi 2015, when 13...e5!N 14.Nxc6 Qxc6 15.Rfc1 Be6 16.Qb4 Rfc8³ would have left White suffering from his inferior queenside structure. 12.d5?! also leads to problems. In Gonzalez Velez – Vidarte Morales, Montcada 1997, it was logical to hit the centre with 12...e6!N, and after 13.Rd1 exd5 14.Bb5 a6 15.Rb1 Bb7 16.Bxc6 Bxc6 17.Rxb6 d4 18.Qxc5 Ba4 19.Rxd4 Bxd4 20.Nxd4 Rfc8 21.Qb4 Qc7µ White does not have enough compensation for the exchange. 12...a6 13.Bg5 Bb7 14.Bxc6 We are following Sarana – Xiong, Saint Louis 2017. This is a good moment to improve Black’s play by means of:

663

14...Qxc6N 15.Rfe1 Rfe8 16.Rad1 cxd4 17.cxd4 b5ƒ Thanks to the two bishops, Black retains some chances to fight for more than equality. C3) 8.Qb3 This has been the most popular choice. White hopes to prove that the queen on d7 is misplaced, but the following lines indicate that this is not the case at all. 8...0-0 9.Be3 A slightly different type of game ensues after: 9.Nf3 b6 10.Bb5!? Only this move has an independent meaning. 10.Be3 leads straight to our main line below, and 10.Bd3 c5 11.Be3 Nc6 12.Rd1 cxd4 13.cxd4 e6 14.0-0 Bb7 gets there after a few additional moves. 10...c6 11.Bc4 c5 12.d5 12.Ne5?! proves to be too aggressive: 12...Bxe5 13.dxe5 Nc6 14.Bd5 Nxe5! 15.Bh6 (15.Bxa8? is refuted by 15...Ba6 16.Bh6 Qg4–+ with a decisive attack) 15...c4 16.Qc2 Bb7 17.Bxb7 Nd3† 18.Kf1 Qxb7 19.Bxf8 Rxf8µ In Elsness – Hammer, Lillehammer 2013, Black had more than enough compensation for the exchange, with White’s misplaced king making it difficult for him to coordinate his forces. 12...Ba6

664

13.0-0 A strong GM went badly astray with 13.a4? Qg4 14.0-0 Qxe4 15.Bxa6 Nxa6 16.Re1 Qd3 17.Re3 in Volkov – Turov, Sochi 2004, when 17...c4!N 18.Rxd3 cxb3 19.Nd4 Nc5 20.Re3 Bf6 21.Nc6 e6 22.dxe6 Nxe6–+ would have left White a pawn down with the inferior structure, due to the weakness on c3. 13...Bxc4 14.Qxc4 b5 15.Qb3 In Michalik – Bulski, Slovakia 2017, Black could have obtained great dynamic play by energetically advancing the queenside pawns as follows:

15...c4!N 16.Qc2 b4 17.Bb2 Rc8 18.cxb4 c3 19.Bc1 a5! Opening more lines. 20.Rb1 665

In the event of 20.bxa5 Rxa5 21.Rb1 Qa4ƒ the a2-pawn is lost and the far advanced c3-pawn will cause White some problems. 20...axb4 21.Rxb4 e6 22.Qb3 Na6 23.Rb7 Qe8 24.Be3 exd5 25.exd5 Nc5 26.Bxc5 Rxc5 27.d6 Qe6ƒ Despite the fact that White has an extra pawn, he is on the defensive due to Black’s far-advanced passed pawn, which enjoys the support of the g7-bishop and the rest of Black’s pieces. 9...b6 10.Nf3 Bb7 11.Bd3 c5 12.0-0 12.Rd1 cxd4 13.cxd4 Nc6 14.0-0 transposes. 12...Nc6 Black has made all the natural developing moves, and now White must decide how to defend the d4pawn.

13.Rad1 White indirectly protects the pawn while placing our queen in an X-ray. White should not repeat 13.d5 Na5 14.Qa3, as played in Den Dunnen – Waardenburg, Rotterdam 2016, in view of 14...e6!N when Black seizes the initiative because defending the d5-pawn is not convenient. 13.Bb5!? Qc7 14.Qa3 was played by a 7.Qa4† expert in Moiseenko – Piorun, Warsaw 2013. My improvement is:

666

14...Na5!N 15.Nd2 cxd4 16.cxd4 Qd8 17.Qb2 Rc8 18.Rac1 Qd6÷ Black has no problems and can look to the future with confidence. 13...cxd4 14.cxd4 e6 Black anticipates the d4-d5 advance. If White plays it too soon, he will be left with an isolated and weak pawn. 15.Bb5 Qd6 White’s task is to prepare d4-d5, while Black’s immediate plan is to take control of the c-file.

16.Bc1 667

Threatening to win the exchange with Ba3. There have been other continuations, but they cannot pose serious problems to Black. 16.e5?! Qb4 17.Qd3 is not really in the spirit of the position. In Czerw – Tomaszewski, Karpacz 2015, it would have been logical to continue with 17...a6N 18.Rb1 Qe7 19.Ba4 b5 20.Bd1 Rad8³ when Black has unpleasant pressure on the d-file. 16.Rfe1 Rac8 17.d5 exd5 18.Bxc6 Bxc6 19.Bd4 Bxd4 20.Rxd4© occurred in Lalic – Gopal, Cappellela-Grande 2011. White has compensation for the pawn, but after 20...Rfe8N he must be careful not to become worse. 16...Rfd8 17.Bg5 It would be risky to try 17.d5N 17...Na5 18.Qa4 exd5 19.e5 Qc7 20.Bg5 Rf8 when White has a certain amount of compensation for the pawn but it’s not so easy to prove, so I would prefer Black’s side. The time has come for Black to improve on a high-level game.

17...Rdc8!N 17...Nxd4!? 18.Bxd8 Rxd8 was the continuation of Ragger – Vachier-Lagrave, Germany 2013. Black’s exchange sacrifice was interesting, but not at all necessary. True, Black gets a pawn plus control over the dark squares, but one should not underestimate the power of rooks when they have open lines at their disposal. I believe the text move enables Black to press for a slight advantage, rather than just adequate compensation. The critical continuation is: 18.d5 exd5 19.Qxd5 Qxd5 20.exd5 Na5 21.d6 Bxf3 22.d7 668

This is a better try than 22.gxf3?! Rc5 23.f4 Rxb5 24.d7 Rxg5† 25.fxg5 Nb7 26.Rfe1 Kf8µ when Black’s minor pieces are stronger than the rook, and the d7-pawn is less of a danger than it may appear at first glance. 22...Bxd1 23.dxc8=Q† Rxc8 24.Rxd1³ Thanks to the two bishops, White retains good chances for a draw, but there is no question that Black is the one playing for a win with his extra pawn. D) 7.Bb5†!?

Rather like the previous variation, this check is an attempt to break Black’s harmony. If Black blocks the check with either his knight or bishop, the piece in question may prove to be suboptimally placed. That is why I prefer to block with the pawn, and I will say more about this below. The text move was first used by the fourth Women’s World Champion, Olga Rubtsova, against Hermanova in Moscow 1950. However, it seems that no one paid attention to it until 1961, when Bronstein, Rabar and Blom all played this way and it gradually became recognised as a serious move. In more modern times, Topalov, Eljanov, Kramnik, Gelfand, Beliavsky and Ivanchuk have all tried it. Still, much like the other moves examined in this chapter, it is something or a rarity and is mostly used as a surprise weapon. 7...c6! As mentioned above, I prefer to block the check with the pawn rather than committing a minor piece. You might expect Black to follow up by castling followed by typical Grünfeld play with ...c5, and this is certainly a possibility. However, there is another tempting option: if White retreats the bishop to a4, as happens most commonly, then Black often does well to chase the bishop again with ...b5 and possibly ...a5 to gain space on the queenside with gain of tempo.

669

8.Ba4 8.Bc4 was Rubtsova’s choice but instead of 8...Nd7, as played by Hermanova, Black can play 8...c5 with a simple transposition to Chapter 15, with one more move played. The text move is the more consistent follow-up to White’s 7th, but the bishop on a4 is not located in the best way, as evidenced by the number of lines below where Black obtains fine counterplay with ...b5. 8...0-0 9.Ne2 White develops the horse in such a way as to prevent the ...Bg4 pin, while also guarding the c3-pawn against moves such as ...Qa5. On the other hand, the knight is slightly less active on e2 than f3. The overoptimistic 9.h4?! was played in D. Toth – Kladiva, Eger 2002, when 9...c5N would have been logical, and after 10.Ne2 cxd4 11.cxd4 Nc6 12.Be3 Qa5† 13.Kf1 Rd8³ the pressure on the d4-point is significant, while White’s attacking chances are not so great. Surprisingly, 9.Ba3?! has been tried by a few strong players, including in Dao Thien Hai – Peng Xiaomin, Kolkata 2000. Nobody played the most accurate 9...Qa5!N, when 10.Bb4 Qc7 leaves White’s bishop on a more exposed square, and after the further 11.Ne2 a5 12.Ba3 b5 13.Bb3 Rd8 14.0-0 e5ƒ Black has excellent prospects. 9.Nf3 is possible although it allows Black to pin the knight whenever he wishes. I suggest: 9...Qa5 10.Bd2 (10.0-0?! Qxc3 11.Bg5 Bg4 12.Bxe7 Re8 13.Bc5 occurred in Tsiganova – Richtrova, Germany 1994, when 13...Bxf3N 14.Qxf3 Qxf3 15.gxf3 Nd7 16.Rad1 Nxc5 17.dxc5 Rad8µ would have left White in an unpleasant situation due to her rotten pawn structure) 10...Bg4 11.0-0 This occurred in Dehaybe – M. Hansen, email 2009, and here I would suggest:

11...Nd7N 12.Rb1 Rab8 13.Bb3 c5÷ Black has an excellent position. 670

White can hardly benefit from delaying the development of the royal knight. 9.Be3 b5 10.Bb3 was played in Malinin – Osoblivi, Novgorod 1999, when I recommend the thematic follow-up to Black’s last move:

10...a5!N Intending to capture even more space on the queenside. 11.a4 (11.Nf3 a4 12.Bc2 Qa5 13.Bd2 c5 14.0-0 Bg4 15.e5 Nc6 16.Be4 Rac8³ gives Black plenty of space as well as pressure against the enemy centre) 11...Nd7 12.Ne2 Ba6 13.axb5 cxb5 14.0-0 Nb6

Black has a promising game, since the central pawns are not going anywhere and Black has good prospects along the c-file and on the queenside in general. An important point is that 15.Rxa5?! allows 15...Na4 16.Rxa6 Rxa6³ when White does not have enough compensation for the exchange.

671

9...b5! Black postpones the characteristic Grünfeld plan of ...c5, instead taking the opportunity to gain queenside space with gain of tempo. The ...c5 break may or may not be played later. 10.Bb3 This is the best and most active choice. 10.Bc2?! was played in Ivanchuk – Kamsky, Dos Hermanas 1996, when 10...b4!N would been an excellent pawn sacrifice to destroy White’s central fortifications. (The game continuation of 10...c5 was not bad at all, but the text move is more energetic.) Play may continue: 11.cxb4 (if White declines the sacrifice with 11.Bb2, then 11...c5 12.cxb4 Qb6! 13.Rb1 cxb4 14.0-0 a5ƒ gives Black excellent play with ...Rd8, ...Ba6 and ...Nc6 on the agenda) 11...a5 12.bxa5 c5„ Black can follow up with ...Nc6 with further pressure against d4, while the pawns on a2 and a5 remain weak. 10...a5 Black continues his plan of queenside expansion and forces White to consider the prospect of ...a4, driving the bishop to a worse position. He can either prevent this by playing D1) 11.a4 himself, or ignore it, most commonly with D2) 11.0-0. 11.h4?! starts an ill-prepared attack. 11...a4 12.Bc2 c5 13.Rb1 (13.h5 looks consistent but 13...cxd4 14.cxd4 Nc6 15.Be3 Qa5† 16.Kf1 Rd8 puts the d4-point under heavy fire) 13...b4 14.d5 bxc3 15.h5 e6 16.Be3 Na6µ In Lenderman – Wagner, Gibraltar 2015, White’s strategy had backfired. There is not much of an attack on the kingside, while Black already has an extra pawn and threats in the centre. 11.a3 enables White to keep the bishop on the a2-g8 diagonal, but costs additional time. In Gleizerov – Sasikiran, Dubai 2002, it would have been good to play: 11...a4N 12.Ba2 c5 13.Rb1 Ba6 14.Be3 c4!? This move is certainly not forced and it has some drawbacks, but I think these are outweighed by the 672

positives of blocking the a2-bishop and securing the b3-outpost. 15.0-0 Nc6³ Black’s plans include ...e5, ...Na5-b3 and ...Qd6, with comfortable play. 11.Be3 was once played by Pavel Eljanov, a 7.Bb5† expert, though only in a blitz game. Play continued 11...Nd7 12.0-0 a4 13.Bc2 Nb6 14.Qc1 Be6 15.Re1 Bc4 16.Bh6 and now in Eljanov – Salgado Lopez, Dubai (blitz) 2014, the most accurate continuation would have been:

16...Qc7!N The queen is optimally placed here, as she supports both of the key pawn breaks, namely ...c5 and ...e5. Play may continue 17.Bxg7 Kxg7 18.Qe3 f6 19.h4 e5³ when I like Black’s position: he has a fine outpost on c4 and his pieces are well positioned, whereas White’s minor pieces have limited scope. D1) 11.a4 This looks like a natural response to Black’s last move, but White will not be able to keep his centre together. 11...b4 12.cxb4 12.Be3 has also been tried by a few strong players. Black should continue 12...bxc3 13.Rc1 Nd7 14.Rxc3 c5 15.0-0 Ba6 16.Bc4 cxd4 17.Bxd4 as in Yakovich – Brinck-Claussen, Koge 1997, at which point it was possible to seize the initiative with:

673

17...Bxd4!N 18.Qxd4 Bxc4 19.Qxc4 Rb8 20.Rd1 Ne5 21.Qd4 Qxd4 22.Rxd4 Rb1† 23.Nc1 g5³ Black’s active rook and well-placed knight enable him to press for more than a draw. 12...axb4 13.Be3 c5 14.Rc1 cxd4 15.Nxd4 This position was reached in Svidler – Grischuk, Sochi 2015. Black’s game can be improved by activating the queen’s rook in an unusual way:

15...Ra6!N Threatening to win a piece with ...Rd6. 16.Nb5 674

White has nothing better. For instance, 16.Rc2 is well met by: 16...Bxd4! 17.Bxd4 (or 17.Qxd4 Qxd4 18.Bxd4 Nc6 19.Bc5 Be6µ when Black removes the defending bishop and gets to the weak pawn on a4) 17...Be6 18.Bxe6 Rxe6 White must jettison the e4-pawn and fight for a draw, as 19.f3?! leads to even greater problems after 19...Rd6 20.Rd2 Nc6 21.Be3 Rxd2 22.Qxd2 Qa5–+ when White loses the a4-pawn for no compensation. 16...Be6 17.Bxe6 fxe6

18.Qe2 Qa5 19.0-0 Qxa4 20.Nc7 Rc6 21.Qg4 Rxc1 22.Bxc1 Qd7 23.Nxe6 Nc6ƒ Due to the powerful passed pawn on b4, Black keeps some initiative. D2) 11.0-0 GAME 6

Vadim Afanasiev – Viacheslav Zakhartsov Split 2013 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bb5† c6 8.Ba4 0-0 9.Ne2 b5 10.Bb3 a5 11.0-0 Castling is the most natural and popular choice. 11...Nd7! 675

The hasty 11...a4?! commits Black to a rigid queenside structure sooner than I would like. 12.Bc2 Nd7 (12...e5 13.Be3 Nd7 14.Bd3ƒ leads to a complex game but I slightly prefer White here too) 13.e5!N 13...Nb6 14.Nf4ƒ White has a small initiative. The text move sees Black develop more flexibly. A key position has arisen, where White has tested a number of continuations. 12.Bg5!? This has been the most popular move, especially among stronger players. White develops actively and intends to trade dark-squared bishops. 12.Be3 transposes to the 11.Be3 line in the note on the previous page. The creative 12.h4 cannot pose problems for Black: 12...e5 13.Bg5 Qc7 14.h5 a4 15.Bc2

15...h6 16.Be7 Re8 17.Bb4 g5 18.Rb1 Nf6 19.dxe5 Ng4 20.Bd6 Qd8 21.Nd4 Bxe5 22.Bxe5 Nxe5³ Black’s accurate play yielded some advantage in Nogler – Wegelin, email 2013. Black’s pieces are well placed and White’s bishop is passive on c2. 12.a4 cannot be recommended for White, as after 12...b4 his centre will fall apart. 13.cxb4 axb4 14.Qc2 was seen in Carnic – Nestorovic, Kladovo 1994, when Black could have developed some initiative with:

676

14...Ba6!N Best play seems to be 15.Be3 e5 16.Rfd1 exd4 17.Nxd4 c5 18.Nc6 Qf6 19.Bd5 Rfc8 20.e5 Nxe5 21.Nxe5 Qxe5 22.Bxa8 Rxa8 23.Qxc5 Qxc5 24.Bxc5 Bxa1 25.Rxa1 b3³ Despite the equal material and opposite-coloured bishops, Black has some winning chances due to his advanced passed pawn, which can be supported by the king after ...f5, ...Kf7 and so on. 12.a3!? is a familiar concept to keep the bishop pointing towards Black’s king. 12...a4 13.Ba2 occurred in Dorcic – Tatisic, Belgrade 2010, when it would have been logical to prepare ...c5 by means of:

13...e6!N 14.Re1 c5 15.Be3 Bb7 16.f3 Qe7 17.Qd2 Rfc8÷ Black can follow up with ...cxd4 and ...Nb6-c4, with a good game. 12.Rb1 has been played, although I am not sure why White wanted the rook to lean against the b5677

pawn. 12...a4 13.Bc2 Nb6 14.Bd3 Be6 15.Qc2 Bc4 16.Ba3 Bxd3 17.Qxd3 Nc4 18.Bc5 was seen in Posedaru – Stella, Skopje 2013, when a good move would be:

18...f5!N 19.Rfd1 Qd7 20.e5 Rad8÷ Black does not have a shadow of a problem, and I would choose his side because of the powerful knight on c4. 12.e5!? was an interesting idea in Lauer – Suarez Rodriguez, corr. 2012. I think it would be worth playing:

12...e6N Stopping the pawn from going any further. A logical continuation would be 13.Ng3 Bb7 14.Ne4 c5 15.Bg5 Qc7 16.Nd6 Bc6÷ with a complex game. Each side has its pros and cons, but overall Black has enough resources with ideas of destroying White’s centre with some combination of ...b4, ...a4 and ...f6. 678

A final alternative is the aggressive 12.f4!?, as played in A. Wang – Rohan, Berkeley 2018. Black’s plans should include ...b4 and ...c5, and 12...Ba6!N would have been a useful precursor to both moves. My analysis continues: 13.e5 b4 14.cxb4 axb4 15.Qc2

15...c5! 16.dxc5 Bxe2 17.Qxe2 Nxc5 18.Qc4 Nxb3 19.Qxb3 Qb6† 20.Be3 Qa5= The mutual pawn weaknesses on a2 and b4 are likely to be destroyed, resulting in a drawish endgame.

12...a4 13.Bc2 Nb6 The horse will go to c4, after which Black can prepare ...e5 or ...c5. 14.Qc1 White intends to trade the dark-squared bishops. 679

14.f4!? f6 15.Bh4 was played in Stocek – Kislik, Pardubice 2012, when 15...Nc4N would have been logical, and after 16.Qd3 f5 17.e5 Qc7÷ Black has a comfortable position. 14...f6! Black provokes his opponent’s plan while preparing to take some space in the centre with ...e5. 15.Bh6 e5 16.Bxg7 Kxg7 17.Rd1 In another instructive example, White transferred his queen to the kingside with 17.Qe3 Nc4 18.Qg3, before going on the attack with 18...Qe7 19.f4, but Black was ready with an excellent counter: 19...exd4 20.cxd4

20...f5! White’s pawns are stopped and Black captures the initiative. 21.Qc3 Qa3 22.Qxa3 Nxa3 23.Bd3 b4 24.Ng3 fxe4 25.Bxe4 Ra7 26.f5 (26.Bxc6 is met by 26...Nc2 27.Rab1 Nxd4µ when Black’s queenside pawns give him clearly better prospects) 26...Bxf5 27.Bxf5 gxf5 28.Rf2 f4µ The advanced queenside pawns defined Black’s advantage in Husemann – Ederer, email 2011. 17...Qe7 18.Ng3 Be6 Developing the bishop while aiming at the c4-square and a2-pawn is natural and good. 18...Rd8!?N was an interesting alternative, postponing the development of the bishop, which might go to g4 in some scenarios. A possible continuation is: 19.Bd3 (or 19.f4 exd4 20.cxd4 Nc4µ when Black has a powerful knight and pressure against the centre) 19...a3 Creating a different outpost for the knight. 20.Qe3 Na4 21.Be2 Be6µ The pawns on a2 and c3 are weak, and Black can reorganize his queenside play with ...Nb2-c4 if the situation demands it. 19.f4 680

White tries to develop counterplay on the kingside but this is not a dangerous action: most of his pieces are on the opposite flank, and Black is well placed to create threats in the centre. 19.Bd3N is a sensible alternative, preparing to trade the passive bishop for the knight, but it does not solve all White’s problems. After 19...Rad8 20.Qe3 Nc4 21.Bxc4 Bxc4 22.Nf1 exd4 23.cxd4 Qd6ƒ the bishop dominates the position and White’s knight cannot find a good square. 19...Rad8 20.Qe3 White pulls the queen closer to the kingside. The immediate 20.f5 is well met by 20...Bf7 21.Qe3 Nc4 22.Qf2 Nb2 23.Rd2 g5µ when White has no serious chances on the kingside and the black pieces are perfectly centralized. 20...Nc4 21.Qf2

21...Qa3 Attacking the c3-pawn is good enough, but it was even stronger to play 21...Bg4! 22.Rd3 (if 22.Re1 then 22...Qd6µ makes it hard for White to keep his centre together) 22...Nb2! 23.Rd2 Qa3 24.dxe5 Rxd2 25.Qxd2 Nc4 26.exf6† Rxf6 27.Qf2 Qxc3µ when Black’s pieces are far more active than their counterparts. 22.f5 Bg8 This move keeps some advantage but 22...gxf5! 23.exf5 Bd5 was even stronger. For instance, 24.Bd3 exd4 and Black’s central play is more significant than White’s vague hopes on the kingside. 23.fxg6 It is logical to open the f-file for the heavy pieces and, on occasion, push the h-pawn to weaken the 681

f5-square. An instructive motif occurs after 23.Rd3 g5 24.Nh5† Kh8 25.h4 when Black should sacrifice a pawn to shut down White’s initiative:

25...g4! 26.Qg3 Qe7 27.Qxg4 b4! 28.Qg3 bxc3 29.dxe5 Rxd3 30.Bxd3 Nxe5µ Black has an excellent knight on e5 and dangerous passed pawn on c3. White has no real attack while his pieces are mostly passive: the bishop is blocked by its own pawns while the rook is stuck defending the a2-pawn. 23...hxg6 24.Rd3

24...Nb2? 682

This poor decision squanders all of Black’s advantage by forcing the white rook to move to f3 where it participates in the attack. The correct idea was to reactivate the bishop with 24...Be6! when Black remains in control. For example: 25.h4 (now 25.Rf3 is meaningless because of 25...Bg4) 25...Nb2 26.Re3 (26.Rf3 is still not dangerous due to 26...Bg4 27.h5 Bxf3 28.Qxf3 g5 29.Qf5 Kh8 30.dxe5 Qc5† 31.Kh2 Qxe5 32.Qg6 Rd7–+ and White gets nowhere) 26...Qd6 27.Rf1 exd4 28.cxd4 Qxd4 29.Ne2 Qe5 30.Nf4 Bc4 31.Qg3 Bf7 32.Nxg6 Bxg6 33.h5 Qg5 34.Qxg5 fxg5 35.Rxf8 Kxf8 36.hxg6 Nc4 37.Re2 Ne5–+ The pawns on a2, e4 and g6 are scattered and weak, while Black’s rook and knight are excellent. 25.Rf3 Qd6 Black connects the queen to the defence. 25...Nc4 might have led to a draw after 26.h4 exd4 27.cxd4 Qd6 28.Rd1 Nb2 29.Rb1 Nc4 30.Rd1= with a repetition. 26.Rb1? There was no need for this move. White should have continued his attack with 26.h4! when best play continues: 26...Bh7! Covering the f5-square. (26...exd4? 27.h5 d3 28.hxg6 Be6 29.Nh5† Kxg6 30.Rg3† Kf7 31.Bxd3! Nxd3 32.Rxd3!‚ gives White a decisive attack) 27.h5 g5 28.Rf1 Nc4 Now White has a choice:

29.h6†!? This leads to a confusing game. (The other option is 29.d5 Qe7 30.h6† Kh8 31.Nf5 Qf7 32.Ne3 Nxe3 33.Qxe3 Rd6 34.Qc5 Qd7= when the position is double-edged and approximately equal.) 29...Kh8 30.Rxf6 Qxf6 31.Qxf6† Rxf6 32.Rxf6 exd4 33.cxd4 Ne3 34.Bd3 Rxd4 35.Be2 Bg8 36.Rxc6 Bxa2÷ A sharp endgame has arisen. The computer offers zeros but in a practical game anything could happen. 683

26...a3? White is given a second chance. Why does Black leave the knight on b2, where it is not participating in the game? Black could have reclaimed the initiative by returning the horse to the battle: 26...Nc4! 27.Rd1 27.h4? gives White nothing after 27...exd4 28.h5 Ne5! when the knight makes its presence felt. 29.hxg6 Nxf3† 30.gxf3 f5! Another important move, after which White’s attack comes to a swift end, for instance: 31.Nxf5† Rxf5 32.exf5 d3 33.Bd1 d2–+ 27...exd4 28.cxd4 c5!

29.Ne2 It is dangerous for White to take the pawn: 29.dxc5?! Qc6 30.Rfd3 (30.h4? proves even worse after 30...Ne5 31.Rc3 b4–+ and White loses the exchange) 30...Rxd3 31.Rxd3 Rc8 32.Ne2 Ne5 33.Rd2 Qxc5–+ Black’s pieces are much better positioned and the weakness of the a2-pawn is acutely felt by White. 29...Ne5 30.Rg3 cxd4 31.Nxd4 Qc7 32.Nxb5 Rxd1† 33.Bxd1 Qc1ƒ Black has full compensation for the pawn, and his active queen along with the weak a2-pawn make life difficult for White.

684

27.d5 This move is not terrible, but it would have been more dangerous to continue 27.h4! with similar ideas as on the previous move – and the inclusion of Rb1 and ...a3 makes the b3-square available for White’s bishop, which may help him in some lines. It is worth mentioning that the attempt to exploit the hanging a2-pawn by means of 27...Kh8 28.h5 g5 29.Rf1 Bxa2 is refuted by 30.Nf5 Qd7 31.dxe5 fxe5 32.Qg3+– with a crushing attack. 27...Nc4 Finally, Black returns the knight to the game. I also analysed an interesting alternative: 27...f5!? 28.exf5 Bxd5 29.Ne4 (or 29.Re3 Bxa2 30.Rbe1 Bc4 31.Rxe5 Kg8 32.Re7 Rd7 33.Re8 a2 34.Ne4 Rxe8 35.Nxd6 Rxe1† 36.Qxe1 Rxd6 37.Qe8† Kg7 38.Qe5† Rf6 39.Qe7† Rf7 40.Qe5† Rf6= and White must give perpetual check) 29...Bxe4 30.Bxe4 Qd1† 31.Qf1 gxf5 32.Rg3† Kh6 33.Rh3† Kg7 34.Rg3†= The game ends in perpetual check. 28.h4 cxd5? It was vital to cover the f5-square. 28...Bh7! is an important defensive resource which we already encountered in the note to White’s 26th move above. A possible continuation is 29.h5 g5 30.Rf1 Qe7 31.h6† Kh8 32.Nf5 Bxf5 33.Rxf5 g4 34.Rxf6 Qxf6 35.Qxf6† Rxf6 36.Rxf6 cxd5 37.exd5 Ne3 38.Be4 Nxd5 39.Rc6 g3 40.Kf1 Nf4 41.Ke1 Nd3† 42.Ke2 Nc1† 43.Kf3 Nxa2 44.Rc7 Nc1 45.Rh7† Kg8 46.Rg7† Kh8 47.Rh7† Kg8= when it should be a draw. 29.h5 685

Now Black’s position is completely lost. 29...Bh7 The bishop must go here to protect the f5-square. 29...Bf7? allows White to break through with 30.Rxf6!+– intending 30...Qxf6 31.h6†. The computer recommends 29...Qb6 but after 30.Qxb6 Nxb6 31.hxg6 dxe4 32.Nf5† Kxg6 33.Bxe4 Bxa2 34.Rg3† Kh7 35.Rxb5+– White has a decisive attack, despite the queen trade.

30.Rxb5 d4 31.Rb7† A cleaner road to victory is 31.hxg6 Bxg6 32.Nf5† Bxf5 33.Rg3†! Bg6 34.Rb7† Rd7 35.Rxg6†! Kxg6 36.Qf5† Kh6 37.Rxd7 Qxd7 38.Qxd7+– and Black can give up. 31...Rd7 32.Rxd7†?! This move complicates the win. More accurate is 32.h6†! Kh8 (or 32...Kxh6 33.Nf5† gxf5 34.Rh3† Kg7 35.Rxh7† Kxh7 36.Qxf5†+– when Black gets checkmated or loses his queen) 33.Rxd7 Qxd7 34.Rxf6 Rc8 35.Bb3 dxc3 (35...Qe7 36.Bxc4+–) 36.Rf8† Bg8 37.Rxc8 Qxc8 38.Qf6†+– and mate next move. 32...Qxd7

686

33.h6†? A final error allows the win to slip away. After the accurate 33.Bb3! Ne3 34.hxg6 Bxg6 35.Nf5† Nxf5 36.exf5 Bh7 37.Be6 Qa7 38.Kh1 Kh8 39.Qh4 dxc3 40.Rxc3+– Black’s position does not hold, as White is too active. 33...Kxh6 34.Rxf6 34.Bb3!? was still possible as a winning attempt, but it is less strong than on the previous move and Black should be okay. The text move leads to a relatively simple draw. 34...Rxf6 35.Qxf6 Ne3 36.cxd4 Qxd4 37.Qh4† Kg7 38.Qe7† Kh6 39.Qh4† Kg7 40.Qe7† ½–½ Even though the theoretical relevance of this game ended at move 21 or 22, due to the existence of clear improvements for Black, the later part of the game and accompanying notes contained several instructive points, especially with regard to White’s attacking resources in such positions and the ways in which Black should deal with them. Conclusion This chapter has dealt with a number of second-tier continuations after the opening moves 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7. We started with two moves from the c1-bishop. First there was 7.Ba3, which impedes the ...c5 advancement, but 7...Nd7! is a convenient reply: Black soon plays ...c5 anyway, when White runs the risk of the bishop being left out of play on a3. We then considered Kruppa’s move 7.Bg5, after which 7...c5 8.Rc1! 0-0 9.Nf3 Bg4! 10.d5 Qd6! 11.Be2 Nd7! enables Black to strike at the centre with ...f5 and/or ...e6, with at least equal prospects. 687

We then considered the disruptive 7.Qa4†!?, when 7...Qd7! is our reply, since a queen exchange will offer Black a safe endgame. Keeping the queens on is more ambitious, and 8.Bb5!?, 8.Qa3 and 8.Qb3 all deserve attention, but in general Black will get a harmonious game with moves like ...b6, ...Bb7, ...c5 and ...Nc6, when the queen is rather well placed on d7. Finally we considered 7.Bb5†!?, with the same general idea of luring a piece or pawn to an unfavourable square. However, after 7...c6! 8.Ba4 0-0 9.Ne2 b5! 10.Bb3 a5 we saw how Black could utilize the c6-pawn to launch a space-seizing operation on the queenside, gaining time against the enemy bishop. All four of the main variations in this chapter are relatively rare, but they are not devoid of poison and can be used to drag an unsuspecting victim into unfamiliar territory. The reader should carefully study this chapter to learn how to respond to each option. Knowledge is power, and good opening preparation combined with study of the illustrative games will enable you to find the right plans, not only in the opening but in the middlegame and possibly even the endgame as well.

688

A) 7.Be3 c5 8.Qd2 Qa5 369 A1) 9.Rb1 369 A2) 9.Rc1 371 B) 7.Nf3 c5 8.Be3 Qa5 374 B1) 9.Bd2 374 B2) 9.Nd2 Bd7!? 10.Rb1 Qxc3 376 B21) 11.dxc5 377 B22) 11.d5!? 379 B3) 9.Qd2 0-0 10.Rc1 Nd7! 380 B31) 11.Bh6 384 B32) 11.h4!? 386 B33) 11.d5 388 B34) 11.Bd3 392

689

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 In this chapter we will consider the early development of White’s bishop to e3. This is certainly a natural continuation, as White develops to a good square while reinforcing the d4-point. We will examine two related set-ups: A) 7.Be3 (delaying or omitting Nf3) and B) 7.Nf3 c5 8.Be3. A) 7.Be3 c5 In this section, we will consider those continuations where White refrains from developing the royal knight for a significant period. White’s plans include Qd2, Rc1 and d4-d5, capturing space. Black’s task will be to explode the centre with moves like ...e6 and ...f5. 8.Qd2 Other continuations will transpose: 8.Rc1 Qa5 9.Qd2 transposes to variation A2 below, while 8.Nf3 is variation B of course. 8.Bc4 is well met by 8...Qa5 (8...Nc6 is also possible, when 9.Ne2 leads to Chapter 15) when 9.Qd2 transposes to the 9.Bc4 line in the notes below. 8...Qa5 We reach an important branching point, where White’s main options are A1) 9.Rb1 and A2) 9.Rc1. 9.Nf3 leads to variation B3, while 9.Rd1 Nc6 10.Bc4 0-0 11.Ne2 transposes to the 9.Bc4 line below. 9.Bc4 makes the bishop a target, and after 9...Nc6 10.Ne2 0-0 11.Rd1 Bd7 12.0-0 cxd4 13.cxd4 b5! Black shoots on the flank and seizes space. 14.Bd5 Qxd2 15.Bxd2 was seen in Kirgeridis – Sidiropoulos, Porto Rio 2014, and here I suggest:

690

15...Rfc8N 16.e5 e6 17.Bf3 Rab8³ Black has slightly the more pleasant endgame, due to his prospects of creating a passed pawn on the queenside. A1) 9.Rb1

This move has been used several times by Karpov, along with several other strong GMs. White activates his rook and threatens Rb5, which explains Black’s reply. 9...b6 Here White has tried various possibilities:

691

10.Bb5† White intends a peculiar manifestation of the ‘disappearing tempo’ in the opening. 10.Nf3 0-0 11.Rb5 Qa4 12.Rb3 Ba6 13.Bxa6 Qxa6 14.Qe2 Qa4 15.0-0 Rc8 16.Rfb1 cxd4 17.cxd4 Nc6÷ gave Black a good game in Probstfeld – Boehme, email 2012. 10.Rc1 Losing a tempo to provoke ...b6 seems weird, but it has been tried by some strong players, including the young Kramnik. 10...0-0 11.Nf3 Nd7 12.d5 12.Bd3 Ba6 13.Bxa6 Qxa6 14.Qe2 was played in Daschian – Najer, Moscow 1999, when 14...Qxe2†N 15.Kxe2 b5÷ would have been logical: Black’s pawn majority on the queenside seems more important than White’s solid pawn centre.

12...Bb7 13.Bh6 Such an attacking attempt causes no more damage to Black’s kingside than a pitchfork striking water. 13.c4 Qxd2† 14.Bxd2 e6! 15.Bd3 f5! sees Black thematically undermining the centre. 16.exf5 gxf5 17.dxe6 Rae8 18.0-0 Rxe6 19.Bf4 Bxf3 20.gxf3 Be5 21.Rce1 Rd6 22.Bxe5 Nxe5 23.Rxe5 Rxd3µ Black’s strategy was fully justified. Aron Nimzowitsch would surely have taken off his hat to Black’s play in Uksini – Colpe, Lueneburg 2014. 13...Nf6 14.Bxg7 Kxg7 15.Bd3 In Kozul – Salgado Lopez, Skopje 2014, the most thematic continuation would have been:

692

15...e6!N Attacking the centre before White can finish developing. 16.c4 Qxd2† 17.Nxd2 Nh5÷ With ...f5 on the way, White’s centre is about to crumble. 10.Rb5 Qa4 11.Rb2 White makes another ‘disappearing move’, threatening to land a bishop on b5, but it is easy to deal with. 11...Ba6 12.Bxa6 It is logical to lure the black knight to a suboptimal square, while avoiding any misplacement of the white king. 12.f3?! was played in Volkov – Roiz, Internet 2004. It is not clear why Black avoided 12...Bxf1N 13.Kxf1 Nc6 14.Ne2 Na5ƒ when his position is excellent. 12...Nxa6 13.Ne2 0-0 14.Bh6 Rad8 15.Bxg7 Kxg7 16.0-0 In Murawski – Zezulkin, Poland 1998, a logical continuation would have been:

693

16...Nc7N 17.Qe3 Qa3÷ Many pieces have left the board, and the pressure on White’s centre is felt more and more.

10...Bd7 11.Be2 11.Bd3 Nc6 12.Ne2 0-0 is pleasant enough for Black. 11...Bc6 I don’t think the bishop is any worse on c6 than it would be on a6. Moreover, Black is developing quickly while White spends one tempo after another moving his bishop. 12.Bd3 Nd7 13.Ne2 Rd8 14.f3 0-0 15.h4 h5 16.Bg5 Rfe8 694

Black has successfully grouped his forces and is ready for action in the centre. 17.Rc1 Bb7 Threatening ...Ne5!. Black is already better, and in the following example White quickly collapsed. 18.d5 Ne5 19.Bb1 Nc4 20.Qf4?? Be5–+ Shaked – Kasparov, Tilburg 1997. A2) 9.Rc1

White removes the rook from the X-ray attack of the g7-bishop, while pre-protecting the c3-pawn and thus preparing d4-d5. 9...0-0 10.d5 Removing the tension from the d4-point is a responsible decision which gains space. On the other hand, the g7-bishop becomes even stronger and Black has chances to hit the centre with ...e6 and ...f5. A more difficult game arises after 10.Nf3, which transposes to variation B3 on page 380. 10.Bd3 Rd8 11.d5 e6 12.c4 Qxd2† 13.Kxd2 This position arose, via a slightly different move order, in Drozdova – Tomashevskaya, St Petersburg (rapid) 2018. Black would have done well to play:

695

13...b5!N With great counterplay. 10.Bc4 Nd7 11.Ne2 Rd8 12.0-0 was played in Stojanovic – Mitrovic, Tivat 1995. I recommend:

12...cxd4N It’s time! 13.cxd4 Qxd2 14.Bxd2 Nb6! 15.Bb3 Bxd4 16.Nxd4 Rxd4 17.Be3 Rxe4 Black has picked up two pawns, although White has enough compensation to maintain the balance. 18.Rc7 Be6 It is best for Black to give the material back in order to activate his pieces. 19.Rxe7 Nd5 20.Rxb7 a5 21.Bg5 Rb4 22.Rxb4 axb4 23.h4 Nc3 24.Be7 Bxb3 25.axb3 Ra3 26.Bxb4 Rxb3= The position is absolutely equal. 10...e6 11.c4 11.d6?! was played in Li Chao – Paragua, Istanbul (ol) 2012. White’s last move condemns the pawn 696

to its fate, although Black needs to be accurate to make the most of his chances. The best move is: 11...Qa4!N Attacking the e4-pawn before focusing on d6. 12.f3 (12.Bd3 Rd8³) 12...Rd8 13.Bxc5 b6 14.Be3 Qa3 Black will quickly equalize the material count while claiming a clear advantage due to his better pawn structure and faster development. To a more complicated game leads: 11.Nf3 exd5 12.exd5 Bf5 13.Be2 13.Nh4 was played in Khenkin – Ilincic, Kecskemet 1990, when Black could have utilized a tactical trick: 13...Bb1!N (13...Be4!?N is also possible although 14.c4 keeps the game unclear) 14.a4 Be4 15.f3 (Black’s previous finesse prevents 15.c4? here due to the simple 15...Qxa4) 15...Qd8 16.c4 Qxh4† 17.g3 Qe7 18.fxe4 Qxe4 19.Bg2 Qf5÷ White has just enough compensation for the pawn, thanks to his bishop pair and passed pawn on d5. Nevertheless, in a practical game I would choose Black’s side. 13...Nd7 14.Nh4 Be4 15.f3 In Wang Yue – Zeng Chongsheng, Ningbo 2011, Black opted for a less attractive version of the ...Bb1 trick. Instead he could have obtained an excellent game with:

15...Rae8!N 16.fxe4 Rxe4 17.Nf5! The best try. Certainly 17.Nf3? Rfe8–+ must be avoided. 17.g3 is not ridiculous, but 17...Rfe8 18.Ng2 Rxe3! 19.Nxe3 Bh6 sees Black win back the knight on e3, leaving him with fabulous compensation for the exchange. 17...Rfe8! 18.Nd6 Rxe3 19.Nxe8 Rxe8 20.Rc2! 20.0-0? Bd4† 21.Kh1 Be3µ 20...Be5 21.0-0 Qc7 22.Qh6 c4© Black has excellent compensation for the sacrificed exchange, due to his safer king and pressure on the dark squares. 11...Qxd2† 12.Kxd2 697

12.Bxd2?! seems illogical. 12...exd5 13.cxd5 Re8 14.Bd3 was seen in Riha – Mellal, Griesheim 2006, when 14...b6!N 15.Ne2 Ba6 16.Bxa6 Nxa6 17.f3 f5!µ would have made it hard for White to defend his centre.

12...b6 13.Ne2 13.Be2 occurred in Korennoi – Alekseenko, St Petersburg 2018, when 13...Nd7!N 14.Nf3 Nf6 15.Bd3 Ng4³ would have hunted down the dark-squared bishop, leaving White to suffer in the hope of making a draw. 13.Bd3 f5! sees Black carry out the familiar undermining plan. 14.Ne2 (14.exf5 gxf5 15.Bg5 Bb7 16.Ne2 exd5 17.Nf4 occurred in Mertanen – Arlandi, Plovdiv 2010, when 17...Nd7 18.cxd5 Ne5 19.Rhe1 Rf7³ would have prepared to destroy the last remnants of White’s centre) 14...fxe4 15.Bxe4 Bb7 16.Nf4 exd5 17.Nxd5 In Kanter – Ni Shiqun, Doha 2014, it would have been best to continue with:

698

17...Rf7!N 18.Rhd1 Na6³ With a promising endgame for Black. 13...f5 14.f3 14.exf5 occurred in Tratar – Bensch, Ranshofen 2012, when 14...exd5!N would have led to complications which favour Black with accurate play: 15.fxg6 d4 16.gxh7† Kxh7 17.Bf4 Nc6 18.g3 Bg4 19.Bg2 Rae8 20.Rhe1

20...Nb4! 21.a3 Na2! 22.Rc2 Nc3ƒ 14...fxe4 15.fxe4 Nd7 16.Nf4?! It is already hard to offer White good advice. 699

16.Nc3N may be a better try although 16...Ne5 17.Be2 Ba6 18.Nb5 Bxb5 19.cxb5 exd5 20.exd5 Rfe8µ keeps White under pressure. 16...Nf6 17.Bd3 17.dxe6 is a move with the recognition of disaster in its essence: 17...Ng4 18.Be2 Nxe3 19.e7 Re8 20.Kxe3 Bb7 21.Nd5 Bxd5 22.cxd5 Bh6† 23.Kf3 Rxe7µ White is clearly in trouble. 17...Ng4 18.Rcf1 In Tregubov – Mamedyarov, Ajaccio 2008, Black exchanged on e3 and kept some advantage, which he subsequently converted. However, he could have secured victory more quickly with:

18...Bh6!N 19.dxe6 Bxe6 20.h4 Nxe3 21.Kxe3 Rad8–+ White has no good ideas and it will not be long before Black starts picking up material. B) 7.Nf3 c5 8.Be3

700

It is only in conjunction with the knight on f3 that the early Be3 is at its most potent. White’s concept involves reinforcing the crucial d4-point, followed by either advancing in the centre with Rc1 and d4d5, or keeping his structure more flexible and relying on fast development. White’s set-up was first played by the Australian master Koshnitsky against Crowl in Sydney 1932/1933. A few years later, in 1936 in Nottingham, Vidmar played it against Alekhine. Later, it was taken up by Korchnoi, Miles, Ribli, Gligoric and Ftacnik. White’s system became especially popular after the World Championship matches of 1990 and 2000, in which Karpov and Kramnik used it against Kasparov. Other modern grandmasters who play it include Mamedyarov, Aronian, Grischuk and Gelfand. Moreover, it is recommended by Alexey Kornev in A Practical White Repertoire with 1.d4 and 2.c4. 8...Qa5 Attacking the c3-pawn immediately is the main line. White may respond with B1) 9.Bd2, B2) 9.Nd2 or B3) 9.Qd2. B1) 9.Bd2 This rather rare move loses a tempo and can hardly cause Black problems. 9...0-0 10.Be2 10.h3 Investing another tempo to prevent ...Bg4 is playable, but rather slow. An interesting reply is: 10...Qa3!? 11.Be2 b6 12.0-0 cxd4 13.cxd4 In Zhao Jun – Zeng Chongsheng, Tianjin 2018, Black traded bishops on a6, but I think Black does better to keep the position more complex with:

701

13...Bb7N 14.Qb1 This avoids moving an already-developed piece, but it feels rather awkward. 14.Bc1 Qd6 15.Qd3 is playable, though after 15...Qc6 16.d5 Qa4÷ Black is certainly not worse. 14.d5!? is interesting but Black is not obliged to accept the sacrifice on a1, and should instead play 14...Na6!÷ when Black will take over the c-file, with the knight coming to c5 at a suitable moment. Black’s position is promising. 14...Rc8 15.Bb4 The calm 15.Rc1 gives Black a pleasant game after 15...Nc6 16.Be3 e6³, when his pieces are active and White’s queen is passively placed, while the d4-pawn remains a target. 15...Qa4 16.Bd1 16.Bxe7 Qe8 17.Ba3 Bxe4³ leaves White worse due to the isolated d4-pawn. 16...Qd7³ Black’s pieces are more harmonious and the d4-pawn is weak.

702

10...Bg4 11.0-0 Rd8 Forcing White to decide how to protect his centre. 12.Rb1 This has been the most popular continuation. 12.Ng5? loses valuable time as well as control over the centre, and 12...Bxe2 13.Qxe2 Qa6 14.Qxa6 Nxa6 15.d5 c4 16.Be3 Rac8 17.Bxa7 Bxc3 18.Rab1 Rd7µ left Black in control in Korchnoi – Smirin, Biel 2002. 12.a4?! tries to seize space but White cannot really justify spending a tempo in this way. 12...Bxf3 13.Bxf3 cxd4 14.cxd4 was seen in Polak – Timoscenko, Austria 2007, and now 14...Qa6!N 15.Be2 Qd6³ would have put White’s centre under heavy fire, forcing him to fight for a draw in some way. 12.d5?! has scored highly but it is not a sound sacrifice: 12...Bxc3 13.Bxc3 Qxc3 14.Rc1 Qa3 15.Ne5 Bxe2 16.Qxe2 Nd7 17.f4 Rac8³ left White without enough compensation in T. Carlsen – Cipolli, corr. 2007. 12.Qc2!? seems playable, although it is understandable that very few players have opted to place the queen on such a seemingly exposed location. 12...cxd4 13.cxd4 Qa3 14.Rab1 Nc6

703

15.d5 (15.Bc1!?N may be the way for White to maintain the balance: 15...Qd6 16.d5 Bxf3 17.Bxf3 Nd4 18.Qd3 Nxf3† 19.Qxf3 b6 20.Bg5 Rac8=) 15...Bxf3 This was Do Minh Phung – Nguyen Viet Chung, Vung Tau 2004. Instead of bizarrely taking with the g-pawn, White should have played 16.Bxf3N, although 16...Nd4 17.Qd1 Nxf3† 18.Qxf3 Qxa2 19.Bf4 b6³ still leaves him under pressure to prove his compensation. 12...cxd4 13.cxd4 Qxa2 14.Ra1 Qb2 15.Rb1 Qa2 16.d5!? Obviously White can force a draw with 16.Ra1= as played in Gelashvili – Mchedlishvili, Lagodekhi 2016, and a few other games. The text move is the ambitious choice, although White takes a significant risk by playing it. 16...b6 17.Bg5 Nc6 18.Rc1 h6 19.Bh4 19.Rxc6?! Bxf3 20.Bxf3 hxg5µ simply gave Black a healthy extra pawn in Degtyaryov – Pellen, corr. 2009. We have been following S. Ivanov – Kantans, Vaujany 2014. My improvement is:

704

19...Ne5!N 20.Bxe7 I also checked 20.Rc2 Qa3 21.Nxe5 Bxe2 22.Qxe2 Bxe5 23.f4 Bf6 24.Bf2 Bg7³ when White’s compensation for the pawn is insufficient. 20...Rdc8 21.h3 Rxc1 22.Qxc1 Qxe2 23.Nxe5 Rc8 24.Qb1 Bxe5 25.hxg4 Rc2 26.Qd1 Qxd1 27.Rxd1 Rc7 28.d6 Rd7ƒ White is in danger due to the connected passed pawns on the queenside, and he will need to play accurately to hold a draw. B2) 9.Nd2

705

This is rather more poisonous than the previous option. White deprives the opponent of the ...Bg4 resource and gets ready to disturb the black queen. In some scenarios, he may also strengthen his centre with f2-f3. 9...Bd7!? Several other moves have been tried by strong players but I like the idea of preparing ...Ba4. 10.Rb1 White attacks the b7-pawn, abandoning the c3-pawn to its fate. 10.Nb3 leads to a forced draw: 10...Qxc3† 11.Bd2 Qb2 12.Bc1 Qc3†= White gains nothing with the alternative development of the rook: 10.Rc1 cxd4 11.cxd4 11.Nc4 Qa4 12.Qxa4 Bxa4 13.cxd4 0-0 is also pleasant for Black. A good example continued 14.Be2 Rd8 15.d5 Na6 16.Na5 Nb4 17.Nxb7 Rdb8 18.Rc7 Nxa2 19.Bc4 Nc3 20.Rxe7 a5 21.d6 Bc6 22.Rxf7 Bd5 23.Rxg7† Kxg7 24.exd5 Rxb7 25.0-0 Ne4ƒ when the complications had subsided and White faced a difficult defence in Kazoks – Morozov, corr. 2017. 11...0-0 12.Bc4 Ba4 13.Bb3 Bb5 Black’s bishop shuffle effectively wins him a tempo, and White will have to make concessions to finish his development.

14.a4 14.Bc4 Bxc4 15.Rxc4 Qxa2 16.0-0 occurred in Y. Vovk – Nepomniachtchi, Berlin (blitz) 2015, when 16...a5!N 17.Rc7 a4 18.Rxb7 Rc8µ would have given White problems fighting against the powerful passed pawn. 14...Ba6 15.Bc4 706

We have been following Deac – Areshchenko, Calimanesti Caciulata 2016. My improvement is: 15...Nc6N 16.0-0 Rac8 17.Bxa6 Qxa6 18.e5 b6 19.h3 Rfd8 20.Rc4 Na5³ Having won the fight for the c-file, Black has a pleasant game, especially as the pawns on a4 and d4 are weak. 10.Be2!? This rare but quite interesting move has been chosen by a few strong GMs. 10...Ba4 11.Qc1 cxd4 12.Bxd4 12.cxd4? Nc6µ puts White under serious pressure in the centre and on the queenside. 12...Bxd4 Black probably rejected 12...0-0N on account of 13.Bxg7 Kxg7 14.Qa3, although the computer is not too concerned about the pin, so you may wish to investigate this as well. 13.cxd4 Nc6 14.d5 Nd4 15.Bd3 0-0 16.Qb2 Here I can offer an improvement on Aleksandrov – A. Zhigalko, Minsk 2018.

16...e5N 17.0-0 Bc2! 18.Nc4 18.Nb3?! Bxb3 19.axb3 Qb4³ favours Black thanks to his strong knight and the weakness of the b3pawn. 18...Qa6 19.Bxc2 Qxc4 20.Bd1 Qd3 21.Re1 Qa6÷ Black can look to the future with confidence, due to his superior minor piece and the possibility of creating a passed pawn on the queenside in the long run. 10...Qxc3 We have reached an important position where White has two interesting moves: B21) 11.dxc5 is relatively solid whereas B22) 11.d5!? is somewhat riskier and more ambitious.

707

B21) 11.dxc5 Bc6 12.Bb5 This is the usual move, which generally leads to exchanges and equality. 12.Bc4!? leads to a more complicated game. 12...0-0 13.0-0 In Illingworth – Khusnutdinov, Sydney 2015, it would have been natural and logical to play:

13...Rd8!N Black has a good game no matter how White plays. For instance: 14.Qe2!? (I also checked 14.Bd5 e6 15.Bxc6 Nxc6 16.Qc1 [16.Rxb7?? is impossible due to 16...Na5–+] 16...Qa5 17.Rxb7 Qxa2÷ when Black’s game looks slightly preferable: he has an outpost on d4 and the bishop breathes fire on the long diagonal) 14...Nd7 15.Rfc1 Qa5÷ The position is rather complicated but I like Black’s chances. His pieces work in harmony and the c5-pawn may later become weak.

708

12...Bd4! This is not the only decent move but it seems the most challenging. 13.Bh6 Another game continued 13.Rc1 Qb2 14.Rc2? (White should settle for repeating the position with 14.Rb1 Qc3) 14...Qxb5 15.Bxd4 0-0 and Black was clearly better in Bluebaum – Schreiner, Meissen 2013. If you prefer to avoid the early repetition, you can also consider 13...Qb4!?N 14.Rc4 Qxb5 when Black is at least equal, whichever way White takes on d4. 13...Qxc5 14.Bxc6† Nxc6 15.0-0 White has some compensation for the pawn but it helps that Black can immediately force the queens off. 15...Qh5 16.Qxh5 gxh5 17.Rfc1 17.Rxb7? loses the exchange to 17...Bb6 followed by ...0-0-0. 17...0-0-0 17...Rg8!?N 18.Rxb7 Rg6 is an ambitious option, completing development in a creative way. A possible continuation is 19.Rc7 Ne5 20.Be3 Bxe3 21.fxe3 h4 22.h3 Rd8 23.Nb3 Ra6, reaching a complex endgame where Black’s chances are not worse. 18.Nf3 We have been following Erdos – Kreisl, Zalakaros 2016. I recommend:

18...Rhe8!?N 19.Rb5 e5 20.Bg7 a6 21.Rd5 Rd7 22.Kf1 Rc7 Black is forced to return his extra pawn in order to neutralize White’s initiative. 709

23.Nxd4 exd4 24.Bxd4 Nxd4 25.Rxc7† Kxc7 26.Rxd4 Re5= Black’s weakened kingside structure is balanced out by his queenside majority and active pieces, so the endgame is equal. B22) 11.d5!?

White sacrifices a pawn, betting on the unfavourable placement of the black queen. 11...b6 12.Bc4 12.Be2?! 0-0 13.0-0 is too passive. In Prizant – Kowsarinia, St Petersburg 2014, the logical and thematic 13...e6!N 14.Re1 exd5 15.exd5 Qa5 16.Qb3 Qa4³ would have left White with insufficient compensation for the pawn. 12...0-0 13.0-0 Qa5 Retreating the queen seems best. White has tried a couple of different approaches from here. 14.Bb3!? This move prepares Nc4, a2-a4 and f2-f4, which seems attractive, but Black is not going to stand still either. 14.Qe2!? This move, which prevents ...b5, was Mamedyarov’s choice. 14...e6 15.e5? This move leads to big difficulties for White. He should have preferred 15.Rfc1!?N 15...exd5 16.exd5 Re8÷ when White has compensation for the 710

pawn but no advantage. I think White is at a slightly higher risk, since Black just needs to get his queen into the game and he will stand better. 15...exd5 16.Bxd5 Nc6 17.Nc4 Qa6 18.Rfd1 In Mamedyarov – Svidler, Riyadh 2017, Black missed the strongest continuation:

18...Rad8!N White collapses in the centre, for instance: 19.Rb2 19.f4 Nb4! gives White huge problems. 19.Bg5 Rde8 20.f4 Nd4 is also depressing for White. 19...Nxe5 20.Bxc5 Bg4 21.f3 bxc5 22.Nxe5 Qxe2 23.Rxe2 Be6 24.Nc6 Rd6 25.Ree1 Bxd5 26.Ne7† Kh8 27.Rxd5 Ra6–+ Black has an extra pawn and a won position. 14...Qa6 Black walks the queen out of the danger zone. 15.a4!? A more recent try was 15.Re1!? in Parligras – Baron, Skopje 2019. I recommend improving the queen: 15...Qc8N White has just enough compensation, but Black has no obvious weaknesses. Many moves are possible and I will just present one illustrative line: 16.Bc4 Na6 17.Nf3 Nc7 18.a4 e5! 19.Qc2 Ne8÷ Followed by ...Nd6. White still has some compensation for the pawn, but in a practical game I would side with Black. The text move was played in Verst – Goffin, corr. 2015. Once again, many continuations are playable but I prefer:

711

15...Rc8!?N Preparing ...c4 is a natural human idea. 16.Qc2 The alternative is: 16.Rc1 Bb2 17.Rc4!? (17.Rb1 Bg7 18.Rc1 Bb2= repeats the position) 17...b5 18.Rc2 Bd4÷ Reaching an unclear position in which Black’s chances are no worse. 16...b5! Black simplifies the queenside and gains some breathing space for his pieces. 17.axb5 Bxb5 18.Bc4 Bxc4 19.Qxc4 Qxc4 20.Nxc4 Nd7 21.Rb7 Nb6 22.Nxb6 axb6 23.Rxe7 Bf8 24.Rb7 Rab8 25.Rxb8 Rxb8 26.Rd1 Bd6 27.f4 f6 28.Rb1 b5 29.e5 fxe5 30.fxe5 Bxe5 31.Bxc5 Bc3= A natural sequence of moves has resulted in a drawish endgame. B3) 9.Qd2

712

This is the main line, and a favourite choice of Kramnik among other strong players. 9...0-0 10.Rc1 The rook vacates the long diagonal and reinforces the c3-point, preparing central expansion with d4d5 and c3-c4. Other moves are not too challenging, for instance: 10.h3?! is too slow, and 10...Nc6 11.Rc1 cxd4 12.cxd4 Qxd2† 13.Kxd2 Rd8 puts White’s centre under serious pressure. 14.Rc4 Nxd4 15.Bxd4 e5 16.Rc7 was seen in Petursson – Borisek, Berlin (rapid) 2015, and now a small finesse is:

16...Bh6†!N Before White’s bishop can go to c4, Black forces the king to choose a square other than d3, which would now block the bishop. 17.Kc2 exd4 18.Bc4 d3†! 19.Bxd3 Be6 20.Rb1 b6 21.a4 713

Rac8µ 10.Be2?! is another weak move which can be met in the same way: 10...Nc6 11.Rc1 cxd4 12.cxd4 Qxd2† 13.Kxd2 Rd8 14.Rc4 (14.Rhd1 Nxd4 15.Nxd4 e5 16.Kc3 exd4† 17.Bxd4 Rxd4! is an important detail; 18.Rxd4 Be6 19.Kd3 Bxd4 20.Kxd4 Bxa2 and although White had a certain amount of activity for the pawn, Black still enjoyed excellent winning chances in Jahn – Pribyl, Vrbno pod Pradedem 1965) 14...Bg4 15.d5 e6 16.Kc1 exd5 17.exd5 Ne5 18.Rc2 Bf5 19.Rc5 b6 20.Nxe5 Bxe5 21.Rc4 Rac8 22.Rxc8 Rxc8† 23.Kd1 Be4 24.Bf3 Bxf3† 25.gxf3 Rd8–+ White’s attempts to avoid the loss of a pawn failed miserably in Arjun – Danin, Koge 2013. 10.Rd1?! is not the most useful move, as the d4-pawn remains paralysed and Black quickly develops pressure in the centre: 10...Nc6 11.Be2 Bg4 12.0-0 Rfd8 13.Qb2

13...Qb6! 14.Qxb6 axb6 15.dxc5 Rxd1 16.Rxd1 bxc5 17.Bxc5 Bxc3 18.h3 Rxa2 19.Bb5 Bxf3 20.gxf3 Bf6 21.f4 Ra5 22.Bxc6 bxc6³ Gunnarsson – Thorsson, Reykjavik 2016. White has some drawing chances, but Black can obviously be happy to have reached a pawn-up endgame. 10.Bc4 This outwardly active continuation permits Black to develop strong counterplay. 10...Nc6 11.Rb1 11.Rd1 Bg4 12.0-0 Rad8 13.Qb2 Bxf3 14.gxf3 occurred in Rafaelic – Gakic, Porec 2011, when 14...cxd4N 15.cxd4 Qh5 16.Be2 Qh4 17.Kh1 Rd7³ would have given White problems both in the centre and on the kingside. The text move is an attempt to slow the development of the c8-bishop, but White is fighting for equality at best.

714

11...a6 12.Bd5!? 12.0-0 makes Black’s task relatively easy: 12...cxd4 13.Nxd4 Nxd4 14.Bxd4 Bxd4 15.Qxd4 Qc7 16.Rfd1 b5 17.Bd5 Rb8 18.Qe3 e6 19.Bb3 a5 20.e5 Bb7 21.Rd4 Rfd8 22.h3 Rxd4 23.cxd4 Rc8µ Despite having ‘helped’ White to repair his pawn structure, Black kept a clear advantage due to his more mobile pawn majority and domination of the open file in Kuiphof – Wister, corr. 2011. 12...Rd8 13.0-0 cxd4 14.cxd4 Qxd2 15.Bxd2 Nxd4 16.Nxd4 Bxd4 17.Bxb7 Rb8 18.Ba5 Rf8 19.Bd5 Be6 Black is comfortably equal, and on the next move White started to go astray. 20.Rfd1?! Rxb1 21.Rxb1 Bxd5 22.exd5 Rc8³ White faced a difficult defensive task in Ballesteros Gonzalez – Suarez Uriel, Madrid 2010. 10.Rb1 Finally, the threat of Rb5 is easy enough to handle. 10...b6 11.Rb5 11.Rc1 Bb7 12.d5 Nd7 has been covered on page 370: see the 10.Rc1 line in the notes to variation A1. 11.Bc4?! was played in Astaneh Lopez – A. Hunt, Bunratty 2009. I am not sure why Black refrained from 11...Bb7N, since 12.Bd3 Rd8ƒ puts strong pressure on White’s centre and 12.e5 Bxf3 13.gxf3 Rd8 gives Black excellent play. 11...Qa4

715

12.Rb3 Not for the first time in this chapter, White shuffles his rook in the hope of harassing the queen, but Black easily gets a comfortable game by exchanging light-squared bishops. 12.dxc5? is a positional blunder: White opens the queenside files and destroys his own centre, all for the sake of an extra pawn which he will not be able to hold on to for long anyway. 12...Nc6 13.cxb6 axb6 14.Rb2 Rd8 15.Nd4 Bxd4 16.cxd4 Nxd4 17.Bd3 Qa3 18.0-0 Ba6 19.Bxa6 Nf3† 20.gxf3 Rxd2 21.Rxd2 Qxa6µ Sifre Calafat – Casas Saez, corr. 2009. 12.Rb2 Ba6 13.Bxa6 Qxa6 is also comfortable for Black. A good example continued 14.Qe2 Rc8 15.Qxa6 Nxa6 16.Kd2 Rd8 17.Ke2 Rac8 18.Rhb1 cxd4 19.cxd4 Rc4 20.g4?! e6 21.h4 when, in Colonna – Rossi, Grosseto 2009, the natural 21...Rdc8N 22.h5 Ra4 23.Rd1 Nb4 24.Rdd2 Bf8ƒ would have forced White into defensive mode. 12...Ba6 13.Bxa6 Qxa6 14.Qe2 Qa4!? A move such as 14...Rc8N is equally playable – it’s a matter of taste whether or not you prefer to allow the queen trade. 15.0-0 Rc8 16.Rfb1 cxd4 17.cxd4 Nc6 18.Qd3 Na5 19.Ra3 Qd7 20.Bf4 Rd8 21.d5 Rac8÷ Black was fine in Probstfeld – Boehme, email 2012.

716

10...Nd7! This is far from the most obvious move, but in recent years it started to become popular. Several strong GMs have played it, with Nepomniachtchi perhaps the biggest champion of Black’s cause. One of the main points of the text move is to move the knight to f6 to challenge the e4-pawn. Another important possibility is to strike in the centre with ...e5. Among the alternatives, 10...Rd8 is the most popular choice. However, after 11.d5 e6 12.Bg5 f6 13.Be3 Nc6 I believe the rare 14.h4!?ƒ to be somewhat unpleasant for Black. We have reached an important branching point, with four main options for White: B31) 11.Bh6, B32) 11.h4!?, B32) 11.d5 and B34) 11.Bd3. All these moves have their plus points, although the last two are by far the main lines in terms of popularity. 11.e5? is extremely dubious. 11...Rd8 12.e6 fxe6 13.Bh6 occurred in C. Varga – Torok, Budapest 2003, when Black could have killed the enemy counterplay with:

717

13...Ne5!N 14.Be2 Nxf3† 15.Bxf3 cxd4 16.Bxg7 Kxg7 17.0-0 Rd6 18.h3 Rb8µ Black has two extra pawns and every chance of converting them to victory. Another rare idea is: 11.Be2 Nf6 12.Bd3 White’s idea seems to be to develop the bishop to d3 without allowing ...e5, but it is hard to believe that losing a tempo in this way can really threaten Black. 12...b6 Other interesting possibilities include 12...Rd8!?N and 12...Ng4!?N. 13.0-0 This position was reached in Vo Kim Cang – Tran Quoc Dung, Ho Chi Minh City 2017. Black has a wide choice of possibilities but I believe the most combative choice to be:

718

13...Bb7N 14.d5 14.e5 cxd4 15.cxd4 Qxd2 16.Bxd2 Nd5³ favours Black thanks to the outpost on d5 and the backward pawn on d4. 14...c4 15.Bb1 e6 16.dxe6 fxe6 17.Bd4 17.Ng5 is an active try but it is not dangerous. 17...Rae8 18.f3 e5 19.Qe2 Qa4 20.Rcd1 Bc8³ leaves the knight on g5 located badly, and the passive bishop on b1 is also not a good piece. 17...Rad8 18.Qe2 Nh5! Black has a promising game. An important line is:

19.Qxc4 Rxf3! 20.Bxg7 Nxg7 21.gxf3 Ba6 22.Qc7 Rc8 23.Qe7 Bxf1 24.Rxf1 Re8 25.Qc7 Qg5† 26.Kh1 Qh5 27.Kg2 Rf8 28.Qg3 Qc5µ With ...Nf4 coming, Black is obviously doing well. 719

B31) 11.Bh6

Interestingly, despite the relative rarity of this move, Grünfeld expert Nepomniachtchi has faced it as many as four times. Twice he played 11...Rd8 but I prefer his other choice: 11...Nf6 12.Bd3 As played by Nakamura. Trading bishops immediately is slightly less accurate: 12.Bxg7 Kxg7 13.Bd3 Rd8 14.d5 14.Qg5?! occurred in Corral Blanco – Salgado Lopez, Linares 2014, when 14...h6!N 15.Qe5 Qxa2 16.Qxc5 Bg4µ would have given Black an obvious initiative, since 17.Ne5?? Nxe4! would be curtains. Now I can offer an improvement:

720

14...e6!N 14...c4 15.Bb1 e6 was seen in Bogosavljevic – Nepomniachtchi, Herceg Novi 2005. Black should be fine here and we will reach the same position in our main line below. However, White’s move order involving an early Bxg7 enables Black to get a slightly improved version. 15.0-0 15.c4 Qxd2† 16.Nxd2 b5! is pleasant for Black. 15...exd5 16.e5 Ng8÷/³ White has a certain amount of play for the pawn but, in contrast to the above game, his compensation is much less clear when he does not have the d4-square available for his pieces. 12...Rd8 13.d5 13.Bxg7 Kxg7 leads back to the previous note. 13...c4 Black could also consider 13...e6!?N, hoping to transpose to the note to White’s 12th move after an exchange on g7. However, after 14.0-0 exd5 15.e5 the play will develop differently, as Black’s knight does not have the g8-square available. This could be analysed more deeply but I prefer the text move. 14.Bb1 e6 We have reached a critical position where the d5-pawn will soon be lost. White must decide whether or not to trade the dark-squared bishops. 15.Bxg7!?N Delaying the exchange until now gives White a better version of the note to move 12 above.

721

15.0-0 was Nakamura’s choice, based on the idea that White’s bishop on h6 is more useful than its counterpart. 15...exd5 16.e5 Ne4 17.Qf4 Bh8 Despite the fact that the bishop remains restricted by the e5-pawn, it does a good job of defending the dark squares around the king. 18.Rfe1 Re8 19.Bxe4 dxe4 20.Qxe4

20...Qxa2!?N (The calm move 20...Be6 led to balanced play and later a draw in Nakamura – Nepomniachtchi, Baku 2015.) The text move is the most ambitious choice. White has reasonable compensation for the pawn but Black has good defensive resources, and will look to get his queen back into play via a6-e6 or possibly a3-e7. Even if Black returns his extra pawn at some point, his bishop pair and passed a-pawn will remain useful. 15...Kxg7

722

The position has a double-edged character. Black is poised to win a pawn, while White will look to develop his initiative on the kingside. We have transposed to a couple of games in which White exchanged on g7 a few moves earlier. (As we saw in the note to move 12, when this happens, Black can maximize his chances with 14...e6!N instead of 14...c4.) 16.0-0 16.d6? simply marches the pawn to its death. 16...e5! 17.Qg5 Now in Farjot – Perron, France 1995, the most accurate continuation would have been 17...Re8!N 18.0-0 Bg4 19.Nd2 Qc5 20.h3 Be6–+ when White will soon be a pawn down without the slightest compensation. 16.Nd4!?N is an interesting move which has yet to be tested. My analysis continues 16...exd5 17.e5 Ne4 18.Qf4 Re8 19.0-0 Bd7 20.f3 Nc5 21.Qe3 Qb6 22.f4 f6÷ when White still needs to prove his compensation and Black’s position seems easier to play. Another idea awaiting a practical test is: 16.Qf4!?N 16...exd5 17.e5 Nh5 17...Ne4?! 18.Bxe4 dxe4 19.Ng5ƒ should be avoided. 18.Qh4 18.Qe3 is safer, though after 18...Kh8 19.Qh6 Qa3 20.0-0 Qf8 21.Qg5 Ng7 22.Qh4 Ne6÷ Black is certainly not worse. Once again, the onus is on White to justify his pawn sacrifice. The text move is more aggressive but Black has an excellent reply in store.

18...d4! 19.0-0 d3! Sacrificing a piece in order to relegate the bishop on b1 to the role of a beautiful picture on the wall. 20.g4 Bxg4! 21.Qxg4 Rac8 22.Qg5 Qd5 23.Qe3 b5© Black has fantastic compensation for the piece. Later he can decide whether to maintain the bind on the queenside or look to break through with ...a5 and ...b4. 723

16...exd5 17.e5 Ne4 18.Qf4 I also considered 18.Qe3!?N 18...Qb6 19.Nd4 Re8 20.f4 Nd6 21.Qf3 Nb5 22.Rcd1 Kh8 23.Kh1 Nxd4 24.Rxd4 Be6 25.f5 Bxf5 26.Bxf5 gxf5 27.Qxf5 Qe6= when White has enough compensation for equality, but nothing more. After the earlier transposition, we have been following Bogosavljevic – Nepomniachtchi, Herceg Novi 2005. Now I can offer an improvement:

18...Nc5!?N The knight heads to e6 to defend the kingside. A check on f6 will not hurt. 19.Ng5 I don’t believe White has anything better than this move, which leads to a forced draw. A riskier alternative is 19.Qh4 Ne6 20.Nd2 Qb6 21.Kh1 Nc5 22.f4 f5÷ when once again White needs to work to prove his compensation. 19...Rf8 20.Qf6† Kg8 Black is ready to repel the attack with ...Qd8, so the next move is more or less forced. 21.Nxh7! Kxh7 22.Bxg6† Kg8 23.Bh7† Kxh7 24.Qh4†= With perpetual check. B32) 11.h4!?

724

Although this move has been a rare choice thus far, the success of Alpha Zero reminds us that we should constantly be on the lookout for attempts to ram us with the rook’s pawn. The text move leads to a lively and challenging battle, and I expect to see more games with it in the future. 11...Nf6 12.Bd3 Rd8 13.h5 This move is not forced, but it is certainly the most consistent and critical choice. I would like to improve on Predke – Ofitserian, Sochi 2019, by choosing the most principled and challenging continuation: 13...Nxh5N White has many possible ways to continue, and I have strived to analyse the most logical continuations according to my understanding of the position. 14.e5 Cutting off the knight’s escape route seems like the most critical try. 14.Bh6? crudely threatens Rxh5 but 14...cxd4 strikes in the centre while conveniently opening the 5th rank for the queen to defend. White has nothing better than 15.Bxg7 Kxg7 16.cxd4 Qxd2† 17.Kxd2 Bg4µ when Black enjoys an active position as well as an extra pawn. 14.Bc4?! also allows an early queen exchange and thus cannot be recommended for White: 14...cxd4 15.cxd4 Qxd2† 16.Kxd2 e5 17.Bg5 Rd6 18.Rh4 h6 19.Be7 Rd7 20.Ba3 a6³ Although the position remains quite lively, White does not have enough activity to claim full compensation for the missing pawn. 14.Rh4?! Nf6 15.Bh6 is another attacking attempt which falls short: 15...Ng4 16.Bxg7 Kxg7 17.Ne5 725

h5 18.Nxg4 Bxg4

19.f3 Be6 20.d5 Qc7! 21.Kf2 c4 22.Bc2 Qc5† 23.Ke2 f6 24.g4 Bf7 25.gxh5 g5 26.h6† Kh8 27.Rh2 b5³ Black is no longer a pawn up, but his king is safe and he is ready to play on the queenside with ...a5 and ...b4. 14.d5!? seems like the most serious alternative to our main line. My analysis continues: 14...Bg4 15.c4 Qxd2† 16.Nxd2 b6 17.f3 Bc8 18.g4

18...Bd4! 19.Bh6 (19.Bxd4? is of course met by 19...Nf4! 20.Bf1 cxd4 when Black boasts a superb knight in addition to the extra pawn) 19...Nf6 20.Nb3 Bb2 21.Rc2 Be5÷ White’s space advantage gives him just enough compensation. However, with an extra pawn and no major weaknesses, Black should not be unhappy with the outcome of the opening.

726

14...Be6 15.Rh4 Threatening to trap the knight with g2-g4. There is no justification for 15.Rxh5? cxd4 16.Bxd4 gxh5µ. 15...Bd5 16.Ng5 h6 17.Ne4 cxd4 18.cxd4 Qxd2† 19.Kxd2

19...Bxa2 It is worth depositing another pawn in the bank, especially as Black now has connected passers on the queenside.

727

20.Nc3 20.g4 Nf4! is a nice resource, and after 21.Bxf4 g5 22.Nxg5 hxg5 23.Bxg5 Rxd4 24.Bxe7 Bxe5 Black keeps a slight plus. 20...Be6 21.Ne2 Rac8 22.g4 Rxc1 23.Kxc1 Rc8† 24.Kd2

24...Nf4! Again this move is key. 25.Nxf4 g5 26.Nxe6 gxh4÷ A complicated endgame has arisen. Any result is still possible, but White needs to be more careful due to Black’s multitude of passed pawns. B33) 11.d5

728

11...Nf6! Several other moves have been tried but the text is the most popular, and the strongest in my opinion. Black attacks the centre with a view to provoking a queen exchange, as occurs after White’s usual reply. 12.c4 12.e5 Nd7 13.Bh6 has been played a few times but 13...Nxe5! is a good answer: 14.Nxe5 Bxe5 15.Bxf8 Kxf8

16.f4 (16.g3 occurred in Sarkar – Gledura, Gibraltar 2017, when 16...Bf5 17.Bg2 Rd8 18.0-0 Bf6© maintains Black’s compensation) The text move was played in Bellaiche – Sunilduth Lyna, Moscow 729

2018. I recommend 16...Bd6N© when Black has fine compensation for the exchange, thanks to the bishop pair and compact pawn structure. Keeping the queens on the board with 12.Bd3!? has been a rare choice, but it has been tried by some strong GMs. A good example continued: 12...c4 13.Bb1 Ng4!? 14.Bd4 (I also considered 14.Bg5!?N which seems more logical, although 14...f6 15.Bf4 e5 16.dxe6 Bxe6 17.Nd4 Rad8 18.0-0 Ne5÷ still gives Black a good position) In Sakaev – Morozevich, Moscow 2008, it would have been good to continue with:

14...Bh6N 15.Qb2 f6! Black’s bishop is too useful to be ‘sacrificed’ for the feeble rook on c1. 16.h3 e5 17.dxe6 Ne5 18.e7 Re8ƒ Black will pick up the e7-pawn while keeping some initiative, as his pieces are located actively and harmoniously. 12...Qxd2† 13.Nxd2 b6 We have reached a double-edged queenless middlegame where Black’s task will be to challenge White’s strong centre with ...e6, and quite possibly ...f5 after a suitable knight move. 14.Be2 The fact that Vachier-Lagrave played this – even if it was only a blitz game – gives us reason to take it seriously. Two other moves deserve attention. 14.f3 This has been the most popular continuation. 14...e6 15.Kf2 15.Be2 transposes to 15.f3 in the notes to the main line below. 15.Nb1 was seen in a battle of strong GMs, Salem – Artemiev, Martuni 2015. I would like to recommend a standard plan for such positions: 15...Ne8!?N 16.Nc3 f5 17.exf5 gxf5 18.Bd3 Nd6 19.0730

0 Bd7÷ Black is at least equal. 15...Ne8 16.a4 f5 17.exf5 A useful improvement is:

17...gxf5!N 17...exd5 led to an eventual success for Black in Ladva – Bok, St Petersburg (blitz) 2016, but White should have taken the opportunity to take on g6, leading to unclear play. 18.g3 exd5 19.cxd5 Bd7 20.a5 Nf6³ Black’s chances are somewhat higher due to White’s questionable pawn structure. 14.h3 This move has also been tried by some strong GMs. Again White covers the g4-square, but this time in a different way. 14...e6 15.Bd3 Instead of exchanging on d5 immediately, Black should maintain the tension with: 15...Re8 16.0-0 Ba6 White faces a difficult decision on how to keep his centre together. 17.dxe6?! I don’t like this move, as White voluntarily surrenders the centre. 17.Bc2 is preferable although 17...exd5 18.exd5 b5„ gave Black a fine, active game in Akobian – Nepomniachtchi, Riyadh (blitz) 2017. 17...Rxe6 18.Rfd1 Rd8 19.Be2 This position was reached in Parligras – Svidler, Germany 2016. Black has a pleasant position after any sensible move, but I think the strongest and most purposeful continuation is:

731

19...Ne8!N Intending ...Bd4 with a favourable trade. The knight may later make its way via e6 to d4. 20.Nf3!? Sacrificing a pawn for some activity may be the best that White can do. A quieter continuation such as 20.Bg4 Re7 21.a4 can be met by 21...Bd4 22.Bg5 f6 23.Bf4 h5 24.Bf3 Nc7µ when Black’s superior pawn structure and outpost on d4 give him clearly better chances. 20...Rxd1† 21.Rxd1 Rxe4 22.Rd7 Bxc4 23.Bxc4 Rxc4 24.Rxa7 h5 25.Ra8 Kf8 26.Kf1 Ke7 27.Ra7† Ke6! 28.Ng5† Kd5 29.Rxf7 Bf6 30.Nf3 Ke6 31.Rb7 Rb4µ Material is now equal, but Black keeps good winning chances thanks to his active pieces and strong passed pawn on c5.

732

14...e6 Black continues with his plan and forces White to make a decision. 15.e5?! Although this move is objectively a mistake, we will take it as the main line as it is the most ambitious and critical choice. 15.0-0N is a logical yet untested move. Play may continue: 15...exd5 16.cxd5 (I also checked 16.exd5 Ng4 17.Bf4 Re8 18.Bd3 Bd7 when Black is fine: his ideas include ...Ne5 or ...Bh6, and in some cases ...b5 can be considered)

16...Re8 17.Bd3 Ng4 18.Bf4 Ne5 19.Be2 Bd7 20.Rfe1 f5÷ Black’s central counterplay is underway and he has a fine position. 15.f3 was played in Vachier-Lagrave – Nepomniachtchi, Berlin (blitz) 2015. I think Black’s most promising continuation is:

733

15...Ne8!N Thematically unblocking the f-pawn while preparing to post the knight on d6. My illustrative line continues: 16.0-0 f5 17.a4 Nd6 18.a5 Re8 19.Bd3 Nb7 20.Nb3 Bd7

21.Rfe1 Ba4 22.a6 Nd6 23.Nd2 exd5 24.cxd5 Rad8³ Black’s position seems slightly preferable due to the activity of his pieces. White’s central pawns are blocked, and if the b-and c-pawns can be mobilized at the right time, White could find himself in real trouble. 15...Ne8 16.f4 White strengthens the e5-pawn as reliably as possible. 16.Bf3 is risky. 16...exd5 17.Bxd5 Rb8 18.Bf4 occurred in Villegas – Vykouk, Pardubice 2017, when Black overlooked a powerful tactical idea: 734

18...Nc7!N 19.e6 Nxd5 20.Bxb8 (20.cxd5? Rb7 leaves White’s central pawns doomed) 20...Nc3 21.Bxa7 Re8 22.Kf1 Na4© Black has more than enough compensation for the sacrificed exchange, thanks to his two powerful bishops, active pieces and the unfortunate position of the white king. Note that 23.exf7†?! Kxf7 24.Re1 Be6µ only adds to White’s troubles, as his bishop may soon be trapped by ...Ra8. 16...f6 17.Nf3

This tense position was reached in Saiyn – Yeletsky, Samara 2016. Black should have continued his mission of undermining the enemy centre by means of:

735

17...g5!N 18.exf6 18.0-0? gxf4 19.Bxf4 exd5 20.cxd5 fxe5 21.Bxe5 Rxf3 22.Bxf3 Bxe5µ turns out unsuccessfully for White. I also considered 18.dxe6, when 18...Bxe6! is a promising pawn sacrifice (although 18...gxf4 is perfectly playable too). For instance: 19.exf6 Nxf6 20.Nxg5 Rae8 21.Nxe6 Rxe6 22.Bf2 Ng4µ Despite White’s extra pawn and temporary possession of the bishop pair, he has problems due to his poorly coordinated pieces and the unfortunate placement of his king. Finally, White can sacrifice a pawn with 18.d6, but after 18...gxf4 19.Bxf4 fxe5 20.Bxe5 Bxe5 21.Nxe5 Nxd6³ he does not have full compensation. 18...Nxf6 19.fxg5 Ne4ƒ Thanks to the unpleasant threat of ...Nc3, Black has more than enough compensation for the sacrificed pawn. B34) 11.Bd3

This is the most natural and popular continuation. White develops the bishop, prepares to castle and protects the e4-point in anticipation of ...Nf6. 11...e5! Black responds by putting maximum pressure on the d4-point. The idea is to meet d4-d5 with ...f5, obtaining counterplay in the spirit of the King’s Indian. 12.d5 736

This ambitious continuation has been White’s most popular choice. 12.Bh6 does not pose a problem for Black. 12...cxd4 13.Bxg7 Kxg7 14.cxd4 Qxd2† 15.Kxd2 exd4 16.Nxd4 b6 17.Ke3 (I also checked 17.Rc7N 17...Nc5 18.Rc1 Bd7 19.Ke3 Rfc8 20.Rxc8 Rxc8 21.f3 Kf6= when Black has no problems in the endgame) 17...Nc5 18.Rhd1 This occurred in Ding Liren – Nepomniachtchi, Internet 2019, when 18...Bd7N would have kept the bishop in touch with both flanks, with equal chances. 12.dxe5 Nxe5 13.Nxe5 Bxe5 Simplifying in the centre also fails to pose many problems. 14.Bc4 Perhaps a more reliable continuation is 14.0-0 Rd8 15.Qc2 as in Zaja – Zufic, Porec 2011. A logical continuation is 15...Qc7N 16.f4 Bg7 17.e5 Be6 18.f5 gxf5 19.Bxf5 Qxe5 20.Bxh7† Kh8 21.Rf3 Qxc3 22.Qxc3 Bxc3 23.Rxc3 Kxh7 24.Rxc5 Rd5= with an approximately equal endgame. It is also worth considering 14.f4!? as played in Oksuz – Esen, Izmir 2016. I like the following active idea: 14...Rd8!?N 15.fxe5 c4 16.Bd4 cxd3 17.Qxd3 Be6 18.0-0 Qxa2÷ Black has no problems: the bishop on d4 is passive and the passed a-pawn may become a powerful force later. 14...Be6 15.Bxe6 fxe6 16.0-0 Qc7 17.g3 This was Gledura – Vidit, Malmo 2018. My improvement is:

17...Rfd8N÷ Black has a comfortable game; his plans may include ...b5, or ...c4 followed by ...Rd3. 12.h4!? This creative idea only appears in two games on my database. 12...cxd4 13.cxd4 Qxd2† 14.Kxd2 exd4 15.Bxd4 Here I recommend an improvement on Aronian – Nepomniachtchi, Khanty-Mansiysk (ol) 2010: 737

15...Nf6!N 16.Rc7 Rd8 17.h5!? This seems more challenging than 17.Bxf6 Bxf6 18.Rhc1 Be6 19.R1c2 Rdc8³ when Black’s bishop pair and potential for a queenside passed pawn give him the upper hand. 17...Nxh5 18.Bxg7 Nxg7 19.Ne5 Be6 20.Rxb7

20...Rac8! 21.Rxa7 Rc5 In the space of a few moves Black has gone from being a pawn up to a pawn down, but his pieces have become considerably more active. 22.f4 f6 23.Nf3 Bc4 24.Ra3 Nh5 25.Rc1 Nxf4 26.Ne1 Nxd3 27.Nxd3 Rc6³ Black has more active pieces, plus a bishop versus a knight and fewer pawn islands. Therefore White faces a tough battle for a draw.

738

12...f5 Black wastes no time in challenging the enemy pawn centre while threatening ...f4. 13.Bh6 Saving the bishop while loosening Black’s control over e5. Several other moves are possible, for instance: 13.exf5 sees White giving up his centre in the hope of exploiting Black’s open kingside. 13...gxf5 14.Bh6?! (a better attempt looks to be 14.Ng5!?N 14...f4 15.Bxh7† Kh8 16.Bc2 fxe3 17.Qxe3 Nf6 18.Qg3 Bh6 19.0-0 Qc7÷ with a challenging game for both sides) In Stone – Bailey, Toronto 1989, Black could have safely grabbed the material:

14...e4!N 15.Bxg7 Kxg7 16.Ng5 exd3 17.Ne6† Kf7 18.Nxf8 Nxf8 19.Qxd3 Qxa2µ White does not have enough compensation, as the knight will go to g6 where it will control the important squares along the e-file. 13.Qe2!? fxe4 14.Bxe4 Nf6 15.d6 led to an interesting situation in Yu Yangyi – Vidit, Danzhou 2018. My improvement is:

739

15...Kh8!N 16.Ng5 h6 17.h4 Nxe4 18.Nxe4 b6 19.f3 Bf5÷ The weakness of the d6-pawn compensates Black for the passivity of certain pieces, especially the queen. 13.Qc2 f4 14.Bd2 b5 15.c4 b4 16.0-0 brings us to another useful novelty:

16...Qd8!N÷ (the inferior 16...Qc7? was played in Nakamura – Nepomniachtchi, Paris [rapid] 2019, when 17.Bxb4!!N 17...cxb4 18.c5ƒ would have given White a formidable initiative for the piece) The text move leads to a complex manoeuvring game where Black’s play seems easier, due to the possibility of storming the kingside with ...g5-g4. It is worth mentioning that 17.Bxb4?! cxb4 18.c5 is no problem here: not only is Black’s queen less exposed on d8 than c7, but it also supports 18...g5!³ with counterplay on the kingside.

740

13.Bg5 This move is recommended by Kornev in A Practical Repertoire for White with 1.d4 and 2.c4. White is trying to set up a wedge with d5-d6 and Be7 to smother Black’s play. 13...c4 14.Bxc4 fxe4 15.Nh4 Kornev recommends 15.d6† Kh8 16.Nh4 but does not consider 16...Qc5. White responded with the natural 17.Bb3 in Neverov – Kulaots, Minsk 2017, and here I suggest: 17...Bf6N 18.0-0 Bxg5 19.Qxg5 Qxd6÷ Reaching a complex position where Black’s chances are not worse. The text move was played in Rocha – Gabrielian, Lisbon 2018. Again we can improve Black’s play:

15...Nc5!N 16.0-0 Qa4 17.Qe2 17.Bb3?! Nxb3 18.axb3 Qxb3 19.c4 sacrifices a pawn in the hope of mobilizing White’s c- and dpawns, but 19...a5! 20.Rc3 Qb4³ highlights the fact that Black has a strong passed pawn of his own. I also checked 17.Be2 Qd7 18.c4 Qd6 19.f3 Bd7÷ when Black has a powerful blockade and stands no worse, especially keeping in mind the offside knight on h4. 17...Bg4! Black should take action before Be7 and/or Bb5 can be played. 18.d6† Be6 19.Bxe6† Nxe6 20.Be7 Rf7

741

21.Rce1 Nf4 22.Qxe4 Qxe4 23.Rxe4 Nd5 24.Nf3 Nxe7 25.dxe7 Rxe7= Black has successfully coped with the d5-d6/Be7 wedge, and has good prospects in the endgame due to White’s split queenside pawns.

13...fxe4 14.Bxe4 The inclusion of 14.Bxg7 Kxg7 also leads to an interesting game: 15.Bxe4 Nf6 16.c4 (I also considered 16.Qe3!?N 16...Qa6 17.c4 Qa5† 18.Nd2 Ng4 19.Qg5 Rf6 20.f3 h6 21.Qh4 g5 22.Qg3 Rf4! 23.Bb1 Qa3 24.Nb3 Qb4† 25.Ke2 Nf6÷ when the position remains complicated and Black’s king feels no worse than White’s) The text move was played in Stupak – Zinchenko, Paleochora 2017, and here I like:

742

16...Qb4!N 17.Qxb4 cxb4 18.Nd2 Bf5÷ Black’s chances are not worse. 14...Qa4! Black activates his queen to annoy the opponent and prevent him from getting fully coordinated. I believe White should improve on the existing game with: 15.Bb1!?N I also considered 15.c4N 15...Nf6 16.Bxg7 Nxe4 17.Qe3 Qa5† 18.Kf1 Kxg7 19.Qxe4 Bf5 20.Qxe5† Kg8 21.h4 Rae8 22.Qb2 Bd3† 23.Kg1 Re2© when Black’s activity is at least as important as his minimal material shortage. The game continued: 15.Bc2?! Leaving the a2-pawn unprotected while allowing the black queen to switch to the desired flank. 15...Qg4 16.Qg5 Nf6 16...Nb6!?N also deserves attention. For instance, 17.Bxg7 Qxg5 18.Nxg5 Kxg7 19.d6 Rd8 20.0-0 Rxd6 21.Rfe1 h6 22.Ne4 Rc6 23.Nd2 Be6³ and White still needs to prove his compensation for the pawn.

743

17.Bxg7?! In the event of 17.0-0 Qxg5 18.Bxg5 Nxd5 19.Rfe1 h6 20.Be4 Be6 21.c4 Nb4 22.Be7 Rf4 23.a3 Rxe4 24.Rxe4 Nd3 25.Rd1 Bf5 26.Re2 e4 27.Nd2 Re8 28.Bh4 g5 29.Bg3 b6³ Black’s outpost on d3 and White’s queenside weaknesses more than make up for the sacrificed exchange. White should have preferred 17.d6!?, although 17...Bxh6 18.Qxh6 Bf5 19.0-0 Bxc2 20.Rxc2 Rad8 21.Rb2 b6 22.Nxe5 Qe4 23.Nf3 Rxd6 24.Re1 Qc4 25.Qc1÷ reaches a position where Black’s superior queenside structure is at least as important as his slightly open kingside. 17...Kxg7 18.Rd1? White spends valuable time defending one pawn, only to lose a different one. He should have settled for 18.0-0 Qxg5 19.Nxg5 Nxd5 20.Rfe1 Re8 21.Ne4 b6 22.Rcd1 Be6³ when Black is a pawn up but White’s activity offers reasonable drawing chances. 18...Qc4 19.Qe3 Ng4 20.Qd3 Qxa2 21.Qe2 I think Black should force matters with:

744

21...e4!N 21...Bf5 was advantageous for Black in A. Onischuk – Wei Yi, Astana 2019, but I like the text move even more. 22.Bxe4 Qxe2† 23.Kxe2 Re8 24.Nd2 Bf5 25.f3 Nf6 26.c4 Nxe4 27.Nxe4 Bxe4 28.fxe4 Rxe4†µ Black has good chances to convert his extra pawn.

15...Qc4 An interesting alternative is: 15...Qg4!? 16.Qg5 Nf6 (but not 16...Nb6?! because of 17.c4²) 17.Bxg7 Kxg7 18.c4 e4 19.Qxg4 Bxg4 20.Nd2 Rae8 21.Nf1 Re7÷ White has a protected passed pawn but the slight vulnerability of the f2- and c4-pawns should compensate for this.

745

16.Bxg7 Kxg7 17.Qe3 Nf6 18.Nxe5 Qxd5 19.0-0 Re8 20.f4 Ng4 21.Qg3 Nxe5 22.Rcd1 Nf3† 23.gxf3 Qf7 24.Rfe1 Bf5= White’s active rooks offer enough compensation for his structural defects, but Black has nothing to fear. Conclusion This chapter has dealt with White’s scheme of seizing the centre with his pawns before supporting the d4-point with a bishop on e3. We started by analysing some options where he delays the development of the royal knight to f3. I see no benefit to this approach, and the analysis showed that Black could obtain good counterplay in all such variations. We then turned our attention to the more popular and challenging option of 7.Nf3 c5 8.Be3 (or 7.Be3 c5 8.Nf3). We respond with 8...Qa5, when 9.Nd2 demands close attention although 9.Qd2 0-0 10.Rc1 is deservedly the main line. I opted to meet this with the not-so-obvious 10...Nd7!, after which the knight can either move to f6 or support central action with ...e5. We examined several options from there, many of which can lead to complicated and somewhat irrational positions. In general, Black’s prospects seem quite bright, especially if he has a well-developed sense of dynamics. The cost of a mistake could be relatively high for both sides, making this an attractive battleground for combative Grünfeld players who seek to fight for more than equality.

746

A) 8.Bc4 400 B) 8.h3!? 403 C) 8.Be2 Nc6 405 C1) 9.Be3 406 C2) 9.d5!? 408 D) 8.Bb5† Nc6 411 D1) 9.d5 412 D2) 9.0-0 413 E) 8.Rb1 0-0 9.Be2 Nc6! 10.d5 Ne5 11.Nxe5 Bxe5 417 E1) 12.Rb3 426 E2) 12.Qc2 427 E3) 12.Qd2 e6 13.f4 Bc7 14.0-0 exd5 15.exd5 Ba5 428 E31) 16.Ba3 433 747

E32) 16.f5 435 E33) 16.d6! Game 7 438

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Nf3 c5 This is one of the most important positions in the Grünfeld. White has established a pawn centre and Black’s task will be to put pressure on it with moves like ...Nc6 and perhaps ...Bg4. This chapter will deal with five main options: A) 8.Bc4, B) 8.h3!?, C) 8.Be2, D) 8.Bb5† and E) 8.Rb1. 8.Bg5 transposes to 8.Nf3 in the notes to variation B of Chapter 12 on page 341, and 8.Be3 is variation B of Chapter 13. Other moves are not dangerous at all, for instance: 8.Qa4†?! Nc6 White already has problems defending the d4-square. 9.Be3 A similar situation arose after 9.Bb5 Bd7 10.0-0 a6 11.Bxc6 Bxc6 12.Qc2 cxd4 13.cxd4 Rc8 14.Qb1 0-0³ in Krasauskas – Labunec, Klaipeda 2002. Black has the advantage of two bishops and his position is much more comfortable to play. 9...Bd7

10.Bb5 10.Qa3!?N looks like a slight improvement although 10...cxd4 11.cxd4 0-0 12.Rd1 Rc8 13.Be2 Qa5† 14.Qxa5 Nxa5„ gives Black a comfortable endgame, as he is ahead in development and has captured the c-file. 10...a6 11.Bxc6 Bxc6 12.Qc2 cxd4 13.Bxd4 Bxd4 14.Nxd4 Rc8 15.0-0 0-0 16.Rab1 Qc7µ White had no compensation for the weak c-pawn in Radosevic – Molnar Gabor, Djenovici 2018. 748

8.Bb2?! This move assigns the bishop to a defensive role with few active prospects. 8...0-0 9.Be2 Other moves lead to even greater problems: 9.Bc4?! Bg4 hits the d4-pawn hard, and White will have to spoil his structure to avoid losing it. 10.00 cxd4 11.cxd4 Nc6 12.e5 Bxf3 13.gxf3 e6 14.f4 occurred in Sopow – Perkovics, Hungary 2009, when 14...Qh4N 15.Qf3 Bh6µ would have made it hard for White to defend the weaknesses on d4 and f4 in a convenient fashion. 9.d5?! e6 10.Bc4 b5! leads to big problems for White. 11.Bxb5 Qb6 12.Qb3 exd5 13.0-0? (13.exd5N was the lesser evil although 13...c4 14.Bxc4 Re8† 15.Kf1 Nd7 16.Qxb6 Nxb6ƒ reaches a position where White’s extra pawn is outweighed by his poorly coordinated pieces and misplaced king) Now in Marumo – Osman, Novi Sad (ol) 1990, the accurate 13...c4!N 14.Qb4 Na6 15.Bxa6 Qxb4 16.cxb4 Bxb2 17.Rab1 Bxa6 18.Rxb2 c3–+ would have won the exchange, the e4-pawn and the game.

9...cxd4 10.cxd4 Qa5† 11.Qd2 Qxd2† 12.Nxd2 12.Kxd2 does not change much. In Tsarouha – Aggeli, Panellhnia 2005, Black should have continued 12...Rd8N 13.Rhd1 Bd7 14.Bc4 a5 15.a4 Na6 16.Ke2 e6 17.Ne5 Bxe5 18.dxe5 Nc5³ when White’s bishop pair is outweighed by his poor pawn structure, with e4 and a4 both seriously weak. 12...Rd8 13.Nf3 In Azzi – Lim, corr. 2013, it would have been best to continue with:

749

13...b6N 14.0-0-0 Bb7 15.d5 Nd7 16.Bxg7 Kxg7 17.Nd4 Nc5³ White has achieved the most he could hope for from this variation. Nevertheless, his king is not entirely safe and his centre is prone to being undermined by ...e6 or ...f5. A) 8.Bc4

At first glance this move appears logical enough, but it does nothing to hinder Black’s plan of pressing the d4-pawn by means of ...Nc6 and ...Bg4. Thus, this move cannot pose serious problems to Black, and White may have difficulty protecting his centre. Despite all that, GM Hillarp Persson has scored 4/4 with it, and some other strong players have tried it from time to time.

750

8...0-0 Be sure to avoid 8...Bg4?? 9.Bxf7†, which has occurred in several games. 9.0-0 9.Be3 Bg4 10.0-0 Nc6 transposes to our main line. 9.Rb1?! can be found on page 418 – see 9.Bc4?! in the notes to variation E. Investing a tempo to avoid the pin deserves attention: 9.h3!? Hillarp Persson has always preferred this plan. However, we now have a slightly unusual version of variation B below, in which White plays an early h2-h3 but generally follows it with a move other than Bc4. 9...Qc7 The more popular 9...Nc6 is certainly playable, but I prefer to threaten White’s bishop. 10.Bb3 I also considered 10.Be2 b6 11.0-0 Bb7 12.Qd3 as played in Istomin – Gomez Celdran, corr. 2008. Here I suggest 12...Nd7N 13.Be3 Rac8÷ with mutual chances. 10...b5!? 11.0-0 Bb7 12.Qe2 This position was reached in Erdei – L. Szabo, Hungary 2005. I suggest: 12...a6N 13.a4 Nd7 14.Re1 e6÷ Reaching a balanced position where the main battle lies ahead. 9...Nc6 10.Be3 Strengthening the d4-square is White’s most natural approach, although he has tried some other moves. 10.e5 cxd4 11.cxd4 Bg4 12.Be3 transposes to the main line below. 10.Bb2?! has the same idea as our main line, except White relegates his bishop to a worse square. 10...cxd4 11.cxd4 Bg4 12.e5 (12.Rb1 Bxd4 13.Bxd4 Bxf3 14.Qxf3 Qxd4 15.Bd5 e6 16.Rfd1 occurred in Andrzejewski – Kolodziejski, corr. 2008, when 16...Qa4N 17.Bxc6 bxc6µ would have left White with insufficient compensation for a pawn) 12...Bxf3 13.gxf3 e6 14.f4 This occurred in Sopow – Perkovics, Hungary 2009, when 14...Qh4!Nµ would have been most accurate; White’s pawn weaknesses and exposed king will cause him no end of problems. 10.Bd5 cxd4 11.Bxc6 (11.cxd4 Bg4 12.Be3 transposes to 11.Bd5 cxd4 12.cxd4 in the notes to the main line below) 11...bxc6 12.cxd4 c5 13.Be3 Bg4³ gave White problems with the d4-square in Levkovsky – Gagnon, Montreal 2011. 10.dxc5 is not terrible although White can hardly fight for the advantage in this manner. 10...Qc7

751

11.Be3 Bxc3 12.Rb1 occurred in Lindstrom – Welford, email 2009, when 12...Bg7N 13.Qc1 Bg4= would have given Black a comfortable position. 10...Bg4 Black develops efficiently, creating pressure against d4 and forcing White to switch to defence – although it is too early to claim that Black has any tangible advantage. 11.e5 White shields the d4-pawn but weakens the square in front of it. Thus, White’s centre loses its elasticity and the d5-square becomes an inviting outpost, while the d4-pawn remains backward and potentially vulnerable to a frontal assault. 11.Rb1?! transposes to the 9.Bc4?! line in the notes to variation E on page 418. 11.d5 is probably White’s best bet but it has no independent significance, as 11...Ne5 12.Be2 leads to variation D2; see page 415 for the continuation. 11.Be2? simply lost a pawn afte11...Bxf3 12.Bxf3 cxd4 13.cxd4 Bxd4µ in Musayev – Agmanov, Pavlodar 2014. 11.dxc5 is harmless. 11...Qc7 12.Rb1 Ne5 occurred in Litvinov – Ganbold, Vung Tau 2008, when White clearly should have played 13.Be2N, although 13...Nd7 14.c6!? (or 14.Qd5 b6!) 14...Qxc6÷ leaves Black without a shadow of a problem. 11.Bd5 gives Black an easy route to a risk-free edge: 11...cxd4 12.cxd4 Bxf3 13.Qxf3 (13.gxf3? Na5 14.Rb1 e6 15.Bxb7 Rb8 16.Ba6 Rxb1 17.Qxb1 Nc6 18.Rd1 Nxd4 19.Bxd4 Bxd4µ gave Black a serious advantage in Ponting – Semenov, email 2012, due to his better pawn structure and safer king; the opposite-coloured bishops only make the coming attack more dangerous)

752

13...Nxd4 14.Bxd4 Bxd4 15.Rad1 Qb6 16.Rb1 Qa5 17.Bxb7 Rab8 18.Bd5 e6 19.Rxb8 Rxb8 20.Bb3 Qc5 21.g3 a5ƒ Nigalidze – Kulaots, Plovdiv 2012. Although the position is close to equal, Black retains some initiative due to his more active pieces, especially his bishop, which exerts pressure against f2 whereas its counterpart is firmly blocked by the e6-pawn. Black’s plans should include ...a5a4, seizing space and making the a2-pawn more vulnerable – especially if Black can land his rook on b2. 11...cxd4 12.cxd4 Rc8 Black develops the rook while threatening to win a pawn via X-ray by ...Nxe5. Although White will defend against this simple threat, it is clear that his central pawns are no threat, and Black has a number of promising ideas including conquering the d5-outpost and advancing his queenside pawns. 13.Rc1 13.Be2 does not change the essence of the position. 13...Be6 14.Qd2 Bd5 15.Rfc1 occurred in De Virgilio – Petzold, corr. 2012, when 15...Qd7N³ would have left Black with the better game thanks to his more flexible pawn structure and the excellent bishop on d5. 13.Bb3 was tried in Geurtsen – van Tuyl, Utrecht 1998, and a few other games. 13...e6!N seems to me to be the most logical reply. My illustrative line continues 14.h3 Bxf3 15.Qxf3 Nxd4 16.Qxb7 Bxe5 17.Rad1 Qf6 18.Qxa7 Ra8 19.Qc5 Rfc8 20.Qb4 Rab8 21.Qa5 Nxb3 22.axb3 Rxb3µ when Black enjoys a healthy extra pawn.

753

13...e6 14.Be2 Qd7 15.h3 Bf5! Black rearranges his minor pieces. The bishop heads for d5 (or c6) and the knight will go to f5, putting pressure on the e3-bishop and d4-pawn. 16.Qd2 Be4 17.Qb2 Ne7 18.Nd2 Bc6

White’s problem is that his central pawns have lost all dynamism and his pieces are unable to exploit the d6- or f6-outposts. Moreover, Black can create a passed pawn on the queenside in the long run, so endgames are likely to favour him. 19.Bd3? 754

This allows Black to increase his domination. 19.Rfd1 was preferable, although 19...Nf5 20.Bf3 Bxf3 21.Nxf3 Rxc1 22.Rxc1 Rc8ƒ still leaves White facing the problems mentioned in the previous note. 19...Bxe5! 20.dxe5 Qxd3 21.Bxa7 Material remains equal but Black’s pieces are much more active and White’s kingside is weak. 21...Nf5! Introducing the tactical possibility of ...Bxg2 followed by ...Nh4† with a formidable attack. 21...Nd5!? is a promising alternative; for instance, 22.Bc5 Rfd8 23.Rc4 g5 24.Rg4 h6 25.Re1 Nf4ƒ and Black’s initiative is rather unpleasant. 22.Rfd1? After this further error, there is definitely no saving White’s position. 22.Bb6?! allows 22...Bxg2 23.Kxg2 Nh4† 24.Kh2 Qxd2 25.Qxd2 Nf3† 26.Kg3 Nxd2 when Black has excellent chances to realize his extra pawn. 22.Rc3! was more resilient. 22...Qd5 23.Nf3 Ra8 24.Ra3 Rfd8µ Black remains clearly better due to his more active pieces, but there is no decisive breakthrough in immediate sight. 22...Rfd8! The immediate 22...Bxg2 is less convincing for tactical reasons but Black can simply strengthen his position in preparation for it, as there is nothing much that White can do to defend. 23.Bb6 No better is 23.Nb3 Qxd1† 24.Rxd1 Rxd1† 25.Kh2 Bd5 26.Qe2 Bxb3 27.axb3 Rcc1–+ with a decisive attack. In Darvas – Bobarnac, corr. 2011, Black missed the chance to strike a decisive blow:

755

23...Bxg2!N 24.Kxg2 Nh4† 25.Kh2 Qd5 26.Rg1 Rxc1 27.Qxc1 Qxd2 28.Qxd2 Rxd2–+ Further material losses for White are inevitable. B) 8.h3!?

This is one of the most important sidelines, which has been used by many strong players including Korchnoi, Svidler, Kramnik and Yu Yangyi. The idea is obviously to prevent ...Bg4 while avoiding disclosing any future plans for now. 8...0-0 9.Be2 9.Be3 Qa5!? (9...Nc6 is not so clear after 10.Bc4!?, although Black should ultimately be fine here 756

too) 10.Bd2 (10.Qd2 has been covered on page 380 – see 10.h3?! in the notes to variation B3 of the previous chapter) 10...Qa3 11.Be2 b6 12.0-0 cxd4 13.cxd4 occurred in Zhao Jun – Zeng Chongseng, Tianjin 2018, when Black traded bishops on a6. Instead I propose:

13...Bb7N 14.d5 Na6÷ With a complex middlegame with approximately equal chances. 9...b6!? Since White has prevented ...Bg4, it is logical to develop the bishop to b7 to put pressure on e4. The engine approves of this plan and there are no long theoretical lines attached to it. In his video lecture, The Grünfeld according to Svidler, the super-GM recommends the sharp 9...cxd4 10.cxd4 Nc6 11.Be3 f5!?, when all kinds of pathways are possible. This is theoretically fine, but it seems unnecessary to prepare such a complicated reply against a sideline when the simple alternative is just as good. 10.Be3 This is the usual choice among strong GMs. Some other games have continued with: 10.0-0 cxd4 11.cxd4 Bb7 12.Qd3 Ba6 Black wins a pawn, although White should obtain sufficient compensation to maintain the balance. 13.Qe3 Bxe2 14.Qxe2 Bxd4 15.Nxd4 Qxd4 16.Bh6 The alternative is 16.Bb2 Qa4 17.f4, as played in Perez Carillo – Corti, Argentina 2000, when 17...Nd7!N would have been the most flexible choice, enabling the knight to go to c5 or f6 according to circumstances. Play may continue 18.f5 Rac8 19.Rad1 Rc2 20.Rd2 Rfc8 21.Rf2 Rxd2 22.Qxd2 Nf6 23.fxg6 hxg6 24.Qh6 Qd1† 25.Rf1 Qh5³ when White does not have full compensation for the pawn. 16...Rc8 17.Rac1 Nd7 18.Rxc8† Rxc8 19.Rd1 Qa4 20.Qg4 757

This position was reached in Nyback – Kulaots, Finland 2018. Black’s best approach would have been to return his extra pawn to exchange queens:

20...f5!N 21.exf5 Qxg4 22.hxg4 Nf6= With a balanced endgame.

10...Bb7 11.Qd3 This is the most popular and logical continuation. 11.Qc2?! enables Black to target the awkwardly-placed queen. 11...cxd4 12.cxd4 Qc8 13.Qb1? (13.Bc4N improves although 13...b5 14.Bd3 Qxc2 15.Bxc2 Nc6³ leaves Black with a lead in development along with easy pressure on the d4-point) 13...Bxe4!–+ White’s position had already collapsed in Altantuya – Stankovic, Veliko Gradiste 2019. 758

11...e6 12.0-0 cxd4 13.cxd4 Nc6 14.Rfd1 Qd6 15.Rac1 We have been following Bluebaum – Van Wely, Brest 2018. I think it would have been logical for Black to fight for the c-file with:

15...Rfc8N 16.Qd2 Ne7 17.Bd3 Rxc1 18.Rxc1 Rc8= Black will exchange the remaining rooks and can look towards a future endgame with confidence. C) 8.Be2

This quiet developing move, which was first played by Albert Becker against Ernst Grünfeld himself at Vienna 1922, should not be too theoretically challenging. Nevertheless, it has been used on an 759

occasional basis by Karpov, Andreikin, Grischuk and others. 8...Nc6 The effectiveness of this move in the present variation explains why 8.Rb1 (as covered later in the chapter in variation E) is such a popular choice. Ideally White would like to meet the knight move with d4-d5, and moving the rook out of harm’s way obviously changes things in that scenario. White’s most popular continuation has been the solid but harmless C1) 9.Be3, defending the centre. However, in spite of the above point about the rook being vulnerable on a1, the aggressive C2) 9.d5!? deserves careful consideration. White risks being worse after: 9.Rb1?! cxd4 10.cxd4 Nxd4 11.Nxd4 Qxd4 12.Bb5†

12...Kf8! With this move you can fight for victory. Avrukh evaluates it as too dangerous, but I believe Black can consolidate safely. If Black is content to play it safe, then 12...Bd7 13.Bxd7† Qxd7 14.Qxd7† Kxd7 15.Rxb7† Ke6 16.Ba3 Bf6 17.Ke2 Rhc8 18.Rc1 Rxc1 19.Bxc1 Rc8 20.Be3 Rc2† 21.Kd3 Rxa2 22.Rxa7= results in total equality. 13.Qe2 Here we can improve on Black’s play in Guthrie – Riveiro, Medellin 1996, by means of: 13...a6!N 14.Bc4 I also checked 14.Bd3 b5 15.h3 (or 15.0-0 Bg4!µ) 15...Be6 16.0-0 h5µ followed by ...Kg8-h7, when Black connects his rooks and remains with an extra pawn. 14...Bg4! 15.f3 Be6 16.Bxe6 fxe6 17.Rxb7 Rc8 Black has returned his extra pawn in order to seize the initiative, and he will soon be a pawn up 760

again. 18.Be3 Qa1† 19.Qd1 Bc3† 20.Kf2 Qxa2† 21.Kg3 Kf7µ The passed a-pawn will cause White a lot of problems. C1) 9.Be3 Bg4

As is typical for the Grünfeld, White defends the d4-pawn, only for Black to attack it again by threatening to remove a key defender. 10.e5 White has to play this move to avoid losing the d4-pawn. However, the d5-square is permanently weakened and the d4-pawn becomes lonely. The pseudo-active 10.d5? is worse: 10...Bxf3 11.gxf3 Bxc3† 12.Kf1 Bxa1 13.Qxa1 Nd4 14.Bb5† Kf8 15.Bxd4 cxd4 16.Qxd4 f6 17.h4 Qb6–+ White had no compensation for the material damage in Shurygin – Lupulescu, Albena 2013. 10.Rb1 gives Black a pleasant choice: 10...0-0 transposes to 10.Be3 Bg4 in the notes to variation E on page 421; and the immediate 10...Bxf3 certainly can’t be bad for Black either. 10...cxd4 11.cxd4 Qa5† Black prepares for an endgame, where his plans include bringing a rook to d8 for even greater pressure on the d4-pawn. It is becoming clear that the opening has turned out somewhat favourably for Black. 12.Qd2 761

Agreeing to a slightly inferior endgame position. 12.Kf1?! can hardly satisfy White. 12...Rd8 13.Qb3 Bxf3 14.Bxf3 Nxd4 15.Bxd4 Rxd4 16.Qxb7 0-0 17.Qxe7 Bxe5 18.Re1 occurred in Bou Janoher – Badosa Romanyo, Girona 2004, when Black should have played:

18...Bd6!N 19.Qf6 Rd2–+ Decisively aiming at the weak pawns on f2 and a2. 12.Bd2!? Qc7 also gives Black good chances, for instance: 13.Rc1 (another game continued 13.Be3 Rd8 14.Qa4 Qd7 15.Rd1? Nxe5 16.Qxa7 Nxf3† 17.gxf3 Bh3–+ when White’s kingside weaknesses were the deciding factor in Sallwey – Koronowski, email 2013) 13...Rd8 14.Qa4 0-0 15.Be3 Qd7 16.Bb5 In Tishin – Zinchenko, Alushta 2006, Black should have played:

762

16...Bxf3N 17.gxf3 Qh3 18.Bxc6 bxc6 19.Qxc6 Bh6!? (Avrukh recommends 19...Rc8 which also gives Black chances to press, but I like the text move even more) 20.f4 Rb8 21.Qe4 Rb2ƒ Black has a nasty initiative for a minimal material deficit. 12.Nd2 Bxe2 13.Qxe2 Qc3 14.0-0 Nxd4 15.Bxd4 Qxd4 16.Nf3 Qa4 17.Rac1 occurred in Istratescu – Gupta, Differdange 2008. It’s hard to guess why a grandmaster rated above 2600 chose such a variation as White, although in the game he obtained enough activity to regain the lost pawn and make a draw. Black could have maximized his chances with:

17...Qa5!N 18.Rfd1 (18.e6 f5! also favours Black) 18...e6³ White’s active rooks give him a certain amount of compensation for the pawn, but not enough for full equality.

12...Qxd2† 13.Nxd2 763

Another game continued 13.Kxd2 Rd8 14.Kc3 0-0 15.Rac1 Rd7 16.h3 Be6 17.Bc4 Bxc4 18.Kxc4 Rfd8³ when Black was the only one who could play for a win in Hafner – Rakay, corr. 2009. 13...Be6! Avrukh quotes a game in which Black exchanged on e2 and was easily equal, but I think he can achieve more by rerouting the bishop to d5. 14.Bf3 14.0-0 occurred in Pyrich – Verlinde, corr. 2015, when 14...0-0-0!?N 15.Nf3 Bd5³ would have given Black a typical edge due to his better pawn structure. 14...Rd8 15.Nb3 It is inadvisable for White to trade off a bishop with 15.Bxc6†?! bxc6 16.Nb3 as occurred in Chojnacki – Plat, Poronin 2017. Here I would suggest 16...Bd5N 17.0-0 f6³ when Black has a pleasant advantage due to the two bishops, while having an isolated pawn on c6 is not so terrible as it is reinforced by the light-squared bishop. 15...Nb4 16.0-0 Black can fight for an edge with the following improvement:

16...Bd5!N After 16...Nc2 17.Rac1 Bxb3 18.axb3 Nxd4 19.Bxd4 Rxd4 20.Rfe1! White held a draw easily enough in Gluhov – Horvat, corr. 2014. Black is in no danger whatsoever here, so you could certainly play this way if you desire an easy draw with no risk. 17.Nd2 Bxf3 18.Nxf3 Nd5 19.a4 Ra8!? 764

The rook could prove useful here in the event that the a-file is opened. 20.Rfc1 b6 21.a5 0-0ƒ White should be able to hold a draw, but Black’s strong knight and ability to mobilize his pawn majority assure him of slightly better chances. C2) 9.d5!?

This is the more challenging continuation, fighting for the initiative. 9...Bxc3† 10.Bd2 Bxa1 11.Qxa1 Nd4 12.Nxd4 cxd4 13.Qxd4 The forced sequence has resulted in a position where White has compensation for the sacrificed exchange due to the weakness of the dark squares. Nevertheless, with accurate play Black can neutralize White’s initiative and fight for the advantage. 13...f6!? Avrukh recommends 13...0-0, which is also quite playable and has yielded good results for Black. However, I like Black’s chances after the text move. 14.0-0 This move has scored well for White, but Black’s play can be improved. I checked two other aggressive continuations: 14.h4!? The idea of h4-h5 certainly looks quite threatening, but Black has time to organize a good defence. 14...Qb6

765

15.Qd3 Another game continued 15.Qa1 Bd7 16.0-0 and now in Bozinovic – Rade, Pula 2000, Black could have played 16...0-0!N without fear. For instance: 17.Rb1 Qc7 18.Qd4 Rac8 19.Qxa7 Qc2 20.Rd1 Qxe4 21.Bf3 Qa4 22.Qxb7 Rb8 23.Qc7 Rfc8 24.Qg3 Qxa2 25.h5 Rb1!–+ White’s attack is at an impasse and Black has a decisive advantage. 15...Bd7 16.h5 Kf7 17.0-0 Rac8 18.Be3 It is worse to play 18.Rb1?! Qc7 19.Rb2 b6 20.Bd1 as in Moutousis – Kouvatsou, Katerini 1993, when 20...Rhd8Nµ would have left White with clearly inadequate compensation. We have been following Kirschbaum – Konopka, Germany 2009. Black’s most accurate continuation looks to be:

18...Qc7N 19.Bd4 e5 20.dxe6† Bxe6 21.Qf3 Qe7³ 766

It is not clear how White can continue his attack, while Black has a natural plan of invading along the c-file. Another active try is: 14.e5 0-0 White has a few possible continuations but Black should remain better with accurate play.

15.0-0 This has been tried at GM level. 15.exf6 exf6 16.0-0 occurred in Dill – Stijve, Lenzerheide 2018, when 16...Qb6N 17.Qf4 g5! 18.Qg3 Bf5 19.Bc3 Rac8µ would have enabled Black to fight off the attack while keeping his extra material. 15.Bc4 looks interesting and active. However, after 15...b5 16.Bb3 a5 17.d6† e6 18.0-0 as played in Plachetka – Pribyl, Trnava 1979, Black can continue 18...a4!N 19.Bc2 fxe5 20.Qxe5 Qf6 21.Qe3 Bd7µ when White’s attack is not dangerous, and Black’s material advantage should eventually make itself felt. 15...fxe5 16.Qxe5 Qd6 17.Qb2 17.Qe3 Bd7 18.Bc3 Rac8 19.Be5 occurred in Danner – Rytshagov, Pula 1997, and now 19...Qc5!Nµ would have White struggling to find compensation. The text move was played in Lputian – Tukmakov, Lvov 1984. Here I recommend:

767

17...a5!?N It is useful to prevent Bb4. Many continuations are possible but it seems to me that White has a hard time proving his compensation. I will mention one illustrative line: 18.Bh6 Rf7 19.Qb5 Bf5 20.Qxb7 Rb8 21.Qc6 Rc8 22.Qb5 g5! 23.Bxg5 Rg7 24.Bh4 Be4ƒ Now Black is the one who is developing activity and fighting for the initiative on the kingside. 14...Qb6 It is useful to develop the queen while gaining a tempo. 15.Qd3 I also analysed: 15.Qa1 Bd7 16.Bc3 (16.Rb1 Qc7 17.Bb4 Kf7 18.d6 was seen in Mikrut – Kosiorek, Rudnik 2000, when 18...Qc2!N 19.Bd1 Qxe4 20.dxe7 Rhe8 21.Rb3 Rxe7 22.Bf3 Qe5 23.Qd1 Rae8 24.Bxe7 Rxe7 25.Re3 Qb5µ would have left Black with a healthy extra pawn) 16...Rc8 17.Bd4 Now in Dziuba – Ipatov, Moscow 2009, the most accurate queen move would have been:

768

17...Qb4!N 18.Rb1 Qa3 19.Be3 b6µ It is not easy for White to develop his initiative, and Black’s extra exchange may come into its own in the future.

15...Bd7 16.Rb1 Qc7 17.Rc1 Qd6 18.Rb1 In the event of 18.Qb3, as played in Mohr – Kuemin, Switzerland 2004, Black should give up a pawn as follows, in order to activate his rook:

769

18...Rc8!N 19.Rb1 b5 20.Bxb5 Bxb5 21.Qxb5† Kf7µ White’s compensation for the exchange is clearly insufficient. We have been following Florez Castro – Cortes Orihuela, Medellin 2017. Here it is worth preventing Bb4, even at the cost of a pawn:

18...a5!N 19.Qd4! This is White’s most stubborn continuation although Black remains somewhat better after it. 19.Rxb7 looks natural but 19...Rb8 20.Rxb8† Qxb8 21.Bxa5 Kf7µ leaves White struggling, despite having a bishop and two pawns versus a rook. He has spent valuable time collecting pawns on the queenside, and Black will develop a lot of activity along the queenside files. 770

19...b5 20.f4 e5 21.fxe5 Qxe5 22.Qxe5† fxe5 23.Bxb5 Bxb5 24.Rxb5 Rf8 25.h3 Rf7³ White keeps some compensation for the exchange, but he will still have to work for a draw. D) 8.Bb5†

White develops the bishop with tempo, trying to break the harmony of the black pieces. If the knight comes to c6 then White will have the option of exchanging it in order to ease the pressure on d4 and weaken Black’s structure. The Belgian master Colle was the first to play this way, against Gerbec at Bad Oeynhausen 1922. Later the same year, Kostic played it against Grünfeld in a high-level game. The variation remained a rarity until the 1970s, when Stein, Tukmakov, Kholmov and Alburt helped to popularize it. In modern times it has been played by Kramnik, Dubov, Sasikiran, Fressinet and Eljanov, as well as Anand, who employed it against Gelfand in Game 1 of their 2012 World Championship match. 8...Nc6 This has been the usual choice among top players, although Black’s other two plausible moves are quite playable as well. White has two main options: D1) 9.d5 and D2) 9.0-0. D1) 9.d5 This seemingly aggressive move usually leads to an equal endgame. 9...Qa5 10.Rb1 Bxc3† 11.Bd2 a6

771

12.Bxc6† The alternative is much worse: 12.Ba4? Bxd2† 13.Nxd2 b5 14.dxc6 bxa4 15.0-0 15.Qc2 is no better. A good example continued 15...0-0 16.0-0 Be6 17.Nc4 Bxc4 18.Qxc4 Rac8 19.Rb7 e6 20.Rd1 Rxc6 21.h4 Rd8 22.Rxd8† Qxd8 23.Qxa4 Rb6 24.Rxb6 Qxb6–+ and the queen ending was hopeless for White in Privitera – Kaufman, Budapest 2010. 15...Be6 16.f4 16.Qc2 is met by 16...Rc8 17.Nf3 0-0 18.Rb7 Rxc6 19.Rxe7 as in David – Erdos, Baile Govora 2017, and now 19...Rd6N 20.e5 Rb6µ would have been clearly better for Black thanks to the passed c-pawn and the superiority of the bishop over the knight. We have been following Lavrov – Kedrov, Moscow 1996. White’s last move looks natural but Black has a powerful idea at his disposal:

772

16...0-0-0!N 17.Rb2 It is important to consider the aggressive: 17.Rb7 Bxa2 Preventing Qb1 is the priority. 18.Rf2 c4 19.Qa1 c3 20.Nf1 Bc4–+ Despite the fact that the rook and pawn have lodged themselves deep in Black’s camp, there is no attack. Thus Black enjoys a decisive advantage due to his extra material and the harmonious arrangement of his pieces. 17...Rd4 18.f5 gxf5 19.exf5 Bd5 20.Qe2 Rg8 21.g3 Rb4 22.Rc2 Qb6 23.Qxe7 c4† 24.Rf2 Rd8 25.a3 Rb3 26.Qb7† Qxb7 27.cxb7† Kxb7 28.Nxc4 Rb1† 29.Rf1 Rxf1† 30.Kxf1 Be4 31.Rc3 Rd3 32.Rc1 Bxf5µ Black has excellent winning chances because of the weak pawn on a3 and the superiority of the bishop over the knight in such an open position. 12...bxc6 13.dxc6 Be6 14.Rc1 Bxd2† 15.Qxd2 Qb4! After this exact move, an equal endgame is practically inevitable. 16.Qxb4 16.0-0 does not change matters: 16...Qxd2 17.Nxd2 0-0-0 18.Nb3 Bxb3 19.axb3 Rd3 20.Rxc5 Rxb3 21.Ra5 Rb6 22.Rfa1 Rd8 23.h3 Rxc6 24.Rxa6 Rxa6 25.Rxa6= reached an absolutely equal endgame in Sasikiran – Grischuk, Eilat 2012. 16...cxb4

773

17.c7 Kd7 18.Ne5† Kc8 19.Nc6 Kxc7 20.Nxb4† Kd6 21.Ke2 Rhc8 22.Rhd1† Ke5 23.Nc6† Kf6 24.a3 Rc7 25.Rc3 Rac8 26.Rdc1 g5= A draw was soon agreed in Martinovic – Bosiocic, Pozega 2018. D2) 9.0-0

This is the more popular sequel. 9...0-0 10.Be3 White naturally develops the bishop and reinforces the d4-point.

774

10.Ba3 cxd4 11.Bxc6 bxc6 12.cxd4 is analysed below under 10.Bxc6 bxc6 11.Ba3 cxd4 12.cxd4. 10.d5?! leads to difficulties for White after: 10...Bxc3 11.Bh6 Nd4 12.Rc1 (12.Nxd4 Bxd4 13.Bxf8 Qxf8 14.Rb1 a6 15.Be2 b5³ gave Black more than enough compensation for the exchange in Gerzhoy – So, Montreal 2012, due to the powerful dark-squared bishop and the mobile queenside pawns) 12...Nxb5 13.Bxf8 Qxf8 14.Qd3 Bd7 15.a4 In Chuprov – Gavryushin, Kurgan 2001, Black could have obtained some advantage with:

15...Qe8!N 16.axb5 Bb4 17.Qe3 Bxb5 18.Rfd1 Ba4³ Materially Black is no worse, and he has the long-term advantages of the bishop pair and three connected passed pawns on the queenside. White can also try to justify the position of the bishop on b5 by exchanging it: 10.Bxc6 bxc6 11.Ba3 cxd4 12.cxd4 12.Nxd4? makes no sense: White leaves himself with a weak c3-pawn and Black gets a big advantage after playing energetically. 12...c5! 13.Bxc5 Qc7 14.Nb3 Ba6 15.Re1 Bxc3 16.Rc1 Bxe1 17.Qxe1 Rfd8–+ Bonin – G. Shahade, New York (rapid) 2002. 12...Bg4 13.Bc5 13.Rc1?! is risky. 13...Qa5 14.Qd3 Bxf3 15.Qxf3 Bxd4 16.Rxc6 Rac8 17.Rxc8 Rxc8 18.Bxe7 occurred in Babychuk – Hess, corr. 2009, when 18...Qxa2N³ would have given White reason to worry about the passed a-pawn. 13...Re8 Now White must decide how to position his queen’s rook. 14.Rb1!? The more popular option, which sets a trap. 14.Rc1 can safely be met by 14...Bxf3 15.Qxf3 Bxd4 16.Rfd1 e5 17.Bxd4 exd4 18.Rxc6 Rc8 19.Rxc8 Qxc8 20.h4 (or 20.Rxd4 Rxe4=) 20...Qc4 21.h5 Qxa2 22.Rxd4 Qb1† 23.Kh2 Qb8† with equal play, Alburt – Kudrin, New York 1986.

775

14...Qd7 14...Bxf3?! is pointless here, as 15.Qxf3 Bxd4? 16.Rfd1 e5 17.Rb7+– reveals the purpose of the rook on b1. 15.Rb4 a5 16.Ra4 Qb7 17.Qb1 17.Qb3?! was the choice of a chess legend but he ran into difficulties after 17...Qb5! 18.Ng5 Qxb3 19.axb3 e5 20.f3 Bd7 21.dxe5 Rxe5 22.Be3 Rb5µ in Korchnoi – Sax, Reykjavik 1988. Black has two powerful bishops and the ability to create a passed pawn on the queenside in the future. 17.h3!? is reasonable, when 17...Bxf3 18.Qxf3 Qb5 19.Qb3 e5 20.f3 occurred in Hlavica – Stancl, Czech Republic 1999. I suggest 20...Bh6N 21.Rd1 Be3† 22.Kh1 Red8 23.Qc2 Rab8 24.Rc4 exd4 25.Bxd4 Bxd4 26.Rcxd4 Rxd4 27.Rxd4 Qe5= when the position is essentially equal, although White has to be slightly more careful due to his vulnerable first rank and the passed c-pawn. 17...Qb5 18.Ra3 We have been following A. Mastrovasilis – Safarli, Patras 2019. This would have been a good moment to eliminate the knight on f3 to facilitate a central pawn break:

776

18...Bxf3N 19.Rxf3 e5 20.Rb3 Qa4 21.dxe5 Rxe5÷ Black is no worse: his excellent pieces are just as relevant as his extra pawn island.

10...Bg4 Once again, White must solve the problem of the pressured d4-pawn. Black’s pieces are arranged harmoniously and White has little chance of emerging from the opening with an advantage. 11.d5 This leads to more complex positions than other options. White captures space and removes the pawn from the fire of the enemy pieces. On the other hand, the central pawns lose their elasticity and the bishop on g7 becomes even stronger. 777

11.dxc5 can hardly cause problems: 11...Bxc3 12.Rc1 Bg7 13.h3 (13.Bxc6?! bxc6 14.Qa4 Bxf3 15.gxf3 Qc7 16.Rb1 Be5 17.h3 Bf4 18.Bd4 Rfd8ƒ left White under some pressure due to his compromised pawn structure in Franic – Rade, Bol 2013) This occurred in Gelashvili – Sutovsky, Minsk 2017, and here I suggest:

13...Bxf3N 14.Qxf3 Nd4 15.Bxd4 Qxd4 16.Rfd1 Qe5= With equal play. White can also relieve the pressure on his centre at the cost of the bishop pair with: 11.Bxc6!? bxc6 12.Rc1 (12.Rb1?! is well met by 12...cxd4 13.cxd4 Qd7 14.Rc1 Rac8³ followed by ...Rfd8 with pressure on the d4-pawn) 12...Qa5 13.Qc2 (13.Qe2?! Rfd8 14.Rfd1 Bxf3 15.gxf3 cxd4 16.cxd4 Bxd4! 17.Bxd4 Rxd4 18.Rxd4 Qg5† 19.Kf1 Qxc1† 20.Qd1 Qc5µ left Black with a healthy extra pawn and good winning chances in M. Ivanov – Khalifman, Minsk 2012) 13...Bxf3 14.gxf3 f5 15.exf5 Rxf5 16.Qb3† Kh8 17.Qb7 In Pechwitz – Hechler, corr. 1999, Black should have continued:

778

17...Raf8N 18.dxc5 Qa4 19.Qxe7 Rxf3= White’s extra pawn is not felt because it is doubled; moreover, White’s king is somewhat exposed.

11...Ne5 12.Be2 White should retreat his bishop to prevent his pawn structure from being ruined. The greedy 12.Bxc5? backfires after 12...Nxf3† 13.gxf3 Bh3 14.Re1 Bxc3 15.Rc1 Bxe1 16.Qxe1 Rc8– + when White finds himself the exchange down in a bad position. 12.Rc1? is also asking for trouble. 12...c4! 13.Bd4 Bxf3 14.gxf3 was played in S. Martin – Schulze, Germany 1997, when Black could have obtained a great advantage with:

779

14...Qc8!N 15.f4 Nd3 16.Bxg7 Kxg7 17.Rb1 a6 18.Ba4 Qh3 19.f3 b5 20.Bc2 Nxf4–+ Black has an extra pawn and a dominant knight, while White’s bishop is terribly passive. 12...Nxf3† 13.Bxf3 Bxf3 14.Qxf3

14...Qa5 Defending the c5-pawn while attacking c3. 15.Rac1 Passive protection with 15.Bd2?! turns out worse: 15...Rad8 16.Qe3 (16.Qd3 occurred in Mollov – Berebora, Balaguer 1996, when 16...Qa4 17.Rab1 c4 18.Qe2 Rd7ƒ would have given Black some 780

initiative, as the c3- and a2-pawns are weak) 16...b6 17.Rfc1 Qa4 18.Qd3 e6 19.Bg5 f6 20.Be3 f5 21.f3 fxe4 22.fxe4 exd5 23.exd5 Rf5 24.d6 Be5–+ In Kaygisizer – Muttoni, email 1999, White was losing the d6-pawn while remaining with weaknesses on c3 and a2. 15...b6 16.Rc2 The position is close to equal, but somewhat easier to play for Black. The weak pawns on a2 and c3 require attention, while White cannot do much with his central majority. Black also has the option of opening the centre with a timely ...e6. 16.Qe2?! Bxc3 17.Rfd1 was an attempt by White to sacrifice a pawn to improve the activity of his pieces in Nowak – Lyashenko, corr. 2007. However, White’s compensation is not sufficient after the accurate continuation:

17...Rac8!N 18.e5 Bd4! 19.d6 exd6 20.exd6 Bxe3 21.Qxe3 Qxa2³ The passed pawn on d6 gives White some compensation, but not enough for full equality. 16.Bg5 Rfe8 17.Rc2 Qa4 18.Rfc1 h6 19.Be3 occurred in Hlavica – Galvin, email 2014, and here I suggest: 19...Qc4N³ Black can look to open the centre with ...e6, while White is deprived of counterplay since his rooks do nothing but defend the weak c3-pawn. 16...Qa4 17.Rfc1 Qc4 18.Bg5 f6 19.Bf4 e5 19...e6!?N is also worth considering. For instance, 20.Re2 Rae8 21.Rce1 Rf7 22.h4 Rfe7 23.dxe6 Rxe6ƒ and Black enjoys the more compact pawn structure, whereas White has weaknesses on e4, c3 and a2. 20.Bg3 Here it would have been best to activate the bishop. 781

20...Bh6!N 20...f5?! is too fast and active. The game continued 21.exf5 with approximate equality in Ftacnik – Rodshtein, Germany 2013, but the more precise 21.Rd1!N would have enabled White to fight for the advantage. 21.Re1 Rf7 22.Qe2 I also considered 22.Qd1 b5 23.f3 a5 24.Bf2 Bf8 25.g3 Bd6 26.Kg2 f5ƒ when Black controls the game. 22...Qxe2 23.Rcxe2 b5 24.f3 Rb7 The engine calls the position equal and it should surely be a draw with best play. However, Black’s mobile queenside pawns make the endgame slightly harder to play for White. Here is a brief illustrative line showing how White may get into trouble if he tries to play too actively: 25.a4 b4 26.cxb4 cxb4 27.Rc2 b3 28.Rc6 Rab8 29.Rxf6 b2 30.Rb1 Be3† 31.Kf1 Bd4µ White’s extra pawn is far less significant than the dangerous pawn on b2. E) 8.Rb1

782

This option is dubbed in the literature as the “Modern Exchange Variation”. White’s idea is to remove the rook from the opposition of the g7-bishop and simultaneously exert pressure on the b7pawn. In some lines the rook may go to b5, especially if the black queen has moved to a5. Overall it can be considered an active and dynamic system, where White routinely sacrifices a pawn on a2 or c3 in order to develop his initiative. This variation was first introduced by GM Shamkovich, against Gheorghiu in Cleveland in 1975. Its popularity took off in the 1980s, with top players such as Kasparov and Yusupov playing it, and especially Boris Gelfand who played it many times with generally excellent results. More recently, Aronian, Kramnik, Mamedyarov, Radjabov, Grischuk and Karjakin are just some of the top players who have used it. As you would expect from a variation with such a rich history, it has accumulated a lot of theory and there are numerous pitfalls for both sides. Edward Dearing’s 2005 book Challenging the Grünfeld is a notable work which advocated this system for White. 8...0-0 Black’s best approach is to castle before committing any other pieces. Compared to variation C2 above, 8...Nc6 9.d5! offers a promising sacrifice of the c3-pawn, without including the rook on a1 as part of the bargain. 9...Ne5 is a safer reaction, when White may or may not transpose to our main line. Either way, it is clear that Black gains nothing from delaying castling. 9.Be2 This is the most popular and logical move, developing a piece and preparing to castle. Other options give Black an easier life: 9.Bc4?! allows Black to exert massive pressure on the d4-pawn. 9...Nc6 10.Be3 Bg4 11.0-0 cxd4 12.cxd4 Nxd4 13.Bxd4 Bxf3 14.Qxf3 Qxd4 15.Bd5 e6 16.Bxb7 Rad8 17.Bc6 Qc4 18.Rfc1 Qxa2µ 783

Black had a healthy extra pawn in Vasilev – Todorov, Sofia 2008. 9.h3?! prevents ...Bg4 but is too slow. 9...Nc6 10.Be3 Qa5 11.Qd2 cxd4 12.cxd4 (no better is 12.Nxd4 Rd8 13.Bc4 e6 14.Rd1 Qc7 15.Bb3 as seen in Hahner – Ganz, email 2014, when 15...b6 16.Qe2 Na5 17.Rc1 Bb7µ would have given White no compensation for his poor pawn structure) 12...Qxd2† 13.Kxd2 Rd8 White’s king is too exposed, for instance: 14.Kc3 Be6 15.Rxb7 Rac8 16.Kb2 Nxd4 17.Bxd4 Bxd4† 18.Nxd4 Rxd4 19.Ba6 Ra4–+ With a double attack on the bishop and the a2-pawn, White’s position collapsed in R. Webster – Armbruster, Isle of Man 2004. 9.Bb2 is hardly logical. 9...Qa5 10.Nd2!?N This is perhaps the only chance to justify White’s previous move. (10.Qb3? simply loses a pawn, and 10...cxd4 11.Qb4 Qxb4 12.cxb4 e5 13.Rc1 Be6–+ left White with no compensation in Lenders – Gijbels, Lommel 2000) 10...Qxa2 11.Bd3

11...b5! 12.Bxb5 Ba6 13.Bxa6 Nxa6 14.Qb3 Qxb3 15.Nxb3 cxd4 16.cxd4 Rfb8³ Black’s activity along the b-file and passed a-pawn gives him slightly better chances. 9.Be3?! is poorly timed. 9...Qa5 10.Rb5 (10.Qd2 transposes to 10.Rb1 in the notes to variation B3 of Chapter 13 – see page 382) 10...Qxc3† 11.Bd2 Qa3 12.Ra5 (12.Rxc5 occurred in Pongratz – Voith, Germany 2002, when 12...Bg4!N 13.Bc1 Qxa2 14.Be2 Na6 15.Rc2 Qa5† 16.Bd2 Qa4µ would have left White with no compensation for the missing pawn) 12...Qb2 13.Rxc5 In Ristic – Nestorovic, Paracin 2009, Black could have claimed some advantage with:

784

13...Bg4N 14.d5 Qxa2 15.Ra5 Qb2 16.Rb5 Qf6 17.Rxb7 a5µ The passed a-pawn has great strength, thanks to Black’s control of the long diagonal. 9.Qd2!? This is the most interesting of White’s rare 9th moves, although Black should still be able to obtain a fine game. The idea is to support the c3-pawn in preparation for d4-d5. 9...Nc6 Black has several other decent options, but I am quite happy to provoke the move that White wants to play. 10.d5 Na5! 10...Ne5 11.Nxe5 Bxe5 gives White the option of transposing to variation E3 below with 12.Be2, but he also has the independent possibility of 12.f4. The text move is more attractive: the knight may be transferred via b7 to d6 later, and in the meantime Black is well placed to attack White’s centre.

785

11.Be2 11.c4 e6 12.Bb2? seems like a natural move but is actually a serious inaccuracy. 12...Bxb2 13.Rxb2 exd5 14.cxd5 Re8 15.e5 Bg4 Just like that, White’s centre collapses. 16.e6 fxe6 17.d6 Bxf3 18.gxf3 Nc6 19.Rxb7 Nd4–+ Knuth – Giemsa, Germany 1999. 11.h4!? was played in Kuerten – Merker, corr. 1997. I think the most practical solution is the simple 11...h5!?N, when nothing scary is happening on the kingside and Black will resume his central play on the next move. For instance, 12.Be2 e6 13.d6 e5 14.0-0 b6 15.Rd1 Bb7 16.Qc2 Qd7 17.Bb5 Bc6³ when the passed d-pawn is not going anywhere, and Black has good chances to surround and eventually capture it. 11...e6 12.c4 exd5 13.cxd5 Re8 14.Qc2

14...Rxe4! 15.Qxe4 Bf5 16.Qa4 Bxb1 17.0-0 Bf5 18.Bg5 f6 19.Bd2 b6 20.Bxa5 bxa5 21.Bb5 Rb8µ 786

In Nocke – Borchers, Germany 1995, Black enjoyed the advantage of two bishops, as well as an extra pawn which – despite being doubled – is still of some value.

9...Nc6! I like the idea of provoking White to advance the d-pawn, as the g7-bishop becomes even more functional. 9...cxd4 10.cxd4 Qa5† is a major line, when 11.Bd2 Qxa2 12.0-0 gives White good compensation for the pawn. Black’s position is sound, and certain lines have been analysed all the way to a clear draw. Although this is theoretically fine, especially for elite GMs and correspondence players, I imagine that most readers will not be so keen on the prospect of memorizing a great deal of theory to draw in such a manner. 9...b6 is a popular alternative, intending to develop calmly. For a while I found this approach rather attractive, but I was put off by 10.0-0 Bb7 11.d5! Bxc3 12.h4 (12.Bc4 is more popular and has also scored well for White) as played in Grischuk – Nepomniachtchi, Internet (blitz) 2018, and various other games. White has a dangerous initiative for the sacrificed pawn and it is not so easy to solve Black’s problems. 10.d5 As mentioned earlier, one of the main points of White’s 8th move is to be able to play this energetic move, offering to sacrifice the c3-pawn, while keeping the rook safe. Other continuations are worse, for instance: The pseudo-active 10.Rb5? leads to trouble after 10...cxd4 11.Rd5 Qc7 12.cxd4 b6 13.Ba3 Be6 when White’s position is in disarray and his rook must lose additional time:

787

14.Rb5 Nxd4 15.Nxd4 Qc3† 16.Kf1 Qxd4 17.Qxd4 Bxd4 18.Bxe7 Rfe8 19.Bg5 Bxa2–+ Kubala – Siroky, Slovakia 2001. Black has an extra pawn and the queenside passers should ensure a relatively simple victory. 10.e5?! is strategically undesirable. 10...Be6! 11.Rxb7 occurred in Herning – Aronow, Mittelfranken 2007, and here I can offer an improvement:

11...cxd4N 12.cxd4 Bxa2 13.0-0 Bd5 14.Rb1 a5³ Black’s passed pawn and strong outpost on d5 make his position much easier to play. 10.dxc5?! leads to a bad pawn structure for White after 10...Bxc3† 11.Bd2 Bxd2† 12.Qxd2 as seen in Rojas – Brajnikov, email 1999, when it would have been good for Black to defend the b7-pawn in order 788

to liberate his remaining bishop:

12...Rb8!N 13.h3 Be6 14.Rb2 Qa5 15.Qxa5 Nxa5³ Black has the better chances due to White’s split queenside pawns. 10.0-0?! is a strange way to sacrifice a pawn. 10...cxd4 11.cxd4 Nxd4 12.Nxd4 Qxd4 13.Qc2 (13.Qxd4?! Bxd4 14.Ba3 Bb6 15.Rfd1 occurred in Koziel – Van Gool, Netherlands 1994, when 15...Re8N 16.Bb4 Be6 17.a4 Rac8µ would have left White with no compensation for the pawn) White has some compensation for the pawn due to his superior development and activity along the queenside files. Black should not try to hold on to his extra material, but should instead return the pawn in order to activate his pieces:

13...b6! 14.Bb2 Qc5! 15.Qxc5 bxc5 16.Bxg7 Kxg7 17.Rfc1 Be6 18.Rxc5 Rfc8 19.Rxc8 Rxc8 20.a4 789

Rc3³ In Beltran Marin – Devine, corr. 2013, Black had an endgame advantage due to his more active pieces and White’s vulnerable a-pawn. 10.Be3 This routine developing move enables Black to develop strong pressure against the d4-pawn. 10...Bg4 11.0-0 11.Rxb7 Bxf3 leads to the same thing as White will have to castle soon enough. 11.e5 may be White’s best, although 11...cxd4 12.cxd4 Qa5† 13.Qd2 Rfd8 14.Qxa5 Nxa5 15.0-0 Rac8³ gave Black the more comfortable position in Vasseur – Salati, corr. 2015.

11...Bxf3 12.Bxf3 cxd4 13.cxd4 Bxd4 14.Bxd4 Nxd4 15.Rxb7 e5 16.Qa4 16.Bg4?! causes the bishop to be left out of play. In Peskov – Zemlyansky, Borisoglebsk 2011, Black could have increased his advantage with 16...h5!N 17.Bh3 Qa5 18.Kh1 Rab8µ when Black’s pieces are much more active and the a2-pawn is an obvious target. 16...Nxf3† 17.gxf3 Qg5† The players agreed a draw here in Merker – Floeter, Germany 2004, but Black could have played on:

790

18.Kh1 a5 19.Rd1 Rac8 20.Rbd7 Rc3 21.R7d3 Rxd3 22.Rxd3 Rc8³ Black can combine threats against the enemy king with attacks on the various pawn weaknesses.

10...Ne5 This is the best and most reliable continuation, exchanging the active knight on f3. Unlike the 9.Qd2 line analysed earlier, 10...Na5?! is inadvisable here because White’s queen is not misplaced and he has instead spent a tempo on the far more useful Be2. Accepting the pawn sacrifice is risky: 10...Bxc3†?! 11.Bd2 Bxd2† 12.Qxd2 Na5 13.h4ƒ The knight on a5 is offside and White has more than enough compensation for the pawn.

791

11.Nxe5 Again this is the clearly best continuation. Here is a brief summary of the rare alternatives: 11.0-0?! offers a pawn without any justification: 11...Nxf3† 12.Bxf3 Bxc3 13.Qc2 Bd4 14.Bh6 Re8 15.Rfd1 Qd6 16.g3 b6µ White had no compensation in Hairapetian – Karapetyan, Yerevan 2016. 11.Qd2?! resembles the main line except that here White’s position will be devoid of harmony. 11...Nxf3† 12.Bxf3 Qa5 13.Rb3 c4 14.Rb4 b5 15.0-0 Bd7 16.Qc2 Qc7 17.Be3 a5 18.Rbb1 Rab8³ Black’s mobile queenside pawns gave him the upper hand in Telman – Tukmakov, Simferopol 1983. 11.c4?! is a questionable pawn sacrifice. 11...Nxf3† 12.Bxf3 Qa5† 13.Bd2 Qxa2 14.Qc1 b6 15.0-0 was seen in Martinez Molina – Suarez Rodriguez, corr. 2014, and now it would have been logical to lean on the c4-pawn:

15...Ba6!N 16.Be2 e6 17.Rd1 exd5 18.exd5 Rfe8 19.Bd3 Qa4 20.Bc3 Bxc3 21.Qxc3 Qd7³ White has a hard time demonstrating compensation for the missing pawn. 11.Qc2?! enables Black to develop easy play on the queenside by means of 11...Nxf3† 12.Bxf3 Qa5 13.Bd2 b5 14.0-0 Bd7. A good example continued 15.Rfe1 Qc7 16.Be2 c4 17.f4 and now in Rodrigo da Silva – Fusco, Sao Paulo 2009, it would have been a good moment to challenge the d5-pawn:

792

17...e6!N 18.dxe6 Qb6†! This small finesse enables Black to recapture in the optimal way. 19.Kh1 Qxe6 20.Bf3 Bc6 21.Be3 Rad8 22.Rbd1 Rxd1 23.Rxd1 a5µ Black is clearly better for structural reasons: White’s pawns on a2, c3 and e4 need to be defended and Black can eventually create a passed pawn with ...b4. 11.Bd2 is an attempt to prepare the advance of the central pawns, but 11...f5! is a fine response. 12.Nxe5 Bxe5 13.f4!? Aiming to shut in Black’s dark-squared bishop. (13.Bf3 Qd6 14.h3 occurred in Sabrowski – Figura, Berlin 2016, and now 14...b5N 15.0-0 Rb8³ is slightly better for Black) 13...Bg7 14.e5 Qxd5 15.0-0 Bakos – Sipos, Budapest 2004. Now I like:

15...Be6N 16.Bf3 Qxa2 17.Ra1 Qc4 18.Bxb7 Rad8÷ Black is not worse; White still needs to prove the existence of compensation. 793

A final option is: 11.Nd2 This has occasionally been tried by strong players. The idea is to leave Black with an unstable knight which can be driven away with f2-f4. 11...f5! Once again, this move works perfectly. 12.0-0 12.f4 Nf7 leaves White’s centre rather shaky. 13.e5 Qxd5 14.0-0 Qxa2 15.Bc4 Qa5 16.Nb3 Now in W. Forster – Schaefer, Auckland 2016, it would have been natural and strong to play 16...Qc7N 17.Qd5 b6! in order to activate the light-squared bishop. 18.Qxa8 Bb7 19.Qxa7 (after 19.Qxf8†? Kxf8 20.Be3 Qc6 21.Rf3 Qa4–+ White’s position collapses) 19...Ra8 20.Qxa8† Bxa8 21.Nd2 Kf8³ Black’s queen and pawn are stronger than White’s pair of rooks, although victory is still a long way off. 12...fxe4 13.Nxe4 Bf5 14.f3 b6 15.c4 15.Bf4 c4 makes the position more dynamic, but still somewhat favourable for Black. 16.Qd2 Rc8 17.Rbd1 b5 18.Rc1 occurred in Perikov – Bernal Varela, email 2010. 18...Qb6†N 19.Be3 Qa5 20.Rfd1 Rfd8 21.Bg5 Qa3³ The d5-pawn, detached from base, makes an inviting target for the black pieces. 15...Bxe4 16.fxe4 Rxf1† 17.Bxf1 In Utasi – J. Roos, Budapest 1985, Black could have established a strong dark-square blockade with:

17...Qd6N 18.g3 Rf8 19.Bf4 g5! 20.Bxe5 Qxe5³ The difference in the strength of the bishops is obvious. 11...Bxe5 We will consider E1) 12.Rb3 and E2) 12.Qc2, followed by the main line of E3) 12.Qd2. 12.c4?! has somehow yielded an impressive plus score for White, but 12...Qa5† 13.Bd2 Qxa2 picks up a pawn for questionable compensation. A good example continued 14.0-0 Bd4 15.Bh6 Rd8 and now 794

16.Kh1?! seems like a logical move to free the f-pawn, but White’s attack is simply too slow. In Wilder – D. Roos, New York 1984, Black missed a strong manoeuvre:

16...Rd6!N Activating the rook in an unusual way. 17.Qd3 Rb6 18.Bg5 e6µ White’s compensation is insufficient. 12.Qd3 is sometimes played but the queen is clumsily placed. 12...f5 13.Bf3 In Gajdamowicz – Nawrocki, Poznan 2011, Black missed a strong idea:

13...b5!N 14.Bh6 Rf7 15.0-0 a6³ Black has a flexible pawn structure and White’s pawns on e4, d5 and c3 may become excellent targets in the future. 12.Bd2 is playable but slightly passive. 12...Qd6 13.c4 (13.Qc1 allows Black to strike at the centre: 13...f5! 14.f4 Bg7 15.c4? [White should prefer 15.e5, even though 15...Qxd5 16.0-0 Qd8³ offers 795

insufficient compensation for the pawn] 15...fxe4µ In Sieglen – Bergner, Deizisau 2014, Black was a pawn up and ready to destroy the centre with ...e6 in the near future.) This position was reached in Krupina – Filatov, Yaroslavl 1995, and here I like:

13...e6!?N 14.h4 exd5 15.exd5 a6 16.h5 b5„ Black is at least equal. White’s ability to open the h-file is not too threatening, as her pieces cannot be quickly transferred to the kingside. Some strong players have tried the following risky pawn sacrifice: 12.0-0!? Bxc3

13.Bh6 This was the choice of a talented young Russian grandmaster in a recent game. I checked three other options: 796

13.Rb3 Bg7 14.Rg3 occurred in Pak – Marin, Almaty 2016, when 14...Qc7Nµ would have clearly favoured Black. 13.Qc2 Bd4 14.Bh6 Re8 15.Kh1 e6 16.dxe6 was seen in Lanchava – Arakhamia-Grant, USSR 1990, when 16...Rxe6N 17.Bc4 Rb6µ would have left White without any visible compensation for the pawn. 13.Be3 b6 14.f4 e6 15.Bc4 may be White’s best attempt, as played in Drenchev – Erturan, Albena 2013. I recommend 15...Bg7!N 16.d6 Bd7 17.e5 Bc6³ when White has some compensation for the pawn thanks to the protected passer on d6 and fact that the g7-bishop is, for the time being, shut out of play. Nevertheless, Black has an extra pawn and an obvious plan to mobilize the queenside pawns, so his chances remain somewhat preferable. 13...Bg7 14.Bxg7 Kxg7 15.Qc2 b6 16.Qc3† f6 17.f4 e6 18.Rbd1 exd5 19.Rxd5 Qe7 20.e5 We have been following Esipenko – Wei Yi, Moscow 2019. In the game Black developed his bishop to b7, only to retreat it to c8 a few moves later after a rook exchange. Although he still stood slightly better and went on to win, a natural improvement would be:

20...Be6!N 21.Rd6 Rad8µ White has scant compensation for the pawn, as his pieces are unable to create any real problems. White’s only remaining option is to defend the c3-pawn with: 12.Qb3!? e6 13.f4 Bg7 14.c4 Re8

797

15.e5!?N 15.Qd3? is a serious mistake which allows a thematic tactical strike: 15...exd5 16.cxd5 Rxe4! 17.Qxe4 Bf5 18.Qc4 Bxb1 19.0-0 Bd4† 20.Kh1 Be4–+ In Ivlev – I. Popov, Sochi 2016, White was a pawn down with a miserable position; the d-pawn is also likely to be destroyed in the near future. The text move seems like White’s best chance, although Black gets to smash up the centre anyway: 15...f6! 16.0-0 fxe5 17.fxe5 exd5 18.cxd5 Bxe5 19.Bb2 Bxb2 20.Rxb2 Qd6

21.Bb5 Re4 22.Qf3 Bf5 23.Bd3 Re5÷ The position remains in a state of dynamic equilibrium, although White is the one who is under pressure to demonstrate compensation for the pawn. It is worth noting that the following tempting capture would be a serious mistake: 24.Rxb7? c4! 25.Bb1 Rf8µ 798

The passed pawn on c4 is more dangerous than the one on d5, which remains weak. Moreover, all of Black’s pieces are active and well centralized. E1) 12.Rb3

This somewhat pretentious move has the idea to defend the pawn before playing c3-c4 and subsequently utilizing the rook along the 3rd rank. 12...e6 12...Qc7 has been more popular but I prefer the text move. 13.f4 Bg7 14.c4 14.d6 is conveniently met by 14...Bd7!, threatening ...Ba4 while preparing ...Bc6. This position was reached in S. Ivanov – Mikhalevski, St Petersburg 1999. I propose a novelty:

799

14...Bd7!?N 15.dxe6 One of the ideas behind Black’s last move is that the greedy 15.Rxb7?! will be met by 15...Bd4! when it becomes hard for White to secure his king. 16.Qd3 (I also checked 16.Bb2 Qa5† 17.Kf1 Rfb8 18.Rxb8† Rxb8 19.Bxd4 cxd4 20.Kf2 Qc5 21.Kg3 Rb2ƒ when Black has a nasty initiative for a mere pawn) 16...Re8

White faces a difficult future, especially as the natural-looking 17.Be3? allows 17...exd5 18.cxd5 Bf5! 19.Bf3 Qxd5–+ when White’s position is falling apart. 15...Bxe6 16.Qxd8 Raxd8 17.Rxb7 Rfe8 18.e5 But not 18.Rxa7? Bxc4! 19.Bxc4 Rxe4† 20.Be2 Rde8–+ when White will find it impossible to 800

restrain the rampage of the raging rooks.

18...f6! Black ensures maximum activity for all his pieces. The best that White can do is to steer the game towards an equal ending, as in the following line. 19.Bb2 fxe5 20.Bxe5 Bxe5 21.fxe5 Bc8 22.Rc7 Rxe5 23.Kf2 Be6 24.Rb1 Rf8† 25.Bf3 Bxc4 26.Rxa7 Re2† 27.Kg3 Rxa2 28.Rc7 Ra3 29.Rbb7 Rf7 30.Rc8† Kg7 31.Rxf7† Bxf7 32.Rxc5= With a thoroughly equal position. E2) 12.Qc2

801

This move has been tried by some strong players but it does not pose serious problems for Black. 12...e6 13.f4 Bg7 14.dxe6 The hapless 14.c4?! was played in Hebert – Luksza, Quebec 1994, when 14...exd5!N 15.cxd5 Re8 would have threatened the thematic ...Rxe4 trick. Therefore 16.Rb3 is almost forced, and here Black can strike with:

16...c4! After this powerful injection of dynamism, White’s position becomes difficult. His best chance is to move his rook somewhere along the 3rd rank, after which ...b5 gives Black a great game. Taking the c4-pawn is worse, as shown after: 17.Bxc4?! (17.Qxc4 proves even more unfortunate after 17...Bf5!–+ threatening to take on e4 as well as ...Rac8) 17...Bf5 18.Re3 Bd4 19.Re2 b5 20.Bd3 Rc8 21.Qb1 Bg4 22.Rc2 Qxd5 23.Rxc8 Bxc8–+ White’s problems connected with the king in the centre are too much to cope with. 14...Bxe6

802

15.Rxb7 Taking the pawn seems most challenging, but Black gets plenty of activity and will soon restore the material balance. White preferred 15.0-0 in Gaprindashvili – Qin Kanying, Shanghai 1992, when 15...Qf6!?N 16.f5 gxf5 17.exf5 Bd5 18.Bf3 Bc4 19.Be2 Bd5= would have repeated the position. 15...Qa5 16.e5 Rab8 17.Rxb8 Rxb8 18.0-0 18.Bd3?! leads to difficulties for White after 18...Bxa2 19.0-0 Bb1 20.Qe2 Qxc3 21.Bc4 Bf5, and after the further error 22.Rd1? Rb1–+ his position collapsed in Fomichenko – Feller, Agneaux 2006. 18...Bxa2 19.Be3 Be6 20.Bf3 f6= The position remained objectively equal in Degerhammar – Cerrato, corr. 2015, although in a practical game I would slightly prefer Black due to the possibility of the passed a-pawn becoming a nuisance for White. E3) 12.Qd2

803

This is the main line by far, and I believe it to be White’s best, despite blocking the dark-squared bishop for the time being. Having defended the c3-pawn, White is ready to castle. He will often seize space in the centre with f2-f4 and, if permitted, e4-e5. The queen on d2 supports these ideas by controlling the dark squares and defending the d5-pawn. 12...e6 Challenging White’s centre in this way is the usual choice, and it seems best to me. 13.f4 White plays consistently and forces Black to make a decision regarding the bishop. If it retreats to the kingside, White can proceed with c3-c4 and perhaps e4-e5. If it goes to the queenside, White may look for opportunities to attack on the kingside with f4-f5. 13.0-0 exd5 14.exd5 just gives Black an easier version of the main line, as he reaches a similar central structure without having to place his bishop on the queenside. 14...Qd6 15.f4 Bg7 16.c4 Bd4† 17.Kh1 occurred in Yang Yijing – Chu Ruotong, Zhongshan 2015, and here Black’s play can be improved by:

804

17...Rb8N 18.Bb2 Bxb2 19.Rxb2 Bd7÷ The d5-pawn is securely blocked, and everything in Black’s position is in order. 13.c4?! runs into 13...Qh4! 14.Bd3 exd5 15.cxd5 Re8 16.Kf1 and now in Schwalen – Viebahn, Germany 2009, Black missed a powerful idea:

16...b5!N 17.Rxb5 (17.Bxb5 Qxe4 18.Rb3 Rd8µ leaves White with a weak pawn on d5 as well as a misplaced king) 17...c4! 18.Bxc4 Qxe4 19.Bd3 Qd4 20.Bb2 Qa4 21.Bxe5 Rxe5µ Black has a strong initiative for a tiny material deficit. 13.Bc4?! Qh4! is also bad for White. 14.Qe3 was seen in Hultin – B. Jaracz, Stockholm 2006, and here I can offer: 805

14...exd5!N 15.Bxd5 Be6! 16.Bxe6? This seemingly natural continuation only makes matters worse. (16.Qg5 Bxc3† 17.Ke2 Qxg5 18.Bxg5 Bxd5 19.exd5 f6µ looks bleak for White, but this would be the lesser evil) 16...fxe6 17.Bd2 Bf4 18.Qe2 Bxd2†–+ White loses material. 13...Bc7 In this variation I prefer the bishop on c7 to g7, due to the possibility of e4-e5, blocking the h8-a1 diagonal. 13...Bg7 14.c4 has also been tested in many high-level games. This also gives Black a sound position, and the choice is largely a matter of taste, but personally I am not so happy to give White such a big pawn centre. The trade-off is that the text move leaves Black’s kingside slightly exposed, but I do not believe this to be much of a problem. 14.0-0 White has nothing better than this natural continuation. 14.c4?? is obviously not an option here, as 14...Ba5–+ wins the queen. 14.Bb2? is a serious mistake. 14...exd5 15.exd5 occurred in Plachkinov – Chalmeta Ugas, Barcelona 2019, when 15...Bf5N 16.Rd1 Qh4† 17.g3 Qh3–+ would have left White with insurmountable problems due to his king being stuck in the centre. 14.Ba3? is another strange choice. 14...exd5 15.exd5 occurred in Kunze – Maruejols, Naujac 2008, when 15...Bf5!N would have best again, since 16.Rxb7 Bb6 17.0-0 Bc8 18.Rxb6 axb6µ leaves White with no real compensation for the exchange. 14.dxe6 is not so bad although Black has little to fear: 14...Bxe6 15.Rxb7 (the weaker 15.0-0?! Qxd2 806

16.Bxd2 b6 17.c4 was played in Thingstad – Arvola, Fagernes 2011, and now 17...Rae8N 18.Rbe1 Bd7 19.e5 f6³ gives Black pressure against the e5-pawn, and the c4-pawn may also become weak) 15...Bb6 16.Qxd8 Rfxd8 17.Kf2 This occurred in Zajsek – Roques, email 2017, and now we can enhance Black’s play by means of:

17...c4†!N 18.Be3 Bc8! 19.Bxb6 axb6 20.Rxb6 Rxa2 21.h3 Kg7 22.Rbb1 Rdd2 23.Rhe1 Rd3 24.Rbc1 Rb2 25.Red1 Be6 26.Rb1 Ra2© Black’s active rooks provide full compensation for the sacrificed pawn. 14.Bc4 Finally we come to White’s second-most-popular continuation, although it still ranks way behind our main line. Black can hit back on the queenside with: 14...a6 15.a4

807

15...b5! 16.axb5 axb5 17.Bxb5 17.Rxb5? is bad for more than one reason. The most convincing refutation is 17...Ra4!, with 18.Qe2 Bxf4!–+ a nice tactical detail. 17...exd5 18.Bc6 Other continuations are weaker, for example: 18.exd5? occurred in Slavin – Hawkins, England 2013, when 18...Bf5!N 19.Rb2 c4! 20.Bxc4 Re8† 21.Kf2 Ra4!–+ would have given Black a decisive attack. 18.Qxd5?! has never been played. My analysis continues 18...Qxd5 19.exd5 Bf5 20.Rb2 c4! 21.Bxc4 Ba5 22.Bd2 Bxc3! 23.Bxc3 Ra1† 24.Kf2 Rxh1³ when White has some compensation for the exchange due to his bishop pair and passed pawn, but Black can obviously press for a win.

18...Ra6 19.Bxd5 19.exd5 Bf5 20.Rb7 Rxc6 21.dxc6 gives Black a choice. The safe option is: 21...Qh4† (a more complicated route is 21...Qe8†!? 22.Qe2 Qxc6 23.Rxc7 Qxc7 when the position remains balanced) 22.g3 Re8† 23.Kf2 Qh3 24.Rxc7 Be4 25.Rg1 Qxh2† 26.Kf1 Qh3† 27.Kf2 Qh2†= Neither side could evade the perpetual check in Tinjaca Ramirez – Povchanic, corr. 2002. 19...Rd6 20.0-0 Bf5 21.exf5 Rxd5 22.Qc2 Rxf5 23.g3 23.c4 does not change much after 23...Qd4† 24.Kh1 Qd7 25.Kg1 Qd4†= with another repetition, Fuller – Brzezinski, corr. 2013.

808

23...c4 24.Be3 Bd6 25.Rbd1 Qe7 26.Rfe1 In Nyvlt – Walczak, corr. 2011, the simplest continuation would have been: 26...Bc5N The position remains equal.

14...exd5 15.exd5 White keeps a central passed pawn which may advance further in the middlegame. 15.Qxd5 fails to put any pressure on Black and is hardly ever played. 15...Qe7 16.Be3 Bb6 17.e5 occurred in Solymosi – Biro, Hungary 2012, when 17...Be6N would have been natural. After something like 18.Qd2 Rfd8 19.Qb2 c4 20.Bxb6 axb6 21.Rf2 b5³ Black has the edge thanks to his active rooks. 809

15...Ba5 Black puts pressure on the c3-pawn and prevents c3-c4, which would have defended the d5-pawn while opening the long diagonal towards Black’s king. True, the bishop is now a long way from the kingside, but it is hard for White to get his pieces into place to exploit the dark-square weaknesses. A critical position has arisen, from which White has tried various moves. We will focus on the three most popular of them: the harmless E31) 16.Ba3, the aggressive E32) 16.f5 and finally the main line of E33) 16.d6!, after first checking some minor options: 16.Bb2?! has been played several times but it makes no sense, as White cannot follow up with c3-c4 anyway. 16...Bf5 17.Rbd1 Qd6 18.Qc1 Rfe8 19.Bf3 c4–+ Faced with the idea of ...Bd3, White’s position had become very difficult in Markotic – Rogulj, Tucepi 1996 16.g4?! prevents ...Bf5 but is too slow and weakening. A good response is 16...Re8N 17.f5 Qd6 18.Bf3 b5 19.Kh1 Ba6ƒ when Black has the upper hand due to his better pawn structure. 16.Bf3?! is doubtful due to 16...Qf6 17.Bb2 c4! when it becomes clear that the bishop on e2 was not so badly placed after all. 18.Rbe1 Bf5 19.Re5 occurred in Cherniak – Baikov, Moscow 1998, and here I suggest:

19...Bb6†N 20.Kh1 Bc5 21.Rfe1 b5 Black is in control. A sample variation is 22.d6 Qxd6 23.Qxd6 Bxd6 24.Bxa8 Bxe5 25.Rxe5 Rxa8 26.Rxb5 Rd8 27.h3 Be4 28.Re5 Bxg2† 29.Kxg2 Rd2† 30.Kg3 Rxb2µ when White may still be able to hold a draw, but he will certainly have to suffer for a long time. 16.Bc4 was tried in Mikrut – Palus, Polanica Zdroj 1999, and a few other games. The purpose of White’s last move is not entirely clear, but I suggest meeting it with: 16...b6N 17.a4 Qd6 18.Bb2 Bf5 810

19.Rbe1 a6 20.Re5 Rfb8ƒ Followed by ...b5 when White’s bishop comes under attack. It is excessively brave and risky for White to hunt the bishop with: 16.Rb5?! b6 17.Rxa5 17.d6 occurred in P. Smirnov – Oral, St Petersburg (blitz) 2014, when 17...Bd7N 18.Rxa5 bxa5 would have resembled the main line. For instance, 19.c4 Rb8 20.Bb2 Rb6 21.Qc3 f6 22.Ba3 Re8 23.Bf3 Rxd6 24.Bxc5 Ra6 25.Bb7 Qc7 26.Bxa6 Qxc5† 27.Kh1 Bc6µ when White has succeeded in winning back the exchange, but his bishop is marooned out of play and his king is vulnerable. 17...bxa5 18.c4 White has questionable compensation for the exchange and Black can fight for the advantage with a novelty.

811

18...Re8!N Most games have continued with the less convincing 18...Rb8?! when 19.Ba3! makes things unclear: 19...Rb4 (or 19...Qd6 20.Qe3© and White has enough compensation) 20.Bxb4 axb4÷ Despite Black’s eventual victories in S. Ernst – Van Beek, Groningen 1999, and one other game, the position is balanced at this stage. 19.Ba3 Qd6 20.Bd3 Qe7 21.Kh1 Qe3µ White’s compensation is insufficient. Another idea is: 16.Rb3!? Qd6 17.Ra3 Less logical would be 17.Qd1 Bd7 18.Rxb7 as in Kunze – Pribyl, Germany 1999, when 18...Bxc3N 19.Be3 Rfe8 20.Bf2 Bg7³ leaves Black with the more pleasant game; White would like to be able to retreat the pawn from f4 to cover some squares and improve the scope of his dark-squared bishop.

17...c4 When followed up correctly, this move should lead to a draw. 17...b6!?N 18.Qb2 c4 19.Be3 Ba6÷ is another way to keep the game going. 18.Rxa5 Qb6† 19.Kh1 Qxa5 20.f5 Rd8 21.Qh6 Bxf5 22.Rxf5 gxf5 23.Qg5† Kf8 24.Qf6 Qa6 25.Qh8† Ke7 26.Qe5† Kf8N 26...Kd7? turned out well for Black in the game, but it should have been the losing move: 27.Qxf5† (27.Bf4!?+– is also good enough) 27...Kc7 28.Bf4†!N (28.Qxf7†? Kb6 should have led to a draw in Haag – Ftacnik, Deizisau 2001, but White erred again and went on to lose) 28...Kb6 29.Qf6†! Ka5 30.Bc7†+– White picks up the rook on d8 and/or catches the enemy king. 27.Qh8†= White should take the perpetual check. (27.Bg5?! runs into 27...Kg8! 28.Bf6 Qd6 29.Qxf5 Kf8µ when Black’s extra material should eventually tell.)

812

E31) 16.Ba3

The idea of this move is to provoke Black’s natural reply: 16...b6 White hopes to prove that the bishop on a5 will be misplaced, but this is not quite the case. The bishop is unpleasantly pressing on c3; besides, the bishop on a3 is hardly the best piece. 17.Rbd1 This has been the most popular choice so we will take it as the main line, although White’s results with it have been far from impressive. 17.Bf3?! is weaker; we already know that the bishop should remain on e2 for a while, to prevent Black from clamping down with ...c4. 17...Qd6 18.Rbc1 occurred in Scharrer – Huth, Hassloch 1998, when Black could have claimed some advantage with:

813

18...Ba6N 19.Rfe1 Rfe8 20.Bb2 c4 21.Rcd1 Bb7µ The d5-pawn is blocked and weak, while the b2bishop is passive. 17.Rb5?! Ba6! leads to a forcing sequence: 18.Rxa5 Bxe2 19.Qxe2 bxa5 20.Bxc5 Qxd5 21.Bxf8 Rxf8 The resulting endgame is surprisingly dangerous for White, as the advance of the f-pawn has greatly weakened his kingside.

22.f5 Rc8 23.Qf2 Rxc3 24.Qxa7 Rc2 25.Qb8† Kg7 26.f6† Kh6 27.Qg3 g5 28.a3 This occurred in Pelletier – Ni Hua, Kallithea 2008, and now 28...Qc5†!N 29.Kh1 Rc3 30.Qe1 Qxa3–+ would have given Black a decisive advantage: he has an extra passed pawn on the queenside, and the f6-pawn is also a weakness. 17.Bb5!? Bf5 18.Rbd1 reached a complex position in Haba – Ftacnik, Koszalin 1998. My suggestion 814

for Black is:

18...a6!?N 19.Bc6 Qf6! 20.Bb2 (20.Bxa8 Bxc3 gives Black excellent compensation after recapturing on a8 next) 20...Rac8 21.Qe3 c4 22.Qe5 Qxe5 23.fxe5 b5÷ We have reached a position of dynamic equilibrium, where any result would be possible in practice. Perhaps White’s safest continuation would be 17.Bb4!? Bxb4 18.cxb4 Qd6 19.Rbc1 as played in Yevseev – Avrukh, Groningen 1996. Here I suggest:

19...Bd7N 20.bxc5 bxc5 21.Bf3 Rab8= Maintaining the balance. Finally, 17.d6!? Bf5 18.Rbd1 Qf6 19.Bb2 Rad8 20.g4 led to a complicated struggle in Alexey Smirnov – Shinkevich, Izhevsk 2013. Best play seems to be: 815

20...Bd7N 21.c4 Qh4 22.Qd5 Bxg4 23.Qe5 f6 24.Qd5† Kg7 25.Bxg4 Qxg4† 26.Kh1 Qe2 27.Ba1 b5! ÷ With the last move, Black increases the range of his bishop. The position remains complicated and dynamically balanced.

17...Qd6 18.Bb2 White tried 18.Bb5 Bg4 19.Rde1 in Zpevak – Hracek, Czech Republic 2000, when Black could have pounced on the d5-target: 19...a6!N 20.Bc6 Rad8³ Intending ...Bd7 to remove the primary defender. 18...Bd7 19.Qc2 Bf5 20.Qc1 Rfe8 Black has a pleasant position and it is not surprising that he soon took over the initiative.

816

21.Bf3?! c4! As usual, this move works well when White removes his bishop from the f1-a6 diagonal. 22.Ba3 Qf6³ The constant need to defend the c3-pawn was inconvenient for White in Pravec – Nemec, Czech Republic 1999. E32) 16.f5

This move amounts to a pawn trade: Black can hardly ignore the pawn, with f5-f6 and Qh6 on the cards, so he is more or less forced to capture it. White will then capture on b7, hoping that the rook on the 7th rank along with the newly opened kingside lines will benefit him. Despite all this, and the fact that it has been used by several strong GMs including Chernin, Bacrot, Aronian and Atalik, the accepted theoretical verdict is that Black has no great problems. 16...Bxf5 17.Rxb7 Too much excitement gets White into trouble after: 17.Rxf5?! gxf5 18.Bb2 (18.Rxb7 Qf6 19.Qg5† Qxg5 20.Bxg5 Bxc3 21.d6 occurred in Katsuta – Baba, Tokyo 2018, when the simple 21...Rfb8N 22.Rxb8† Rxb8–+ would have left White with no compensation for the exchange, as d6-d7 can always be met by ...f6) 18...Re8 19.Bb5 In Dilleigh – Tikkanen, Gatwick 2011, Black had more than one route to victory, but the most accurate would have been:

817

19...Re4!N 20.Bd3 Rd4! 21.Qe3 Rxd5 Black’s material advantage will decide. For instance, 22.Qg3† Kf8 23.c4 Rxd3 24.Qg7† Ke8 25.Qe5† Qe7 26.Qh8† Qf8 27.Qe5† Kd8–+ and White’s attack comes to an end. 17...Qd6 White has tested several continuations, without proving any more than equality. 18.Bc4 This has been the usual choice, guarding the d5-pawn and hoping to unleash the bishop’s power against f7 at some point. Another game continued 18.Qh6 Qe5 19.Bb2 Bxc3 20.Bxc3 Qxc3 21.Rb3 Qg7 22.Qxg7† Kxg7 when Black won a pawn, but White was just in time to get it back:

818

23.Rc3 Rab8 24.Rxc5 Rb2 25.Bc4 Be4 26.Rf2 Rb1† 27.Rf1 Rb2= In view of the repetition, the players agreed a draw in Bravo – Gauci, Istanbul (ol) 2012. 18.Rb5 Bc7 19.g3 has been played at GM level. Black’s most militant continuation would be:

19...Bh3!?N (the game ended quickly following 19...c4 20.Rxf5 gxf5 21.Ba3 Qxa3 when a draw was agreed in Kiriakov – Mikhalevski, Dieren 1997, without waiting for 22.Qg5† and a perpetual) 20.Rf2 Rab8 21.Rxb8 Rxb8 22.c4 Qe7= There is some tension in the position, although it remains in approximate equilibrium.

819

18...Rfb8 19.Rxb8† 19.Rb3 Qe5 20.Re1 Qf6 21.Re3 Re8 22.Qe1 Rxe3 23.Qxe3 was also approximately equal in Staniszewski – Van Unen, corr. 2012; improving the bishop with 23...Bc7!?N= would be my choice here. 19...Rxb8 20.Qe1 The queen gets out of the way of the c1-bishop. White has also tried: 20.Qg5 Bxc3!? 20...Rb1= is also satisfactory, and it was not long before a draw was agreed in Bacrot – Sutovsky, Khanty-Mansiysk 2005. 21.Rxf5 Bd4†

820

22.Kh1!N 22.Rf2? was a mistake in Dubessay – Molner, Reykjavik 2014, which could have been refuted by: 22...h6!N 23.Qxh6 Rb1 24.Kf1 Bxf2 25.Kxf2 Qf6† 26.Qf4 Qxf4† 27.Bxf4 Rb4–+ White loses one of his bishops, resulting in a hopeless endgame. 22.Kf1?! is a less convincing improvement due to 22...Rb1 (22...f6 23.Qf4 Qxf4† 24.Rxf4 Rb1 also gives Black an extra pawn although White should be able to hold) 23.Bd3 Rxc1† 24.Qxc1 gxf5³ when White will have to suffer to make a draw. 22...Rb4! 23.Bb3 23.Bd3 f6 24.Qg4 Be5 25.Bc4 transposes. 23...f6 24.Qg4 Be5 25.Bc4 Bd4= With a repetition. (If Black is desperate to play on he can try 25...Rxc4 26.Qxc4 gxf5 27.Be3 Bd4 28.Bxd4 cxd4 29.Qxd4 Qe5, but the position remains equal and White would have to play remarkably badly to lose.)

821

20...Rb1 21.g3 Since Black is going to play ...Bc7, this move is needed anyway. Another game continued 21.Qe3 Bc7 22.g3 Qd7 23.Bd3 Bxd3 24.Qxd3 Ra1 25.Qe2 Qxd5 26.Qe8† Kg7 27.Bh6† Kxh6 28.Qe3† when the rivals agreed a draw in Krzyzanowski – Ankay, corr. 2012. The likely finish would have been:

28...Kg7 29.Rxa1 Bb6 30.c4 Qxc4 Black is materially no worse, having a bishop and two pawns against a rook, so White should take perpetual check while he has the chance: 31.Qe5† Kg8 32.Qe8† Kg7 33.Qe5†=

822

21...Bc7 22.Qe8† Qf8 23.Qe2 Keeping some tension on the board. 23.Qxf8† Kxf8 24.Be3 Bd6 25.Rxb1 Bxb1 26.Kf2 Ke7 27.Ke2 Kf6 28.a3 Be4 29.Bh6 Ke7 30.Ke3 Bg2= gave Black no problems in Vasin – Norchenko, corr. 2009. 23...Bh3 24.Re1 h5 25.Bd3 Rb8 26.Qe7 Ba5!? There is nothing wrong with this, but a simpler route to clear equality would be 26...Qxe7N 27.Rxe7 Bd6 28.Re1 (but not 28.Rxa7?? c4–+) 28...Kg7 29.c4 f6 30.Re2 Kf7= when it is hard for either side to improve their position. 27.Qxf8† Kxf8 28.Be3 Bxc3 29.Bxc5† Kg7 White has a passed pawn in the centre but it will not get much further. Black has good control over the position and his rook may go to b2, threatening the white king as well as the a2-pawn.

30.Rc1 Bf6 31.Kf2 a5 32.Rb1 Rc8 33.Be3 Be5 34.Rb7 Rc3 35.Rb3 Rc8 36.Ra3 Rb8 37.Rxa5 Rb2† 38.Kg1 At this point a draw was agreed in Ham – Dothan, corr. 2014. Black has a number of acceptable continuations but 38...h4 is the most human choice, when the game is likely to end in a perpetual check on the kingside within a few moves. E33) 16.d6! GAME 7 Gordon Dunlop – Piotr Oskulski 823

Correspondence 2012 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Nf3 c5 8.Rb1 0-0 9.Be2 Nc6! 10.d5 Ne5 11.Nxe5 Bxe5 12.Qd2 e6 13.f4 Bc7! 14.0-0 exd5 15.exd5 Ba5 16.d6!

This move poses the maximum difficulty for Black. Advancing the central passed pawn makes it harder for Black to bring the desired harmony to his position. On the other hand, Black also has certain advantages: the c3- and a2-pawns are weak as usual, and even the powerful d6-pawn may become vulnerable if White misplays his hand. 16...Rb8 By defending the b7-pawn, Black prepares the bishop’s development to f5. 17.Ba3 This has been the most common choice. Other ideas include: 17.Bb2?! Bf5 18.Rbd1 occurred in Chiku Ratte – Cardin, Montreal 2013. White’s play is almost comical, as he will not be able to play c3-c4, which renders the b2-bishop useless. A logical continuation is:

824

18...b5!N 19.g4 Bd7 20.f5 Qh4³ Black has every reason to be happy. 17.g4?! is too aggressive. 17...b5 18.f5 (18.Rxb5?! is obviously bad in view of 18...Rxb5 19.Bxb5 Bxg4µ when the white king is devoid of cover) 18...c4 19.Ba3 This was Chorfi – Hage, email 2000, and here I can recommend:

19...gxf5!N 20.gxf5 (or 20.Bc5 Qf6 21.Rf3 Bd7 22.gxf5 Bb6³ and White remains with weak pawns on c3, d6 and f5) 20...Bd7 21.Bc5 Qf6 22.Bd4 Qxd4†! 23.Qxd4 Bb6 24.Rbd1 Rfe8 25.Bf3 Re5³ Black is somewhat better due to White’s pawn weaknesses. 17.Bf3?! has been tried by some strong players, but need I repeat that White should not give up control over the f1-a6 diagonal? 17...b5! 18.Ba3 (the passive 18.Bb2 Bf5 19.Rbd1 c4 20.Qd5 was played in 825

Riemersma – Sedlak, Amsterdam 2011, when 20...Be6N 21.Qe5 Re8 22.Rfe1 Qh4 23.g3 Bb6† 24.Kh1 Qh3µ would have left White in a difficult position; again, the bishop on b2 is comical) 18...Bf5 19.Rbd1 This occurred in Xu Jun – Ni Hua, Wuxi 2006, and here Black’s play can be enhanced by means of:

19...Qb6!N 20.Kh1 b4 21.cxb4 Bxb4 22.Bxb4 Qxb4ƒ The d6-pawn is not going anywhere, but the prospects for the c5-pawn are rather rosy. 17.Rb5!? The idea of this quite aggressive move is that White is ready to sacrifice the exchange on a5, or possibly by leaving the rook to be taken on b5. 17...b6 18.Ba3!? This was the choice of Aronian who, in addition to being an elite GM, is an expert on the 8.Rb1 system. 18.Rxa5 bxa5 19.Ba3 is possible although Black has more than one way to maintain approximate equality. For instance: 19...a4!? 20.Bxc5 (20.c4?! is risky, and after 20...Qf6 21.Rd1 Bd7 22.Bf3 Rfc8 23.Qe3 Qf5 24.Qc3 Rb1ƒ Black’s initiative was becoming dangerous in Mukherjee – Holroyd, corr. 2014) 20...Qa5 21.Qd4 Bd7 22.Bb4 Qb6 23.Bc5 Qa5 24.Bb4= Due to the repetition, the players agreed a draw in Diermair – Barbosa, Graz 2014. 18...Ba6 19.f5 19.Rxa5?! is worse in view of 19...Bxe2 20.Qxe2 bxa5 21.Bxc5 as played in Cech – Cernousek, Rakovnik 2013, followed by 21...Qd7!N 22.c4 Rfc8 23.Qe5 Qf5 when White’s compensation is insufficient. 19...Bxb5 20.Bxb5 Qf6 21.fxg6 Bxc3 22.gxh7† Kh8 23.Rxf6 Bxd2

826

24.Bb2 White deviated with 24.Rf3 in Johnson – Holroyd, corr. 2015. It seems to me that 24...Bh6N is a natural reply, and after 25.d7 Rbd8 26.Kf2 f5 27.Bb2† Bg7 28.Bc1 Bd4† 29.Kf1 a6 30.Bc6 Kxh7 31.Rh3† Kg6 32.Rg3† Kh7= the game ends in perpetual check. 24...Be3† 25.Kf1 Kxh7 26.d7 Rbd8 27.Rf5 Bd4 28.Bxd4 cxd4 White has to be slightly careful, but his strong passed pawn enables him to maintain the balance.

29.Re5 Kg7 30.Ke2 f5 31.Re8 d3† 32.Kd2 a6 33.Bc6 b5 34.h4 Kf6 35.a3 Rh8 36.h5 Kg5 37.Re6 Rh7 38.Rd6 Rxh5 39.Kxd3 Rh4= The endgame remained balanced and a draw was soon agreed in Aronian – Vachier-Lagrave, Bilbao 2013.

827

17...Bf5 18.Rbd1 18.Rb3?! places the rook in a passive position. 18...b6 19.g4 Bd7 20.f5 was the continuation of Yordanova – Evans, corr. 2011, and now 20...Qh4!N 21.Bb4 gxf5 22.Bxa5 bxa5 23.gxf5 Rb6 24.c4 Qf6³ would have given Black somewhat better chances due to White’s exposed king and weak pawns. 18.Rb5!? is rare but not bad. 18...a6 19.Rxc5 Bb6 20.Kh1 Bxc5 21.Bxc5 Rc8 22.Bb4 occurred in Zuferi – Carow, Neustadt an der Weinstrasse 2017, and here I suggest: 22...Qb6N 23.c4 Rfe8 24.c5 Rxe2 25.cxb6 Rxd2 26.Bxd2 Rc6 27.Bb4 Rxb6 28.a3 Rc6 29.Kg1 b6 30.Rf3 a5 31.Rc3 Bd7 32.Rxc6 Bxc6 33.Bc3= With an obviously drawn endgame. 18.Rbc1 Resting the rook against the c3-pawn does not look beautiful. 18...Re8 19.Bf3 19.g4!? is an attempt to play more actively, but 19...Be4 20.Bxc5 Re6 21.f5 gxf5 22.gxf5 Bb6 23.Bd4 Rxd6 24.Rf4 Bxd4† 25.cxd4 Bc6 26.Rg4† Kh8 27.Qf4 Qf6 28.Bc4 Rbd8 29.Qe5 Rg8 30.Rxg8† Kxg8 31.Rc3 Kf8÷ left Black with no problems in Bjuhr – Hebels, corr. 2007. In such positions, White’s isolated pawns may become targets for attack in the future. 19...b6 20.g4 Bd7 21.Bb4 Bxb4 22.cxb4 Qf6 23.Rfd1 Worse is 23.bxc5?! bxc5 24.Rxc5 Rb2 25.g5 Qxg5† 26.fxg5 Rxd2 27.Rd5 Rxa2³ when Black has some winning chances. We have been following Iotov – Muneret, corr. 2009. My improvement is:

828

23...Ba4!N 24.g5 White would prefer not to commit his pawns to a rigid formation, but he has to drive the black queen to a different square in order to avoid losing the d6-pawn. 24...Qf5 25.Re1 Rbd8 26.Bg2 Rxe1† 27.Rxe1 h6 28.gxh6 Qf6 29.bxc5 bxc5 30.Qd5 Qd4† 31.Kh1 Qxd5 32.Bxd5 Rxd6³ White will have to defend well to hold the endgame. 18...Qf6! This move was introduced by Ftacnik in 2006, and was not considered by Dearing in Challenging the Grünfeld, which was published the previous year. 18...b6?! was played in an older game, and after 19.h3 Qf6 20.Bb2 Rbd8 21.Bb5± Black paid the price for weakening the light squares in Khalifman – D. Roos, Germany 1998, as the bishop on b5 prevents Black from putting a rook on e8. 18...Rc8!? has been chosen by some strong players and also seems quite acceptable for Black. 19.Bxc5 So far, this move has been the unanimous choice. 19.Bb2?!N is a plausible but inaccurate move and 19...Rbd8 already favours Black. For instance:

829

20.Bf3 (our ‘iron friend’ recommends 20.g4 Be4 21.Bf3 Bxf3 22.Rxf3 Rfe8µ but White’s position remains uncertain to say the least) 20...c4! 21.Qd5 Bb6† 22.Kh1 Bd3µ The d6-pawn is weak and the bishop on d3 cements Black’s position. 19...Bxc3 20.Bd4 This continuation leads to simplifications. 20.Qd5!? keeps things more complex although the position remains balanced: 20...b6 21.Bd4 (21.Ba3!? was played in Wenzel – M. Petrov, corr. 2016, and here I suggest strengthening Black’s play with 21...Rfe8!N 22.Bb5 Re3 23.Bc1 Re6 24.Bd7 Re4 25.Kh1 Rd8 26.Bxf5 gxf5 27.Rd3 Bb4 28.d7 Kf8 29.a3 Bc5 30.Rfd1 Qe6= when Black has solved all his problems) 21...Bxd4† 22.Rxd4 Rfc8 23.Rfd1 This was Otalora – Molina, Medellin 2017, and here I suggest: 23...Rc2N 24.R4d2 Be6 25.Qd3 Rc5 26.Qe4 Rd8= The position remains equal and it is difficult for either side to make any progress.

830

20...Bxd4† 21.Qxd4 Qxd4† 22.Rxd4 b5 Black intends to make a passed pawn of his own on the queenside. 23.Rc1 Be6 24.Rd2 24.Rc5 Bxa2 25.Bxb5 Be6 26.Bd3 Rfd8 27.f5 gxf5 28.Bxf5 Bxf5 29.Rxf5 Rb6 30.Rfd5 was dead equal in Gross Winter – Oger, corr. 2010. In the future there will be an exchange of a- and d-pawns, and there will be nothing else to play for. 24...Rfd8 25.Rc6 b4 26.Ra6 Rb6 27.Rxb6 axb6 The position has simplified to an endgame – but even here, there is a fork in the pathway, with games proceeding in different directions.

831

28.Kf2 The stem game with 18...Qf6! continued 28.h3 Kf8 29.Bf1 h5 30.h4 Ra8 31.Rd4 Rd8 32.Rd2 Ra8= with a repetition and subsequent draw in Mikhalevski – Ftacnik, Bellevue 2006. 28...f5 29.Bb5 Kf7 30.Kg3 White can try 30.Bc6N but after 30...Rc8 31.d7 Rd8 32.Rd4 Ke7 33.Rxb4 Bxd7 34.Bxd7 Rxd7 35.Rxb6 Rd2† 36.Kg3 Rxa2 37.Rb7† Kf6 38.Rxh7 Ra3† 39.Kf2 Ra2†= the result is the same. 30...b3 31.axb3 Bxb3 32.d7 Be6 33.Rd6 Ke7 34.Rxb6 Bxd7 35.Rb7 Kd6 36.Bxd7 Rxd7 37.Rxd7† Kxd7 Surprisingly for a correspondence game, the players continued in a dead drawn pawn endgame. 38.Kf3 Ke6 39.Ke3 Kd5 40.Kd3 h6 41.h4 h5 42.g3 Kc5 43.Kc3 Kd5 44.Kd3 Kc5 45.Kc3 Kd5 46.Kd3 Finally the inevitable draw was agreed. ½–½ Conclusion This chapter has dealt with a variety of options after 7.Nf3 c5 (excluding 8.Bg5?! and 8.Be3, which were covered in Chapters 12 and 13 respectively). Starting with the rare 8.Bc4 and 8.h3!?, we saw that Black was generally able to press against the d4-point with no problems, with fair chances to take over the initiative in the event of the slightest slip by White. We then looked at two bishop moves. 8.Be2 is well met by 8...Nc6, when 9.Be3 Bg4 gives Black no 832

problems at all, leaving 9.d5!? as the critical choice. Theoretically Black is at least equal here and I think he should obtain some advantage with precise play, but it is important to be well prepared. We then moved on to 8.Bb5† Nc6, when 9.d5 leads practically by force to an endgame where Black is not worse at all. 9.0-0 keeps more tension in the position although Black keeps a pleasant game: the bishop on b5 does not bother him at all. The final part of the chapter examined 8.Rb1, one of the most strategically rich and well-analysed branches of the Grünfeld. After the standard 8...0-0 9.Be2 I recommend 9...Nc6!, which offers – at least in my opinion – the best balance of theoretical soundness, a manageable body of theory and certain chances to play for a win in the event that White mishandles the position. The main line continues 10.d5 Ne5 11.Nxe5 Bxe5 12.Qd2 e6 13.f4 Bc7 14.0-0 exd5 15.exd5 Ba5 when 16.d6! is White’s most ambitious choice, although a combination of analysis, grandmaster praxis and correspondence games makes it clear that Black’s position is completely sound.

833

A) 9.d5 444 B) 9.Be3 0-0 447 B1) 10.Rb1!? 448 B2) 10.Rc1!? cxd4 11.cxd4 Qa5† 451 B21) 12.Qd2 452 B22) 12.Kf1 Game 8 454 B3) 10.0-0 b6! 459 B31) 11.h4!? 461 B32) 11.Rb1 463 B33) 11.Rc1 466 B34) 11.Qd2 Bb7 468 B341) 12.Rac1 469 B342) 12.Rad1 Game 9 471

834

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 This is arguably the most critical test of the Grünfeld Defence, as well as one of White’s most popular choices. White develops the bishop actively, while preparing to reinforce the d4-pawn with Ne2. Compared to the various Nf3 lines already examined, Black will no longer have the option of pinning and subsequently eliminating the knight with Bg4, due to the possibility of f2-f3. Moreover, the knight will not be as passively located on e2 as initial appearances may suggest. In the future it can jump to f4, controlling some central squares and/or participating in a kingside attack. Sticking with the kingside theme, White can sometimes consider attacking with h4-h5 or, after castling, f4-f5. 7.Bc4 was first played at a high level by Erik Lundin against Karl Helling, at the 1931 Olympiad in Prague. It remained a rarity for a few decades thereafter, but eventually it found favour among players such as Gligoric, Simagin, Boleslavsky, Bronstein, Geller and especially Spassky, who won many bright games with it. Other fierce fights in this variation took place at the World Championship match of Seville 1987, where Karpov employed it as his primary weapon against Kasparov. More recently, Carlsen, Topalov, Kramnik, Aronian, Morozevich, Ivanchuk, Radjabov and Yakovenko are some of the leading players who have used it. In the literature, Grandmaster Sakaev recommended it in two catchily-titled books: How to Get the Edge Against the Gruenfeld (2004) and An Expert’s Guide to the 7.Bc4 Gruenfeld (2006). Unsurprisingly, theory has been enriched with new ideas and plans since then. 7...c5 8.Ne2 8.Nf3 leads back to variation A of the previous chapter on page 400. 8.Be3 has been mentioned via the 7.Be3 c5 8.Bc4 move order on page 369.

8...Nc6 The similarly popular 8...0-0 is likely to transpose to one of the lines analysed below. White may respond to the attack on the central pawn with A) 9.d5 or B) 9.Be3. 835

A) 9.d5 This move captures space but expands the activity of the g7-bishop, while the white centre loses some of its flexibility. 9...Na5 Confronting White with a choice of what to do with his attacked bishop. 10.Bd3!? Avrukh does not consider this move in Grandmaster Repertoire 9, but some strong GMs have played it since his book was published. 10.Rb1 was played in a game between two American GMs, I. Ivanov – Yermolinsky, USA 1992. A simple solution was 10...Nxc4N 11.Qa4† Bd7 12.Qxc4 b6, seizing the advantage of the bishop pair. After 13.0-0 0-0 14.Bg5 h6 15.Be3 e6³ I prefer Black’s prospects. 10.0-0 offers the same piece trade but on this occasion 10...0-0! looks best, forcing White to move his bishop. 11.Bd3 (11.Bb3 occurred in Yakobson – Kuzenkov, corr. 1986, and here I like 11...e6!N 12.dxe6 Bxe6 13.Bxe6 fxe6 14.Qc2 Qc7³ when each side has structural defects, but Black benefits from the strong bishop on g7 and a knight which may go to the excellent c4-square) 11...e6 12.Bc2 exd5 13.exd5 Now I can offer an improvement on Jordaan – Pott, corr. 2010:

13...b6!N 14.Re1 Bb7 15.d6 Re8 16.Bf4 Qd7µ Black’s powerful bishops shoot from the flanks, and the d6-pawn is more of a weakness than a strength. The other natural continuation is: 836

10.Bb5† Bd7 11.Qa4 White gains nothing from 11.a4 0-0 12.0-0 b6 13.Bg5 Bxb5 14.axb5 Qd7 15.Rb1 e6 16.Qd3 exd5 17.exd5 Rfe8³ as played in Komljenovic – Herrera, Malaga 2002, when I agree with Avrukh’s assessment that Black has the slightly preferable game. 11...b6 12.0-0 a6 13.Bxd7† Qxd7 14.Qc2 I also checked 14.Qxd7† Kxd7 15.Rb1 Nc4 16.f4 b5³ when Black’s active pieces gave him the better chances in Muci Kuechler – Vargas Solano, Hermosillo 2002.

14...0-0 15.Bg5 Nc4 16.a4 We have been following Christiansen – Gulko, Estes Park 1987, and here I like the idea of taking action on the queenside: 16...b5!?N³ Black has the more active minor pieces and can look to create a passed pawn on the queenside in the future.

837

10...0-0 11.Rb1!? I believe this to be White’s most challenging continuation. The idea is to remove the rook from the long diagonal, thus preparing to strengthen the centre with c3-c4. 11.0-0 transposes to the 10.0-0 line in the notes above. 11.Be3 This move has been chosen by a few strong GMs. 11...b6 12.Rc1 e5 13.0-0 f5 This position has arisen several times, usually via a different move order. Black’s position seems promising to me, thanks to the simple plan of developing his initiative on the kingside. 14.f3 This seems most logical. Other tries include: 14.exf5?! gxf5 is an unfavourable trade for White. The d5-pawn is vulnerable, and defending it with 15.c4 allowed 15...e4 16.Bb1 Ba6µ when the c4-pawn became the target in Krush – Ganguly, Gibraltar 2009. 14.f4?! allows 14...fxe4 15.Bxe4 Nc4 when White is in trouble. For instance, 16.Qd3 Nxe3 17.Qxe3 Qd6 18.fxe5 Rxf1† 19.Rxf1 Qxe5 20.Qf3 Ba6 21.d6 Rd8–+ and Black already had a decisive advantage in S. Andreev – Todorov, Sofia 2008. Black’s bishops dominate the board and the d6-pawn is a serious weakness which will be destroyed in the near future. 14.c4?! f4 15.Bd2 occurred in Alberio – Stella, Lugano 2011, when 15...f3!N 16.gxf3 Bh3 17.Re1 Rxf3 would have been powerful, based on the following point: 18.Bxa5 (18.Be3 avoids the worst although 18...Qh4 19.Ng3 Raf8 20.Be2 R3f7‚ still leaves Black with excellent attacking prospects) 18...Qh4! 19.Ng3 Raf8 20.Bf1 Rxf2 21.Rc2 Bg4 22.Rxf2 Rxf2 23.Kxf2 Bxd1–+ Black’s kingside threats have been converted into a decisive material advantage. 14...f4 15.Bd2 838

15.Bf2 g5 16.g4 h5 17.h3 occurred in Luciani – Formento, Imperia 2012, and now it would have been good to capture a bit of space on the queenside: 17...c4!N 18.Bc2 Nb7 19.Kg2 Nd6 20.Rh1 Rf6ƒ Black has an excellent position, with chances to disturb White both on both flanks, although the kingside is likely to remain his main focus for a while. 15...c4 16.Bc2 g5 17.g3 This position was reached in Krasenkow – Zulfic, Canberra 2017. I suggest:

17...Rf7N 18.Kh1 Kh8ƒ Black enjoys an improved version of the Kings Indian and it is hard for White to find any counterplay. 11...e6 12.c4

839

12...f5!? Smashing up the centre is certainly thematic, although 12...b6 also gave Black a decent position in the aforementioned Ivanov – Yermolinsky game. 13.0-0 fxe4 14.Bxe4 Nxc4 15.Nf4 Nd6 16.Bf3 We have been following Yu Yangyi – Vachier-Lagrave, Khanty-Mansiysk 2019. I propose an improvement:

16...e5!?N 17.Ne6 Bxe6 18.dxe6 e4 19.Bg4 b6 20.Qd5 Nf5 21.Qxe4 Qd4 22.Re1 Qxe4 23.Rxe4 Bf6= The position is dynamically equal. The e7-square is under tight control, meaning that White’s passed pawn will go no further. Later Black can occupy the d-file. B) 9.Be3

840

This is the main line by far, and a move that will be played automatically in most cases. 9...0-0 Now B1) 10.Rb1!? and B2) 10.Rc1!? warrant some attention, but the natural B3) 10.0-0 remains the clear main line. 10.Qd2 Qa5 has been covered on page 369 – see 9.Bc4 Nc6 10.Ne2 0-0 in the notes to variation A of Chapter 13. 10.h4 cxd4 11.cxd4 Qa5† 12.Kf1 h5 13.Rc1 Qa3 transposes to 13.h4 h5 in the notes to variation B22 on page 455. 10.f3 is playable although it seems unnecessary to play it at this early stage. 10...b6 is one of many reasonable moves, when 11.0-0 would transpose to the 11.f3 note on page 460. 10.e5?! sees White trying to limit the g7-bishop, but he sacrifices the flexibility of his pawn centre along with control of some key squares. 10...Qc7 11.Rc1 Rd8 12.f4 was the continuation of Travnik – Hlavac, Klatovy 2007, and now Black can get a better position with:

841

12...Na5!N 13.Bd3 Be6 Targeting the weakened light squares. 14.Qa4 f5 15.0-0 c4 16.Bc2 Bd5µ White’s central pawns are well blocked by the powerful bishop on d5, and Black has an easy plan of mobilizing his queenside pawns. 10.d5 is not dangerous: 10...Ne5 11.Bb3 b6 12.0-0 f5 (12...Ba6!?N is also interesting) 13.Bc2 Qd6 14.Bf4 Ba6 15.exf5 Rad8 16.Re1 Qxd5 17.Qxd5† Rxd5 18.Be4 Rd7 19.Bxe5 Bxe5 20.fxg6 hxg6 21.Bxg6 Kg7 22.Be4 Rd2 23.f4 Bf6 24.Bf3 Rfd8ƒ Black had more than enough activity to compensate for the sacrificed pawn in Voro – Devine, corr. 2015. 10.dxc5 White wins a pawn but gives up plenty in return: his pawn centre has come apart, his c-pawns are doubled and the g7-bishop gains even more power. 10...Qa5 11.0-0 White tried 11.Qb3 Ne5 12.Bd4 in Rebollo Peralta – Pujol Codinach, Tarragona 2003, when Black could have simplified advantageously with: 12...Nxc4!N 13.Bxg7 Kxg7 14.Qxc4 Rd8 15.Qb4 Qc7 16.0-0 e5 17.Rfd1 Bg4 18.f3 Be6 19.Rd6 Rdc8³ The c5-pawn must fall, leaving Black with the better structure and minor piece. 11...Ne5 12.Bb3 Ng4 13.Qc1 In Trumpf – Schmid, Zurich 1991, Black’s most accurate continuation would have been:

842

13...Qc7!N 14.f4 Another interesting continuation is 14.Ng3 Nxe3 15.Qxe3 Bd7 16.Rad1 Bb5 17.c4 Bc6 18.f4 Rad8³ when Black’s strong bishops give him the better play, while White’s extra pawn is not felt. 14...Nxe3 15.Qxe3 Bg4 16.e5 Bxe2 17.Qxe2 Qxc5†³ Despite the opposite-coloured bishops, White faces a long struggle for a draw due to his inferior pawn structure. B1) 10.Rb1!?

10...Na5! Avrukh’s recommendation of 10...cxd4 11.cxd4 Qa5† is perfectly playable but the text move is more 843

ambitious. 11.Bd3 This is the most logical and popular choice. It seems strange for White to voluntarily part with a bishop: 11.Bb3?! Nxb3 12.Qxb3 In Rojas Solano – Justo, Calvia (ol) 2004, Black could have obtained a great position with:

12...b6!N Preparing to activate the unopposed bishop. 13.Qa4 Preventing ...Ba6. 13.dxc5 is strongly met by 13...Qd3! 14.c4 Qxe4 15.0-0 (15.f3?! Qh4† 16.Bf2 Qg5 17.0-0 bxc5µ gives Black an extra pawn as well as the bishop pair) 15...Be6ƒ when Black has the two bishops and the more active game, so her position is clearly better. 13.0-0 Ba6 also gives White problems. For instance, 14.Rfe1 Bd3 15.Rbd1 c4 16.Qa4 Bxe4 17.Qxc4 Bd5µ when Black has the better pawn structure and excellent play on the light squares. 13...cxd4 14.cxd4 a5! Not a standard move, but the logic is clear: the bishop is going to a6. 15.d5 15.Qb5 is convincingly met by 15...Ba6! 16.Qxb6 Bd3 17.Qxd8 Rfxd8 18.Rd1 Bxe4µ when Black’s bishops are great and the isolated d-pawn will remain a target. 15.0-0 solves the problem of the king but after 15...Ba6 16.Qd1 b5µ followed by ...b4. Black is clearly better because she has the advantage of the two bishops and the possibility of creating a passed pawn on the queenside. The text move looks critical as White attacks the b6-pawn, but Black can sacrifice it to get a powerful initiative: 844

15...Ba6! 16.Rxb6 16.Bxb6? Qd6 17.Bxa5 Bd3 18.Rd1 Qa6 gives Black a decisive advantage.

16...Qc7! Black must show some accuracy in developing the initiative. 16...e6?! 17.d6!÷ is less convincing. 17.Rc6 Qd7 18.Qd1 18.Bd4? is unplayable due to 18...Rfb8–+. 18...Rfc8 19.Rxc8† 19.Nd4?! e6 20.dxe6 fxe6–+ leaves White’s king stuck in the centre with Black’s bishops shooting from both flanks. 19...Qxc8 20.0-0 Qc4 21.Re1 Qxe4µ Black is clearly better thanks to the bishop pair and the vulnerability of the d5-pawn. 11...e5! Black invites a blocked centre resembling a Kings Indian, but in an improved version, since White will not have his usual counterplay on the queenside. 12.d5 This natural move has been the unanimous choice so far. I also considered: 12.dxe5!?N 12...b6 13.f4 Be6 14.a4 Nc4 Black has sacrificed a pawn in return for a high level of activity. 15.Bc1 15.Bf2?! Qe7 16.0-0 (16.Qc1?! prevents the pin but after 16...Rad8 17.Bc2 Rd2–+ White is in big trouble) 16...Rad8µ places White in an unpleasant pin. 845

With 15.Bxc4 Bxc4 White eliminates the strong knight but gives Black the advantage of two bishops and pressure on the light squares. 16.Qxd8 Rfxd8 17.Ng3 h5 18.h4 Rd3ƒ Black has an unpleasant initiative and White’s extra pawn is absolutely not felt. 15...g5! 16.0-0 gxf4 17.Bxf4 Nxe5ƒ Black’s pieces are more active than their counterparts, especially his knight. 12...b6

13.c4 It seems logical to strengthen the centre and prevent Black from playing ...c4. White fails to pose any problems with 13.Qc1 f5 14.Bg5, as played in Dreev – Thorsteins, Reykjavik 1990, on account of 14...Bf6!N 15.Bh6 Rf7 16.Bd2 (or 16.c4 f4µ when the bishop on h6 has no moves while Black keeps all his other advantages) 16...c4 17.Bc2 Nb7 18.0-0 Nc5µ when Black clearly has the more pleasant play. 13.Qa4 f5 14.f3 was tried in Poblete – Hadley, corr. 2003, when Black’s game can be improved with:

846

14...c4!N 15.Bxc4 (15.Bc2 f4 16.Bf2 g5µ leaves White with no counterplay, while Black can develop his initiative on the kingside and potentially even the queenside as well) 15...fxe4 16.fxe4 Qh4† 17.Ng3 Bg4ƒ White’s king is stuck in the centre and Black seizes the initiative. 13...f5 14.Nc3 f4 15.Bd2 This position was reached in Dreev – Kamsky, Reykjavik 1990. I offer an important novelty:

15...Nc6!N If Black can install the knight on d4 he will have a great game, so the following line is critical. 16.dxc6 Qxd3 17.Qe2 Qd6 18.Nd5 Rf7 19.c7 Bb7 20.f3 Bxd5 21.exd5 847

21...e4! Another important move: Black is not interested in counting pawns but instead seeks to activate his pieces. 22.fxe4 Bd4 23.Kd1 Qxc7ƒ White’s central pawns cannot advance, due to the fact that Black’s bishop has taken up a dominant position. The white king is stuck in the centre, and Black can look to play on either flank with ...g5-g4 or preparing ...b5. B2) 10.Rc1!?

848

The rook vacates the long diagonal while occupying a file which could be opened any time in the event of an exchange on d4. One possible reason for White’s delaying castling is to ram us on the kingside by launching the h-pawn, although it can also be played with less aggressive intentions. 10...cxd4 11.cxd4 Qa5† This check disturbs the harmony of the white pieces. In principle, it is advantageous for Black to exchange queens and other pieces, so if White wishes to play as ambitiously as possible he will have to make a minor concession of some kind. B21) 12.Qd2 is a rare move which deserves attention as it has been tested at a high level, but the more aggressive B22) 12.Kf1 is the clear main line. 12.Rc3!? looks odd but is playable enough. 12...Rd8 13.Qd2 occurred in Nett – Strebel, corr. 2015, and here I like:

13...Bg4!?N 14.f3 Be6 15.Bxe6 fxe6 I have already mentioned that exchanges tend to help Black, and the doubled e-pawns are of no great concern here because White’s centre is about to be liquidated anyway: 16.0-0 Nxd4 17.Nxd4 Bxd4 18.Bxd4 e5 White has nothing better than 19.Qg5 Rxd4 20.Qxe7 Qxc3 21.Qe6† Kh8 22.Qxe5†= with perpetual check. 12.Bd2 Qh5 13.d5 13.Be3 meekly offers a repetition. If Black aspires to more than a draw, he has the option of 13...Qg4 14.Ng3 Qxd1† 15.Rxd1 as played in Gipp – Grellmann, Mecklenburg 2003, at which point 15...Bd7N 16.0-0 Rfc8÷ leaves Black without a shadow of a problem. In a practical game I would choose Black’s side: White’s pieces are not optimally located and Black a simple plan of developing an initiative on the queenside with moves like ...Na5 and ...b5. 13...Ne5 14.Qb3 In Berglund – Klauner, Sweden 1999, Black overlooked a powerful idea: 849

14...b5!N 15.Bd3 15.Bxb5?! Rb8 16.f4 Ng4ƒ puts White in danger. 15.Qxb5?! a5! also gives Black a powerful initiative. A sample continuation is 16.Nf4 Qh4 17.Qb3 g5 18.g3 Qh6 19.Ne2 Qh3–+ when White’s king is stuck in the centre and his light squares are too weak. The text move is safer although Black should be happy after: 15...Nxd3† 16.Qxd3 a5ƒ With the bishop pair, plus good chances to create a passed pawn on the queenside in the future. B21) 12.Qd2 Qxd2† 13.Kxd2 Kramnik played this way with White against Svidler some years ago, but no other top players have tried it.

850

13...Rd8 14.Rhd1! White’s only decent idea is to sacrifice the central pawn. 14.Bd5?! Bd7 15.Rhd1 e6 forces the bishop to retreat, admitting the fallacy of White’s 14th move. 16.Bc4 Now in Zivanovic – Boskovic, Serbia 2006, Black could have developed a queenside initiative with:

16...Bf8N 17.Ke1 Na5 18.Bd3 Ba4 19.Bc2 Bxc2 20.Rxc2 b5³ With ...Nc4 on the way, Black has every reason to be happy. 14.d5?! leads to difficulties for White: 14...Na5 15.Rhd1 (15.Bd3?! is even worse in view of 15...e6 16.Nf4 exd5 17.Nxd5 Be6 18.Ne7† Kh8 19.Ke2 as played in Pabon Sarmiento – Bolivar, Bogota 851

2012, when 19...Bxa2N 20.Rc7 Be6µ would have been a safe pawn-grab) In Mermer – Isik, Antalya 2019, Black could have secured some edge with:

15...Nxc4†N 16.Rxc4 e6 17.Ke1 exd5 18.Rxd5 Be6 19.Rxd8† Rxd8³ With the bishop pair along with a pawn majority on the queenside, Black has good chances to increase his advantage. 14.Kc3?! The king can hardly feel safe after moving here, and White’s pieces will have difficulty achieving harmony. 14...Bd7 15.Rhd1 Rac8 16.Kb2N 16.Kd2? occurred in Runic – Abdil, Neum 2008, when Black missed the chance to strike a powerful blow: 16...Nxd4!N 17.Nxd4 (no better is 17.Bxd4 Bxd4 18.Nxd4 Bg4 19.Bd5 Bxd1 20.Rxc8 Rxc8 21.Kxd1 Rd8 22.Ke2 e6–+ when Black obtains a decisive material plus) 17...Bxd4 18.Bxd4 Ba4 19.Bb3 Rxd4† 20.Ke3 Rxc1 21.Rxc1 Rb4µ Black has a healthy extra pawn and good chances to convert it. The text move is the lesser evil but Black still has a great game after:

852

16...b5! 17.Bd3 a6 18.f3 Nb4 19.Rxc8 19.Bb1?! saves the bishop but at too high a cost, as 19...Be6 20.Rxc8 Rxc8 21.Rc1 Bc4µ sees Black’s pieces take up dominating positions. 19...Rxc8 20.a3 Nxd3† 21.Rxd3 Be6³ Black will press with the two bishops. 14...Bxd4! This is preferable to 14...Nxd4 15.Nxd4 Bxd4 16.Ke1 e5 17.Bxd4 exd4 18.Bd5 when Black’s position is certainly holdable, but he has no winning chances and he has to be slightly careful not to come under pressure. 15.Nxd4 Nxd4 16.Ke1!? Avrukh does not consider this move, which Kramnik introduced in 2013, but it is not too threatening. In the event of 16.Kc3 Nc6 17.Rxd8† Nxd8 18.Rd1 Nc6 19.Bd5 Be6 I agree with Avrukh’s assessment that Black is comfortably equal. White also gets nowhere after: 16.Bxd4 Rxd4† 17.Ke3 Rxd1 18.Rxd1 Bg4 19.f3 Rc8 20.Bd5 Be6 21.Bxe6 fxe6

853

22.Rd7 (the passive 22.Rd2? Rc3† 23.Kf4 Kf7 24.h4 occurred in Kosek – Hosek, Czech Republic 2007, when 24...Ra3!µ would have enabled Black to ask his opponent some questions) 22...Kf7 23.Rxb7 Rc2 24.Rxa7 Rxg2 25.h4 Rh2 26.Ra8 e5 27.a4 Rxh4= The passed a-pawn enabled White to draw without much difficulty in Kaczmarek – Vieites, corr. email 2017. 16...Nc6 17.Rxd8† Nxd8 18.Bd5 e6!? The stem game continued 18...Nc6 19.Bxc6 bxc6 20.Rxc6 Bb7 21.Rc4 a6 22.f3 Rc8 23.Rb4 Bc6= and was subsequently drawn, Kramnik – Svidler, Nizhnij Novgorod 2013. Although the position is essentially equal, Kramnik used his marginally more active pieces to keep the game going for a while, whereas the text move gives White nothing whatsoever. 19.Bb3 Nc6 Black’s extra pawn is balanced out by White’s bishop pair and more active pieces. 20.Rd1 Kf8 21.Ba4 Ke8 22.Rb1 Kd7 The players agreed a draw in Fuzishawa – C. Gilbert, corr. 2015. B22) 12.Kf1 GAME 8 Douglas Smith – Andreas Brugger Correspondence 2006

854

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.Rc1!? cxd4 11.cxd4 Qa5† 12.Kf1 The king moves to the safe zone, leaving the rook to find purpose along the h-file.

12...Qa3! The queen stays in touch with the a2-pawn and prepares ...Na5 in some lines. It may also regroup to d6 in order to exert pressure on White’s centre. 13.Rc3 It seems advisable for White to drive the queen away. 13.Qd2?! Rd8 quickly becomes unpleasant for White: 14.Rc3 (even worse is 14.d5?! Ne5 15.Bb5 b6 16.f4 Ng4 17.Bd4 as in Ftacnik – I. Gurevich, Biel 1993, when 17...Bxd4!N 18.Qxd4 Ba6µ would have given White problems due to his misplaced king and undeveloped rook on h1) Now we can improve on Black’s play in Bernadskiy – Manigandan, Chennai 2017, by means of:

855

14...Qa4!N (in the game Black preferred the thematic 14...Qd6 and was at least equal – but with White’s queen already having committed to the questionable d2-square, it seems better for Black to keep his queen on the flank) 15.Bb3 Qa6 16.Bc4 b5 17.Bd5 Bb7³ Black has an obvious edge in development. 13.Qb3 helps Black to achieve one of his goals: 13...Qxb3 14.Bxb3 Rd8 15.d5 Na5 16.Ba4 b6 17.f3 e6 18.Bg5 f6 19.Bd2 exd5 20.Bxa5 bxa5 21.Bc6 Rb8 22.Bxd5† Kh8 23.Kf2 f5ƒ The two bishops gave Black fine play in Grego – Wisham, corr. 2010. 13.h4 h5 14.Qb3 (14.Rc3 is preferable, when 14...Qd6 transposes to the main game below) 14...Qxb3 15.Bxb3 occurred in Swiercz – Bobras, Warsaw 2010, when a natural and promising continuation would have been:

856

15...Na5!?N (15...Rd8N is also decent) 16.Ba4 a6 17.f3 (17.Bd2 b5 is good for Black) 17...b5 18.Bc2 Nc4÷ Black has excellent prospects on the queenside. 13...Qd6 14.h4! As mentioned earlier, White’s scheme involving the king on f1 requires him to attack on the kingside, otherwise there will be no way to justify the location of the rook on h1. White can hardly be satisfied by 14.f3 Rd8 15.Rd3?! Na5 16.Bb3 as played in Brettschneider – Schmenger, Binz 1995, when 16...Nxb3N 17.Qxb3 b5!³ would have favoured Black, for the usual reasons of the bishop pair and potential passed pawn on the queenside. Another unfortunate choice would be 14.Qd2?! Rd8 15.Rd3 as played in Franic – Rade, Mravinci 1993, when Black has a powerful resource at his disposal:

857

15...b5!N 16.Bxb5 Rb8 17.a4 a6 18.Bf4 Qd7 19.Bxb8 axb5 20.Bf4?! (the lesser evil seems to be 20.axb5 Nxb8 although White’s prospects remain bleak to say the least) 20...bxa4–+ Black has tremendous compensation for the exchange thanks to the powerful bishops and passed pawn combined with White’s misplaced king. 14...h5! It is well worth preventing h4-h5. 15.f3 Intending Kf2, freeing the rook from captivity. 15.Qd3?! was a strange choice in Quintiliano Pinto – Utiyama, Itapetininga 2012. 15...Rd8N looks natural and strong, and 16.e5 Qb4 17.e6 Ne5 18.exf7† Kf8! 19.Qc2 Bf5 20.Qb3 a5!ƒ gives Black a powerful initiative:

858

Note that 21.dxe5? allows the simple yet beautiful 21...a4! when Whites dies because of the weakness of his first rank. 15.Qd2 Rd8 16.Rd3 Qb4 17.Qxb4 Nxb4 18.Rd2 was also unimpressive from White in Borboni – Pinna, Vescovato 2014. I would continue:

18...Bd7N 19.f3 Rdc8 20.Bb3 a5 21.Kf2 a4 22.Bd1 b5³ Black’s pieces are better developed and he has some initiative on the queenside. 15.Kg1 is not bad, although it’s hardly the most human move. 15...Nxd4 16.Nxd4 e5 17.Rd3 exd4 18.Bxd4 Qe7 19.Bxg7 Kxg7 20.Bd5 occurred in Gunes – Underwood, corr. 2015, when the simplest continuation would be: 859

20...Be6N The position is equal, although White still has to solve the mini-problem of getting the h1rook into the game. 15.Rc1 is also playable but nothing for Black to worry about: 15...Rd8 (if Black is satisfied with a draw, then 15...Na5!?N 16.Bd3 Nc6 leaves White nothing better than repeating with 17.Bc4=16.Qb3 This occurred in Aliyev – Doshi, St Petersburg 2017, and here I offer:

16...Bxd4!?N 17.Bxf7† Kg7 18.Nxd4 Nxd4 19.Qd5 Ne6 20.Qxd6 Rxd6 21.Bxe6 Bxe6= With an equal endgame.

860

15...Rd8 16.Bd5 This is the best way to protect the d4-pawn. 16.Rd3?! is well met by 16...Na5 17.Bb3 Nxb3 18.Qxb3 as in Lillevold – Thingstad, Oslo 2009, when 18...b5!N would have been excellent for Black. The questionable pawn sac 16.Kf2?! Nxd4 17.Nxd4 Bxd4 18.Bxd4 Qxd4† 19.Qxd4 Rxd4 20.Ke3 was played in Nenciulescu – Barbosa, email 2008. Black’s best winning chance looks to be:

20...Rd6!?N 21.Rhc1 Be6 22.Bxe6 Rxe6 23.Rc7 b6³ Despite his active rooks, White still has to battle for a draw.

861

16...Nb4!? After banishing the bishop from d5, Black plans to mobilize his queenside pawns and explode in the centre with a timely ...e5. 16...e5!? The immediate central break is also perfectly playable. We will follow an instructive example. 17.Bg5 Bf6 18.dxe5 Qxe5 19.Bxf6?! This move leads to problems for White. White also got into difficulties after 19.Kf2?! Bxg5 20.hxg5 Be6 21.f4 Qd6 22.Qb3 Na5 23.Qb2 Bxd5 24.Rd1 in Giussani – Gellert, corr. 2010, when Black overlooked 24...b5!N 25.exd5 (25.Qxb5? Bc6–+) 25...Qb6† 26.Kg3 Nc4 27.Qc1 Re8 28.Nd4 Re4 29.Rcd3 Rae8–+ with a decisive advantage, due to the active rooks and superb knight on c4. 19.Qb3!?N is best, for instance: 19...Be6 (a safe alternative is 19...Bxg5 20.Bxf7† Kg7 21.hxg5 Na5 22.Qc2 Kxf7 23.g4 b5 24.Rc5 Qd6 25.Rd5 Qc6 26.Rc5 Qd6= when neither side can avoid repeating moves) 20.Bxf6 Qxf6 21.Bxe6 fxe6 22.Kf2 Rd7÷ Reaching a complex position in which Black’s chances are not worse. 19...Qxf6 20.Qb3 Ne7 21.g3 Nxd5 22.exd5

22...b5! 23.Kf2 Bb7 24.Rd1 Qb6† 25.Kg2 b4 26.Rcd3 Qd6 27.Rd4 a5 28.a3 Ba6 29.Nf4 bxa3³ All this happened in Scheer – A. Mueller, email 2013. Black benefits from the more mobile passed pawn, the stronger minor piece and the safer king. 17.Bf4?! This natural continuation leads to difficulties. The best continuation was: 17.Bb3 a5 18.Kf2 (but not 18.Qb1?! when, in Vaccher – Dugo, corr. 2008, Black could have played 18...e5!N 19.d5 Bd7 20.Kf2 b5ƒ with a promising initiative on the queenside, 862

while White’s queen is far from optimally located on b1) 18...e5 19.d5 Bd7 20.Qd2 This occurred in Santo – Laczay, corr. 2009, and here I suggest 20...b5!?N, reaching an unclear position in which Black’s chances are not worse. 17...Qa6! 18.Bc4 White had no real choice, as the bishop and a2-pawn were both under fire. 18...b5 19.Bb3 19.Bxf7†? Kxf7 20.Qb3† may appear tempting, but it soon resulted in a hopeless position for White after 20...Be6 21.Qxb4 Bc4 22.Ra3 Qb6 23.d5 a5 24.Qb1 e6–+ in Ugge – Pasko, corr. 1999. Black has two great bishops and a lot of activity, while White’s position is devoid of harmony.

19...Nc6 Another tempting option is 19...Nxa2!?N 20.Rc7 b4 21.Kf2 Be6 22.d5 Nc3 23.Nxc3 Qb6† 24.Kg3 bxc3³ when Black has a small but stable advantage due to the queenside passed pawns. 20.Rd3 An earlier correspondence game continued: 20.Be3 b4 21.Bc4 Qb6 22.d5 Qb8 23.Rc1 Ne5 24.Kf2 a5 25.Nd4 Saunders – Krueger, corr. 2004. Here I like a dynamic approach:

863

25...e6!N 26.dxe6 Bxe6 27.Bxe6 fxe6 28.Qb3 Rd6 29.Rhd1 a4ƒ Black has the initiative due to his mobile queenside pawns, while White is in no position to exploit the weakened kingside structure. 20...b4 21.Kf2 Qb6 22.Be3 Ba6 23.d5 Qc7 24.Rd2 Ne5 The knight on e5 strongly upsets the harmony of the white pieces, as the d3-square must constantly be defended. The knight cannot really be driven away either, as f3-f4 will allow the knight to settle on g4. 25.Rc2 Qd6 26.Qd2 Rac8

27.Rhc1?? 864

27.Rb1 was a more stubborn defence, although 27...Rxc2 28.Bxc2 Rc8 29.Bb3 Nd3† 30.Kg1 Bc3! 31.Nxc3 bxc3 32.Qd1 Nf4µ leaves Black clearly better thanks to his passed pawn and active pieces. The text move must have been the result of a ‘mouse slip’ or some other moment of carelessness. White blunders the exchange and he saw no reason to postpone his inevitable resignation after Black’s reply. 27...Nd3† 0–1 B3) 10.0-0

10...b6! This move has become one of the major Grünfeld trends of recent years. Previously it had a dubious reputation. For example, Sakaev (2006) only considers it in a note, while Avrukh (2011) covers both 10...Bg4 and 10...Qc7 11.Rc1 b6 from Black’s perspective; however, it later became clear that Black does not have to commit his queen before playing ...b6. The text move was first played in the game Kirsch – Gavey from the Skopje Olympiad in 1972, but it lay dormant for several years until Miles used it against Polugaevsky at Wijk aan Zee 1979. Then it was adopted by GMs Jansa, Ftacnik and Smejkal, but Black ran into problems and the move remained under a cloud until mid-2011, when Grischuk unveiled an important improvement against Aronian. Since then, it has been taken up by a host of other top players including Ding Liren, Svidler, VachierLagrave, Le Quang Liem, Gelfand, Anand, Giri, Nepomniachtchi and So. Black’s last move implies a pawn sacrifice, and the critical response of 11.dxc5! will be given special 865

attention in the next and final chapter. In the remainder of this chapter, we will analyse the following alternatives: B31) 11.h4!?, B32) 11.Rb1, B33) 11.Rc1 and B34) 11.Qd2. We should also get acquainted with some minor options at White’s disposal: The database indicates that 11.Bd3? was played in the original game Kirsch – Gavey, Skopje (ol) 1972, followed by 11...Na5? 12.Rc1. Perhaps the moves were entered incorrectly (11.Rc1 Na5 11.Bd3 being the more plausible sequence), as it seems strange that both players could overlook the straightforward 11...cxd4N 12.cxd4 Nxd4–+ when White has no compensation for the pawn. 11.Bd5 seems odd, as the bishop will be driven away by means of ...e6 sooner or later. 11...Qc7 12.Bf4 Qd7 gives Black a good position, and after the further 13.Bxc6?! Qxc6 14.d5 Qb5 15.Re1, as seen in Simson – Kulaots, Puhajarve 2012, the straightforward 15...Ba6Nµ would have put White in trouble due to Black’s bishop pair and superior pawn structure. 11.Qa4!? deserves attention, although it has only been tried in one game, Smolonogov – Fedotov, Samara 2015. I recommend 11...Na5N 12.dxc5 Nxc4 13.Qxc4 Be6 14.Qb5 bxc5 15.Qxc5 Qd7 16.Bd4 Rfc8 17.Qa3 Bc4© with excellent compensation: White’s extra pawn is not felt and Black’s pieces are perfectly located, especially the bishop on c4. 11.e5 is a speciality of GM Naumkin: he first played it in Naumkin – Gutenev, Chelyabinsk 2017, and has repeated it several times since. My recommendation is: 11...Qc7N 12.Rc1 (or 12.Rb1 Na5 13.Bd3 Bb7÷ when Black’s pieces are well placed) 12...Bb7 13.Nf4 Rad8 14.Qe2 e6÷ Black has a comfortable position. An overaggressive try would be: 11.f4?! White alludes to the possibility of attacking with f4-f5. However, this will create additional weaknesses for him, while Black is ready to put pressure on the enemy pawn centre. 11...cxd4 12.cxd4 e6 13.Rc1 Na5 14.Bd3 Bb7 This position has arisen in quite a few games, via different move orders. 15.f5 This is the most consistent continuation of White’s plan. 15.d5?! exd5 16.e5 proved unsuccessful after 16...Qe7 17.Nd4 Nc4 18.Qe2 Nxe3 19.Qxe3 Rac8µ in Krzywda – Pawlowski, Lublin 2006. White’s compensation for the pawn is not visible and Black has the additional advantage of two bishops. 15.Qe1 Rc8 16.Qf2 was seen in Goroshilov – Radivilov, Moscow 2008, when 16...Qd7!N would have been logical, preparing ...Qa4 in some lines, as well as exchanging rooks along the c-file. Play may continue 17.f5 Rxc1 18.Rxc1 exf5 19.exf5 Re8µ when Black is clearly better thanks to his active pieces and better pawn structure. 15...exf5 16.exf5 Re8 17.Qd2 In Hecht – Anderberg, Berlin 1997, Black’s best continuation would have been:

866

17...Rc8N 18.Rxc8 Bxc8! This is the stronger recapture, as Black creates a threat of ...Bxf5, intending to meet Bxf5 with the crushing ...Nc4. Play may continue: 19.Bg5 Qd5 20.f6 Bb7 21.Rf2 Bf8 22.Bh6 Bd6³ Black is somewhat better thanks to his active pieces: his bishops are aimed at the kingside and his rook occupies an open file, while White’s pawn structure is fragmented. 11.f3 Strengthening the centre does not pose any problems for Black. 11...Bb7 12.dxc5 12.Qd2 Rc8 13.Rac1 is covered on page 466 – see 12.f3 Rc8 13.Qd2 in the notes to variation B33. 12...Qc7 This type of position will be discussed in more detail in the next chapter, but this is quite a pleasant version for Black.

867

13.cxb6 13.Qa4 is not scary in view of 13...Ne5÷ when Black’s position is certainly not worse. Another possible continuation is 13.Nf4 Bh6 14.cxb6 axb6 15.Nd5 Bxe3† 16.Nxe3 Ne5© when Black has excellent compensation due to his activity along the a- and c-files. 13...axb6 14.Qc2 Na5 15.Bb5 Ba6 16.a4 Nc4÷ Black had full compensation for the pawn in Chekletsov – Wagener, Pardubice 2015. His pieces are perfectly located: the bishop on g7 presses against the queenside and the knight occupies a formidable outpost. 11.Bb5!? Despite being a rare choice, this move is a typical idea for such positions. White removes the bishop from a possible ...Na5 attack. 11...Bb7 12.d5 12.dxc5N is worth considering, although Black obtains typical compensation due to his pressure on the queenside: 12...Qc7 13.Nf4 e6 14.Qd6 Qxd6 15.cxd6 Na5 16.f3 Bc6 17.Bxc6 Nxc6 18.Rac1 Rfd8 19.Rfd1 Bf8 20.d7 Ne5÷ The d7-pawn falls and, thanks to the weak c3-pawn, Black can try for more than a draw. 12...Ne5!? This leads to double-edged play. The untested 12...Na5N leads after 13.Rc1 e6 14.c4 a6 15.Ba4 exd5 16.exd5 b5 17.cxb5 axb5 18.Bxb5 Qxd5 19.Qxd5 Bxd5 to an equal endgame. For instance: 20.Bxc5 (or 20.a4 c4=) 20...Rfb8 21.a4 Nb3 22.Nf4 (22.Rcd1? Rxb5! 23.axb5 Bc4 is close to winning for Black) 22...Nxc5 23.Nxd5 Nxa4= With an absolutely equal position. 13.Nf4 c4 14.Rb1 Qd6 15.Qe2 Rac8 16.Rfd1 We have been following Peralta – Krysa, Resistance/Saenz Pena 2013. I recommend:

868

16...Rfd8N Bringing the last piece into play. A possible continuation is: 17.Bd4 Bh6! 18.Be3 Black can either repeat the position or keep playing with: 18...Qf6 19.Nh3 Bxe3 20.Qxe3 Nd3÷ Black’s chances are no worse: the knight on d3 is strong, and its counterpart on h3 is misplaced. B31) 11.h4!?

The idea of attacking on the kingside with h4-h5 is a serious idea in such positions. At the time of writing it has only been played in four games in this exact position, but all of them involved strong 869

GMs. 11...e6 12.h5 Qh4 One of the advantages of 10...b6 compared to 10...Qc7 11.Rc1 b6 is that Black can respond to White’s attacking plan with this queen move. 13.hxg6 hxg6 14.f3 cxd4 15.cxd4 Rd8 16.Qd2 16.Bb5? was a significant error in Timofeev – Kurnosov, Khanty-Mansiysk 2012. Black missed the opportunity to deliver a damaging blow with:

16...Nxd4!N 17.Nxd4 Qh8! (17...e5 is playable but White can maintain the balance in a few ways; the text move is much stronger) 18.Bc6 Rxd4! 19.Qe2 Rb8 20.Rad1 Bb7µ Black has a healthy extra pawn and good chances to realize it. 16...Bb7 17.Rac1 Be5!? Another possibility is 17...Qh8!?N 18.Rfd1 Rac8 19.Qb2 Qh4÷ with a complex position of dynamic equilibrium. 18.f4 Bg7 19.Rf3 Rac8 20.Rh3 At this point it is worth suggesting an improvement on Dubov – Svidler, Hamburg 2019:

870

20...Qe7!N The queen returns to the safe zone and Black is poised to trade some pieces. The critical line continues: 21.f5 exf5 22.exf5 Na5 23.Bd3 Rxc1† 24.Nxc1! This is the most dangerous continuation. Let’s consider the other possibility: 24.Qxc1 Rc8 25.Qb1 (25.Qd2?! is worse in view of 25...Nc4 26.Bxc4 Rxc4 27.fxg6 fxg6 28.Qd3 Rc6 29.Bh6 Bc8! 30.Re3 Re6 31.Qc4 Bxh6 32.Qxc8† Kf7 33.Rxe6 Qxe6ƒ when White will have to defend accurately, despite the limited material) 25...g5÷ The position is complicated but objectively equal. White has chances on the kingside, but Black has the edge in terms of static elements due to his favourable pawn structure.

871

24...gxf5 25.Ne2 Rd6 26.Bg5 26.Ng3? would be an error due to 26...Qxe3†! 27.Qxe3 Bxd4 28.Qxd4 Rxd4 29.Nxf5 Ra4µ when White faces a tough battle to save the endgame. White can play more modestly with 26.Bf2!?, when 26...Bc8 27.Re3 Qd7 28.Rf3 Qd8 29.Bxf5 Nc6 30.Bxc8 Qxc8÷ reaches a complex but objectively equal position, as the weakness of the d4-pawn is balanced out by Black’s slightly exposed king. 26...Qd7 27.Qf4 Be4 28.Bxe4 fxe4 29.d5 Nb7 The time has come to bring the knight back into the game. The position remains equal in several lines, one example being: 30.Rg3 Kf8 31.Nc3 Nc5

872

32.Nxe4 Nxe4 33.Qxe4 Rg6!= A final accurate move assures Black of equality. (The trick is to avoid the greedy 33...Rxd5? due to 34.Be7†!+– when Black pays the price for spoiling his coordination.) B32) 11.Rb1

White takes his rook off the long diagonal and places it on the b-file, forcing Black to pay extra attention to the possibility of dxc5. 11...Na5 12.Bd3 This leads to a quite common position, although it has usually arisen via a slightly different move 873

order. 12.Bd5 was played in S. Andreev – Samaridis, Rhodes 2019, when Black could have made use of an interesting tactical opportunity: 12...cxd4!N

13.cxd4 (13.Bxa8?! dxe3 14.Qxd8 Rxd8 15.Rbd1 exf2† 16.Kxf2 Rf8³ gives Black more than sufficient compensation for the exchange, due to his excellent bishops and White’s multitude of pawn islands) 13...Bb7 14.Bxb7 Nxb7 15.d5 Qd7 16.Nd4 Nc5 17.f3 e6÷ Black successfully challenges the enemy pawn centre and reaches a comfortable position. 12...cxd4 13.cxd4 Bb7 This position has been reached numerous times at grandmaster level, with White trying a variety of ideas. 14.Qa4!? This move has scored quite impressively although I believe Black should be fine. A natural-looking continuation is 14.Qd2 Rc8 15.Rfc1 Qd6 16.d5 e6 17.Bf4 as seen in Voiteanu – David, Baile Olanesti 2013, when Black should have traded bishops:

874

17...Be5!N 18.Bxe5 Qxe5 19.dxe6 Qxe6÷ Black’s pieces are located harmoniously and he does not have a shadow of a problem. 14.d5 is a tempting way to capture space. Black should react energetically with: 14...e6 15.Bd2 (15.dxe6N 15...fxe6÷ gives Black enough activity to make up for the slight damage to his pawn structure) 15...exd5! Provoking interesting complications. 16.Bxa5 dxe4 17.Bc2 We are following Corrales Jimenez – Koban, Harkany 2008, and here I propose an improvement:

17...Qg5!N Threatening ...Qxa5 and setting up ...e3 ideas. 18.Bb4 Rfd8 19.Qe1 Rac8÷ Black has a promising position, with two pawns for the piece plus an initiative which is not easily neutralized. For instance, the natural-looking 20.Bb3?! fails to solve White’s problems after: 20...e3 21.f3 a5 22.Ba3 Rd2 23.Bc1 875

23...Ba6! Not content with sacrificing a piece, Black makes it a rook! 24.Bxd2 exd2 25.Qf2 Bf8! 26.Kh1 Bc5 27.Qg3 Qf5 28.Ng1 Bxf1 29.Rxf1 Re8µ White will find it hard to cope with the d2-pawn. 14.h4 As mentioned, this type of battering ram has become a standard device in such positions. 14...Rc8 15.h5?! This consistent move turns out badly. 15.Rc1!?N is an improvement although 15...e5÷ leads to an unclear position where Black’s chances are no worse. 15...Nc4 16.Bxc4 Rxc4 White’s kingside play is not enough to offset Black’s advantage of two bishops. We will follow a good example. 17.Qd3 Ra4 18.h6 Bh8 19.Qb3 Qd7 20.Nc3 Ra5 21.d5 Ba6 22.Rfd1 Rc8 23.Rbc1 Bc4 24.Qb4 Be5 25.a4 Bd6 26.Qb2 f6 27.f4 In Porper – Antal, Edmonton 2009, Black could have struck hard on the kingside with:

876

27...g5!N 28.fxg5 Qg4 29.Qf2 29.gxf6? Qg3 30.f7† Kf8! wins. 29...fxg5 30.Qf5 Qxf5 31.exf5 Bb3 32.Ne4 Bc5ƒ White suffers from weak pawns on a4 and d5. 14.f3 This move strengthens the e4-point but is rather innocuous, and Black gets good play by fighting for the c-file. 14...Rc8 15.Qa4 15.Qd2?! Nc4 16.Bxc4 Rxc4 17.Rfc1 Ba6 18.d5 occurred in Dinh – N.H. Le, Ho Chi Minh City 2005, when 18...e6! 19.Rxc4 Bxc4 20.Nf4 Qd6³ would have successfully challenged White’s centre, leaving Black with two bishops and good chances to make a passed pawn on the queenside. 15...e6 16.Rfc1

877

16...Rxc1† 17.Rxc1 a6! Black temporarily gives up the c-file but improves his queenside prospects in other ways. 18.Qa3 This position occurred in Molzahn – Jean, corr. 2017. I recommend: 18...Nc6N 19.e5 Qd7 20.Be4 Rc8÷ Black has a good position. The d4-pawn remains weak, while Black can play ...Bf8 any time in order to mobilize the queenside pawns.

14...e6 15.Rfd1 15.Rbc1 can be met by 15...Rc8, when 16.Rxc8 Qxc8 17.Rc1 reaches the same position as 15.Rfc1 below.

878

15.Rfc1 Rc8 16.Rxc8 Qxc8 17.Rc1 Qa8 18.f3 Rc8 should be fine for Black, for instance:

19.Rb1 (19.Rxc8† Qxc8 20.Kf2 Bf8 was also equal in Shabaev – Bobel, corr. 2012) 19...Bc6 20.Qb4 Qb7 21.Rc1 Qd7 22.g3 Bb7 23.Rxc8† Qxc8= The position remained balanced in Mills – Hernandez, corr. 2017. 15...Qh4!? Some strong GMs have played differently, but I like this move which aims to provoke some weakening of White’s kingside. 16.f3 Rfd8 17.Bf2 Qe7 18.Be1 Nc6 19.Bb5 Rac8÷ Black’s position is quite comfortable. In the one practical example that reached this position, White tried to play too actively and got into trouble:

879

20.d5? exd5 21.exd5 Na5 22.Bxa5 bxa5 23.Kh1 Rxd5 24.Rxd5 Bxd5 25.Ba6 Rc5–+ Black’s bishop pair and superior activity gave him a decisive advantage in Riha – Kantor, Brno 1961. B33) 11.Rc1

This is another quite logical move, which has been tried by several strong GMs including Giri, Nakamura and MVL. The theory of it continues to be enriched by new practical examples. 11...Bb7 12.d5 As usual, this move has mixed consequences: White gains space but improves the scope of the g7bishop and gives Black a central outpost for his knight. 880

12.Qd2 transposes to variation B341 below. 12.f4?! Na5 13.Bd3 cxd4 14.cxd4 e6 transposes to the 11.f4?! line given in the notes on page 460. 12.f3 With this move, White reinforces the e4-point before further developing his play in the centre. Black should counter by exchanging heavy pieces along the c-file. 12...Rc8 13.Qd2 e6 14.Rfd1 Na5 15.Bd3 15.Bb5!? occurred in Bluebaum – Warmerdam, Karlsruhe 2019, when 15...Bc6N 16.Ba6 Ba4 17.Re1 Rc7÷ would have reached a double-edged position in which Black’s chances are no worse. 15...cxd4 16.cxd4 Qd7 17.Qb4 17.Rxc8 Rxc8 18.Rc1 Rxc1† 19.Qxc1 Qa4 20.Qc2 Qxc2 21.Bxc2 Ba6÷ was also fine for Black in Meiers – Pos, Amsterdam 2015. 17...Rxc1 18.Rxc1 Rc8 19.Rxc8† Qxc8 20.Qc3 Also after 20.Qa4 h5 Black had equalized without any problems in Agmanov – Mamyrbay, Kokshetau 2015. 20...Qd7 21.Kf2 Nc6 22.h3 a6= The game remained balanced in Kracht – Gburek, email 2008. 12.Bb5 This was the choice of Giri, as well as some correspondence players. White moves the bishop away from a possible ...Na5 jump, while trying to provoke a subtle weakening of Black’s queenside with ...a6. 12...Rc8 13.d5 This is the most ambitious choice, gaining space before Black can open the c-file for his rook. 13.Qd2 e6 (13...Na5 also led to acceptable play for Black in Giri – Vachier-Lagrave, Saint Louis 2019) 14.Rfd1 a6 15.Bxc6 Bxc6 16.d5 Ba4 17.Re1 exd5 18.exd5 Re8 19.c4 Qd6 20.Bf4 Qf6 21.h3 Bd7 22.Ng3 Qd4 was objectively equal in Moeckl – Ter Steeg, corr. 2014, although Black’s position would be the more pleasant in a practical game, due to the bishop pair and the possibility of creating a passed pawn on the queenside with ...b5. 13...Ne5 The knight occupies the square which was made available by White’s previous move. 14.c4 a6 15.Ba4 e6 16.Nf4 exd5 17.exd5 Qh4!? Provoking complications which are not unfavourable for Black.

881

18.h3 Bh6 19.g3 Qd8 20.Nxg6 Bxe3 21.Nxe5 Qg5 22.Kh2 Bxc1 23.f4 Qg7 24.Qxc1 White has sacrificed the exchange for a pawn and some initiative. His compensation appears decent, but the correspondence game demonstrated that Black is not worse at all: 24...Kh8 25.Bd7 Rcd8 26.Bg4 f6 27.Nf3 Bc8 28.Bxc8 Rxc8÷ All this happened in Haznedaroglu – Tiemann, corr. 2015. It is not clear if White’s compensation is enough for full equality. Black can look to prepare ...b5, creating a passed pawn on the queenside. 12...Ne5 13.Bb3 A Hungarian chess legend tried 13.Bb5 in Portisch – W. Schmidt, Jakarta 1983, when Black could have struck a powerful blow to the centre: 13...f5!N 14.Nf4 fxe4 15.Ne6 Qxd5 16.Qxd5 Bxd5 17.Nxf8 Kxf8³ Black has two pawns for the exchange and a lot of activity for his pieces. 13...c4 14.Bc2 e6 15.dxe6 Nd3! This position has been tested a few times by strong players. Black has sacrificed a pawn to activate his pieces, and White must decide whether to exchange on f7, or d3, or both. 16.exf7† 16.Bxd3 occurred in Polugaevsky – Miles, Wijk aan Zee 1979, when 16...cxd3N would have been best. My analysis continues:

882

17.Nd4 (17.exf7† Rxf7 transposes to the main line below) 17...Bxe4 18.f3 Bf5 19.Nxf5 (19.exf7† Rxf7 is another transposition to the main line below – as is the case on the next move) 19...gxf5 20.exf7† (in the event of 20.Qd2?! fxe6 21.Rfd1 Rc8 22.Qxd3 Qxd3 23.Rxd3 Rc7³ Black retains somewhat better chances thanks to the weak c3-pawn) 20...Rxf7÷ The position is about equal although the strong d3pawn gives Black some practical chances to be better. 16...Rxf7

17.Bxd3 17.f3?! is risky; 17...Nxc1 18.Qxc1 gave White some compensation for exchange and he went on to win in Korobov – Schreiner, Yerevan 2014, but the sacrifice is not really correct. Black’s strongest 883

continuation is: 18...b5!N With two possible follow-ups in mind. The first is to prepare ...a5 and ...b4, creating a passed pawn on the queenside. The second is ...Qa5, attacking the pawns on c3 and a2. For instance: 19.Rd1 (in the event of 19.f4 Black should activate the queenside pawns with 19...Qe7 20.e5 a5 21.Nd4 b4µ when he has good chances to realize his material advantage) 19...Qa5 20.Qb2 a6 21.Nd4 Bf8µ It’s not easy for White to improve his position, and Black can slowly advance on the queenside. Eliminating the strong knight is the best that White can do, even though it gives Black the advantage of two bishops. In Krasenkow – Troff, Cappelle-la-Grande 2016, the correct recapture would have been: 17...cxd3N 18.Nd4 Bxe4 19.f3 Bf5 20.Nxf5 The active 20.g4 Bc8 21.Qxd3 Bb7³ leaves Black with more than sufficient compensation for a pawn, due to the advantage of two bishops and the open position of the white king. 20...Rxf5 20...gxf5!? transposes to the note to White’s 16th move above. I decided to analyse the text move anyway, as it seems equally playable. 21.Qb3† Kh8 22.Rcd1 Qd7 23.Kh1 Rc8 24.Bd4 Rb5 25.Qa3 Ra5 26.Qb3 Rb5= The game should end in a draw by repetition. If White wants to play to win, then he risks becoming worse. For example: 27.Qa4?! Bxd4 28.cxd4 28.Rxd3?? Rb1!–+ is an important tactical point. 28...Rc3 29.h3 Kg7³ Black’s position is somewhat preferable due to his more active pieces, the passed d3-pawn and the potential weakness of the d4-pawn. B34) 11.Qd2

884

White develops his queen and introduces the possibility of Bh6 to exchange the Grünfeld bishop. 11...Bb7 White must make an important decision concerning how to position his rooks. We will consider B341) 12.Rac1 and B342) 12.Rad1 as the main possibilities. 12.f3 leads back to the 11.f3 line in the notes on page 460, while 12.Rfd1 cxd4 13.cxd4 Rc8 14.Rac1 e6 transposes to variation B341 below. 12.Rfe1 Rc8 13.Rac1 cxd4 14.cxd4 occurred in Enchev – Kanmazalp, Albena 2013, when 14...e6N 15.f3 Qe7÷ would have been logical, intending ...Qb4 or ...Na5 with exchanges along the c-file. 12.Rab1 Rc8 13.Rfc1 e6 14.f3 was played in Mihajlov – Abdumalik, Fornebu 2016, and now I like 14...Qe7!?N 15.Qb2 Rfd8÷ when it is difficult for White to achieve anything concrete, while Black can exchange on d4 at any moment and begin trading rooks. 12.f4?! is too ambitious. 12...Rc8 13.d5 Na5 14.Bd3 occurred in Schmitt – Bensch, Untergrombach 2015, and now it would have been logical to hit the centre by means of:

885

14...f5!N 15.Ng3 fxe4 16.Bxe4 e6 17.dxe6 Qxd2 18.Bxd2 Bxe4 19.Nxe4 Rce8µ Black will pick up the e6-pawn to equalize the material balance, at which point he will enjoy the much better game due to his superior pawn structure and more active pieces. The knight will be perfect on c4, while the bishop exerts pressure on the c3-pawn. 12.Bh6?! is also risky, as White leaves his centre with insufficient protection. 12...Bxh6 13.Qxh6 cxd4 14.Rad1 (14.f4? is even worse. 14...Qd6 15.e5 Qc5 16.Rf3 Bc8 17.Ng3 was seen in Williams – Melkumyan, London 2012, when Black already has many continuations leading to a decisive advantage, the strongest being 17...d3†!N 18.Kh1 Qxc4–+ when White’s position is hopeless.) 14...Ne5 15.Rxd4 Qc7 16.Bd5 In Mandiza – Thybo, Riga 2017, the most convincing continuation would have been:

886

16...Ba6!N 17.Qd2 Ng4 18.f4 Rac8µ Black is clearly better thanks to his powerful bishop and pressure on the c3-pawn. B341) 12.Rac1

This move is logical enough, although Black may be able to utilize the position of the rook to make beneficial exchanges on the c-file. 12...e6 13.Rfd1 This is the most logical and popular choice. 13.h4? is optimistic to say the least. 13...cxd4 14.cxd4 Qxh4 15.f3 occurred in Jakobsen – Simonsen, Denmark 2013, when 15...Qe7Nµ would have left White with no compensation for the pawn. 13.Bb3?! is too passive. 13...Na5 14.Ng3 (14.f3N seems like a slight improvement although 14...cxd4 15.cxd4 Rc8 16.Rxc8 Qxc8 17.Rc1 Qd7 18.h4 Rc8 still gives Black a promising position, and he can secure the advantage of the bishop pair at any time) 14...Rc8 15.Rfd1 cxd4 16.cxd4 Rxc1 17.Rxc1 Qd7³ In Molvig – Schmied, Helsingor 2013, Black was ready to proceed with ...Rc8, controlling the important c4-square. It is also worth noting that the knight on g3 is located badly. 13...cxd4 14.cxd4 Rc8 As is often the case, Black’s main idea is to remove some of the tension in the position by forcing exchanges along the c-file. 15.Bh6 This is among the more challenging of the moves White has tried. 887

15.f3 is a popular choice but 15...Na5 will quickly transpose elsewhere: 16.Bb5 reaches the Muzyka – Bolignano game below, under 15.Bb5; and 16.Bd3 transposes to the 12.f3 line in the notes to variation B33 on page 466. 15.d5?! is too active. 15...Na5 16.Bb5 exd5 17.exd5 Rxc1 18.Nxc1 occurred in Degtiarev – Ragger, Skopje (blitz) 2018, when the most accurate continuation would have been:

18...Qc8!N 19.Ne2 Nc4 20.Bxc4 Qxc4³ Black benefits from the bishop pair and the weakness of the d5-pawn. 15.Bb5!? has been tried in correspondence play. 15...Na5 16.f3 Rxc1 17.Rxc1 a6 18.Bd3 was seen in Muzyka – Bolignano, corr. 2014, and here I suggest:

888

18...Nc6 19.Bb1 b5÷ Black has a good position with pressure on the d4-pawn, while it’s not easy for White to threaten anything. Another interesting possibility is: 15.Bg5!? Qd7 16.Bb5 a6 17.Ba4 b5 18.Bb3 Ne7 Black plays according to his plan of exchanging heavy pieces along the c-file. 19.f3 Rxc1 20.Rxc1 Rc8 21.Rxc8† Nxc8 22.Be3 This position arose in Kummer – Huber, Feldkirch 2013, when 22...Nd6N= would have maintained the balance.

15...Na5 16.Bxg7 Kxg7 17.Bd3 Qe7 18.Rxc8 18.Qf4?! seems doubtful. 18...Nc6 19.Bb1 (no better is 19.d5N 19...exd5 20.exd5 Ne5³ when the d5pawn is weak) This was Jozefek – Banas, Slovakia 2013, and now 19...e5!N³ would have favoured

889

Black, thanks to the strong b7-bishop which puts pressure on the e4-pawn. 18.Nf4 is not dangerous. In Cherepanov – R. Volkov, Moscow 2019, it would have been good to play 18...e5!N 19.dxe5 Qxe5÷ when Black has perfectly placed pieces and pressure on the e4-pawn. Finally, another game continued 18.Qb2 Kg8 19.Rxc8 Rxc8 20.Rc1 Rxc1† 21.Qxc1 Qb4 22.Qc3 Qa4÷ and Black was again doing well in Siekierski – Jarmula, Warsaw 2013. 18...Rxc8 19.Rc1 In Sai Agni – Pranav, Kolkata 2013, Black exchanged on c1 and was about equal. However, I would be tempted to play more ambitiously with: 19...Nc6!?N 20.f3 Qd6÷ Black has an excellent position, with the usual pressure on the d4-pawn. B342) 12.Rad1 GAME 9 David Navara – Ding Liren Prague (Rapid–1) 2019 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.00 b6 11.Qd2 Bb7 12.Rad1

890

By developing the rook in this way, White avoids exchanges along the c-file. Once Black has played ...e6 – which will have to be played at some point, to fight for the centre – White can answer with d4d5; and after an exchange of pawns on that square, the second rook can be activated by Rfe1. 12...Na5 Black continues logically, aiming for the c4-square. A reasonable alternative is: 12...e6!? I only found a few examples which reached this position, but Black’s set-up is sound and deserving of further attention. 13.e5!? This ambitious move aims to shut the g7-bishop out of play, while taking a grip on the dark squares. A less committal alternative is 13.Bh6 Rc8 14.Bxg7 Kxg7 15.Rfe1 Na5 16.Bd3 Qf6 (Black can also play 16...h6!?N 17.f3 Qf6÷ to reach a complex position in which his chances are no worse) 17.Qe3 cxd4 18.cxd4 Nc4÷ when Black had a decent game in Amico – Dubko, corr. 2012. 13...cxd4 14.cxd4 Rc8 15.Bd3 Qe7 16.Bg5 Qd7 White’s play can be improved with:

17.Bb1!?N 17.Qf4?! Nb4³ was inconvenient for White in Sultan – Gopal, Al Ain 2014. I also considered 17.Be4N 17...Ne7 18.Bxb7 Qxb7 19.Nc3 Nd5 20.Ne4 Qa6 21.h4 h5 22.Rc1 Rxc1 23.Rxc1 Rc8= when further simplifications occur and the position remains balanced. 17...Ba6 18.Rfe1 Bxe2 19.Rxe2 h6! A nice trick to simplify the position. 20.Bxh6 20.Be3 is possible but 20...Rfd8= shows the benefit of driving the bishop away from g5; Black is 891

fine. 20...Qxd4! 21.Qxd4 Nxd4 22.Rxd4 Bxh6= With level material and opposite-coloured bishops, a draw is likely. 13.Bd3 cxd4 14.cxd4 Rc8 Black has achieved a lot: he owns the c-file and controls the c4-point, and his bishops exert their usual pressure on White’s pawn centre.

15.Bh6 By exchanging dark-squared bishops in this way, White reduces the pressure on the d4-pawn. However, he does not fight for the c-file at all. 15.Qb4 Nc6 16.Qa3 was played in an older game, Johannessen – Larsen, Halle/Saale 1963. It would have been quite logical to control the d5-square with 16...e6N, and after something like 17.Ba6 Bxa6 18.Qxa6 Qd7÷ Black’s position is no worse. Fighting for the c-file with 15.Rc1 deserves some attention, although Black was fine after 15...Qd6 16.f3 Rfd8 17.Rfd1 e6 18.Kh1 Rxc1 19.Rxc1 Nc6 20.Bg5 Rc8 21.e5 Qd7ƒ in Shishkin – Sukhorskij, corr. 2011. Despite the engine’s evaluation of 0.00, Black has some initiative because the d4-pawn is backward and weak, while Black owns the d5-outpost along with a mobile queenside pawn majority. Capturing space with 15.d5!? is also possible: 15...Nc4 16.Bxc4 Rxc4 17.Bd4

892

17...e5! 18.dxe6 fxe6 19.Bxg7 Qxd2 20.Rxd2 Kxg7 21.f3 e5= Orekhov – Kuznetsov, Ekaterinburg 2007. Although Black’s e-pawn is detached from the pawn base, it controls the d4- and f4-squares, thus limiting White’s knight. Black’s queenside pawns are strong and White’s a-pawn could become vulnerable, so overall the position remains balanced. 15...Nc4 16.Bxc4 Rxc4 17.Bxg7 Kxg7 Black has an excellent position: the bishop puts pressure on the e4-pawn, and the rook is active on c4. White has kept his pawn centre but it is hard to do much with it, as any pawn advances will create additional weaknesses. 18.Ng3 The knight performs only protective functions on this square, but White has nothing better. 18.e5?! is doubtful as White gives away too many central light squares. In Naumkin – Shishkin, Padova 2012, it would have been logical to play 18...Qd7!N 19.h3 Rd8µ when White has problems with the d4pawn and all of Black’s pieces are active. Naumkin improved to some degree with 18.Qe3 in Naumkin – Massironi, Nova Gorica 2019. Nevertheless, after 18...e6 19.Rc1 Ra4 20.Rc2 Qe7 21.Rfc1 Rd8 22.f3 Ba6³ Black is slightly better due to his pressure against the weak pawns on d4 and a2. 18...e6 19.d5! White should play dynamically to avoid becoming worse. The calmer 19.Qe3 was played in Naumkin – Lin, Nova Gorica 2019, when 19...Qe7N would have been promising for Black. For instance, 20.Rc1 Rfc8 21.Rxc4 Rxc4 22.Rc1 Ra4! 23.Qd2 h5³ when Black exerts pressure against the weak pawns on a2, d4 and e4, and his bishop is somewhat stronger 893

than White’s knight. 19...exd5 20.exd5

20...Re8 Developing the rook is logical enough, but alternatives also deserved attention: 20...Rc5 21.Qd4† f6 22.d6 Re8 23.f3 Qd7 24.Qf4 Rce5 25.Rf2 Re1† 26.Rf1 R1e5= is a possible route to a repetition. 20...Qf6!?N may be marginally more accurate: 21.Qd3 Qf4 22.d6 Rd8 23.Rfe1 Bc6 24.f3 b5 25.Ne2 Qf6 26.Nd4 a6ƒ The passed d-pawn is more of a weakness than a strength, so White needs to defend with care. 21.Rfe1 Rxe1† 22.Qxe1 Qd6 23.Qe2 Rc7 24.Qb2† Qf6 25.Qa3 Rd7

894

26.Qxa7? Now White gets into trouble. 26.d6!N 26...Bc6 27.Qb4 h5 28.h4 Qe5÷ would have kept things balanced. The passed d-pawn is unlikely to promote, and White’s pieces need to defend it; on the other hand, it smothers Black’s rook and generally limits his activity, so the position is about equal. 26...Rxd5 27.Rb1 Qc6 28.f3 We have reached a crucial moment, where Black needed to play accurately to exploit his opponent’s mistake from a few moves earlier.

895

28...Qc2? This seemingly logical and active move allows the victory to slip away. 28...Rd2!N was correct: 29.Ne4 (29.Qxb6? loses more quickly to 29...Rxg2†! 30.Kxg2 Qxf3† 31.Kh3 Bc8†–+ with mate to follow) 29...Rb2! 30.Rf1 Qc2 31.Nf2 Bd5–+ White’s pieces are too passive and the a2-pawn is about to fall; Black should win with accurate play. 29.Qxb6 Rd2 30.Ne2! White is saved by a remarkable defence. 30...Ba6 30...Rxe2 allows 31.Qd4† Kg8 32.Qd8† Kg7 33.Qd4†= with perpetual check. 30...g5!? is a possible winning try, creating room for the king, although White should be able to hold. For instance, 31.Rc1 Qd3 32.Qxb7 Qxe2 33.f4 Qf2† 34.Kh1 Re2 35.h3 Qxf4 36.Qc8 Rxa2 37.Qc4³ when Black keeps an extra pawn, but his winning chances with three versus two on the kingside are minimal. 31.Nf4 Qxa2 32.Qb4 Qc2

33.Ne6†! Forcing a perpetual check. 33...fxe6 34.Qe7† Kh6 35.Qh4† Kg7 36.Qe7† Kh6 37.Qh4† Kg7 38.Qe7† ½–½

896

Conclusion White’s set-up with 7.Bc4 and 8.Ne2 is a popular choice which has been studied extensively by theoreticians. After 8...Nc6 we looked at a few rare lines, including 9.Be3 0-0 10.Rc1!? where White delays castling in order to go on the offensive with h2-h4. However, Game 8 revealed a good response, where Black delivered a check on a5 to nudge the white king to f1, followed by rerouting the queen to d6 and meeting h2-h4 with ...h5. Thereafter, Black’s king is in no danger while White has problems getting the h1-rook into the game. We then moved on to the most popular 10.0-0, when 10...b6! has emerged as the most accurate move and the favourite choice of most leading Grünfeld connoisseurs. In this chapter, we looked at the various ways in which White may proceed without capturing the c5-pawn. He may try a number of slightly different plans and move orders, each with their own pros, cons and subtleties. As a general rule, Black will develop his queenside with moves like ...Bb7, ...Rc8, ...e6 and ...Na5, in conjunction with a timely exchange on d4 and trading at least one pair of rooks along the c-file, reaching a safe, harmonious position with good long-term prospects.

897

A) 12.cxb6 480 B) 12.Nd4! Ne5! 13.Nb5 Qb8 483 B1) 14.Bd5 Ng4 15.g3 Nxe3 16.fxe3 a6 484 B11) 17.Bxf7† 485 B12) 17.Bxa8 485 B13) 17.Nd4! 487 B2) 14.Be2 bxc5! 488 B21) 15.Bxc5 488 B22) 15.f4 489 B23) 15.Rb1 a6! 492 B231) 16.Na3!? 492 B232) 16.Nd4 Game 10 494

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.0898

0 b6! 11.dxc5! This is the most fundamental and principled continuation: White simply captures the pawn and challenges Black to prove his compensation. 11...Qc7 This is the only decent move. Black doesn’t try to win back the pawn at once but instead aims to improve his pieces, relying on his activity and lasting pressure against White’s queenside pawns. White’s last move picked up a pawn but increased the scope of the g7-bishop, while also freeing the e5square for Black’s pieces, especially his knight. The present position is the main tabiya for the chapter, and it has featured in numerous games involving elite players: Karjakin, Jakovenko, Bacrot, So, Carlsen and Aronian have all tried to prove something for White, while Svidler, Grischuk, Gelfand, Giri, Le Quang Liem and Nepomniachtchi are among the defenders of Black’s cause. Vachier-Lagrave has played it for both sides but mainly for Black, as you would expect from a Grünfeld specialist. In this critical position, White’s main tries are A) 12.cxb6 and B) 12.Nd4!, although several other moves have been tested. 12.Rb1 should be met by 12...Ne5 13.Bd5 Rb8, when 14.cxb6 axb6 transposes to variation A below. 12.Bd4?! occurred in a contest of strong grandmasters but has not been repeated, for good reason. In Tarjan – Smejkal, Vrsac 1983, Black could have obtained the upper hand with 12...bxc5N 13.Bxg7 Kxg7. A sample line continues 14.Qc1 Ne5 15.Qf4 Rb8 16.Qg3 h5 17.h4 Kh7 18.Rab1 Rxb1 19.Rxb1 Qd6 20.Bb3 Rd8µ when Black’s control over the d-file and powerful knight give him clearly better play.

White’s position could deteriorate further after 21.f4? (even worse is 21.Rd1?? Qxd1† 22.Bxd1 Rxd1† 23.Kh2 Ng4†–+ and White can resign) 21...Ng4 22.Bxf7 Ba6 23.c4 Qd2 24.Qe1 Qe3† 25.Kh1 Nf2† 26.Kh2 Rd2–+ when Black has a winning attack. 899

12.Qd2 has only been played in one game. My suggestion would be:

12...bxc5!N (the tempting 12...Rd8 was played in Zhukova – Nebolsina, Moscow 2010, when White could have grabbed a pawn: 13.cxb6N 13...axb6 14.Qb2 Ne5 15.Qxb6 Qxb6 16.Bxb6 Rd2 17.Bb5 Bd7 18.Nd4 Bxb5 19.Nxb5 Rdxa2 20.Rxa2 Rxa2 21.Bd4 Nc4 22.Bxg7 Kxg7 Black has enough activity to hold a draw but I would rather keep some winning chances alive) 13.Bf4 Ne5 Black has a lot of activity and the engine finds nothing better for White than trading off her bishop: 14.Bxe5 Bxe5 15.f4 Bg7 16.Rae1 Qa5 17.e5 Rd8 18.Qe3 Rb8³ Black is slightly better due to her bishop pair, active queen and rooks occupying the two open files. 12.Bd5!? was played in Ploehn – Schindler, Bavaria 2012. I suggest meeting it with 12...Ba6!?N 13.cxb6 axb6 14.Rb1 e6 15.Bxb6 Qc8, when Black seems fine, for instance: 16.Bxc6 (16.Bb3 Rb8 followed by ...Rd8 gives Black full compensation) 16...Qxc6 17.Qc2 Bb5 18.Rb4 Qxb6 19.a4 Bxa4 20.Rxa4 Rxa4 21.Qxa4 Qb2 22.Qd1 Bf6 23.Re1 Rd8 24.Qb1 Qd2 25.g3 Kg7 26.Kf1 Bxc3 27.Nxc3 Qxc3= With an absolutely equal position. 12.Qc2!? I only found two games with this move but it seems reasonable enough. 12...bxc5

900

13.f4!?N This new move seems quite logical: White prevents the knight from coming to e5 and if it moves somewhere else, he can play f4-f5 with attacking chances. 13.Bxc5? Ne5 14.Bd5 was played in Poinsignon – Laurent, France 1995, when 14...Qxc5!N 15.Bxa8 e6 would have refuted White’s play. For instance, 16.Rfb1 Ba6 17.Bb7 Bd3 18.Qd1 Qc7 19.h3 Rb8 20.f4 Bxb1 21.Rxb1 Rxb7 22.fxe5 Rxb1 23.Qxb1 Qc4–+ when White has insurmountable difficulties due to the large number of weak pawns and the superiority of the bishop over the knight in the open position. 13...Na5 14.Bd5 Rb8 15.f5 Nc6! The knight makes its way to a better square. 16.Nf4 Ne5 17.h3 Ba6 18.Rfd1 Nc4 19.Bxc4 Bxc4÷ Black is in perfect order: his king is safe and the two bishops are strong, especially the one on g7 which eyes the c3-pawn. 12.Bf4 This move has been tested several times in practice but nobody responded in the best way:

901

12...Be5!N 12...Ne5? is the only move that has been played. However, after the strong 13.c6!N± White is much better, since the c6-pawn cannot be taken by the queen due to Bd5. 13.Qc1 After 13.cxb6 play may continue: 13...axb6 14.Bb5 Bxf4 15.Nxf4 Qxf4 16.Bxc6 Ra5 17.Qd3 Ba6 18.c4 Qc7 19.Bb5 Bxb5 20.cxb5 Rfa8 21.Qb3 Ra3= The white pawns on a2 and b5 are weak, so Black has excellent compensation for the pawn. 13...Bb7 14.cxb6 axb6 15.Rd1 Ra4 16.Bd5 e6 17.Bxc6 Bxc6 18.f3 Bb5© Black has active pieces and significant pressure on the queenside pawns. 12.Nf4!? Another rare but interesting move. From here, the knight can jump to d5 or d3. 12...e6 13.Nd3 13.Qd6?! Qxd6 14.cxd6 is not too impressive. In Herzog – Chamberlain, corr. 2012, Black could have played 14...Rd8N 15.Rfd1 Na5 16.Be2 Bxc3 17.Rac1 Be5³ when White will have to play resourcefully to get any compensation at all for the doomed d6-pawn. This position was reached in Kotanjan – Khairullin, Yerevan 2014. I can offer an improvement:

902

13...Bb7!N In the game Black played 13...Na5 but I would rather prepare this move by adding a veiled threat against the e4-pawn. It is also useful to keep other options available for the knight. 14.f3 Strengthening the e4-pawn is the most natural reaction to Black’s threat. 14.Qc2?! is clumsy, and 14...Na5 15.Nb2 Qe5³ attacks c3 and e4, with excellent play for Black. 14.cxb6 axb6 15.Rb1 is met strongly by 15...Ne7!, attacking both the bishop and the e4-pawn, and the knight may go to d5 later. After 16.Qb3 Bxe4 17.Bxb6 Qc6 18.f3 Bd5 19.Bc5 Rfd8 20.Rfd1 Bxc4 21.Qxc4 Nd5 22.Nb4 Qa4µ White’s extra pawn will not last for long and Black’s active pieces put a lot of pressure on White’s position. 14...Bxc3 15.cxb6 axb6 16.Rc1 Bg7 Black has won back the pawn and the position is equal. A sample continuation is: 17.Qd2 Rfc8 18.Rfd1 Qd6!? Setting a small trap. 19.Bb3 19.Bxb6?? is punished by 19...Ne5!–+ and White suffers a fatal loss of material. 19...Na5 20.Rxc8† Rxc8 21.Qb4 Qxb4 22.Nxb4 b5 23.Rd7 Bc3 24.Bd2 Bxd2 25.Rxd2 Kf8 26.Rd6 Rc3 27.Rb6 Nxb3 28.axb3 Rxb3 29.Rxb5 Bc6 30.Rb8† Be8= The cloud of smoke has dissipated and the position is equal. Finally, another rare but interesting move is: 12.f4!? This was played by Carlsen so it’s surprising that no one has repeated it. 12...bxc5 13.Rb1 I also considered 13.Qa4!?N 13...Na5 14.e5 Nxc4 15.Qxc4 Be6 16.Qxc5 Rfc8 17.Qxc7 Rxc7 18.Bd4 Rb8© when Black has sufficient compensation for the pawn, due to his bishop pair, active rooks 903

and pressure against White’s queenside. 13...Rd8 14.Qa4 Na5 15.Bd5

15...Bd7 An interesting alternative is 15...Ba6!?N 16.c4 Rxd5 17.exd5 Nxc4 18.Rfc1 Nxe3 19.Qxa6 Nxd5 20.Qc4 Qd8 21.Rd1 e6 22.Qxc5 Bf8© when Black has good compensation: he has a pawn for the exchange, along with well-placed pieces and a solid structure. 16.Qa3 Rac8 17.f5 e6 18.Bf4 Be5 19.fxe6 fxe6 20.Bb3 Nxb3 21.axb3 Rf8÷ The position remained balanced in Carlsen – Svidler, Moscow 2011. A) 12.cxb6 axb6

904

Solidifying White’s extra pawn obviously presents a challenge to Black’s scheme, but analysis and practical testing have demonstrated that Black has excellent compensation. His minor pieces all have good prospects, and he will obtain plenty of activity along the open queenside files. 13.Rb1 White removes the rook from the long diagonal and attacks the b-pawn. This natural move has been the most common choice, although several others have been tested: 13.Qc2 removes the queen from the open file while defending both queenside pawns. In Schucman – Sena Moura, Sao Paolo 2012, it would have been good to play 13...Ba6!N. It may seem paradoxical to trade minor pieces while a pawn down, but doing so will expose White’s weaknesses even more. For instance: 14.Bxa6 Rxa6 15.Bf4 Ne5 16.Rfb1 Rfa8 17.Rb4 Qd7 18.Nd4 Nc6 19.Nxc6 Qxc6 20.e5 Ra5© Black has excellent compensation for minimal material damage, as the pawns on a2, c3 and e5 are all weak. 13.Bd5 removes the bishop from the path of Black’s queen, but it lacks stability on d5 because ...e6 can come at any moment. 13...Ba6 14.Rb1 Rab8 15.Qc2 Ne5 16.Rfd1 Rfd8 17.Bf4 was seen in Elancheralathan – Mesaros, Maribor 2012, when a good continuation would have been:

17...e6N 18.Bb3 h6!? Other moves are playable, but I want to show that Black can take his time and slowly improve his position. One of many possible continuations is 19.h3 Kh7 20.Rd4 Rbc8 21.Qd2 g5 22.Bg3 Bxe2 23.Qxe2 Qxc3 24.Rxd8 Rxd8 25.Bxe5 Bxe5 26.Rd1 Rxd1† 27.Bxd1= when Black is marginally more active but White should have no trouble holding a draw. 13.Rc1 The idea of this move is to remove the rook from the long diagonal while protecting the c3-pawn, but the rook is rather passive and Black easily develops his initiative. 905

13...Na5 14.Bd5 Bb7 It is already becoming difficult for White to defend his weak pawns on a2, c3 and e4. We will follow a practical example where White soon fell into trouble. 15.Rb1 Nc4 16.Bd4? This is a grave error which offers Black a decisive advantage. White needed to simplify with 16.Bxc4 Qxc4 17.Rb4 Qxa2 18.Bxb6 Ba6 19.Re1 Bxe2 20.Rxe2 Qa1 21.Rb1 Qxc3= when the position is absolutely equal. 16...e5 17.Bxc4 Black mishandled his advantage and even went on to lose in G. O’Connell – McElligott, Drogheda 1999. A simple improvement is: 17...exd4N 18.Bb3 dxc3–+ Material is equal but Black’s bishop pair and powerful passer on c3 should add up to a decisive advantage. 13.f3 White reinforces his remaining central pawn but gives Black plenty of freedom. 13...Ne5 With this typical move, Black activates the knight and prepares to trade some minor pieces.

14.Bd5!?N This seems like the most interesting reply. 14.Bb5 was played in Walker – Kaufman, Parsippany 2013, when 14...Ba6N 15.Bxa6 Rxa6 16.Bd4 Nc4© would have seen Black fulfil his plan by exchanging some minor pieces. White’s queenside pawns are completely immobilized while Black’s knight is excellent, and he can easily improve his position by doubling with ...Rfa8. White’s extra pawn is not felt at all, and there is a good chance that Black will regain a pawn on the queenside while keeping some degree of initiative. 14...Rb8 15.Bd4 e6 16.Bb3 Rd8 17.Qe1 Ba6 18.f4 Nd3© 906

Black’s active pieces offer full compensation. An illustrative line is:

19.Qd2 Bxd4† 20.Nxd4 Nc5 21.Rfe1 Rbc8 22.f5 Nxb3 23.axb3 Qxc3 24.Qxc3 Rxc3 25.Rxa6 Rxd4 26.f6 h5 27.Ra8† Kh7 28.Ra7 Rd2!? 28...Kg8= is quite adequate but Black can also draw in a more active manner. 29.Rxf7† Kh6 30.Rf8 Rcc2 31.Rh8† Kg5 32.f7 Rxg2† 33.Kh1 Rxh2† 34.Kg1 Rcg2†= Before White can make a new queen, Black delivers perpetual check. A final alternative is: 13.Bf4 White attacks the queen but Black was planning to put his knight on e5 anyway, and it is not at all obvious that the bishop will be any better located on f4 than on e3. 13...Ne5 14.Bb5 This was played in a high-level game. The seemingly more active 14.Bd5 Ra5 15.Qc2 Ba6 16.Rfd1 e6 17.Bb3 g5 18.Bg3 was played in C. Hernandez – Lorenzo de la Riva, Barcelona 2019, when Black would have done well to expand further on the kingside with: 18...h5!N 19.h3 Rc8ƒ The plan of ...Qe7 followed by ...h4 and ...g4 is unpleasant for White to deal with.

907

14...Ra3 15.Qc1 I also checked the active 15.Qd4!?N when 15...Be6 16.Qb4 Rfa8÷ gives Black a great game: his pieces are active and the a2-pawn may soon be destroyed. 15...Qc5 16.Qb2 Ba6 17.Bxa6 Rxa6 18.Bxe5 Bxe5 19.Rfb1 Rfa8 20.g3 Qc6© Black had more than enough play for a pawn in Timofeev – Nepomniachtchi, Moscow 2015. The bishop is stronger than the knight and the pawns on a2, c3 and e4 all require defending.

13...Ne5 14.Bd5 Rb8 Another important position has arisen, where White has tried various moves but with generally poor results.

908

15.f4 This should lead to equality with best play. 15.Qb3? was a mistake made by a talented Russian junior in Simonyan – Triapishko, Taganrog 2018. The most convincing refutation would have been: 15...e6!N (15...Ba6 is another strong move which brought Black a quick win in the game, but the text move is objectively even better) 16.f4 exd5 17.fxe5 Ba6 18.Rf2 Bc4 19.Qb4 dxe4–+ Material is equal and Black’s powerful bishops are perfectly placed to target the weak pawns on a2, c3 and e5. Another game continued 15.Qc2 Ba6 16.Rfd1 e6 17.Bb3 Rfd8 and here White must improve with:

18.Bf4N (18.Bd4? Ng4 19.g3 e5 20.Be3 Nxe3 21.fxe3 Qc5–+ left Black with two great bishops along with numerous weak pawns to pick off in Schucman – Sena, Sau Paulo 2012) 18...Rbc8 19.Rxd8† Rxd8 20.Rd1 Rxd1† 21.Qxd1 Bxe2 22.Qxe2 Qxc3= Black wins back the pawn and has no problems. 15.f3 strengthens the e4-pawn. As we have already learned, Black’s best plan is to trade some minor pieces, as shown after 15...e6 16.Bb3 Ba6 17.Bd4 Rfd8 18.Qe1 Nc4 19.f4 Bxd4† 20.cxd4 as played in Karayilan – Beltins, corr. 2012. Here Black’s game can be improved by means of:

909

20...Na5!N 21.Qd2 Qc6 22.f5 Qxe4 23.fxe6 fxe6 24.Rfe1 Nxb3 25.Rxb3 e5÷ Material is equal and the game should end in a draw, but Black has pressure against the d4-pawn and the stronger minor piece, so White still has to be careful to avoid being worse. 15...Ng4 16.Bd4 Bxd4† 17.cxd4 Ne3 18.Qc1 Nxd5 19.exd5 Qd6

20.Qc6 Qa3 21.Nc3 Bd7 22.Qxd7 Qxc3 23.Qxe7 A draw was agreed here in Bares – Pecis, corr. 2016. The likely continuation would be: 23...Qxd4† 24.Kh1 Qxd5= With level material and a simplified position. 910

B) 12.Nd4!

Jumping the knight to the centre poses the biggest problems. On the most basic level, White threatens to exchange on c6 and follow up with Bd5 to win the rook on a8. A secondary idea is to move the knight to b5 in order to harass the black queen. The text move first appeared in 1982, when Christiansen used it to defeat Jansa at the Lucerne Olympiad. The present position was tested a few more times in the 1980s but it pretty much vanished until around 2010, mostly because the entire line with 10...b6 was reckoned to be dubious for Black. 12...Ne5! This strong move should be an automatic choice. The knight attacks the bishop on c4 and will sometimes go to g4, threatening mate on h2 while also hitting the bishop on e3. 13.Nb5 This is the most ambitious continuation. Other moves are not dangerous, for instance: 13.Bb5?? proved to be positional suicide after 13...Ng4 14.Nf3 Nxe3 15.fxe3 Qxc5 16.Qe2 Bb7 17.Bd3 Rad8 18.Rac1 Bh6–+ in Neale – Gagliardi, corr. 2015. Black has a decisive advantage due to his powerful bishops and White’s atrocious pawn structure. 13.Be2 bxc5 14.Nb5 GM Sakaev recommended this for White in How to Get the Edge Against the Gruenfeld (2004) and An Expert’s Guide to the 7.Bc4 Gruenfeld (2006). The position was only recently tested in Dziuba – Molenda, Katowice 2019, when Black should have continued:

911

14...Qa5!N In the game Black opted for 14...Qc6, which is the only move mentioned by Sakaev in both of the aforementioned books. The text move is a clean equalizer, and probably one of those novelties that all the top Grünfeld players have known about for years without ever having a chance to play it. 15.Bxc5 White wins a pawn but his pieces located on c5 and b5 are unstable, and the c3-pawn is weak. 15...Nc6! Threatening ...a6, since Bb4 will no longer be a convenient reply for White. My analysis continues: 16.Be3 Rd8 17.Nd4 Qxc3 18.Rc1 Qb2 The knight on c6 is attacked twice but Black avoids losing material due to some simple tactics. 19.Rxc6 Bxd4 20.Rc2 Qb6 21.Rd2 e5 22.Bc4 Be6 23.Bxe6 Qxe6 24.Bxd4 exd4 25.Rxd4 Rxd4 26.Qxd4 Qxa2= The resulting endgame is equal, although White needs to be slightly more careful due to the passed apawn supported by the rook. 13...Qb8 Again Black does not have much of a choice. On b8 the queen remains safe from harassment, while also keeping the ...Ng4 threat alive. Now White must decide how to deal with the threat to his bishop. B1) 14.Bd5 is an aggressive continuation but the calmer B2) 14.Be2 is considered the main line at present. 14.Bf4 does not really put Black under pressure and is hardly ever played. A rare exception at GM level continued 14...bxc5 15.Rb1 Bb7 (15...a6 and 15...Qb6!?N also seem fine) 16.Qe2 Qc8 (16...e6N also looks decent) 17.Bd5 Bc6= and Black had no particular problems in Postny – Ragger, Austria 2014.

912

B1) 14.Bd5

This continuation leads to complications in which both sides must know the theory deeply. 14...Ng4 15.g3 Nxe3 16.fxe3 a6 This position has been tested at elite GM level as well as in correspondence games. White is at a crossroads: he can grab a pawn with the tactical shot B11) 17.Bxf7†, eat the rook with B12) 17.Bxa8, or recentralize his knight with B13) 17.Nd4!. B11) 17.Bxf7† Kh8 17...Rxf7 18.Qd5 Bf6 19.Nd6 e6 20.Qd1 Rf8 21.e5 Be7 22.Rxf8† Bxf8 23.Qf3 Ra7 24.Rf1 Bh6 25.cxb6 Qxb6 26.Nxc8 Qxe3† 27.Kg2 Qxf3† 28.Rxf3 leads to an endgame where Black should make a draw, but he has to be slightly careful due to White’s extra pawn.

913

18.Nd4 A fairly short draw ensued after: 18.Bd5 Rxf1† 19.Qxf1 axb5 20.Bxa8 Qxa8 21.cxb6 h5 22.Qxb5 Qxe4 23.Qe8† Kh7 24.Qxc8 Qxe3† 25.Kg2

Already in this position, the opponents agreed a draw in Kloster – Gromark, corr. 2012. Black will makes an easy perpetual check after 25...Qe2†, since 26.Kh3?? loses to 26...g5 27.Qf5† Kh6 28.g4 e6– + when White must give up his rook to avoid an immediate mate. 18...bxc5 19.Bd5 Rxf1† 20.Qxf1 cxd4 21.Rb1 Qa7 22.Qf7 Be6! This precise move solves all of Black’s problems.

914

23.Qxe6 dxc3 24.Bxa8 Qxe3† 25.Kg2 Qe2† 26.Kh3 Qh5† 27.Kg2 Qe2†= With perpetual check unavoidable, the players agreed a draw in Vachier-Lagrave – Gelfand, Tromso 2013. B12) 17.Bxa8 axb5 I was surprised to see that, so far, this position has exclusively been tested in correspondence games and never over the board. 18.Bd5 Qc7 19.cxb6 Qxb6 The results of the correspondence games show that Black’s position is in perfect order. Black has yet to taste defeat; he has excellent compensation for the exchange due to White’s numerous weak pawns and exposed kingside. 20.Qe2 This has been the usual choice, defending e3 while eyeing b5. Other moves also keep the game dynamically equal, for instance: After 20.Qd2!? Bh3 21.Rfc1 Qd6 22.Qe2 b4 23.cxb4 Bxa1 24.Rxa1 Qxb4 White has returned part of his material advantage and simplified the position, but Black still has full compensation for a pawn:

25.Rd1 e6 26.Rd4 Qb1† (a possible winning attempt is 26...Qa5!?N intending ...Qg5, ...h5 and ...Rc8) 27.Rd1 Qb4= With a repetition and a draw in De Carlos Arregui – Aguiar Garcia, corr. 2015. 20.Qf3!? e6 21.Bb3 b4 22.Rab1 bxc3 23.Bc2 Qa7 reaches another interesting position:

915

24.a4 Ba6 25.Rfd1 Be5÷ In Gluhov – Leal, corr. 2012, Black retained full compensation for the exchange. The bishops are powerful and the c3-pawn is strong, while White has to keep an eye on his weak e-pawns as well. 20...e6 21.Bb3 b4 22.a4 White also got nowhere with 22.cxb4 Bxa1 23.Rxa1 Qxb4 24.Qc4 Qa5 25.Rd1 Bb7 26.Qd4 when a draw was agreed in Hunger – Eljoseph, corr. 2016. It is hard for White to come up with any kind of active plan due to the need to keep his weak e-pawns safe. 22...Qc6 23.Qc4 Qb6 24.Kf2 Ba6 25.Qxb4

916

25...Qd8 26.Rad1 Qf6† 27.Kg2 Bxf1† 28.Rxf1 Qd8 29.Rf2 29.a5 is also not dangerous. In Wurtinger – Loeffler, email 2016, the simplest solution would have been:

29...Qd2† 30.Rf2 Qxe3 31.Rf3 Qe2† 32.Rf2 Qe3= With a repetition and a draw. 29...Qc7 30.Bc4 Rb8 31.Bb5 Bxc3 32.Qc4 Qxc4 33.Bxc4 Rb4 34.Bb5 Rxe4 35.Kf3 Rb4 36.Ra2 Rb1= The endgame was equal in Macayo – Haugen, corr. 2016. True, White has a passed a-pawn, but it is practically impossible to advance it any further due to the cooperation of Black’s rook and bishop. B13) 17.Nd4! White can pose the maximum difficulties with this move, although Black is still fine with accurate play. 17...Bh3! Everything else is worse and gives White a better game. For example, one of the best Grünfelders in the world erred with 17...bxc5? and paid the price after 18.Rb1 Qa7 19.Nc6 Qc7 20.Nxe7† Qxe7 21.Bxa8 Bh3 22.Bd5 Bxf1 23.Qxf1 Bxc3 24.Qxa6± when White was much better in Karjakin – Vachier-Lagrave, Saint Louis 2019. 18.Bxa8 Taking the rook is the most obvious idea. 18.cxb6 Bxf1 19.Bxa8 Qxa8 20.Kxf1 transposes to the main line below.

917

18.Nc6!?N awaits a practical test. Best play continues 18...Qe8! 19.cxb6 Rc8 20.b7 Rxc6 21.Rb1 Be5 22.Qa4 Rf6 when the b7-passer should cancel out Black’s extra piece:

23.Qxe8 Rxe8 24.Rxf6 exf6 25.Rb6 a5 26.Ra6 Bc7 27.Rc6 Be5 28.Ra6 Bc7= The game should end in a draw by repetition, since it is not profitable for White to play 29.Rxf6?! due to 29...Be6µ and Black is clearly better. 18.Rb1!? leads to interesting play but Black has at least enough resources: 18...Bxf1 19.Qxf1 b5 20.a4 e6 21.Bxa8 Qxa8 22.axb5 axb5 23.Qd3 Rd8

24.Rxb5?! (24.Qc2N is an improvement although 24...e5 25.Nxb5 Qc6 26.c4 Qxc5 27.Qe2 Bh6 28.Kf2 f5„ offers Black a fine initiative for the pawn, and it is White who must look for a way to equalize) 24...e5 25.Rb6 exd4 26.cxd4 Qa5³ This occurred in Sande – Tiemann, corr. 2013. White’s 918

three extra pawns are the material equivalent of Black’s bishop, but the exposed position of the white king and the inability of his pawns to advance give us reason to evaluate the position as offering winning chances for Black. 18...Bxf1

19.Kxf1 I also checked 19.cxb6N 19...Qxa8 20.Qxf1 (20.Kxf1 transposes to the main line below) 20...Rb8 21.Qb1 e5 22.Nf3 Qc6 23.Qd3 Qxb6 24.Rd1 Bf6 25.Kf2 Qc6© when Black has excellent compensation for the missing pawn, due to White’s many pawn weaknesses. 19...Qxa8 20.cxb6 This position was reached in Reichgeld – Nagel, email 2013. My preferred continuation is: 20...Qxe4N 21.Kf2 Rb8 22.Qb3

919

22...Bf8!? This is not the only adequate move, but it’s an interesting and unusual way to activate the bishop, which is heading for c5. A sensible conclusion could be: 23.Re1 Qb7 24.Rb1 e5 25.Nf3 Bc5 26.Nxe5 Qe4 27.Qxf7† Kh8 28.Qf6† Kg8 29.Qf7†= With perpetual check. B2) 14.Be2 This move has been tested in many more practical games than 14.Bd5. Rather than hunting the rook in the corner, White prevents ...Ng4 and keeps his kingside structure intact. Next he plans f2-f4 to chase the active knight away.

920

14...bxc5! First played by Grischuk against Aronian in 2011, this was the key move that revived the entire 10...b6! system at the top level. Thereafter, players all over the world had to rethink the long-held evaluation of that move. One of the early games continued 14...a6?! 15.Na3 b5, and now in Knaak – Ftacnik, Bratislava 1983, White could have secured a substantial advantage with 16.f4N 16...Nc6 17.e5±. White has three serious options here: B21) 15.Bxc5, B22) 15.f4 and B23) 15.Rb1. B21) 15.Bxc5 This natural-looking capture enables Black to claim typical compensation for the pawn. 15...a6 Now White must decide on a route for his knight. 16.Na3!? This doesn’t look beautiful, but White intends to manoeuvre the knight via c2 and perhaps b4. The natural-looking 16.Nd4 was well handled in a GM game: 16...Qc7 17.Nb3 a5 18.Bd4 a4 19.Nc5 Rd8 20.Qc1 Ng4 21.Bxg4 Bxg4 22.h3 e5 23.Be3 Be2 24.Re1 Bc4 25.Qc2 Bf8 26.Nxa4 Bd3 27.Qb3 Qc6 28.Nb6 Ra3ƒ Black’s pieces enjoyed a lot of activity in Yakubboev – Le Quang Liem, Sharjah 2019; White’s eventual victory in the game does not reflect the outcome of the opening. 16...Qc7 17.Bd4 Rd8 18.Qc1 Ng4 19.Bxg4 Bxg4 20.Qe3 Rab8 21.e5!? 921

21.Bxg7 Kxg7 22.Nc2 Rbc8 23.Qg5 Be2 24.Rfe1 h6 25.Qc1 Bd3 26.Nb4 Qxc3 27.Qxc3† Rxc3ƒ occurred in Bacrot – Vachier-Lagrave, Biel 2013, when Black’s more active pieces gave him slightly better chances in the endgame. The text move avoids simplifications and leads to a more complex battle. In Hagen – Kislik, Budapest 2010, a natural continuation would have been:

21...Rb2!N 22.h3 Be6 23.Rfb1 Rxb1† 24.Rxb1 Qd7 25.Qe2 Qa4 26.Qb2 a5© Black has good compensation for a pawn, since he has two bishops and quite active pieces. White, on the other hand, lacks harmony for his pieces and needs to take care of the weak pawns on a2, c3 and e5. B22) 15.f4

922

This is a natural and principled move, but GM and correspondence praxis has proved the reliability of Black’s position: he has yet to taste defeat from here. 15...Ng4 16.Bxc5 a6 17.Na3 Qc7 18.Bd4 e5 19.fxe5 Nxe5 After a few forced moves, we reach a crossroads where White has tried a couple of different queen moves. 20.Qc1 Several strong GMs have chosen this move. The alternative is: 20.Qd2 Rb8 21.Nc2 Rb2

923

22.Kh1 This was Bacrot’s choice. 22.Rfb1 was tried a few years later in Shchepetnev – Barata, corr. 2017, when in my opinion it was best to play a small combination: 22...Nf3†!N 23.gxf3 Bxd4† 24.cxd4 Rxc2 25.Qe3 Bd7 Despite the improvement to White’s pawn structure, Black’s activity is enough to maintain the balance. For instance: 26.Rc1 Rc8 27.Rxc2 Qxc2 28.Bxa6 Qb2 29.Re1 Bh3 30.Bf1 Bxf1 31.Rxf1 Rc3 32.Qh6 Rc2 33.d5 Qd4† 34.Kh1 Qd3 35.Kg1 Qe2 36.Qh3 Qe3† 37.Kh1 Rxa2= White cannot make progress as Black’s pieces are so active. 22...Rd8 The move chosen by MVL is one of several options giving Black an acceptable game. For example, it’s quite possible to play 22...Ng4N 23.Bxg4 Bxd4 24.Qxd4 Rxc2 25.Bxc8 Qxc8= when Black will soon win a pawn back, with an absolutely equal position. I also like 22...Be6!? 23.Rfb1 Rfb8 24.Rxb2 Rxb2© when, due to Black’s considerable activity, White’s extra pawn is not felt. 23.Bd1? White should have preferred 23.Qc1!N 23...Rxc2 24.Bxe5 Rxc1 25.Bxc7 Rxf1† 26.Bxf1 Rd2= when the endgame is equal, as Black will soon win a pawn back. The text move was too passive in Bacrot – Vachier-Lagrave, Dubai (rapid) 2014. A nice idea for Black is:

924

23...a5!N There is nothing wrong with the natural 23...Be6, which gave MVL some advantage in the game, but I like the bishop on a6 even more. 24.Qf4 Ba6 25.Rf2 h5 26.h3 Rd7µ Black’s pieces are extremely active and he can slowly increase pressure, while the white pieces suffer from clumsiness; in particular, the bishop on d1, knight on c2 and rook on a1 should be pitied.

20...Bg4 21.Bxg4 Nxg4 22.Qf4 Qxf4 23.Rxf4 Ne5 The position is becoming increasingly simplified and White still has an extra pawn, but let’s take a closer look at the details. White lacks harmony and his pieces are scattered seemingly at random across the board, so it will take time for them to get coordinated. The pawns on a2, c3 and e4 also need to take care of themselves, as Black will start hunting for them. Black’s play is generally quite simple; due to 925

the activity of his pieces and White’s pawn weaknesses, he easily proves his compensation. 24.Rb1 This was played in a game between two top Russian GMs. 24.Nc2 was a more recent try, hoping to improve the failed knight. 24...Nc6 25.Rb1 (I also checked 25.Bxg7 Kxg7 26.Rd1 Rac8 27.Ne3 Ne5 28.Nd5 Rc4 29.Rd4 Rxd4 30.cxd4 Nc6 31.Rf6 Rc8 32.Kf2 Nxd4 33.Rxa6 Rc2†= with a likely draw)

25...Nxd4 26.cxd4 Rac8 27.Rf2 Rc4 28.Rb7 Bxd4 29.Nxd4 Rxd4= A draw was agreed a few moves later in Wang Yue – Xu Xiangyu, Hangzhou 2018. I also considered 24.Rff1!?N with the idea of improving the rooks first: 24...Rac8 25.Rab1 Nd3 26.Rb7 Be5

926

27.Ra7 (27.Rd1 is not dangerous in view of 27...Nc5 28.Re7 Bd6 29.Ra7 Nxe4 30.Rxa6 Rfd8 31.Nb5 Bc5 32.Rf1 Nxc3 33.Bxc5 Ne2† 34.Kh1 Rxc5 when Black wins back the pawn and even has some winning chances due to White’s vulnerable king) 27...Rfd8 28.Nc2 Bxd4† 29.Nxd4 Rxc3 30.Ne2 Rc2 31.Rfxf7 Rxe2 32.Rg7†= With perpetual check. 24...Rad8 24...Rac8 has also proved to be a reliable choice in a couple of games.

25.Rff1!?N This seems slightly more ambitious than 25.Nc2 Nc4 (rather than 25...Nd3?! 26.Rff1 when the purpose of the knight on d3 was not clear in Aronian – Grischuk, Kazan 2011) 26.Rb4 Rc8 27.Ra4 927

Bxd4† 28.Nxd4 Ne5 29.Rxa6 Rxc3 30.Rf1 Rc4 31.Nf3 Rxe4 when the players agreed a draw in Jakovenko – Svidler, Tbilisi 2015. 25...Rd7 26.Rb6 Nd3 27.Bxg7 Kxg7 28.Rd1 I also checked 28.Rxa6 Nc5 29.Rc6 Nxe4 30.Nb5 Ra8 31.a3 when Black would be equal even if he did nothing and offered White a free move. The simplest continuation looks to be 31...Rb8 followed by ...Nxc4, winning back the pawn with an obvious draw. 28...Re8 29.Rxa6 Rxe4 30.Nb5 Re3 31.a4 Rde7 32.h3 Nc5 33.Ra8 Re2 34.Nd4 Ra2© With ideas such as ...Rb7-b2 on the agenda, White should look for a way to return his extra pawn and simplify to a draw. B23) 15.Rb1

This continuation is the most ambitious. White places the rook opposite the enemy queen to create some unpleasant tension. 15...a6! It is best to kick the knight away without delay, before improving the position of the queen and completing the development of the queenside pieces. White needs to determine the future of his knight: B231) 16.Na3!? is a tricky option which has been tested at the highest level, but I would consider B232) 16.Nd4 to be the main line. B231) 16.Na3!?

928

White places the knight at the edge of the board but intends to bring it to the more active c4-square at some point. Another advantage of this move is that White saves a tempo compared to variation B232 below, as the knight will not be attacked after Black moves his queen. 16...Qc7 17.f4 This is the only ambitious move, so nothing else is ever played. 17...Rd8 18.Qc2 Ng4 19.Bxg4 Bxg4 20.f5 After a few relatively forcing moves, an important position has arisen. 20...gxf5!? This move has occurred in practice a couple of times. It leads to a complicated battle where any result is possible, and the cost of a mistake is high for both sides. I found another idea which has not yet been tested in practice: 20...Qa5!?N This is a relatively safe option which should lead to an equal position. 21.Rb3 Qa4 Threatening ...c4. 22.Bxc5 22.c4?! is worse: 22...Qd7! With the surprising threat of ...Bd1. 23.Qf2 Qc6 24.Bxc5 Qxe4 25.fxg6 fxg6 26.Qf7† Kh8 27.Qxe7 Qxe7 28.Bxe7 Rd2ƒ White faces a difficult defence due to Black’s powerful bishops and excellent rook on the 2nd rank.

22...Qd7! Again Black threatens to win the exchange in a most unusual fashion with ...Bd1. 23.Qf2 Qd2 24.Nc4 Qxf2† 25.Kxf2 929

I also considered 25.Rxf2 Rd1† 26.Rf1 Rxf1† 27.Kxf1 Bd1 28.Rb7 Bc2 29.Nb6 Bxe4 30.Rxe7 Bd3† 31.Kf2 Rd8 32.fxg6 hxg6 33.c4 Bf5© when Black has full compensation due to his active pieces: the two bishops are strong as always, and the rook may burst into action on d2 or d1 at any moment.

25...gxf5 26.Bxe7 Re8 27.Nb6 Ra7 28.Nd5 Raxe7 29.Nxe7† Rxe7 30.Rb8† Bf8 31.h3 Bh5 32.exf5 f6 33.Rfb1 Bf7 34.Rc8 Be8 35.g4 Bd7 36.Rcb8 Bb5= Despite the unbalanced material situation of rook and two pawns versus two bishops, the endgame is dynamically balanced and should be drawn after accurate play from both sides. 21.Nc4 e6 22.h3 Bh5 23.exf5 exf5 24.Bg5

We have reached a key position where Black needs to be accurate – and this is after 24 moves! The Grünfeld is an ambitious opening and this is one of those critical lines where you need a good memory 930

in order to navigate through the various nuances, traps and important manoeuvres. 24...Rf8! This move would be hard to find over the board; but once you see the problem with the alternatives, the logic for it becomes clear. 24...Re8? runs into trouble after 25.Nb6 Rab8 26.Nd5± when Black has serious coordination problems and must worry about a possible check on e7. This could have been avoided if e8-square was available for the queen’s rook! 24...f4?! was MVL’s choice in the original game with 16.Na3, but 25.Qf2 gave White some advantage – especially after the further mistake 25...f3? 26.Bxd8 Rxd8 27.Qh4 fxg2 28.Rfe1+– when White went on to convert his material advantage in Karjakin – Vachier-Lagrave, Stavanger 2018. 25.Bf4 Now 25.Nb6 can be conveniently met by 25...Rae8 26.Nd5 Qd6 when Black gets a comfortable position. After something like 27.c4 Re2 28.Qd3 Qe5÷ the position is theoretically equal but I would prefer Black in a practical game, due to the bishop pair and active rook on e2. 25...Qc6 26.Rbe1 Rfe8÷ An unclear position has arisen, where Black has a full share of the chances. 27.Be5?! This was played in the only game to reach this position, but it enables Black to claim a slight plus. It was more accurate to play 27.Nd6!? Rxe1 28.Rxe1 Bg6 29.Qd3 Rd8 30.Qg3 h5÷ when White’s active pieces and better structure give him enough compensation for the pawn. Still, Black has his own trumps: he keeps an extra pawn for the time being, his pieces are decently placed (even the passive bishop on g6 performs a useful function in gluing the kingside together) and the queen may disturb the a2- and c3-pawns in the future. 27...Bg6 28.Qf2 In Arvola – Tari, Larvik 2019, Black played 28...f6 to preserve his bishop pair. Although his position remained quite playable, I think the top priority should be to complete development with:

931

28...Rad8!N 29.Bxg7 Kxg7 30.Ne3 Re5³ Despite Black’s fractured kingside structure, his extra pawn is definitely of some value and White will have to battle for a draw. B232) 16.Nd4 GAME 10 Surya Ganguly – Anish Giri Leon 2012 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.00 b6 11.dxc5 Qc7 12.Nd4 Ne5 13.Nb5 Qb8 14.Be2 bxc5 15.Rb1 a6 16.Nd4

932

The knight is heading for b3, at which point the pressure on the c5-pawn will force it to advance. This in turn will weaken certain squares, which White will try to exploit. 16...Qc7 17.Nb3 Rd8 18.Qc2 c4 After a few forced moves, White must decide what to do with his knight. 19.Nc5 Here the knight stands most actively, getting on Black’s nerves. I checked two untested possibilities: 19.Nd4N 19...Bg4 20.f3 Bd7 21.Rfd1 Rab8÷ and Black’s position is no worse. An interesting possibility is ...Nd3, sacrificing a pawn in return for the advantages of the bishop pair and pressure against the c3-pawn. 19.Nd2N

933

19...Rd6! Defending against Bb6. 20.h3 Be6 21.Bd4 Rad8 Black has a fine position, especially in view of the following attractive idea which I discovered: 22.f4 Nc6 23.Be3

23...Rd3! 24.Bxd3 Rxd3 25.Bb6 Qd6ƒ Black has excellent compensation for the exchange: his pieces are extremely active and the c3-pawn is about to drop, leaving Black with a strong passed pawn and chances to become even more active with ...Nd4 or ...Nb4-d3.

934

19...Nd3! This simplifying move solves the last of Black’s opening problems. 20.Nxd3 cxd3 21.Bxd3 Qxc3 22.Qxc3 Bxc3 23.Bc4 Bd7!? 23...Be6 24.Bxe6 fxe6 is safe enough, as the doubled pawns are not really weak at all. 25.Rb7 Kf7 26.Rfb1 occurred in Fleetwood – Sethuraman, corr. 2015, when 26...Bd4N= would have been the simplest way to steer towards a draw. The text move is a better choice if Black wishes to keep some winning chances alive, although the position is still within the limits of equality of course.

935

24.Rfc1 Bf6 Despite the symmetrical pawn structure and limited material, the position is not yet a dead draw, as the game demonstrates. 25.Kf1 Centralizing the king is a reasonable choice. White has tried a few other ideas in correspondence games: 25.Bf4 e6 26.e5 Bg7 27.g3 h6 28.h4 Rdc8 29.Rb6 Bb5 30.Bxb5 Rxc1† 31.Bxc1 axb5 32.Rxb5 Rxa2= was absolutely equal in Plomp – Nataf, corr. 2014. 25.g3 Bb5 26.Kg2 Kg7 27.Bb6 Rdc8 28.Bd5 Rab8 29.Ba7 Rxc1 30.Rxc1 Rd8 31.Rc7 Rd7 32.Rxd7 Bxd7 33.Kf3 e6= and the position was approaching the inevitable draw in Ackermann – Loeffler, email 2016. 25.f4 e6 doesn’t really change anything: 26.g3 Bb5 27.Kf2 Bd4 28.Rc2 Rac8 29.Rbc1 Rb8 30.Bb3 Bxe3† 31.Kxe3 Rd3† 32.Kf2 and it was clearly equal in Mary – Wettstein, corr. 2015.

25...Bb5 26.Be2 This is not a mistake exactly; but if White was happy with a draw, then 26.Bxb5 axb5 27.Rxb5 Rxa2= would have been a simpler way of achieving it. 26...Bd4 27.Bxd4 Rxd4 28.Bxb5 axb5 29.Rc2 But not 29.Rxb5? when 29...Rxa2µ creates problems. 29...Rxe4 30.Rxb5 Ra7 936

Even though the endgame should clearly be a draw with best play, Black has reason to play on. White’s a-pawn is going nowhere and Black can easily attack it before advancing his kingside pawns.

31.g3 Rea4 32.Rbb2 With hindsight, it was worth considering 32.h4!?, simply sacrificing the a-pawn to reach a standard endgame a pawn down where the drawing technique is well known. 32...g5 33.Kg2 Kg7 34.Re2 f6 35.Re6 Kf7 36.Rbe2 h5 37.h3 h4 38.g4 38.Kh2!?³ is the engine’s preference but it is not easy to play such moves.

38...Ra3 39.Kh2 R3a6 40.Rxa6 Rxa6 41.Kg2 Ra3 42.Rb2 e5 43.Rb7† Ke6 44.Rb6† Kf7 45.Rb7† Kg6 46.Rb2 Ra6 47.Re2 Kf7 48.Rd2 Ke6 49.Rb2 Kd5 50.Rd2† Ke4 51.Rb2 Ra4 52.Re2† Kd5 937

53.Rd2† Ke6 54.Rb2 Ra6 55.Rd2 f5 56.gxf5† Kxf5 57.f3 Ra3 58.Re2 Ke6 59.Rd2 Kf6 60.Rb2 Ke7 61.Rd2 Ke6 Having defended well for a long time, Ganguly makes a decisive error.

62.Kf2?? White could have drawn with 62.Re2, or even 62.Rd8!? Rxa2† 63.Kg1 when Black has no real way to make progress. 62...e4! White’s previous move proves to be doubly unfortunate: the h3-pawn is no longer defended, and the position of the king on the f-file means that Re2 is not a valid defence, as the f3-pawn will fall with check. 63.fxe4 Rxh3 Black’s two connected passed pawns will decide the outcome of the duel. 64.Rd5 Kf6 65.a4 Ra3 66.e5† Ke6 67.Rd4 Kxe5 68.Rb4 Kf5 69.Rc4 Kg6 70.Kg2 Kh5 0–1 Conclusion We meet the ambitious 7.Bc4 set-up with 7...c5 8.Ne2 Nc6 9.Be3 0-0 10.0-0 b6!, after which 11.dxc5! is the critical continuation which has been the focus of this final chapter. After our automatic reply of 11...Qc7, we started by examining 12.cxb6 axb6, reaching a thematic structure where Black has excellent compensation for the sacrificed pawn due to his pressure on the queenside, thanks to the open

938

a- and c-files and the g7-bishop which drills along the h8-a1 diagonal. 12.Nd4! is a better move, when 12...Ne5! 13.Nb5 Qb8 leads to another branching point. 14.Bd5 Ng4 15.g3 Nxe3 16.fxe3 a6 reaches a sharp position where 17.Bxf7†, 17.Bxa8 and 17.Nd4! all deserve attention. Black is fine in all cases, although a certain amount of memory work is essential. 14.Be2 bxc5! reaches another critical juncture. Then 15.Bxc5 a6 and 15.f4 Ng4 16.Bxc5 a6 both offer Black fine compensation for the sacrificed pawn, so 15.Rb1 can be considered the main line, although 15...a6! gives Black good prospects after either knight retreat. Here too, deep preparation is needed, but the same will be true for White as well, and I was unable to find any advantage for him in any of the variations.

939

Variation Index Chapter 1 1.d4 Nf6 2.c4 g6 3.f3 c5 A) 4.dxc5!? 18 B) 4.d5 d6 5.e4 e6 19 B1) 6.Ne2!? 19 B2) 6.Nc3 exd5 7.cxd5 Bg7 22 B21) 8.Be3 22 B22) 8.Nge2 26 B23) 8.Bg5 31 Chapter 2 1.d4 Nf6 2.c4 g6 3.g3 c6 4.Bg2 d5 A) 5.Qa4!? 37 B) 5.Nf3 Bg7 6.0-0 0-0 39 B1) 7.Nc3?! 41 B2) 7.Nbd2 43 B3) 7.b3 dxc4! 8.bxc4 c5! 49 B31) 9.e3 49 B32) 9.Bb2 50 B4) 7.Qb3 54 C) 5.cxd5 cxd5 6.Nf3 Bg7 62 C1) 7.0-0 62 C2) 7.Nc3 Ne4 66

940

C21) 8.Qa4† 67 C22) 8.Qb3 68 Chapter 3 1.d4 Nf6 2.c4 g6 3.Nc3 d5 A) 4.f3?! 75 B) 4.h4?! 76 C) 4.Nf3 Bg7 79 C1) 5.g3 dxc4 6.Qa4† Nfd7 7.Qxc4 Nb6 79 C11) 8.Qb3 80 C12) 8.Qd3 81 C2) 5.h4!? c6! 82 C21) 6.Qb3!? 83 C22) 6.Bg5!? 84 C23) 6.cxd5 85 Chapter 4 1.d4 Nf6 2.c4 g6 A) 3.Nf3 Bg7 4.e3 91 B) 3.Nc3 d5 4.e3 Bg7 96 B1) 5.Qb3 97 B2) 5.Nf3 0-0 101 B21) 6.b4 103 B22) 6.cxd5 104 B23) 6.Be2 c5! 110 B231) 7.dxc5 111 B232) 7.0-0 112

941

Chapter 5 1.d4 Nf6 2.c4 g6 3.Nc3 d5 A) 4.Bg5 Bg7!? 119 A1) 5.e3 119 A2) 5.Bxf6 Bxf6 121 A21) 6.Nxd5?! 121 A22) 6.cxd5 c6! 121 A221) 7.dxc6 122 A222) 7.Rc1!? 123 A223) 7.e4 0-0 125 A2231) 8.Bc4?! 126 A2232) 8.Nf3 127 A2233) 8.e5 129 B) 4.Nf3 Bg7 5.Bg5 Ne4! 133 B1) 6.Bf4 134 B2) 6.Bh4 Nxc3 7.bxc3 dxc4! 142 B21) 8.e3 143 B22) 8.Qa4† 144 B3) 6.cxd5 Nxg5 7.Nxg5 e6 148 B31) 8.Qd2 150 B32) 8.Nf3 156 Chapter 6 1.d4 Nf6 2.c4 g6 3.Nc3 d5 A) 4.Bf4 Bg7 5.e3 0-0 164 A1) 6.Qb3 166 A2) 6.cxd5 168 A3) 6.Rc1 Be6! 170

942

A31) 7.cxd5 171 A32) 7.c5 Game 1 173 A33) 7.Qb3 c5! 8.Qxb7 Qb6 9.Qxb6 axb6 180 A331) 10.dxc5 182 A332) 10.Nf3 184 B) 4.Nf3 Bg7 5.Bf4 0-0 187 B1) 6.Rc1 187 B2) 6.e3 c5! 189 B21) 7.Qb3 190 B22) 7.dxc5 Ne4! 192 B221) 8.Be5 193 B222) 8.Rc1 194 Chapter 7 1.d4 Nf6 2.c4 g6 3.Nc3 d5 A) 4.Qa4†!? Bd7 5.Qb3 dxc4 6.Qxc4 Bg7 202 A1) 7.Bf4!? 203 A2) 7.e4 205 B) 4.Nf3 Bg7 5.Qa4† Bd7 6.Qb3 dxc4 7.Qxc4 0-0 208 B1) 8.g3!? 209 B2) 8.e4 210 B3) 8.Bf4 214 Chapter 8 1.d4 Nf6 2.c4 g6 3.Nc3 d5 A) 4.Qb3!? dxc4 5.Qxc4 Bg7 6.e4 0-0 220 A1) 7.Bf4 220 A2) 7.Be2!? 221 943

B) 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 224 B1) 7.Bf4 Na6 8.e4 c5! 225 B11) 9.Rd1!? 226 B12) 9.dxc5 Game 2 227 B2) 7.e4 a6! 233 B21) 8.Qa4!? 236 B22) 8.Bf4 b5! 9.Qxc7 Qxc7 10.Bxc7 Bb7 238 B221) 11.e5 241 B222) 11.Bd3 243 Chapter 9 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 0-0 7.e4 a6! A) 8.e5 b5 9.Qb3 Nfd7 247 A1) 10.Be2!? 249 A2) 10.Ng5!? 251 A3) 10.h4!? Game 3 253 A4) 10.Be3 c5 11.e6! cxd4! 258 A41) 12.exf7† 261 A42) 12.Nxd4 261 A5) 10.e6 fxe6 264 A51) 11.Qxe6† 266 A52) 11.Be3 Nf6 269 A521) 12.h4 270 A522) 12.a4! 271 B) 8.Be2! b5 276 B1) 9.Qd3 276 B2) 9.Qb3 c5! 10.dxc5 Be6! 11.Qc2 Nbd7 277

944

B21) 12.c6!? 279 B22) 12.Be3 Game 4 281 Chapter 10 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 A) 5.Nf3 Bg7 291 A1) 6.Qa4†!? 292 A2) 6.Bg5!? 294 B) 5.g3 Bg7 6.Bg2 Nxc3 7.bxc3 c5 298 B1) 8.e3 299 B2) 8.Nf3 Nc6 301 B21) 9.e3 302 B22) 9.Be3 304 C) 5.Na4!? e5!? 307 C1) 6.e4!? 309 C2) 6.dxe5 Nc6 311 C21) 7.f4 312 C22) 7.a3 313 Chapter 11 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.Bd2 Bg7 A) 6.Nf3 317 B) 6.e4 Nxc3 7.Bxc3 0-0 320 B1) 8.Bc4 322 B2) 8.Rc1 324 B3) 8.Nf3 326 B4) 8.Qd2 c5! 9.d5 Bxc3 10.bxc3 Qd6! 327

945

B41) 11.Nf3 329 B42) 11.f4 330 Chapter 12 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 A) 7.Ba3 Nd7! 8.Nf3 c5! 9.Qb3! 0-0 333 A1) 10.Bd3 337 A2) 10.Be2 338 B) 7.Bg5 Game 5 341 C) 7.Qa4†!? Qd7! 350 C1) 8.Bb5!? 351 C2) 8.Qa3 353 C3) 8.Qb3 354 D) 7.Bb5†!? c6! 8.Ba4 0-0 9.Ne2 b5! 10.Bb3 a5 356 D1) 11.a4 359 D2) 11.0-0 Game 6 360 Chapter 13 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 A) 7.Be3 c5 8.Qd2 Qa5 369 A1) 9.Rb1 369 A2) 9.Rc1 371 B) 7.Nf3 c5 8.Be3 Qa5 374 B1) 9.Bd2 374 B2) 9.Nd2 Bd7!? 10.Rb1 Qxc3 376 B21) 11.dxc5 377 B22) 11.d5!? 379 946

B3) 9.Qd2 0-0 10.Rc1 Nd7! 380 B31) 11.Bh6 384 B32) 11.h4!? 386 B33) 11.d5 388 B34) 11.Bd3 392 Chapter 14 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Nf3 c5 A) 8.Bc4 400 B) 8.h3!? 403 C) 8.Be2 Nc6 405 C1) 9.Be3 406 C2) 9.d5!? 408 D) 8.Bb5† Nc6 411 D1) 9.d5 412 D2) 9.0-0 413 E) 8.Rb1 0-0 9.Be2 Nc6! 10.d5 Ne5 11.Nxe5 Bxe5 417 E1) 12.Rb3 426 E2) 12.Qc2 427 E3) 12.Qd2 e6 13.f4 Bc7 14.0-0 exd5 15.exd5 Ba5 428 E31) 16.Ba3 433 E32) 16.f5 435 E33) 16.d6! Game 7 438 Chapter 15 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 A) 9.d5 444 947

B) 9.Be3 0-0 447 B1) 10.Rb1!? 448 B2) 10.Rc1!? cxd4 11.cxd4 Qa5† 451 B21) 12.Qd2 452 B22) 12.Kf1 Game 8 454 B3) 10.0-0 b6! 459 B31) 11.h4!? 461 B32) 11.Rb1 463 B33) 11.Rc1 466 B34) 11.Qd2 Bb7 468 B341) 12.Rac1 469 B342) 12.Rad1 Game 9 471 Chapter 16 1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0-0 10.00 b6! 11.dxc5! Qc7 A) 12.cxb6 480 B) 12.Nd4! Ne5! 13.Nb5 Qb8 483 B1) 14.Bd5 Ng4 15.g3 Nxe3 16.fxe3 a6 484 B11) 17.Bxf7† 485 B12) 17.Bxa8 485 B13) 17.Nd4! 487 B2) 14.Be2 bxc5! 488 B21) 15.Bxc5 488 B22) 15.f4 489 B23) 15.Rb1 a6! 492 B231) 16.Na3!? 492

948

B232) 16.Nd4 Game 10 494

949