Playing 1.e4 e5: A Classical Repertoire 1784830143, 9781784830144

Citation preview

Contents   Title Page Key to Symbols used Foreword by Parimarjan Negi Bibliography Introduction Summary of Recommendations Open Games 1 Early Deviations and Gambits The Portuguese Opening The Nakamura Variation The Ponziani Opening The Göring Gambit The Centre Game The King’s Gambit 2 Bishop’s Opening and Vienna Game The Bishop’s Opening The Vienna Game 3 Four Knights – Introduction The Halloween Gambit 4.a3 4.d3 4.h3 4.Be2 The Italian Four Knights The Glek System 4 Four Knights – 4.d4 and 4.Bb5 The Belgrade Gambit The Scotch Four Knights The Spanish Four Knights 5 Scotch Game On the Way to the Main Line The Main Line 6 Two Knights – 4.d4 and 4.d3 The Original Morphy Attack The Improved Morphy Attack The Modern Variation – Introduction 4.d3 – Introduction and 5.Nc3: Main Lines after 5.0–0 0–0 7 Two Knights with 4.Ng5 The Kieseritzky Variation 8.Bb5 8.Qf3 8.Bd3! – The Modern Trend

Spanish 8 Exchange Variations The Exchange Variation, 5.Nc3 f6 The Simple 5.d4 The Main Line The Delayed Exchange Variation 5.d4 – The Central Attack 9 On the Road to the Main Line Spassky’s 5.Nc3 The Worrall System Just Before the Main Lines 10 The Trendy d2-d3 5.d3 The Fashionable 6.d3 Anand’s concept with 8.a3! 11 Breyer – 10.d3 and Sidelines after 10.d4 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8 Introduction to 10.d4 12 Breyer – Alternatives to 13.Nf1 13.b4 Bf8 14.a4 Nb6 15.a5 Nbd7 16.Bb2 Rb8 Introduction to 13.a4 The Super-Critical 20.b4! 13 Breyer Main Lines 14.Bg5 14.Ng3 g6 15.Bg5 15.b3 The Sophisticated 18.Bg5 15.a4   Index of Main Games Variation Index

A Classical Repertoire

Playing 1.e4 e5 By

Nikolaos Ntirlis

  Quality Chess http://www.qualitychess.co.uk       First edition 2016 by Quality Chess UK Ltd Copyright © 2016 Nikolaos Ntirlis     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 978-1-78483-014-4

Hardcover ISBN 978-1-78483-015-1

 

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 America by National Book Network   Distributed in Rest of the World by Quality Chess UK Ltd through

Sunrise Handicrafts, ul. Poligonowa 35A, 20-817 Lublin, Poland

  Typeset by Jacob Aagaard

Proofreading by Colin McNab, Danny McGowan and John Shaw

Edited by Andrew Greet

Cover design by adamsondesign.com

Cover Photo by Capture365.com

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

 

Acknowledgements I’d like to express my gratitude to: Sabino, Sune, Sam, John, Elshan, Ioannis and Sotiris, for sending me important feedback

on various lines of the book.

Jacob and George, for promoting my work and for their general support.

Last but not least, Andrew, for being a most helpful and understanding editor.

Foreword by Parimarjan Negi   These days, the art of chess analysis is completely different from that of playing the game. You have to be creative, diligent, and constantly guide the computer in the direction you want – but often, you let the machine do the heavy lifting. This is in stark contrast to the practical skills possessed by tournament players – indeed, it can often be detrimental to one’s own game. Practical players therefore have to be wary about delving too deeply into the art of analysis, and a natural consequence is that their skills and understanding of engines will not be as refined as those of correspondence specialists.   Nikos Ntirlis made his decision to be an analyst early. The first time I met Nikos was at the 2006 Greek Team Championship. I had just become a Grandmaster, while he was an enthusiastic member of the local club. I had never focused too much on opening theory before then, and I was surprised to hear Nikos talk enthusiastically about the intricacies of deep Grünfeld lines that even I wasn’t familiar with. Now, almost a decade later, Nikos has improved as a player, but it’s as an opening analyst where he has channelled his passion to produce outstanding results. Since he has clearly defined his priorities, he has extensively honed his art of finding and developing ideas using the engine.

  In this book, I was particularly impressed by Nikos’s direct, dynamic ideas for Black against the various non-Spanish variations. Black players, and books, often give these lines generic, regurgitated treatment. Nikos, on the other hand, provides the ideal, principled solutions to White’s various options in the Open Games. One example is the 4.Ng5

variation of the Two Knights Defence. The critical 9.h4! has been giving Black a lot of trouble lately, but Nikos analysed it extensively and came up with the almost unknown 9...Qc7!?, which leads to vibrant play for Black – and might even shut this line down as a try for a White advantage.   Against the Spanish, Nikos has wisely avoided any of the ultra-dynamic set-ups which can flutter in and out of fashion. Instead he advocates the classical Breyer, which has been tried and tested by numerous world champions and other leading players. The arising positions are hard to analyse, as engines tend to lose some of their effectiveness in closed structures. I have often used that to my advantage with White, winning many nice games by strangulating my opponents from positions that were supposedly close to ‘0.00’. Nikos, being Nikos, has done much more than present a bunch of lines with dry, superficial engine assessments. Instead he has looked more deeply, utilizing high-level correspondence games (including a few of his own) and sophisticated analytical tools to refine his ideas. The result is a bombproof repertoire which is solid enough never to be refuted, yet complex enough to offer plenty of winning chances.   Parimarjan Negi Stanford, December 2015

Bibliography     Opening Books Alterman: The Alterman Gambit Guide – Black Gambits 2, Quality Chess 2012 Alterman: The Alterman Gambit Guide – White Gambits, Quality Chess 2010 Bologan: Bologan’s Ruy Lopez for Black, New in Chess 2015 Bologan: Bologan’s Weapons in the Open Games, New in Chess 2014 Brunello: Attacking the Spanish, Quality Chess 2009 Burgess, Emms, Gallagher & Nunn: Nunn’s Chess Openings, Cadogan 1999 Emms: Beating 1 e4 e5, Everyman Chess 2010 Emms, Flear & Greet: Dangerous Weapons: 1 e4 e5, Everyman Chess 2008 Emms, Kosten & Cox: Dangerous Weapons: The Ruy Lopez, Everyman Chess 2012 Kaufman: The Kaufman Repertoire, New in Chess 2012 Khalifman: Opening for White According to Anand 1.e4, Volume 2, Chess Stars 2003 Kindermann: The Spanish Exchange Variation: A Fischer Favourite, Olms 2005 Lokander: Opening Repertoire: The Open Games with Black, Everyman Chess 2015 Lysyj & Ovetchkin: The Open Games for Black, Chess Stars 2012 Marin: Beating the Open Games (2nd Edition), Quality Chess 2008 McDonald: Concise Chess Openings, Everyman Chess 2001 McDonald: The Ruy Lopez: Move by Move, Everyman Chess 2011 Mikhalevski: Grandmaster Repertoire 13 – The Open Spanish, Quality Chess 2013 Muller & Voigt: Danish Dynamite, Russell 2002 Obodchuk: The Four Knights Game, New in Chess 2011 Ovetchkin & Soloviov: The Modern Vienna Game, Chess Stars 2014 Panczyk & Ilczuk: Ruy Lopez Exchange, Everyman Chess 2005 Pavlovic: Fighting the Ruy Lopez, Everyman Chess 2009 Pinski: The Four Knights, Everyman Chess 2003 Pinski: The Two Knights Defence, Everyman Chess 2004 Sakaev: The Petroff: An Expert Repertoire for Black, Chess Stars 2011 Shaw: The King’s Gambit, Quality Chess 2013 Watson: Mastering the Chess Openings Volume 1, Gambit 2006   Other Books I consulted Botvinnik: One Hundred Selected Games, Dover 1960 Capablanca: Chess Fundamentals, Everyman Chess 1994 Donaldson & Minev: The Life and Games of Akiba Rubinstein (Volume 2), Russell 2012 Fischer: My 60 Memorable Games, Batsford 2009 Giddins: 50 Essential Chess Lessons, Gambit 2006 Giddins: 50 Ways to Win at Chess, Gambit 2007 Karolyi: Mikhail Tal’s Best Games Volumes 1-2, Quality Chess 2014-2015 Kasparov: My Great Predecessors, Parts 1-5, Everyman Chess 2003-2006 Kasparov: Kasparov on Modern Chess, Parts 1-4, Everyman Chess 2007-2010 Kasparov: Kasparov on Kasparov, Parts 1-3, Everyman Chess 2011-2014 Nimzowitsch: Chess Praxis, Quality Chess 2008 Polgar: Judit Polgar Teaches Chess, Volumes 1-3, Quality Chess 2012-2014

Pritchett: Chess Secrets: Giants of Innovation, Everyman Chess 2012 Smyslov: Endgame Virtuoso, Everyman Chess 1997 Tarrasch: Three Hundred Chess Games, Hays Pub 1999 Terekhin & Bronznik: Techniques of Positional Play, New in Chess 2013 Yusupov: Build Up Your Chess 2, Quality Chess 2013   DVDs Gustafsson: Black Repertoire against 1.e4 Vol.2, ChessBase 2010 Eljanov: The Ruy Lopez Breyer Variation, ChessBase 2015 Shirov: Guide to the Tkachiev Ruy Lopez, ChessBase 2010 Kasimdzhanov: Endgames for Experts, ChessBase 2004   Periodicals New In Chess Magazine New In Chess Yearbook Chess Informant ChessBase Magazine Secrets of Opening Surprises Chess Evolution   Websites chess.com chess24.com chess-structures.com (Mauricio Flores Rios blog) chesshistory.com chesspublishing.com qualitychess.co.uk (blog) skakistiko.com thechessmind.net (Dennis Monokroussos blog)

Introduction     In 2012, the year after my first book (a collaboration with GM Jacob Aagaard on the Tarrasch Defence) was published, I was hired as the openings coach of the Danish national team during the Istanbul Olympiad. I was already a pretty respectable opening analyst back then, but I knew I could improve. I followed the discussions in specialized forums regarding chess software and opening analysis, and noticed that the main contributors were usually strong correspondence players. I therefore made the decision to become one! This made a lot of sense, as I don’t have time for over-the-board competitions, and correspondence chess has enabled me to test my abilities as an analyst against the world’s best.  

My Personal Story with 1.e4 e5 As I was improving my rating and facing stronger and stronger opposition, I realized that in order to be successful as Black, I needed to play sound but complicated openings, otherwise my winning chances would be close to zero. After a lot of deliberation and experimentation, I concluded that 1.e4 e5 would be an ideal choice. The big problem, of course, is what to do against the Spanish, but I was able to find a solution which satisfied me at once.  

The Breyer System The cornerstone of any Black repertoire with 1.e4 e5 has to be the chosen defence against the Spanish. In this book, we will follow the traditional main line to reach the following position: 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3

At this point 9...Nb8! introduces the Breyer System, which sees Black reroute his knight to the flexible d7-square. The Breyer has been used by the world’s elite for decades, so its soundness is not in question. In some theoretical lines there is not a single piece or pawn exchange for more than 25 moves, which tells you something about its complexity. The Breyer is the perfect choice for the ambitious, strategically-oriented player who wants to fight for the win with

Black. This is proven by the list of our Breyer ‘heroes’, which includes Smejkal, Spassky, Portisch, Karpov, and Carlsen. I might add that Adams, Leko and Svidler have all employed the Breyer when they have wanted a break from the Marshall.   The Breyer has been tested for several decades at high levels of play, so there was already an abundance of material. The challenge for me was to work on the numerous variations using the software and tools at my disposal, with an especially deep focus on those variations which are currently regarded as theoretically critical.  

Testing and Refining my Ideas   At the time of writing I am closing in on the master title at correspondence chess, and I have tested the Breyer (as well as several of the other recommendations in this book) against formidable opponents. I have also had the privilege of working on opening ideas with many strong players, including several grandmasters. I am happy that I have shared my files with these players, who have then come back to me with important feedback, enabling me to refine my analysis. I also work with club players and I have presented some of the material in this book in lectures. This experience has helped me to understand which variations may prove difficult for players to fully grasp, and some of my recommendations have been modified accordingly.  

Repertoire Choices This book is intended to be useful for ambitious correspondence players and grandmasters, while also teaching amateurs how to play 1...e5 without overloading them with useless information. It sounds like an impossible task, and readers will have to judge for themselves how close I came to succeeding, but I am satisfied that I gave my best efforts.   In the chapters that follow, you will find antidotes to all of White’s main attempts after 1.e4 e5. All authors write their books in the way they like to read them; I appreciate chess analysis of the highest possible quality, but I also like books with more general instructive and entertainment value as well. This is why I chose a format of complete games, which has the advantage of showing how the play may develop after the opening. Also, a certain characteristic of 1...e5 is that it has been played by all the World Champions and their challengers since the dawn of time, so history has presented us with some incredibly instructive and beautiful games, which deserve to be admired in full. Some of them are classic, famous games, in which modern analysis software can shed new light on the work of previous commentators. In this book you will find new games, old games, correspondence games and even the occasional blitz encounter; if I thought a particular game showed in the best possible way what is happening in a particular variation, I went ahead and included it.

Summary of Recommendations   The King’s Gambit After 1.e4 e5 2.f4 exf4 3.Nf3 I have used 3...g5 with success in my correspondence praxis, and I consider this the ultimate test of White’s opening, but I do not believe it to be the smartest choice for a practical player, especially if you don’t encounter this opening too often. Instead I am attracted by the simplicity of the Schallopp Defence with 3...Nf6! 4.e5 Nh5!, followed by moves like ...d6, Nc6, ...g6 and ...Bg7, in order to challenge the advanced e-pawn. Chapter 1 will provide further details of this, along with the 3.Bc4 King’s Gambit and other gambits and rare lines.   Four Knights 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 reaches an important position. 4.d4 is known as the Scotch Four Knights, which received some attention over the last few years after Kramnik surprised Aronian with an unusual but sound method of handling the white position. Black has to remember a couple of nuances to equalize fully, but the nice thing about studying this line is that it enables you to handle most of White’s rare fourth moves with 4...d5!, intending to play Kramnik’s line with reversed colours! 4.Bb5 is the Spanish Four Knights, and the traditional main line. In that case I suggest following the example of Karpov against Hou Yifan, where the former World Champion gives a model demonstration of Black’s chances.   Scotch Game After 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 I favour 4...Nf6! for a number of reasons. The main line continues 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 and now I advocate 9...Ba6!, intending to follow up with ...Qe6, ...Bb4 and ...d5. I call this the Professional Variation, as it is the main reason why elite players’ enthusiasm for the Scotch has stalled.   Two Knight’s Defence After 1.e4 e5 2.Nf3 Nc6 3.Bc4 I sincerely believe you should avoid the ‘boring’ 3...Bc5 and play 3...Nf6! instead! It is well known that Black will have to say goodbye to one of his precious pawns after 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6, but he gets dynamic compensation in the form of a lead in development, which often transforms to a more stable type of positional compensation such as the better pawn structure. In practical play White’s position often proves more difficult. Even after the currently critical 8.Bd3!, I discovered a nice way to solve Black’s problems while creating some for White.   Spanish Game 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 leads to what has become known as the “Spanish Torture”. For the practical player, it might feel like torture to study how to get here, as White has a multitude of deviations along the way, from 4.Bxc6 to the modern 6.d3, the latter being especially popular among grandmasters at present. I have covered these and other sidelines in Chapters 8-10, proposing solutions which are both effective and as straightforward as possible to learn and play.   When it comes to the main line, as I mentioned earlier, we will study the Breyer System starting with 9...Nb8! 10.d4 Nbd7. Over the next few moves, Black will generally continue with ...Bb7, ...Re8, ...Bf8 and ...g6, intending to fianchetto his bishop on g7. Unlike some other systems in the Closed Spanish, Black will not touch his c-pawn yet, as he wants to be ready to meet the space-gaining d4-d5 with ...c6, challenging White’s pawn centre. Even if White reinforces his pawn chain with b2-b3 and c3-c4, Black will always have counterplay on the queenside, and can sometimes even sacrifice a piece to destroy White’s centre, as shown in Game 64 in Chapter 13.  

The Breyer is such a rich opening that it is difficult to pinpoint one or two variations that are most important to study, as you really need to absorb the whole repertoire to appreciate how Black’s flexible system can be made to work against White’s various ideas. The traditional main line is covered in Chapter 13, and it goes without saying that this should be studied carefully. It is also worth mentioning that my biggest analytical challenge came after 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.b3 g6, as covered in Chapter 12. (By the way, the recent Breyer works by Eljanov and Bologan do not cover this line in the depth it deserves.) I scored a good victory with Black in a correspondence game in this line, but during the game I realized that my opponent could have handled the position in a more challenging way. I thought I had found a good solution for Black, until Sune Berg Hansen, while preparing this line for the recent European Team Championship in Reykjavik, discovered a counter-improvement for White. This caused me a lot of headaches, but I eventually found a way to make Black’s position work, while even putting some pressure back on White. Full details can be found in Chapter 12.  

Enough Talk – Let’s See Some Chess!   At the start of each chapter you will find a set of diagram previews, which feature some of the most important and/or entertaining positions that lie ahead. You can use them as a training tool, by trying to find the best continuation, or as memory markers to help you recall the most important theoretical nuances while preparing for a game.   I hope that you will find this book interesting and entertaining, and I will be more than happy to help you score some brilliant wins after 1.e4 e5!.   Nikos Ntirlis Patras, Greece December 2015

  1.e4 e5   The Portuguese Opening 2.Bb5 page 15   The Nakamura Variation 2.Qh5 page 16   The Ponziani Opening 2.Nf3 Nc6 3.c3 page 18   The Göring Gambit 2.Nf3 Nc6 3.d4 exd4 4.c3 page 20 4...dxc3 page 20 4...d5 Game 1, page 22   The Centre Game 2.d4 exd4 3.Qxd4 (3.c3 page 26) 3...Nc6 4.Qe3 Nf6 5.Nc3 (5.e5? page 31) 5...Bb4 6.Bd2 0-0 7.0-00 Re8 page 27 8.Qg3 Game 2, page 28 8.Bc4 Game 3, page 31   The King’s Gambit 2.f4 exf4 page 33 A) 3.Bc4 Qh4†! 4.Kf1 d6! page 34 A1) 5.Nc3 page 34 A2) 5.Nf3 page 36 B) 3.Nf3 Nf6! page 38 B1) 4.Nc3 page 38

B2) 4.e5 page 40 & Game 4, page 42  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How does Black secure enough

compensation? (page 18)  

How should Black secure the f4-pawn? (page 34)

Can you suggest a good move for Black? (page 29)

Find a beautiful resource for Black. (page 30)  

How should Black continue? (page 36)  

How does Black demonstrate an edge? (page 39)  

How does Black obtain an advantage? (page 32)  

What is the trap that Black must avoid? (page 41)

1.e4 e5 After these opening moves it is generally accepted that the most critical, and frankly the best, move is 2.Nf3. White is not obliged to follow this advice all the time though, and has the right to throw us a surprise now and then. The real challenge for Black is not to discover a good way to play against those rarities, but to find a line that will be simple enough to be remembered any time it is needed for years to come. The Ponziani Opening and the Göring Gambit are two systems that may leave White with some difficulties if Black remembers the correct theoretical continuation, which is not a difficult thing to do. The remaining systems are not particularly dangerous unless you fall into a trap. I was pleased to discover a simple and effective way to meet the Centre Game, especially because my editor Andrew Greet produced a serious repertoire for the White side of this opening in Dangerous Weapons: 1 e4 e5 and thus made it quite popular again, giving us extra opportunities for easy points! (I really hope that Andrew won’t erase this comment and thus gives me a chance to tease him!) John Shaw’s landmark The King’s Gambit also presented some challenges, but I found a pleasant solution for Black which does not require too much work.

The Portuguese Opening 1.e4 e5 2.Bb5?! With this move White already abandons hope of obtaining any sort of opening advantage.   2...Nf6 3.d4 3.Nc3 c6 4.Ba4 loses a pawn after 4...b5 5.Bb3 b4 6.Na4 Nxe4 as in Damaso – Tisdall, Manila (ol) 1992. With ...d5 coming next, White has zero compensation. 3...c6 4.dxe5 4.Ba4? Nxe4 5.dxe5 was met by 5...Nxf2!?µ in Lima – Cordovil, Portugal 1992, but one can argue that 5...Bc5!N 6.Nh3 d5! is even stronger!   4...Nxe4 5.Bd3 5.Qe2 Qa5† 6.c3 Nxf2 was another embarrassment for White in Camejo Almeida – Izeta Txabarri, Andorra 1987, as White lost a pawn without getting any real compensation for it.   5.Be2 Bc5 also favours Black, as White will have to make a concession in order to defend f2.   5...Nc5 6.Nf3 d5 Black has already achieved full equality, if not a slight advantage. We will follow a game of Zoltan Almasi, which I had the privilege of watching live in the playing hall at the Tromso Olympiad.

7.0-0 7.exd6 Bxd6 gives Black an easy game; White will have to either give up the bishop pair or waste time moving the d3-bishop, handing Black a lead in development.   7...Be7 8.h3 0-0 9.Nc3 Nbd7 10.Ne2 f6! Black already has a sizeable advantage and he went on to win in Rocha – Almasi, Tromso (ol) 2014.

The Nakamura Variation 1.e4 e5 2.Qh5?! The relevant Wikipedia page calls this move the Wayward Queen Attack, but I think it makes sense to name it after Nakamura, who famously played it against Sasikiran and got an advantage out of the opening, although the Indian GM eventually prevailed. The American wrote a short article about it in Secrets of Opening Surprises (SOS) 7, in which he credits Bernard Parham from Indiana as the first player to show how White’s second move could be used as a serious system as opposed to a beginner’s attempt to land a quick mate on f7.

2...Nc6 2...Nf6?! 3.Qxe5† Be7 is an attempt to sacrifice a pawn for the initiative, but there is no need for Black to resort to a speculative gambit. According to Nakamura, 4.Qf4! leads to a better position for White.   Although it has no relevance to our study of the opening, somehow I cannot resist pointing out that 2...Ke7?? 3.Qxe5# would be an amusing helpmate! 3.Bc4 Setting up Scholar’s Mate, or Napoleon’s Mate (as it is called in Greece), or Shepherd’s/Shoemaker’s Mate, as it is known in other countries.   3...g6! 3...Qe7 is also fine, and was Kasparov’s choice in a consultation game against the actor Woody Harrelson! White managed to draw, as he had a bit of help from Yasser Seirawan.   4.Qf3 Nf6 4...f5!? was Sasikiran’s suggestion after his famous game against Nakamura. 5.Ne2 (5.exf5? Nd4! 6.Qg3 Qf6 7.Bb3 Ne7 is given by Nakamura as a line White should avoid) 5...Nf6 6.d3 In Rios – Whatley, Dallas 2001, Black played the calm 6...d6, but I would prefer the following move which is not mentioned by Nakamura:

6...f4!N I claim a slightly better game for Black due to his extra space, although a complicated struggle lies ahead, with the centre closed and all the pieces still on the board. This would be my choice in a correspondence game, and you can certainly play this way if it appeals to you. However, I will keep the text move as the main line as it provides a relatively easy and convenient route to at least an equal game.  

5.Ne2 5.Qb3? can be refuted by 5...Nd4! (5...Qe7 is perfectly fine for Black, but sometimes the best defence is a counterattack!) 6.Bxf7† Ke7 7.Qc4 b5! and White loses his bishop.   5...Bg7 6.d3 In his SOS article, Nakamura gives this as a more accurate move order than 6.Nbc3 d6 7.d3 Bg4 8.Qg3, as played in Nakamura – Sasikiran, Copenhagen 2005. Personally I am not convinced that the early ...Bg4 helps Black, so I would

suggest either 7...0-0 or the immediate 7...Na5, with similar play to the main line below.

6...d6 7.h3 Nakamura suggests this to rule out 7.0-0 Bg4, but I would prefer 7...Na5 anyway. Compared with our main line, White can save a bit of time by omitting h2-h3, but Black’s position remains comfortable and it is not clear if White has anything better than playing h2-h3 anyway within a few moves.   7...Na5 8.Nbc3 After 8.Bb5†?! c6 9.Ba4 b5 10.Bb3 0-0³ Black will exchange on b3, then play ...Nd7 and ...f5.   8...0-0N 9.0-0 Although I found no games from this position in my database, Nakamura reaches it in his notes in the SOS article.   9...Nxc4 10.dxc4 The position resembles a Vienna game where Black has gained some time to fianchetto his bishop. White may argue that the g7-bishop could have been better developed elsewhere, but I found a good way for Black to make sense of it.

10...Nd7!³ Black intends to fianchetto his other bishop on b7 and play ...f5.

The Ponziani Opening 1.e4 e5 2.Nf3 Nc6 3.c3 The Ponziani is a rare opening at all levels; neither I nor any of my numerous students have ever faced it. Still, Magnus Carlsen gave it a try once, as has my friend and colleague Jacob Aagaard.   3...Nf6 3...d5 is the other main line, but after 4.Qa4 I don’t think White is worse. Since there is a lot of theory here, I decided to skip it.   4.d4 Nxe4 5.d5 Bc5! Black sacrifices a piece for two pawns and excellent attacking chances against the exposed white king. Results have been mixed, but I believe this is mainly because Black has often been too hasty to reclaim some material.

6.dxc6 Bxf2† 7.Ke2 Bb6! Most games have continued with 7...bxc6, but the text move is more troublesome for White.   8.Qd5 8.Be3? is the only other move to have been tried, but it is not much good. After 8...bxc6 Black plans ...d5 and ...Ba6† with excellent play. The one practical example continued 9.Bxb6 axb6 10.Nbd2 Nf6 11.Kf2 Qe7 12.b4 e4–+ and Black was already winning back the piece in Krabbe – De Rooi, Amsterdam 1964.   The text move produces a double attack on the e4-knight and the b7-pawn, but Black should not be worried.   8...Nf2! 9.Rg1 9.Bg5? f6 10.Nxe5 Qe7 was winning for Black in Soderberg – Engman, Sweden 2010.   9.Qxe5† Kf8 10.Rg1 was seen in Bosch – Lombaers, Wijchen 2014, and now 10...Ng4!N followed by ...dxc6, protecting the knight and attacking g1, is excellent for Black.  

9.cxb7 Bxb7 10.Qxb7 Nxh1 11.Be3 Bxe3 12.Kxe3 0-0 followed by ...d5 was also excellent for Black in Csapo – Rachela, Hungary 2008. The white king is ridiculously exposed.

9...0-0! 9...dxc6 is playable, but being materialistic after you have played 5...Bc5 is not a good strategy.   10.cxb7 Bxb7 11.Qxb7 Qf6 Even though Black is temporarily two pieces down, White is still the one who must fight for survival! Even if he does everything correctly, the worst outcome for Black is an endgame with rook and one or two (passed) pawns versus two minor pieces. My database contains mostly computer games from this position, and it is hardly surprising that no one has had the courage to test White’s position over the board.   12.Qd5 12.Qa6 has done well for White in some computer games, but the untested 12...e4!N gives White serious problems, a sample line being: 13.Nd4 c5 14.Nc2 d5µ   12.Na3N is probably White’s best, although none of the engine games have continued this way. The main line of my analysis runs: 12...e4 13.Nc4 Rab8 14.Qd5 exf3† 15.gxf3 Rfe8† 16.Kd2 Ne4† 17.fxe4 Bxg1

Black will almost certainly pick up another pawn, most likely on h2. Engines tend to slightly prefer Black’s position after something like 18.Bd3 c6 19.Qf5 Qxf5 20.exf5 Bxh2 but I would estimate the chances as being close to equal – certainly this is the nearest White can get to equality after 5...Bc5!.  

12...c6 13.Qd2 e4 14.Nd4 e3 15.Qc2 Rae8 16.Nf3 Re6 17.a4 a5 18.Na3 d5 19.Rb1 So far we have been following the weird encounter sniper – gaga, Internet 2014, but now I suggest another path for Black:

19...Rfe8N My analysis indicates that Black’s chances are better, although of course the position remains complicated.   20.b4 After 20.g3 c5 21.Rg2 c4 22.Rg1 g5 23.Bg2 g4 24.Nd4 Nd3³ Black is dominating, despite being two pieces down!   20.g4 Ne4 also favours Black according to the engines.   20...Ne4 21.Bb2 Nd2 22.Re1 Qf4³ Apart from the obvious ...axb4, Black’s ideas also include ...g5-g4 or ...Qg4 threatening ...Nxf3. White will have to give back material in order to survive.  

The Göring Gambit 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.c3 4.Nxd4 is the Scotch Game (Chapter 5), while 4.Bc4 Nf6 can be found in Chapter 6 under the 3.Bc4 Nf6 4.d4 exd4 move order.

The text move is rarely seen at a high level nowadays, and Black has more than one good way to meet it.   4...dxc3 4...d5 is a solid alternative which will be covered in Game 1 below.   I would also like to show how you can get a good game by accepting the gambit, should you prefer to do so. I wouldn’t capture this pawn only to give my opponent a comfortable position with the initiative – I’d take it in order to give it back to get a positional advantage!   5.Nxc3 5.Bc4 can be met by 5...cxb2 6.Bxb2 d6! when Black will neutralize the c4-bishop with ...Be6. 7.0-0 (7.Qb3!? is a clever attempt to switch up the move order, but 7...Be6! 8.Bxe6 fxe6 9.Qxe6† Qe7 10.Qf5 Qf7 works out well for Black) 7...Be6 8.Bxe6 fxe6 9.Qb3 Qd7 10.Ng5

10...0-0-0! (This is cleaner than the often-recommended 10...Nd8, when 11.f4 leads to a messy position) 11.Nxe6 Re8 12.Nxf8 Rxf8 13.Nc3 Nf6 14.Rab1 Qf7 White had about enough compensation for equality, but nothing more in Herzog – Farkas, email 2006.

5...Bb4 6.Bc4 Bxc3†! This is a useful way to reduce White’s options. Other moves allow him to play Qb3, after which he can seriously consider recapturing on c3 with the queen.   7.bxc3 d6 8.Ng5 8.0-0 Nf6 9.Ba3 Bg4! is known to be great for Black (the last move is designed to avoid 9...0-0? 10.e5!‚ as Alekhine once played with White). 10.Bb5 (10.Qb3 is met by 10...Bxf3! 11.Bxf7† Kf8 12.gxf3 Ne5µ) 10...0-0 11.Bxc6 bxc6 12.e5 Nd5µ Penrose – Smyslov, Munich (ol) 1958.  

8.Qb3 Qe7 9.0-0 Nf6 10.e5 looks scary, but Black is known to be doing fine after:

10...Nxe5 11.Nxe5 dxe5 12.Ba3 c5 13.Bb5† Kf8 14.f4 Be6! The game continues, but White will struggle to justify the two-pawn investment.

8...Ne5 9.Bb3 h6 10.f4 I found this to be the most critical variation after we accept the Göring Gambit. Best play continues:   10...Qe7 11.fxe5 hxg5 12.Qd5 dxe5 13.Ba4†! This disruptive check is an important resource.   13.Bxg5 Nf6 simply leaves White a pawn down for questionable compensation.

13...Bd7 14.Qxb7 Rd8 15.Qxa7 Bxa4 16.Qxa4† Qd7 17.Qxd7† Rxd7 18.Bxg5 f6 19.Be3 Rh4 Black is about to restore material equality, and his compact structure ensures him of at least equal chances. We will follow a correspondence game for a few more moves.

20.a4 Rxe4 21.Kf2 Rd6 22.Rhb1 Rc4 23.a5 Ra6 24.Rb8† Kf7 25.Rb7 Ne7 Amazingly, this was the knight’s first move of the game!   26.Ra7 Rcc6

The players agreed a draw ten moves later in Berza – Silin, corr. 2012. In a practical game, Black could certain keep pushing for something more.   Having shown that taking the pawn on c3 is a viable option, I will now present my main recommendation of declining the gambit with 4...d5.     GAME 1 Dimitrios Mastrovasilis – Zurab Sturua Istanbul 2003   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.c3 d5 5.exd5 Qxd5 6.cxd4 Bg4 This is the recommended approach for Black in most of the main 1.e4 e5 sources. There is no mystery behind this: Black’s game is easy and White has to be careful not to end up worse.  

7.Be2 Bb4† 8.Nc3 Bxf3 9.Bxf3 Qc4 Capablanca introduced this move against Marshall in 1926, and it remains as problematic as ever for White to achieve something against it.   10.Qb3 A different type of endgame occurs after: 10.Bxc6† bxc6 11.Qe2† Qxe2† 12.Kxe2

Quite a lot of players have attempted to play for the draw as White in this line, but with generally poor results. Black has easy play, and small inaccuracies tend to harm White more than Black. 12...Ne7 13.Be3 0-0-0 13...Kd7!? is another sound move given by Marin, but I am happy to recommend the main line. 14.Kd3!

Improving the king seems best. Another continuation is: 14.Rhd1 Rhe8 15.Rac1 Nf5 (15...Ba5!? is a logical alternative, intending to put the bishop on b6 before playing ...Nf5) 16.Kf3 Rd6 (16...h5 was also successful in Pirrot – Sturua, Biel 1996, but there is actually no need to prevent White’s next) 17.g4?! White saw ...Rf6 coming and he panicked. 17...Nxe3 18.fxe3 Rh6 White lost a pawn and eventually the game in Heppell – Haslinger, Isle of Man 2009.

14...Nf5! Bologan believes Black should play 14...c5 and liquidate into a drawn rook endgame a pawn down, but Black is not worse after the text move, and can keep some winning chances. 15.Kc4 Ba5 16.g4 Otherwise Black will improve his position with ...h5 and ...Rhe8, perhaps followed by doubling on the e-file. 16...Nd6† 17.Kd3 Bb6

18.a4 a5 19.Rad1 f5 20.h3 Kb7 21.Kc2 Rhf8

Black was at least equal in Dos Santos – Salzmann, email 2007.   10...Qxb3 11.axb3 11.Bxc6†?! bxc6 12.axb3 is hardly an improvement: 12...Ne7 (12...a5 also gives Black a small edge) 13.0-0 Kd7!? Black has done extremely well from this position in practice. One example continued: 14.Rd1 Rhb8 15.Rd3 Bxc3 16.Rxc3 Rb4 17.Rd3 Rab8³ Firnhabe – Schlemmer, corr. 1998.

11...Nge7 12.0-0 a6 Black can consider another move order: 12...0-0-0!? This gives White a couple of extra options, but neither of them is any good. This way Black also gets an extra option, but I doubt that it is any better than our main line. 13.Be3! The dodgy alternatives are: 13.d5?! Nd4 14.Rxa7 Nxf3† 15.gxf3 Bxc3 16.bxc3 Nxd5³ Maric – Veroci, Belgrade 1989. 13.Bxc6?! Nxc6 14.d5 Bxc3 15.dxc6 Bd4³ Van den Berg – L’Ami, Delft 2015.  

13...a5!? 13...a6 transposes to our main line. The text move was recommended by Sakaev. 14.Rfd1 Rhe8 14...Kb8 is the older move, when 15.Rac1 Nf5 16.Nb5 Rd7! 17.Bxc6 bxc6 18.Rxc6 Kb7 19.Rcc1 Re8 was excellent for Black in Mastrovasilis – Skembris, Agios Nikolaos 2000. However, White has two improvements: 17.Bf4! is at least okay for him, while earlier 15.Na2 Bd6 16.Nc3 leaves Black with nothing better than 16...Bb4 repeating. Sakaev ends his analysis after the text move. I guess his idea is to be ready for White’s queenside activity after: 15.Rac1 Kb8 16.Nb5 Nd5 Black is solid and well coordinated.   Overall I prefer the feel of the ...a6 set-up which doesn’t give away the b5-square, although I cannot prove analytically that it is superior to the ...a5 concept, which might have the advantage of being simpler. However, even if you intend to play the ...a6 line, you might want to use the 12...0-0-0!? move order anyway, to keep your opponent guessing while giving White an opportunity to choose one of the inferior options on move 13.  

13.Be3 0-0-0 14.Rfd1 Kb8 15.g3 Rhe8³ Black has slightly better chances. Ideally he would like to follow the well-known principles of fighting against an IQP, by trading a few minor pieces while retaining his structural superiority. Another strategy is to safely blockade the d-pawn (either on d4 or on d5) and start expanding on the kingside.   16.d5 White should advance this pawn while he has the chance, as it frees a bit of space for his pieces while making the pawn harder to attack.   The following is a nice example where White left the pawn on d4: 16.Bg2 Bd6 17.Ne4 Nf5 18.Rac1 Be7 19.Rc4 Na5 20.Rc3 Bb4 21.Rcd3 Nc6 If White does nothing, Black will slowly improve his position, for instance by playing ...h6 and doubling on the d-file.

22.Nc5 A logical try, but it gives Black a chance to simplify to a favourable endgame. 22...Bxc5 23.dxc5 Rxd3 24.Rxd3 Ne5 25.Rc3 Ng4 White cannot keep his pair of bishops. 26.Bd2 Rd8 27.Be4 Rxd2 28.Bxf5 Nxf2 29.Rf3 Nd1 30.Bxh7 f6 31.Rd3 Rxd3 32.Bxd3 Nxb2 33.Be2 a5 Black went on to convert his advantage in D. Mastrovasilis – A. Mastrovasilis, Porto Carras 2005.

16...Ne5 17.Bg2 A slightly better move is 17.Be4 with the idea that, after 17...f5 18.Bg2, Black cannot use the f5-square for his e7knight. However, there are other ways to improve this piece: 18...Ng8! is my preference, although 18...Nc8!? also has its merits.   17...Nf5 18.Bf4 A couple of months after this game, Dimitrios tried to improve with 18.Bg5. Play continued 18...f6 19.Bc1 Bc5 20.Na4 Ba7 21.b4 Nd4 22.Nc5 when Black lost the thread, dancing around with his knights for seven more moves without achieving anything concrete in D. Mastrovasilis – Vouldis, Athens 2003. Instead he should have played:

22...Ndf3†!N 23.Kh1 Ng4 24.Bxf3 Nxf2† 25.Kg2 Nxd1 26.Bxd1 Bxc5 27.bxc5 Rxd5³ Black will take on c5 next, leaving himself with rook and three extra pawns against two bishops.

18...Bc5 19.Ra4 Bb6 20.Bh3 Nd6 Having regrouped his minor pieces, Black goes on to improve his position beautifully.   21.Kg2 h6 22.Bc1 f5 23.Kf1 Re7 24.Rh4 Rde8 25.Rh5 g5! This is a picture to remember. Black has a perfect blockade on d6 and his expansion is ready to bear fruit, as the next note demonstrates.

26.f4 26.Rxh6 g4 27.Bg2 Nd3µ was Black’s idea.   26...Nd3 27.Bd2 g4 28.Bg2 Be3 29.Ke2 Nxb2 White resigned – the final position is a rare picture of domination. 0–1

The Centre Game 1.e4 e5 2.d4 exd4

3.Qxd4 Of course we should briefly consider the Danish Gambit:   3.c3 dxc3!? 3...d5 will most probably transpose, after 4.exd5 Qxd5 5.cxd4 Nc6 6.Nf3, to my main recommendation against the Göring Gambit. 4.Bc4 4.Nxc3 Nc6 5.Nf3 transposes to the Göring Gambit as covered earlier.

4...cxb2 5.Bxb2 d5 6.Bxd5 6.exd5?! avoids simplifications but blocks the light-squared bishop. 6...Nf6 7.Nf3 Bd6 8.0-0 0-0 9.Nc3 Bg4 Perfectly logical play by Black.

10.Qd4? (10.h3 Bxf3 11.Qxf3 Nbd7 leaves White a pawn down for not much compensation; 10.Nb5 Nbd7! 11.Nxd6 cxd6 also favours Black) Surprisingly, in Opocensky – Reti, Baden 1914, both players overlooked 10...Bxf3 11.gxf3 Nc6!µ when Black is close to winning, as he will follow up with ...Ne5 and ...Nh5, with a strong attack on top of the extra pawn. 6...Nf6

7.Bxf7† 7.Nc3 is met by 7...Nxd5! 8.Nxd5 Nd7 followed by ...c6, and if the knight moves then ...Nc5 will give Black an excellent position. White can sacrifice a piece with 9.Nf3 c6 10.0-0!? but Black comes out on top after 10...cxd5 11.exd5 Be7 12.Bxg7 Rg8 13.Re1 Nb6 14.d6 Be6 15.Be5 Bf6µ as in Guidoni – Laurent, corr. 2009. 7...Kxf7 8.Qxd8 Bb4† 9.Qd2 Bxd2† 10.Nxd2 Re8 11.Ngf3 Nc6=

Black will complete development with ...Bg4 and ...Rad8. Extensive practice over the decades (as well as my own analysis) has shown that Black’s queenside majority is at least as strong as White’s kingside one – but not stronger, unfortunately.   The present position is the starting point for the Centre Game.

3...Nc6 4.Qe3 Nf6 5.Nc3 Bb4! Black can develop in a few different ways, but I like this simple and straightforward plan.   6.Bd2 0-0 7.0-0-0 Re8 8.Qg3 This is traditionally the main line, although 8.Bc4 might be objectively slightly superior. Black should play the simple 8...d6 intending ...Be6, but if White goes for the sensible-looking 9.f3 then Black can already get more ambitious with 9...Na5!. Game 2 will show this in more detail, along with a few of White’s earlier deviations.

8...Nxe4! 8...Rxe4!? is an attractive alternative, but I think the text move is more convincing if followed up correctly.   9.Nxe4 Rxe4

Now White must avoid the bishop exchange if he hopes to get any sort of initiative.   10.Bf4 10.c3 Be7 is also excellent for Black, as shown in the notes to the following game.   10...Qf6 11.Nh3 d6 12.Bd3

12...Nd4! In his video lectures for Chess24, GM Vallejo Pons goes as far as to call this the refutation of White’s system. After analysing this line I cannot help but agree with him! Further details can be found below.   GAME 2 Ian Nepomniachtchi – Francisco Vallejo Pons Moscow 2007   1.e4 e5 2.d4 exd4 3.Qxd4 Nc6 4.Qe3 Nf6 5.Nc3 Bb4 6.Bd2 0-0 7.0-0-0 Re8 8.Qg3 Nxe4 9.Nxe4 Rxe4

10.Bf4 After 10.c3 Black has a few decent moves, but I like 10...Be7 the most. The idea is to harass the white queen with ...Bh4 and force it to go to f3, where it will deprive the g1-knight of its more natural square. White has several choices but none is too impressive.   a) 11.Bd3 allows Black to demonstrate his ideas: 11...Bh4 12.Qf3 Re8! Threatening ...Ne5. 13.Qh5 g6 14.Qh6 Ne5 This was Nadj Hedjesi – Pap, Valjevo 2011. Black was ready for ...d6 and ...Bf6, and White had hardly any compensation for the pawn deficit.   b) 11.f4 stops any ...Ne5 ideas but weakens the light squares, enabling Black to counter with an attractive idea: 11...d5 12.Bd3

12...Bf5! An excellent exchange sacrifice, which has only been played in one game, Rivas – Kokkala, email 2012. Black

can continue calmly with ...Qd7 and ...Bf6, as taking the e4-rook almost equates to suicide for White.   c) 11.Nf3 was recommended in Dangerous Weapons: 1.e4 e5 by Andrew Greet (and not by Emms, as Bologan writes in his book!). This move stops ...Bh4, but allows: 11...d6! 12.h3

12...Ne5! This leads to simplifications which will underline Black’s material advantage. Surprisingly, this natural continuation had never been played when the Dangerous Weapons book was published, but it has since been seen in several games, and has achieved a heavy score for Black. 13.Nxe5 Rxe5 14.f4 Ra5 Black’s only potential problem is the awkward position of his rook, but White has no way to take advantage of it. An example from correspondence praxis continued: 15.a3 Bd7 16.Bd3 Ba4

17.Bc2 (17.Rde1 loses an exchange to 17...Bh4; if 17.Rdf1 d5 the piece sacrifice ...Bxa3 is in the air, while Black’s other ideas include ...c5 followed by some combination of ...c4, ...Bf6 and ...d4.) 17...Bxc2 18.Kxc2 d5 19.Qd3 Qd7 Black was dominating in De Blasio – Diaz, email 2009.

10...Qf6 11.Nh3 11.Bxc7? d6 traps the bishop.   After 11.Bg5?! Qe6 12.Bd3 the simplifying 12...Re1!? kept Black on track to convert his pawn advantage in Hernandez Munoz – Galan Martin, Linares 2007, although 12...Qxa2N 13.c3 Bd6 might have got the job done more quickly.   11...d6 12.Bd3 Nd4! This is Vallejo’s refutation as described above.   13.Kb1 13.Be3 was seen in Mieses – Capablanca, Berlin 1913, where the great Cuban played one of the few blunders of his chess career: 13...Bg4? allowing 14.Ng5! – although he still managed to come back and win from a losing position the exchange down. A much better move is 13...Nf5!N as given by both Kaufman and Vallejo, and after 14.Qf3 Re8 15.Bg5 Qe5 White has no real compensation for the pawn.   13...Ne2 14.Qf3

14...Ba3!! 15.bxa3 No better is: 15.Bc1 Qxf3 16.gxf3 Re7 17.Ng5 Nxc1 18.Rxc1 Bc5 19.Rce1 Rxe1† 20.Rxe1 Bd7 21.Re7

21...Re8! 22.Bxh7† Kh8 23.Rxe8† Bxe8 24.Be4 c6 25.Bd3 d5 Black had an easily winning endgame in Freund – Dothan, corr. 2014.   15...Nc3† 16.Kc1 Nxa2† 17.Kb1 White has no choice, as 17.Kd2 Qc3 is mate.   17...Nc3† 18.Kc1 Bxh3  

19.Bxe4 19.Bd2 Ne2† 20.Kb1 (20.Bxe2 Qa1#) 20...Qxf3 21.gxf3 Re6 is another winning endgame for Black.   19...Bxg2! 20.Qxg2 Qxf4† 21.Kb2 Nxe4 21...Qxe4 22.Qxe4 Nxe4 is also good enough, as Black has three pawns for the exchange.   22.Rd4 Qf6 23.Qxe4 c5 24.Re1 h5 25.Qf4 Qxd4† 26.Qxd4 cxd4–+ The activity of White’s rook is not enough to make up for a three-pawn deficit.   27.Re7 b6 28.a4 g5 29.Kb3 Kg7 30.Rc7 Re8 31.Rxa7 Re2 32.Rb7 Rxf2 33.Rxb6 Rxh2 34.a5 Re2 35.a6 Re8 36.a7 Ra8 0–1 This game and the accompanying analysis have effectively put the 8.Qg3 pawn sacrifice out of business. The next game will deal with the various ways in which White may deviate from it.     GAME 3 Robert Rabiega – Andrei Maksimenko Berlin 2009   1.e4 e5 2.d4 exd4 3.Qxd4 Nc6 4.Qe3 4.Qa4 is occasionally seen, but 4...Bc5 5.Nc3 Nge7 is fine for Black. An entertaining example continued: 6.Bg5 0-0 7.Nd5 Kh8!? (7...d6³ was more to the point, with ...Bd7 coming next; the text move sets a nice trap though!) 8.b4? White probably expected the bishop to retreat to d6 (which is actually not bad at all) as going to b6 allows Nxb6 when Black must spoil his structure. Black had another idea though:

8...Nxd5!! 9.Bxd8 Bxb4† 10.c3 Nxc3 11.Qb3 Rxd8–+ Black had an incredibly strong initiative in Kramer – Firmenich, corr. 1960, not to mention enough material compensation for the queen already.   4...Nf6 5.Nc3 An inferior alternative is: 5.e5? Ng4 6.Qe4 Black has a few good ways to deal with this line, my preference being: 6...h5! 6...d5 7.exd6† Be6 is the old main line where Black has excellent compensation, although things are getting hot after 8.Ba6 Qxd6 9.Bxb7 Qb4† and so on. 6...Ngxe5!? was shown to be perfectly playable by Gustafsson. The idea is 7.f4 d5 8.Qe3 (8.Qe2 Bg4 is excellent for Black) 8...d4 when Black does well in the complications, although some memorization is needed. The text move is simple and strong: Black just takes a moment to defend the knight in order to prepare ...Bc5.

7.Nc3 7.h3 Ngxe5! is now perfectly safe, as 8.f4 Qh4† 9.Kd1 Qf2!µ works out well for Black. 7...Bc5 8.Nh3 Ngxe5 If you are in any doubts about the complications below, you can safely play 8...0-0 with a good game. 9.f4 d5! 10.Qxd5

10...Qe7! 11.fxe5 Bxh3 12.Ne4 12.gxh3?? Qh4† 13.Kd1 Rd8 wins the queen. 12...Bb4† 13.c3 Be6µ Black had a dangerous initiative in Sutorikhin – Pirrot, Bad Woerishofen 2005.   5...Bb4! 6.Bd2 0-0 7.0-0-0 Re8 8.Bc4 With the pawn sacrifice not working, nowadays the text move is the only option which allows White to survive at

correspondence level. White develops a piece and prevents the threatened ...d5.

8...d6! Please don’t be tempted to try 8...Bxc3?! 9.Bxc3 Nxe4, as after 10.Qf4 White’s compensation is rather dangerous.   “[...] keeping the attack on e4 and preparing to neutralize White’s bishop with ...Be6” is how Neil MacDonald sums up Black’s last move in Concise Chess Openings and I would agree that this is all you need to know to play this line over the board. Still, since I have researched this variation deeply for my correspondence praxis, I thought it would be nice to show you how Black can get the most out of the position.   9.f3 9.Nf3 Be6 10.Bxe6 Rxe6 11.Ng5 Re8 12.f4 h6 13.h4 Qd7! By covering the h3-square, Black prepares to grab the material on offer. (The immediate 13...Bxc3?! 14.Bxc3 hxg5? 15.hxg5 Nxe4 16.Qh3 wins for White) 14.Qf3

14...Qg4! 15.Nh3 Nd4 16.Qd3 Bxc3 17.bxc3 Ne2† 18.Kb2 Nxe4 19.Rhe1 Nc5 20.Qd5 Na4† 21.Kb3 Nb6 Black easily converted his huge advantage in Winawer – Janowski, Vienna 1896.   9...Na5! 9...Be6 10.Bxe6 Rxe6 is only equal.   10.Bd3 This is a better try than 10.Bb3, which was met convincingly by 10...Nxb3† 11.axb3 a5! 12.Qf2 Bd7 13.Nge2 a4 14.bxa4 Bxa4 15.Nb1 Bc5 16.Be3 b6µ in Romero Holmes – Karpov, Madrid 1992.

10...d5! 11.Qf2 11.Qg5 h6 12.Qh4 d4 13.Nce2 Bxd2† 14.Rxd2 c5 was miserable for White in Von Feilitzsch – Keres, corr. 1932-33.   11...d4 12.Nb1 Bxd2† 13.Nxd2 c5 Just like in the above Keres game, White is already in deep trouble. Rabiega tries to slow down the pawn assault on the queenside but to no avail.   14.Nb3 Nxb3† 15.axb3

15...a5! 16.Bb5 Bd7 17.Qf1 Qb6 18.Ba4 Qb4! Black crashes through.   19.c3 dxc3 20.Bxd7 Qxb3 21.Qe2 Red8 0–1

The King’s Gambit 1.e4 e5 2.f4 The fearsome King’s Gambit! Scottish Grandmaster John Shaw wrote the modern bible of this opening a few years ago, and this publication marked a new phase in the modern practice of this ancient opening. Many fascinating new games have taken place, with various articles full of GM analysis being published. Shaw’s verdict was that the King’s Gambit is not refuted and, in the hands of the right player, can produce interesting positions rich in possibilities. As much as I’d love to provide you with a refutation of 2.f4 (and you can be certain that I did my best to find it!) I don’t think this is the right attitude for a practical player. The King’s Gambit is an amazingly complex opening, with computers still offering little help to someone trying to make sense of what is going on. Even if you study it carefully with Black and manage to understand all the opening nuances, you can still end up in the kind of position where a single mistake might send you into a deep hole from which there is no escape.   My approach in this book is different. I suggest that against Shaw’s 3.Nf3 we should take a slice of humble pie and play for equality, with the possibility of more if White plays inaccurately. Against 3.Bc4 I will offer quite a novel approach, which is essentially a refinement of a line that is slowly becoming popular among GMs and high-level correspondence players. In both cases, the arising positions are simple enough to understand and play without much precise opening knowledge.   2...exf4 We start with this natural move. White’s main options are A) 3.Bc4, the Bishop’s Gambit, and B) 3.Nf3, the Knight’s Gambit, which is the big main line.   Instead of that White can play some weird gambits which remind us of the way Steinitz approached the King’s Gambit in the 1860s and 1870s, when he had no qualms about his king strolling around the centre of the board – a strategy that Kasparov called “nonsense”.   3.Nc3?! Qh4† 4.Ke2 Qe7! The young Etienne Bacrot made people aware of the strength of this move.

5.d4 Nf6 6.e5 (Korchnoi’s suggestion of 6.Kf2 can be met by 6...d5!, attacking e4 while threatening a deadly check on

g4; 6.Qd3 is another try, but 6...d5! 7.e5 c5 is a strong reply) 6...d6! 7.Nf3 dxe5 8.dxe5 g5!µ White was facing the highly unpleasant prospect of ...g4 in Bauer – Bacrot, Enghien-les-Bains 1999.   3.d4?! Qh4† 4.Ke2 Qe7! should transpose, as White can hardly have anything better than 5.Nc3.   A) 3.Bc4

With this move White tells us he is not bothered by a check on h4, because he can move his king to a not-so-bad square and then gain a tempo with Nf3. OK, no problem then...   3...Qh4†! 4.Kf1 d6! Black intends to justify his early queen move by quickly developing his queenside pieces and castling there. A nice feature of this plan is that ...Be6 will gain a tempo. White may proceed with A1) 5.Nc3 or A2) 5.Nf3.   5.d4 Be6 6.Qd3 Nd7 7.Nc3 c6 leaves White with nothing better than 8.Nf3, when 8...Qh6 reaches variation A2 below.   A1) 5.Nc3   This was recently used by Ivanchuk to defeat Karjakin. White’s idea is to follow up with d2-d4 and only then play Nf3, in order to meet ...Qh6 with the strong Ne2.   5...Be6 6.Bb3 Nd7 7.d4

7...Ngf6!N I love this method of countering White’s scheme: the idea is simply to play ...Nh5, defending f4 in case White goes for the Ne2 plan. If White does not put his knight on e2, Black will play ...Nh5 anyway, threatening a big check on g3.   7...0-0-0 8.Nf3 Qh6?! 9.Ne2! sees White execute his plan, but it is worth mentioning that 8...Qf6! is better, when 9.Nd5 Bxd5 10.Bxd5 g5 11.c3 Ne7 12.Qb3 Nxd5 13.Qxd5 h6µ was the continuation of Exler – Thurner, Vienna 2006. White’s play can obviously be improved, but Black’s position at move 8 is fine; I just happen to like the main line even more.   Karjakin’s 7...g5!? is the option I like the least, but even this is playable if followed up correctly: 8.Nf3 Qh5!? Black keeps the option of putting a pawn on h6 rather than his queen, but he should not have carried out that plan on the next turn. 9.h4 h6?! (9...Bg7! equalizes according to Mikhalevski on ChessPublishing)

10.Kg1! By threatening hxg5 White gained the initiative. I will show you a bit more of the game, as Ivanchuk played fantastically in the next phase: 10...g4 11.Ne1 Bxb3 12.axb3 Ngf6 13.Nd3! g3 14.Qf3! Qxf3 15.gxf3 Nb8 White won in style after 16.Ne2 in Ivanchuk – Karjakin, Jurmala (rapid) 2015, although 16.Nxf4!N would have been quicker.   8.Qf3 8.Nf3 Qh6 9.h4 Nh5³   8.Nb5 0-0-0! 9.Nxa7†? Kb8 10.Nb5 Nxe4 would be disastrous for White.   8...Nh5 9.g3 9.Nge2 g5³ 9.Nb5 Kd8! is an important detail. 10.Bxe6 fxe6 11.e5

11...Kc8! 12.Qc3 c6 13.Nxd6† Bxd6 14.exd6 Qg5 Black intends ...Qd5 and ...Qxd6, when White will struggle to prove compensation for the pawn.   The text move is a typical King’s Gambit motif, but Black has a good way of meeting it.

9...fxg3 10.Kg2 Nhf6! Black plays ...Qg4 next and stands fine. A few more details:   11.hxg3 11.Bxe6 fxe6 12.hxg3 Qg4 13.Nb5 0-0-0! (13...Qxf3† 14.Nxf3 Kd8? 15.Ng5±; but 13...Kd8!? is playable, the main line being: 14.e5 dxe5 15.Qxb7 Rb8 16.Qxc7† Ke8„ Crazy stuff!)  

14.Qxg4 Nxg4 15.Nxa7† Kb8 16.Nb5 c6 17.Nc3 e5 18.d5 Rc8³ As often happens, White has regained the pawn but is positionally slightly worse.   11...Qg4 12.Rh4

12.Nb5 0-0-0! 13.Nxa7† Kb8 14.Qxg4 Bxg4 15.Nb5 Nxe4µ   12...Qxf3† 13.Nxf3 a6

14.e5 Bxb3 15.axb3 dxe5 16.dxe5 Ng8 White has a bit of activity for the pawn, but only Black can be better.   A2) 5.Nf3 Qh6  

6.d4 6.Nc3 Be6 7.Bb3 Nd7 8.d4 is a slightly different version of Ivanchuk’s plan, but Black can counter it in the same way as the previous variation, i.e. 8...Ngf6! intending 9.Ne2 Nh5!³.

  6...Be6 7.Qd3 Both 7.Bxe6 fxe6 and 7.d5 Bd7 8.Nc3 Nf6, intending ...Ng4, are fine for Black. The text move resembles an earlier Ivanchuk win over Nikolic in Antalya 2004, although in that game Black posted his queen on the awkward g4-square rather than the more convenient h6.   7...Nd7! 8.Nc3 Just as in the previous variation, Black has more than one good solution.

8...c6!?N Stopping any Nb5 ideas before we play ...Nf6.   I wouldn’t recommend 8...0-0-0 on account of 9.a4!, which created an attack and made the position highly unclear in Constance – Richard, email 2004.   However, Black might well consider: 8...Ngf6!?N 9.Nb5 9.Ne2 Nh5 is our main idea. 9.Bxe6 fxe6 10.Nb5 Kd8 also works out well for Black. 9...Kd8! Both kings are rather misplaced but Black remains a pawn up. The position is complicated of course, with all the pieces still on the board, but my analysis indicates that Black is at least equal. Here are a few sample lines:

10.Bxe6 10.e5 Nd5! 11.Bxd5 Bxd5 12.exd6 Bxd6 13.c4 Qg6! and the bishop goes to e4, since 14.Qxg6 allows 14...Bxc4†. 10...fxe6 11.e5 dxe5 11...Nh5!?÷ 12.dxe5 12.Nxe5?! Nxe5 13.dxe5† Nd7³ Black plays the king to c8 next and stands excellently. 12...Nd5 13.c4 Nb4 14.Qe4  

14...Nc5 14...Kc8 15.Bxf4 Nc5 16.Bxh6 Nxe4 17.Bc1 Bc5 is also nothing more than equal. 15.Qxf4 Qxf4 16.Bxf4 Nbd3=

  9.Ne2 This is White’s only critical idea, but Black has a nice way to counter it.

9...d5! 10.exd5?! This is the only ambitious move, but it turns out well for Black. 10.Bb3 is better, but after 10...dxe4 11.Qxe4 Ngf6 12.Qxf4 Qxf4 13.Bxf4 Bxb3 14.axb3 Be7= Black has no problems.   10...cxd5 11.Bb5 11.Bb3 Bd6 12.Qb5? Ngf6 13.Qxb7 0-0 is awful for White, as his pieces are poor and his king is unsafe.   11...Bd6 12.Ne5 Bxe5 13.dxe5 Qh4 14.Nxf4 14.Bxf4 Ne7³

14...a6 15.Bxd7† Bxd7 16.Qg3 Neither 16.Qxd5?! Rd8!µ nor 16.Nxd5 Rc8³ is an improvement.   16...Qxg3 17.hxg3 Ne7³ Black’s better pawn structure gives him a risk-free edge.   B) 3.Nf3 Nf6!

The Schallopp Defence! The exclamation mark is for the practical value of this move: it’s a rare choice, so White players are unlikely to know it as well as the main lines, and it offers Black at least equality in all lines, without the need to learn much theory.   White may respond with B1) 4.Nc3 or B2) 4.e5.

  B1) 4.Nc3 d5!   The most natural and energetic reaction.   5.exd5 5.e5 Ne4 gives Black an easy game, for instance: 6.d3 (6.Qe2 Nc5 7.d4 Ne6 is also at least equal for Black; if 8.g3 then 8...Nc6! hits the d4-pawn before playing ...g5 next) 6...Nxc3 7.bxc3 g5 8.h4 g4

9.Nd4 (9.Ng1 f6!?N 10.Bxf4 Bg7³) 9...c5 10.Nb3 Nc6 11.Bxf4 Bg7 12.d4 c4 13.Nc1 White had regained his pawn in Haida – Molisch, Brno 1937, but 13...f6!³ would have been positionally favourable for Black.   5...Nxd5 This position has often been reached via the move order 3...d5 4.exd5 Nf6 5.Nc3 Nxd5. By aiming for this position via the Schallopp Defence, we have avoided the major option of 5.Bb5†, which Shaw regards as White’s best, although he analyses the present variation in some detail as well.  

6.Bc4!? This may well be the best move, and it is certainly the most interesting.   The other main option is: 6.Nxd5 Qxd5 7.d4 Be7! This is an important detail to remember! 7...Bd6 8.c4 Qe6† 9.Kf2! may lead to serious complications, especially after 9...Qf6 10.c5! Be7 11.Qd2 g5 12.b4!, which Shaw analyses in some detail. By developing the bishop on the more modest e7-square, Black avoids being hit by c4-c5 and prepares to develop smoothly. 8.c4 Qe4† 9.Be2 Nc6 10.0-0 10.d5 Ne5 11.Nxe5 Bh4†! 12.Kf1 Qxe5µ

10...Bg4! Black has a slight edge due to his pressure in the centre. Shaw analyses several moves to confirm this assessment, but I will just mention his main line, which is based on a correspondence game: 11.d5 Ne5! 12.Kh1 0-0 13.Nxe5 Bxe2 14.Re1 Bxd1 15.Rxe4 Bc2 16.Re2 Bd6³ Tellier – Schmidt, email 2009.   6...Nxc3 The engines’ initial reaction is 6...Qe7† 7.Qe2 Nxc3 8.Qxe7† Bxe7 9.dxc3 g5, but I think Shaw is correct to say that both 10.h4 and 10.g3 give White enough compensation – although Black is not worse either.   7.dxc3 After 7.bxc3?! Bd6 8.Qe2† Qe7 9.Qxe7† Kxe7 10.0-0 Be6 11.Re1 Black is better after 11...Kf6, a move awarded an exclamation mark by Shaw, or 11...Kd7, which looks natural to me.   7...Qxd1† 8.Kxd1 Bd6 9.Re1† Kf8 White would have great compensation for the pawn if only he had a good way to bring the c1-bishop and a1-rook into the game. As things stand, Black is ready to develop his queenside pieces with a dominating position, so the following forcing line is White’s only serious attempt to cause problems.

10.Ne5 Bxe5 11.Rxe5 f6 12.Rc5 c6 13.Bxf4 b6 White’s rook is trapped, but he can get a knight and a pawn for it with the help of the following trick.   14.Rd5! cxd5 15.Bxd5 g5 16.Be3 Nc6 17.Bxc6 Bg4† 18.Kd2 Rd8† 19.Bd4 Rd6 20.Bb5 Bd7 21.Bd3

Shaw writes: “The machine does not rate White’s compensation as entirely adequate, but to my eyes the strong bishops make a good impression and the plan of a4-a5 looks rather irritating for Black.”   This seems true, but we have some ways to irritate White as well.   21...Bc6! 22.g3 h5! Black has active prospects on the kingside. With best play the position is on the verge of being equal, but I prefer Black’s chances.   B2) 4.e5 Nh5!   As we saw in the Bishop’s Gambit section, the knight stands quite nicely on h5 in the King’s Gambit. From this position Black can obtain a fully equal game with a defensive scheme that takes literally five minutes to learn.  

5.Be2 According to my experience, only well-prepared King’s Gambiters will play this move over the board. It’s the main line of theory, but does it look scary to you?   The more frequent continuation is: 5.d4 d6! This is Black’s usual plan, undermining White’s pawn centre. It is worth considering both a) 6.Qe2 and b) 6.Bc4!?.   a) 6.Qe2 d5! 7.c4 After other moves White’s queen would simply be misplaced. 7...dxc4 8.Qxc4

8...c6!

Shaw’s recommendation is indeed Black’s best. I tried to improve with 8...Nc6!?, threatening ...Be6, based on the fact that 9.d5? Nb4 is terrible for White. However, 9.Nc3 Nb4 10.e6! is an annoying resource. Best play continues 10...Bxe6N (10...Nc2†? 11.Kd2 Nxa1 12.exf7† was winning for White in Kawagoe – Nieuwoudt, email 2010) 11.Qb5† c6 12.Qxh5 Nc2† 13.Kf2 and now 13...Nxa1 14.Bd3 and 13...Nxd4 14.g3 are both rather unclear. This could be explored in more detail, but there is no point when the text move gives Black a more solid advantage with less risk. 9.Qb3 Be7 10.Bc4 0-0 11.0-0 Na6 12.Nc3 Nc7 13.Ne2 g5! Both sides have developed logically. With the white queen far away on b3, Black can afford to weaken his kingside slightly to preserve the f4-pawn. 14.Bd2 Ng7 15.Bd3 This position was reached in Baumann – Sueess, corr. 2010. Once again, Shaw suggests an improvement at exactly the right moment:

15...Bf5!N 16.Bxf5 Nxf5 17.Qxb7 Ne6 By returning the extra pawn, Black has achieved excellent coordination with useful pressure against the d4-pawn.   b) 6.Bc4!? This move is not that strong, but I want to draw your attention to it because it sets a nasty trap which, I’m sorry to say, I stupidly fell into!

6...dxe5? I will give this as the main line to show you what to avoid. 6...Be7 followed by ...0-0 was fine for Black in Piskov – Smejkal, Germany 1996. (Smejkal also happens to be a huge Breyer hero!) 6...g6!?N is another good option, intending ...Nc6 and ...Bg7 as in the main line below. 7.Bxf7†! 7.Nxe5? Qh4† was my idea, but I completely missed White’s typical pseudo-sacrifice. 7...Kxf7 8.Nxe5† Kg8 9.Qxh5 I got killed in a few more moves in Spyropoulos – Ntirlis, Patras 2015. At least I was in good company, as Schallopp himself once fell for the very same trick in 1885!

5...d6!

Lokander also gives this line; it stands out in Shaw’s book as a low-theory equalizer.   6.0-0 Shaw mentions 6.d4? is inaccurate due to 6...dxe5 7.Nxe5 (7.dxe5 Qxd1† 8.Bxd1 Nc6 intending ...g6 and ...Bg7 is bad for White) 7...Qh4† 8.Kf1 Ng3†, and I picked up a few easy points this way in my training games.   6...Nc6 7.d4 g6 8.Nc3 Bg7

Shaw admits White cannot claim any advantage. This is excellent news for Black, whose play was quite straightforward. The following game illustrates how Black’s kingside majority can produce a dangerous attack.    GAME 4 Mauro Marchisotti – Alena Lukasova Correspondence 2013   1.e4 e5 2.f4 exf4 3.Nf3 Nf6 4.e5 Nh5 5.Be2 d6 6.0-0 Nc6 7.d4 g6 8.Nc3 Bg7 9.exd6 cxd6 10.d5 Ne5 10...Ne7 is also playable according to Shaw. Alena Lukasova is a strong theoretical researcher, so I trust her choice.

11.Nd2 11.Nd4?! allows 11...Bg4! due to the hanging state of the d4-knight. 11.Nxe5?! dxe5 is excellent for Black as 12.Bxh5 gxh5 13.Qxh5 is met by 13...Qb6†µ followed by ...Qg6. Shaw ends his analysis at this position, stating that White can restore material equality but cannot claim any advantage.   11...Nf6 11...0-0!? is also interesting, as 12.Bxh5 gxh5 13.Rxf4 Ng6 14.Rf1 Bg4 gives Black a good deal of activity to make up for the structural damage.   12.Rxf4 0-0 13.Nf1 Re8 14.Be3 a6 15.a4 b6 16.Bf2 Nfd7 17.Ne3 Bb7 18.Rb4 18.Nc4 Qg5 19.Bg3 Nxc4 20.Rxc4 Rac8³

18...f5! I like Black’s position very much. Having developed most of her pieces, Lukasova starts to take some space on the kingside.   19.Nc4 Nxc4 20.Rxc4 Rc8 21.Rxc8 Qxc8 22.Qd2 Be5 23.Bd4 Qc7 In the next phase of the game Black organizes her kingside assault. The purpose of the last move seems to be to protect b6 in order to transfer the knight to the kingside. However, considering that the knight returns to d7 a few moves later, Black might have saved time with 23...Qd8!?.

24.Rd1 Nf6 25.Bf3 Bxd4† 26.Qxd4 Nd7 27.Be2 Qd8 28.Rd3 Qg5 29.Kf2 An awkward move to make, but White – a strong correspondence player himself – found no other way to deal with the threat of ...Qc1†.   29...Qc1 30.Re3 Ne5 31.Qxb6 Bc8 31...Qxc2? 32.Rxe5! (but not 32.Qxb7?? Ng4†) 32...Rxe5 33.Qxb7 wins for White.   32.h3! White has to stop the potential ...Ng4†.  

32...Bd7! 32...Qxc2?! 33.Qxd6 shows the importance of White’s previous move. 33...Qd2 34.Qc5²   The text move sees Black abandon her queenside in order to play ...f4 and go for an all-out attack against the exposed white king. It is also directed against the defensive resource of Rxe5, after which Black can take with the rook and block the check on d8 with ...Be8.   33.Bxa6 f4 34.Re2 f3 35.gxf3 Qf4 In a practical game, Black’s attack would be extremely difficult to repel. Armed with silicon assistance, however, White manages to defend with precise simplifications.   36.Rxe5 Rxe5 37.Qd8† Be8 38.Ne2 Qh2† 39.Ke1 Qg1† 40.Kd2 Qe3† 41.Kd1 Qxf3

42.Bd3 Qh1† 43.Kd2 Qxh3 44.c4 Re3 45.Nc1 Qh6 46.a5 Kg7 47.Qb6 Re7† 48.Kc2 ½–½   Conclusion   In this chapter my aim has been to provide strong yet easy-to-learn antidotes to a variety of tricky lines White may venture.   Against 1.e4 e5 2.Bb5 I suggest the simple 2...Nf6 3.d4 c6 followed by 4...Nxe4. This might not be a refutation, but Black tends to end up with a slight edge after the opening.   Nakamura’s 2.Qh5 leads after 2...Nc6 3.Bc4 g6 4.Qf3 to a situation where Black can either gain space with the ambitious 4...f5!? or develop solidly with 4...Nf6, when a subsequent ...Na5 leads to a pleasant version of a Vienna Game.   We then moved on to two sidelines after 2.Nf3 Nc6. The Ponziani is not a terrible opening, but the accurate 3.c3 Nf6 4.d4 Nxe4 5.d5 Bc5! puts serious pressure on White, who struggles to obtain equal chances even with best play. Against 3.d4 exd4 4.c3, the Göring Gambit, I examined what can happen if we take the pawn (and I suggest giving it back quickly!) but the real main line is 4...d5!, when Black equalizes easily and gets good chances to play for the full point, as Game 1 shows.   The Centre Game can be tricky to deal with, but Game 2 shows a near-refutation of the 8.Qg3 pawn sacrifice, courtesy of Vallejo’s bomb 14...Ba3!. This means that 8.Bc4 is White’s soundest move, but Game 3 shows that 8...d6 9.f3 Na5! gives Black excellent prospects.   The King’s Gambit is another opening that requires respect, but after 1.e4 e5 2.f4 exf4 my recommendations tick the necessary boxes for being theoretically sound and offering Black at least equal chances, without being overly complicated or difficult to learn. After 3.Bc4 Qh4† 4.Kf1 d6! our plan involves ...Be6, ...Nd7 and ...0-0-0. If White reacts with Nf3 we retreat the queen to h6, and if Nc3-e2 then ...Nf6-h5 secures the f4-pawn. By coincidence, exactly the same square is reserved for the knight in the 3.Nf3 variation. After 3...Nf6 4.e5 (4.Nc3 d5! also works out nicely for Black) 4...Nh5! Black has an easy plan of targeting the e5-pawn with ...d6, ...g6 and ...Bg7. Game 4 is a fine example of how Black may utilize the kingside pawn majority to build an attack.

  1.e4 e5   The Bishop’s Opening 2.Bc4 Nf6 3.d3 (3.d4 page 47) 3...c6 page 47 4.Nf3 (4.Qe2 page 47) 4...Be7 5.0-0 d6 6.h3 Game 5, page 48 (6.Re1 page 48; 6.c3 page 47 & 49)   The Vienna Game 2.Nc3 Nf6 page 51   A) 3.g3 page 51   B) 3.Bc4 Nxe4! 4.Qh5 (4.Nf3?! page 52) 4...Nd6 page 52 5.Qxe5† (5.Bb3 page 52) 5...Qe7 6.Qxe7† Bxe7 7.Bb3 Nf5! Game 6, page 53   C) 3.f4 d5! 4.fxe5 (4.exd5?! page 57) 4...Nxe4 page 57 5.Nf3 (5.d3 page 57; 5.Qf3 page 57) 5...Be7 6.Qe2 (6.d4 page 58) 6...Nxc3 7.dxc3 c5! 8.Bf4 Nc6 9.Qf2 Game 7, page 58 (9.0-0-0 page 60)  

Diagram Preview

On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

What is Black’s plan? (page 48)  

Can you find Rubinstein’s move? (page 54)

How does Black fully equalize? (page 49)

What was Reti’s idea to equalize? (page 58)  

Which is the most flexible move? (page 52)  

Should Black play 8...c5, or something else? (page 59)  

How does Black equalize? (page 53)

White has just played 9.Qf2.

Can you find Karpov’s reply? (page 60)

  This chapter will deal with two related systems. The Bishop’s Opening is a sensible way for White to get either a Vienna Game or an Italian-like position, so it cannot be bad. Instead of going for a sophisticated move order with many possible transpositions, I recommend a simple and direct solution which has been used by many strong players. The Vienna Game is a serious system which was examined thoroughly in a recent book by Ovetchkin and Soloviov; their suggestions forced me to treat the Vienna with respect, and I would advise the reader to do the same.

The Bishop’s Opening 1.e4 e5 2.Bc4

The Bishop’s Opening was recommended by Philidor in the 18th century as a way to avoid the good positions that arise for Black after 2.Nf3 d6! followed by ...f5!, as Philidor didn’t fancy Black’s chances after 2...Nc6 3.Bb5!. Opening theory has moved on a lot since Philidor’s time, but we continue to see this opening from time to time.   2...Nf6 3.d3 The Urusov Gambit, which arises after 3.d4 exd4 4.Nf3, can conveniently be met by 4...Nc6, immediately transposing to the Two Knights Defence (see Chapter 6). If you feel brave and want to hang on to the extra pawn, a reliable line is 4...Nxe4 5.Qxd4 Nf6 6.Bg5 Be7 7.Nc3 Nc6 8.Qh4 d6, followed by ...Be7 and ...Qd7, after which Black will decide where to castle.

3...c6 4.Nf3 4.Qe2 is the kind of move of which Philidor would have approved, but Black’s plan remains the same: 4...Be7 5.Nf3 0-0 6.Bb3 d6 7.c3 Nbd7 8.0-0 Nc5 9.Bc2 Bg4 With equal chances, like in the main line.   4...Be7 5.0-0 d6 I like this solid approach more than rushing to gain space in the centre with ...d5.   6.c3 6.h3 is a way to avoid the ...Bg4xf3 exchange, but Black is fine after 6...0-0 7.a4 Nbd7 8.Nc3 Nc5 9.Re1 Be6. It is worth spending some time familiarizing yourself with the critical position arising after 10.Bxe6 Nxe6 11.d4 Qa5! as featured in Game 5.   The text move leads to a harmless variation of the Philidor where White has timidly moved his pawn to d3 instead of d4. GM Yusupov covers this line nicely in Build Up Your Chess 2 from Black’s perspective.   6...0-0 7.Re1 Nbd7 8.Bb3 Nc5 9.Bc2 Bg4 10.Nbd2 Ne6 11.h3 Bh5 12.Nf1 This is one of the critical variations of this opening, and the following manoeuvre should be remembered.

12...Nd7! 13.Ng3 Bxf3 14.Qxf3 g6! Stopping Nf5 and preparing ...Bg5, as first played by Yusupov in 1994. For more details, see the note to move 6 in the game below.   GAME 5 Pierangelo Turati – Marek Sadowski email 2009 1.e4 e5 2.Bc4 Nf6 3.d3 c6 4.Nf3 4.Qe2 was seen in the famous game Polgar – Gelfand, Khanty-Mansiysk (3.2) 2009, which Polgar won in spectacular style to equalize the match and force a tie-break. However, Black gets a good game by following our normal plan: 4...Be7 5.Nf3 0-0 6.Bb3 d6 7.0-0 Nbd7 8.c3 Nc5 (Although Gelfand’s 8...a5 was fine, I see no reason to deviate from the plan featured in the main game) 9.Bc2 Bg4! This is one of the defining moves of our system. In the Spanish this move is rarely good, but with the knight going to e6 the placement of the bishop on the kingside is justified.

10.h3 Bh5 11.d4 exd4 12.cxd4 Ne6 13.Rd1 (13.g4?! Bg6³ followed by ...d5 is fine for Black as well) 13...d5! 14.e5 Ne8! 15.Nc3 N8c7 Black is ready to play ...f6 with an excellent game. Let’s follow a practical example for a few more moves: 16.g4 Bg6 17.Bf5

17...f6! 18.exf6 Rxf6 19.Bxg6 hxg6!µ In Ivanovic – Nikcevic, Cetinje 2012, Black had won the opening battle and was ready to put pressure on White’s weakened kingside.   4...Be7 5.0-0 d6 6.h3 I consider this to be the only really critical continuation, as practice has shown that Black is fine if he is allowed to play ...Bg4xf3.   6.Re1 0-0 7.Nbd2 Nbd7 8.Nf1 allowed Black to eliminate an important bishop with 8...b5 9.Bb3 Nc5 in Spasov – Eljanov, Turin (ol) 2006.

6.c3 0-0 7.Re1 Nbd7 8.Bb3 Nc5 9.Bc2 Bg4 10.Nbd2 Ne6 11.h3 Bh5 12.Nf1 Nd7! 13.Ng3 13.g4 Bg6 14.Ng3 Ng5 and Black is fine according to Yusupov. 13...Bxf3 14.Qxf3 g6! The importance of Black’s last few accurate moves was noted earlier. Let’s see how the game may develop: 15.Bh6 15.Be3 Bg5 gave Black no real problems in Gelfand – Yusupov, Munich 1994.

15...Re8 15...Ng7!?N is also okay according to Yusupov. Black’s main idea is still ...Bg5 but, with the rook on f8, Black might also have the chance to play ...f5. 16.Rad1 Bg5 Yusupov evaluates the position as equal, and I agree.

6...0-0 7.a4 Nbd7 8.Nc3 Nc5 9.Re1 Be6

Black should be fine, but of course anything can happen with a board full of pieces. The only theoretically challenging continuation is:   10.Bxe6 Nxe6 11.d4 White tries to take the initiative in the centre, but Black’s next move keeps everything under control. 11...Qa5! Black takes advantage of the slight queenside weaknesses and gets ready to centralize his rooks.   12.Be3 Rfe8 12...exd4 13.Nxd4 Nxd4 14.Bxd4 Rfe8 is close to equal, but it is more interesting to maintain the central tension.   13.d5!? White decides to force things rather than allow ...exd4 followed by ...Rad8 at Black’s convenience. However, the resulting structure holds no great dangers for Black, who gets his own chances on the kingside.   13...cxd5 14.Nxd5 Nxd5 15.exd5 Nc5 16.c4 Rac8 16...g6!?N 17.Rb1 Qb4! 18.b3 f5 is a logical alternative. It is worth mentioning the following forcing line:

19.Nxe5!? dxe5 20.d6 Bf6 21.d7 Nxd7 22.Qd5† Kh8 23.Qxd7 Red8 The position remains balanced.   I also found nothing wrong with 16...Nd7!?N. The knight is flexibly placed on this square, restraining White’s queenside pawns while also eyeing the f6- and e5-squares after a future expansion with ...f5 and/or ...e4.

17.Qc2 e4!? 17...g6N is a good alternative. A possible continuation is 18.b4 Qxb4 19.Reb1 Qa5 20.Rb5 Qd8 21.Bxc5 Rxc5 22.Rxc5 dxc5 when White has some play for the pawn, but no advantage.   18.Bxc5 exf3 19.Bd4 fxg2 19...Qxd5 20.Qxh7† Kxh7 21.cxd5 fxg2 leads to equality, one possible continuation being: 22.Bxa7 Bf6 23.Rxe8 Rxe8 24.Be3 Re4 25.a5 Rb4=   20.Re2 Bf8 21.b3 a6 Black wants to be able to move his queen, so he makes sure the a-pawn won’t hang.

22.Re4

White wants to take over the e-file by putting his queen on e2. He has slight pressure, but Black holds the position comfortably.   22.Rae1 would allow 22...Qxe1†!? of course.   22...Rxe4 23.Qxe4 Qb4 24.Qe3 Rc7 25.Kxg2 Re7 26.Qc3 After 26.Qd3 Qa5 Black brings his queen back to d8 and might even dream about ...Re8 and ...Qg5†. He stands fine anyway.

26...Qxc3 27.Bxc3 f5 28.Bb4 Re4 29.Re1 Rxe1 30.Bxe1 Kf7 31.Kg3 White decided it was not worth trying to turn his tiny advantage into something more. If you encounter this variation over the board, I would suggest going for one of the alternatives given on moves 16 and 17, when White has even less. ½–½

The Vienna Game 1.e4 e5 2.Nc3 Nf6! Moves like 2...Nc6 and 2...Bc5 are playable, but there is a risk of transposing to a variation of the King’s Gambit which falls outside our repertoire, and where White has realistic chances of an opening advantage. I find my recommended line to be both theoretically sounder and easier to play.

White’s three main options are A) 3.g3, B) 3.Bc4 and C) 3.f4.   A) 3.g3   This is a reasonable move, but we will not spend much time on it here, as it should transpose to a system considered in a later chapter.   3...d5 4.exd5 Nxd5 5.Bg2 Nxc3 6.bxc3 Bc5! 7.Ne2 White’s best move is 7.Nf3, which immediately transposes to the Glek System, coverage of which begins on page 77. I decided to cover this system via the Four Knights move order as, in my experience, that is how the Glek System is usually reached. Putting the knight on e2 leaves the path clear for the f-pawn, but the knight has less influence over the centre and Black obtains a promising game by simple means.   7...0-0 8.0-0 Nc6 9.d3

9...Bg4! 10.h3 Be6 11.c4 Qd7 12.Kh2 Rad8 13.Be3 Bxe3 14.fxe3 Ne7 15.Nc3 f5³ Black was better in Liu Qingnan – Harikrishna, Beijing 2012.   B) 3.Bc4

3...Nxe4! The widely recommended course for Black is 3...Nc6 4.d3 Na5, when 5.Nge2 Nxc4 6.dxc4 has occurred in a lot of games. Theoretically Black is fine here, but White’s position is easy to play and in a practical game Black might find it difficult to choose the right plan. The authors of The Modern Vienna also convinced me that 5.Bb3!? is a serious move. The text move seems like a better practical decision, as it leads to simpler positions where the chances of forgetting one’s preparation or struggling to understand the middlegame position are much lower.  

4.Qh5 4.Nf3?! is a rare old line which is closely related to another dodgy gambit in the Italian Four Knights. After 4...Nxc3 5.dxc3 f6 6.0-0 d6 Black is a clear pawn up, and it is not clear how White can utilize his lead in development. His best try is:

7.Nh4 Otherwise Black will play ...c6 and ...d5, with an almost winning position. The text move prepares to launch the f-pawn, but after 7...g6 8.f4 Qe7 The queen goes to g7 and Black has everything covered. This is as much as you need to know to meet this rare line, but I’ll provide a few more details all the same: 9.f5 Qg7 10.Be3 c6 11.Qe2 (11.Be6 Bxe6 12.fxe6 Na6! followed by ...Nc7xe6 was close to winning for Black in Thorn – Sward, Sweden 1993) 11...d5 12.Bb3 In Crepan – Rezonja, Bled 2000, Black should have played:

12...Be7!N Preparing to meet c3-c4 with ...d4, while also getting ready to castle. White has no real compensation for his missing pawn.  

4...Nd6 5.Qxe5† After 5.Bb3 Be7! 6.Qxe5 0-0 7.d4 I like the suggestion of the Greek IM Panayotis Frentzas:

7...b5! An excellent move, preparing queenside expansion with ...a5 while waiting for White to occupy the f3-square with his knight before hunting the b3-bishop with ...Nc6-a5. See the notes to Game 6 for further details.   5...Qe7 6.Qxe7† Bxe7 7.Bb3  

7...Nf5! 8.Nf3 8.Nd5 Bd8 is fine for Black, who intends ...c6 and ...Nd4.   8...c6! Black has equalized. The position might seem rather dry, but the following game provides a classic demonstration of

simple endgame concepts from an outstanding genius, reminding us that an equal position is not the same as a drawn one.    GAME 6 Stefano Rosselli del Turco – Akiba Rubinstein Baden-Baden 1925   1.e4 e5 2.Nc3 Nf6 3.Bc4 Nxe4 4.Qh5 Nothing else makes much sense here; 4.Nxe4 d5 is more than fine for Black, and we have already seen that 4.Nf3?! is inadequate.   4...Nd6 5.Qxe5† This is the safe option, which also happens to be White’s best.   5.Bb3 This has been a more popular choice, but it is well met by: 5...Be7! 5...Nc6 6.Nb5 g6 7.Qf3 f5 8.Qd5 Qe7 9.Nxc7† Kd8 10.Nxa8 b6 is the so-called Frankenstein-Dracula Variation. Black is temporarily a rook down, but he will pick up the a8-knight and he has some attacking chances. Unfortunately, after the precise 11.Nxb6 axb6 12.Qf3 Bb7 13.d3 Nd4 14.Qh3 e4 15.Be3!± White plans to castle long and Black’s compensation is inadequate.

6.Qxe5 6.Nf3 Nc6 7.Nxe5 g6 is fine for Black. The main line runs 8.Nxc6 dxc6 9.Qf3 0-0 10.0-0 and now Black’s best idea is 10...Nf5!= intending ...Nd4. 6...0-0 7.d4 b5! This nice idea was proposed by IM Frentzas on the ChessPublishing forum. Black is waiting for White to occupy the f3-square before going after the b3-bishop.

By contrast, 7...Nc6 8.Qf4 Na5 9.Bd5 Ne8 10.Bf3² was slightly better for White in Caruana – Giri, Wijk aan Zee 2014. 8.Nge2 White logically keeps the f3-square available for the bishop in case Black chases it. 8...Nc6 9.Qf4 b4 10.Nd5 In this position Frentzas found an important improvement over Keller – Royer, Champs sur Marne 2007:  

10...a5!N With strong queenside counterplay.   5...Qe7 6.Qxe7† Bxe7 7.Bb3 For several decades this line was considered equal, but Ovetchkin and Soloviov shared some new analysis demonstrating that White can apply some pressure if Black is not accurate.   7...Nf5! 7...c6 allows 8.d4! Nf5 9.d5².   8.Nf3 c6 9.0-0 d5 10.Re1

10...Na6! It is curious that the authors of The Modern Vienna missed this particular move order and examined only 10...0-0. Instead of castling, the black monarch can find a better home on f7 or d7. Black’s last move is an efficient way to defend e6 (because Black will play ...f6 soon) as well as the d5-square (to be able to defend comfortably against c3-c4).   11.d4 h5!? I could certainly stop at this point, claim equality and move on, but I think you will find it useful to have a model game for such a position – and one could not ask for better than this Rubinstein masterpiece.   12.Ne2 Nc7 13.c3 f6 14.Ng3?! This is White’s first inaccuracy. 14.Nf4, 14.Bf4 or even 14.Bc2 would all have kept things about equal.   14...Nxg3 15.hxg3 g5! Rubinstein takes the first step towards exploiting the slight weakening of White’s kingside.

16.Bd2 Bf5 17.Re3 Kd7 18.Rc1 Bd6 19.Bc2 Bxc2 20.Rxc2 Rae8 21.Bc1 Rxe3 22.Bxe3 Ne8! The knight is no longer needed on c7, so Rubinstein brings it to f5.   23.Re2 Ng7 24.Bd2 Nf5 25.Re1 c5! The famous Soviet trainer Razuvaev commented: “Taking the decision to play such a move is harder than sacrificing twenty bishops on h7!”

26.dxc5 Otherwise Black will have the option of playing ...c4, followed by gaining additional space on the queenside and creating a weakness on c3.   26...Bxc5 27.Kf1 h4! 28.gxh4 g4! 29.Nd4 Bxd4 30.cxd4 Rxh4 31.Bc3

31.Be3 g3! undermines the bishop and keeps White under pressure.   31...Rh1† 32.Ke2 Rh2 33.Rg1 Nh4 34.g3 This weakens the light squares, but 34.Kf1 Ke6 leaves White terribly passive, while Black can improve his king and eventually attack the d4-pawn.

34...Nf5 35.b3 Ke6 36.Bb2 a6 37.Bc3 Nd6 38.Ke3 Ne4 39.Be1 Kf5 40.Rf1 Rh8 41.Kd3 b5 The black pawns cooperate nicely with the knight, controlling both light and dark squares. Meanwhile White has no targets, no counterplay and no way to stop the potentially killing ...f5-f4 break.

42.a3? This move not only places another pawn on the same-coloured square as White’s bishop, but also leaves both of his queenside pawns vulnerable on the third rank, which Rubinstein exploits in the game. There is actually a little story

about this move as well. For many years it had been recorded erroneously in many sources (including Tarrasch’s tournament book) that White had played his pawn to a4, but this made little sense as it is difficult to believe that Rubinstein would have foregone the opportunity to take on a4 and put his rook on c4. It was only an investigation by the famous chess historian Edward Winter that revealed that Rosselli had actually moved his pawn one square rather than two!   42...Rh7 43.Ba5 Rh8 44.Bb4 Rc8 45.Ba5 Ng5 46.Bd2 Ne4 47.Ba5 Kg6 48.Bb4 f5 After going back and forth, Rubinstein realizes he cannot make progress if he doesn’t use his f-pawn.

49.Ba5 Rh8 50.Ke3 Re8 51.Kd3 Rc8 Black could have played 51...f4!, just as he did on move 55, the only difference being the rook’s position on e8 instead of e7, which does not change anything. Perhaps Rubinstein had not yet decided on the winning plan, or maybe he just wanted to play a few more moves to tire his opponent.   52.Bb4 Rc7 53.Ba5 Rh7 54.Ke3 Re7 55.Kd3 f4! At last the decisive thrust comes! One way or another, White will be left with crippling pawn weaknesses.

56.gxf4 If 56.Bd2 then 56...Nxd2! 57.Kxd2 f3 and the rook penetrates to e2.   56...Rh7 57.Bd2 Nxd2! According to Donaldson and Minev in their book about Rubinstein: “Boris Gelfand points to this game as an antecedent for Fischer’s similar trade of a beautiful knight for a bad bishop in game seven of his match with Petrosian.” Boris is referring of course to the famous 1971 Fischer – Petrosian game where Fischer, at move 22, made one of the deepest and most famous positional decisions in the history of our game.   58.Kxd2 Rh3

59.f3 gxf3 60.Rf2 Kf5 61.Ke3 Kg4 62.b4 Rh1 63.f5 Re1† 64.Kd3 Re4

White’s passed pawn can be neutralized by ...Rf4, whereas Black’s is decisively supported by the king, so White resigned. 0–1   C) 3.f4   This is traditionally regarded as the main line of the Vienna Game. White is trying to play an improved King’s Gambit, but Black himself has an improved version of the ...d5 defence.   3...d5! Black would be ill-advised to try and repeat the recipe against the King’s Gambit, as 3...exf4? 4.e5 forces the humiliating 4...Ng8. Then 5.Nf3 d5 6.d4 g5 7.h4! g4 8.Ng1!± is a brilliant piece of preparation from my junior years when I was venturing this line with White!

4.fxe5 After 4.exd5?! Nxd5 5.fxe5 Nxc3 6.bxc3 Qh4† 7.Ke2 Bg4† 8.Nf3 Nc6 White is struggling to equalize. His best hope is to hang on to his extra pawn for a while with:

9.Qe1 Qh5 10.d4 0-0-0 11.Kf2 f6! 12.Bb5! fxe5 (Black can also consider 12...Bd7!?N 13.e6 Qxb5 14.exd7† Rxd7 when he remains slightly better due to his superior pawn structure) 13.Bxc6 bxc6 14.Nxe5 Bd6 15.h3 Rhf8† 16.Kg1 Bxe5 17.Qxe5

17...Qxe5 18.dxe5 Rd1† 19.Kh2 Rxh1† 20.Kxh1 Rf1† 21.Kh2 From this position both 21...Be6 and 21...Bf5 have scored positively for Black, although White should be able to hold with accurate defence.   4...Nxe4 5.Nf3 Alternatives will be discussed in the notes to Game 7. Here is a brief summary:   5.d3 Nxc3 6.bxc3 Be7!? followed by castling and ...f6 is a sound and simple way to play for Black.   5.Qf3 Nxc3 6.bxc3 (6.dxc3 is similar to the main line; Black seeks to castle queenside after playing ...Be6 and ...Nc6) 6...c5! 7.Qg3 Bf5! 8.Bd3 Be6 is an important line which will be covered in more detail below.

5...Be7!? This move was recommended by the Hungarian GM Gyula Breyer, whose ideas are being taken rather seriously in this book! This system is not the most ambitious available to Black, but it is extremely simple to learn. The main idea is to castle and play ...f6.   6.Qe2 The most popular continuation.   6.d4 0-0 7.Bd3 f5 is a well-known equalizing variation. The main line runs 8.exf6 Bxf6 9.0-0 Nc6 10.Nxe4 dxe4 11.Bxe4 Nxd4 12.Ng5 and now Reti discovered the neat idea of:

12...Bf5! and Black equalizes with ease. This is just about the only concrete line that you need to memorize before playing Breyer’s 5...Be7.

  6...Nxc3 7.dxc3 c5 Practice has shown that Black is fine. He will castle queenside and target the e5-pawn or try to create further weaknesses on White’s queenside. The following game is a model demonstration of Black’s play.    GAME 7 Andrei Istratescu – Anatoly Karpov Bucharest (rapid – 5) 2005   1.e4 e5 2.Nc3 Nf6 3.f4 d5 4.fxe5 Nxe4 5.Nf3 5.d3 Nxc3 6.bxc3 Be7 (6...d4 followed by ...Nc6 and ...Bc5 is the main line and is good for Black, but it is more complicated than I would like a rare variation in my repertoire to be. Who is going to remember the details?) 7.Nf3 0-0 8.Be2 f6 Black’s plan is sound and simple. 9.exf6 (Also 9.0-0 fxe5 10.Nxe5 Nd7= is absolutely fine for Black) 9...Bxf6 10.d4 The position is equal, and the following game shows one good way for Black to proceed:

10...c5 11.0-0 Qa5!? I like this move, which stops any Ba3 ideas. 12.Bd2 Nc6 13.Kh1 Bf5! In Nagy – Laszlo, Hajduboszormeny 1997, Black had the better structure and more active pieces. Over the next few moves the rooks will go to the centre and the f5-bishop to e4.   5.Qf3 Nxc3 I’d like to stick to the ‘exchange on c3 when the e4-knight is attacked’ policy to keep things simple. 6.bxc3 6.dxc3 is similar to our main game; the queen looks strange on f3, although it does at least force our bishop to go to e6 instead of f5. 6...Be7 7.Bf4 Be6 8.0-0-0

8...Nd7! (Not 8...c5? 9.Bc4!) Black is planning ...c5, ...Qa5 and long castling. 9.Qg3 g5! 10.Be3 c5 11.Kb1 Qa5 Black was ready to castle and soon took over in Wang Yu – Ganguly, Mumbai 2003.

6...c5! This fine waiting move is directed against d2-d4, and keeps the plan of ...Be7 and ...0-0 in reserve for now. 6...Be7 7.d4 0-0 8.Bd3 f6 9.Qh5 creates some complications which I’d like to avoid. In the event of 9...g6 10.Bxg6 hxg6 11.Qxg6† White has at least a perpetual. 7.Qg3 7.d4? is met by 7...cxd4 8.cxd4 Qh4† winning a pawn. The text move has been the main try for White. The idea is to make it harder for Black to castle, but we have an excellent resource: 7...Bf5! I very much like this move in connection with the follow-up below. Apart from interfering with White’s

development, Black prepares to castle on the queenside (after ...Nc6 and ...Qd7 or ...Qa5) or on the kingside (after ...Bg6 preparing ...Be7). What more can you ask for from a single move? 8.Bd3N This is the best my engines can come up with for White. 8.Bb5† Nd7 9.Nf3 a6 is fine for Black.

8...Be6! 9.Nh3 In the event of 9.Nf3, Black stops Ng5 with 9...h6 and continues with ...Nc6, ...Qc7, ...0-0-0, ...g5 and so on. 9...c4! 10.Be2 Qc7 11.Ng5 Nd7 It is White who is struggling to keep the equilibrium.   5...Be7 6.Qe2 This is the most ambitious way to play for White, but as we saw in the variations above, Black shouldn’t have any problem exchanging on c3 and thus “helping” White open lines for his c1-bishop.   6.d3 Nxc3 7.bxc3 transposes to 5.d3 above.   6...Nxc3 7.dxc3 c5! 8.Bf4 Nc6 9.Qf2 White intends Qg3 followed by Bd3 and 0-0.   9.0-0-0 Be6 10.h4 10.c4?! is only dangerous for White. 10...d4 11.Nd2 Qa5 12.a3 0-0-0 13.Ne4 Qa4!? Black was slightly better in Hector – Ivkov, Cannes 1989, the main point being: 14.Nd6†N 14...Kb8! 15.Nxf7 Rdf8! 16.Ng5 (16.Nxh8 Rxf4–+) 16...d3!–+ with a deadly initiative for the slight material deficit. 10...h6 11.g3 Qd7 12.Bg2 0-0-0 13.h5 Having made all the obvious developing moves, Black must decide on a plan. Anand comes up with a clever idea:

13...Na5!? 14.b3 Nc6 The knight returns, having provoked a significant weakening of White’s queenside. 15.Qf2 Qc7 Black wants to put his queen on a5 and most probably hide his king on a8 before looking for a suitable moment to break with ...d4 or ...c4. 16.Kb1 Kb8 17.Nh4?! This makes things easier for Black. 17...d4! White was under a bit of pressure in Adams – Anand, Linares 1994, although Anand played slightly inaccurately and agreed to a draw a few moves later. In our main game White is not treated so kindly!

9...h6! White’s last move prepared Qg3, so Karpov gets ready to meet that with ...g5. The ...h6 move is often useful in this structure, preventing the enemy pieces from using the g5-square at any point in the near future.   10.Be2 10.Qg3?? g5 11.Bd2 g4 wins a piece, as retreating the knight would allow ...Bh4 winning the queen.   10...Qb6 11.Rb1 11.0-0-0 Be6 12.Qg3 0-0-0 gives Black a slightly improved version of the Adams – Anand game quoted above.   11...Bf5!? 11...Be6 followed by ...0-0 – or even ...0-0-0 and ...g5 – was not a bad alternative, but Karpov wants to win a pawn!

12.Bd3 12.Bd1?! is pathetic. After 12...Qa6 13.a3 0-0 14.Qg3 Kh8 15.Kf2 White is ready to evacuate his king from the centre, but it is already too late: 15...d4! 16.cxd4 cxd4 17.Re1 Qb6 and with ...d3† coming, White is in serious trouble.   12...Bxd3 13.cxd3 Qa6 14.Rd1 Qxa2 15.0-0 d4!? Karpov offers to return the extra pawn in order to take control over the central light squares.   According to the ruthless engines, it would have been better to play 15...0-0!N, and if 16.Qg3 Kh7, with chances for a more substantial advantage.   16.c4?! This must have been a difficult decision, as both of White’s options look unpromising. Still, simplifying and regaining the lost pawn would have been preferable: 16.cxd4 Qd5! (after 16...cxd4 17.Nxd4 Bc5 18.Be3 White is doing fine) 17.dxc5 Bxc5 18.d4 Bb6 Black has a nice blockade in the centre, but White has a central majority and his pieces are active; the engines confirm that the position is roughly equal, although we should remember that Karpov excels in such positions, so it is understandable that Istratescu preferred to keep the position more complex.

16...Nd8 17.Bc1? This is a more serious mistake.   Playing more actively with 17.e6 Nxe6 (17...0-0 18.Ne5! Nxe6 transposes) 18.Ne5 was to be preferred although, being two pawns up, Black can afford to give up the exchange:

18...0-0! 19.Nd7 Qb3!? The computer calls it equal, but I would consider Black to be the slight favourite with two pawns for the exchange.

17...Qa6! 18.e6 The right idea, but it’s a move too late.   18...0-0 19.Qg3 19.exf7† Nxf7µ   19.Ne5 Qxe6! Black takes control over d7 and after 20.Rde1 Bd6µ White doesn’t have a good discovered attack.   19...Nxe6 20.Rde1 White has to remain two pawns down and hope for a cheapo, as 20.Bxh6? is refuted by: 20...Bd6! 21.Qg4 f5 22.Qh5 (22.Qg6 Rf6–+) 22...gxh6 23.Qxh6 Bf4 24.Qg6† Ng7–+

20...Qd6 21.Ne5 Bg5

Karpov is in full control, and he converts his material advantage with his customary exemplary technique.   22.Bxg5 Nxg5 23.h4 Nh7 24.Qf3 Rae8 25.Ng4 Rxe1 26.Rxe1 b6 27.Re5 a5 28.Rd5 Qe7 29.Re5? Blundering a third pawn, but the result was a foregone conclusion in any case.   29...Qxh4 30.Re2 Qg5 31.Nf2 Nf6 32.Qc6 Ng4 33.Nxg4 Qxg4 34.Re1 Qg3 35.Qe4 f5 36.Qe6† Kh7 37.Re5 Rf6 0–1   Conclusion   The Bishop’s Opening is often used as a transpositional weapon but my recommended medicine, which was also prescribed by Yusupov, avoids any such tricks. The key manoeuvre for Black is ...Nd7-c5-e6, but since this would seriously limit the scope of the c8-bishop, we should exchange this piece for either the b3-bishop or the f3-knight. In the latter case, it is important to follow up by trading the dark-squared bishops with a quick ...Bg5.   The Vienna Game is a respectable opening. After 1.e4 e5 Nc3 Nf6 3.Bc4 I like the simplicity of 3...Nxe4; even though it is not the most ambitious move available, it takes away White’s attacking potential and gives Black equal chances from an early stage in the game.   In the event of 3.f4 d5 4.fxe5 Nxe4 there is something to be said for 5.Qf3, which I suggest meeting with 5...Nxc3 6.bxc3 (for 6.dxc3 check the course of the Wang – Ganguly game mentioned in the notes) 6...c5! 7.Qg3 Bf5! intending 8.Bd3N 8...Be6! with fine prospects for Black. The main line is 5.Nf3, where I like the straightforward 5...Be7, and after 6.Qe2 Nxc3 7.dxc3 c5 Black is doing well, as Karpov demonstrated in Game 7.

1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6   The Halloween Gambit 4.Nxe5 Nxe5 5.d4 Game 8, page 65   4.a3 d5! 5.Bb5 (5.exd5 page 69) Nxe4 page 69 6.Nxe5 (6.Qe2 page 69) 6...Qf6! Game 9, page 69   4.d3 Game 10, page 71   4.h3 page 73   4.Be2 page 74   The Italian Four Knights 4.Bc4 Nxe4! page 75   The Glek System 4.g3 d5! 5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3 Bc5! page 77 8.0-0 0-0 9.Re1 Qf6! 10.d3 (10.Qe2 pages 79 & 80; 10.Rb1 page 80) 10...Bb6! 11.Bg5 Game 11, page 80 (11.a4!? page 81) 11.Be3 Game 12, page 83  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following

chapter.   Black is to move unless otherwise indicated.  

How does Black refute White’s play? (page 67)  

Which is the correct square for

the queen? (page 81)

How did Nezhmetdinov punish

White’s timid opening play? (page 72)

How does Black take the initiative? (page 82)  

What is the most accurate move? (page 75)  

How does Black launch an attack? (page 84)  

White has just played 6.dxc3. How should Black proceed

How should Black continue?

in this critical position? (page 76) (page 84)   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 The Four Knights Game is an old and dependable opening system which seems to return to fashion every ten years or so, when someone comes up with a slight improvement over a main line or even discovers a whole new concept. It was rather popular in the early 1990s, but by the turn of the millennium it was believed to have been completely neutralized. When analysis engines became stronger and all opening lines were re-evaluated, the Four Knights started to reappear in

the repertoires of top grandmasters, although that trend has declined a little in the last few years. Up until quite recently, the term “Four Knight’s Game” was widely used to describe what is more accurately called the Spanish Four Knights, which arises after 4.Bb5!?, as all White’s alternative had been dismissed as harmless. Today though, following the likes of Kramnik, 4.d4 has become trendy again, and you can find these two moves covered separately in the next chapter. Before then, we will cover an assortment of alternative options for White. Moves like 4.a3 and 4.h3 have their merits, and even 4.Nxe5 is not as ridiculous as it looks. We also must not forget 4.g3, the Glek System, named after the German Grandmaster who popularized it. Theoretically Black has more than one route to equality, but a long and complex fight still lies ahead.

The Halloween Gambit We begin our investigation with an outrageous-looking piece sacrifice.    GAME 8 Gennady Obukhov – Steffen Schubert Email 2009   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Nxe5 Nxe5 5.d4 To give up a knight in this way looks like a beginner’s mistake, but it should not be underestimated. According to the database White scores 48%, and many Black players have been demolished after hanging on to the extra piece with seemingly natural moves. The gambit can be refuted, but it is worth paying close attention to the method.

5...Nc6! I like this move for two reasons. First and foremost, I consider it the strongest route to an outright refutation of the Halloween Gambit. Secondly, it has the practical advantage of giving you an easy way to return the material and get a positional plus in the event that you forget some of the details of the more complicated refutation.   5...Ng6 6.e5 Ng8 7.Bc4 d5!? 8.Bxd5 c6 is a well-known line which is credited to Euwe, and which has been reproduced in several modern sources. Although Black might be objectively better here too, I find the resulting positions more unclear than I would like, and White’s strong central pawns offer him ongoing compensation.   6.d5 6.e5? Ng8 7.f4 was played by Svidler in one of his ‘Banter Blitz’ sessions for Chess24, when he ventured the gambit to mark the 31st of October! However, this is the wrong way to advance the pawns, a fact which becomes especially apparent after 7...d5!N followed by ...Nh6, when White’s pawn majority is going nowhere and he is simply a piece for

a pawn down.   6...Ne5 6...Nb8?! 7.e5 Ng8 should be avoided, as 8.d6! puts Black in real danger.   6...Bb4?! 7.dxc6 bxc6 takes us to the Scotch Four Knights but, if you wish to return the piece for safety, there are more favourable ways of doing so.   7.f4 Ng6 8.e5 We have reached a critical position – this is the moment where Black must choose between aiming for an outright refutation or settling for a smaller but safe advantage.

8...Ng8!? This is the ambitious choice. If you want a simpler solution, the following alternatives both work well:   8...Bc5 9.exf6 0-0! left White struggling to contain Black’s lead in development in Kopylov – Biriukov, Tiumen 2013.   8...Bb4 9.exf6 Qxf6 also gives Black an easy game with level material.   9.d6 This is the only serious attempt to justify White’s gambit. After 9.Qe2 Bb4 Black is ready to develop the g8-knight and castle. The only way to disrupt this plan is 10.f5 N6e7 11.f6, but after 11...gxf6 12.exf6 Nxf6 13.Qe5 Rg8 14.Qxf6 Nxd5µ White is in a terrible situation.

9...cxd6 10.exd6 Qf6 11.Nb5 The idea of penetrating to c7 is something of a trademark for the Halloween Gambit, but Black’s powerful reply kills the whole idea.   11.Qe2†!N is the best try, but the queen exchange makes it easier for Black to convert his extra material. My analysis continues: 11...Qe6 12.Nb5 Qxe2† 13.Bxe2 Rb8 14.Be3 b6 15.0-0-0

15...Nh4! 16.g3 Nf5 This is the most important manoeuvre to remember; the remaining moves are pretty easy. 17.Bf2 Nf6 18.Rhe1 Kd8 White’s initiative has run out of steam, for instance 19.Nxa7 Ne4!µ and Black follows up with ...Nfxd6 and ...f5.

11...Nxf4! This is the crucial move, sacrificing material for an attack to turn the position on its head!   The materialistic 11...Kd8 has been more popular but after 12.Be3, followed by Qd2 and 0-0-0, White has a serious initiative.   12.Nc7† Kd8! 13.Nxa8 13.Qf3 Bxd6 14.Nxa8 b6 is dead lost for White as well.   13...Qe5†! White is the exchange up for the moment, but that’s the end of the good news for him. The white king is forced to move, and will soon come under a fierce attack. Black is poised to win the d6-pawn, and will reclaim a material advantage after picking up the trapped knight on a8.

14.Be2 14.Kf2 Bxd6 15.Bd3 occurred in Feng – Jara, Internet 2010, when the simplest winning continuation would have been: 15...Nxd3† 16.cxd3 (16.Qxd3 Nf6–+) 16...Nf6–+   14...Nxg2† 15.Kf1 Another game continued: 15.Kf2 Nh4 16.Qg1 Nf5 17.Qg5† f6 18.Qf4 Bxd6

19.Qxe5 In Hoynck van Papendrecht – Hameister, email 2007, Black went on to win after recapturing with the bishop, but it feels more harmonious to play 19...fxe5!N 20.Bg5† Nf6 21.c3 Bc5† followed by picking up the knight on a8, with a material advantage and central domination.   15...Nf4 16.Qd2 Another massacre ended in just five more moves: 16.Bf3 Bxd6 17.c4 Nf6 18.Qc2 Bc5

19.Qc3 Qf5 20.Bxf4 Qxf4 21.Kg2 Ne4 0–1 Reppert – Dehaybe, email 2009.   16...Nxe2

17.Qg5†!? 17.Qxe2 gives Black a choice between 17...Bxd6, leading to a winning endgame, and 17...Qxd6, when the combination of White’s exposed king and his trapped knight will decide the game. The text move enables the a8-knight to escape, but it is not enough to save White.   17...Qxg5 18.Bxg5† f6 19.Be3 b6 20.Nc7 Bb7 21.Kxe2 Bxh1 22.Rxh1 Bxd6 23.Nb5 Bb8–+ White is two pawns down and has no counterplay. The rest is easy.

24.h4 Ne7 25.Kf3 Re8 26.h5 Nc6 27.Rg1 g5 28.hxg6 hxg6 29.Rh1 g5 0–1

4.a3 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.a3

This odd-looking move had a brief period of popularity, and is sometimes referred to as the Gunsberg Variation. In general, I propose to meet all irregular 4th moves with 4...d5, seizing space in the centre and asking White how he intends to justify his eccentric play.   4...d5! 5.Bb5 This is White’s main idea.   5.exd5 Nxd5 6.Bb5 Nxc3 7.bxc3 Bd6 reaches a reversed Scotch Four Knights where White’s extra a2-a3 move has hardly any value.   5...Nxe4 6.Nxe5 6.Qe2 Qd6 7.Nxe4 dxe4 8.Qxe4 Bd7 gives Black a comfortable game – see the notes to Game 9 for more details.   6...Qf6! Black attacks the e5-knight and threatens mate on f2, thereby forcing the knight to retreat.   6...Qg5!? is a good alternative which is also discussed in the notes below.   7.Nf3 Be6 Black continues with ...Bd6 and is at least equal, as the following game demonstrates.    GAME 9 Cees Raijmaekers – Guntis Gerhards

Correspondence 2014   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.a3 d5 5.Bb5 Nxe4 6.Nxe5 6.Qe2 Qd6 7.Nxe4 dxe4 8.Qxe4 Bd7 is obviously fine for Black, and after 9.d4 (9.0-0 0-0-0 10.Bxc6 Bxc6 11.Qxe5 Qg6! is good for Black) he can choose between an ambitious continuation and an easy equalizer:

9...f5!?N (After 9...0-0-0 10.dxe5 Nxe5 11.Bxd7† Rxd7 12.0-0 Nxf3† 13.Qxf3 Qd5 14.Qxd5 a draw was agreed in Caldwell – Weiss, email 2008) 10.Qe2 e4 11.Bxc6 Bxc6 12.Ne5 My suggestion is:

12...Qe6!? Intending ...Bd6. (On the other hand, 12...Qxd4?! 13.Qh5† g6 14.Nxg6 hxg6 15.Qxg6† gives White at least a draw and possibly more.)

6...Qf6! 6...Qg5!? is also promising. The critical line continues: 7.Nxc6 Qxg2 8.Rf1 a6 9.Nxd5 axb5 10.Nxc7† Kd7 11.Nxa8 Kxc6 Black is an exchange down for the moment, but White’s king is in danger and the a8-knight might drop off at some point.

12.Qe2 Bc5! (12...Bh3 13.a4! is a draw according to my analysis) 13.d4 (13.a4 Re8!) In I. Schneider – Levushkina, Nuremberg 2011, Black could have kept a dangerous initiative with 13...Be6!N. I analysed this more deeply and found that it should eventually be a draw, but in a practical game the position is much more dangerous for White, so take your pick between this and the game continuation.   7.Nf3 Be6 8.0-0 Bd6

Black has a considerable development advantage, so White has to start creating threats if he doesn’t want to be on the wrong side of a miniature.   9.Re1 0-0! 9...Nxc3 10.dxc3 0-0 11.Bg5 Qg6= as in Van den Berg – Op den Kelder, Hengelo 2005, is a solid way to avoid sacrificing a pawn. The game continuation gives Black excellent compensation though.   10.Nxe4 dxe4 11.Bxc6 bxc6 12.Rxe4 Bd5 13.Re3

13...c5! Black gains some space and makes room for his bishop to remain on the long diagonal in case it is attacked.   14.d3 Rad8

14...a5!? intending ...a4 is an interesting alternative. Another logical continuation is 14...Rab8 15.Rb1 Bf4 16.Re1 Bxc1 17.Qxc1 Bxf3 18.gxf3 Qxf3 19.Qe3 Qc6. In all cases, Black has full compensation.   15.Qe1 Qg6 15...c4!? is also possible, but Gerhards keeps his focus on the kingside.

16.g3 Bb7 17.Nh4 It is not so easy for White to develop his queenside pieces and coordinate his position. An important point is that 17.Bd2? would allow the unpleasant 17...Bf4!.   17...Qg4 18.Qe2 Qg5 19.Ng2 Qd5 20.f3 Qd4 21.Qf2 Qf6 White still can’t develop his queenside and Black is ready to increase the pressure with ...Rfe8. Raijmaekers decides to bail out by returning the extra pawn before things get any worse.

22.Re1 Qxf3 23.Qxf3 Bxf3 24.Rb1 Rfe8 25.Bf4 Kf8 26.Bxd6† cxd6 In this correspondence game the players saw no reason to continue, but there is still a lot of play in the position, and in a practical game Black could try to exploit the superiority of his bishop over the knight. ½–½

4.d3 Another meek move. You don’t really need to remember anything special against it – just play 4...d5 and enjoy the fact that you are effectively playing as White. Although we could leave it at that, I cannot resist including a short game which I have often presented to my beginner students as a model example of how to carry out a direct attack. GAME 10 N. Kosolapov – Rashid Nezhmetdinov Kazan 1936   Nezhmetdinov never quite broke through to the elite level, but he was a gifted attacking player who is best known for his winning record (+3 =0 –1) over Mikhail Tal. Their game from Baku 1961 (and, to some extent, Moscow 1959) was remarkable, as Nezhmetdinov obliterated Tal in Tal’s own style, sacrificing material for a ferocious initiative and smashing through the defences – something which hardly any other player managed to achieve against Tal. I would recommend looking up those games on the database, but for now we will see how Nezhmetdinov destroys a less accomplished opponent.   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d3 d5 5.exd5 Nxd5 6.Be2 Bb4 7.Bd2 0-0  

8.Ne4?! The position can be classed as a reversed Steinitz Defence of the Ruy Lopez. White’s best continuation in that variation is: 8.Nxd5 Qxd5 9.Bxb4 Nxb4 10.a3 Otherwise Black can gain space with ...c5. 10...Nc6 11.0-0 Bg4 12.Re1 Rfe8 White’s extra tempo does not change the fact that he is slightly worse. See the game Capablanca – Thomas, Hastings 1919, for a model exploitation of the space advantage in such positions.   8...Be7 9.0-0 f5 10.Ng3

10...g5 11.Ne1 Nf4 12.f3? Black’s play has been exemplary and he already stood better, but this weakening move allows him to create a masterpiece.   12...Bc5† 13.Kh1 Rf6! 14.c3 Be6 15.Nc2 Rh6 16.Be3 g4!? Nezhmetdinov casually leaves his bishop hanging, as he has seen he can easily regain the piece while bringing his queen into the attack.   17.Bxc5 Qh4 18.Bg1 Qxg3 19.Ne1 Bd5 20.b4

20...Nh5 21.Bf2 Qf4! White miraculously escapes the mate after: 21...Qxh2† 22.Kxh2 Ng3† 23.Kxg3 f4† 24.Kxg4 Be6† 25.Kg5 Rg6†

26.Kh5 Kg7

Black threatens to deliver mate with ...Nd8, ...Rh6† and ...Nf7, but White has a fantastic defensive resource in 27.Qb3!! when Black must take a draw.   22.fxg4? Missing Black’s beautiful idea. 22.Qc1 was essential, although 22...g3 keeps White under serious pressure.

22...Qxh2†! Now it is time!   23.Kxh2 Ng3† 24.Kxg3 f4# A theoretically irrelevant game, but what a pleasure it has been to study it!

4.h3 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.h3 This move didn’t receive any serious tests until Giri ventured it in the rapid section of the 2014 London Classic, scoring a superb 2½/3 against Kramnik, Caruana and Adams! It has a slight advantage over 4.a3, as explained in the comment to move 10 below.

4...d5! All three of the aforementioned super-GMs preferred 4...Bb4, when 5.Bd3!? was Giri’s idea. Black should be fine here too, but I see no reason to refrain from our standard response.   5.exd5 After 5.Bb5 Nxe4 I don’t see why Black should be any worse off than in the similar 4.a3 variation examined earlier.   5...Nxd5 6.Bb5 Nxc3 7.bxc3 Bd6 8.0-0 0-0 9.d4 exd4 10.cxd4 We have reached a reversed Scotch Four Knights where White’s extra h2-h3 move has some value, as it prevents 10...Bg4, which would otherwise have been one of the main candidate moves. However, as we will see in the Scotch Four Knights section, Kramnik popularized a new concept for White involving h2-h3, Qf3 and Bf4, so we can do the same thing with reversed colours.

10...Bf5 I think this is the most flexible, although other move orders lead to similar play and may transpose.   There have not been many practical tests of this variation, but one such example continued: 10...Qf6 11.c3 (I think the best continuation would have been 11.Bd3N 11...h6 12.Re1 Bf5 13.Bxf5 Qxf5 14.c4, but after 14...b6= I see no problems for Black) 11...h6 12.Bd3 Bf5 The ‘Kramnik plan’ is in full swing. 13.Rb1 b6 14.Rb5 Bxd3 15.Qxd3 In Mammadova – Stetsko, Chakvi 2015, the best continuation would have been:

15...Rae8N 16.c4 a6 17.Rb2 Re7 Followed by ...Rfe8.   11.Re1 h6 Black can follow up with ...Qf6 or first ...Re8. We will follow an engine game for a few more moves.  

12.Bd3 Qd7!? It turns out that the queen has a second good square! 12...Qf6 is also fine of course.

13.c4 Rfe8 14.Bxf5 Rxe1† 15.Qxe1 Qxf5 16.Be3 b6 Black was fine in Stockfish – Houdini, Internet (blitz) 2014.

4.Be2 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Be2 This is a more serious move than it might appear, and has been used by such uncompromising players as Grischuk and Short. Still, it is easy to handle if we understand a thing or two about the Scotch Four Knights.

4...d5! 4...Bb4 can be met by 5.Nd5 with interesting play, while most other moves allow 5.d4!, with good chances of an advantage for White.   5.exd5 Nxd5 6.0-0 Nxc3 7.bxc3 7.dxc3? is a strategic blunder, and after 7...Qxd1 8.Rxd1 f6³ Black has the structural advantage of a reversed Exchange Spanish, without the loss of the bishop pair.   7...Bd6 7...Bc5?! 8.Bb5! puts unpleasant pressure on the e5-pawn.   7...e4 is playable, but after 8.Ne1 White has the plan of f2-f3, so I prefer the main line.   8.d4 0-0 Compared with the main line of the Scotch Four Knights, the bishop is on a less active square than b5, so Black is not obliged to exchange on d4 just yet.   9.Re1

9...h6! Once again we will use Kramnik’s set-up with colours reversed. Black takes control over g5 in order to continue with ...Bf5 and ...Qf6.   9...Bf5 10.Bd3! led to a slight plus for White in Grischuk – Fressinet, Riga 2013. The move Black would like to play is 10...Qf6N, but Fressinet may have been put off by 11.Bg5 Qg6 12.Nh4!? Qxg5 13.Nxf5. Black actually seems to be fine after 13...g6, but as a general rule I would prefer to have played ...h6 before putting the queen on f6, to eliminate Bg5 ideas altogether.   10.h3 10.Bb5 has been played a handful of times but it makes no sense to me. After 10...exd4 11.cxd4 Black can choose between 11...Bg4 12.c3 Qf6= with a reversed Scotch Four Knights and 11...Qf6N followed by ...Bf5 à la Kramnik.   10...Re8!? 10...Qf6N is fine too, but there is nothing wrong with developing the rook before proceeding with the main plan.

11.Rb1N 11.dxe5? Nxe5 12.Nxe5 Bxe5 was simply bad for White in Cook – Boettcher, email 2010.   11...b6 12.Bd3 exd4 13.cxd4 Rxe1† 14.Qxe1 Qe7=

The Italian Four Knights 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Bc4 This move arguably does White more harm that the ‘spike’ moves 4.h3 and 4.a3.

4...Nxe4! This is why!   5.0-0 5.Bxf7†? Kxf7 6.Nxe4 d5µ is rubbish for White.   5.Nxe4 d5 6.Bd3 gives Black a choice:

a) Old textbooks give 6...dxe4 7.Bxe4 Bd6 as better for Black, but it is actually not so bad for White after 8.0-0 0-0 9.Re1. I was surprised to see that Ivanchuk once played this with White against Karjakin, though it was only a blitz game. Black can choose between 9...Re8, with rough equality, and the more dynamic 9...Nb4!?, intending to meet either 10.c3 or 10.a3 with 10...f5!.   b) 6...Nb4!? is an interesting alternative: 7.Ng3 (7.Nxe5?! dxe4 8.Bxe4 Qd4³ was Birisan – Kovalenko, corr. 2014) 7...e4 8.Be2 exf3 9.Bxf3 Qe7† 10.Be2 Bg4 11.0-0 Bxe2 12.Nxe2 0-0-0 This has occurred in a few engine games, and the position seems dynamically balanced.   5...Nxc3 6.dxc3 f6! The most ambitious move, preventing Ng5 and solidifying the centre. If left in peace, Black will develop with ...Qe7, ...Bd7 and ...0-0-0.  

7.Nh4 Threatening Qh5† and preparing to launch the f-pawn into the attack, but Black has everything under control.   7...g6 8.f4 Qe7 9.b4 White has to spend a tempo anyway to defend against ...Qc5†, so at least this move gives him the option of advancing on the queenside in the event that Black castles on that flank.   9...d6 10.f5 Qg7 11.Be3 Ne7 12.Bd3

12...d5! 12...g5 is also good, and led to a victory for Black in Doettlaff – Kuhlmann, Germany 1992, but I like the text move even more.   13.Bc5N 13.Be2 c6 left White in serious trouble in Sroka – Borys, Pszczyna 2007; he has no attack and the f5-pawn is about to fall.   The text move seems like the best try, but Black gets a decisive advantage by smartly returning the extra pawn.   13...c6 14.fxg6 hxg6 15.Bxe7 Bxe7 16.Nxg6 Rh6 17.Nxe7 Kxe7–+ Black dominates the centre and will soon develop a crushing attack with ...Be6 and ...Rah8.

The Glek System 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.g3 d5! 4...Bc5 5.Bg2 d6 is another respectable system, but I suggest sticking to our usual strategy of playing ...d5 when White allows it.   5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3

7...Bc5! I think this move is the most accurate. The f2-pawn is vulnerable and the bishop can retreat to b6, offering a convenient shield for the b7-pawn, which often finds itself under fire from a rook on b1.   The other main move is 7...Bd6. This is playable too, but there are some pitfalls, for instance: 8.0-0 0-0 9.Rb1 Rb8 10.d4 Re8?! (10...Bg4 11.h3 Bh5 12.g4 Bg6 13.Nxe5! is good for White; 10...h6 is the most reliable move, but I was still not completely happy with it and prefer the positions in my main line) 11.Ng5! Bf5

12.f4! This was first played by the Ukrainian-Italian IM Elena Sedina in 2004 but, surprisingly, it has only been repeated in two games since. Even GM Sakaev overlooks this move when covering this line in his Petroff book (although it should be mentioned that he mainly focuses on 4...Bc5, and only gives 4...d5 as a secondary option for Black). I discovered the strength of Sedina’s move one ‘sunny night’ during the Tromso Olympiad (due to its northerly location, Tromso only gets dark for a few hours each night in the month of August, which was when the Olympiad took place) while assisting with the preparation of a member of the Danish national team. When I finally gave up trying to make this line work for Black, I had only a couple of hours to find a satisfactory alternative. I was pleasantly surprised to find that 7...Bc5 was not only theoretically reliable but also relatively easy to learn and understood.   8.0-0 0-0 9.Re1   The ‘fork trick’ with 9.Nxe5 Nxe5 10.d4 leads nowhere after 10...Bd6 11.dxe5 Bxe5; the computer calls it equal, but in a practical game Black would have some chances to exploit his superior pawn structure, and it is hardly surprising that no White player has ventured this.   9.d3 Sometimes White leaves the rook on f1, perhaps in the hope of supporting a future f2-f4. In that case I propose: 9...Bb6 Lokander gives 9...Bg4, which is fine of course, but it gives White the chance to chase the bishop with h3/g4 and follow up with Ng5-e4. This is hardly lethal of course, as Black can hit back with ...f5, but I prefer to keep things more flexible. The text move also fits in better with the style of the system I am advocating in the main line. 10.a4 10.Re1 Qf6 transposes to our main line. 10.Nd2 can be met by 10...Be6. At the risk of stating the obvious, there is no need to fear 11.Bxc6 bxc6; Black will follow up with ...Qd7 and ...f6, with an excellent game.

10...Re8! 11.Nd2 11.Re1 was played by Mamedyarov against Leko in a blitz game in 2014. Black should play 11...Qf6!N, which immediately transposes to 11.a4!? Re8! as given in the notes to Game 11. 11...Be6 Black can consider 11...Qf6!?N here too, inviting 12.Re1 with a transposition to the line mentioned in the note above. White can deviate with 12.Nc4 (12.Ne4 is met by 12...Qg6 followed by ...h6 and ...f5 with an excellent game) but 12...Na5! 13.Nxb6 axb6 sees Black giving up the bishop pair in order to improve his structure and target the weak a-pawn. 12.Qh5

12...Qd7!N Improving over Acs – Balog, Hungary 2008, where 12...g6?! unnecessarily gave White attacking chances after 13.Qh6 with ideas of Ne4-g5.

13.Ne4 f6 With a healthy game for Black.

9...Qf6! Black restrains the d2-d4 advance and on a good day he can hope to play ...Bg4, ...Re8 and ...e4.   9...Re8 allows 10.d4 exd4 11.Rxe8† Qxe8 12.cxd4 Bb6 13.c4 when White has a certain initiative.   10.d3 This is White’s most flexible and best move.   10.Rb1 merely encourages 10...Bb6, which Black would like to play anyway. After 11.d3 Black has more than one good option, as mentioned in the notes to Game 11 below.   10.Qe2 doesn’t strike me as particularly clever. A good reply is: 10...Bd7 (10...Bg4 is also fine, as mentioned in the notes to Game 11 below) 11.d3 e4! 12.dxe4 (12.Qxe4?? is skewered by ...Re8, which is why the bishop went to d7) 12...Qxc3 Black was better in Virovlansky – Pokrovski, Moscow 1999.

10...Bb6! Flexible play by Black. This is a move he’ll have to play anyway, so why not do it right now? It is useful to compare what happens if Black plays moves that were playable against 10.Rb1: 10...Bd7 is fine in the position with Rb1 and ...Bb6 interpolated, but now White can save time with 11.Be3! Bb6 12.Nd2² as in Tischbierek – Hebden, Guernsey 2012.   10...h6 allowed White to develop some pressure with 11.Be3 Bb6 12.Nd2 Bf5 13.Rb1 Rad8 14.c4 Bc8 15.Nb3 in Lu Shanglei – Sachdev, Jakarta 2012.   10...Bg4 11.h3 Bh5 12.g4 Bg6 is a trendy line but, instead of 13.Be3 Bb6, which transposes to Game 12, White can consider 13.g5!? or 13.Bg5!?, both of which lead to double-edged play. My proposed move order is intended to rule out such options.   After the text move there are two main branches that need to be studied:

a) 11.Bg5 Qd6 12.Nd2 Qg6 13.Be3 Ba5! 14.c4 b6 is Game 11, where I have inserted some more explanations about some earlier sidelines, as well as the rare but potentially dangerous 11.a4!?.   b) 11.Be3 Bg4 12.h3 Bh5 can be regarded as the main line, for the simple reason that Mamedyarov has played it with White (by transposition, after a slightly less accurate move order by Black) against a couple of strong opponents. 13.g4 Bg6 is normal, and now White’s main plan is to play a2-a4 and Nd2-c4 in order to challenge Black’s solid set-up on the queenside. As 14.a4 is met by the strong 14...e4!, White has to start with 14.Nd2 but this is not optimal either. Black plays 14...Qh4! followed by ...f5 with excellent counterplay, as shown in Game 12.   GAME 11  Lu Shanglei – Evgeniy Najer China 2013   1.e4 e5 2.Nc3 Nf6 3.g3 d5 4.exd5 Nxd5 5.Bg2 Nxc3 6.bxc3 Nc6 7.Nf3 Bc5 8.0-0 0-0 9.Re1 Qf6

10.d3 10.Qe2 This has been seen a few times in practice, but it gives Black a pleasant choice: 10...Bg4 10...Bd7!? is also good, as mentioned in the introductory comments above. Here I will show that the other bishop development is equally promising. 11.h3 Bh5 12.g4 Bg6 13.g5 13.d3 e4! was good for Black in Blimke – Matras, Polanczyk 2000. 13...Qd8 14.Nxe5 This position occurred in Bosak – Pytlik, email 2003. I am amazed that Black refrained from:

14...Nxe5N 15.Qxe5 Bd6 16.Qa5 Bxc2

Black has a dangerous attack against the weakened white king. A brief sample line continues: 17.Bxb7 Rb8 18.Bg2 Bd3 Black has great play for a pawn, with ...Rb5 and ...f6 being just a couple of ideas White must worry about.   Also harmless is: 10.Rb1 Bb6 White’s last move has only driven the bishop where it wants to go anyway. 11.d3 Re8 11...Bd7 is also playable, intending to finish development with ...Rad8. After 12.Bg5 Qg6 13.Qd2 h6 14.Be3 Rfe8 15.Rb5 Qf6 16.Rb2 Rad8 Black had fine play in Buljubasic – Nezirovic, Tuzla 2010.

12.Bg5 Qd6 13.Nd2 Be6 Given the time, Black would play ...Bd5, so a logical continuation is: 14.c4 Rab8!? 15.Be3 Bxe3 16.Rxe3 Nd4 17.Nf3 Nxf3† 18.Qxf3 b6 With equal chances, Sielaff – Stewart, corr. 2009.   10...Bb6! As explained in the introductory section, I think Black should insert this move before deciding whether to develop the c8-bishop or the f8-rook next.

11.Bg5 White wins a tempo in order to continue developing his pieces, but this move actually achieves nothing, other than proving that Black really doesn’t need to play ...h6 early in the game!   11.Be3 will be examined in the next game. Other alternatives are:   11.a3? must have been a mistype in the email (we are looking at a correspondence game played between two highly rated players), but the game serves as an example to show what Black is threatening if he gets a free move: 11...Bg4 12.h3 Bh5 13.g4 Bg6 14.Be3 e4µ Bortnik – Sage, email 2003. 11.a4!? This is a serious alternative. I believe the correct reaction is: 11...Re8! 11...h6 once again strikes me as a loss of time, and White can take advantage of it with: 12.Qe2 Re8 13.Nd2 In this system the optimal place for the white queen is, surprisingly, the f1-square, but the first player rarely has time to place it there without consequences. Here he can do so, precisely because of Black’s slow 11th move. 13...Bg4 (13...Be6 planning ...Bd5 would have been nice if it wasn’t for 14.Bxc6 bxc6 15.Ne4± followed by a4a5) 14.Qf1 Rad8 15.Nc4 White successfully implemented his ideal plans, and thus he stood better in Pridorozhni – Ganguly, St Petersburg 2014. 12.Bg5 Black’s previous move ensures that 12.Qe2 can be met by 12...e4!.

12...Qd6!N 12...Qg6 was seen in the only game from this position, but after 13.Qc1 Qh5 14.Be3 Bh3 15.Bh1 White had some pressure in Sazon – Pinho Rego, corr. 2015. The text move is better, and can be compared with our main game. 13.Nd2 Be6! The main advantage of the queen on d6 is that it overprotects the e5-pawn. This enables Black to develop the bishop on e6 without fearing a4-a5, which will divert a defender away from the important central pawn. Now Black is ready to play ...Bd5, neutralizing White’s strong bishop. 13...Qg6 14.Be3 Ba5 is similar to the main game, but here White can play 15.Ra3, showing another use of the a2a4 move. Black may not stand worse after 15...Be6, but it still feels like White has gained something. 14.c4 Stopping ...Bd5, but Black can regroup with:

14...Qd7 15.Nb3 f6 16.Bd2 Nd4 With a fine position. Still, it is clear that 11.a4 requires some care.   11...Qd6 Against a neutral move Black will play ...f6 and ...Be6, while the move chosen by White in the game gives Black an opportunity for something else.   12.Nd2

12...Qg6! 12...Be6!? is fine too, but the text move is stronger.   13.Be3 Ba5! 14.c4 b6 Black has solved all of his opening problems, and is ready to finish development and centralize his rooks.   15.Rb1

15...Bg4! 16.f3 Bd7 17.Rf1 Rae8 18.Nb3 Bb4 19.Bd2 Bd6 20.Be3 Ne7³ Exemplary play by Najer. The plan is ...c5 and ...Nf5, after which ...h5-h4 can be considered. White tries to stop this but he only worsens his position.   21.c5?! Nd5! 22.Qd2 22.Bd2 bxc5 is a safe extra pawn.

22...Nxe3 23.Qxe3 Be7 24.f4 Otherwise ...Bg5 would have been unpleasant.   24...Qh6 25.Kh1 exf4 26.Qxf4 Qxf4 27.Rxf4 bxc5 28.Bd5 Bd6 29.Rf2 Re5 30.Bc4 g6! Black continues to play excellently. His bishop pair dominates the board, and his last move prepares to put the king

on g7 in order to play ...f5.

31.Na5 Kg7 32.Kg2 f5 Already a pawn down and faced with the threat of ...f4, White makes a tactical oversight which loses an exchange.   33.Bb5? Rb8 34.Bxd7 Rxb1 35.Nc6 Ree1 36.Nxa7

36...c4! 37.d4 Rg1† 38.Kf3 Ra1 Black has played beautifully, and he converts his advantage without giving his opponent the slightest hint of counterplay.   39.c3 Rgf1 40.Rxf1 Rxf1† 41.Ke3 f4† 42.gxf4 Bxf4† 43.Ke2 Ra1 44.a4 Ra3 45.Bb5 Rxc3 46.a5 Rc2† 47.Kf3 c3 0–1  

GAME 12 Shakhriyar Mamedyarov – Dmitrij Jakovenko Tbilisi 2015   1.e4 e5 2.Nc3 Nf6 3.g3 d5 4.exd5 Nxd5 5.Bg2 Nxc3 6.bxc3 Nc6 7.Nf3 Bc5 8.0-0 0-0 9.Re1 Qf6 10.d3 Bb6! The game actually continued with 10...Bg4 11.h3 Bh5 12.Be3 Bb6, but I have changed the move order to emphasize the more accurate move order which reduces White’s options, as explained in the introductory section.   11.Be3 Bg4 12.h3 Bh5 We have arrived back at the position from the game. The pin on the knight is awkward for White, and Black is already threatening ...e4, which explains White’s next two moves.

13.g4 Bg6 14.Nd2 White has neutralized both the pin and the ...e4 plan, and is ready to play Nc4 and a2-a4, with dangerous play on the queenside. Black has to seek his chances on the other flank.

14...Qh4!„ Preparing ...f5.   15.a4 Continuing his plan as if nothing has happened – typical maximalist Mamedyarov play! Three alternatives also deserve attention:   15.Rb1?! f5 16.Qf3 was seen in Stevic – Brkic, Zadar 2010, when Black missed a good opportunity:

16...e4! 17.dxe4 If the queen moves, then ...exd3 followed by ...Rad8 puts White under heavy pressure. 17...Ne5 18.Qg3 Qxg3 19.fxg3 fxg4µ   15.Nf3 gives Black the option of repeating the position, but he can also play for a win with 15...Qe7 intending ...f5. The

only example in my database continued 16.Bxb6 axb6 when the correct continuation is: 17.Nxe5!N (I suspect that 17.Nd4? was a mouse slip, as after 17...Qc5 Black was clearly better in Lifshitz – Silva, email 2007) 17...Nxe5 18.d4

The position is complicated, but after the accurate 18...Ra5! 19.f4 (19.Bxb7? Rd8µ) 19...f6 my analysis indicates that Black is at least equal.   Just a couple of week after the main game, Mamedyarov reached the same position and deviated with: 15.Qf3 This is the top choice of the engines, but Black should not be worried. 15...f5!

16.Qg3 This is better than 16.gxf5?! when, in Mamedyarov – Fridman, Jurmala (rapid) 2015, Black should have played 16...Rxf5!N 17.Qg3 Qa4!³ followed by ...Raf8 with the more comfortable position.

16...Qxg3 17.fxg3 fxg4 18.hxg4

18...Ba5 19.Rab1 Bxc3 20.Rxb7 Nd4 21.Rxc7 Rac8 22.Bd5† Kh8 23.Rxc8 Rxc8 The moves so far have followed the engines’ top choices, but Black has still emerged with some initiative. 24.Bxd4 Bxd4† 25.Kh1 Rxc2 26.Nf3 h6 White is still under pressure and finally blunders.

27.Nxe5? Bxe5 28.Rxe5 Rc5 Black followed up with ...Bf7 to exploit the pin, and he eventually won in Pridorozhni – Landa, Kaliningrad 2015.  

15...f5 16.gxf5?! Mamedyarov probably overlooked Black’s next move.   16.a5N is offered by Roiz as an improvement, based on 16...Bxe3 17.Rxe3 fxg4 18.Qe2 gxh3 19.Rxh3 Qf6 20.a6 b6 21.d4! when White has compensation for the pawn. Indeed, Qc4† is a threat, but Black keeps a good position with:

21...Bf5! Hitting the rook while preparing to block the potential queen check. Best play continues 22.Ne4 Qe7 23.Re3 exd4 24.cxd4 Qf7 25.Ng5 Nxd4 26.Nxf7 Nxe2† 27.Rxe2 Kxf7 28.Bxa8 Rxa8 with a roughly equal endgame, although I think Black’s position is more pleasant; he has two pawns for the exchange, pawn majorities on both sides of the board and some weak white pawns to target.   16...Bh5! This strong move forces White to enter defensive mode. The idea is to prevent Qg4 while forcing White to make a

difficult decision about how to deal with the attack on his queen. Black will then take back on f5 and double rooks on the f-file.   17.Nf3! Qe7 Black still wants to take on f5 with the rook, which is why he avoided putting the queen on f6. Mamedyarov continues to offer strong resistance though, and with his next few moves he tries to take advantage of the placement of the queen on the e-file.   18.Bxb6 axb6 18...Rxf5!N 19.Be3 Raf8 would have posed more difficult problems. The game continuation allows White to escape.   19.d4 Rxf5 20.Qd3 Raf8 21.Nxe5 Nxe5 22.Rxe5 Rxe5 23.dxe5 Qxe5 24.Qd5†³/= White has avoided the worst. The remainder of the contest is an interesting endgame which you can find analysed by Roiz in ChessBase Magazine 165, but it does not have much do with our opening study.

24...Qxd5 25.Bxd5† Kh8 26.Bxb7 Rf5 27.Re1 Rc5 28.Re3 Kg8 29.Be4 Bf7 30.Bd3 Kf8 31.Bb5 Bg6 32.Bd3 Bxd3 33.Rxd3 Rc4 34.Kg2 Rxa4 35.Kg3 Ke7 36.h4 g6 37.f3 Ke6 38.Kh3 Rc4 39.Kg3 Ra4 40.Re3† Kf6 41.Rd3 Rc4 42.Re3 Kf5 43.Rd3 c6 44.Re3 b5 45.Rd3 Rc5 46.Re3 Re5 47.Rd3 Rd5 48.Re3 Rd2 49.f4 Kf6 50.Re5 Rd5 51.Re3 Rc5 52.Kf3 Rc4 53.Kg4 h6 54.Kg3 Kf5 55.Re5† Kf6 56.Re3 h5 57.Kf3 Kf5 58.Re5† Kf6 59.Re3 c5 60.Rd3 Ke6 61.Re3† Kd6 62.Rd3† Kc6 63.Ke3 Ra4 64.Rd8 Rc4 65.Rg8 Rxc3† 66.Kd2 Rh3 67.Rxg6† Kd5 68.Rg5† Kd4 69.Rxh5 b4 70.Rg5 c4 71.Rg1 Rxh4 72.f5 Rf4 73.Kc1 Rxf5 74.Kb2 Ke3 75.Re1† Kd2 76.Re4 Rc5 ½–½   Conclusion   In this chapter we examined several tries for White after 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6.   The Halloween Gambit is more dangerous than it looks, but Game 8 shows how Black can make a counter-sacrifice to get a decisive initiative. The notes also contain a couple of less complicated ways to get at least a slight advantage.   4.d3, 4.a3, 4.h3 and 4.Be2 should all be met by 4...d5. In each case Black gets easy play, and the ensuing positions often resemble a Scotch Four Knights with colours reversed.

  The Glek System with 4.g3 is also met by 4...d5, but some important decisions lie ahead. The main line continues 5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3, when I favour 7...Bc5!. Then after 8.0-0 0-0 9.Re1 Qf6 10.d3, the move order nuance 10...Bb6! gives Black good chances, with ...Bg4 being the likely follow-up.

1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6   The Belgrade Gambit 4.d4 exd4 5.Nd5 Nb4! page 89 6.Bc4 (6.Nxf6†? page 90; 6.Nxd4?! page 90) 6...Nbxd5 7.exd5 Bb4† 8.Bd2 Game 13, page 90 (8.Kf1!? page 90; 8.c3 page 91)   The Scotch Four Knights 4.d4 exd4 5.Nxd4 page 93 5...Bb4! 6.Nxc6 bxc6 7.Bd3 (7.Bg5?! page 93; 7.Qd4 page 93) 7...0-0 8.0-0 d5 (8...Re8 page 94) 9.exd5 cxd5 page 94 10.h3!? Game 14, page 96 10.Bg5 c6 11.Na4 Game 15, page 98 (11.Qf3 page 95; 11.Ne2 page 98)   The Spanish Four Knights 4.Bb5 Bb4 5.0-0 0-0 page 102 6.d3 (6.Bxc6 page 103) 6...Bxc3! (6...d6 page 104) 7.bxc3 d6 8.Bg5! Qe7! 9.Re1 a6! (9...Nd8 page 105) 10.Bxc6 Game 16, page 106 (10.Ba4 page 106; 10.Bc4 page 106)  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.    

How could Black have put

the young Tal under pressure? (page 90)  

How can the queen be diverted from

its powerful outpost? (page 91)

Find a strong move for Black. (page 99)  

Play a move à la Karpov. (page 106)  

How should Black unpin the f6-knight? Can you find a clear way to equalize? (page 93)  

How does Black avoid any problems? (page 97)

(page 106)  

And here, how should Black

unpin the knight? (page 106)

  1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6   In this chapter we will deal with the two most important options of 4.d4 and 4.Bb5. Just as in the previous chapter, my aim is to provide quality solutions without making things unnecessarily complicated. Take for instance the Spanish

Four Knights with 4.Bb5: despite being a respectable system, it remains a pretty rare occurrence, and I can recall several occasions when students of mine have been left pondering over their fourth move, trying to remember what on Earth we had prepared after 4...Nd4 (a strong but theoretically demanding variation) some afternoon in the club half a year ago. Such experiences taught me the importance of choosing simple enough systems that you won’t forget their main ideas many months (or years!) from the last time you studied them. I would go so far as to call this kind of pragmatic approach one of the key attributes of a successful practical player.

The Belgrade Gambit 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nd5 This gambit has been played by the young Tal among others, but it is rarely seen today, for good reasons. The most challenging reply is:

5...Nb4! 5...Be7 is a popular recommendation. It is solid enough, but the arising positions are only equal and too dry for my taste. 5...Nxe4?! 6.Qe2 f5 is playable, but it leads to unclear positions where Black has to know his theory. The text move is easier to learn and just a stronger move overall.   6.Bc4 This is the main line and the move you should expect to face. 6.Nxf6†? Qxf6 is simply bad for White. Black plays ...Bc5, ...d6 and ...0-0, with an extra pawn and a solid position.   6.Nxd4?! Nxe4 7.Nf5 c6 is not much fun for White either, as shown in the analysis in the notes to Game 13.   6...Nbxd5 7.exd5 Bb4† 8.Kf1 8.Bd2 Qe7† 9.Qe2 leads to mass exchanges, but Black can fight for more than equality with the following idea: 9...Bxd2† 10.Kxd2 Qxe2† 11.Kxe2

11...c5! 12.dxc6 bxc6 13.Nxd4 d5 14.Bb3 c5 Though it may not seem like much, Black is comfortably better, as shown in Game 13.   The text move is not the most obvious choice, but is has some tricky points to it, so it is important to remember a small detail:

8...0-0 9.Qxd4 h6 10.h4 Black’s only real problem is the excellent placement of the queen on d4. Once you appreciate this, the following solution makes perfect sense:   10...c5! One way or another, the queen will be driven away. Further details can be found in the notes to the game below.  

GAME 13  Rok Hrzica – Tamas Banusz Trieste 2008   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nd5 Nb4! 6.Bc4 6.Nxf6†? Qxf6 is poor for White, as mentioned earlier. A good example continued 7.Bc4 Bc5 8.0-0 d6 9.Ng5 0-0 10.e5 Qg6 11.a3 Nc6 12.Bd3 Bf5 and Black was a healthy pawn up in Van der Weide – Golod, Groningen 1995, and went on to win an excellently played game.   6.Nxd4?! Nxe4 7.Nf5 c6 8.Nxb4 Bxb4† 9.c3 Qf6 10.Qf3 10.Nxg7† Kf8! is excellent for Black, as Karolyi points out. 10...Nxc3 11.a3

11...Qe5†! 11...Ba5 12.Bd2 was much less clear in Tal – Averbakh, Riga 1954, a game in which the young Tal went on to score his first ever win over a grandmaster. 12.Ne3N After 12.Kd2? Ne4† 13.Kd3 d5 White already felt compelled to resign in Adamczyk – Schenning, email 1998. 12...Ba5 13.Bd2 Na4 Unlike the Tal – Averbakh game, the a5-bishop is defended, and Black should eventually prevail with his two extra pawns.

6...Nbxd5 7.exd5 Bb4† 8.Bd2 8.Nd2 is met by 8...Qe7†.   The first choice of the engines is 8.Kf1!? 0-0 9.Qxd4 h6 10.h4 as played in Galloway – Gray, Sunningdale 2007.

I agree with Bologan’s recommendation of 10...c5!N, in order to drive the queen away from its powerful outpost on d4: 11.Qd3 (11.dxc6 dxc6!) 11...d6 12.Bg5!? This is the only real attempt to cause problems, but Black has a good reply: 12...a6 13.a4 Bg4!µ Black is better as the white king is awkwardly placed.   Finally, White can make a real gambit out of the opening with: 8.c3 Qe7† 9.Be2 dxc3 10.0-0

White has a slight lead in development, but Black has a solid position and should be better with careful play. A good example continued: 10...0-0 11.a3 Ba5 12.b4 Bb6 13.Bg5 d6 14.Bd3 Bg4 15.h3 Bh5 16.Re1N 16.Qc2 was played in Evertsson – Johansson, Stockholm 1994, when Black could have obtained a near-decisive advantage with 16...Bxf3N. The text move is a logical attempt to improve White’s play, but Black can keep the upper hand with:

16...Qd7 17.Bxf6 gxf6 Black will exchange rooks along the e-file and the bishop will go to g6 to provide cover for the king. However, White does have a trick that should be considered: 18.Ne5!? fxe5 19.Bxh7† Kg7 19...Kxh7?? 20.Qxh5† Kg7 21.Qg5† Kh8 22.Qf6† Kh7 23.Re4 wins for White. 20.Qxh5 f5 21.g4

White is doing his best to cause problems, but Black remains in control after:

21...Qe8! 22.Qxe8 Raxe8 23.Bxf5 c6 24.Rac1 Bd4 25.dxc6 bxc6µ Black will improve his position with ...d5, then march his king to g5 and sacrifice the exchange on f5.

8...Qe7† 9.Qe2 Bxd2† 10.Kxd2 Qxe2† 11.Kxe2 As mentioned in the introductory section, Black can cause a surprising amount of problems with the following continuation:   11...c5! 12.dxc6 bxc6 13.Nxd4 d5 14.Bb3 c5

15.Ba4† This is met by powerful play by Black in the next few moves, but the alternatives do not bring White any joy either.   15.Nb5 Ba6 16.a4 0-0 followed by ...Rab8 and ...c4 is clearly better for Black.   15.Nc6! is the least of the evils. 15...Bd7N (The ambitious choice is 15...Kd7!?N, and after 16.Ne5† Black should sacrifice the f7-pawn and play energetically; objectively this is promising, although it is not to everyone’s taste to give up a pawn when there is a simpler route to a slight edge.) 16.Ne5 c4 17.Nxd7 Kxd7 18.Ba4† Kd6 Black has a small, safe edge. He has good central control, and will play on the queenside with ...Rab8 and ...Rhc8.   15...Ke7 16.Nf3 Rb8 17.Rhe1 Kd6 18.b3 Ne4! 19.Kd3

19...f5! 20.c4 d4!

This is even stronger than taking on f2 with check!   21.Re2 Nc3 22.Re5 Bb7 23.Ree1 23.Rae1 would be defeated in the same way.   23...Be4† 24.Kd2 Nxa4 25.bxa4 Rb2†–+ It is amazing how quickly Black has obtained an overwhelming position after the queen exchange. The rest is easy.  

26.Kd1 Rxf2 27.Ng5 Re8 28.Nxe4† Rxe4 29.Rg1 Ree2 0–1

The Scotch Four Knights 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nxd4 This is the Scotch Four Knights Variation – an almost perfect option for a beginner to learn as his first opening against 1.e4 e5. White gets a safe, weakness-free position with a slight lead in development. Black has several ways to equalize in the main lines, but the real challenge is to find an option offering sufficient possibilities to unbalance the game.

5...Bb4! 6.Nxc6 bxc6 7.Bd3 I have often seen 7.Bg5?! played at club level. This move allows Black to gain the upper hand with 7...h6 8.Bh4 Qe7! followed by ...g5 and snatching the e4-pawn.   7.Qd4 is played occasionally, with the idea of castling on the queenside. A good reply is: 7...Qe7 8.f3

8...Bc5 9.Qd3 a5!? One practical example continued: 10.Be3 Ba6 11.Qd2 Bxf1 12.Rxf1 0-0 13.0-0-0 Rfb8 14.Bxc5 Qxc5 15.Qd4 Qb4! Black was at least equal in Kantans – Bojkov, Germany 2014.   7...0-0 8.0-0

8...d5 The traditional move is also the strongest. 8...Re8 This move has become quite trendy recently, and I wondered if it might appeal to readers wanting to take some risks in order to increase their winning chances. Unfortunately, I doubt that Black has many of those after: 9.Qf3! 9.Bg5 h6 10.Bh4 Bxc3 11.bxc3 g5 12.Bg3 Nxe4 13.Qh5 d6 14.f4 Bf5 is sharp, but Black is doing fine. The flexible 9.Re1!? is a second potentially problematic move for us. I found nothing better for Black than 9...d6

10.Bg5! h6 11.Bh4 Qe7! planning ...Qe5 to break the pin. There are too few practical examples to make any definite conclusions, but I am not particularly keen on Black’s position.

9...Bd6 A weird-looking move, but it is one of the main ideas behind Black’s last move. 9...Rb8 10.Be3!? looks promising for White. 10.Bf4! Bxf4 11.Qxf4 d6 12.Rae1 Nd7 13.Qd2 In Naiditsch – Aronian, Baden-Baden 2015, White held a small edge without any risk.   9.exd5 9.e5?! is risky for White. After 9...Ng4 10.Bf4 f6! Black has superb piece play.   9...cxd5

10.Bg5 This is the old main line and a decent move of course, but the pin does not achieve much for White, and attention has recently shifted in another direction. 10.Re1?! was mentioned by Negi as a possible refinement of the Kramnik plan mentioned below, but 10...Bg4! is an excellent reply, when White must either meekly trade bishops or weaken his kingside with f2-f3.   10.h3!? This is Kramnik’s concept. The idea is to put the queen on f3 and keep the dark-squared bishop flexible. For instance, this piece will often go to f4 rather than g5. 10...Re8! Black gets ready to meet Qf3 with ...Qe7 in order to take control of the open e-file. A good alternative is 10...Rb8 11.Qf3, after which Anand has ventured both 11...Bb7 and 11...h6 against Karjakin and Adams respectively, with good results, albeit only in blitz games. Even at faster time limits, Anand’s stamp of approval is a good indication of the quality of these moves, so you could certainly investigate either or both options in more detail. However, as the main line is easy to understand and entirely unproblematic, I see no reason to spend any more time on them here.

11.Qf3 c6 12.Bf4 12.Bg5 h6 leads us to the same type of endgame as the bolded line below. 12...Qe7! According to Negi, this is the most logical antidote to Kramnik’s concept. Indeed, Black’s control over the e-file is a serious annoyance to White; see Game 14 for more details.   10...c6 10...Bxc3!? 11.bxc3 h6, followed by ...Qd6, has been played by Lasker, Capablanca, Alekhine, Pachman and Smyslov, so it cannot be too bad. I mention this as a possible way to unbalance the game and try to outplay your opponent. Still, the main line is far from a dead draw, so I will not take up valuable space analysing a second option.   11.Qf3 11.Na4 prepares to advance the c-pawn and play on the queenside. This move is examined in Game 15.   11...h6!

11...Be7 is not a bad move, but it is slightly passive and gives White some chances of obtaining a small edge. The text move forces an endgame where Black’s strong bishops fully compensate for his broken kingside structure.   12.Bxf6 12.Bh4 g5 13.Bg3 Bg4µ loses material for White. 12.Bf4 is inconsistent, and Black is more than okay after 12...Bd6 followed by ...c5, ...Be6 and ...Rb8. A model game showing how Black can exploit this kind of central superiority is Lutz – Yusupov, Munich 1992.   12...Qxf6 13.Qxf6 gxf6 14.Ne2

This position has been debated in a few hundred games, most of which have featured one of the following three scenarios:   (1) White transfers his knight to f5 and Black exchanges it, resulting in an opposite-coloured-bishop endgame with an extremely high probability of a draw.   (2) White does not want to allow an easy draw, so he postpones the knight manoeuvre to f5 and instead tries to strengthen his position with moves like b2-b3, c2-c3, f2-f4, g2-g3, and of course centralizing the rooks at some point. In this scenario Black may obtain counterplay with ...a5-a4, and can either put his bishop on e5 or double rooks on the e-file.   (3) White transfers his knight to f5 but Black refuses to exchange it. This is the way for Black to play if he is not satisfied with a draw. The bishop pair is a long-term asset, so Black doesn’t have to ‘prove something’, but can instead wait for the right opportunities to come by themselves.   14...Re8 14...Bd6 15.Nd4 c5 16.Nf5 Bxf5 17.Bxf5 has resulted in many quick draws.   Another Rubinstein masterpiece in the open games continued: 14...Rb8 15.Ng3 Ba5 16.Rab1 Re8 17.Nf5 Bxf5 18.Bxf5 Re2 19.g4 Rb4 20.Kg2 Rf4 21.Rbd1 h5 22.h3 Bb6 Black skilfully obtained an advantage in the opposite-colouredbishop endgame, although White managed to hang on for a draw in Spielmann – Rubinstein, Berlin 1926.   15.Nd4 Bb7!

Avoiding 15...Bd7 16.Bf5. This would not be too bad for Black, but he would obviously prefer to keep both bishops on the board.   16.Nf5 Bf8 17.Rae1

17...Re5! This is a nice way to utilize the doubled f-pawns.   18.Re3 Bc8 19.Nd4 Bd7 The position is still close to equal of course, but I would always favour the two bishops in such situations. Black has succeeded in driving the knight back from f5, and can continue improving his position in small steps.   20.Rfe1 c5 21.Nf3 Rxe3 22.fxe3 a5 23.b3 a4 24.Nd2 Rb8 25.c4 Be6 26.cxd5 Bxd5

We have been following Novikov – Sakaev, Krasnoyarsk 2007, which ended in a draw sixty moves later. Obviously both sides had some alternatives along the way, but I don’t see much value in going into too many details – the main thing is to understand that the early queen exchange does not have to mean a quick draw, and that there are many ways to improve Black’s position with a view to eventually exploiting the bishop pair.   We will now look at some illustrative games, beginning with Kramnik’s set-up with h2-h3 and Qf3.  GAME 14 David Navara – Chanda Sandipan Gibraltar 2013   1.e4 e5 2.Nf3 Nf6 3.Nc3 Nc6 4.d4 exd4 5.Nxd4 Bb4 6.Nxc6 bxc6 7.Bd3 d5 8.exd5 cxd5 9.0-0 0-0 10.h3 c6 11.Qf3 Re8 12.Bf4

12...Qe7! Negi recommends this for Black, and I agree. Black has absolute control over the e-file, so he has nothing to worry about. 12...Bd7 was played by Aronian against Kramnik in 2012, in the game that triggered the newfound popularity of this variation. Other moves are also playable, but the text is the most convincing.   13.Ne2 This is a typical move in White’s set-up.   13.a3 Bd6 14.Bxd6 Qxd6 This way White reduces Black’s control over the e-file but it comes at a price, as now Black gets counterplay along the b-file. 15.Qg3!? 15.Rfe1 Bd7 threatens the awkward ...Rxe1† followed by ...Rb8, when b2-b3 is impossible as the a3-pawn will hang. White can avoid this with 16.Qg3 Qxg3 17.fxg3, but after 17...c5 18.b3 Kf8 19.g4 h6 Black is at least

equal. 15...Qb8!? Exchanging on g3 is also fine of course. 16.Rae1 Be6 17.b3 17.b4 is met by 17...a5. 17...c5 18.Qh4 Qc7 The threat of ...c4 means that White has nothing better than to repeat with:

19.Nb5 Qb8 20.Nc3 Qc7 A draw was agreed here in Klein – Van Kampen, Wijk aan Zee 2015, although Black could have kept the game going with 20...Qd6!? 21.Nb5 Qb6 if he wished.

13...Bd6 13...Ne4 is a logical alternative, intending 14.Ng3 f5. A game continued: 15.c3 Bc5 16.Be3 Bxe3 17.Qxe3 Qf7

18.Bxe4 fxe4

19.f3 Ba6 20.Rf2 exf3 21.Qxf3 Qxf3 22.Rxf3 g6 23.Kf2 Kg7 Black’s better minor piece gave him a small edge in Lugovskoy – Buhmann, Vienna 2013.   14.c4 After something like 14.Bxd6 Qxd6 15.Qf4 Qxf4 16.Nxf4 Rb8 17.b3 Rb4 18.Ne2 c5 Black has some optical pressure, but of course the position is nothing more than equal really.   The text move gives White about as much as he can hope for in this variation. He wants to create a weakness on d5 and exchange some pieces via f4, but it is not enough for an advantage.

14...Bxf4 15.Qxf4 Ba6 16.Ng3 Qe5 17.Qxe5 Rxe5 The queen exchange was well-timed; White was not given time to play Nf5, and now his knight is slightly misplaced

on g3.   18.Rac1 Rb8 19.Rfd1 Bxc4 20.Bxc4 dxc4 21.Rd2! 21.Rxc4 Rxb2 22.Rxc6 Rxa2 23.Rc8† (23.Nh5 Ne8!) 23...Re8 leads to an endgame where White will be tortured for quite some time, with no guarantee of a draw at the end.   21...Rc5 Black’s extra pawn is not the most healthy, so a draw is to be expected, but it’s hardly a success for White’s opening system.   22.b3 h5 23.bxc4 Rb4 24.Rdc2 Ra5 25.Rd1 Raa4 26.Rdc1 The endgame still looks a bit unpleasant for White, but Navara defends coolly and the draw soon follows.

26...g6 27.Ne2 c5 28.Nf4 Ra3 29.Re1 Raa4 30.Rec1 Ra3 31.Re1 Raa4 ½–½   GAME 15 Hikaru Nakamura – Magnus Carlsen Zurich (blitz – 5.3) 2014   Despite the fast time control, these two great players produced a game packed with instructive content for us.   1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nxd4 Bb4 6.Nxc6 bxc6 7.Bd3 0-0 8.0-0 d5 9.exd5 cxd5 10.Bg5 c6 11.Na4 11.Qf3 was discussed in the introductory section.   11.Ne2 does not pose much of a threat. 11...Bd6 is one good reply, preparing ...c5 to deprive the knight of the d4-square without shutting our bishop out of play (there is also an incidental threat of ...Bxh2†!). 12.Ng3 h6 13.Bd2 Black can

seize the initiative with:

13...Ng4! 14.Be2 Qh4 15.Bxg4 Bxg4 Black gained full control with his pair of bishops in Steinitz – Zukertort, USA (2) 1886, and brought home the full point in exemplary style.

Nakamura’s move is a good positional choice. White hopes that the knight, in conjunction with a subsequent c2-c4 and Rc1, will help to control the c5-square and the queenside in general. I must admit I had a hard time finding a convincing way to meet this simple plan. Eventually I decided to investigate a new idea given by my analytical engines, which had hardly ever been tried according to my database. I became enthusiastic after analysing it, and this positive impression was reinforced when the World Champion played it in the present game.   11...Bd6! Apart from the general purpose of centralizing the bishop, this move prepares to meet c2-c4 with ...Bg4, which is quite an annoying move for White to face.

  12.c3!? This weird-looking move is the kind of thing Nakamura often tries in blitz games.   12.c4 is the obvious move but, in well over fifty games in my database, so far nobody has played the excellent move: 12...Bg4!N. The tactical point is that 13.Bxf6 Bxh2†! 14.Kh1 Bxd1 sees Black walk away with an extra pawn, as the knight on a4 hangs. Therefore White must settle for 13.f3 Bd7 (or 13...Be6), when Black has a comfortable game and can play ...Qc7 to provoke further kingside weaknesses at a moment of his choosing.   12.Re1 is a logical move which forces Black to modify his plan. A good reply is: 12...h6! (12...Bg4?? 13.Bxf6 Bxh2† just leaves Black a piece down after 14.Kf1 or 14.Kh1, as he will not be able to pick up the knight on a4) 13.Bh4 White has to maintain the pin, otherwise he will be slightly worse. 13...c5 14.c4 d4 Black has a good position and can start thinking about ...g5, as well as improving his position with normal developing moves.

15.b4!? This is a reasonable attempt to complicate the position, but after 15...cxb4 16.c5 Bc7 17.Rc1 Be6³ Black’s extra pawn gave him the upper hand in Yandemirov – Yemelin, Krasnoyarsk 2003.

12...Re8 The immediate 12...Bg4!? is also good, but Carlsen prefers to keep his options open for another move.   13.Bh4 Bg4! 14.f3 14.Qc2 avoids weakening the kingside, but the queen does not stand particularly well on this square. 14...h6 15.Rfe1 (15.c4?! is good for Black after 15...Rc8 or 15...d4) 15...Be6! Threatening ...Bxh2†!. 16.h3 (16.Bg3 c5³) 16...g5 17.Bg3 Nh5! The knight is heading for f4, which would not have been possible with the queen on d1!   14...Bh5 15.Qc2 Bg6³ White’s opening has not been a success at all. His knight is misplaced and ...Qc7 will provoke a further weakening of his kingside.

16.Rae1 Here is a sample line to demonstrate some of White’s problems: 16.Bxg6 hxg6 17.Bf2 Qc7 (17...Nd7 is also good) 18.g3 c5!? 19.Rac1 White would like to play b2-b4 to solve the problem of his misplaced knight, but Black can cut across this plan with:

19...Nd7! 20.b4 (20.Rfe1 Ne5 21.Kg2 Qc6µ) 20...c4! Black intends ...Ne5, ...Qc6 and ...a5, and will choose the right moment to break with ...d4.   16...Qc7! 17.g4!? 17.h3 Bxd3 18.Qxd3 Nh5!µ After 17.g3 Bxd3 18.Qxd3 Nd7 the bishop on h4 is out of play.   In the next phase of the game Nakamura will offer the h2-pawn constantly, but Carlsen will refuse to take it and give White the (admittedly distant) hope of potential counterplay along the h-file. Both players might have done things differently at a longer time control, but the general standard of play continues to be quite high.

17...Bxd3 18.Qxd3 Nd7 19.Qc2 Nf8 20.Rxe8 Rxe8 21.Re1 Rb8!? 21...Rxe1† 22.Bxe1 Ne6µ is certainly good for Black, but Carlsen’s choice also makes some sense. White’s rook has no entry points on the e-file, while Black remains in control of the position and can look for ways to combine threats in different parts of the board.   22.Kg2 Ne6 23.Bg3 c5 24.b3 Otherwise Black would have played ...c4 to isolate the knight.

24...c4! It is important to play this before White gets a chance to consolidate with c3-c4 and Nb2-d3.   25.Rd1 cxb3! 26.axb3 d4

Just as White created his first real threat of the game, Carlsen made it clear who is boss in the position.   27.Bxd6 Qxd6 28.cxd4 Nf4† 29.Kh1 Qd5?! As Mikhalevski points out, it was more accurate to continue with 29...Re8! 30.Nc5 Qc6!, threatening ...Re2. (30...Qxd4 is a bit too clever, and 31.Ne6! enables White to simplify) White can stay in the game with 31.d5! Nxd5, but he remains under pressure due to his weakened kingside.

30.Qe4! This queen centralization avoids the worst of White’s difficulties.   30...Qxb3 31.Re1? 31.Nc3 Qxc3 32.Qxf4 is given as ‘0.00’ by the ruthless engines; Black is unable to exploit White’s airy kingside due to the passed d-pawn and his own weak back rank.   31...g5! 32.Nc5 Qa2! 33.Rg1 Nh3! Magnus plays with fantastic accuracy in this crucial phase of the game.   33...Rb2 is met by 34.Qe8† Kg7 35.Qe5† f6 36.Qe7† Kg6 37.Qe8† Qf7 38.Qe4† with a draw.

34.Nd3 34.Rf1 Nf2† wins.   34...Nxg1 35.Qe5 Nxf3! 35...Rb6 is also winning but Carlsen’s solution is cleaner.   36.Qxb8† Kg7 37.Ne1 Qa1! 38.Kg2 Nxe1† 39.Kf2 Nd3† 40.Ke3 Qc3! An excellent and instructive game by Carlsen, whose handling of the opening deserves to be remembered. 0–1

The Spanish Four Knights 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Bb5 As I mentioned in the chapter introduction, 4...Nd4 is theoretically healthy for Black, but it requires you to learn a lot of sharp variations, which is not ideal against an opening you may only face once or twice per year. That is why I prefer to recommend the other main continuation, which is solid and easier to learn.

4...Bb4 Playing symmetrically may not seem inspiring, but don’t worry – we will soon get a chance to unbalance the game in one way or another.   5.0-0 Lokander points out that White can head for a boring position with 5.Nd5, but I am not at all bothered by this. After 5...Nxd5 6.exd5 e4 7.dxc6 dxc6 8.Bc4 exf3 9.Qxf3 Qe7† White has abandoned all hope of an opening advantage, and there are enough pieces on the board for Black to try and squeeze out more than half a point from the position.   5...0-0 This is the main starting position of the 4...Bb4 Four Knights. This system produces rich strategic positions, full of subtleties and hidden nuances, which are even beyond the scope of modern analysis engines.   5...d6?! is an inaccurate move order because of 6.Nd5! Ba5 7.d4!, as Botvinnik pointed out.

6.d3 6.Bxc6 dxc6 can be compared with the Delayed Exchange Variation of the Spanish, as covered later in Chapter 8. White may try:   a) 7.Nxe5 gives up all hope of an advantage, and Black’s only challenge is how to play for a win. 7...Re8 8.Nd3 Bxc3 (8...Ba5?! 9.f3 Nh5 intending ...f5 led to an eventual win for Black in Von Gottschall – Rubinstein, Hannover 1926, but Black’s pawn sacrifice is objectively dubious) 9.dxc3 Nxe4 10.Qf3 Nd6

11.Be3 (11.Bf4! is the accurate way to maintain equality) 11...Bf5 12.Rfe1 Qh4 13.Qf4 Qh5 14.Bd4 f6 15.f3 Qg6 16.Qg3 b6 17.Qxg6 Bxg6 18.Bf2 Nc4³ Little by little, White was outplayed in Koenig – Rubinstein, Vienna 1922.   b) 7.d3 Bxc3!? I like the plan used by Korchnoi in the following game. (7...Nd7 8.Ne2 Re8 9.Ng3 Nf8 is a reliable setup which we will encounter later in the section on the Spanish) 8.bxc3 Bg4! 9.h3 Bxf3 10.Qxf3 Qd6 The last two

bishop exchanges introduce certain imbalances to the position, which are not unfavourable for Black.

11.Qe3 Rfe8 12.Kh1 Re7 13.g4 Nd7 14.f4 exf4 Black obtained a good position by placing his pawns on dark squares with ...b6, ...c5 and ...f6 in Kasparian – Korchnoi, Moscow 1952.

6...Bxc3! Here is one of the important nuances I mentioned at the start of this section. If Black intends to go for a ‘Metger regrouping’ (read on and you will soon understand what I mean by that!), it is best not to delay this move.   6...d6 This has been the more popular choice, but it gives White the chance to play: 7.Ne2 7.Bg5 Ne7! has proven to be a reliable choice for Black, while he also has the option of transposing to our main line with 7...Bxc3 8.bxc3 Qe7.

7...Ne7 8.c3 Ba5 9.Ng3 When Anand approves this continuation for White and uses it to crush one of the world’s top players, it means that Black has to run to the lab and create an antidote as quickly as possible. 9...Ng6 9...c6! is more accurate, as it reduces White’s options with his bishop. After 10.Ba4 Ng6 11.d4 Bb6!? followed by ...d5 Black is on the path towards equality, although he may have to suffer a bit along the way.

10.d4 Bb6 11.Re1 c6 12.Bd3 If Black had played 9...c6! White wouldn’t have been able to put the bishop on this square and thus reserve c2 for his queen without losing a tempo. 12...Re8 13.h3 h6 14.Be3 Be6 15.Qc2 Qc7 16.a3 Mikhalevski remarks: “A subtle move. If I understand it correctly Anand wants to play 17.c4 with the idea to meet 17...c5 with 18.d5 and subsequent b2-b4. That’s where 16.a3 is useful.” 16...a5 17.c4 a4 18.Red1 exd4 19.Bxd4 Ne5 20.Be2 Bc5 21.Rd2 Nfd7 22.Rad1 Red8?! 23.Nh4! Bxd4 24.Rxd4 c5!? 25.Rxd6 Nc6 26.Nhf5 Nd4 27.Qd2 Ne5

28.Rd5! White went on to win a beautiful game in Anand – Mamedyarov, Shamkir 2015. Obviously Black’s play can be improved, e.g. at move 9, but I think it makes more sense to exchange on c3 and rule out altogether the plan used by Anand.   7.bxc3 d6 Black can start with 7...Qe7 with the idea to play ...h6 next, but after 8.Bg5 we will soon transpose to the main lines.   8.Bg5! This pin is one of the main strategic characteristics of the position. Black cannot easily get rid of it, as ...h6 followed by ...g5 weakens the kingside considerably, and also runs the risk of being met by an effective piece sacrifice on g5.  

8...Qe7! This is the starting point of a manoeuvre known as the Metger Unpin. The idea is to free the d8-square for the c6knight in order to continue with ...Nd8-e6, which may be followed by ...Nf4-g6 if necessary, finally forcing White’s bishop to release the pin.   9.Re1 Black requires several moves to realize his plan, so White gets ready for central play with d3-d4.   9.d4 is a rare move order. I propose: 9...a6!? 10.Bd3 (10.Ba4 h6 11.Bh4 Nd8 and 10.Bxc6 bxc6 resemble the main lines, and may transpose if Re1 is played at some point) 10...h6 11.Bh4 Bg4! Black has great counterplay.

9...a6! This important move considerably improves Black’s chances of implementing the Metger manoeuvre under favourable conditions.   9...Nd8 10.d4 Ne6 This has been the traditional continuation, but White has a clever way to meet it. 11.Bd2! 11.Bh4 is met by 11...Nf4 followed by ...Ng6. 11.Bc1 allows 11...c5 12.Bf1 Rd8, although 13.g3 may well be slightly better for White anyway.

11...Rd8 11...c5 12.Bf1! comes with the threat of dxe5, which cannot be parried with ...Rd8 as would have been Black’s choice with the bishop on c1. 12.Bd3 Nf8 13.Nh4!² White follows up with Qf3, as in Cheparinov – Jakovenko, Jermuk 2009.   10.Bxc6 The alternatives have never been well analysed, but you can find 10.Bc4 and 10.Ba4 discussed in the notes to the game below.   10...bxc6 11.d4 This is assessed as slightly better for White in ECO, but Karpov showed that this is not the case after:

11...a5!„ In the illustrative game below, we will see how Black can obtain excellent play on the light squares, especially c4.   GAME 16  Hou Yifan – Anatoly Karpov Sanya City (1) 2010   1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.Nc3 Bb4 5.0-0 0-0 6.d3 Bxc3 7.bxc3 d6 8.Bg5 Qe7 9.Re1 a6! 10.Bxc6 This leads to a position with knights and opposite-coloured bishops, where Black can utilize the e6-square for his queen in order to get rid of the annoying pin. Can White do better by retreating the bishop?   10.Ba4 This allows 10...Nd8! when the bishop is unable to return to f1 after d2-d4; this has been known since the game Schlechter – Tarrasch, Cologne (5) 1911. A possible continuation is: 11.d4!?N Everything else is met by ...Ne6, and if Bh4 then ...Nf4-g6 will break the pin. The text move seems to me to be the only critical move, but 11...Ne6 12.Bc1 b5! 13.Bb3 Bb7 is excellent for Black.   10.Bc4 This move keeps the bishop closer to the centre, but it has other drawbacks. 10...h6 11.Bh4 Na5 12.Bb3 g5! White faces a tough dilemma: sacrifice a piece for questionable compensation, or allow Black to build a strong attack.

13.Bg3 13.Nxg5 hxg5 14.Bxg5 must always be considered in these situations, but Black can eliminate one bishop and consolidate on the kingside with: 14...Nxb3 15.axb3 Qe6 16.Re3 Nh7 17.Rg3 Kh8 18.Qh5 f6 The queen will defend everything from f7, so White is in trouble. 13...Bg4 14.h3 If 14.d4 c5! the b3-bishop starts to feel uncomfortable.

14...Bh5„ Black’s ideas include ...Nd7 and ...f5. The pawn is currently pinned, but Black can get around this either by exchanging on b3 or by moving his king to g7. In the meantime, d3-d4 can still be met by ...c5!.   10...bxc6 11.d4 If White does not take the opportunity to advance in the centre, she won’t have the slightest claim for an edge. This move comes with a downside though, and Karpov doesn’t need to be asked twice to take advantage of it.   11...a5! A great move, after which Black is already at least equal.   12.Rb1 Qe6 13.a3 Nd7 14.Nh4 After something like 14.h3 f6 White can stop the ...Nb6-c4 idea with 15.Be3, but Black can prepare it again with 15...Ba6 followed by putting a rook on b8 before proceeding with ...Nb6. Also 15...Qa2!? is possible – but not 15...Nb6? 16.dxe5 followed by Bxb6 and White wins a pawn.   14...f6 15.Bc1 Nb6 16.Nf5 Nc4 17.Ne3 Hou Yifan’s manoeuvre with Nh4-f5-e3 doesn’t make much of an impression, but at least she exchanges the strong black knight, leading to a position with opposite-coloured bishops. Unfortunately for the world’s best female player, Karpov is arguably the greatest master in chess history when it comes to positions with this particular piece distribution. For more about this theme, see the relevant chapter in Mihail Marin’s excellent book Learn from the Legends.

17...Ba6 18.f4 Nxe3 19.Bxe3 Qc4 White’s weaknesses on the queenside are accessible to Black’s pieces, while the same cannot be said about the other side.   20.Qd2 Rae8 21.fxe5 fxe5 22.Bf2 h6 23.h3 Rf7 24.Bg3 Kh7 25.Kh2 Bb5 26.Rbd1 Re6 27.Ra1 Ba4 27...Rfe7! is a better way to provoke the exchange on e5 while preventing the possibility mentioned in the next note.

28.dxe5 28.Qd3! Bb5 29.dxe5 dxe5 30.Qxc4 Bxc4 31.Rad1 would have been equal, as Black does not have the chance to take over the d-file in the way he does in the game.   28...dxe5 29.Rac1 Qc5 30.Qe3 Qxe3 31.Rxe3 Rd7 Now Karpov takes over the d-file and beautifully increases his advantage.   32.Rd3 Rdd6 33.Kg1 Bb5 34.Rdd1

34...Be2! 35.Re1 Rd2 36.Bf2 Bb5 37.Be3 Rd7 38.Rcd1 Red6 39.Rxd6 It is understandable that White wished to trade a pair of rooks, but 39.Rb1! would have been more resilient.  

39...cxd6! Improving the pawn structure is more important than controlling the d-file. Another important point is that Black’s bishop controls the f1-square, so White’s remaining rook will stay passive.   40.Bb6 a4 41.Be3 c5 This move restricts White’s bishop even further. Thanks to the ...cxd6 move, Black only has one weak pawn, on d6, which can easily be defended by the rook and later the king. White, on the other hand, has serious weaknesses on e4 and c2.

42.Kf2 Rf7† 43.Kg3 Rf6 Karpov activates his rook and cuts off the white king from the centre, while keeping the d6-pawn securely defended. The next step will be to bring the king to e6, which will allow the rook to roam free and start harassing White’s queenside pawns.   44.Bd2 Kg8 45.Re3 Bc6 46.Re2? Passive defence with 46.Re1 Kf7 47.Bc1 Ke6 48.Bd2 would have forced Karpov to work to improve his position.   46...Rf1–+ Now the rook penetrates and it will soon be over.

47.Be1 Kf7 48.c4 Ke6 49.Bd2 49.Bc3 Rf4 picks up the e4-pawn.   49...Ra1 50.Re3 Ra2 51.Kf3 Rxc2 52.Bc3 Bb7 53.g3 Ba6 54.Be1 Bxc4 55.Rc3 Be2† 56.Ke3 Ra2 57.Bd2 Bf1 58.h4 Bh3 59.Bc1 Bf1 60.Bd2 Kd7 61.Bc1 Kc6 62.Kf3 Be2† 63.Ke3 Bh5 64.Bd2 Bg4 65.Kd3 Kb5 66.Ke3

White may have lost on time, or she may have resigned after spotting that 66...c4 would be zugzwang. 0–1 Conclusion   The Belgrade Gambit with 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nd5 is a rare guest today in all levels of play. 5...Nb4! is both simple and strong, and you only need to familiarize yourself with a few details as explained in Game

13.   The Scotch Four Knights is popular at club level, so a deeper study of the introductory remarks together with Games 14 and 15 and their notes will certainly pay off. Generally speaking, Black’s central control offers him good chances, and you can play for a win even in those lines where White aims for simplifications.   The Spanish Four Knights contains many nuances, but I have mapped out a clear path to a healthy position. The first important point is to play 6...Bxc3!, cutting out the Ne2 plan which Anand used to defeat Mamedyarov. The next step is to begin the Metger unpinning manoeuvre with ...Qe7, not forgetting to flick in 9...a6!, when Black can modify his plan depending on what White does with his bishop. Exchanging on c6 was reckoned to be the most challenging option, but Karpov’s 11...a5! gives Black fine prospects, as shown in Game 16.

1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4   On the Way to the Main Line 4...Nf6 5.Nxc6 bxc6 Vallejo’s 6.Qe2!? page 113 (6.Bd3 page 113) Sidelines after 6.e5 6.e5 Qe7 7.Qe2 Nd5 page 116 8.Nd2 Game 17, page 117 8.c4 Nb6 9.b3 Game 18, page 120 (9.Nd2 page 120)   The Main Line 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 page 124 14.Qf3 dxc4 15.Rc1?! Game 19, page 128 (15.Be2 page 128) 14.Qh4 dxc4 page 126 15.Be2 Nd5 16.Bd4?! Game 20, page 131 16.Bxc4 Game 21, page 132 15.Rc1 Qg6! 16.Bb4 Game 22, page 135 16.g3!? Game 23, page 139  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up

to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

Is 6...d5 playable? (page 114)  

How should Black continue? (page 131)  

Can you find a strong move for Black

in this sharp position? (page 123)

Which is Black’s best here? (page 135)  

Can you find Karjakin’s amazing concept? (page 125)  

Can you reproduce Kasimdzhanov’s prep? (page 136)  

What was Ponomariov’s flexible solution? Can Black profit from the white queen (page 139)

not being on h4? (page 128)   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 It was around the end of 2011 when Jacob Aagaard asked me to prepare some ideas in the Scotch for him to use as White. I soon became fascinated by this opening which had been around for so much time, but still was relatively unexplored. Adding the fact that most chess engines at the time had virtually no clue as to what was going on, I had to use all the analysis techniques I was aware of (creating endless engine tournaments, analysing with the IDeA tool, using

Monte Carlo analysis...) and was proud to discovered a few interesting novelties. To my disappointment, they did not remain novelties for long, as many bright chess minds around the world were also working on the same opening and had the chance to use these ideas in real games. Still, those “some ideas” developed into quite an extensive file which Jacob put to good use. To take one example, in the last round of the 2012 Scottish Championship against GM Hebden, Jacob obtained a slightly better position before offering a ‘draw gambit’. Mark declined and the rest is history: Jacob won a good game and became Scottish Champion.   Four years of analysing this opening (as well as playing it myself, though mainly at correspondence) taught me that the Scotch is a pretty annoying opening to face – especially over the board. To put it simply, White intends to use his halfmove advantage to create a slight weakening in Black’s pawn structure and then make sure (or hope!) that nothing else happens until late in the game. This long-term strategy is something that even the strongest chess engines fail to appreciate correctly, instead indicating that Black is equal in a huge number of positions where he really is not. Realizing this, many White players started reviving variations which used to be considering unambitious, but which actually force Black to defend pretty accurately for a long time with few prospects of winning the game. I checked a lot of sources when writing this chapter, and was surprised to find that several authors had underestimated White’s prospects, often proposing equality in slightly dubious lines. Take a look at Negi’s The Modern Scotch Opening DVD and Shaw’s upcoming Playing 1.e4 – Caro-Kann, 1...e5 & Minor Lines to see how much trouble this opening can cause.   Considering all the above, I had to make a crucial decision over which line to propose for Black. There were many serious candidates, but all of them had flaws. Many of them contained a lot of untried possibilities for White which needed to be taken into account and analysed carefully. Another common characteristic was that if White wanted to play safe, Black could no nothing about it and had to settle for a draw. Finally, I decided that the best line to propose to my readers would be what I call “the Professional Variation”. This involves a series of direct, forcing moves, with the main line leading to a drawn endgame. The reliability of Black’s system has been proven in several super-GM games over the past few years, and this goes a long way towards explaining the drop in popularity of the Scotch among top players.  

The Professional Variation   4...Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5

I call this the Professional Variation because it is mainly used by super-GMs against super-GMs, when drawing with Black is just another day in the office. I always thought that the usual ambitious club player would prefer something less worked-out-to-a-draw, but then I realized that this variation has some obvious advantages over others:   1) The initial moves are forcing and easy to understand, so it is unlikely that you will forget an important move or be hit with a huge novelty. 2) If White doesn’t force us into one of the few drawing endgames, Black will have a pleasant game, rich in active possibilities. 3) If White goes for the endgame in Game 22, Black will have to make some cool moves to reach a fortress – something that is always pleasant to execute over the board. 4) The endgame with rook and opposite-coloured bishops (see Game 23) is equal but far from dead – you can still try to outplay your opponent. 5) Having a super-safe line in your repertoire always gives you an air of confidence. And finally... 6) Don’t worry, I’ll give you a simpler alternative anyway. I am a tournament player as well, so I understand that sometimes you might just not want to risk allowing your opponent to memorize his way to a draw.

On the Way to the Main Line 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 4...Bc5 is the other main option, but my recommended system is easier to learn, with fewer lines to memorize. I also consider the resulting positions to be generally richer and more interesting.  

5.Nxc6 As mentioned earlier, the main purpose of the Scotch Game is to inflict some structural damage on Black. The drawback is the time it costs him; once Black recaptures, he will have one more piece in play than White.   5.Nf5? is met strongly by 5...d5.   5.Nc3 Bb4 transposes to the Scotch Four Knights, as examined in the previous chapter.   5...bxc6 The usual move is 6.e5, and we will discuss it immediately after the following brief diversion.  

Vallejo’s 6.Qe2!?   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.Qe2!? This weird-looking move is advocated by Paco Vallejo in his Chess24 lectures on the Scotch. I would also like to mention a playable but unambitious sideline, which is not really important enough to warrant its own section:   6.Bd3 In his DVD entitled The Deadly Scotch, Andrew Martin points out that one of the good things about the Scotch is that White can choose, if he wishes, to kill the game and make it too dry for Black to win. I guess that’s what the title “Deadly” refers to. Joking aside, let’s see what happens if White aims for that type of game. 6...d5 7.exd5

7.e5?! Ng4 8.0-0 Bc5 9.Bf4 gives Black more than one good reply. My preference is: 9...f6!? (9...g5 is theoretically better for Black, but requires some precise knowledge) 10.h3 (10.exf6 0-0!) 10...Nxe5 11.Bxe5 fxe5 12.Qh5† Kf8 13.Qxe5 Qf6! (13...Bd6!? is also possible) A trap to remember is 14.Qxc7? (14.Qxf6† gxf6³) 14...Bd6! 15.Qa5 Bxh3! and Black was already winning in Couspeyre – Razumov, email 2011. 7...cxd5

The game is equal and relatively dry, but there is ample opportunity to outplay an opponent with an asymmetrical pawn structure and board full of pieces. You don’t need any special preparation to play such a position, so I will just present a couple of examples in which White appeared to be intent on drawing the game – a strategy which often gets punished in practice. 8.Bb5† 8.0-0 Be7 9.c4 0-0 10.cxd5 Nxd5 11.Be4 Be6 12.Nc3 Nxc3 13.bxc3 Qxd1 14.Rxd1 Rad8 15.Be3 c5 16.Bf3 Rxd1† 17.Rxd1 Rb8 18.h3 Rb2 19.Bd5 Bxd5 20.Rxd5 Rxa2 Black won a pawn and later the game in Nimzowitsch – Tarrasch, San Sebastian 1911. 8...Bd7 9.Bxd7† Qxd7 10.0-0 Be7 11.Nd2 0-0 12.b3 a5 13.a4 Bb4 14.Bb2

14...Ra6! A nice way to activate the rook while guarding f6. Black has excellent central control and Adams soon takes over. 15.Nf3 Re8 16.Qd3 Rd6 17.Rad1 c6 18.Ne5 Qc8 19.Qf3 Rde6 20.Nd3 Bd6 21.h3 Qb8 22.Rfe1 Ne4 23.Qg4 Bh2† 24.Kf1 Rg6 25.Qf5 Ree6 26.Qf3 Rgf6 0–1 Mellado Trivino – Adams, France 2001.   Let’s return to Vallejo’s 6.Qe2!?.

6...Bc5!? Black simply wants to castle and play ...Re8, asking White why on Earth he chose the awkward e2-square for his queen. I like this simple approach very much, although I would like to mention a few alternatives; the first is an example of what to avoid, but the other two seem quite promising.  

The following recent example shows the venom in White’s set-up: 6...Qe7 7.Nc3 (7.e5 would lead back to the main lines, but White does not have to be so cooperative) 7...Qe6 8.a3!? (The more popular 8.Bd2 is also quite critical) 8...a5 9.Bf4 Bc5 10.0-0-0 0-0 11.Bxc7 Re8 12.e5 White had the upper hand and eventually won a wonderfully played opposite-coloured-bishop endgame in Andreikin – Kramnik, Baku (rapid – 3.4) 2015.   6...a5!? plans ...Ba6, effectively forcing 7.e5 Nd5. This must surely be fine for Black as well, but I like my main suggestion even more.   6...d5!? is an intriguing attempt to seize the initiative: 7.exd5† (7.Nc3?! Be7 was simply better for Black in Lerner – Mikhalevski, Givataim 2000, as there was no sense in the queen’s position on e2) 7...Be7 8.dxc6 0-0 The engines call this equal, but my good friend Sune Berg Hansen told me he thinks Black can get a fantastic attack for the two pawns. I will leave this for the reader to investigate.   7.e5 One of the advantages of Black’s last move is that the plan involving 7.Nc3 and long castling is no longer good because of 7...0-0 followed by a quick ...Re8 and ...d5.   7...Nd5 8.c4 Nb6 9.Qg4 White has to do something quickly, otherwise Black castles and continues with ...Re8 and ...Qe7.  

9...g6 Yes, this weakens the kingside slightly, but Black is well ahead in development and is poised to gain additional time with ...d5.   10.Bg5 Be7 11.Bh6 d5 12.Qf3 12.Qg3 is worse, as after 12...Nxc4 13.Bxc4 dxc4 14.0-0 Qd3 15.Qf4 Be6 Black will play ...0-0-0 with a promising initiative.   12...Bg5 12...Nd7!?N is also good according to the engines.

13.Bg7 Rg8 14.Bf6 Bxf6 15.exf6 Bf5! Improving over 15...Be6 16.Nc3 Qd6 17.Be2 0-0-0 which was okay for Black, but slightly easier to play for White in Vallejo Pons – Jakovenko, Khanty-Mansiysk 2013.   16.Nc3 Qxf6 17.cxd5 So far we have been following Das – Pruijssers, Dieren 2014, where Black inserted a queen check before recapturing on d5. Instead I recommend the improvement:

17...cxd5!N 18.Nxd5 Qxb2 19.Bb5† Kf8 20.0-0 Qxb5 21.Nxc7 Qd3 22.Qxd3 Bxd3 23.Nxa8 Nxa8 Black has some winning chances with two minor pieces against a rook.  

Sidelines after 6.e5

  1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5

6...Qe7 6...Nd5?! 7.c4 followed by Bd3 is a dream position for White.   6...Ne4 is a respectable sideline, but I found a nice antidote: 7.Be3! Intending Nd2 to exchange the active knight. I shared this idea with Jacob, who used it to good effect. 7...Qe7 8.Qd4! Qb4† 9.Qxb4 Bxb4† 10.Nd2 Bxd2† 11.Bxd2 Nxd2 12.Kxd2 White has a typical Scotch endgame advantage due to his better structure. 12...Rb8 13.b3 d6 14.Re1 Be6 15.Bd3 Kd7 16.f4 c5 17.Rhf1 In Aagaard – Hebden, Glasgow 2012, White went on to win a nice game, and with it the title of Scottish Champion.   7.Qe2 Nd5 8.c4 Moves like 8.h4 f6 (8...g6 9.Bg5 is a bit awkward for Black) 9.c4 Ba6, and 8.g3 g6, followed by ...Bg7 and ...f6, are fine for Black. You don’t need to know any more about these rare lines.   8.Nd2 is a more significant sideline, which can be compared with 8.c4 Ba6 (instead of my recommended 8...Nb6) 9.Nd2, which is quite a serious option for White. In the present position Black does well with 8...g6! 9.Nf3 (or 9.c4 Nf4!) 9...Bg7 10.c4 Nb6! 11.Bg5 Qb4† 12.Qd2 a5! as proposed by Van der Wiel. See Game 17 for further details.

8...Nb6! 8...Ba6 is the other big main line, but there’s a ton of new ideas (for both sides) which you will not find mentioned in manuals authored just a few years ago. One modern move is 9.Nd2!?, which contains a wealth of interesting possibilities. The text move has the advantage of being more ‘concrete’, and gives considerably less scope for original play.   9.Nc3! With Black’s knight no longer controlling c3, this is the most ambitious continuation for White. We’ll cover this is some detail in the next section.   9.b3 (9.Nd2 will generally lead to the same thing) 9...Qe6 10.Bb2 Bb4† 11.Nd2 0-0 12.0-0-0 is an older line which has a drawish reputation, but we will see 12...Qe7! leading to fireworks in Game 18. If you feel intimidated by White’s piece sacrifice in that game, there is always Piket’s 12...d5 13.exd6 Qxd6 as given in the notes to Black’s move 12 of Game 18.   GAME 17 John van der Wiel – Anastasia Savina Antwerp 2011   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.Nd2 As I mentioned earlier, I actually rate the 8.c4 Ba6 9.Nd2 variation quite highly for White. The present move order is slightly inaccurate.   8...g6! The most principled choice, although Negi’s proposal of 8...Bb7, intending ...0-0-0, ...f6 and ...Rde8, is also good.   9.Nf3

9.c4 Nf4! The position of the knight on d2 gives us this convenient option. 10.Qe3 10.Qe4 Ne6 followed by ...Bg7 and ...f6 is fine for Black. 10...Bh6! The 10...Ne6/...Bg7/...f6 plan is playable here too, but the text move is more ambitious.

11.Kd1 11.Qc3 Bg7 does not help White. 11...0-0 12.h4! 12.Nf3 was seen in Boros – Meier, Austria 2010, when 12...d5!Nµ would have been strong. 12...d5 13.g3 13.h5 d4 (13...Bg5!N actually looks even stronger) 14.Qg3 g5 15.Nf3 Kh8 was clearly better for Black in Chigaev – Smirnov, Kemerovo 2013. 13...d4 14.Qf3 14.Qxd4? Ne6 followed by ...Rd8 is horrible for White.

14...Nh5 15.g4 Bxd2 16.Bxd2 Ng7 17.Bd3 White’s inventive play kept him in the game in Howell – Kryvoruchko, Warsaw 2011. Still, Black has a few good options, my preference being: 17...f5!?N With at least equal play for Black.

9...Bg7 10.c4 Nb6! Planning ...d6 or ...f6. In his series of NIC Yearbook surveys on the Scotch, the Dutch Grandmaster Van der Wiel opined that Black’s system is so effective, that he doesn’t imagine White will be ever be able to show something good against it.   11.Bg5 11.b3 0-0 12.Bb2 (or 12.Bg5 Qe6 followed by ...d6) 12...d6! sees Black successfully undermine the centre.

  11.h4 0-0 12.h5 d6 is also excellent for Black.   11...Qb4† 12.Qd2 12.Kd1? was seen in Moreno Chacon – Teran Abreu, Caracas 2014, when Black missed the powerful 12...Na4!Nµ.  

12...a5! 13.0-0-0 13.a3 is nothing after 13...Qxd2†³ followed by either ...d6 or ...Na4.   Van der Wiel points out that 13.Bd3 Ba6 (or 13...Na4!?) 14.Rc1 0-0, followed by ...Rfe8 or ...d5, gives White some problems.   13...0-0 14.Bf6 Black threatens to open the centre with ...d5 or ...d6 (possibly after a preliminary ...Re8), so White decides to complicate matters before it is too late. Van der Wiel writes that he doesn’t believe in 14.h4, but it should be checked all the same.

My main line continues 14...d6 15.h5 dxe5 16.hxg6 hxg6 17.Bh6 Qxd2† 18.Nxd2 Bxh6 19.Rxh6 Kg7 20.Rh4 Be6³ and White faces a tough defence a pawn down.

14...Ba6 Overlooking the strong 14...d5!N. Van der Wiel mentions this strong move, adding that he had intended 15.Bxg7 Kxg7 16.a3 to limit the damage. Let me continue this line a bit further: 16...Qb3

Black threatens ...Na4 and ...Rb8. White should probably look for simplifications with 17.cxd5 cxd5 18.Qd3 Qa4 19.Qd4, but after 19...Qxd4 20.Nxd4 c5 21.Nb5 Be6 Black is at least slightly better.   The rest of the game is not so relevant for our opening preparation, but I will include some light annotations as the structure and general nature of the position is quite typical for the Scotch.   15.Bxg7 Kxg7 16.b3 16.Qc2 can be met by the standard 16...Rab8„ intending ...Na4.   16...Qa3† 17.Qb2 Qxb2† 18.Kxb2 Rfe8 18...c5 19.Bd3 Bb7 20.Rhe1 Rfe8 is a better move order.

19.Bd3 c5 20.Rhe1

White could have played 20.Be4, which is why 18...c5 was preferable.   20...Bb7 21.Be4 Reb8?! 21...Bxe4! 22.Rxe4 a4= was more accurate.   22.Kc3?! 22.Bxb7! Rxb7 23.Kc3 a4 24.Nd2 would have been slightly unpleasant for Black, although 24...axb3 25.axb3 d5 would have kept her “within the drawing range” according to Van der Wiel.

22...Bxe4 23.Rxe4 a4 24.Nd2 At this point the players agreed a draw, although 24...axb3 25.axb3 Ra3 would have forced White to be careful. ½–½   In the next game we will look at some of White’s secondary options after 8.c4.   GAME 18 Vassily Ivanchuk – Zoltan Almasi Havana 2014   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6

9.b3 9.Nd2 Qe6 is another direction, when White has two main moves:   a) 10.b3 is the most popular choice, when 10...Bb4! 11.Bb2 transposes to the main game.   b) White can also try 10.Qe4, when the rare 10...g6! is strong. This move often works well when White’s queen has gone to e4, because Black no longer has to worry about Ne4 followed by Bg5.

11.Bd3 Bg7 12.f4 0-0 13.0-0 d5 14.Qe2 f6 Black was more than fine in Sanchez Castillo – Vajda, Bled (ol) 2002.   9...Qe6 10.Bb2 Bb4†! 11.Nd2 0-0 12.0-0-0

12...Qe7! This move frees the diagonal of the c8-bishop and sets up the attacking plan of ...a5-a4 and ...Ba3.   12...a5 is playable but premature, as it allows White additional options such as Korchnoi’s 13.a3!?.   12...d5!? This move is also worth mentioning, as it offers Black a solid alternative in case he wants to avoid the madness of the main game. 13.exd6 Qxd6 14.Qe3!?N This was mentioned by Piket as a possible improvement over 14.Nf3 Qh6† 15.Qe3 Qxe3† 16.fxe3 Re8, when Black was fine and only lost because of some later errors in Kasparov – Piket, Dortmund 1992. 14.Ne4 is also nothing to worry about after 14...Qe7 intending ...a5.

14...Ba3 15.Bd3 15.Qc3 f6 16.Bd3 a5„ 15...Bg4 16.f3 Rae8 17.Ne4 Bxb2† 18.Kxb2 Qe5† 19.Kc2 f5 Piket’s line ends here; Black is more than holding his own.

13.Ne4 This is the most ambitious option by far; White is threatening a deadly check on f6.   Another game continued: 13.f4 a5 This gives Black serious counterplay now that a2-a3 is unavailable to White. 14.Qh5 I found a useful improvement:

14...d5!N 14...Ba3 15.Bd3 h6 16.Qf5 g6 17.Qf6 d5 18.Qxe7 Bxe7 19.h3 a4 20.g4 Rd8 21.f5 gave White strong counterplay in Markus – Hartl, email 2007. 15.exd6 15.Bd3 is met by 15...g6. 15.f5 may appear dangerous, but Black is better after either 15...Ba3 or the simpler 15...g6!?. 15...Qxd6 16.Be5 This is necessary to prepare Bd3, which seems to be the only way for White to threaten anything. 16...Qe7 17.Bd3 g6 18.Qe2 Ba3† 19.Kc2 a4 The position remains complicated, but Black has the initiative.

13...Re8 14.Nf6†!? Objectively the attack is unsound, but it is hard to blame Ivanchuk for going for it, as his moves up to this point all seemed to be leading up to this sacrifice.   14.Qc2?! is too tentative. 14...d5! 15.cxd5 cxd5 16.Nc3 Bxc3! 17.Bxc3 c5 18.Bd3 g6 19.f4 c4! Black’s powerful play has yielded a significant advantage. The way Bruzon handles the black pieces from here is a delight to watch:

20.bxc4 Bg4 21.Rde1 Rac8 22.Kb2 Nxc4† 23.Ka1 d4! 24.Bxd4 Red8 White is almost paralysed. 25.Qf2 In Yu Yangyi – Bruzon, Tsaghkadzor 2015, Black eventually lost due to a combination of missed opportunities and heroic defence by White. However, he could have obtained a winning position with:

25...Nd2!!N As pointed out by Krisztian Szabo. 26.Qxd2 Rxd4 27.Re3 Qc5–+ White’s pinned bishop and exposed king give him insurmountable problems.   14.Qh5N is also met by 14...d5 15.cxd5 cxd5. Of course 16.Nf6† is the critical move, but it leads to a forced line which ends in Black’s favour: 16...gxf6 17.exf6 Qc5† 18.Kb1

18...d4! 19.Qh6 Qf8 20.Qg5† Kh8 21.Rxd4 Qg8 22.Qxg8† Kxg8 23.Rxb4 Bf5† 24.Kc1 Re1† 25.Kd2 Rb1 26.Bd4 Rd8 27.Ke2 Nd5 28.Rb7 c5 29.Bxc5 Nxf6µ There is no way to stop ...Bd3†, winning the f1-bishop.   14...gxf6 15.Qg4† Kh8 16.exf6 Qf8 17.Qh4 Black is a piece up but he faces serious threats such as Bd3 or a rook lift at the third rank. Almasi keeps a calm head though and refutes the attack with apparent simplicity.

17...d5! 18.Bd3 Qg8 19.c5! White has burnt his bridges so giving up another pawn hardly matters – it is more important to stop Black from playing ...dxc4.   19...Bxc5 20.Rhe1?! Choosing the wrong rook makes things easier for Black.

  20.Rde1! This has the same idea but does not allow Black to demonstrate the brilliant refutation seen in the game. Nevertheless, he can still repel the attack with the following precise continuation: 20...Re4! Giving back a bit of material to exchange an important attacking piece and free the d5-square for the knight. 21.Bxe4 White has to take the rook, otherwise Black will just consolidate with ...Bf5. 21.Rxe4? dxe4 22.Bxe4 Nd5 is useless for White. Black is not only a piece up, but also has the initiative, with ideas like ...Bd7 followed by ...a5-a4, or ...Bd6 intending ...Bf4, or ...Re8 and ...Be5. 21...dxe4 22.Rxe4 Nd5 23.Rhe1 Bf5 23...Bd7?? 24.Re7+– 24.Re5 Bg6 25.f4 Bb4 26.Rf1 Bf5 27.Rf2 Rd8! Black wins, as the bishop is untouchable:

28.Rxf5 28.h3 is met by 28...Qf8 29.g4 Qc5† with mate to follow. 28...Ne3 29.Qh5 Qg6 Black wins due to the possibility of mate on d1.

20...Bg4! 21.f3 Bd7 22.Re7 Be3†! 23.Kb1 c5 White’s attack has run out of steam, and Almasi gives his opponent no further chance.   24.Qh5 Rxe7 25.fxe7† Bd4 26.Bxd4† cxd4 27.Qe5† Qg7 28.Qxc7 Qf6 29.Re1 Re8 30.Qxa7 White has picked up a few pawns, but Black has prepared a decisive counterattack.

30...Bf5! 31.Bxf5 Qxf5† 32.Kb2 Qg5 33.Kb1 Qg6† 34.Kb2 Qxg2† 0–1  

The Main Line 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Ba6! This move order is slightly unusual, but there are a couple of reasons why I favour it. Firstly, it allows us to enter our main line without having to worry about too many weird alternatives from White. It also gives us the secondary option of a neat sideline of our own, which has recently been tested by a few top players.   9...Qe6 is the most popular move, when 10.Qe4 Ba6 leads to our main line. However, White gets some additional options like 10.Bd2, 10.f4 and 10.b3. True, White also has a couple of ways to deviate against my recommended line below, but each of them has a clear drawback.

10.Qe4 White has to defend c4, and this is the best way to do so.   10.b3?! I dare to call this move inaccurate, despite the fact that it has been used by such great players as Karjakin, Wang Hao and Nepomniachtchi. The key improvement for Black was found by Karjakin at move 16. 10...0-0-0! 11.Bb2 Re8 12.f4 f6

13.0-0-0 13.Ne4 is less explored, but 13...fxe5 14.fxe5 d5 15.Nd2 g6 16.0-0-0 Bg7³ was a good example for Black in the game Stilman – Krol, email 2011. 13...fxe5 14.Qxe5 Qxe5 15.fxe5 Rxe5 16.Bd3 White almost has the makings of serious compensation for the pawn deficit, but the following idea virtually refutes his scheme.

16...Re3! Threatening ...d5. 17.Kd2 Chasing the rook away is the obvious response. White has tried a few other moves, but the overall score currently stands at +4 =0 –0 to Black. 17...Re8

By luring the king to d2, Black makes the coming ...d5 a lot more effective. The point is that the king rather than the rook will end up on d3, so Black will not lose the d5-pawn. 18.Rde1 Be7 19.Ne4 d5 20.cxd5 Bb4† 21.Nc3 Bxd3 22.Kxd3 cxd5µ White had no compensation for the pawn in Wang Hao – Karjakin, Moscow 2013.   10...Qe6 By attacking c4 again, Black forces White’s hand.   If you want to avoid the theoretical lines presented in the rest of the chapter, you can consider: 10...f6!? This has only been seen in ten games at the time of writing, so it is ripe for further investigation. 11.exf6 11.c5!? Bxf1 12.Kxf1 Qxc5 13.exf6† Kf7 14.Qf3 gxf6 15.g3 h5 16.Ne4 Qe5 17.Bd2 Kg6 18.Nxf6 Bg7 19.Bc3 Qf5 was the remarkable continuation of Vachier-Lagrave – Adams, Bilbao 2014. Black’s last move forces a queen exchange, and the game was eventually drawn in 45 moves. According to the engines, the evaluation from move 11 onwards hardly deviated from the typical ‘computerish’ 0.00. 11...Qxe4† 12.Nxe4 gxf6

13.b3 13.Nxf6† Kf7 14.Nh5 Bb4† 15.Kd1 Bxc4 gave Black at least equal chances in Hracek – Karjakin, Tromso (ol) 2014. 13...0-0-0 14.Bb2?! 14.Be3!N is preferable, so that 14...Bb4† can be met by 15.Nd2. This may be the most important direction for further investigation of 10...f6!?. Black’s pieces are active and the engines call it equal, but his pawn structure could become a long-term concern. Still, if your goal is to unbalance the game and avoid the potential for a forced draw in the main line, then this would be a suitable outcome from the opening. The text move was played in Sapa – Bak, Poland 1992 – which, by the way, transposed to this line via a totally different move order of 8.c4 Ba6 9.Qe4 Nb6 10.Nc3 f5 11.exf6 gxf6 12.b3 Qxe4† 13.Nxe4 0-0-0. At this point Black could have seized the initiative with: 14...Bb4†N 15.Kd1 d5 16.cxd5 Bxf1 17.Rxf1

17...Rxd5†! 18.Kc1 Re8³ Black’s tremendous piece activity is more important than his dismal pawn structure.   11.b3 Negi has proposed 11.f4 but I don’t think Black has any problems after: 11...Bb4 (Another good option is: 11...d5 12.f5 dxe4 13.fxe6 fxe6 14.b3 Bb4 15.Bb2 Nd7 16.0-0-0 Nc5 17.Be2 Negi ends his analysis here, but I think Black is fine after 17...Bb7 intending ...a5-a4) 12.f5 Bxc3† 13.bxc3 Qe7 White’s e- and f-pawns look impressive, but he is taking a risk playing so ambitiously while behind in development.

From this position Negi tried to pose problems for Black with 14.Be3 Qa3 15.Rc1 Qxa2 16.Bd3 Bxc4 17.0-0 Bxd3 18.Qxd3. Although White has some attacking prospects for the two pawns, it seems to me that only Black can be better after something like 18...Qc4 19.Qd2 Nd5 20.Bd4 c5.

11...Bb4 12.Bd2 Bxc3 13.Bxc3 d5 We have reached the tabiya of the Professional Variation. First played by Spassky in 1979, it was not repeated until the 1990s, but was subsequently taken up by a host of top players, including Timman, Adams, Karjakin, Bacrot, Ponomariov and Kasimdzhanov. So far Black’s position has withstood all White’s attempts to break it, leading most super-GMs to abandon this line as a winning attempt for White. Naturally it remains popular below the elite level, so I recommended a careful study of the following analysis and games.   14.Qh4! This excellent move was first played by Kasparov to defeat Adams in 1999.   The older 14.Qf3 is inaccurate as it gives Black the useful option of long castling. See Game 19 (Wang Hao – Kasimdzhanov) for more about this.

14...dxc4 15.Rc1!? This was considered an important novelty when Macieja used it to beat Korchnoi in 2004. A year later Radjabov played it successfully against Karjakin, and it remains the most theoretically challenging option – albeit one that can be defused with precise play.   The more popular 15.Be2 Nd5 is not too dangerous. 16.Bd4 was used by Kasparov in the stem game against Adams, but soon afterwards an improvement was found by Piket. See Game 20 (Morozevich – Piket) for further details. 16.Bxc4 is a better try, but 16...g5! should be fine for Black. See Game 21 (Pavasovic – Mastrovasilis) and the accompanying notes for the proof.

15...Qg6! Karjakin’s idea, stopping White’s normal development with Be2.   The idea behind 15.Rc1 is shown after 15...Nd5?! 16.Bxc4 when, compared with the 15.Be2 line mentioned above, White is ready to meet 16...g5? with 17.Qxg5 as the c3-bishop is protected.   16.g3!? This is White’s latest attempt; he prepares Be2 and 0-0, but weakens the long diagonal.   16.Bb4 f6 17.Qh3 Qe4† 18.Be2 Kf7 is examined in Game 22 (Karjakin – Kasimdzhanov). The main line continues 19.Qe3 Qxe3 20.fxe3 Rhe8 21.exf6 Rxe3 22.fxg7 Kxg7 23.Kf2 Rxe2† 24.Kxe2 cxb3† when the endgame should be drawn.   16...0-0 17.Be2 c5 18.0-0 Bb7 Black hurries to place his bishop on its ideal square before deciding where to put his rooks.

19.bxc4 19.Rfe1 cxb3 20.axb3 Rae8 21.Bg4 Qc6 is also fine for Black, as shown in the notes to Game 23.   19...Rfe8! 20.Rfe1 Nd7 The knight heads for f8, where it covers the e6-square. Take a look at Game 23 (Yu Yangyi – Ponomariov) for more details on this important variation.   GAME 19 Wang Hao – Rustam Kasimdzhanov Tashkent 2012   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 9...Ba6! is my preferred move order, as explained in the introductory section.   10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 14.Qf3 dxc4

15.Rc1?! This seems like a normal move, especially considering that it is regarded as White’s best try for an advantage in the similar position with the queen on h4. However, it allows Kasimdzhanov to demonstrate his profound understanding of chess by highlighting the downside. He does this by castling queenside – which would not have been possible had the queen been standing on h4.   A better move is: 15.Be2 As recommended by Negi on his DVD. One obvious advantage is that the rook may be more useful on a1 in case Black castles on the queenside. Despite this difference, Black can still get a playable game whichever side of the board he chooses for his king. Let’s look at both castling options:

a) 15...0-0

This is the usual move. 16.0-0 In my database I found games of grandmasters Spassky, Bacrot and Gupta, but none of them chose the continuation I believe to be the most accurate. 16...c5!N I think this is the simplest way for Black to equalize. The rook comes to b8 and the knight to d5, after which the threat of ...cxb3 becomes real (as the queen will defend the a6 bishop along the sixth rank). A possible continuation is: 17.Rfe1 Rab8 18.Qg3 Nd5 19.bxc4 Nxc3 20.Qxc3 Rbd8 Intending to plant the rook on d4. 21.Rad1 21.Qa3 Bxc4 22.Bxc4 Qxc4 23.Qxa7 Rd2 gives Black enough counterplay to hold the balance.  

21...Rfe8 22.Bf1 h6 22...Rxd1 23.Rxd1 Qxe5 would have been desirable if there was not a back-rank mate, so Black’s last was directed against this horror scenario. 23.f4 Bc8 Black has enough activity to balance out his inferior pawn structure, one tempting plan being ...Qb6 and ...Rd4.   b) 15...0-0-0!?N This may appear risky and the engines don’t like it at first, but in fact Black will gain absolute control of the light squares after the a6- and e2-bishops get exchanged, and this will make his king’s position relatively safe. 16.0-0 Kb7 17.Rfc1 17.Rfb1 can be met in exactly the same way.

17...h5! Black’s plan is to improve his position further with ...g5-g4 and exchange light-squared bishops with ...cxb3 whenever it is convenient – which will most likely be when the queen’s rook steps away from the a-file. Once the bishops are off the board, Black’s control over the light squares will give him excellent counterplay. A sample continuation is: 18.Qe4 g5 19.Rab1 After 19.bxc4 Bxc4 the black king will be safe on a8 after a piece (most probably the queen) comes to d5. 19...cxb3 20.Bxa6† Kxa6 21.axb3 In general, in this kind of situation the king can return to b7 and, when a7 comes under pressure, the knight can guard it safely from c8. In this particular position though, Black can afford to be more energetic

21...Rd5! 22.Ra1† Kb7 23.Qe2 Rb5 24.Ra3 Nd5 Black has taken over the initiative. After long castling by Black the engines evaluate White’s position as better,

without providing any special reason for it. As an extra precaution I decided to leave the IDeA tool running on the position for two whole nights in an attempt to discover a way for White to increase the pressure, but all that happened by the end of that period is that the evaluation dropped significantly, thus confirming my assessment that Black is fine.  

15...0-0-0! Kasimdzhanov’s move gives Black a superior version of variation ‘b’ in the note above.   16.Be2 Kb7 17.0-0 cxb3 18.Bxa6† Kxa6 19.axb3 Kb7 Just as in the aforementioned variation, Black has complete control over the light squares. In the game White’s rook has already committed itself to the c1-square – the undesirability of which is confirmed by Wang Hao’s next move.

20.Ra1 White has nothing better. If he tries to justify the placement of the rook on c1 with 20.Rc2 Rd5 21.Rfc1, Black can

play 21...Rhd8 22.Bb2 Rd3!? 23.Qe2 Qd5µ with a dominating position.   20...Rd5 21.Qe2 Rb5 22.b4 Ra8 23.Rfc1 a6 Black consolidates his position beautifully and does not get too greedy – he is already one pawn up after all.   24.h3 24.Rab1 would reduce White to passive defence, leading to a miserable position for him after something like 24...Nd5 25.Bd2 Re8 26.Re1 Nb6 27.f4 f6µ.   24...Nd5 25.Bd2 Nxb4 Having secured his position as much as possible, Black captures a second pawn. White probably hoped to obtain some counterplay using the extra open file on the queenside, but Kasimdzhanov keeps control and converts his advantage skilfully.

26.Rc4 a5 27.f4 Qd5 28.Rac1 Rd8 29.Be3 Nd3 30.R1c2 c5 31.Ra4 Kb6 32.Rca2 Nb2 33.Qxb2 Qd1† 34.Kh2 Qxa4 35.Bxc5† Kb7 36.Qf2 Qb3 37.Ra3 Qc4 38.Be3 g6 39.Qe1 Rd3 40.Qe2 Rbd5 White had seen enough. 0–1   GAME 20 Alexander Morozevich – Jeroen Piket Internet (blitz) 2000   Despite being a blitz encounter, this game was of considerable theoretical importance as it showed a significant improvement for Black, which essentially put White’s 16th move out of business.   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 14.Qh4 dxc4 15.Be2 Nd5

This is the best reaction to the placement of the queen on h4.   16.Bd4?! Kasparov used this to score a fine victory over Adams in the original game with 14.Qh4, but things move on. Indeed, less than a year after the Adams game (but before the present game took place), Kasparov switched to 16.Bxc4 against Timman, probably because he already knew an improvement was lurking. See the next game for more about the latter move.

16...Qf5! Piket’s powerful novelty sent White players looking elsewhere for a theoretical advantage.   The stem game continued: 16...c5?! 17.Bxc5 Nc3 18.Bxc4 (18.Bf3! was even stronger) 18...Qxe5† 19.Be3 Ne4 20.0-0 Bxc4 21.bxc4 0-0 22.Rfe1 Rfe8 23.f3 Nd6 24.Bf2 Qf5 25.c5 Nb5 26.Qb4 Qd3 27.Red1 a5 28.Qa4 Qe2 29.Re1 Qd3? (29...Nc3 is the only move, although Black is still under heavy pressure) 30.Rxe8† Rxe8 31.Rd1 1–0 Kasparov – Adams, Sarajevo 1999.   16...Ne7!?N 17.0-0 Nf5 18.Qf4 Nxd4 19.Qxd4 0-0 was given by Kasparov as fine for Black, but I don’t think this is entirely accurate, as 20.bxc4 f6 21.f4 gives White some advantage.   17.0-0 No better is: 17.bxc4 Nf4 18.Bf3 0-0! 19.0-0 (19.Bxc6? loses to 19...Rad8 20.Be3 Qd3! 21.Qxf4 Bxc4 when White has to give up the queen) 19...Bxc4 20.Rfe1 Bd5³ Black is a pawn up and his pieces are excellently placed.

17...Nf4 17...Qf4!?N may be even more accurate: 18.Qxf4 (18.Qg4 Qxg4 19.Bxg4 cxb3 gives White compensation but nothing more) 18...Nxf4 19.Bf3 0-0-0 20.Bxa7 Nd3 Black has no problems at all.   18.Bf3 0-0 19.Rfd1? A tactical oversight.   19.Rfe1 would have been better, but Black is still fine after: 19...Rfd8 20.Bc5 (20.Rad1 h5! 21.Be4 Qg4=) 20...Ng6 21.Qg3 Rd3 22.Be4 (22.h4 Nf4) 22...Rxg3 23.Bxf5 Rg5 24.Bxg6 hxg6=   Mikhalevski’s suggestion of 19.b4 can be met by 19...Rfd8 20.Rfd1 Rd5! planning ...Re8 and ...Nd3.

19...cxb3! 20.Re1

20.axb3? Be2! is nasty.   20...Rfd8 21.Be4 21.Bc3 Ng6 keeps the pressure for Black, so White decided to seek salvation in an endgame.   21...Rxd4 22.Bxf5 Ne2† 23.Rxe2 Rxh4 24.Rd2 g6 25.axb3 Bb5 26.Bc2 Rb4µ White has no real compensation for the pawn, as Black’s pieces are active and his pawn weaknesses are not easy to collect. The rest of the game is of minimal interest to our opening study, but it is still a decent example of how to convert an endgame advantage – especially considering the fast time control.

27.Rad1 c5 28.g3 c4 29.bxc4 Rxc4 30.Bb3 Rc3 31.Ba2 Re8 32.f4 a5 33.Rb2 Bc6 34.Bd5 Bxd5 35.Rxd5 Ra8 36.Rd7 a4 37.Rbd2 a3 38.Rd8† Rxd8 39.Rxd8† Kg7 40.Ra8 h5 41.Kg2 Rc2† 42.Kf3 a2 43.h3 c5 44.g4 h4 45.f5 gxf5 46.gxf5 c4 47.Ra7 c3 48.e6 Rc1 49.e7 Re1 50.Rxa2 Rxe7 51.Kg4 Re4† 52.Kg5 f6† 53.Kh5 Rc4 54.Rg2† Kf7 55.Rc2 Ke7 56.Kg6 Rc6 57.Re2† Kd7 58.Rc2 Kc7 59.Kh5 Kd6 60.Kxh4 Ke5 61.Kg4 Ke4 62.h4 Kd3 63.Rc1 Kd2 64.Rf1 c2 65.h5 c1=Q 66.Rxc1 Rxc1 67.h6 Rh1 0–1    GAME 21 Dusko Pavasovic – Dimitrios Mastrovasilis Topola 2004   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 14.Qh4 dxc4 15.Be2 Nd5 16.Bxc4  

16...g5! This was first played by Timman against Kasparov.   17.Qd4 This is the most ambitious reply. 17.Bxd5 This is given as White’s best by Dembo and Palliser, but they do not consider the following reply, probably because the initial engine assessment is not so great. 17...gxh4! 18.Bxe6 fxe6

We have reached an interesting opposite-coloured-bishop endgame. Black’s pawn structure is not pretty, so I analysed the position as deeply as I could to make sure Black is fine. Both IDeA and Monte Carlo analysis confirmed my initial suspicion (which was strengthened by the fact that Kasparov avoided this with White) that

the endgame is a draw. Here are some examples to show how we should handle the position: 19.Rd1 19.f3 0-0-0 20.Kf2 Rd3 21.Rhc1 h3! From White’s perspective, it is quite annoying that Black can flick in this move at any time. 22.g3 c5 23.Ba5 Rf8 24.f4 Rd5 Black had good counterplay and eventually won by raiding the h2-pawn in Krucko – Husak, email 2010. 19.Bd4 0-0-0! 20.Bxa7 Rd5 21.Be3 At this point Black captured on e5 and made an easy draw in Homont – Husak, corr. 2012. Another good option is 21...Rg8!?N 22.Rg1 c5 with ideas of ...c4. 19...Rg8 20.g3 This position was reached in Nepomniachtchi – Ponomariov, Beijing (blitz) 2013, which was White’s only victory to date in this endgame. At this point Black could have obtained a fully satisfactory position with:

20...c5!N Intending ...c4. You may find it helpful to play out the endgame against friends and engines to get a feel for how to handle it. Overall Black should not be in much danger of losing, and there is a serious possibility that White may overestimate his chances and get into trouble – which would explain Black’s superb results (+2 =1 –0) at correspondence level in the endgame.   17...Bxc4 18.Qxc4

18...0-0-0! Once again it was Piket who offered a much-needed improvement over a Kasparov game.   After 18...Nf4 19.Qxe6† Nxe6 20.0-0-0

White is better because of the simple plan of Rhe1-e4-a4, which Kasparov implements perfectly. 20...Ke7 21.Rhe1 Rhd8 22.Rxd8 Rxd8 23.Re4 Rd5 24.Kc2 c5 25.Ra4 White was clearly better and went on to win another positional masterpiece in Kasparov – Timman, Wijk aan Zee 2000. After this we can appreciate the value of the game continuation, as now the black king will be able to cover the queenside weaknesses.   19.Bd4 White has also tried: 19.0-0 Nf4 20.Qxe6† Nxe6

20...fxe6!? 21.Rfe1 Rd3 was also fine for Black in Brod – Neumeier, Hartberg 2004. The text move gives Black an improved version of the Kasparov – Timman game.

21.Rac1 21.Rfe1 Rd3 22.Rac1 Rhd8 23.Kf1 h5 24.Re4 c5 was unproblematic for Black in Macieja – Timman, Curacao 2002, as now 25.Ra4 is met by 25...Kb7. 21...Rd3 22.Rfd1 Rd5 23.Kf1 Rhd8 24.Re1 h5 25.g3 g4 26.Kg2 Ng5 The position is level here, but Black got the upper hand quickly after:

27.e6?! Nxe6 28.Bf6 Rg8 29.Re4 Rg6 30.Be5? Nc5! 31.Rxc5 31.Ree1 Nd3–+ 31...Rxc5 32.Bf4 Re6 33.Ra4 a5 34.a3 Rb5 0–1 Mamedov – Karjakin, Calvia (ol) 2004.

19...c5! 19...Nf4?? 20.Qa6† Kd7 21.0-0-0+–   19...Kb7 led to a draw in Baklan – Piket, Germany 2000, but Pavasovic won a game from this position a few months before the present game. The game continuation is a strong improvement from the Greek GM, which buries another try for White.   20.Qxc5 After 20.Bb2 Black can play 20...Nf4 since the a7-pawn is not hanging.   20...Qg4! 21.Qxa7 Qe4†! 21...Nf4? allows 22.Qa6† when the king cannot go to b8 because of Ba7† with mate to follow. Therefore 22...Kd7 is forced, when 23.Rd1! is winning for White.   22.Be3 Nf4! Threatening ...Qb4† followed by ...Qb5† and ...Ne2† with a quick checkmate.   22...Nxe3!? 23.Qxe3 Qxg2 24.Ke2 Rhe8 is another route to an equal game.   23.Qa6† White defends against the mating threat with gain of tempo.   23...Kb8 24.Rg1!? Rather than take a perpetual White defends g2, but a new wave of attack comes.

24...Rd3! Threatening to double on the d-file as well as to take on e3. Pavasovic wisely decides to settle for a draw.   25.Kf1 Rxe3 26.Qb5† Kc8 27.fxe3 Nd5 28.Qa6† Kb8 29.Qb5† Kc8 30.Qa6† ½–½   GAME 22 Sergey Karjakin – Rustam Kasimdzhanov Zug 2013   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 14.Qh4 dxc4 15.Rc1 This is the most challenging move of all. White prepares to meet ...Nd5 with Bxc4, as the ...g5 response will no longer work with the c3-bishop defended.  

15...Qg6! Kasimdzhanov plays the move that his opponent introduced five years earlier against Wang Hao.   16.Bb4 Karjakin repeats the move which was played against him by the Chinese super-GM, as well as by Carlsen in a blitz game in 2010. White prevents short castling, while incidentally threatening mate.   16.g3 is examined in the next game.   16...f6! Black prepares to castle queenside.   17.Qh3 17.exf6? 0-0-0! is too dangerous for White.   17.bxc4?! 0-0-0! also favours Black. A sample line is: 18.g3 Rhe8 19.Bc3 fxe5 20.Be2

20...Rd4! 21.Bxd4 exd4 22.Qxd4 c5 Black has a dangerous initiative. Depending on where the white queen goes, Black may proceed with ...Qd3, ...Qe4 or simply ...Bb7.   The text move is better. White prevents castling and threatens Qe6†, while also defending g2 and thus enabling the f1bishop to move. When Karjakin had this position as Black, he played 17...Bc8 and drew both of the aforementioned games against Wang Hao and Carlsen. He must have believed he could get something from the other side of the board, but his opponent was ready with a radical new idea.   17...Qe4†! Kasimdzhanov has unearthed an amazingly deep idea which leads to a drawing endgame.   18.Be2 18.Qe3N is an obvious alternative, but Black has a strong reply: 18...Qxe5! (18...Qxe3† 19.fxe3 0-0-0 is close to equal but Black can be more ambitious) 19.Qxe5† fxe5 20.bxc4 Rd8! (20...Kf7!? is also okay) 21.Be2

21...Bxc4! 22.Bxc4 Rd4 23.Bb3 Rxb4 24.0-0 Kd7 Black emerges with two extra pawns at the end of this long line given by Mikhalevski. True, his structure is bad and White can probably hold with accurate play, but Black can certainly press.

18...Kf7 Preventing Qe6† and enabling Black to continue development.   19.Qe3 I don’t think White has anything better.   19.0-0!?N Objectively this is not such a strong move, but it should definitely be checked, as it is the kind of move a reckless attacker might try.

19...Qxe2 20.exf6 20.Rfe1 runs into 20...Bc8! 21.Rxe2 (21.Qc3 Qd3) 21...Bxh3 22.gxh3 Rhe8 and Black is a pawn up.

20...Bc8 21.Qg3 21.Qh4 Qg4 22.fxg7 Re8 23.Qxh7 Qxg7 24.Qh4 Bf5 is clearly better for Black; he may not have any pawns in front of his king, but his active pieces provide adequate cover. 21...Qg4 22.Qxc7† Bd7 23.h3

23...Nd5! 24.fxg7 Rhe8 25.Qd6 Qe6 Black emerges from the complications with a clear advantage.

19...Qxe3 20.fxe3 Rhe8 This is the right rook, as White’s pawn will go to g7 and it would be better for us if this does not come with tempo.   21.exf6 21.Bc3 gives Black a choice between 21...fxe5, 21...Nd5!? and 21...Kg8, all of which are fine for him.   The text move is the only ambitious continuation, leading to a forcing sequence.   21...Rxe3 22.fxg7 Kxg7 23.Kf2 Rxe2† 24.Kxe2 cxb3† 25.Kf2 25.Kd2 would help Black to develop his rook with tempo: 25...Rd8† 26.Kc3 Rd3† 27.Kb2 Na4† 28.Kb1 c5 29.Ba3 c4 30.Rhe1

30...Bb5! Safeguarding the bishop from a check on the sixth rank. 31.Re7† Kg6 32.Re6† Kg7= White had better take a

draw while it is available, otherwise Black will play either ...Nc3† or the even stronger ...c5 followed by ...Rd2 (the first move is necessary to ensure that White will not be able to play axb3 followed by Rxc7†).

25...bxa2 26.Rxc6 Nd5 27.Bc3† 27.Rxa6 Nxb4 28.Ra5 a6! (28...Re8 29.Rxa7 Kf6 is also a draw but the text move is more convincing; Black intends ...Rb8-b5) 29.Ke2 Rb8 30.Kd2 Rb5 31.Rxb5 axb5 32.Kc3 c5 33.Rf1 Kg6 34.Kb2 Kg5 and Black is ready to push his pawn down the h-file.

35.Rf7 c4 Threatening ...c3†. 36.Rf1 Kg4 37.Kc3 Nd5† and White has to go back and agree to a draw, as 38.Kd4? b4 39.Kxd5 b3 leaves him helpless against the pawns.

27...Nxc3 28.Rxc3 Bc8! Black transfers the bishop to e6 and brings the rook to b1, thus establishing a fortress.   29.Rxc7† Kh8 30.Re1 Be6 31.Rc6 Rf8† The immediate 31...Rb8?? would have allowed 32.Rcxe6 Rb1 33.Ra6+– and the a-pawn is stopped.   32.Ke3 32.Ke2 is strongly met by: 32...Bd5! 33.Rc3 (If 33.Rcc1 Bxg2 Black picks up a pawn while preserving his passed pawn on a2) 33...Bc4†! (33...Bxg2 34.Ra3 Bd5 should hold, but White can play on for a while)

34.Kd2 Rf2†! 35.Ke3 Re2†! Black forces the a-pawn through, but it’s just a draw anyway. 36.Rxe2 a1=Q 37.Rxc4 Qg1† 38.Rf2 Qxh2=

32...Rb8 Now after 32...Bd5? White has 33.Ra6, which was impossible with White’s king on e2.   33.Ra6 Rb1 34.Kd2 Rb2† 35.Kc3 Rb1 36.Kd2 Rb2† 37.Kc3 Rb1 38.Kd2 ½–½   GAME 23 Yu Yangyi – Ruslan Ponomariov Danzhou 2014   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Nf6 5.Nxc6 bxc6 6.e5 Qe7 7.Qe2 Nd5 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 14.Qh4 dxc4 15.Rc1 Qg6 16.g3!? This is White’s latest attempt to prove something.   16...0-0 17.Be2 c5! Ponomariov plays the most flexible move.   17...Rad8 18.0-0 c5 was seen in Caruana – Bacrot, Thessaloniki 2013, when the then Italian player came up with a poisonous idea in 19.Rfe1!, planning Bg4 and eventually e5-e6.   18.0-0 Bb7 The bishop settles on its ideal square. Now White must worry about the possibility of ...cxb3, and ...Qc6 might prove useful, especially in response to a future Bh5.  

19.bxc4 Yu Yangyi does not want to risk being a pawn down for the rest of the game.   If White plays in the style of Caruana with 19.Rfe1 intending Bg4 and e5-e6, Black can benefit from not having committed his rook yet: 19...cxb3 (19...Rfe8!? 20.Bg4 Bd5 is another sensible idea) 20.axb3 Rae8! 21.Bg4 Qc6 22.f3

22...h6! This useful little move anticipates the possibility of Bf5 (which will no longer win a tempo) followed by Be4. Black is at least equal, a sample line being: 23.e6 fxe6 24.Qh5 Nd7 25.Qg6 Nf6³ Black’s prospects are somewhat better in this complex situation.   19...Rfe8! The most accurate rook move, keeping in mind the following manoeuvre.  

20.Rfe1 Nd7! The knight is heading for f8 and e6. If Black can complete this manoeuvre his coordination will be perfect, so it’s now or never for White to do something.

21.e6!? This is the only serious attempt to cause problems, but Black is well placed to meet it.   21...Rxe6 22.Bh5 White has to force the play, otherwise ...Rae8 is coming.   22...Qh6 23.Rb1 Rxe1† 24.Rxe1 Rf8! Excellent play by Ponomariov. White is more or less forced to simplify into a worse but tenable endgame.

25.Bg4 Qxh4 26.gxh4 Bc6 26...Nb6!? 27.Re7 Bc8 is marginally better for Black, but I think Ponomariov’s move is the best winning try. White is forced to bail out to a slightly worse endgame with rook and opposite-coloured bishops.   27.Bxd7 Bxd7 28.Re5 f6 29.Rxc5 c6 30.Ra5 Be6 31.Rxa7 Rd8 32.Re7

32...Bxc4 32...Rd1†! 33.Kg2 Bxc4 would have given White a bit more to think about, although it should still be a draw of course.   33.a4 Rd3 34.Re3 Rd1† 35.Re1 Rd3 36.Re3 Rd1† 37.Re1 Rd3 38.Re3 ½–½   Conclusion   The Scotch is one of the most challenging openings Black can face after 1.e4 e5. A quality solution is needed, and the Professional Variation fits the bill.   After examining several unthreatening sidelines, we moved on to 8.Nd2 in Game 17. This method of developing the knight can work well if Black has committed his bishop to a6 but, as things stands now, Black gets a good game with 8...g6!. In Game 18 we examined what happens if White reacts to 8.c4 Nb6 by developing his knight to d2 and the bishop to b2. Ivanchuk’s piece sacrifice is interesting, but if Black keeps a cool head he emerges with excellent chances to convert his material advantage. If you forget the analysis or don’t feel comfortable playing this way over the board, you have a safe alternative available in the form of Piket’s 12...d5.   The rest of the chapter took place in the deep waters of the main line. Before getting too deep, however, I mentioned 10...f6!? as an interesting alternative; the games Hracek – Karjakin and Vachier-Lagrave – Adams, both from 2014, are a useful starting point for further investigation.   The first major crossroads in the Professional Variation comes at move 14, when White has to make up his mind between the older 14.Qf3 and Kasparov’s 14.Qh4. The latter move is stronger, for the reason demonstrated in Game

19: the older move allows Black to castle on the queenside and obtain great counterplay.   The second crossroads comes when White chooses between 15.Be2 and the more modern 15.Rc1.   After 15.Be2 Nd5! Black is fine after both 16.Bd4 Qf5! (Piket) and 16.Bxc4 g5! (Timman). Games 20 and 21 provide the details.   After 15.Rc1 Timman’s idea no longer works because the c3-bishop is defended, but Karjakin’s 15...Qg6! is an excellent solution. White can play 16.Bb4 (Game 22) or the more recent try 16.g3 (Game 23) but in each I find no reason for Black to be worried, unless he needs to avoid a draw at all costs – in which case 10...f6!? could be preferred.

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6   The Original Morphy Attack 4.d4 exd4 5.Ng5?! page 145 The Improved Morphy Attack 5.0-0 Nxe4 6.Re1 d5 7.Bxd5 page 146 (7.Nc3?! page 146) The Modern Variation – Introduction 4.d4 exd4 5.e5 d5! 6.Bb5 Ne4 7.Nxd4 Bc5! 8.0-0?! page 151 (8.Nxc6? page 151) 8.Be3 Game 24, page 153   4.d3 – Introduction and 5.Nc3: 4.d3 Be7 5.Nc3 page 155 Main Lines after 5.0-0 0-0 page 157 The c2-c3 Plan 6.Bb3 (6.c3 page 157) 6...d6 7.c3 Game 25, page 157 The a2-a4 Plan 6.a4 page 160 The Right Move Order 6.Re1 d6 7.a4 page 161 The Critical Na3 Plan 7...Na5 8.Ba2 c5 9.Na3! page 163  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

What is the most accurate way to play? (page 145)

How can White be put under pressure? (page 146)

What is the accurate way for

Black to equalize? (page 154)

How should Black meet Tiviakov’s plan? (page 159)  

How should Black play? (page 151)  

How can White be made to suffer for his sins? (page 152)

Find a plan for Black. (page 163)  

What is the key move for Black here? (page 164)

  1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6! I consider the Two Knights Defence the best weapon against White’s last move, especially if you are looking for a game full of excitement!   3...Bc5 is the main alternative. My problem with this move is that it gives White a choice between a slow, closed game,

which is reminiscent of the Spanish but with the bishop on c5 (which is not how I recommend playing against the Spanish itself), and a bunch of gambits and other tricky tries, all of which try to take advantage of the placement of the bishop to c5 in some way, with either b2-b4 or d2-d4 at some stage.   By contrast, in the Two Knights, the closed lines after 4.d3 are not that different from the way we will meet the Spanish, while the attacking lines after 4.d4 lack punch. As often happens in chess, by gaining in some areas it means we have to give up something else: in this case, a big fat pawn after 4.Ng5!. That move will be discussed separately in the next chapter, once we have dealt with the following moves:   4.d4 is the Morphy Attack, to use Alterman’s terminology. Black has to know a bit of theory to deal with it, but I have managed to reduce the workload rather significantly, by recommending a move which has proven to be reliable at correspondence level, and whose surprise value could easily catch out your opponents.   4.d3 keeps the e-pawn solidly defended for the time being. The ensuing positions resemble the d2-d3 system in the Spanish, as discussed later in Chapter 10.   4.Nc3 transposes to the Italian Four Knights, where 4...Nxe4! is the answer.   4.0-0?! Nxe4 followed by ...d5 is simply bad for White. His best attempt is 5.Nc3, with another transposition to the Italian Four Knights, as covered earlier.   The only other sideline worth mentioning is Hector’s pet line: 4.Qe2 According to my friend, Sune Berg Hansen, who has spent many hours in his life preparing against the creative Swedish grandmaster, Black’s best is: 4...Bc5! It is no coincidence that another strong grandmaster and theoretician, Ivan Sokolov, chose this move against Hector the last time they met.

5.c3 The ‘trap’ 5.Bxf7†? Kxf7 6.Qc4† backfires after 6...d5! 7.Qxc5 Nxe4 when Black is much better, as has been known since Albin – Schlechter, Vienna 1914! 5...0-0 6.d3 d5! 7.Bb3 We have transposed to a line of the Giuoco Pianissimo which actually scores quite well for White, but this is one

of those cases where the statistics don’t tell the whole truth. As we will see in Chapter 9, the ...d5 break works well in positions where the queen has gone to e2, as White is unlikely to be able to play exd5 to good effect. 7...Be6! I like this simple solution. 8.Ng5 Ne7!? 8...Bg4 is also good, and in fact I don’t see anything better than 9.Nf3.

9.Nxe6 What else? 9...fxe6 10.0-0 Ng6 11.g3 c6 12.Nd2 a5 13.Nf3 Qb6 14.Ng5 Rae8 Black was doing fine in Hector – I. Sokolov, Helsingor 2012; his eventual defeat was the result of later mistakes.

The Original Morphy Attack 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d4 exd4 White has opened the centre, and he hopes to create problems using his slight lead in development. In this section we will look at the way Morphy originally handled the position.   5.Ng5?! The database shows that Morphy won some games with this line in 1858, but it is simply a poor cousin of the 4.Ng5 line which will be analysed later.   5...d5 6.exd5  

6...Qe7†! 7.Kf1 7.Qe2?! Qxe2† 8.Kxe2 Nb4 already favours Black, who will take on d5 next.   7...Ne5 8.Qxd4 8.Bb3 h6 9.Nf3 Nxf3 10.Qxf3 Bd7 is unpleasant for White; he must defend against ...Bb5†, while Black is ready to castle and target the misplaced white king.   Grabbing the d4-pawn is White’s only chance to make sense of his opening play, but Black has great compensation thanks to the bishop pair and the king on f1.

8...Nxc4 9.Qxc4 h6 10.Nf3 Qc5 11.Qe2† Be7 12.c4 After 12.Be3 Qxd5 13.Nc3 Qa5 White’s king is misplaced and Black’s bishops have great potential. White can grab a pawn with 14.Re1 0-0 15.Bxh6, but after 15...Re8 16.Bd2 Bd7 Black had more than enough compensation in Amann – Janosi, email 2008.

12...b5! 13.Nc3 Qxc4 14.Qxc4 bxc4 15.Bf4 Bb7 16.Bxc7 Nxd5 17.Nxd5 Bxd5 18.Re1 Kd7 19.Be5 f6 20.Bc3 Bc5 Black’s bishop pair gave him the better chances in the endgame in Bjornsson – De Jong, Porto Mannu 2015.

The Improved Morphy Attack 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d4 exd4 5.0-0 5.e5 is covered next.   The text move does not have the best theoretical reputation, but its devotees do not get discouraged by such things. They usually know the main lines inside out, some of which extend deep into the endgame. I was happy to find a solution which should not be too difficult to remember, and which has a decent chance of taking your opponents out of their preparation.  

5...Nxe4 6.Re1 d5 7.Bxd5 This small tactical trick is what justifies White’s opening play.   7.Nc3?! This is a tricky sideline, but Black gets an advantage with a few decent moves. 7...dxc3! 7...dxc4 is also playable, but the text move is stronger. 8.Bxd5

8...Be6 9.Bxe4 Qxd1! 10.Rxd1 cxb2 11.Bxb2 f6 Black has an extra pawn and the better structure, and he will follow up with ...Bd6 and ...0-0-0 if permitted. 12.a4!? 12.Re1 Kf7! is a necessary deviation from the master plan, as Bxc6† was threatened. 13.Bc3 Ba3! (13...Bd6 is possible, but the text move is even stronger as it leaves room for the rooks to be exchanged along the d-file) 14.Rab1 Rab8 15.Red1 h6 Black stands better and is ready for ...Rhd8. This is the only move which has survived for White in correspondence games. For some reason, however, nobody has played the obvious reply: 12...Bd6N Black holds a healthy advantage. Of course, if 13.a5 then simply 13...a6 keeps things solid.   7...Qxd5 8.Nc3 Here I am pleased to recommend a rare move which could easily catch your opponents off balance.

8...Qd7! This move is virtually unknown, having been played in less than thirty games, out of nearly four thousand after White’s last move! It is fully sound, contains a drop of poison and does not allow any forced draw.   8...Qa5 is by far the most popular move, and 8...Qh5 is the main alternative. Interestingly, Lokander offers the latter move as his main recommendation, but mentions my recommended move in his notes and even awards it an exclamation mark! I used to consider 8...Qd8!? a decent alternative, but am not completely happy with Black’s position after 9.Rxe4† Be7 10.Nxd4 f5 11.Rf4 0-0 12.Nxc6 Qxd1† 13.Nxd1 bxc6. White’s score in my database is mediocre (slightly below 50%) and the engines’ evaluation is slightly in Black’s favour. However, Black’s structure is not great at all, and it would not take many inaccuracies for him to end up worse because of it.   9.Nxe4 With the queen on d7 instead of d8, taking with the rook does not work as well for White: 9.Rxe4† Be7 10.Nxd4 10.Nb5 0-0 11.Bf4 is another try, but Black is fine after 11...Bd8. 10...f5! Now it is obvious that the queen is better on d7.

11.Rf4 11.Rxe7†N 11...Nxe7! 12.Bf4 (12.Qh5† Ng6 13.Nf3 Qc6 14.Ne5 Qe6 15.Bf4 0-0µ) 12...Ng6 13.Qe2† Kf8 14.Be5 Qe7³ Black defends against White’s temporary initiative and remains material up. 11...0-0 12.Nxc6 Qxc6 In Hund – Zahn, Erfurt 2014, Black enjoyed the advantage of the bishop pair without having had to compromise her structure.   9...Be7 10.Bg5 This kind of attack would be bad with the queen standing on d8, but here it is White’s best.   10...0-0 10...f6? would have been fine with the queen on d8, but here it allows 11.Bxf6! 0-0 12.Bxe7 Qxe7 13.Nxd4 when Black does not have enough compensation for his pawn deficit. The text move is therefore forced.   11.Bxe7 Qxe7 12.Nxd4 Bd7 Black is easily equal, and in the long term he can try to exploit the superiority of his bishop over a white knight.   13.Nb5 13.Nb3 Rae8 14.Qd2 Bc8 was also fine for Black in Krzyzanowski – Anokhin, email 2011. The fact that Krzyzanowski, a strong correspondence player, was unable to challenge Black’s set-up, is an indication of the soundness of 8...Qd7.

13...Rad8 14.Qh5 14.Nxc7?? Be6 wins for Black.   14...Ne5 I have a handful of centaur (human plus engine) games on my database from this position, contested on the PlayChess server, all of which ended in Black’s favour.   15.Nxc7N This was not played in any of the aforementioned games, but it is the only truly critical move so it makes sense to check it.

15...Bc6 16.Ng5 h6 17.Nge6 fxe6 18.Qxe5 Rd2

White has saved his knight with some clever tactics, but Black has a strong counterattack. I will analyse a bit deeper, as most of White’s moves are forced.

19.f3 Qh4 20.Re2 Rxe2 21.Qxe2 Qe7 21...Qd4† followed by ...Qxb2 is fine as well.   22.Qxe6† 22.Nxe6? Re8 is obviously bad, while 22.Qe5 Rf5 gives White no choice but to transpose to the main line with 23.Qxe6† Rf7.   22...Rf7 23.Qxe7 Rxe7 24.Na8

24...Rd7!

24...b6?! 25.Rd1! is more complicated, as White will defend his knight with Rd8. The text move prepares ...b6, when White will have to give up his knight and fight for a draw.

The Modern Variation – Introduction 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d4 exd4 5.e5

The term “Modern Variation” refers to the fact that this move was adopted by Lasker, Marshall and others towards the end of the 19th century, as they attempted to improve on the older Morphy games.   5...d5! Interestingly, this was played successfully by Morphy himself, as well as by Anderssen, his main rival. Based on their impressive victories with the black pieces, it is hardly surprising that the whole 5.e5 variation was dismissed as offering nothing for White. Nowadays though, it is the only challenging continuation after 4.d4 exd4, and I must admit it cost me quite a lot of time and energy to reach the conclusions I eventually did. Before moving on, let me show you some of the problems Black may face after other moves.   5...Ng4 6.0-0 d6 6...Be7 might well be Black’s best move, and I noticed it was recommended by Lokander not long before this book went to print. The problem is that after 7.Re1 d6 8.exd6 Black has to compromise his structure with 8...cxd6. Lokander thinks Black is doing okay and he may be right, but I do not fully trust the position, and in a practical game it would only take one or two small errors for Black to become worse with the isolated d-pawn. 7.exd6 Bxd6 7...Qxd6 8.Re1† Be7 (8...Be6 might be better, but this is another indication that something is wrong with Black’s position. After 9.Bg5!? there is no convenient solution to Black’s opening problems.) 9.b3!? 0-0 10.Ba3 Qd8 11.Bxe7 Nxe7 12.Qxd4 Qxd4 13.Nxd4 Even though Black is so far undefeated from this position, he is a touch worse due to White’s annoying lead in development. 8.Re1† Kf8 8...Be7 9.Bg5 0-0 10.Bxe7 Nxe7 11.Qxd4 Qxd4 12.Nxd4 produces the same kind of problems for Black as in the 7...Qxd6 variation. This is a popular line for Black with lots of games played, but I was unable to solve the problems after: 9.Nbd2 Nge5 10.Nxe5 Nxe5 11.Bb3! Qh4 12.h3 Bxh3 13.Rxe5! Bxe5 14.Qf3 White is clearly better, as has been shown in a bunch of correspondence games.

  5...Ne4 There is nothing terribly wrong with this move, but I don’t really like the position reached in the line below. 6.Qe2 Nc5 7.0-0 Ne6 7...Be7 8.Rd1 Ne6 9.Bxe6 is the same. 8.Bxe6

8...fxe6 8...dxe6 9.Rd1 Bc5 10.c3! is good for White. 9.Rd1 Be7 10.Nxd4 Nxd4 11.Rxd4 0-0 12.Nc3 b6 12...d6 13.exd6 cxd6 14.Be3 is slightly better for White. 13.Be3 Bb7 14.Rad1 Bc6 15.b4!? White is pressing. Black should be able to hold if he defends accurately for some time, but this is not my cup of tea.   6.Bb5   6.exf6? dxc4 7.0-0 Qxf6 8.Re1† Be6 9.Bg5 Qf5 leaves White two pawns down for no compensation; Black will easily complete his development with ...Bc5 and ...0-0.   6...Ne4 7.Nxd4

7...Bc5! This was played twice by Anderssen in 1861 – what else would you expect from him? Even though Black seems to be sacrificing material, the text move is actually the safest, most forcing route to an equal position.   7...Bd7 Interestingly, this was Morphy’s choice in 1857. It has occurred in roughly three times as many games as our main line, but I was not completely happy about the following continuation: 8.Bxc6 bxc6 9.0-0 Bc5 10.Be3! Bb6 11.Nd2!? A correspondence game of mine continued 11.f3 Nc5!? (11...Ng5 is the main line) 12.f4 0-0 13.c4 Ne6 14.f5 Nxd4 15.Bxd4 c5 and Black was at least equal in Nilsson – Ntirlis, corr. 2015. I discovered the text move when analysing for the above correspondence game, and was quite relieved when my opponent opted for the main line. I will show you why I was less than satisfied with Black’s prospects.

11...c5 11...Nxd2 12.Qxd2 0-0 (12...c5 13.Nb3 transposes to the line below) 13.Nb3 White establishes a blockade on the queenside dark squares. 12.N4b3 Nxd2 13.Qxd2 Bb5 13...d4 is the other idea, but after 14.Bg5 Qc8 15.c3 Bc6 (threatening ...Bxg2 and ...Qg4 with a draw) 16.e6! White keeps a slight plus.

14.Bxc5!N I was quite alarmed when I discovered this. 14...Bxc5 15.Nxc5 Bxf1 16.Rxf1 0-0 17.f4° Although Black should be able to hold with accurate defence, White has full compensation and a troublesome initiative, especially in a practical game.   8.0-0?! 8.Be3 is the main line, and will be covered in Game 24 below. 8.Nxc6? This move must be checked of course, but it turns out badly for White. 8...Bxf2† 9.Kf1 bxc6 9...Qh4 has also been tried but the text move is my preference, and leads to a big initiative for Black.

10.Bxc6† 10.Bd3 Qh4! is devastating. 10...Kf8 11.Qd3 According to my computer, this is the only chance for White to save himself. 11.Bxa8? is met by 11...Ba6† of course. 11...Bf5 12.Qf3N 12.Qxd5? Qh4 13.Bxa8 Bb6–+ was already curtains in Tomczak – Motyka, Kolobrzeg 2001.

12...Bd7! 13.e6! This is another difficult move that White must find in order to keep himself in the game. 13...Bxe6 14.Bxa8 Qxa8 15.Nc3 Bc5! 15...Bb6 has the downside that after 16.Nxe4 dxe4 17.Qc3 White can neutralize the strong bishop with Be3. The text move allows Black to drop the bishop back to d6, keeping a powerful initiative for the exchange.

8...0-0! Once again, Black does well to ignore the apparent threat to take on c6.   9.Nxc6 9.Bxc6 bxc6 10.Nxc6 (10.Be3 is well met by 10...Qe8!³ when White must worry about several ideas, including ...Qxe5, ...f6 and ...Ba6)

10...Qd7! 11.Nd4 Qe7! 12.Bf4 f6!³ Neuvonen – Timmerman, corr. 1991. Timmerman is a living legend among correspondence players and I always trust his analysis in the open games, as he never wastes a chance to punish an opponent for venturing a dubious sideline.   9...bxc6 10.Bxc6

10...Ba6! 11.Bxa8 The other line continues: 11.Qxd5 Bxf1 12.Qxe4 (12.Bxa8? Bc4 13.Qxd8 Bxf2† 14.Kh1 Rxd8 was winning for Black in Sifrer – Mauhler, corr. 1976) 12...Bb5! 13.Nc3 Bxc6 14.Qxc6 Bd4 15.Bf4 Black has several interesting moves here, but I think the most accurate is:

15...Rb8! 16.b3 Qe8 17.Qf3 Rb6µ In Oestergaard – Daus, email 2005, Black was ready to swing his rook to the kingside and post his queen on c6.   11...Bxf1 12.Be3 Objectively this is not the best, but it is worth checking it closely to get an understanding of Black’s tactical resources.   12.Kxf1 is the lesser evil for White, but after 12...Qh4 13.Qf3 Rxa8 Black was clearly better in Maxfield – Morss, corr.

1992, as he was poised to restore material equality while keeping a significant lead in development.   12...Bxe3 13.fxe3 Bxg2! 13...Qg5!? led to an eventual victory for Black in Dufresne – Anderssen, Rotterdam (1) 1861, but the text move is better.   14.Qg4 14.Kxg2 loses to 14...Qg5† followed by ...Qxe3.   14...Bh3!

15.Qxh3 15.Qh5 prevents the queen check, but 15...Be6 leaves White’s bishop in trouble. 16.Bc6 Qb8! 17.b3 Qb6 and White can resign.   15...Qg5† 16.Kf1 Rxa8–+ In Herrmann – Keres corr. 1934, Black soon won by taking on e5 and bringing his rook into the attack. In the following game we will consider 8.Be3, which seems to be White’s only way to maintain equality.   GAME 24 John Menke – Fred Kahl Correspondence 2014   1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Bc4 Nf6 5.e5 d5 6.Bb5 Ne4 7.Nxd4 Bc5! 8.Be3

8...0-0! Black has tried 8...Bxd4, but practice has shown that it does not equalize.   8...Bd7 transposes to the 7...Bd7 line which I prefer to avoid, for reasons explained earlier.   9.Nxc6 9.Bxc6?! bxc6 10.Nxc6? (10.0-0 is better, but we have already encountered this position in the note to White’s 9th move on the previous page via the move order 8.0-0?! 0-0! 9.Bxc6 bxc6 10.Be3, where it was pointed out that 10...Qe8!³ is excellent for Black) 10...Bxe3! 11.Nxd8 Bxf2† White will have to return the extra queen:  

12.Kf1 Ba6† 13.Qd3 Bxd3† 14.cxd3 Raxd8 15.dxe4 Bd4 16.Nc3 dxe4 In Zadruzny – Castaneda, Dos Hermanas 2004, Black remained a pawn up with the better minor piece as well.  

9...bxc6 10.Bxc5 Nxc5 11.Bxc6 Rb8

12.Qxd5! This natural move is the only real try for an advantage.   12.Bxd5?! Ba6 leaves White’s king precariously placed in the centre.   12.0-0 gives Black the chance to equalize at once with: 12...Ba6! (12...Rxb2 also proved to be playable in Nakamura – Onischuk, Saint Louis 2015, but the text move is more straightforward) 13.Re1 Rxb2 14.Bxd5?! (14.Qd4 Rxc2 15.Na3 is White’s best, and now 15...Nb3!= is the nice trick which solves all Black’s problems) White’s last move was slightly careless, and after 14...Nd3! Black had slightly the better of it in Collins – Wells, England 2015.   12...Qe7 13.0-0 Rxb2 14.Nc3 Rd8 15.Qf3 15.Qc4 is also possible, but after 15...Ba6 16.Bb5 Rxc2! Black is fine according to Mikhalevski.

15...Ne6! Rather than capturing a pawn on e5 or c2, Black hurries to put his knight on the excellent d4-square. The continuation of our correspondence game confirms that White has no advantage.   16.Rfd1 Nd4 17.Qe4 Nxc6 18.Qxc6 Bb7 19.Qc4 Re8 19...Rf8!?N intending 20.Re1 Qg5 seems also fine for Black.   20.e6 Qxe6 21.Qxe6 fxe6 22.Rab1 Rxc2 23.Rxb7 In view of 23...Rxc3 24.Rxa7 Rc2, the players agreed a draw. ½–½

4.d3 – Introduction and 5.Nc3 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 With this move White defends the e4-pawn solidly and aims for a slower game, which bears some resemblance to the d2-d3 lines of the Spanish, as covered in Chapter 10.

4...Be7 One of the advantages of 3...Nf6 over 3...Bc5 is that Black has the option of developing his bishop on the e7-square, avoiding a completely symmetrical position. The c4-bishop is more active than the one on e7, but it is also more vulnerable, so White will have to invest a tempo at some point to ensure it will not be eliminated by a timely ...Na5.   5.Nc3 This is not the most popular but it’s a perfectly sound move, which might be played by anyone from juniors and club players to experienced tournament players who wish to play solidly and avoid mainstream theory.   The usual continuation is 5.0-0 0-0, when White normally safeguards his bishop with either c2-c3 or a2-a4 at some point. Both of these plans will be discussed later. In the event of 6.Nc3 d6 the position will soon transpose to the present variation.   5...d6 6.a3! Otherwise Black will play ...Na5 and eliminate the active bishop.   Here is a model game involving that theme: 6.h3 Na5 7.Bb3 0-0 8.0-0 c6! Note that Black is not in a hurry to exchange on b3; he can eliminate the bishop any time, so it makes sense to keep his opponent guessing for a few more moves. 9.d4 Qc7 10.Be3 Re8

11.Re1 Nxb3 (11...h6 or 11...b5 could also have been played, but it makes no real difference, as Black will almost certainly play ...Nxb3 sooner or later) 12.axb3 b5 13.b4 Bf8 14.Bg5 Nd7 15.Qd3 Bb7 16.d5 a6 17.dxc6 Bxc6 18.Nd5 Qb7 19.Red1

19...Bxd5! 20.Qxd5 Qxd5 21.Rxd5 h6 22.Be3 Nf6 23.Rd3 d5! White was in trouble in Boskovic – Gustafsson, Gibraltar 2010.  

6...0-0 7.0-0 Nd7!? Other moves are possible of course, but I like this ECO recommendation. Black intends ...Nb6, ...Kh8 and ...f5, although he might include ...Be6 along the way, depending on how White plays.   8.Nd5 Nb6 9.Nxb6 After 9.Nxe7† Qxe7 10.Bg5 Qe8 11.Ba2 Be6 Black had no problems in Hauchard – Hebden, Cappelle la Grande 1989.   9...axb6

10.Be3! 10.h3?! Kh8! 11.d4 f5! gives Black excellent play: 12.dxe5 (12.exf5? e4!) 12...fxe4 13.exd6 exf3 14.dxe7 Qxe7

15.Re1

15...Qf6!N (15...Qh4? ran into the nasty surprise 16.Qxf3! when Black was simply a pawn down in Managadze – Mastrovasilis, Achaea 2012) 16.Qxf3 Qxf3 17.gxf3 Bxh3 Black’s superior structure outweighs White’s bishop pair.   10...Kh8 11.d4 Bg4! 11...f5 is playable, but the text move makes things easier.

12.dxe5 After 12.d5 Nb8 Black has great prospects for counterplay with ...f5.   12...Nxe5 13.Be2 Bxf3 14.Bxf3 Nxf3† 15.Qxf3 Bf6 Black was fine in Weill – Hebden, Montpellier 1988.

Main Lines after 5.0-0 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 Be7 5.0-0 0-0

It is worth castling before playing ...d6, as Black may be able to seize the initiative in the centre with ...d5 in the event that White plays too slowly. Generally speaking though, his next move will be ...d6, by which time he will be threatening ...Na5 to eliminate the light-squared bishop. The material from here can be classified according to how White chooses to safeguard this piece. We will consider c2-c3 set-ups followed by those involving a2-a4.  

The c2-c3 Plan   GAME 25 Edward Barr – Robert Zsifkovits email 2007   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 Be7 5.0-0 0-0 6.Bb3 6.c3 is slightly less accurate as it gives Black the additional option of 6...d5!?, which you may wish to investigate independently. The simplest repertoire choice is 6...d6 though, when 7.Re1 (7.Nbd2 will almost certainly transpose after a subsequent Re1; 7.Bb3 leads back to the main game) 7...Na5 8.Bb5 a6 9.Ba4 c5 10.Nbd2 b5 11.Bc2 transposes to the d2-d3 Spanish, as covered in Chapter 10.

6...d6 7.c3 Na5!? I favour this classical way of handling Black’s position. More modern ideas include ...h6 and ...Nh7, or ...Kh8, ...Ng8 and ...f5, but some of the top Chinese players continue to find ways to challenge these plans. The game continuation also bears a close resemblance to some of the Spanish lines examined in Chapter 10, making it an ideal repertoire choice for us.   8.Bc2 c5 9.Nbd2 As we will see, White’s most challenging plan involves a2-a3 and b2-b4 in the near future, so it makes sense to check whether he can get an improved version of it by playing those moves immediately.   9.a3 This has been used successfully by Tiviakov and was recommended by Emms in Beating 1 e4 e5. However, Black can use the same plan as in the main game. 9...Nc6 10.b4 a6 11.Re1 11.Nbd2 allows 11...d5! in even more favourable circumstances, as there is no pressure on e5.

11...d5! 12.exd5 Nxd5 13.h3 13.Nxe5? Nxe5 14.Rxe5 Bf6 is the point of Black’s play. 13...cxb4! Emms only mentions 13...Bf6. 14.axb4 Be6 15.Bb2 Qc7 16.Nbd2 Rfd8 In Savchak – Callow, Internet 2005, the weakness of the c3- and d3-pawns gave Black the better chances.

9...Nc6 10.Re1 Re8 11.a3 Compared with the last note, White could argue that his queenside expansion comes with better timing, as Black has been lured into spending a tempo putting his rook on e8. Nevertheless, Black still gets an excellent game by getting ready to meet b2-b4 with ...d5.   11.Nf1

This is the traditional way to play for White, but it has been known for decades that Black gets a good game by preparing ...Bf8 and ...d5. 11...h6! The immediate 11...Bf8? would allow White to take control of the d5 square with 12.Bg5 and Ne3. 12.h3 12.Ng3 Bf8 13.h3 d5 is mentioned as equal in both ECO and NCO. If 14.Qe2?! then 14...d4! gives Black a space advantage; an instructive example from here is Jakubiec – Beliavsky, Warsaw 2004.

12...Bf8 13.N3h2 d5! 14.Qf3 14.Ng4 d4 15.Bb3 Be6 16.Bxe6 Rxe6 17.c4 Rb8!?³ Vratonjic – Nikolic, Kraljevo 2015. 14...d4 15.Bb3 Na5 15...Be6!? is also fine. 16.Ba4 Bd7 17.Bxd7 Nxd7 Black is slightly better of course due to his extra space.

18.c4 a6 19.Ng3 b5 20.b3 g6 21.Ng4 Qh4 22.Bd2 Nc6 23.Rec1 Bg7 24.Qd1 b4 25.Qf3 a5 Here too, Beliavsky eventually prevailed in a strategically rich game, Nevednichy – Beliavsky, Herceg Novi 2001.   11...a6 12.b4

12...d5! 13.exd5 13.Bb3 h6! 14.exd5 Nxd5 15.Ne4 cxb4 16.axb4 Be6 is also pleasant for Black.   13...Nxd5 14.Bb2 cxb4! 15.axb4 Bg4! Black is already at least equal, and it is not clear how White should deal with the pressure on his queenside pawns.

16.h3 Bh5 17.g4 17.Ne4 Rc8 18.Ng3!? is a thematic plan, but it is not terribly effective here due to a tactical detail: 18...Bxf3 (18...Bg6 is also fine for Black) 19.Qxf3

19...Nd4! 20.cxd4 Rxc2 21.Rxe5 Nxb4 22.Ba3 Nc6 23.Rd5 White has just enough activity to make up for his dodgy pawn structure. 23...Qc8 24.Bxe7 Nxe7 25.Rc5 Rxc5 26.dxc5 Qxc5 27.Qxb7 a5=   17...Bg6 18.Nxe5 Nxe5 19.Rxe5 Bf6 20.Rxe8† Qxe8³

21.d4? This looks like a typical case of a correspondence player trying to complicate matters, relying on the fact that he is material up for the moment. 21.Ne4 Bxe4 22.dxe4 Nxb4 (22...Nxc3!?=) 23.cxb4 Bxb2 is more comfortable for Black, but White should hold fairly easily at correspondence level.   21...Bxc2 22.Qxc2 Qe2 23.Qd1 Re8 24.Qxe2 Rxe2 25.Nc4 b5! 26.Rxa6 h6 Now White has to give up some material.

27.Kf1 Rc2 28.Nb6 Nxb6 29.Rxb6 Bh4! 30.Rb8† Kh7 31.Ba3 Rxf2† 32.Kg1 Ra2 33.Ra8 Rc2 34.Rc8 Bf2† 35.Kg2 Bxd4† 36.Kf3 Rxc3† 37.Rxc3 Bxc3 38.Ke4 Twenty moves ago White grabbed an extra pawn, hoping to withstand Black’s initiative. Now the material balance is

reversed, and Black has no trouble converting the endgame.

38...Kg6 39.Kd5 Kg5 40.Kc5 Kh4 41.Kxb5 Kxh3 42.Kc4 Bf6 43.b5 Bd8 44.Bf8 Kxg4 45.Bxg7 h5 46.Bd4 h4 47.b6 h3 0–1  

The a2-a4 Plan   Instead of the Spanish-like c2-c3, White can also keep his bishop on the a2-g8 diagonal by advancing his a-pawn. We will start by examining a slightly inaccurate version of this plan, where White plays it prematurely.   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 Be7 5.0-0 0-0 6.a4 6.Re1! d6 7.a4 is the more accurate move order, and is discussed below.

6...d5! This is made possible by the slowness of White’s last move.   7.exd5 Nxd5 8.Re1 Other moves allow Black to put his king on h8 and defend the e5-pawn with ...f6. A good example of this plan is: 8.h3 Kh8! 9.Re1 f6 10.a5 a6 11.c3 Be6 Black is ready to complete development with ...Qd7 and ...Rad8. 12.Qb3  

In Movsesian – Sriram, Pardubice (rapid) 2010, Black ventured an interesting pawn sacrifice with 12...b5!?, but 12...Qd7N could also have been played, as the b7-pawn is poisoned.   8...Bg4 9.h3 9.c3 a5!? 10.Nbd2 Nb6 11.Bb5 f6 was fine for Black in Semmler – Schulz, corr. 2014.

  9...Bxf3 10.Qxf3 Nb6 11.Nd2 White can preserve his bishop at the cost of a pawn with: 11.Ba2 (11.Bb3? is punished by 11...Nd4!) 11...Nxa4 White has compensation for the pawn, but nothing more. 12.c3 was seen in Czebe – Wells, Budapest 1993, and now Black could have consolidated with:

12...Nc5!N 13.Rd1 a5 White’s compensation is in question.   11...Nxc4 12.Nxc4 f6= Black had no problems in Kuderinov – Beliavsky, Moscow 2011.  

The Right Move Order 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 Be7 5.0-0 0-0 6.Re1 d6 7.a4 This way White keeps his bishop on the active diagonal, having also prevented the ...d5 break.

7...Na5 7...Be6 is a solid alternative, while other ideas include 7...h6, intending ...Nh7-g5, and 7...Kh8, intending ...Ng8 and ...f5. The text move works fine though, and it ties in nicely with our play in the c2-c3 section, as well as in Chapter 10 when we deal with the d2-d3 Spanish.   8.Ba2 c5 9.Bg5!? This move has been favoured by Anand, a connoisseur of the ‘knight versus bishop’ imbalance.   9.Na3! is the critical move, and will be analysed separately below.   9.Nc3 Nc6 10.Nd5 is another logical continuation. 10...Nxd5 11.Bxd5 Be6 12.c3 Qd7 13.Be3 was seen in Adams – Naiditsch, Dortmund 2006, and now I like Marin’s recommendation: 13...Bf6!? Black is ready to play ...Ne7 at a moment of his choosing.

9...Nc6 10.Bxf6 If White does not make this trade immediately, Black might be tempted to retreat the knight from f6.   10.Nc3 We will follow an instructive example by the Cuban GM Bruzon. 10...Nb4! 11.Bc4?! 11.Bb3 is better, but the same idea of 11...Ne8!N is fine for Black. 11...Bg4! There is nothing wrong with the immediate 11...Ne8!?, but the text move exploits the fact that the c2-pawn is insufficiently defended.

12.Bb3 Ne8! 13.Bxe7 Qxe7 14.h3 Bh5 14...Be6 would have been fine for Black, but Bruzon decides to avoid further exchanges in order to play for a win

from a more complex position. 15.Nd5 Nxd5 16.Bxd5 Nc7 17.Bc4 Rad8 18.Qe2 Kh8 19.Qe3 f6 20.c3 d5 Black took over the initiative in De la Paz Perdomo – Bruzon Batista, Merida 2007.   10...Bxf6 11.Nc3 At this point I think Black should mimic Bruzon’s idea from the note above.  

11...Nb4! 11...Ne7 12.Nd5 Nxd5 13.Bxd5 Be6 14.Bxe6 fxe6 15.c3 gave White a nagging edge in Anand – Naiditsch, Mainz (rapid) 2009.   12.Nd5 12.Bb3 is met by 12...Be6, when Black has an improved version of the above game. If White exchanges on e6, the pawn structure will be the same but the presence of an extra pair of knights makes a big difference.   12...Nxa2 13.Rxa2 Be6 14.c3 This position was reached in Vocaturo – Mazur, Aix-les-Bains 2011, when Black exchanged on d5. I think it would have been better and more interesting to play:

14...g6!N The knight is strong on d5, but Black can play around it and aim for counterplay with ...Bg7-h6 and ...f5. Peter Svidler recommended something similar in a variation of the 6.d3 Spanish, which you can read about in Chapter 10.  

The Critical Na3 Plan 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3 Be7 5.0-0 0-0 6.Re1 d6 7.a4 Na5 8.Ba2 c5 9.Na3! This is the most challenging way for White to develop (9.c3 Nc6 10.Na3! is the same thing). The idea is to send the knight towards e3 without obstructing the c1-bishop along the way, as would occur after 9.Nbd2.

9...Nc6 10.c3 h6 A typical useful move. Black prepares the thematic regrouping with ...Re8, ...Bf8 and ...Be6, without running into a

disruptive Ng5 or Bg5.   Lokander mentions that he rejected the 7...Na5 variation because he did not like Black’s position after 10...Be6 11.Bxe6 fxe6 12.Nc4, but there is no reason for Black to trade bishops this early.   11.Bd2! Saric recommends this move in his Chess24 lectures, and I agree that it is White’s best. Since the g5-square is no longer available to the bishop, the next best choice is d2. While e3 might seem more natural, this square is reserved for the knight, which can get there via c4 or c2. The bishop on d2 also serves the important function of supporting a queenside initiative with b2-b4 at some point. 11...a6! I regard this little move as an important nuance. Since White’s last move introduced the idea of queenside play, Black prepares to fight back in that area of the board. If left unchallenged, he will continue with ...Rb8 and ...b5.   Saric only mentions 11...Be6?! in his video lecture. This move would have been fine if White’s knight had already gone to c4 or c2 but, as things stand, 12.Bxe6 fxe6 13.Nc4! is good for White.   My initial choice was 11...Re8 but after 12.Nc2 Bf8 13.Ne3 Be6 14.Bxe6 Rxe6 White has a strong continuation in 15.c4!, closing the centre in order to attack on the kingside with f2-f4 without having to worry about a ...d5 break. After 15...a5 16.Rf1! Re8 17.Ne1 g6 White can play 18.f4, as in Olofsson – Busemann, corr. 2014, or 18.g3N followed by N1g2, both of which are quite troublesome for Black.   12.Nc4 This is the most natural move, but I also considered two alternatives:   12.Nc2 Be6 is fine for Black, as White’s knight will have to lose precious time to reach the optimal c4-square.   12.h3 Rb8 13.Nc2 (After 13.Nc4 b5 14.axb5 axb5 15.Ne3 Black should play: 15...Re8!„ This move stops 16.Nh4 due to 16...Nxe4! when 17.Ng6 does not gain a tempo on the rook. Black continues with ...Be6 and ...Qd7 and equalizes gradually, as my IDeA analysis has shown.) 13...Be6 14.Bxe6 fxe6 I found a blitz centaur game from this position, which continued:

15.b4 cxb4 16.cxb4 b5 17.axb5 axb5 18.d4 Qd7= Black had no particular problems.

  12...Be6 12...b5 is also possible.   13.h3 13.b4 b5! is similar to the main line.   13...Re8 14.Qb1 Bf8 15.b4

15...b5! The usefulness of ...a6 is on full display. Black has carried out his normal kingside regrouping, while preventing White from taking control over the queenside.   16.Ne3 Bxa2 17.Rxa2 cxb4 18.cxb4 We are following Udeshi – Panchanathan, New Delhi 2013. Here Black should have played:

18...bxa4!N Intending ...d5 in order to target the e4- and b4-pawns. A possible continuation is:   19.Qc2 Rc8 20.Qxa4 d5 21.Nxd5 Nxd5 22.exd5 Qxd5 23.Rc2

23...Nd4! 24.Nxd4 Qxd4 25.Rxc8 Rxc8 26.Qxa6 Rc2 Black will soon pick off the b-pawn, leaving an equal position.   Conclusion   In this chapter we began our coverage of the Two Knights by considering both 4.d4 and 4.d3. After 4.d4 exd4, Morphy’s original idea of 5.Ng5 is dangerous only for White after 5...d5! 6.exd5 Qe7†!. A better version of White’s

idea is 4.d4 exd4 5.0-0 but, after the normal sequence of 5...Nxe4 6.Re1 d5! 7.Bxd5 Qxd5 8.Nc3, we saw that 8...Qd7! gives Black a healthy position while sidestepping any long theoretical lines.   4.d4 exd4 5.e5 is a much more serious variation. The widely recommended 5...d5! 6.Bb5 Ne4 is my choice, but after 7.Nxd4 I strongly advise you to study 7...Bc5! (instead of 7...Bd7). White must walk a narrow path to avoid becoming worse, and even after best play he gets no more than a level endgame.   Whereas 4.d4 initiates an early conflict, 4.d3 leads to a slower-paced struggle. 4...Be7 is a good answer, which also fits in well with the Spanish set-ups examined later. If White chooses the Spanish-style plan of c2-c3, Re1 and Nbd2, Black gets a good game with ...Na5 and ...c5, followed by ...Nc6 in order to prepare ...d5. Black’s central expansion may be further supported by ...Re8, ...h6 and ...Bf8 if needed.   White’s other main plan after 5.0-0 0-0 is to play 6.Re1 (to prevent a quick ...d5) 6...d6 7.a4, in order to keep his bishop on the a2-g8 diagonal. Once again Black plays in the style of the Closed Spanish with ...Na5 and ...c5. It is worth paying particular attention to the method employed Bruzon Batista, as mentioned in the notes. In the event that White plays Bg5xf6 to exploit the d5-square, it is worth remembering that ...Nb4 is a useful resource to eliminate the second white bishop, after which Black can play around the strong d5-knight and aim for kingside counterplay with ...f5. We finished up by checking the quite challenging Na3 plan, but Black should be fine as long as he plays a timely ...a6 in order to fight back on the queenside.

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5   4...d5 5.exd5 Na5   The Kieseritzky Variation 6.d3 Game 26, page 170   6.Bb5† c6 7.dxc6 bxc6 8.Qf3 Be7! 9.Bxc6†?! Game 27, page 177 9.Bd3! Game 28, page 180 8.Be2 h6 Steinitz’s 9.Nh3 Game 29, page 185 9.Nf3 – The Main Line page 188 8.Bd3! – The Modern Trend 8...Nd5! page 191 9.Ne4!? page 191 9.Nf3 Game 30, page 192 The Critical 9.h4! page 197  

 Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up

to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How does Black defuse Bronstein’s idea? (page 170)  

How should Black proceed? (page 187)  

Which rook should be put on e8? (page 173)

What is Black’s plan to obtain compensation? (page 190)  

How does Black take over the initiative? (page 178)  

What powerful move should Black play? (page 184)

Find a clean solution for Black. (page 194)  

Which is Black’s most accurate move? (page 200)

  1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5! This is the most critical way for White to meet the Two Knights Defence.

4...d5! 4...Bc5!? may not be so bad, and I can recommend The Alterman Gambit Guide – Black Gambits 2 for those who are interested in it. I prefer to stick to the main line though.   5.exd5 Na5! 5...b5 and 5...Nd4 have occasionally been tested by strong players, but these lines should only be tried as surprise weapons.   5...Nxd5?! is a natural move, but both 6.Nxf7!? and 6.d4! give Black a dangerous position to defend.   6.Bb5† 6.d3 is the Kieseritzky Variation, and is not such a bad sideline. 6...h6 7.Nf3 e4 8.Qe2 Nxc4 9.dxc4 Bc5 10.Nfd2 0-0 11.0-0 Bg4 12.Qe1 Qd7 13.Nc3 Rae8! is my recommendation, further details of which can be found in the first illustrative game of the chapter.   6...c6 My initial intention was to recommend 6...Bd7!?. Unfortunately, after 7.Qe2 Be7 8.Nf3! Black’s position is miserable in the sense that he has virtually no winning chances, and I was not even able to find full equality.   7.dxc6 bxc6

This is our main tabiya for the 4.Ng5 complex. There are three logical moves, all of which have been tested extensively:   8.Qf3 threatens to win a second pawn on c6, and Black happily donates it with 8...Be7!.   8.Be2 h6 gives White a choice between 9.Nh3, which was the favourite of Steinitz and Fischer, and 9.Nf3, which is objectively stronger. In both cases I like putting my bishop on d6, where it exerts annoying pressure on White’s kingside.   8.Bd3 is the recent trend, which I propose to meet with the natural 8...Nd5. From there we will deal with 9.Nf3 before moving on to the more critical 9.h4!.

The Kieseritzky Variation GAME 26 Marcel Griesmann – Robert Dubois Correspondence 1990   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.d3 Bologan calls this the Morphy Variation but other sources name it after Kieseritzky. The latter makes more sense to me, as the Russian-turned-French player used it a few times before the great American prodigy tried it. Besides, Kieseritzky was one of the leading players of his time, and deserves to be known for something other than being on the losing side of the Immortal Game. White’s last move has been criticized by many sources, both old and new, but undeservedly so. Over the years, such great fighting spirits as Tartakower, Korchnoi and Bronstein have used it successfully, while more recently the French correspondence Senior International Master Patrick Spitz has shown repeatedly that White’s position is not worse, despite the initial engine evaluation. Considering that White has an extra pawn and no real weaknesses, I would argue that the Kieseritzky Variation could be used as a serious practical weapon, although I am happy to say I found an excellent solution for Black.

6...h6 7.Nf3 e4 8.Qe2 8.dxe4?! Nxc4 9.Qd4 Bronstein won a famous game with this experiment against Ernst Rojahn of Norway in the 1956 Moscow Olympiad, but later it was discovered that Black can essentially refute the sacrifice with: 9...Nd6! 10.Nc3

10...Nfxe4! 11.Nxe4 Qe7! This temporary counter-sacrifice is a key idea to remember. 12.0-0 Nxe4 13.Re1 f5 14.Nd2 14.Ne5 threatens Ng6, but 14...Qf6 defends and leaves Black a piece up for scant compensation, as in Bilek – Zika, Czech Republic 1995. After the text move I found an improvement over Arens – Ludigk, corr. 1996.

14...Kd8!N Black takes the king off the e-file and virtually forces White to exchange on e4, as otherwise he will be hit by ...Qc5, ...Qb4 or ...Qf6. 15.Nxe4 fxe4 16.Rxe4 Qf6! 17.Qxf6† If White avoids the exchange of queens, Black will play ...Bd6 and ...Bf5 with a strong attack. The king can step up to d7 for a brief moment to enable the a8-rook to join the party.

17...gxf6 18.Rf4 f5

Black will follow up with ...Bg7 and ...Bd7, and should eventually convert his extra piece.   8.Nd4 also exists and it is not so bad. I think Black’s best is 8...Bc5, as our Breyer hero Smejkal has played. After 9.Be3 though, in Trapl – Smejkal, Prague 1986, the Czechoslovakian GM rushed with 9...Ng4. I think it would have been better to play:

9...0-0! Black intends to play simple moves like ...exd3 and ...Nxd5, with an excellent position. The open e-file might come into play if White gets careless and postpones castling.   8.Bb5† c6 9.dxe4 cxb5 10.Qd3 is a variation on Bronstein’s piece sacrifice, which used to be Spitz’s pet line until a score of ½/3 convinced him to look elsewhere. Play continues: 10...a6 11.h3 Bc5 12.0-0 0-0 13.a3

13...Nc4! (13...Ba7 14.b3 Re8 also favoured Black somewhat in Spitz – Sinkovics, France 2007) 14.b3 Nd6 15.e5 Bf5 16.Qd1 Nfe4 17.exd6 Qb6 18.Qe1 Rfe8 19.b4 Bxd6 20.Qd1 Bc7 21.a4 Bxh3 22.gxh3 Qg6† 23.Kh1 Qh5 Black has handled the position brilliantly, returning his extra piece and sacrificing another one to launch a vicious attack. It is worth including the last few moves of this beautiful game:

24.Kg2 Rad8 25.Ra3 Rxd5 26.Rd3 Qg6† 27.Kh1 Rh5 28.Nh2 Bb6 29.Rf3 Qc6 0–1 Spitz – Martin Clemente, corr. 2007.

8...Nxc4 9.dxc4 Bc5 10.Nfd2 0-0 11.0-0 11.Nb3 This looks a bit strange to me, but the engines see nothing greatly wrong with it. 11...Bg4 12.Qf1 Bb4†! The white king is the target, so Black provokes the following pawn move to weaken the d3-square.

13.c3 Other moves allow Black to open the centre with lethal consequences: 13.Bd2? Bxd2† 14.N1xd2 Re8 15.h3 e3 and White was busted in Steiger – Loeffler, Lauda 1986. 13.N1d2? Re8 and 13.N3d2? Re8 are also bad for White, as ...e3 is coming quickly. 13.Nc3 c6 14.h3 Bh5 15.g4 Bg6 16.dxc6 bxc6 17.Bd2 e3! was Luckis – Keres, Buenos Aires 1939. Once again,

Black played simple moves to open the position and crush his opponent. 13...Bd6 14.Be3 Bf5 Black guards the e4-pawn in order to prepare ...Nd7(g4)-e5-d3†.

15.N1d2 Re8 Black brings another piece into play before committing his knight. 16.h3 Nd7 17.0-0-0 Ne5 18.Kb1 a5 In Llorach Gracia – Vertongen, corr. 2014, Black had a strong initiative. Later he played ...a4 and opened lines with ...c6. When White went into defensive mode on the queenside, Black took over in the centre and eventually prevailed.

11...Bg4 12.Qe1 Qd7 13.Nc3! 13.Nb3? This natural-looking move is a big mistake due to: 13...Bf3!!

The engines have difficulty spotting this idea, but it is an important attacking plan you should know about.

14.Bf4 14.h3 doesn’t help: 14...Qf5 15.Nxc5 Qg6 16.g3 In Alves – Fernandes, email 2010, the most accurate continuation would have been 16...Qh5!N 17.h4 Qf5 18.Kh2 Ng4† 19.Kg1 Ne5 20.Kh2 g5! and Black is winning. 14...Qg4 15.Bg3 Nh5 16.Nxc5 Nf4

17.Qc3 This is forced, as 17.Nd2 Qh3! forced White to resign in Bakos – Le. Toth, Hungary 2011. 17...Ne2† Black picked up the queen and was clearly better in Wydornik – Szulc, email 2003.

13...Rae8!! This is the right rook, but it takes some effort to understand why this is the case. We will begin by analysing the alternative rook move.   13...Rfe8 This seems more natural, but Patrick Spitz discovered a nice way to deal with it. 14.Ndxe4! Bologan doesn’t like this move and Lokander doesn’t mention it at all, but I would not want to face it in an overthe-board game. The Moldovan GM gives 14.Nb3 as White’s best, but he overlooks the strong retort 14...Bf3!. We will see the same idea in our main game as well. 14...Nxe4 15.Nxe4 Qf5 16.Nxc5 Rxe1 17.Rxe1 White’s idea to trade the queen for an approximate material balance is not bad at all, despite the initial disapproval of the engines.

17...Qxc2 18.b3 b5 19.h3 Bf5 20.Bf4 bxc4 21.Na6! Another key move. 21...cxb3 22.axb3 Qxb3 23.Nxc7 Rd8 24.d6 In the high-level correspondence games Spitz – Butnorius, corr. 2011, and Spitz – Rubinas, corr. 2011, it was demonstrated that this position is essentially a draw. In a practical game, of course, any result would be possible.

24...a5!?N I checked this move to see if Black’s play could be improved. I will give a few more moves of my analysis, as it is relevant when comparing the two rook moves at move 13. 25.Re5! 25.Rxa5? allows 25...Qb4 with a triple attack! 25...Bd7 26.Rexa5 Bc6 Black would like to attack g2, but the computer defends rather easily.

27.Bg3 Qd3

28.Re1! Stopping ...Qe4. The analysis goes on, but the rest is not relevant to our discussion. The chances are balanced, and in a practical game any result would be possible.

14.Nb3 The big question of course is what happens after 14.Ndxe4. We will follow the same forcing sequence as above, up to the point where the difference becomes clear: 14...Nxe4 15.Nxe4 Qf5 16.Nxc5 Rxe1 17.Rxe1 Qxc2 18.b3 b5 19.h3 Bf5 20.Bf4 bxc4

21.Na6 (21.Rec1 Qb2 22.bxc4 might be better but after 22...Re8, planning ...Re2 or ...Re4 according to circumstances, Black is at least equal) 21...cxb3 22.axb3 Qxb3 23.Nxc7  

Finally the difference becomes clear: Black does not have to lose a tempo moving the rook from a8 and can instead play 23...a5! 24.d6 a4!, when the advancing a-pawn limits White’s active possibilities a great deal. Black’s material advantage is felt much more here and he might even be winning with precise play.

14...Bf3! This is a thematic idea in these positions and you should always be on the lookout for it.   15.Bf4? An understandable reaction in pre-computer times, but it loses in a typical way for this line. Other moves would have forced Black to work harder for the full point:   15.h3 Qf5 16.Ne2! (16.Nxc5? is still not possible: 16...Qg6 17.g3 Qf5 18.Kh2 Ng4† 19.Kg1 Qh5 20.h4 g5–+) 16...Bd6 17.Ng3 Qg6 White faces the lethal threat of ...Nh5. The computer’s best defensive attempt is:

18.gxf3 exf3 19.Qd1 Bxg3 20.Qxf3 Bxf2† 21.Kh2 Bb6 22.c5 Bxc5 23.Nxc5 Qxc2† 24.Kh1 Qxc5 25.Bxh6 Qxd5 White faces a struggle to survive the endgame.  

15.Ne2!N This is White’s best defensive try.

15...Bd6 16.Nbd4 16.Ng3 also enables White to survive with accurate play: 16...Bg4 17.Be3 (17.Nd4?! c6! [17...c5 is not good this time due to 18.Nb5!] 18.dxc6 bxc6 19.Be3 h5 20.h4 Nh7!‚ White is in trouble as ...f5 comes) 17...h5 18.h4 b5 19.Qc3 bxc4 20.Nc5 Qe7 21.Qxc4 Bc8 22.Rae1 Ng4 23.Bg5 f6 White is not off the hook yet. 16...c5 17.dxc6 This is forced, as 17.Nb3? Bxh2† 18.Kxh2 Qg4 wins. 17...bxc6

18.Qc3! 18.h3 c5 19.Nb5 Bb8!? 20.Ng3 (20.Be3? Nh5! makes room for either the rook or the queen to come to g6 via e6; this, along with ...f5-f4, will lead to a fierce attack) 20...Nh5 21.Nc3 Qe6 22.Nxh5 Bxh5 23.f4 exf3 24.Qxe6

Rxe6 25.g4 Bg6 26.Rxf3 Be5 27.Be3 Bxc3 28.bxc3 Rc8³ White has avoided being mated but he faces a miserable endgame. 18...Be5! An important resource. Best play continues: 19.gxf3 c5 20.f4 cxd4 21.Qg3 d3 22.cxd3 Nh5 23.Qh4 exd3 24.Qxh5 dxe2 25.Qxe2 Bxf4 26.Qd1 Qc7 27.Bxf4 Qxf4 28.b3 Rd8 29.Qh5 Rd2 Black keeps a bit of pressure although White should be able to hold.

15...Qg4 16.Bg3 Nh5! We have already seen the same idea in the note to White’s 13th move. Credit to Monsieur Dubois for finding this amazingly effective attacking plan in the pre-computer era.   17.Kh1 17.Nxc5 Nf4 18.Nd1 (otherwise ...Qh3!! ends the game) 18...Qh3 19.Ne3 Nxg2 20.Qd2 f5 followed by ...f4 wins.

17...Bxg2†! 18.Kxg2 Qf3† 19.Kg1 Nxg3 20.hxg3 Qxg3† At this point White resigned, rather than waiting for 21.Kh1 Qf3† 22.Kg1 Re5. I find this an incredibly valuable game in terms of dealing with the Kieseritzky Variation. 0–1

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 As mentioned earlier, there are three main ways in which White may deal with the attack on his bishop. We will start by considering the indirect defence using the queen.  

8.Qf3   GAME 27 Sergey Turko – Nikolay Trofimov Correspondence 2012   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6

8.Qf3 Be7! As you will see in the 8.Be2 section, as well as in Game 30, I generally like to put my bishop on d6 in the variations where White retreats his bishop and Black has to search for compensation. The reason for this is clear: Black is targeting the sensitive h2-pawn, which will lack the natural defence of a knight on f3. Here though, the bishop must avoid the d6-square for tactical reasons. 8...Rb8 is a playable alternative, avoiding a potential fork on c6 and thereby preparing ...Bd6 on the next move. However, I prefer to save time and develop the bishop immediately.   9.Bxc6†?! White greedily accepts the second pawn. I will go so far as to declare this move an objective mistake, not to mention an excessively risky policy in a practical game. The superior 9.Bd3! is seen in the next game.   9...Nxc6 10.Qxc6† Bd7

It was essential to keep the bishop off the d6-square earlier so that it would not hang at this point.

11.Qf3 11.Qa6 places the queen far away from the kingside and also gives Black the possibility of gaining time with ...Rc8c6. A nice example continued: 11...0-0 12.Nc3 h6

13.Nf3 (13.Nge4?! Nxe4 14.Nxe4 Rc8! 15.Nc3 Rc6 is great for Black, as 16.Qxa7 is refuted by 16...Bc5 17.Qb7 Rf6, followed by ...Bc6 and possibly ...Qa8, when White’s kingside falls apart) 13...Qc7 14.d3 Rab8 15.0-0 Bc6³ In Gulbis – Schuster, corr. 2014, White was struggling to develop his remaining pieces, while Black had many ideas like ...Bxf3, ...Rfc8 and ...Rb6.   11...0-0 12.d3 Rc8 13.Nc3 So far so good. Having made all the obvious developing moves, Black has to find a plan to pursue his initiative, and Trofimov finds an excellent one.

13...Ng4! Black gains a tempo against the knight on g5, while preparing to launch the f-pawn as well as to swing the c8-rook to the kingside.   14.h3 No better is: 14.Nh3 Rc6 15.0-0 Rg6 Threatening ...Bc6.

16.Qe2 f5 With the simple but brutal idea of ...f4-f3. 17.Kh1 17.f3 runs into 17...Bc5† 18.Kh1 Nxh2! 19.Kxh2 f4 and ...Qh4 is coming.

17...f4 18.f3 White had to stop ...f3, but Black has something else in store.

18...Nxh2! 19.Kxh2 Rf5 20.Ne4 Rh5 21.Nef2 Rg3! 22.Re1 White makes a last-ditch defensive try, preparing to meet 22...Bxh3 23.Nxh3 Qd7 with 24.Qf1, although even here Black has plenty of initiative after 24...Qf5. However, there is something even stronger.

22...Bc5! And in view of the unstoppable mating attack with ...Qh4 coming, White called it a day in Butov – Sychov, corr. 2012.   14...Nh6 15.Nge4 Nf5 Now White has to worry about ...Nh4 and ...Nd4, as well as the ever-present possibility of ...Rc6-g6.  

16.0-0 16.Qd1 Nh4 17.g3 Nf5 18.Nd5 Qa5† 19.Nec3 Bd6 also gives Black a strong initiative. An instructive game continued:

20.Bd2 Nd4 21.Ne3 Qb6 22.Ncd5 Qxb2 23.c4 Ba4 24.Qc1 Qb7 25.Rb1 Qd7 26.Bc3 Bc6 27.Qd1 Rce8 With ...f5-f4 on the way, White found nothing better than:

28.0-0 Qxh3 29.Bxd4 exd4 30.Qg4 Hoping to find solace in an endgame, but White has serious problems there too. 30...Qxg4 31.Nxg4 Re2 32.Rfe1 Rfe8 33.Ndf6†!? White would also have had a hard time defending with two knights against two bishops. 33...gxf6 34.Nxf6† Kf8 35.Nxe8 Rxe8 36.Rxe8† Kxe8 Black’s bishops proved too strong for the rook in Kouba – Shishkov, corr. 2009.

16...Nd4 17.Qd1 Rc6 18.Be3 Rg6 Black’s initiative is already close to decisive.   19.Bxd4 19.Kh1 f5 20.Nd2 f4 21.Bxd4 exd4 22.Nce4 it another attempt to defend, but after 22...Qc8 23.Qf3 Rh6 White cannot prevent a powerful sacrifice on h3.   19...exd4 20.Ne2 f5

21.Nf4 21.N4g3 f4 followed by ...f3 was even worse.  

21...fxe4 22.Nxg6 hxg6 23.dxe4 Bb5 24.Re1 Bc5 Materially White is doing well, but ...d3 is a terrible threat, so White found nothing better than simplifying to an unpleasant endgame.

25.c4 dxc3 26.Qb3† Kh7 27.Qxb5 Bxf2† 28.Kh1 Bxe1 29.Rxe1 Qd2! An important detail.   30.Qe2 Qxb2µ Even with the help of his engine collection, White found it impossible to defend against the monstrous pawn on c3. The remaining moves were:

31.e5 Rc8 32.Qg4 Qf2 33.Rc1 Qe3 34.Rf1 Qd3 35.Rc1 Rc4 36.Qd1 Qxd1† 37.Rxd1 g5 38.Rc1 Kg6 39.e6 c2 40.Kg1 Kf6 41.Kf2 Kxe6 42.Ke2 Kd5 43.Kd3 g4 44.hxg4 Rxg4 45.Kxc2 Rc4† 46.Kd2 Rxc1

0–1   The above game, along with those mentioned in the comments, makes a strong case against White’s decision to capture a second pawn on c6. Let’s now see what happens if he declines it by retreating his bishop.   GAME 28 Torgny Skogfeldt – Kari Kuosmanen Correspondence 2013   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Qf3 Be7 9.Bd3! Declining the second pawn is safer, and just a better move overall. Black still has nice prospects though, as his pieces are active and it will take White a long time to develop his queenside pieces.

9...0-0 10.Nc3! This is the most accurate move order.   10.0-0 is more popular but less challenging, as White loses out on some of the benefits of a timely h2-h4 in some lines. 10...h6 11.Ne4 Nd5 12.Ng3 Nb4 Black follows exactly the same course as in the main game, with White’s options being slightly reduced.

13.Nf5 It makes sense to play this before Black gets active with ...f5. 13...Bg5! After this logical sequence, it is worth considering three main possibilities:   a) 14.Re1?! Nxd3 15.cxd3 was seen in Kaidanov – Onischuk, Tulsa 2008. Onischuk proposed the improvement:

15...Bxf5!N 16.Qxf5 Nb7! Black does not have to defend the e-pawn, as 17.Qxe5 Qd7 leaves him with more than enough compensation for the two sacrificed pawns. He can put his rooks on e8 and d8, and his knight will make its way to d4 via d6 and f5 (or b5).   b) 14.h4?!N This is initially one of the top choices of the engines, but Black gets the upper hand after: 14...Nxd3! 14...Bxh4? 15.Qg4! Bxf5 16.Bxf5 allows White to claim an advantage because of his bishop pair and better structure.

15.cxd3 15.hxg5 Nf4! favours Black. Black can restore material equality by taking on f5 and then h4, but an even stronger continuation is:

15...Bf4! 16.Ne3 16.Nxg7? Kxg7 17.g3 Kh8! 18.gxf4?? Rg8† wins for Black. 16...Qxh4 16...Bb7!? 17.d4 f5 also looks promising. 17.g3 Qh3 18.b4 Otherwise White will be crushed by ...c5 followed by ...Bb7.

18...Nb7 19.Qxc6 Bg5 20.Nc3 Nd8 21.Qg2 Qxg2† 22.Kxg2 Ne6³ White’s structure is a mess. c) Preserving the bishop with 14.Be4 may well be White’s best, but 14...g6 still gives Black excellent prospects. The

following forcing line is critical:

15.a3 gxf5 16.axb4 fxe4 17.Qg3 (After 17.Qxe4?! Nb7 White does not have enough for the piece) 17...Nb7 18.h4 f5 19.hxg5 f4 20.Qh4 hxg5 21.Qh6

21...Bf5! (21...Nd6 essentially forces White to take the perpetual with 22.Qg6† but Black can play for more) 22.Qxc6 Rf7 Black has excellent compensation for his missing pawn, with great prospects on the kingside.   10...h6 Bologan proposes 10...Rb8 but I think White stands better after 11.b3!N, delaying castling and having similar ideas as in the main line below. As we will see, b2-b3 is likely to be played at some point anyway, while there is no guarantee that the rook will be best placed on b8, especially at this early stage.   10...g6 would have been playable if White had castled instead of putting his knight on c3 on the last move. In this

position, however, it allows White the dangerous option of 11.h4!.

11.Nge4 Nd5 12.Ng3 Nb4 13.Nf5 Bg5 14.h4! White takes advantage of the fact that he has yet to castle.

14...Nxd3† 15.cxd3 Be7 15...Bf4? is no good here due to 16.Nxg7! Kxg7 17.g3, as there is no counterattack along the g-file.   16.b3 Bxf5 17.Qxf5 c5 18.Bb2 Nc6 This is a typical situation in the critical 8.Bd3 variation. White has the worse structure, but is a pawn up for the time being. His remaining pieces are decently placed, and he can target the pawns on c5 and e5. Black has more space though, and easy play for his pawn. Engines tend to evaluate these positions as slightly better for White, but Black has not experienced any problems in correspondence play.

19.Rc1 The most natural and ‘human’ move.   19.Kf1!? is evaluated as White’s best by the engines. The chances of encountering such a move over the board are pretty low, but I will mention a few possibilities just in case:   a) 19...Nd4 is not quite as strong as in the main game, but it is still playable: 20.Qxe5 Bf6 21.Qd5 Nc2 22.Rb1! (In the event of 22.Rc1 Nb4 Black easily has enough compensation) 22...Nb4 23.Qc4 (23.Qxd8 Rfxd8= is fine for Black) 23...Nxd3 24.Rh3 (24.Ba1 Re8 gives Black plenty of play for the pawn; the immediate threat is ...Nxf2!) 24...Nxb2 25.Rxb2 Bd4! It is obvious that Black has some compensation for the pawn, but is it enough? My deep computer analysis indicates that the position should eventually be a draw, but I guess Black is the one who must be more accurate.   b) 19...Qc8!? is a safe way to play. White cannot really avoid the queen exchange, as in that case a rook comes quickly to d8 and the queen to a6. 20.Qxc8 Raxc8 21.Rh3 f6 22.Na4 Nb4 (22...h5!?; 22...Rfd8!?) 23.Bc3 Rfd8 24.Ke2

24...h5! Black fixes the h4-pawn on a dark square and managed to draw quite easily in Sanner – Roques, Internet 2011, after White exchanged on b4. More critical would have been 25.Nb2!, maintaining the tension, although Black’s compensation is fine after 25...Rd5, intending to double on the d-file.   c) The flexible 19...Rc8!?N might be best of all:

Black still has in mind the ...Nd4 idea (after 20.Kg1, for example, it is playable) but he also prepares to bring the rook into play, as seen after 20.Rc1 Rc7! followed by ...Rd7.

19...Nd4! This powerful move offers Black at least equal chances.   20.Qh5 This looks like a strange square for the queen, but at least it cannot be attacked there.   20.Qe4 just invites a timely ...f5. 20...Re8 21.Rh3 Bf8 22.Kf1 Qa5 23.Qg4 Rad8 24.Re3 f5 25.Qh5 Qa6= Black had at least enough play for the pawn in Schmid – Nichols, email 2010.   20.Qxe5N looks risky but it is not as bad as it looks: 20...Re8 21.Kf1 Bxh4

22.Qh5 (22.Qxc5? asks too much of White’s position and 22...Rc8µ leaves him in trouble; 22.Qd5 Bg5 23.Qxd8 Raxd8 24.Ne4 Ne6 is more than okay for Black) 22...Bg5 23.Ne4 Ne6 24.Qf3 Qa5!? Black will get his pawn back pretty

soon.

20...f5 21.Rh3 Rc8 22.Kf1 Rc6 Black is more than okay and can hope to put some pressure on his opponent.   23.Na4 Qd5 24.Re1 Re6 Black is ready to build patiently with ...Kh7, ...g6 and so on.

25.Bxd4 cxd4 26.Nb2 White responds by trying to fit in Rf3 and Nc4. 26...Qa5! Black does not pass up the opportunity to restore material equality.  

27.Nc4 Qxa2 Black is close to being better, but White finds a forcing continuation with a draw at the end of it.

28.f4 e4 29.dxe4 fxe4 30.Qd5 Rxf4† 31.Rf3 exf3 32.Qxe6† Kh7 33.Qe5! 33.Qxe7? fxg2† 34.Kxg2 Qc2! followed by ...Qd3 or ...Qf5 produces a winning attack.

33...Rf7 34.Qe6 Rf4 35.Qe5 ½–½   That concludes our coverage of 8.Qf3. We will now turn our attention to 8.Be2, which has been White’s most popular continuation. Black responds with 8...h6, leading to another division according to which retreat square White chooses for his knight.

Steinitz’s 9.Nh3 GAME 29 Vassily Ivanchuk – Alexander Beliavsky Dortmund 1998   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Be2 h6 9.Nh3!?

This move was first played by Anderssen, but it was made popular by Steinitz and later approved by Fischer. White’s idea is simple: rather than having his knight chased around after 9.Nf3 e4, he puts it on a more stable square. He is not worried about the damage to his pawn structure that could occur after ...Bxh3, because he will have the bishop pair and his king will feel safe enough on the light squares. If Black ignores the knight, hoping it will prove to be misplaced on h3, White can eventually relocated it to a better square after 0-0, Kh1 and Ng1. In the meantime, Black has no obvious targets and will have to find a way to prove his compensation.   9...Bd6! In the 4.Ng5 Two Knights, putting the bishop on d6 is generally a good way for Black to play against anything apart from 8.Qf3. On this square the bishop is both active and safe, and it allows the c6-pawn to advance in order to provide the a5-knight with a route back to the centre.   9...g5 is recommended by Davies, and 9...Bc5 by both Bologan and Lokander, but somehow they both seem unnatural to me, and my analysis indicates that White can hope for a small edge against both of them.   10.d3 0-0 11.Nc3 Nd5

12.Bd2 White decides to prevent any damage to his queenside structure. The following moves have also been tried:   12.Ne4 is the only line given in NCO. 12...Bc7 Now Black is ready to play ...f5 and ...Qh4 with a strong attack, so White should hurry to create counterplay: 13.c4 Bxh3 14.gxh3 Nf4 15.b4 Nb7 16.Bxf4 exf4 17.0-0 White appears to be okay, at least until you notice the following manoeuvre:

17...Nd6! 18.Bf3 Nf5 19.Nc5 Qd6 20.Na6 Rad8 21.Re1 Nd4 White was in trouble in Tima – Schmezko, email 2000.   12.Ng1!? I encountered this move in a recent correspondence game. I have already mentioned the plan of 0-0, Kh1 and Ng1 to reroute the knight, so it makes a certain amount of sense to carry out the manoeuvre before castling. This way White saves a tempo with his king, but he also commits his knight to the f3-square rather early.

12...Nxc3 13.bxc3 Qc7 14.Nf3 Be6 15.0-0 Rad8

I see no reason for Black to rush things; he is better developed, he has more space and his pawn structure is no longer any worse than White’s (apart from the fact that he has one pawn less). 16.Nd2 f5 17.g3 c5 17...e4 was also fully equal according to my analysis. However, I wanted to introduce the possibility of ...c4, which my opponent was quick to prevent. 18.c4 e4! The game Rashitov – Ntirlis, corr. 2015, has reached this position at the time of writing. I do not believe Black is at all worse; a possible continuation is:

19.f4 Bf7 20.Bb2 Rfe8 21.Nb3 Nc6 The plan is ...Bf8 and ...Nd4. White’s extra pawn is virtually meaningless and I would choose Black’s position any time.

  12.0-0 This natural move has been played by Short. 12...Nxc3 12...Rb8 13.Kh1 Nxc3 14.bxc3 was seen in Short – Wedberg, Malmo 2002, one of the games that enriched our understanding of this variation. The experienced German GM played the natural 14...Be6?!, but after 15.f4! he found nothing better than exchanging on h3, losing a tempo and being slightly worse. The logical conclusion is that many other 14th moves are better and allow Black to remain about equal. 14...c5N, 14...Nb7N, 14...Qe7N and 14...Qc7 are all sensible. 13.bxc3

13...Qh4 13...Nb7!?N should also be fine for Black. 13...Qc7?! can be compared with my game against Rashitov, as referenced above. Unfortunately White has the annoying 14.Bg4! f5 15.Bh5, when he is threatening f2-f4. Black can prevent this with 15...g5, but then 16.f3 prepares Nf2, when White may seize the initiative on the kingside with g2-g4. 14.Kh1! A good defensive move, but Black does not have to hurry to take on h3. 14...Re8!„ Black’s last move prepared a quick ...e4, with good counterplay in De Bari – Salvatore, email 2011. (Please note that 14...Bxh3? 15.gxh3 Qxh3 would have backfired after 16.Rg1!, when it is White who has all the chances on the kingside.)

12...Rb8! Black wins a tempo. Obviously White can defend the b-pawn easily enough, but Black’s idea is to swing his rook via one of the squares on the b-file towards the centre or kingside.   13.b3 13.Rb1 Nb7 is similar.   13...Nb7 14.Ng1 Nc5 15.Nf3

15...Nxc3 If Black wishes to play more ambitiously then 15...Ne6!?N intending ...f5 could be the way to go.  

16.Bxc3 e4 17.Nd2 exd3 18.cxd3 Bf5= Black is poised to win back the d3-pawn, resulting in a drawish position.   19.Nc4 After 19.0-0 Nxd3 20.Nc4 Rb5!?„ Black can plant the rook on d5.   19...Re8 Black could have withdrawn his bishop to c7, but he doesn’t mind trading it for the strong enemy knight.   20.0-0 Nxd3 21.Ba5 Qd7 22.Bxd3 Bxd3 23.Nxd6 Qxd6 24.Bc7 Qxc7 25.Qxd3

At this point a draw was agreed. Black has the marginally worse pawn structure, but after something like 25...Rbd8 26.Qc2 c5 27.Rac1 Qa5 it is obvious that he is in no danger. ½–½  

9.Nf3 – The Main Line 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Be2 h6 9.Nf3 e4 10.Ne5 This is one of the most important variations of the Two Knights.

10...Bd6! I think that this is the simplest route to a good position. Most modern sources recommend 10...Bc5 11.c3 Bd6, the point being that White has lost his main plan of developing his bishop on b2; an added benefit is the slight weakening of the d3-square. This makes a lot of sense, but White’s extra tempo also has some value in controlling the d4-square. Besides, I have no problem allowing White to play b2-b3 and Bb2, as we will be able to counter with the simple plan of ...c5-c4 with a rook coming to d8, which conveniently makes use of the knight’s position on a5.   11.d4 This is the main line nowadays.   11.f4 was popular decades ago, but practice has established that White’s priority should be to develop quickly, rather than having a nice pawn centre but doing nothing with it: 11...exf3 12.Nxf3 0-0

13.0-0 (13.d4 Re8 14.0-0 c5! gives Black excellent piece play, as shown in Vukcevich – Romanishin, Hastings 1976, and many subsequent games) 13...c5 14.Nc3 Nc6 15.d3 Re8 Despite his extra pawn, White’s position is already difficult. A good example continued:  

16.Kh1 Rb8 17.b3 Nd4! 18.Nxd4 cxd4 19.Nb1 Nd5 20.Bf3 Qh4 White’s position collapsed in Zvan – Mikhalchishin, Nova Gorica 1999.   11...exd3 12.Nxd3 0-0 Apart from being theoretically sound, I like that fact that Black’s position is so easy to play. We have a couple of useful developing moves in ...Re8 and ...Bf5, waiting to win a tempo with ...Qc7 when White castles. After that, we will put our other rook on d8. As soon as b2-b3 is played, we can respond with ...c5.

13.0-0 This has been by far the most popular move, although White sometimes delays it:   13.Bf4 Bxf4 14.Nxf4 Qc7 15.Nd3 c5! is fine for Black, for instance: 16.0-0 Bf5 (16...Rb8!?N has the idea to meet b2b3 with ...c4, and if 17.Qc1 Black plays 17...Nc6 followed by ...Nd4, ...Re8 and so on) 17.Nc3 Rad8 18.Qc1 Nc6 Black had at least enough compensation in Socko – Tomashevsky, Moscow 2004.   13.b3 gives Black a choice:

a) 13...Re8!? is interesting, as White can fall into a positional trap which I have used in many online games. The best continuation is: 14.0-0! (14.Bb2? Bg4! 15.f3 Bf5µ is the trap I was referring to) 14...Qc7 15.g3 c5! 16.Bf3 Rb8„ and ...c4 is coming.   b) 13...c5!? is the most logical move in terms of the guidelines laid out earlier. 14.Bb2? (14.Nd2 is correct, when

14...Bf5 gives Black the usual compensation)

14...c4! 15.Ne5 Qe7 Black intends ...Rd8, with a huge initiative; I have caught out some of my online opponents with this trick too!   13...Qc7 14.h3 c5„

15.b3 According to the engines, this natural move is already an inaccuracy. If that is the case, it means Black should be happy with the outcome of the opening.   15.Bd2 c4 16.Ne1 is how the engines want to play the position, but after 16...Re8N 17.Nf3 Bc5 Black is at least equal.   15...c4 16.bxc4 Nxc4 17.a4

17.Nd2 Nxd2 18.Bxd2 Bf5³ Liu Qingnan – Yu Yangyi, Jiangsu 2015.   17...Bf5 18.Nd2 Nxd2 19.Bxd2 Rac8

20.Ne1 Be5!? 21.Ra3 Ne4 22.Bd3 Nxd2 23.Qxd2 Be6³ White’s position was becoming rather unpleasant in Videnova – Guramishvili, Reggio Emilia 2012, and she soon ended up material down.

8.Bd3! – The Modern Trend 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Bd3!

This move was introduced by the English player Isidor Gunsberg, who is best known for (narrowly!) losing a match for the World Championship against Steinitz in 1891. For a long time it did not have the best reputation – especially after the game Cataldi – Keres, Stockholm 1937, in which White suffered a crushing defeat. Indeed, it looks as though White is just inhibiting his queenside development, but he hopes that the bishop’s influence over the kingside light squares will slow down Black’s counterplay. Unsurprisingly, modern analysis engines have uncovered new resources for White, showing that the whole variation had been seriously underestimated for more than a century. Many critical lines have yet to be tested in practice, and I believe there is a huge amount still to be discovered. Naturally I have endeavoured to cast some light on some of the most important new directions.   8...Nd5! The first move is a natural choice.   8...h6 9.Ne4 Nd5 10.0-0 g6 is a highly complex alternative; I think White is better, but not by much.   9.Ne4!? This move was played in the game that gave the entire 8.Bd3 variation a bad reputation, but White’s position at this stage is not so bad at all.   Around the year 2008 a lot of players became enthusiastic about 9.Nf3, but in Game 30 we will see Garry Kasparov quashing White’s hopes for an opening advantage.   White players did not get discouraged though, and in 2011 the amazing 9.h4! took over as the main line. After painstaking analysis, I am proud to say I discovered what I believe to be a bulletproof way to play as Black, which will finally bury White’s dreams.

9...f5 10.Nec3!? This leaves White’s pieces a bit congested on the queenside, but White is hoping that the knight will prove more useful on c3 than on g3.   10.0-0? is more imaginative than good: 10...fxe4 11.Qh5† Todorov – Rusev, Albena 2011. 11...g6!N 12.Qxe5† Kf7 13.Bxe4 (13.Qxh8 exd3–+) 13...Bg7 14.Qg3 Re8µ   The alternative is: 10.Ng3 g6! (10...Nf4?! was seen in Castaldi – Keres, Stockholm 1937, where White went passive with 11.Bf1?, but 11.Bxf5! Bxf5 12.Nxf5 Nxg2† 13.Kf1 is much more challenging. 13...Qd7!?N 14.Qg4 Nh4 15.Qxh4 Qxf5 16.Nc3 Qxc2 17.Qh5† Kd7 18.d4!‚ is one sample line demonstrating a whole new level of complexity and ideas that modern engines can reveal.) 11.0-0 Bg7 12.Re1 0-0 13.Bf1 c5 14.d3 Nc6

15.c3 Qc7 16.Nd2 Be6

It is obvious that Black has excellent compensation for the pawn, and he went on to win in Morau – Brower, corr. 2013.   10...Nf4 11.Bf1 11.0-0 Nxd3 12.cxd3 Bd6 (12...Be7!? is another idea) is similar to some of the other lines examined in this section. One of the many possible continuations is 13.Qh5† g6 14.Qh6 Bf8 15.Qe3 Bg7, when Black is doing fine.   11...Bc5 12.a3 This is another incredibly rich position, of which we might see more practical examples in the future. I think Black’s best is:

12...Bb6!N Leading to complex play, where Black has full compensation for the pawn.   12...Bd4 13.g3 Qb6 14.Qf3 Ng6 has been seen in a handful of computer games in my database. So far Black has scored highly, but after 15.b4 Nb7 16.Na4 Qc7 17.Bb2 Be6 18.Bxd4!N Bd5 19.Qe3 Bxh1 20.f4 a complicated position arises, in which White’s chances might be better after all.   GAME 30 Nigel Short – Garry Kasparov Leuven (blitz – 8) 2011   1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Bd3 Nd5 9.Nf3 As mentioned earlier, this move became popular for a few years, but Kasparov handles it perfectly.   9...Bd6 10.0-0

10...Nf4! Black will damage White’s structure to obtain long-term positional compensation for the pawn.   11.Re1 11.Be2?! has been played twice, but in both games Black overlooked 11...Nxg2!N, with the point that 12.Kxg2 e4 regains the piece with excellent attacking chances.   11.Nc3 This move has been tested quite a lot as well. 11...Nxd3 12.cxd3 0-0 13.Ne4 13.Re1 transposes to the main game.

13...Bc7!

The most accurate. 13...c5 14.b3 Nc6 15.Ba3 Nb4 16.Bxb4 cxb4 17.Re1 Bb7 18.Nxd6 Qxd6 19.Nxe5 Rfe8 20.Nc4 was clearly better for White in Zhigalko – Ragger, Tromso (ol) 2014. 14.Qc2 Protecting d3 is essential. 14...c5! Black reverts to the plan of giving up the c5-pawn to win the one on d3, as the inclusion of the last two moves works in his favour.

15.Qxc5 Both 15.b3 Bg4! hand Black the advantage. 15...Ba6 16.Re1 Bxd3 17.b3 Re8 18.Bb2 Bb6 19.Qa3 f6 20.Rac1 Nb7 21.b4 Bxe4 22.Rxe4 Qd5 23.Re2 Nd6 It is obvious that Black has full compensation for the pawn, and in the following correspondence game White did not even attempt to prove anything.

24.d3 Rec8 25.Qa6 h6 26.Rxc8† ½–½ Goncharov – Tomson, corr. 2012.

11...Nxd3 12.cxd3 0-0 13.Nc3 A tricky move which has done well in correspondence games is: 13.Nxe5 Black must make an important decision. I think the best choice is: 13...c5! Planning a quick ...Bb7. 13...Re8 has been more popular but 14.d4 c5 15.d3! cxd4 16.Nf3 is problematic, as White is a pawn up and may use the e4- and c4-squares as outposts for his knights. The bishop pair only gives Black partial compensation for the pawn.

14.d4! cxd4 15.d3!

White goes for the same plan as in the note above, and is ready to transpose to it after 15...Re8, a move which is, by the way, the top choice of most engines. However, I found a better idea: 15...Bxe5!N Another example of engine misevaluation is 15...Qc7 16.Nf3 Bg4 17.h3 Bxf3 18.Qxf3 Rfe8 19.Qd1! Rxe1† 20.Qxe1 Qc2 21.Qd2 and Black was in trouble in Rogos – Grabner, corr. 2010. This looks to me like a typical case of a correspondence player putting too much trust in his computer. 16.Rxe5 Nc6 Black has given up one of his bishops in order to mobilize his pieces quickly and put White under pressure.

17.Rg5 This is the top choice of the engines, but I consider it an encouraging sign if White has to play such a weirdlooking move.

17.Re1 can be met by 17...Bf5, when the natural-looking 18.Qf3?! Qd7 gives Black a strong initiative. The threat is ...Nb4, and if 19.a3? Rfe8 Black’s huge lead in development will decide the game. 17.Rc5 Qb6 18.Rg5 Re8 19.Na3 Ba6 is another computer-generated line. The stupid box calls this equal, but to the human eye Black’s position is more pleasant. 17...Re8 18.Na3 This is the best that my computer could do for White, and it is definitely not much. 18...f6 19.Rg3 Qd5  

20.Bh6! This may be the only way for White to avoid slipping into a highly unpleasant position. The point is to stop Black’s h-pawn from advancing. 20.Nc4 h5! (20...Ne7 is a good alternative, planning ...Nf5 and ...Bb7, but the text move puts more immediate pressure on White) 21.Bh6 (21.h3 h4 22.Rf3 Ne5 23.Nxe5 Rxe5, followed by ...Bb7 and ...Rae8, gives Black a strong initiative) 21...Re7 22.h4 Bg4 23.f3 Bf5 24.Qb3 Kh7 White is in some trouble, as his rook is seriously misplaced and Black can invade with ...Re2. 20...Re7 21.b3 Bf5 22.Nc4 Bg6 23.Bd2 Bh5 24.Qc1 Rae8 Even without the h-pawn involved in the attack, Black has an excellent position.

13...Re8 14.h3? This slow move has never been repeated and, had the game been contested over a slower time limit, Short would surely have chosen a stronger move. Nevertheless, it makes for an ideal illustrative game for us, as we get to see Black’s ideas in their purest form.   14.Ne4 This is the only theoretically challenging continuation, but Black is fine after: 14...c5! 15.b3 15.Nfg5!? is a tricky move to threaten Qh5, but Black can defuse it with: 15...Bf5! 16.b3 (16.Nxf7 Kxf7 17.Qf3 Ke6 is also better for Black) 16...Be7 17.Nxf7 Qd4 18.Ba3 Kxf7 19.Qf3 Ke6µ In Stockfish – Stockfish, engine game 2014, Black was ready to play ...Rf8 with a better position. The king is relatively safe for the time being, and should be able to hide on the kingside in the near future.

15...Bf8 16.Ba3 The attack on the c5-pawn forces Black to tread carefully. 16...Nc6!? I consider this the cleanest way for Black to play. 16...f5!? 17.Ng3 Nc6 18.Rc1 f4 19.Ne4 Bg4 is an interesting idea from the strong Italian correspondence player Nightingale (who happens to be a great King’s Gambit specialist with White!). Black wants to go ...Nd4 and is ready to abandon another pawn in order to achieve that. See the game Schroeder – Nightingale, corr. 2014, which ended in a draw after some complications.

17.Rc1! 17.Bxc5? f5! was the nice trap White fell into in I. Popov – Ragger, Tromso (1.3) 2013. The best White can do from here is 18.Bxf8N 18...fxe4 19.Bxg7 Kxg7 20.Rxe4 Kh8, but he remains clearly worse, despite having four pawns for a bishop. 17...Nb4! 18.Bxb4 cxb4 19.Ng3 19.Nc5 Qd5 20.d4 Bg4 21.h3 Bxf3 22.Qxf3 Qxf3 23.gxf3 exd4 is another equal position.

19...Bb7! 20.Nxe5 Qd6 21.Nc4 Qd5 22.f3 Qxd3= With two bishops against two knights, Black has plenty of activity to offset White’s extra pawn. 23.Kh1 Rad8 24.Rxe8 Rxe8 25.Qf1 Qd5 26.Qf2 A draw was agreed here in Bocanegra Moreno – Martin Sanchez, email 2011. Both 26...a5 and 26...Bc5 27.Ne3 Qe5 are fine for Black.

14...c5 15.b3 Ba6! Thanks to White’s slow 14th move, Black will restore material equality while keeping some initiative.   16.Ba3 Bxd3 17.Ne4 Otherwise Black goes ...e4.   17...Bxe4 18.Rxe4 f5 19.Ra4!

The rook appears oddly placed here, but Short’s intuition was correct.   19.Re1 e4 20.Nh2 Nc6 is similar to the game, without any tricks available for White.

19...e4 20.Nh2? This passive retreat gives Black everything he wants.   20.Nd4! would have kept White in the game: 20...cxd4 21.Bxd6 Nc6 (After 21...Qxd6 22.Rxa5 f4 Black’s pawns look impressive, but the engines say the position is just equal) 22.Bh2 Qf6 23.Rc1 Re6 Black has some attacking chances, but there is no clear way through and the engines indicate that White is okay.   20...Nc6 21.Rc1 Ne5 22.d4 Nd3 23.dxc5 Bf4 24.Rc2 e3–+ Even in a blitz game, you simply cannot let Kasparov get this kind of position against you! The former world champion impressively wraps up the win.

25.f3 Bg3 26.Nf1 Bf2† 27.Kh1 Nf4 28.Qxd8 Raxd8 29.Rxf4 Rd1 30.Re2 Bg3 31.Kg1 Bxf4 32.g4 Red8 33.Bb4 R8d4 34.Ba5 Rd5 35.Bb4 a5 36.Bxa5 Rxc5 37.Bb6 Rcc1 38.Bxe3 Rxf1† 39.Kg2 Bxe3 40.Rxe3 Rfe1 41.Rd3 f4 0–1  

The Critical 9.h4! 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5† c6 7.dxc6 bxc6 8.Bd3 Nd5 9.h4 Bologan actually dismisses the 8...Nd5 variation because of this move. This is understandable to some degree; if you start checking some lines with an engine you can easily get the impression that Black is struggling. Don’t lose hope though!

9...Qc7!

This has only been played in a tiny minority of games, but I think it is Black’s best. The idea is to pre-empt White’s attacks on the f7-pawn with Qf3 or Qh5.   9...Be7 is proposed by L’Ami, but 10.Qf3 f5 11.Bxf5 Rf8 12.Qh5† g6 13.Bxg6† hxg6 14.Qxg6† Kd7 reaches I position I would not enjoy defending in a correspondence game, let alone over the board:

15.d3 Kc7 16.0-0 Kb7 17.Qe4 Black’s king has found a reasonably safe home, but White is poised to pick up a fifth(!) pawn for the piece, and he has no obvious weaknesses.   9...Bc5 is a serious alternative, but 10.Qf3 f5 11.Nc3!N is rather annoying. (11.Bxf5 0-0 12.Bxh7† Kh8 13.Nf7† Rxf7 14.Qxf7 Nf4 led Black to an impressive win in A. Vovk – Halkias, Metz 2014)

11...0-0 12.Nxd5 Qxd5 13.Qxd5† cxd5 14.c3 The change in the pawn structure means that Black is closer to claiming full compensation for his pawn, but I don’t think he is quite there yet.

10.Bxh7 This is the only critical idea. Against most other moves, Black will chase the knight away with ...h6 (since the reply Qh5 no longer works), then develop his bishop to d6 or e7 and play ...f5.   The following game illustrates the above point quite well: 10.0-0 h6 11.Nf3 Bd6 12.Nc3 0-0 13.a3 Bg4 This is a great square for the bishop after h2-h4 has been played.

14.b4 Rae8!? (Obviously there is nothing wrong with 14...Nb7N followed by ...f5) 15.Be2 Nb7 16.Nxd5 cxd5 17.d3

17...e4 (Holding the pawn back for the moment with 17...Qe7!?N is also fully playable) 18.Nd4 Bxe2 19.Nxe2 Bh2† A draw was agreed here in Pellen – Vinchev, corr. 2014, in view of 20.Kh1 Be5 21.Rb1 Nd6 when the game is level.   10.Nc3 This move has only occurred in computer games in my database. 10...h6

11.Nxd5!? 11.Nf3 Bd6! 12.Ne4 0-0 13.Nxd6 Qxd6 should not worry Black; true, he has allowed a bishop to be exchanged for an enemy knight, but the rest of his pieces work well together and White’s kingside has been permanently weakened by h2-h4. Here is a sample line: 14.0-0 c5! Intending to bring the offside knight back to c6. 15.Re1 Bg4 16.Be2 Nf4 17.d3 Nxe2† 18.Qxe2 Nc6 19.c3 f5 White has some problems on the kingside. 11...cxd5 12.Qh5!?

12.Bb5† Bd7 13.Bxd7† Qxd7 14.Nf3 Bd6 15.d3 0-0 16.0-0 Nc6 gives Black more than enough compensation for the pawn. The text move is tricky, and requires a specific response:

12...Bc5! With the following point: 13.Nxf7 0-0! 14.Nxe5 Rf5!? 15.Qe8† Rf8 16.Qh5 Bxf2†!? Playing for compensation. Black can repeat with 16...Rf5= if he wants. 17.Kd1 Nc6

18.Nxc6 Qxc6 19.c3 Rf6! Preventing White from forcing a queen exchange with Qg6. 20.b3 Be6

Black maintains adequate compensation.

10...g6 11.Qf3 This is the most challenging continuation.   11.h5 This is an obvious contender, but Black has a clever way to deal with it. 11...Qe7! 12.d3 Nf4 13.Bxf4 exf4† 14.Qe2 Kd7! This position has been tested in a correspondence game, in which White now erred.

15.hxg6?! This complicates matters, but the one that can benefit the most is Black. Correct is: 15.Qxe7†N 15...Bxe7 16.Nxf7 (16.hxg6?! Bxg5 17.gxf7 Kc7 is a worse version of the endgame for White, despite having four pawns for the piece at this point) 16...Rxh7 17.Ne5† Kc7 18.Nxg6 Bg5 19.Nd2 Bg4

20.Nf3 Rxh5 21.Rxh5 Bxh5 22.Nxg5 Re8† 23.Kf1 Bxg6 24.Re1 Rxe1† 25.Kxe1 Kd6 White has three pawns against a bishop, but the pawns are a long way back and none of them are passed. It should be a draw, although in a practical game Black might have some winning chances. 15...Qxg5 16.gxf7 Kc7 17.Nc3 Rb8

This position can be compared with the one resulting from 9...Be7. Here Black is more active, better developed and has chances to pick up the f7-pawn. The pin along the h-file is also highly annoying for White, and it is no surprise that he eventually went down. Here are the remaining moves: 18.Qd2 Qe5† 19.Kf1 Nb7 20.Re1 Qg7 21.Ne4 Nd6 22.Qa5† Kb7 23.Nxd6† Bxd6 24.Qh5 Qxb2 25.Kg1 Qc3 26.Rf1 Kc7

27.Qg6 f3 28.Rh6 Qe5 29.g3 Bd7 30.Qh5 Qxh5 31.Rxh5 Be6 32.Bg6 Rxh5 33.Bxh5 Bxa2 34.Bxf3 a5 35.Kg2 Rb1 36.Rxb1 Bxb1 White called it a day in Tanti – Mouron, corr. 2013.

  11...Bf5 12.Nc3

12...Nb4!N 12...0-0-0 13.d3 f6 has occurred in some centaur games on the PlayChess server. According to my analysis, the most annoying move is: 14.Bxg6 (14.Nxd5 and 14.g4 also lead to complicated positions which, on the surface, look okay for Black – although I cannot claim to have analysed either of these lines exhaustively) 14...Bxg6 15.Ne6 Qd7 16.Nxd8 Bh5 17.Qh3 Nb4! I think Black can draw the game with a bit of accuracy. My main proposal is simpler though, and gives Black more winning chances.   13.g4! This is the only decent move, as 13.Kf1 0-0-0µ gives Black a dangerous initiative.

13...Nxc2† 14.Kf1 Nd4 15.Qg3! Otherwise White might get into trouble.   15.Nb5?! This looks tempting, but White is being too clever for his own good. 15...cxb5! 16.Qxa8† Bc8! The queen is in danger and Black has plenty of play for the exchange.  

17.d3 17.b3? is a mistake, despite being one of the top choices of the engines. 17...Nac6 18.Bb2 (18.h5 Ke7 gives Black an improved version of the main line below) 18...Kd7!µ (18...Ke7? is wrong due to 19.Rc1) Black threatens ...Bb7, and if 19.Ne4 Be7 Black picks up the h7-bishop. 17...Nac6 18.h5 Ke7! 18...Kd7? allows 19.Ne4 Be7 20.hxg6 and the bishop is saved. 19.Ne4 Rxh7 Black is better, although the position remains complicated.

15...Bc2! This looks like a weird square for the bishop, but it does the job.   16.h5 16.d3 0-0-0 is likely to transpose.   16...0-0-0„ Black is no worse in this complex position. I will give one illustrative line representing logical, ‘human’ play by both sides.   17.d3 Be7 18.Kg2!? Rdf8 19.Nce4 Bxg5 20.Nxg5 f6

21.Be3! The only move, but adequate for White.   21...fxg5 22.Bxd4 Rxh7 23.hxg6 Rxh1 24.Bxe5 Qd8 25.Rxh1 Bxd3

26.g7 Be4† 27.f3 Bxf3† 28.Qxf3 Qd2† 29.Kg1 Qe1† 30.Kg2 Qd2† The complications end in a draw.   Conclusion   In this chapter we studied 4.Ng5, the most critical test of the Two Knights, which we meet with 4...d5 5.exd5 Na5. Then Kieseritzky’s 6.d3, is better than its reputation, although Black is at least equal with correct play. The main line is of course 6.Bb5† c6 7.dxc6 bxc6, when White has three main options.   8.Qf3 Be7! allows White to take a second pawn, but in that case Black gets a huge initiative. 9.Bd3! is a better move, especially if followed by the accurate 9...0-0 10.Nc3!. Still, the resulting positions are fine for Black if he follows the example of Game 28.   8.Be2 has been the main line since... forever actually! After 8...h6, Steinitz’s 9.Nh3 is not as bad as it looks, although we saw that 9...Bd6! 10.d3 0-0 11.Nc3 Nd5! gives Black a nice game. 9.Nf3 is still the main move of course, when I recommend 9...e4 10.Ne5 Bd6 11.d4 (11.f4 is worse) 11...exd3 12.Nxd3 0-0 as a simple solution. Black has an active position and intends ...c5-c4 in the near future.   8.Bd3! has become the trendy move over the last six or seven years. This is the top choice of the engines, and you have to take them quite deep into the game before they admit that Black has full compensation. We start with 8...Nd5, when 9.Nf3 Bd6 10.0-0 Nf4 11.Re1 Nxd3 12.cxd3 0-0 13.Nxe5 c5! 14.d4 cxd4 15.d3 Bxe5!N is one important line where Black solves all his problems. 9.h4! is the most critical direction of all, but I discovered that the rare 9...Qc7! is a good answer. The most forcing line is 10.Bxh7 g6 11.Qf3 Bf5 12.Nc3, when the strong 12...Nb4!N gives Black a fine position in all variations.

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6   The Exchange Variation 4.Bxc6 dxc6 page 205 5.Nc3 f6 6.d3?! Game 31, page 206 6.d4 page 210 The Simple 5.d4 page 211 5.d4 exd4 6.Qxd4 (6.Nxd4?! page 211) 6...Qxd4 7.Nxd4 Bd6 8.Nc3 Ne7 9.Be3 Game 32, page 211 Capablanca’s Method 9.0-0 0-0 page 214 The Main Line 5.0-0 page 216 5...Bg4 6.h3 h5! 7.d3 (7.c3 page 216) 7...Qf6 8.Be3 Game 33, page 216 8.Nbd2 Game 34, page 219   The Delayed Exchange Variation 4.Ba4 Nf6 5.0-0 (5.Bxc6 page 225) 5...Be7 6.Bxc6 dxc6 page 225 7.Re1 Game 35, page 225 (7.Qe1 page 226; 7.Qe2 page 226; 7.Nc3 page 226) 7.d3 Game 36, page 229   5.d4 – The Central Attack 4.Ba4 Nf6 5.d4 exd4 6.0-0 Be7 page 231 7.Re1 page 232

7.e5 Game 37, page 233

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How should Black deal with the

typical f2-f4 break? (page 207)  

Suggest a strong manoeuvre for Black. (page 221)  

Suggest a plan of action for Black. (page 212)  

What did Nimzowitsch suggest here for Black? (page 215)  

Suggest a plan for Black. (page 222)  

How should Black deal with the

threat of f2-f4? (page 226)  

Can you find Black’s strongest idea? Suggest a good manoeuvre for Black. (page 217) (page 233)   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 In this chapter we will examine those variations where White trades his bishop for the c6-knight. This might come as soon as move 4, but White can also play more subtly, delaying it until Black has committed some of his other pieces, usually with ...Nf6 and ...Be7. All of these lines have common strategic elements and it makes sense to group them in one chapter. The dominant theme, of course, is the struggle of Black’s bishop pair against White’s healthier pawn structure. There are all kinds of strategy manuals offering guidance on how to exploit the bishop pair and I don’t intend to expound too much on this topic. Rest assured, you will find plenty of practical examples in this chapter to show how Black can put his formidable bishops to good use. That being said, I would like to mention the often-overlooked point that the bishop pair is an incredibly useful defensive tool. As I have experienced in my own practice, simplifying the position by sacrificing a pawn but obtaining the bishop pair is one of the most common defensive techniques among the top masters of correspondence chess. One of the reasons why this strategy is so successful is that even the strongest engines are unable to correctly evaluate the defensive capabilities of the bishops. Extensive texting of the Marshall Attack has confirmed this point of view, so it came as no surprise to me when I saw, to take one example, Aronian (a Marshall expert) making Black’s position in some of these exchange variations look easy, even when the engines evaluate the situation in White’s favour. The single biggest piece of general advice I can offer about these positions is that Black does not need to go crazy in his attempts to ‘unleash the bishop pair’, for instance by castling on the queenside and opening the kingside with ...f5, in the spirit of how Alekhine and later Fischer crushed their opponents. Such a strategy can work of course, but there is no need to force the issue. Instead Black can be patient, castle kingside and wait for the right moment to open the position. Even if the right moment never comes, the bishop pair will always be useful in some capacity. With that in mind, let’s stop the talking and see some chess!

The Exchange Variation 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 At this point White has three main continuations, but the great majority of games continue with Fischer’s favourite:

5.0-0 5.Nc3 f6 is seen in the first game of this chapter: a beautiful win by Botvinnik over Romanovsky from 1935. This serves as an instructive example of what Black is generally aiming for if White keeps the position closed. As you will see in the analysis section after the game, I found a nice way to avoid today’s main line by meeting 6.d4 with 6...Bg4!, by which Black forgoes the bishop pair in order to cripple White’s majority on the kingside.   5.d4 is the oldest of the three main lines. Here I have chosen a well-played game by two Greek masters, Ignatiadis and Mihos, from Athens 1998, which features an old but high-quality method of handling Black’s position. After 5...exd4 6.Qxd4 Bd6 7.Nc3 Ne7 8.Be3 0-0 9.0-0-0 we will fight for the initiative with ...f5!. White can castle kingside of course, but in that case Capablanca (and others after him) showed that Black can get a fine game by playing ...f6 and developing patiently.   5...Bg4! My main sources of study on the Exchange Variation are Kindermann, Kaufman, Panczyk & Ilczuk and Shirov, and all of them consider the text move Black’s best! I had to spend quite a lot of time and energy to find a line that I can recommend without reservations, but I am more than happy with the results of my investigation.   6.h3 h5 7.d3 Qf6 8.Nbd2! 8.Be3 was used by Kasimdzhanov to beat Adams at Tripoli 2004 and gain the title of FIDE World Champion, but Game 33 will show that Black does not have much trouble getting a fine game.

As I will demonstrate later, the most popular lines from this position do not give Black full equality, although he remains well within the drawing zone. Happily for us, there is another move which enables us to play for three results rather than just two.   8...g5! If White allows ...Be6 and ...g4 then he risks falling into deep trouble, so he is virtually forced to enter the following endgame.   9.Nc4 Bxf3 10.Qxf3 Qxf3 11.gxf3 f6 Some theoretical texts have evaluated this position as much better for White, but a study of Game 34 will make it clear that Black has excellent resources. If White is not careful he may encounter serious problems – not only in a practical game, but even at correspondence level.   GAME 31 Peter Romanovsky – Mikhail Botvinnik Moscow 1935   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.Nc3 5.Nxe5?! is met by 5...Qd4, so White first defends his e4-pawn in order to threaten to take on e5.   5...f6 This method of defending e5 is the main line and my preferred choice.

6.d3?! This method of playing the Exchange Variation was quite popular in the 1920s and 1930s and, according to Botvinnik (who annotated this encounter in his 100 Selected Games of Chess), Romanovsky had published some theoretical articles on it.   6.d4 is the most popular move, which I will discuss in the next section.   6...Bd6 7.Be3 c5! This move prevents d3-d4 for quite some time and prepares the manoeuvre of the g8-knight to d4. I have seen quite a lot of games at club level being lost for White after castling kingside and having no particular plan in mind, while Black plays ...Be6, ...Qd7 and ...0-0-0, followed by attacking with ...g5, ...Ne7-g6 and so on.   8.Ne2

Romanovsky’s patent, which is still seen today in similar positions (the 4.d3 Anti-Berlin comes to mind). Romanovsky’s reasoning was that White should keep the position closed to prevent Black’s bishops from becoming strong. His last move prepares two possible plans. The first is to play for f2-f4 after 0-0 and Nd2; the second is to play c2-c3 to stop the black knight from coming to d4, and prepare either d3-d4 or a2-a3 and b2-b4. (A good example of the latter plan is the game Romanovsky – Smorodsky, Moscow 1924.) According to Botvinnik, Romanovsky was correct to avoid playing for the first plan, as Alekhine had already shown an effective way to counter it from a similar position some years beforehand.   8...Ne7 9.Ng3 In the event of 9.Nd2 with the idea of f2-f4, Botvinnik gives the following instructive continuation: 9...Ng6! 10.0-0 0-0 11.Kh1 b6 12.Nb3 (12.f4 is still met by 12...f5!) 12...Qe7 13.f4 White has accomplished his plan, but Black has a strong antidote:

13...f5! Botvinnik mentions that the same method was played by Black in Tartakower – Alekhine, Semmering 1926. 14.fxe5 Nxe5 (modern engines indicate that 14...Qxe5!? is even stronger) 15.exf5 Ng4 is given by Botvinnik, with advantage to Black.   9...Be6 10.c3 Qd7?! 10...Nc6! would have been more accurate, in order to rule out the d3-d4 plan entirely. This move has the added benefit of enabling Black to put his queen on e7 instead of d7, especially if White opts for the plan of a2-a3 and b2-b4.  

11.0-0 0-0 Botvinnik does not mention the possibility of 11...0-0-0!? intending ...g5, ...h5, ...Ng6 and so on. This leads to more double-edged play, but is not at all bad for Black, and would perhaps have been the best way to ‘punish’ White for neglecting to play d3-d4 when he had the chance.   12.Qc2 Nc6 13.Nd2 I’m sure that Botvinnik would have responded to 13.a3 with 13...a5!, stopping any ideas of b2-b4.   13...Rad8 14.Rad1 b6 15.f3 Be7 16.Nb3

Romanovsky has managed to keep the position closed but he doesn’t seem to have a particularly strong plan, and from now on he plays a waiting game. I find this game extremely instructive as Botvinnik does find a plan, thus making his position more pleasant to play, even though the objective evaluation remains about equal for a long time.   16...a5 17.Nc1 Bd6 18.Qf2 Romanovsky may well have intended Rd2, Rfd1 and maybe even Nce2 to prepare d3-d4, but Botvinnik’s plan is much quicker.   18...Ne7 19.Rd2 f5! “Black’s move 19 looks rather risky, as the e5-pawn is weakened and White gets e4. But in reality Black can avoid the weakness of e4 by transferring the knight to f4, where White will have to exchange it, while the e4-square is far from Black’s camp, and its occupation by White is more than offset by the opening of the f-file and Black’s freer position.” – Botvinnik.   20.exf5 Nxf5 21.Ne4 h6 22.Qe1 Be7

Botvinnik criticizes his move, commenting: “Tactically unsound; it would have been better to keep the bishop on d6 and double the rooks on the f-file, preparing a pawn attack on the kingside. After the text move Black has to lose time defending his e5-pawn. Now White manoeuvres skilfully, gradually intensifying the pressure on this pawn”. Judging from the continuation of the game, however, Botvinnik’s move is not that bad, as the bishop found its way to f8 and later to g7, as the ...g5 advance is a part of Black’s future plans anyway.   23.Bf2 Bd5 24.Qe2 Rfe8 25.Rdd1 Qe6 26.Rfe1 Bf8 27.Qc2 Qf7 28.Rd2 Re6 29.Rde2 Rde8 30.Qa4 Ne7 The position remains more or less balanced, but Botvinnik is on track to realize his plan of transferring his knight to the f4-square. This puts pressure on Romanovsky, who starts committing slight inaccuracies and soon blunders a pawn.   31.Bg3 Bc6 32.Qc2 Nd5 33.Nb3 g5 34.Nbd2 Bg7 35.Nf1 Botvinnik proposed 35.Nc4!, intending 35...Nf4 36.Bxf4 exf4 37.Nf2, when he states that the rook exchanges should lead to a draw.   35...R6e7 36.Rd2 Rd7

37.Ne3? 37.h3 was suggested by Botvinnik, while the engines propose 37.b3 or 37.Qb1, protecting a2 in order to play Ne3 next.   37...Nxe3 38.Rxe3 Qxa2 Black has won a pawn, but White’s position is still salvageable at this point.   39.h4? 39.c4! was necessary, threatening to trap the queen with Nc3. Black has to play 39...a4 to create the possibility of ...Qb3, but then 40.h4 would have given good drawing chances according to Botvinnik.

39...gxh4! 40.Bxh4 Qf7

With the queen back on a good square, Botvinnik is in full control, with two bishops and an extra pawn. In the next phase of the game he methodically organizes a powerful attack along the g-file.   41.Rde2 Rf8 42.Bg3 Qg6 43.Kf2 h5 44.Nd2 Bh6 45.Rxe5 Rg7 The defences are collapsing and the rest is easy.

46.Nf1 h4 47.Bxh4 Bxf3 48.Kg1 Bxe2 49.Rxe2 Qh5 50.Re4 Rf4 51.Rxf4 Bxf4 52.Qb3† Kh7 53.Bf2 Qf3 54.g3 Qxd3 55.Qe6 Bh6 56.Qh3 Rf7 57.Ne3 Qb1† 58.Nf1 Qf5 59.Qh2 a4 0–1  

5.Nc3 – Main Line with 6.d4   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.Nc3 f6 6.d4 From this position Black usually exchanges on d4, but I have another idea.

6...Bg4! I am surprised that 6...Bg4 is rather popular in the main line after 5.0-0 f6 (don’t forget that I recommend 5...Bg4 here) 6.d4. With the knight still on b1, White can react to the pin on the f3-knight by playing c2-c3, or by exchanging on e5 and putting the queen’s knight on d2. In the present position with the knight already on c3, neither of these options is available, so I don’t understand why the text move is not more popular.   7.dxe5 7.Be3 makes it simple for Black to get a fine game: 7...exd4 8.Qxd4 Bxf3 9.gxf3 Qxd4 10.Bxd4 Bd6 11.Ne2 Kf7 12.0-0-0 Ne7  

Although the much higher-rated player won with White in Kokarev – Zhdanov, Ufa 2004, I see no reason for Black to complain about his chances at this point.

  7...Qxd1† 8.Nxd1 Bxf3 9.gxf3 fxe5 White’s pawn majority has been crippled so he cannot expect any advantage.

10.f4 The most popular and logical move, otherwise ...Nf6 will stop this idea for good.   10...exf4 11.Bxf4 0-0-0 12.Ne3 Re8 13.f3 Nf6 Black has an active position with at least equal chances. After completing development with ...Bc5 and ...Rhf8, he might later consider ...g5, ...h5 and ...g4 to break up White’s pawn centre.

The Simple 5.d4 GAME 32 Konstantinos Ignatiadis – Thomas Mihos Poros 1998   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.d4 exd4 6.Qxd4 As mentioned in the introduction, this is the oldest way for White to attempt to justify giving up the bishop pair. If everything goes his way, he will swap off all the pieces and win the pawn endgame!   6.Nxd4?! should be avoided as after 6...Nf6! White already has to be careful to avoid being worse. 7.0-0 c5 8.Nb3 Qxd1 9.Rxd1 Be6 10.Nc3 (In the event of 10.f3?! Bd6 11.Nc3 0-0-0 12.Be3 b6 Black can play ...Nd7, ...f6, ...g5 and ...Ne5, after which he stands somewhat better.)

10...Nd7 11.Nd5 Bd6 12.Nf4 (12.Na5!?N 12...Bxd5 13.Rxd5 0-0-0 14.Nc4 Nf6 reaches an equal but lively position) 12...Bxf4 13.Bxf4 0-0-0 14.Rd2 b6 15.Rad1 In Iordachescu – Bacrot, Baden-Baden 2013, Black played 15...Nb8 and agreed a draw on the next move, but he could have played for more with 15...f5!N or even 15...Rhg8!?N intending ...g5.   6...Qxd4 7.Nxd4 Bd6 I couldn’t help but recommend this move, as it has been my personal favourite ever since reading Capablanca’s Chess Fundamentals, one of the few classical chess books translated into Greek. Another reason is that we will see similar positions in the Delayed Exchange Variation section, where the black king is on the ‘right’ side of the board to deal with White’s pawn majority.   7...Bd7 followed by long castling has been the modern preference after Alekhine and later Fischer demonstrated the virtues of this approach. This is playable too of course, but there is no need for us to know two systems against this minor line, so I will not go into further details about it.

8.Nc3 Ne7 9.Be3 Most White players have opted for long castling in order to be able to advance the kingside majority without any fears.   In the most famous game in this variation, Lasker – Capablanca, St Petersburg 1914, White preferred 9.0-0, and I have chosen to examine this line separately in the next section.   9...0-0 10.0-0-0 The Capablanca method of ...f6 and ...Re8, although possible, is not as effective when the white king is on the queenside. Black does better with the following active move:

10...f5! This move is reminiscent of the classic Bondarevsky – Smyslov game, which can be found in the section covering the

Delayed Exchange Variation, although here White cannot really be lured into the ‘positional mistake’ of advancing his pawn to e5.   11.exf5 11.f3 gives Black a choice between exchanging on e4 or aiming for ...f4 with a more complicated game. 11...c5 (11...fxe4 12.Nxe4 Nd5 13.Bf2 Nf4 is a simpler equalizer, as shown in Smyslov – Gligoric, Belgrade 1998) 12.Nde2 b6 13.Bf2 A nice game continued: 13...Bb7!? (13...f4 followed by ...Be6 is also possible)

14.exf5? Black was indeed threatening to take on e4, but White should have dealt with it in some other way, as the game continuation shows. 14...Nxf5 15.Ng3 Nh4 16.Rhg1 Bf4† 17.Kb1 Nxg2! 18.Rxg2 Bxf3 19.Rgg1 Bxd1 20.Rxd1 Rad8 According to my database, the game score ended here with 0–1 in Schoenmann – Grünfeld, 1920. It seems awfully premature to resign, but there is no doubt that Black should eventually win the endgame.   11...Nxf5 12.Nxf5 Bxf5 13.h3 It may appear that White is better with the mobile 3–2 majority, but the presence of the black king in that area will make it hard for White to promote a pawn. Moreover, the bishop pair and open f-file offer us plenty of counterplay, and I would argue that Black has slightly better chances, especially from a practical point of view.

13...a5!? Mihos may have had the following game in mind when choosing his plan: 13...Rae8 14.a3?! Providing a ‘hook’ accelerates Black’s queenside attack. 14...a5! 15.Rde1 b5 16.Bd2 b4 17.Nb1 Rb8 18.Re3 c5 19.b3 c4!µ Kieninger – Wildschuetz, Essen 1947.   14.Na4 White wants to take advantage of the c5-square by putting a knight or bishop there, taking advantage of the fact that Black’s last move has rendered ...b6 unplayable. However, Black’s next move convincingly prevents White’s plan.   White should probably prefer something like 14.Rhe1 Rae8 15.Kb1, but the 15...b5 plan remains strong for Black, similar to the game.   14...Rae8! 15.Rhe1 15.Bc5? Re2 and 15.Nc5? Rxe3 show why White was unable to execute his plan of occupying c5.

15...b5 16.Nc3 b4 17.Na4 Be4! 18.Rg1 18.g3? Bf3 is bad for White, while 18.f3 can be met by 18...Bg6 with ideas of ...Bg3.   18...Bd5 19.b3 19.Kb1 can also be met by 19...Bg3! and after 20.Rge1 Rxe3! 21.Rxe3 Bxf2 Black is poised to pick up both the g2and h3-pawns. Being an exchange down is no problem when you have passed pawns supported by such strong bishops, especially with White’s knight being offside.  

19...Bg3! 20.Rd2 Rxe3 21.fxe3 Bf2 22.Rf1 After 22.Rxf2 Rxf2 the g2-pawn is under fire, and if White moves it then ...Be4 will be crushing.   22...Bxe3 23.Rxf8† Kxf8 24.g3 Ke7 25.Kd1 Bxd2 26.Kxd2 Kd6–+

Black is a pawn up with the better minor piece. It should not be too difficult to win this endgame, notwithstanding the doubled c-pawns.   27.Ke3 c5 28.Nb2 Kc6 29.h4 Kb5 30.Kd3 Be6 A nice zugzwang position!

31.c4† Moving the king would allow ...c4, but now the a2-pawn becomes terribly weak.   31...Kc6 32.Ke4 Bf7 33.Nd3 Bg6† 34.Ke3 Kd6 35.Nc1 Be8!? 35...Ke5 is arguably simpler, when White will not be able to defend both flanks. The text move is also fine though.   36.Kf4 a4 37.bxa4 White decides to weaken his structure rather than allow the pawn to get to a3, but the position is losing either way.   37...Bxa4 The rest is easy.

38.Nb3 Bc6 39.Nd2 Ba4 40.Nb3 Be8 41.Nd2 Bf7 42.Ke4 Bg6† 43.Kf4 Bc2 44.g4 Bd1 45.g5 Bc2 46.Ne4† Ke7 47.Nd2 Ke6 48.h5 Bd1 49.h6 gxh6 50.gxh6 Kf6 0–1  

Capablanca’s Method   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.d4 exd4 6.Qxd4 Qxd4 7.Nxd4 Bd6 8.Nc3 Ne7 9.0-0 0-0 Capablanca’s idea is that his king is better placed on the kingside in order to defend against the march of White’s pawn majority.

10.f4

10.Be3 gives Black a pleasant choice between 10...f5!? and 10...f6, with similar ideas as in, respectively, the previous game and the main line below.   10...Re8! Capablanca points out that this move prepares ...Bc5, as the reply Be3 can be met by ...Nd5.   10...f5 was proposed by Tarrasch but he may have been unduly influenced by the final result of the game. In fact there is no need for an improvement at this stage.   In a modern example Black played 10...f6 11.Be3 c5 followed by ...b6, ...Bb7, ...Nc6 and ...Re8 in Bierkens – Kaufman, Arlington 2013. Here too, Black had slightly better chances.   11.Nb3 f6 12.f5 This is a famous and widely praised move. Kramnik explains: “Lasker realized that the e5-square could be weakened because it was difficult to exploit. [...] He grasped a deep concept, which is being automatically employed now: he gave up the e5-square and ‘fenced in’ the c8-bishop.”  

Many sources make it seem like the text move was a surprising choice, but in fact it was rather typical of Lasker, who had previously played f4-f5 in similar situations in his games against Salwe and Janowski in 1909, as well as against Alekhine in an earlier round of the same tournament before the Capablanca game!   If White tries something else then Black’s plan will remain the same: ...b6, ...Bb7, ...c5 and ...Nc6.   12...b6! 13.Bf4

At this point the famous game Lasker – Capablanca, St Petersburg 1914, continued with 13...Bb7?!, but later Capablanca proposed:   13...Bxf4!N 14.Rxf4 c5 15.Rd1 Bb7 16.Rf2 Now 16...Rad8 is the usual move given in modern sources, as it was the main line of Lasker’s and Capablanca’s annotations, but Black has more than one way to improve.

16...Rac8!³ This amazing move was proposed by Nimzowitsch, the idea being to prepare ...Nc6 without having to worry about Nd5 or Rd7. Capablanca writes that White will have a hard time drawing this position because the plan of ...Nc6-e5-c4 is so unpleasant to meet. If White tries to take the sting out of Black’s knight manoeuvre by moving the b3-knight, then Black switches to ...Nd4 and the e4-pawn becomes a serious target. I am surprised that so many modern sources (including Giants of Innovation by Pritchett and Techniques of Positional Play by Terekhin & Bronznik) have quoted

the above game while omitting Nimzowitsch’s outstanding suggestion.   Interestingly, modern engines assess 16...Nc6! as being even more heavily in Black’s favour than Nimzowitch’s move, the main point being that 17.Nd5 Rac8 leaves White vulnerable in the centre. Nevertheless, I decided to keep Nimzowitch’s move as the bolded one, as it has tremendous instructive value as an improvement over the automatic development of the rook to the open file. Whichever option Black chooses, the important thing for us is that he has clearly won the opening battle.

The Main Line: 5.0-0   GAME 33 Bartlomiej Heberla – Martin Kraemer Muelheim 2014 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 Bg4 6.h3 If White does not play this, Black will have the strong plan of ...f6, ...Qd7 and ...0-0-0 available.   6...h5!

7.d3 Of course White cannot take the material, as Polerio and Greco taught us in the 17th century: 7.hxg4 hxg4 8.Nxe5 Qh4 9.f3 g3 and checkmate comes next.   In the event of 7.c3 I propose: 7...Qf6!? (7...Qd3 is theoretically fine, but it gives White the option of 8.hxg4 hxg4 9.Nxe5 Bd6 10.Nxd3 Bh2† 11.Kh1 with a well-known draw, which may not suit ambitious players) 8.d4 then 8...Bxf3 9.Qxf3 Qxf3 10.gxf3 exd4 11.cxd4 Ne7 (or 11...0-0-0) Black has a fine game.   7...Qf6 8.Be3 Avoiding the doubled pawns with 8.Nbd2! is the main line, and we will deal with it in the next game.   8.hxg4? is still too dangerous: 8...hxg4 9.Ng5 Qh6 10.Nh3 Qg6! (This is more accurate than 10...Qh4, which was proposed by Keres) 11.Ng5 Qh5 12.Nh3 Nf6 In Matijosius – Braziulis, corr. 2014, Black followed up with ...Bc5 and ...0-0-0, with a much better position.

8...Bxf3 9.Qxf3 Qxf3 10.gxf3 Bd6 Other moves would allow White to play f3-f4 with a clear advantage.   11.Nd2 White’s set-up must be taken seriously, and he has several possible ways to improve his position. Generally his knight will move to c4 and the f1-rook will go to either d1 or b1. Once the rook has moved, White can bring his king to f1 and e2. If the rook has gone to d1, he can advance in the centre with d3-d4 (which can be prefaced with c2-c3 if necessary). If the rook has gone to b1, he can aim for b2-b4, a2-a4 and eventually b4-b5.   11...Ne7 This developing move is an automatic choice. However, Black’s plan over the next few moves will depend on whether White starts by developing his knight or his rook – and in the case of the rook, we will modify our plans depending on whether it goes to d1 or b1.

12.Nc4 This move keeps the rook flexible for the moment, but the drawback is that the f3-pawn becomes slightly vulnerable. Let’s consider the two most logical rook moves:   12.Rfd1 Ng6 12...c5 followed by ...Nc6 is the other main plan, which will often be followed up by moves like ...a5 and ...f6. This is certainly a solid way to play, but deploying the knight on the kingside gives us more chances to put White under pressure. 13.Kf1 Compared with the main game, White wants to play Nc4 while being ready to meet ...Nh4 with Ke2. It sounds logical, but I think that there is a concrete argument against it.  

13...f5!? The most ambitious, although 13...c5 is a solid alternative which has been played by Adams. 14.Nc4 In the event of 14.exf5 Nh4 Black regains the pawn with a fine game. 14...b5 15.Na5 f4 16.Bd2 c5 The position was complex but fine for Black in Filippov – Predke, Moscow 2011.   The other main rook move is: 12.Rfb1 Ng6 After 12...c5 13.a3 a5 14.b4 cxb4 15.axb4 Nc6 16.bxa5 Rxa5 17.Rxa5 Nxa5 Black went on to hold in Anand – Kasimdzhanov, Moscow 2007, but I am not so keen on the idea of defending a slightly worse position with no winning prospects. Besides, White can try to improve on Anand’s play with 18.Ra1!?N. 13.b4 13.Kf1 can be met by 13...f5!, just as in the line above.

13...f6 14.Kf1 14.a4 Nh4 15.Kf1 g5 16.b5 axb5 17.axb5 Rxa1 18.Rxa1 Kd7 is also fine for Black. 14...Nh4 15.Ke2 g5 16.Nc4 Ng6 17.c3 Kd7 18.Rb2 Ke6 19.Rg1 b6 Black was at least equal in Ibarra Jerez – A. Muzychuk, Gibraltar 2013.   Let’s return to the game continuation of 12.Nc4.

12...Ng6! 13.Rfd1 Black’s plan remains the same in the event that the rook goes to b1 instead.   White also gets nothing from playing on the g-file: 13.Kh2 f6 14.Rg1 Kf7 15.Rg3 (15.Rg2?! would have allowed 15...Nh4 followed by ...g5)

15...Nf4! I like this move as it forces White’s hand and leads to a comfortable position for Black. 16.Bxf4 Otherwise Black can play ...g5, or perhaps ...Ne6 first. 16...exf4 In Rusov – Smirnov, St Petersburg 2001, White took the dubious decision to exchange on d6, but even after the superior 17.Rgg1 c5 Black has prevented d3-d4 and stands at least equal.

13...Nh4! The most forcing continuation, attacking f3.   14.f4 The only alternative that makes any sense is 14.Nd2 but it wastes time, and after 14...g5!? Black is ready for ...f6 and ...c5. I actually reached this position in an internet game, and after 15.d4? I could have obtained a big advantage with 15...f5!µ.   14...exf4 15.Nxd6† cxd6 16.Bxf4 Kd7 White may appear to have the better minor piece, but Black’s plan to attack on the kingside with ...g5, ...f5 and ...g4 is more dangerous than it may look at first sight. Objectively it would be too much to claim that Black has the advantage, but in a practical game he certainly has more chances to pose problems, as the present game shows.

17.Kh1? It would have been wiser to evacuate the king from the danger zone with 17.Kf1!. Black has several playable replies, but perhaps the most purposeful is 17...Rag8!?, intending ...g5, ...Ng6 and ...g4 when the time is right.   17...Raf8 18.Rg1 Rh7 Keeping the rook on the same file as the white king.   19.Rg3 f6 20.Rag1? This is the first move one would consider, but it is a serious mistake as it leaves White’s rooks clumsily placed and enables Black’s attack to gather a lot of momentum. Any neutral move would have been an improvement, although Black would still be able to play ...g5 with the better chances.

20...g5 21.Bc1 g4! 22.Re3 f5 23.exf5 Nf3 24.Rf1 Rxf5 25.hxg4 Opening the h-file is nasty for White, but allowing ...gxh3 would hardly have been an improvement.   25...hxg4† 26.Kg2 Rh3 27.c3 Rfh5 28.Re4 This loses a pawn and the game, but 28.d4 d5 would also have been unpleasant and most likely losing for White.

28...Nh4† 29.Kg1 Nf3† 30.Kg2 Nh4† 31.Kg1 Nf5 32.Bf4 Rxd3 33.Rfe1 c5 34.Kf1 Rf3 35.Kg1 d5 36.Rd1 Nd6 37.Re5 Rxe5 38.Bxe5 Nc4 39.Bh8 Kc6 40.b3 Nd6 41.Re1 Ne4 42.c4 Nxf2 43.Re6† Kd7 44.Rg6 d4 45.Kg2 Rf7 46.Be5 Ne4 47.Rxg4 d3 0–1   GAME 34 Hendrik Sarink – Soren Larsen email 2001   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 Bg4 6.h3 h5 7.d3 Qf6 8.Nbd2 g5! As I mentioned in the introductory section, I was unable to find full equality after any of Black’s alternatives. 8...Ne7 This has been the main line since... forever actually. I will skip some of the details and go straight to the problem continuation. 9.Re1! Ng6 10.d4!

10...Nf4 Kaufman’s recommendation. 10...Bd6 is the other main option, but after 11.hxg4 hxg4 12.Nh2 Rxh2 13.Qxg4 my analysis indicates that White remains somewhat better with accurate play. 11.hxg4 hxg4 12.g3 The only decent move, but it is good enough for a small edge. 12...gxf3 Black has also tested 12...Nh3† but the knight is misplaced here, and Black usually experiences problems with it. 13.Kg2 exd4 14.Rh1² 13.Qxf3 Ne6 14.dxe5

Black has a choice here, but I was not completely happy with either of the following options:  

a) 14...Qh6!? This is an interesting move according to Shirov, and it was Panczyk & Ilczuk’s main line in their pretty good Exchange Variation book. Still, I think White maintains a small advantage. 15.Nb3 g5 16.Be3

16...Qh3 16...Bg7 can be met by 17.Kf1! Bxe5 18.c3 0-0-0 19.Ke2! Qg6 20.Qf5² as in Socko – Olszewski, Warsaw 2008. 17.Qg2 Qh5 18.f3 Bg7 19.Kf2 Bxe5 20.Rh1 Qg6 21.Rxh8† Bxh8 22.c3! So far this has only led to a series of draws, but I find Black’s position a tad unpleasant. 22...0-0-0 23.Qh3 Bg7 24.Nd2! 24.Qf5 Qh5 gave Black sufficient counterplay in Macieja – Adams, Rethymnon 2003. 24...Rh8 25.Qf5 Rh2† Now 25...Qh5 can be met by 26.Rg1 followed by Nf1 if needed. 26.Ke1 Bf8 27.Rb1 Qh5 28.Nf1 Rg2 29.Bf2 b6 30.Rd1 a5 31.Rd2  

At this point a draw was agreed in Turkov – Nilsson, corr. 2011. However, I think White’s chances should be regarded as somewhat better, as Black has to remain focused and play accurately to avoid falling into a worse endgame. In the correspondence game White trusted in Black’s engine-assisted technique, but in a practical game it would not be so easy.   b) 14...Qxe5 15.Nb3! 15.Qf5 Qb5 16.Nf3 g6 17.Qxb5 axb5 18.c3 b4! was fine for Black in Vachier-Lagrave – Shirov, Bundesliga 2009-2010. 15...g6 This was Shirov’s choice in two rapid games against Karjakin in 2010, which took place after his encounter with MVL. 15...Bd6 has also been played, but after 16.Kg2 0-0-0 17.Qf5 White is slightly better.

16.Bd2! c5 Another game continued 16...Bg7 17.Bc3 Qg5 18.Bxg7 Nxg7 19.Rad1 Qh6 20.Kf1 Ne6 21.c3!² and White was ready to walk the king to the safety of the c2-square in Korber – Kruschinski, Internet 2012. 17.Rad1 Rd8 Black remains marginally worse. Of course the position is well within the drawing zone, but the fact that Karjakin won both of the aforementioned rapid games (first with 18.Kg2 and then with 18.Bc3) is a sign which I don’t think should be ignored.

9.Nc4 White does not have much of a choice, as Black threatens to drop the bishop back and play ...g4.   9.b3 is met by a strong manoeuvre:

9...Qg7! 10.Bb2 f6µ It is hard to suggest a good way for White to meet the plan of ...Bc5, ...Be6 and ...g4.

  9.Re1 is also bad for White, as Fischer pointed out in My 60 Memorable Games. 9...Be6 10.d4 g4 11.Nxe5 gxh3 (11...00-0! is even better, as has been shown in a few high-level correspondence games in my database) 12.g3 h2† 13.Kg2 h4 Black has a strong attack according to the American legend.   9...Bxf3 10.Qxf3 Qxf3 11.gxf3 f6 12.h4 It is natural to break up the kingside structure, as White can hardly hope for an advantage if Black is given time to place his knight on g6. Still, on this occasion the ambitious move may not be the best, as Black can obtain excellent counterplay along the open g-file.   The most logical alternative is: 12.Be3 Ne7 Planning to put the knight on g6. 12...c5!? is a rare move in my database, but it looks like a fully viable alternative. 13.d4 exd4 14.Bxd4 Bg7!? I like this idea of Bronstein. 14...Kf7 is playable, but Bronstein probably decided it would be useful to maintain the option of long castling as a convenient way to defend the b7-pawn against Na5. The text move also avoids the potential draw after 15.Rfd1 Ng6 16.Be3 Be7 17.Na5 Rab8 18.Ba7, which has occurred in a couple of games.

15.Ne3 15.Rad1 should be met by 15...0-0-0. 15...Kf7! Now that Na5 is not an option, this makes perfect sense. 16.Rad1 Rhd8 17.Bc3 Ng6 18.Nf5 Bf8 19.Bd2 Bd6 Black was at least equal in Husenov – Bronstein, Moscow 1967. The main plan from here will be to play ...h4 and ...Be5 (or ...Bf4) and ...Nf8-e6.   12...gxh4 13.f4 At this point Panczyk and Ilczuk claim a big advantage for White, but the further course of the game casts grave doubt on their assessment.

13...b5! 14.Na5 14.Ne3 exf4 forces White to waste additional time recapturing the f-pawn: 15.Ng2 (15.Nf5 Nh6 16.Nxh4 Bd6 occurred in Barbalic – Pazderski, corr. 2011, when Black kept some initiative by playing on the g-file) 15...Bd6 16.Nxf4 Ne7 17.Kg2 Kf7 Once again Black had the upper hand in Lahdenmaeki – Nisula, corr. 2013, with the next moves being ...Rag8 and quite possibly ...f5. (17...0-0-0!?N was also possible, with similar ideas.)  

14...exf4! I consider this the most effective practical choice, preventing White from closing the position with f4-f5.   14...Kd7 15.f5 Nh6 16.Nb3 Rg8† 17.Kh2 Nf7 18.Bd2 Bd6 is slightly favourable for Black due to his potential play down the g-file. However, without the ...f5 break available, White is not in too much danger, and he went on to draw comfortably in Schuster – Oreopoulos, corr. 2014.

  15.Bxf4 Kd7 Black’s position is easy to handle. He will gain a tempo with ...Bd6 next, then play ...Ne7 and ...Rag8.   16.Nb3 Bd6

17.Bxd6 17.Be3 Nh6! It is possible to develop the knight to e7, but I would prefer to keep that square free for the king in case Nc5† comes. 18.Kh1 Rag8 Once again Black has fine prospects.   17...cxd6 18.Nd4 Ne7 19.Kh2 Rag8 The reason why I chose this game is that it clearly shows the usefulness of the doubled h-pawns in controlling important squares along the g-file, enabling Black to double his rooks.   20.Rg1 20.f3 stops ...Rg4 but Black still obtains excellent play after 20...Rg5 21.Rg1 Rhg8 22.Ne2 Ng6 intending ...Nf8-e6.

20...Rg4 Also possible is 20...c5 21.Ne2 f5, when an illustrative (human!) continuation is: 22.f3 Nc6 23.a3 Otherwise the constant threat of ...Nb4 would be annoying. 23...Ne5 24.Rxg8 Rxg8 25.Rf1 Rf8 26.Kg2 fxe4 27.fxe4? h3† 28.Kg1 h2† and Black wins.   21.Rg2 White’s position is not easy to defend. Sarink prepares to meet ...Rhg8 with Rag1, but Larsen cleverly changes plan. 21.Kh3 This is the obvious alternative, but Black can continue to cause problems: 21...Rhg8 22.Rxg4 Rxg4 23.f3 White has to do something before ...Ng6-f4† comes. 23...Rg3†! 24.Kxh4 Rg2 Threatening ...c5.

25.Kh3 Rd2 What is White to do about ...c5? One idea is: 26.Re1 c5 27.Re2 However, Black can turn up the heat with: 27...Rd1! 28.Nf5 Also after 28.Nb3 Nc6 White must worry about ideas like ...a5-a4 or ...Nb4. 28...Nxf5 29.exf5

29...a5! White is in a kind of zugzwang, while Black has the simple plan of ...a4, ...b4 and ...Ra1. 30.Kh4 Rh1† 31.Kg3 a4 32.Kg2 Ra1 33.a3 Ra2 34.c3 b4 White is in serious trouble.

21...c5 22.Nf3 f5! 23.Rxg4 In the event of 23.exf5 Nxf5 Black threatens ...Re8-e2, and 24.Re1 runs into 24...Rhg8 25.Reg1 Re8! when White is in trouble. Once again we can see the usefulness of the doubled isolated h-pawns.   23...hxg4 24.Nd2 f4 It is amazing how quickly the position has become hopeless for White, especially as this was a correspondence game.

25.Rg1 g3† 26.Kh3 Nc6 The knight is ready to join the attack from b4, d4 or e5, so White decided enough was enough. 0–1

The Delayed Exchange Variation 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 5.Bxc6 makes far less sense, as now Black’s bishop will be able to go to the optimal d6-square in one go.   The so-called Central Attack with 5.d4 is a rare guest in modern practice, but somehow it finds a place in this chapter, as one of the most popular variations involves Bxc6 and ...dxc6 a few moves down the line. Game 37 at the end of the chapter will provide further details about this.   5...Be7 6.Bxc6 dxc6

This is sometimes referred to as the Doubly Deferred Exchange Variation. White has spent an extra tempo with his bishop in the hope that the f6-knight and e7-bishop will be less than optimally placed in the resulting structure. The material from this point can be divided into two categories.   A) Firstly there are the lines where White plays d2-d4. These positions can be compared with the Ignatiadis – Mihos game where Black used the Capablanca Method of putting his king on the side of the board where it can help to defend against White’s pawn majority. This time we will examine one of Smyslov’s most beautiful games, his classic victory over Bondarevsky at Moscow 1946, while in the notes you will find a more recent and balanced game (but no less instructive, at least in the opening phase) between Grischuk and Aronian from Wijk aan Zee 2011. This section will provide all the essential knowledge you will need to handle the structure which is likely to arise after 7.Re1, 7.Nc3, 7.Qe2 or even 7.Qe1.   B) We will then examine the closed structure involving d2-d3. Interestingly, a similar idea is currently rather popular against the Berlin Defence, specifically after 3.Bb5 Nf6 4.d3 Bc5, when 5.Bxc6 dxc6 leads to something similar to what we get here. Fortunately for us, the present variation compares slightly favourably with the Berlin version, as we have the useful ...Nd7-c5 manoeuvre. Game 36 will show that Black does not have the slightest theoretical problem in this variation, and that, if anything, White is the one who is under pressure to justify his opening play.  

GAME 35 Igor Bondarevsky – Vassily Smyslov Moscow 1946   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Bxc6 dxc6 In this game we will deal with positions where White plays d2-d4, leading to exactly the same structure as the Lasker – Capablanca game discussed earlier. The present game was fantastically annotated by Smyslov himself in Endgame Virtuoso so it was a tough task for me to try and shed any new light on it. Still, my trusty analysis partners (Stockfish and co.) were able to point out some small improvements over the original annotations from the World Champion.   7.Re1 7.Qe1 and 7.Qe2 are not fundamentally different from the other lines we will examine, so I will take the opportunity to save some space and avoid overloading you with variations and examples of similar play. If White continues with d2-d4 the play will be similar to the current main game, whereas d2-d3 will resemble the following one. That being said, I will just mention that both queen moves give Black the extra option of preventing the d2-d4 plan with 7...c5!?, as 8.Nxe5 can be met by 8...Qd4 regaining the pawn. Obviously this is not an option for us after the move played in the main game.   A recent high-level game continued: 7.Nc3 Nd7 8.d4 exd4 9.Nxd4 Taking with the queen would have been similar. 9...0-0 10.Qe2 Bd6 11.Be3 Nf6!? 11...Ne5 transposes to an Ivanchuk – Leko rapid game, and is recommended by Gustafsson in his DVD on the Open Games. White may try 12.h3N, with the idea of f2-f4 without allowing ...Ng4.

However, Gustafsson gives 12...Ng6! followed by ...Qh4 with an active game for Black. (13.f4? just loses a pawn as the d4-knight is hanging.) Although Gustafsson’s suggestion seems satisfactory, I like Aronian’s move even more, as it turns out that Black does not have to fear f2-f4.

12.f4 12.f3 is not really in the spirit of the position and after 12...Re8, followed by ...Qe7 and maybe ...c5 and ...b6 later, Black is doing fine. 12.h3 is nothing either as 12...Re8 (or 12...Qe7) hits the e4-pawn, thus preventing a quick f2-f4. 12...Ng4 13.Rad1 Nxe3 14.Qxe3 f6

15.Nce2 Qe7 16.Kh1 Re8 17.Ng3 Qf7 White had the space but Black had the bishops and eventually prevailed in Grischuk – Aronian, Wijk aan Zee 2011. I chose not to examine this game separately due to spatial constraints, but I would strongly advise you to look up the game and study it closely.   7...Nd7 This is the typical method of protecting e5.   7...Bg4 is also playable, as after 8.h3 Bh5 9.g4 Black can sacrifice a piece for two pawns with 9...Nxg4 10.hxg4 Bxg4. Personally I am not so keen to sacrifice my precious pieces, even if the idea is theoretically sound, especially against certain types of opponents. If you want to know more about this type of sacrifice you can check the works of Marin and Kaufman among others, or simply investigate the idea yourself now that you know that it is possible.

8.d4 exd4 9.Qxd4 0-0 10.Bf4 10.Nc3 Nc5 11.Qxd8 Rxd8 12.Bf4 Ne6 13.Bg3 was seen in Gdanski – Almasi, Moscow 1994.

The Hungarian super-GM suggested 13...Bf6 in his old annotations for Chess Informant, but I prefer 13...b6!N intending ...Bb7 and ...c5, with an excellent game due to the strong bishops.   10...Nc5! 11.Qxd8 Avoiding the queen exchange leads to the type of position seen in the Grischuk – Aronian game, as discussed in the note to move 7.   11...Bxd8 12.Nc3

12...f5! Black wants to open the game for his bishops, just like in the Ignatiadis – Mihos game in the previous section when the white king was on the other side of the board.   12...Ne6 13.Bg3 Bf6 is possible, and is similar to the aforementioned Almasi proposal. This would keep Black close to equality, but Smyslov’s move is in a different class.   13.e5? We have reached an instructive moment of the game. White’s decision might look logical, but the e-pawn is safely blockaded and Black becomes stronger on the light squares, where White lacks the services of a bishop.   13.Bg5 is better. 13...Bxg5 14.Nxg5 h6 15.b4! is given by Smyslov, but after 15...Nxe4 (15...hxg5 16.bxc5 g4!? is also interesting, although White should be okay) 16.Ngxe4 (16.Ncxe4 a5!?) 16...fxe4 17.Nxe4 a5 Black has slightly the more comfortable side of a probable draw.

13...Ne6 14.Bd2 g5! “White’s difficulties stem from the fact that he has no strong points in the centre for his minor pieces” – Smyslov.   15.Ne2 c5! 16.Bc3 b5! 17.b3 At last White has managed to ensure a safe retreat square for the f3-knight, but in the meantime Black has claimed space all over the board.

17...Bb7 18.Ng3 g4 19.Nd2 Be7 20.Nh5?! This move is part of a plan that has been criticized everywhere I have seen this game annotated. The knight proves to be more of a target than a strength on f6, but how else was White supposed to obtain any chances of counterplay? If Black is left untouched he will centralize his rooks and possibly double on the d-file before bringing his king to g6. The same thing actually happens in the game, but in an even better form.

  20...Kf7 21.Nf1 Kg6 22.Nf6 Rad8 23.Rad1 Rxd1 24.Rxd1 Rd8 25.Rxd8 Bxd8 Black doesn’t really need rooks on the board, so Smyslov exchanges them to deprive White of any potential counterplay along the d-file.

26.Ne3?! 26.Bb2 was suggested as a better try by Smyslov; indeed, at least now White is in a position to meet ...Nf4 with Ne3 as there is no fork on e2. Nevertheless, Black remains much better after 26...Bxf6 27.exf6 Nd4 when the f6-pawn will perish.   26...f4! 27.Nd1 27.Nexg4? h5 traps the knight.   27...Bxf6 28.exf6 Be4 29.Bb2 b4! The c2-pawn is going nowhere, so Smyslov takes the opportunity to restrict White’s knight even further before capturing.   30.f3 30.c3? is impossible as 30...Bc2 traps the knight.   30...Bxc2 31.Nf2 gxf3 32.gxf3 Bb1 33.Ne4 Bxa2 34.Nd2 a5! This move ensures that the a2-bishop won’t be trapped. Black could also have won with 34...Bxb3! 35.Nxb3 c4 36.Nc1 a5 when the pawn mass is unstoppable, but Smyslov’s move is more didactic.

35.Kf2 Nd4 36.Bxd4 cxd4 37.Ke2 Kxf6 38.Kd3 Ke5 39.Kc2 a4 This is where the comment to move 34 becomes relevant.   40.bxa4 c5 41.a5 c4 42.a6 d3† Black plays ...c3 next and can bring back his bishop to d5 to guard against the promotion of the a-pawn. White saw no reason to continue the game. 0–1   GAME 36 Mateusz Bartel – Evgeny Tomashevsky Eilat 2012   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Bxc6 dxc6 7.d3 Nowadays this structure is occurring more and more through the Berlin Defence. The difference between the two variations is that in the Berlin the bishop goes to c5 whereas here we have the bishop on e7. This difference might seem irrelevant, considering that Black’s optimal set-up involves ...Bd6 (so in both cases Black will have to lose a tempo), ...c5 and rerouting the knight to either e6 or c6 in order to be ready to jump to d4. However, the advantage of having the bishop to e7 is shown in this game, where Black is able to use the c5-square for his knight. So, whatever difficulties Black is facing in the Berlin (and it seems that he does face some challenges, as more and more White players try this method of avoiding the endgame and find new resources) are not existent in the Delayed Exchange. Phew!

7...Nd7 8.Nbd2 8.b3 c5 (8...0-0 9.Bb2 f6 followed by a quick ...Nc5, just like our main line, is a good alternative and was recommended by Gustafsson) 9.Bb2 f6 10.Nbd2 Nf8! is thematic and fine for Black. A typical continuation would be 11.Nh4 g6 12.f4 exf4 13.Rxf4 Ne6 followed by ...0-0, ...Nd4 and ...f5.   8...0-0 9.Nc4 Another possible direction is: 9.a4 c5!? (9...a5 is more typical) 10.a5 f6 11.Nh4 Rf7!? (11...g6 would have been fine, but it’s nice to see other ideas work) 12.Nf5 Nf8 13.Nc4 Ne6 14.Nxe7† Qxe7 15.Be3 Bd7= Ivanchuk – Topalov, Nice 2009. Similar play was seen in Mestel – Geller, London 1982, a game worth looking up.   9...f6 10.Nh4 This is the only critical move, as otherwise Black plays ...c5 and ...Nb8-c6 as we have seen already.   10...Nc5! Other methods exist, but this one is the simplest and most forcing.

11.Nf5 11.f4 is no problem, as after 11...exf4 12.Rxf4 g6! 13.Be3 Ne6 14.Rf1 f5 Black equalized with ease in Ermenkov – Smejkal, Polanica Zdroj 1972. Note the similarity with the 8.b3 variation in the notes above.   11.Qf3 is a popular but unthreatening alternative. After 11...Ne6 12.Nf5 Nd4 13.Nxd4 both recaptures are fine, but I will trust both Geller and the database statistics and recommend 13...exd4!. The position is equal, but I like Black’s chances with the bishop pair. The doubled c-pawns are not a big problem as Black’s cluster of queenside pawns is mobile and flexible.   11...Kh8!? 11...Bxf5?! 12.exf5 is criticized by Marin in A Spanish Repertoire for Black. Gustafsson, on the other hand, believes that the game is equal after 12...Qd5 13.Qg4 Rfe8 14.Re1 Rad8 15.Nd2 Qf7 when Black’s ideas include ...Rd4 or even ...Bf8 and ...e4. However, after analysing the position quite deeply, I found that 16.b3! a5 Bb2 a4 18.Bc3! leaves White with somewhat better prospects. One annoying plan involves Nf3 and d3-d4, and White can also consider strengthening his position with h4-h5 first.   11...Ne6 is recommended by Kaufman, and seems like a fine alternative, with ideas of ...c5 and ...Nd4. The game continuation has the same idea if White continues with a neutral move, but why not let White execute his ‘threat’?

12.f4 This is the critical move of course, but Tomashevsky displays excellent preparation.   12...Bxf5 13.exf5 Qd5! Black threatens ...e4, which may or may not be prefaced by ...Rae8.   14.fxe5 fxe5 15.Qg4 e4³

16.b4 White wants to oust the knight from an outpost that puts his d-pawn under more pressure, but this move creates additional weaknesses.   16.Ne3 is met by 16...Qd4 intending either ...Rad8, or ...Bd6 and ...Rae8.

  16...Bf6 17.Ne3 Qd4 18.Rb1 Na4 19.Rb3 Rae8 20.Bd2

20...Nb6? Tomashevsky makes his only significant mistake of the game. Instead 20...c5!N followed by ...b5 would have been close to decisive.   21.Kh1? 21.Re1! would have enabled White to hold his position together.   21...Nd5 22.Qg3 exd3 23.Rxd3 Nxe3 24.Bxe3 Qxb4 25.Qxc7 Qc4 The rest of the game is not so important for us. Black still has some work to do, but Tomashevsky maintains control and picks off White’s queenside pawns before simplifying to a winning endgame.

26.Bg1 Qxc2 27.Rh3 Qxa2 28.Qxb7 Qc4 29.Rd1 Re2 30.Rg3 Rfe8 31.h3 Qb5 32.Qd7 c5 33.Qa7 Re1 34.Rxe1 Rxe1 35.Re3 Rxe3 36.Bxe3 h6 37.Qxc5 Qxc5 38.Bxc5 a5 39.g4 Kg8 40.Kg2 a4 41.Kf3 Kf7 42.Ba3 Be7 0–1

5.d4 – The Central Attack 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d4 The so-called Central Attack can also be employed after 5.0-0 Be7 6.d4 and is a popular first weapon in the Spanish, as it leads to open positions with active possibilities for White. In a recent article in CBM 156, GM Krasenkow made a good case for it, showing that Black’s lesser options actually give White a sizeable advantage.

5...exd4 6.0-0 Be7 At this point White has two major options. If he wishes to chase the knight away then 7.e5 is the last chance to do so. This natural move will be covered in Game 37, but first we will deal with the more restrained alternative:   7.Re1 According to Krasenkow, this is White’s best option.   7...b5! 7...0-0!? 8.e5 Ne8 followed by ...f6 or ...d6 seems reasonable, but the annoying pawn sacrifice 9.c3! gives White pretty good compensation.   8.Bb3 The following alternative has occurred in hundreds of games, but it is not impressive at all. I will just mention a few key points and leave it to the reader to investigate in more detail if he wishes: 8.e5 Nxe5 9.Rxe5 9.Nxe5 bxa4 10.Qxd4 0-0 gives Black at least equal play. The text move is the only real attempt to disrupt Black’s position by exerting pressure along the e-file. 9...d6

10.Re1 10.Rg5?! bxa4 11.Rxg7 is interesting but not quite correct: 11...Nh5 12.Rg5 Bxg5 13.Bxg5 Qd7! 14.Qxd4 (14.Nxd4 Qg4 15.Qe1† Be6³ may be White’s best according to Mikhalevski) 14...0-0 15.h3 Bb7µ Rogers – Howell, London 1988. 10...bxa4 11.Nxd4 11.Bg5 0-0 12.Qe2 h6! 13.Bh4 g5! 14.Qxe7 Qxe7 15.Rxe7 Nd5! 16.Re4 c5! favours Black, as shown by Mikhalevski. 11...Bd7 12.Qe2 Also after 12.Qf3 0-0 13.Nc6 Bxc6 14.Qxc6 d5 Black is more than okay. 12...c5 13.Bd2?! 13.Nf3 h6! 14.Nc3 Be6 15.Nxa4 0-0³ De Blasio – Colucci, corr. 1994. The text move looks tempting, but Mikhalevski shows how Black can refute it.

13...0-0!! 14.Qxe7 Qb8 15.Qe3 Re8 16.Qc3 Rxe1† 17.Bxe1 Nd5! 18.Qa3 cxd4 19.Nd2 Qb6!µ Black keeps an extra pawn.

8...d6 Black stabilizes his position in the centre. Obviously White cannot recapture on d4 because ...c5-c4 will trap the bishop, so the great majority of games from this position have continued as follows:   9.Bd5 Nxd5 10.exd5 Ne5 11.Nxd4 0-0 12.a4 12.f4?! Bg4! followed by ...Nc4 is excellent for Black.   12...Bd7 13.Nc3 b4! 14.Ne4 14.Nce2 can be met in exactly the same way.

14...Qb8! 15.b3 I have followed Krasenkow’s guidelines for White, as this indeed seems to be the best way for him to play. In the event that White deviates, the following plan should work well for Black anyway.   15...Qb7!³ The queen is excellently placed on b7, where it simultaneously attacks d5 and covers the sensitive spots on c6, a6 and b4. Black’s plan for the next few moves will be to improve his pieces with ...f5 and Rae8, followed by either ...Bf6 or ...Bd8 and possibly ...c5 later.   Let’s now see what happens if White takes the opportunity to kick the knight away from f6.   GAME 37 Michele Godena – Vassily Ivanchuk Reggio Emilia 2011   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d4 exd4 6.0-0 Be7 7.e5 Ne4 No problem! The knight will go to c5 with tempo whenever it is attacked.

8.Nxd4 White almost always goes for this obvious recapture. Alternatives do exist, but they offer White nothing and are easy to meet. If you are looking for a complete coverage I could never beat Mikhalevski’s analysis in Grandmaster Repertoire 13 – The Open Spanish, a truly outstanding piece of work.   8...0-0 9.Nf5 Nc5! I like this move the most, although 9...d5 is a solid alternative. The usually reliable ECO underestimates the text move, evaluating it as slightly better for White in all variations. However, the books of Lysyj and Ovetchkin and later Mikhalevski have convincingly shown that it is actually Black who can press for the advantage.

  10.Qg4 White should take the opportunity to provoke ...g6, otherwise he will just get a worse version of the game.

10...g6 11.Bxc6 dxc6 Finally we reach the pawn structure that accounts for my decision to include this variation in the present chapter. By contrast with the Exchange and Delayed Exchange variations, here White’s pawn has advanced to e5, giving the position some of the characteristics of a Berlin but without the queen exchange. This slight change in the structure gives Black excellent prospects on the light squares, and also renders White’s pawn majority more difficult to advance in the long run.   12.Nxe7† 12.Nh6†?! Kg7 leaves the knight in danger; for instance, if 13.Qg3 Black has 13...Bh4! with ideas of ...Bg5.   12...Qxe7 13.Qe2 13.Qg3 is the recommendation in ECO, but it does not really change the assessment: 13...Re8 14.Re1 Bf5 15.Nc3 Ne6! Black wants to bring his second rook into play on d8, so his last move is necessary to prevent the reply Bg5. The obvious continuation is:

16.Ne4 Bxe4 17.Rxe4 Rad8 The only question is whether Black is merely comfortably equal or already better. 18.Be3 is a logical move, when either 18...Qd7 (intending ...Qd5) or 18...Rd5!?N give Black good chances according to Mikhalevski.   13...Re8 Mikhalevski proposes to skip this move and go for the same type of position with 13...Bf5 14.Nc3 h5 followed by ...Rad8 and ...Ne6, which he assesses as slightly better for Black. Ivanchuk’s move also leads to a slight edge though, and the general character of the game is about the same in both cases.

14.Re1 14.Nc3!N 14...Qxe5 15.Qxe5 Rxe5 16.b4! is given as best by Mikhalevski, leading to equal play. This seems correct, but I think Black can improve with 14...h5! intending ...Bf5, which is likely to transpose to the main game.  

14...Bf5 15.Nc3 Rad8 Black has completed development and his coordination is excellent. Ivanchuk proceeds to improve his position with ...Ne6 and ...Qb4, with the annoying threat of ...Nd4.   Lysyj and Ovetchkin recommend 15...h5! first, and I would agree that this is a slight improvement to rule out the possibility in the next note.

16.h3 Mikhalevski points out that 16.g4!? Be6 17.Be3 could have been considered.   16...h5 17.Bh6? This accomplishes nothing.   17.a3 would have made more sense, although Black has more than one good continuation. The simple option is 17...Ne6 18.Be3, when White is a tempo up on the game and has covered the b4-square, but Black still has a pleasant game after 18...Nd4 19.Bxd4 Rxd4. A more ambitious alternative is 17...Qh4!?, intending to put the king on h7 and then attack with ...g5-g4! The opposite-coloured bishops will greatly favour Black if he can uncover some targets on the light squares.   17...Ne6 18.Be3 There is no time for 18.a3? due to 18...Nd4.   18...Qb4 19.Rab1 Nd4 20.Bxd4 Rxd4µ White’s strategy has clearly failed. The most obvious threat is ...f6.

21.a3 Qc5 22.Qe3 Rc4!? 22...Bxc2?? is obviously out of the question as 23.Rbc1 Bf5 24.Ne2+– completely turns the tables.   However, 22...f6!? would also have been strong. The main idea is revealed after 23.f4 Qb6! with threats of ...Bxc2 and ...Rxf4.   23.Rbc1 Kg7 Ivanchuk plays solidly. Over the next few moves he maintains his advantage and squeezes bit by bit.   24.Qg3 h4 25.Qh2 Re6 26.Re3 a5 27.g4 This may seem like desperation, but White couldn’t just wait for ...b5-b4 or ...g5.   27...hxg3 28.Qxg3 Qe7 29.Ne2 g5 30.c3 c5 Preventing Nd4.

31.f4? White is probably not yet losing after something like 31.Qg2, but his position would remain uncomfortable. 31...Bg6 is a good reply, threatening the e5-pawn.   31...Rg6! 32.Kh2 gxf4 0–1     Conclusion   We started with the Exchange Variation proper, when 4.Bxc6 dxc6 gives White three main ways to play. After 5.Nc3 f6 we learned from Botvinnik how to handle the closed positions after d2-d3, while the main line of 6.d4 can be effectively met by the rare 6...Bg4!, when my analysis indicates that Black has nothing to fear. White’s most simplistic way of handling the Exchange Variation is 5.d4 exd4 6.Qxd4 Qxd4 7.Nxd4, when I recommend 7...Bd6 followed by ...Ne7 and ...0-0. If White goes for long castling then opening the position with ...f5 leads to a great position for Black, while short castling can be met by the restraining ...f6 and ...Re8, possibly followed by ...b6, ...Bb7 and ...c5 when Black is more than okay.   The most challenging move is 5.0-0, when I recommend 5...Bg4 6.h3 h5 7.d3 Qf6. Now 8.Be3 is met by exchanging on f3 followed by ...Bd6 and ...Ne7-g6, giving Black the useful option of a quick ...Nh4. The most theoretically critical move is currently thought to be 8.Nbd2, but I believe that 8...g5! has been underestimated by many sources. If White wishes to avoid being attacked he is advised to go for an endgame with 9.Nc4 Bxf3 10.Qxf3 Qxf3 11.gxf3 f6. From here White faces an awkward dilemma: if he allows ...Ne7-g6 then Black will have a comfortable game with no problems, but the ambitious 12.h4 gxh4 13.f4 allows Black to seize the initiative with the dynamic 13...b5! 14.Na5 exf4 15.Bxf4 Kd7 followed by ...Bd6 and ...Ne7, when the seemingly weak doubled and isolated h-pawns provide Black with vital support for his operations along the g-file. In the rest of the chapter we considered other variations involving Bxc6, the main one being the Delayed Exchange Variation after 4.Ba4 Nf6 5.0-0 Be7 6.Bxc6 dxc6, when there are two main plans to take into account. If White defends the e4-pawn with his knight, queen or rook, then after the typical 7.Nc3 Nd7 (defending the e5-pawn and preparing to reroute the knight to e6 or c6), White can play d2-d4, reaching a structure examined earlier in the chapter. After studying games such as Bondarevsky – Smyslov and Grischuk – Aronian in the notes, we can see that Black has

nothing to fear. White’s most popular plan involves 7.d3 Nd7 8.Nbd2 0-0 9.Nc4 f6 10.Nh4, but after 10...Nc5! we can get an excellent position by following Tomashevsky’s example in Game 36.   In the final part of the chapter we dealt with the Central Attack. This can be a dangerous weapon at club level, but after 6.d4 exd4 Black should be at least equal. The analysis section showed that Black has fine prospects after 7.Re1 b5, regardless of whether White chooses 8.Bb3 or 8.e5. The final game dealt with 7.e5 Ne4 8.Nxd4 0-0 9.Nf5 Nc5! which usually leads to a Berlin-type structure after White exchanges on c6. White can provoke a dark-square weakening on the kingside, but Black’s rapid development and light-square domination more than make up for it.

  1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6   Spassky’s 5.Nc3 page 240   The Worrall System 5.Qe2 page 241 5.0-0 Be7 6.Qe2 Game 38, page 242   Just Before the Main Lines 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 page 246 8.c3 (8.h3 page 246) 8...0-0 page 246 9.a4 Game 39, page 248 9.d4 Bg4 10.d5 Game 40, page 250 10.Be3 exd4 11.cxd4 d5! 12.e5 Ne4 13.Nc3 Game 41, page 255 13.h3 Game 42, page 257  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.  

Black is to move unless otherwise indicated.  

How should Black respond to the

advance of the white knight? (page 240)  

How can Black gain the initiative? (page 241)  

Suggest a course of action for Black. (page 249)  

Should Black take the exchange on offer? (page 255)  

How should Black continue? (page 245)  

Suggest a plan for Black. (page 256)  

Suggest a good manoeuvre for Black. (page 258)

Suggest a plan for Black. (page 246)   The main line of the Closed Spanish is reached after the opening moves 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3, when 9...Nb8 introduces the Breyer System, which will be covered in Chapters 11-13.   The next chapter will give special attention to 5.d3 and 6.d3, the latter being an especially popular move nowadays.  

Before then, we will look at the various ways in which White may deviate from the above sequence (except for those which involving exchanging on c6, which were covered in the previous chapter). In this chapter we will examine options like 5.Nc3, 5.Qe2, 6.Qe2, 8.a4 and 9.d4, the last of which is especially important. There are no special positional themes that apply to all the systems in the chapter; you simply have to get to know the pros, cons and specific details of each one. The Spanish has been around for centuries, and many of its sub-systems have gone in and out of fashion. All of the lines examined here have developed their own body of theory and many of them have been tested at the highest levels, so none of them should be taken lightly.

Spassky’s 5.Nc3 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.Nc3

This move does not really warrant a full illustrative game as it is hardly ever played nowadays. Still, it is worth having a quick look at the best way for Black to handle it.   5...b5 6.Bb3 Be7 7.d3 d6 The main way for White to utilize the knight’s position on c3 is to play Nd5 to eliminate the e7-bishop. However, we can counter by preparing ...Na5 to eliminate one half of White’s bishop pair.   8.Nd5 8.a4 b4 9.Nd5 Na5 gives Black an even more comfortable version of the main line.   8...Na5! 8...Nxd5?! is not advisable as White plays 9.Bxd5 with tempo, then plays c2-c3 and d3-d4 as in the classic encounter Spassky – Beliavsky, Reykjavik 1988.   9.Nxe7 Qxe7 10.0-0 0-0 Black is poised to eliminate the b3-bishop, with comfortable equality. We will follow a model game in which he went on to seize the initiative.   11.Bd2 11.Bg5 looks more natural, but Black can go for a similar plan with: 11...h6 12.Bh4 Nxb3 (12...g5 is given by Mikhalevski intending to reroute the knight to g7 via e8 – another attractive idea which gives Black a good game) 13.axb3 Qe6!? Intending ...Nd7 and ...f5.   11...Nxb3 12.axb3 Nd7 13.Re1

13...Bb7 This final developing move is necessary, as the immediate 13...f5? runs into 14.exf5 Rxf5 15.Nd4! followed by Nc6.   14.d4 f5! Black obtained the initiative and went on to win in Spassky – Yusupov, Linares 1990. If you remember this simple plan, it becomes pretty easy to deal with the 5.Nc3 variation. If you want more complete coverage of this line from Black’s side, Mikhalevski’s Open Spanish book is excellent.

The Worrall System This refers to set-ups involving an early Qe2, which may be played either before or after castling. Committing the queen is a slight concession, but White hopes to deploy his rook on the active d1-square later.   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.Qe2 5.0-0 Be7 6.Qe2 will be covered in Game 38 below.

5...Be7! 5...b5 6.Bb3 Bc5 is perfectly respectable, but the bishop on c5 does not really tie in with the rest of our repertoire; besides, we would have to learn a different set-up against 5.0-0 Be7 6.Qe2. 5...b5 6.Bb3 Be7 is not bad, but it allows White the extra option of 7.a4.   6.c3 The justification of Black’s last move is that 6.Bxc6 dxc6 7.Nxe5 Qd4 sees Black regain the pawn with an easy game.   6...b5 7.Bb3 7.Bc2 is an attempt to exploit the delayed ...b5, but it looks odd to put the bishop on a closed diagonal. 7...d5! is a good response; that being said, 8.d4!? leads to tricky play. A logical continuation is 8...Bg4!? 9.exd5 Bxf3! 10.Qxf3 Qxd5 11.Qxd5 Nxd5 12.dxe5 Nxe5 13.0-0 0-0 when Black’s lead in development and centralized knights offered full compensation for White’s bishop pair in Bisguier – Matanovic, Zagreb 1955.   7...d5! 7...0-0 invites a transposition to Game 38 below, but White has the extra option of 8.d4, a favourite line of Tiviakov. The text move does not give him so much freedom.

8.d3 This has been by far the most popular choice.   8.exd5 Nxd5 9.Nxe5 Nxe5 10.Qxe5 Nf6 11.0-0 0-0 12.d4 Bd6 can be compared with the note to White’s 9th move in the main game below. Black has lost a bit of flexibility, as he was forced to play ...Nf6 instead of ...Bb7, but he still has full compensation for the sacrificed pawn.   8...0-0 9.Nbd2 9.0-0 leads straight to Game 38 below, so the question is whether White can exploit the fact that he has not yet castled. Transferring the knight to g3 is the obvious try, but Black is well placed to meet it.   9...Be6 10.Nf1

In Klausch – Schulz, Hamburg 2014, Black could have seized the initiative with:   10...a5!N 11.Ne3 11.Ng3 a4 12.Bc2 dxe4 13.dxe4 Bc4µ   11...a4 12.Bc2 d4 13.Nf5 Bxf5 14.exf5 a3µ White has problems on the dark squares. It is not even clear if the last move is Black’s best, as 14...Qd7 and 14...dxc3 15.bxc3 b4 are also clearly better for him, so take your pick.   We will now turn our attention to the main line of the Worrall where White castles early.   GAME 38 Judit Polgar – Michael Adams Las Palmas 1994   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Qe2

6...b5! Perhaps the exclamation mark is a bit generous, but I want to remind you of why it was worth inserting 3...a6 as soon as the bishop landed on b5. Since 6.Qe2 (as well as 6.Re1, 6.d3 etc.) defends e4 and thus threatens to capture on c6 and e5, it is useful to have ...b5 available to stop this idea altogether. Having said that, 6...d6 is also quite a respectable way for Black to play. However, after 7.Bxc6†!? bxc6 8.d4 I think White is a bit better. A really nice example continued: 8...exd4 9.Nxd4 Bd7 10.c4 0-0 11.Nc3 Re8 12.h3 Bf8 13.Qf3 h6 14.Bf4 In Fine – Bernstein, New York 1941, White went on to centralize his rooks and complete the harmonious development of his forces.   7.Bb3 0-0 8.c3 d5! A continuation closer to the spirit of our Breyer main line would have been 8...d6, which was actually my original

intention. Now 9.a4 Bg4! was seen in the beautiful game Fine – Keres, Netherlands 1938, and I had no problem at all in entering the main line with 9.d4 Bg4!. What I didn’t like was the quite clever move order starting with 9.Rd1!, which was recommended by my editor Andrew Greet in Play the Ruy Lopez. Black has nothing better than to enter the closed position arising after 9...Na5 10.Bc2 c5 11.d4 Qc7 12.d5. This is quite playable for Black, and such positions are covered more fully in Marin’s A Spanish Repertoire for Black. However, the position is quite far away from the spirit of the Breyer complex, as we always want to be able to challenge the advanced d-pawn with the ...c6 break.

9.d3 9.exd5?! has been played in a lot of games, but Black gets an excellent version of the Marshall Attack: 9...Nxd5 10.Nxe5 Nxe5 11.Qxe5 Bb7 12.d4 Qd7 Black intends ...Rae8 and ...Bd6 with more than enough compensation for the pawn, as has been demonstrated in many practical games.   9...Be6! If it wasn’t for this move I would have found it hard to recommend something against the Worrall that I find completely fine for Black, at least in the practical sense. There are, of course, other respectable ways for Black to play, such as Karpov’s 9...d4 or the Kaufmann-recommended 9...Bb7, but both of them allow White to maintain annoying pressure along the a2-g8 diagonal. I would advise you to check Greet’s book to see exactly how White can cause problems.   To see the advantages of the text move, it is worth comparing it to the alternative: 9...Re8 This is actually what Adams played in the main game, but I have taken the liberty of changing the move order to show my preferred sequence. Essentially, the drawback of the rook move is that White gets to take advantage of his powerful bishop on b3, whereas my main line strives to neutralize this piece as quickly as possible. 10.Nbd2 Greet recommends 10.Re1!? to maintain the option of meeting ...Bf8 with Bg5, but I think the text move is trickier. 10...Bf8 10...h6 does not work due to 11.exd5 Nxd5 12.Nxe5 Nxe5 13.Qxe5 Bb7 14.Ne4 when Black’s compensation is inadequate. It is also a bad time for 10...Be6 in view of 11.Ng5 Bg4 12.Ndf3! h6 13.exd5! Nxd5 14.Qe4!.

11.exd5! 11.a3 h6 12.Ba2 Be6 13.Rd1 was the continuation of Polgar – Adams. I would be happy to reach this position with Black, but I will do so via the bolded move order to rule out the tricky option given in the main line of this note. 11...Nxd5 If my opponent captures a central pawn, I would prefer to be able to take it back. 11...Na5 is most likely a better move, when White can choose between trying to hold on to his extra pawn or retreating his bishop and playing an interesting middlegame. Maybe it’s equal, but I would rather not give my opponent this kind of choice. 12.Ng5!N This strong move puts Black under some pressure. Now you can see why neutralizing the b3-bishop is my top priority.

10.Rd1 This has been White’s most popular choice, and it keeps us on track to reach our target position from Polgar – Adams.   Of course 10.Ng5 makes little sense as 10...Bg4 is annoying for White.   It is too early for 10.exd5 Nxd5 11.Nxe5 Nxe5 12.Qxe5, as Black has excellent compensation after 12...Bd6 or 12...Re8.   Greet points out that 10.Nbd2 Nh5!? is irritating for White. Black could also play 10...h6 with a likely transposition to our main line.   Finally, 10.Bg5 usually only makes sense when Black’s bishop has retreated to f8. A strong reply for Black is: 10...Re8 11.Nbd2 h6 12.Bh4

12...Nh5! Black has an excellent position. A good practical example continued 13.Bxe7 Rxe7 14.g3 Nf6 15.a4 b4 16.a5 Rb8 17.Rfb1 Qd6 and Black went on to win the a5-pawn in Paehtz – Hebden, Lausanne 2001.   10...Re8 11.Nbd2 It is important to recognize that this move turns Ng5 into something of a threat, as ...Bg4 can be conveniently met by Ndf3. This explains Black’s next move.

11...h6! 12.a3 This is a typical Worrall move, ensuring that the bishop can remain on the a2-g8 diagonal in the event of ...Na5. However, it is slow and not the most testing move overall.   12.Nf1 was recommended by Greet, and subsequently tested by Adams. 12...Bd6 (Greet gives 12...Bf8 13.Ng3 Qd7 14.d4!? with interesting play; Black’s best reply looks to be 14...Bd6 with approximate equality) 13.Ng3 Ne7!? 14.d4 White eventually won in Adams – Gustafsson, Gibraltar (rapid playoff) 2010, but Black’s play can be improved with:

14...Bg4!?N (This is the most forcing move, although Black can also keep things tense with 14...Ng6!?N) 15.h3 Bxf3 16.Qxf3 exd4 17.Bxh6!? Bxg3 18.Qxg3 Ng6 19.Bg5 Nxe4! 20.Bxd8 Nxg3 21.Bxc7 Ne2† The tactical skirmish has resulted in a lively and roughly equal queenless position.   12...Bf8 13.Ba2 Finally we have transposed to the main Polgar – Adams game.   13...Qd7 Black is absolutely fine, the immediate plan being to bring his second rook into play on d8.   14.exd5 Nxd5 15.Ne4 This is a typical plan in the Worrall, but it tends to work better when Black’s bishop has gone to b7 rather than e6.   15...Nf4! Polgar was probably hoping to provoke the double-edged ...f5, but Adams’ move is much better.   16.Bxf4 exf4 17.Bxe6 Rxe6 18.Qc2 Rae8³ Black is in control; his excellent pieces more than make up for the slight damage to his pawn structure.

19.b4 Ne5 19...g5!? followed by a pawn assault with ...f5 and ...g4 looks gruesome for White.   20.Nxe5 20.Nd4 is well met by 20...Rg6! with various attacking ideas.   20...Rxe5 21.Nd2 21.c4 is met by 21...f3! and White is in trouble. Polgar tries to avoid this without resorting to the ugly f2-f3.   21...Re2! 22.Qb1 Qc6 23.Qc1 Qb6 24.d4 g5‚ Black is dominating, and is ready to play ...g4 followed by breaking through on the kingside after suitable preparation. Polgar tries to create some breathing room but only accelerates Black’s attack. Still, in a bad position, every move looks bad with hindsight.

25.h3?! h5 26.Qb1 Qf6 27.Nf1 g4 28.hxg4 hxg4 29.Qd3 Bd6 30.Rd2 f3 31.Rad1 Qh4 White resigned. Adams played excellently, with the exception of his slightly inaccurate move order in the opening. Remember, 9...Be6! is the way to go! 0–1

Just Before the Main Lines   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6  

8.c3 Otherwise Black will be able to eliminate the ‘Spanish Bishop’ with ...Na5.   8.h3 Na5 9.Nc3 0-0 10.d3 has occurred in quite a number of games over the years, but I will just show one nice example: 10...Bb7 (10...Re8 was played in Korchnoi – Portisch, Belgrade 1970, but the text move is more dynamic) 11.Bg5 c6 12.Ne2

12...Nh5! 13.Bxe7 Qxe7 In Fontaine – Sasikiran, Internet (blitz) 2004, Black was ready to take over the initiative with ...Nxb3, ...c5 and ...f5.   8...0-0 9.d4 This is one of the most important subvariations before the main line, and we will concentrate on it for the remainder of this chapter. The alternatives are:   9.h3 Nb8 is our main Breyer System, which will be covered in the next Chapters 11-13.   9.d3 will transpose to the system with d2-d3 and c2-c3, which is examined in Game 44 in the next chapter.   9.a4 is generally not a problem when ...d6 has been played, as Black has the convenient reply: 9...Bd7! (I have played 9...Bg4 successfully in my correspondence practice, but I now believe the text move to be more accurate) 10.d4 h6! Black intends ...Re8, ...Bf8 and – if allowed – ...exd4, ...Nb4 and ...c5. What I like about this line is that if White plays d4-d5, we always have the typical Breyer break with ...c6. Further details can be found in Game 39 below.

9...Bg4 Pinning the knight and putting pressure on White’s centre is the only way to question White’s last move. Although my database shows that fifteen(!) different moves have been played by White, I will only examine the two most logical ones which are played almost exclusively these days: 10.d5 and the more dangerous 10.Be3.   10.d5 Na5 11.Bc2 reaches a popular position, where I recommend 11...Qc8! followed by ...c6. The Norwegian analysts Johnsen and Johannessen claimed that White was better after 12.Nbd2 c6 13.b4 Nb7 14.dxc6 Qxc6 15.Bb2 but it seems to me that this assessment has not stood the test of time. In the notes to Game 40, you’ll see that I recommend 15...a5! 16.h3 Be6 17.a3 g6! when Black plans to transfer the f6-knight to f4 and the b7-knight to f7 at some point in the future. With the exception of this concrete line, where Black has to remember some specific moves, the second player is generally comfortable after 10.d5, as indicated by the Gashimov – Svidler game.   10.Be3 is actually quite a critical continuation. There are many systems available to Black, but none of them offer straightforward equality. After a preliminary check of various sources combined with some database searching and engine checking, I got the dark feeling that Black was playing with fire and might even be objectively worse already. White seemed to get relatively easy play and, even at high-level correspondence play, Black was managing to draw

positions which didn’t look at all suitable for practical over-the-board players. It was then that I remembered I had seen this variation covered in Sabino Brunello’s 2009 book Attacking the Spanish, where he reaches the same position via an Anti-Marshall move order. I have known Sabino since 2011 and have worked with him for about two years; we still meet from time to time at various chess events, where we discuss chess as well as life in general. As well as being a pleasant person and a good friend, I regard Sabino as a player with a fantastic understanding for the game. His score of 11/13 in the ‘C’ group at Wijk aan Zee 2013 was one of the best performances in the 75-year history of the tournament, and it was certainly not an accident. So, my analysis of this important variation is based on Sabino’s analysis and comments from his own book, double-checked of course with the help of the latest analytical tools. It came as no surprise to me that Sabino was correct in 2009 and his suggestions still seem strong. Sabino mentioned a game between Li and Tomashevsky as a model example of Black’s strategy, and it remains the best example I could find, so I have included it in this book as well.   GAME 39 Ole Norregaard Hansen – Mircea Bucur email 2008   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.a4 Bd7 10.d4 h6 In Chess Informant 126, Inarkiev points out that 10...Re8? loses to 11.Bxf7† Kxf7 12.axb5 axb5 13.Qb3† etc.

11.Nbd2 11.h3 Re8 12.Bc2 (12.Na3 Qb8 also causes no problems for Black) 12...Bf8 13.Bd2 was played in Quesada – Inarkiev, Baku (1.3 – rapid) 2015, and now according to Inarkiev himself, it is simplest to play 13...Qb8 14.Na3 Qb7 with a good position for Black.   11...Re8 12.Nf1 Bf8 What we have here is a version of the Smyslov System of the Spanish, where White has played a2-a4 in place of h2h3. Superficially, this might sound like an achievement for White, but in fact the opposite is true, as the early a2-a4 increases Black’s possibilities. The point is that the c6-knight tends to be Black’s ‘problem piece’ in most Closed

Spanish systems. In the Breyer we try to solve the problem as quickly as possible by investing two tempos with ...Nb8d7; in the Chigorin Black plays ...Na5 to gain time for ...c5 before redeploying the knight somehow; while in the Smyslov Black often goes for the manoeuvre ...Na5-c4-b6 at some point. In the present position though, Black has a much easier solution to the problem: he can either play ...Na5 and ...c5, with a pleasant version of a Chigorin due to the softening of White’s queenside, or play ...exd4, intending to meet cxd4 with ...Nb4, exploiting the premature a2-a4.   13.Bc2 This move is designed to meet ...Na5 with Ne3, while against other plans the knight can go to g3. Another idea is to prepare d4-d5 and try to hold on to the d5-pawn with b2-b3 and c3-c4.   13.d5 Ne7 followed by ...c6 is fine for Black.   13.Ng3 Na5 14.Bc2 c5 15.d5 c4!

This reaches a structure that we will encounter later in the Breyer section; this is an improved version as the bishop is already on the optimal d7-square. 16.h3 Qc7 Black can consider following up with ...Reb8 and maybe even ...b4, while the plan played in the game is also thematic and fine. 17.Be3 Nb7 18.Nd2 Nc5 19.Qf3 Nh7 20.Ra2 Reb8 21.axb5 axb5 22.Rea1 Qc8 23.Qd1 Rxa2 A draw was agreed in Dimitrov – Delchev, Bulgaria 1995. After 24.Rxa2 a logical continuation would be:

24...Be7! Intending ...Bg5 with a favourable bishop trade. This manoeuvre is worth remembering, as it is applicable in many similar situations.

13...exd4! 14.Nxd4 After 14.cxd4 Nb4 15.Bb1 bxa4 (15...c5!? is also fine) 16.Ng3 c5 Black had a great game in Bernal Varela – Jimenez, email 2006.   14...b4! 15.Bd2 Qb8! 15...Nxd4 16.cxd4 c5 would have been okay if it were not for 17.e5!.   15...d5!? is a decent alternative. 16.exd5 Nxd4 17.Rxe8 was seen in another high-level correspondence game, Serafim – Lafarga Santorroman, email 2006, with the White player being the only Greek player so far to achieve the Correspondence GM title. Here the engine evaluates 17...Qxe8N as equal, although capturing with the bishop was not

at all bad, and enabled Black to neutralize his opponent’s slight initiative and draw the game.   Still, I like the game continuation even more, and I think that the last two strong moves completely bury White’s hopes of achieving anything from this variation. The Danish Senior International Master on White’s side failed to prove anything different in this game.

16.Nxc6 Otherwise Black can think about ...Nxd4 followed by ...c5.   16...Bxc6 17.f3 a5 17...d5!? is also fine for Black.   18.Be3 bxc3 19.bxc3 d5 20.exd5 Nxd5 21.Qd3 g6 The position is absolutely level, although in a practical game either side could keep playing for a win. In this correspondence game, however, the game soon ended in a surprising draw.

22.Rab1 Qd8 23.Bb3 Rb8 24.Qc4 Qd7 25.Rbd1 Re6 26.Rxd5 Bxd5 27.Qxd5 Rd6 In view of 28.Qc4 Rc6 with an unavoidable repetition, the players agreed a draw. ½–½   GAME 40 Vugar Gashimov – Peter Svidler Nice (blindfold) 2010   It is unusual to have a blindfold encounter as a main illustrative game, but the quality of the game was generally high and the way Black handled the opening was exemplary.   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.d4 Bg4 10.d5 Na5 11.Bc2  

11...Qc8! This is the simplest and most reliable path to equality. Black just wants to play ...c6 and take with the queen.   11...c6 12.h3 is the other main line, when Black usually takes on f3. In case you haven’t looked at this line before, I will show you an important tactical point: 12...Bd7?! allows 13.Nxe5! dxe5 14.d6 when White gets an easy edge with the bishop pair – this is why it helps to put the queen on c8 before playing ...c6.   12.h3 12.Nbd2 c6 13.b4 Nb7 14.dxc6 Qxc6 15.Bb2 is another popular continuation, when I propose:

15...a5! (15...Nd8 followed by ...Ne6-f4 has been much more popular, but I prefer the plan seen in this correspondence game) 16.h3 Be6 17.a3 g6! Black plans to put his knight on h5 and play either ...f6 or ...f5. Later the b7-knight (which for the moment does a good job of covering a5 and d6) can go to f7. 18.Bd3 White usually plays this move with the

idea of playing Rc1 and c3-c4 but this plan is less dangerous when Black has played ...a5.

18...Nh5 19.Nf1 Nf4 20.Ne3 Rfc8 21.Qd2 f6 22.Bc2 Bf8 23.Red1 Nd8 Black’s regrouping worked out perfectly and he went on to win in Yloenen – Laube, corr. 2010.   12...Bd7 Black’s plan is simple: play ...c6 and recapture with the queen.

13.Nbd2 Sometimes White develops the bishop first: 13.Bg5 c6 14.dxc6 Qxc6 15.Nbd2 This has been tested by both Kamsky and Grischuk, but Black has no problems after: 15...Be6! Intending a quick ...Nc4.

16.Rc1 h6 17.Bh4 17.Bxf6 Bxf6 allows Black to prepare ...d5, for instance: 18.Bb1 Rfd8 19.Qe2 Rac8 20.Nf1 Nc4 21.b3 Nb6 22.Ne3 d5 23.Ng4 Bxg4 24.hxg4 In Kamsky – Gustafsson, New York (blitz) 2004, Black missed the chance to get a large advantage with 24...d4!N, although he won the game anyway. 17...Nc4!? Black can also play a waiting move with the idea of meeting Nf1 with ...Nc4, but the text move works just fine. 18.b3 After 18.Nxc4 bxc4 Black will apply pressure along the b-file.

18...Nxd2 18...Nb6!?N is also fine for Black, but there is no particular need to deviate. 19.Qxd2 Black was comfortable in Kamsky – Johannessen, Kemer 2007. There are several decent moves, but my

suggestion would be: 19...Rfc8N 20.Bd3 Qb7 Black has the more pleasant position.

13...c6 14.dxc6 This is the usual move, and there is nothing better.   14.b4 Nb7 15.dxc6 Qxc6 16.Bb2 This has been played quite a few times, but White has a slightly worse version of the similar variation from the note to move 12.

16...a5!? Black can also play the immediate 16...Nd8 followed by ...Ne6 without fearing that his bishop will be left short of squares on g4.

17.a3 Nd8 This works out well because the plan of Bd3, Rc1 and c4 is not dangerous at all. For example: 18.Bd3 Ne6 19.Bf1 axb4 20.cxb4 Qb7 21.g3 Bc6 22.Bg2 Bd8 23.Nh4 g6 24.Rc1 Ra6

25.Qe2N In Smagin – Yemelin, Elista 1995, White went wrong with 25.Qf3?! allowing 25...Ne8! followed by ...Bxh4, ...Nf4 and ...f5, with a clear advantage for Black. The text move was recommended by Dolmatov, with the idea to have the f3-square available for the knight in the event of ...Ne8, but Black keeps an excellent position after: 25...Bb6! Black intends a combination of ...Bd4 and ...Qa7, in order to obtain the d4-outpost for the knight and weaken the a3pawn. I prefer Black.   14...Qxc6 15.Nf1 Nc4 Black reroutes the knight, which is traditionally his problem piece in the Spanish, to the superior b6-square.   15...Nb7!? 16.N3h2 Nc5 is an equally tempting alternative – here too, Black improves the knight and intends to push the queenside pawns. After 17.Ng3 Rfe8 18.Qf3 a5 19.Nf5 b4 20.Nxe7† Rxe7 21.Bg5 Re6 Black was more than okay in McShane – Sokolov, Bundesliga 2002.   15...Rfc8!? followed by ...Bd8!? worked well for Black in Espinosa Veloz – Karjakin, Baku (1.2) 2015, and is another line that merits further investigation.   16.Ng3 16.a4 Nb6 17.a5 Na4! was at least equal for Black in Vogt – Smagin, Berlin 1988, a game where White experienced problems defending his a-pawn.

16...Rfe8 17.b3 17.a4 brought White an eventual victory in Kamsky – Svidler, Nalchik 2009, but I guess had the Russian super-GM faced this in the current game, he intended to use Smagin’s plan of 17...Nb6! 18.a5 Na4!, when Black faces no problems whatsoever.   17...Nb6 18.Be3 18.Bb2 is probably better, although after 18...a5!„ Black remains at least equal.   18...Rab8! There is nothing wrong with 18...h6!? 19.Nh4 Be6 20.Nhf5 Bf8 intending ...d5. However, Svidler’s move threatens to take on c3 and thus puts White under immediate pressure.

19.Rc1 White decides to sacrifice the c3-pawn. He did not have much choice, as 19.Qd2 Rec8 renews the attack, and 20.Ne2? would lose the more valuable e4-pawn.   19...Qxc3 20.Bd3 Qa5 21.Bd2 21.a4 is offered by my machine, the point being that White will have more chances of counterplay if he opens another file with axb5. Thus I think the most ambitious reply is 21...b4! where it is not easy for White to claim enough compensation. For example, 22.Qe2 Be6! 23.Nd2 allows 23...d5!.   21...Qa3 22.Rc2 Rec8 23.Bc1 Qa5 24.Bd2 b4 25.Nh4 Rxc2 26.Qxc2 Rc8 27.Qb1 g6 28.Nf3 Bb5µ Black has done excellently and stands clearly better with his extra pawn. The rest of the game contains some interesting moments, but is not so important for our opening study. Please also remember that this was a blindfold game, so I will keep the annotations fairly light.  

29.Bc2 Nbd7 30.Rc1 d5 31.exd5 Nxd5 32.a4 Bc6 33.Be4 Bb7 34.Rxc8† Bxc8 35.Qe1 N5f6 36.Bd3 Bb7 37.Bc4 Qc5 38.Ng5 Qc6 39.Nf3 a5 40.Bc1

40...Nb6? Svidler tries to remove the annoying c4-bishop, but commits a tactical oversight.   40...Ba6? also doesn’t help because of 41.Nxe5.   So, logically 40...Bd6! is best, defending e5 before playing either ...Ba6 or ...Nb6.   41.Bxf7†! Kg7 41...Kxf7? loses the queen to 42.Nxe5†.   42.Qe3 Qc3?! 42...Nfd7! 43.Be8 Kf8 would have kept some advantage for Black, as White would have had to relinquish the bishop pair.   43.Be6 43.Nxe5! was a safe pawn grab, although 43...Qxe3 44.Bxe3 Nbd7 45.Nxd7 Kxf7 46.Ne5† Ke6 47.Nc4 Bd5! reaches an endgame where Black’s bishop pair enables him to hold. A similar scenario actually occurred in the game a few moves later.

43...Qxe3 44.Bxe3 Nbd7 45.Kf1 e4 46.Nd2 Nc5 47.Bc4 Nfd7 48.Bb5 Kf6 49.Bxd7 Nxd7 50.Ngxe4† Ke6 Black is a pawn down, but Svidler holds the draw comfortably.   51.f3 h5 52.Kf2 Bd8 53.Bd4 Bd5 54.Ke3 Bc7 55.h4 Bd8 56.g3 Bc7 57.Ng5† Kf5 58.Nge4 Ke6 59.Ng5† Kf5 60.Nge4 Ke6 ½–½   Despite the imperfections in the above game, I think it is fair to conclude that Black is doing well in the theoretical battle. Let’s turn our attention to the more critical 10.Be3.   GAME 41 Li Chao – Evgeny Tomashevsky Ningpo 2008   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.d4 Bg4 10.Be3

10...exd4 11.cxd4 d5! 11...Na5 has been the most popular choice, but I was not completely happy with it. The text move leads to a blocked centre where Black’s active pieces give him fine prospects.   12.e5 Ne4 13.Nc3 13.Nbd2!? also exists, and is actually quite a sneaky move. The key point occurs after: 13...Nxd2 14.Qxd2 Bxf3 15.gxf3 Bb4 16.Qc2

Instead of winning the exchange, Black needs to remember to play 16...Na5! with at least equal chances.   By the way, the inclusion of 13.h3 Bh5 makes no great difference, and will either be similar to the above line after 14.Nbd2 or transpose to the next main game if White goes for 14.Nf3.  

13...Nxc3 14.bxc3 Na5 At first I was tempted by the idea of 14...Qd7, which is designed to meet 15.Bc2 with 15...Bf5, until I discovered that 16.Bg5! gives White an annoying initiative. Therefore I decided to stick with Brunello’s original suggestion.

15.Bc2 Nc4 16.Qd3 16.h3 is the other main option, when 16...Bh5 transposes to the next game.   16...g6 17.Bh6 Re8 18.Qe2 a5! Kaufmann recommends 18...Qd7 19.h3 Bf5 and says that “Black is close to equal.” This has been tested in many correspondence games and seems reliable enough, but the game continuation is more ambitious and troublesome for White. It also ties in nicely with the next main game, where the same plan again works well.

19.h3 Be6 20.Bd3

20.Nd2N gives White a small edge according to Kaufman, which is why he prefers 18...Qd7.

However, I don’t see any problem after 20...c6, intending ...b4 with similar ideas as in the main game. (20...Rc8 is the mysterious move given by Kaufman without further details.)   20.Nh2 is another move mentioned by Kaufman, who gives 20...b4 21.Ba4 as slightly better for White. However, Black can improve with 20...Qd7N followed by ...b4, while 20...c6!?N 21.Ng4 b4 also seems fine.   20...c6 21.Nh2 b4! The idea is to play ...bxc3 followed by ...c5.   22.cxb4 axb4 23.Ng4

23...Ra3! 24.Rac1 Qb6

This is a logical move, but it gives White a chance to liquidate to a level position.   If you want a winning try, then 24...Qa8!N is my suggestion. 25.Rc2 (25.Bb1 is met by 25...Qa7! when 26.Nf6† does not work due to 26...Bxf6 27.exf6 Qxd4µ) 25...Bd8!? 26.Rd1 (26.Nf6† Bxf6 27.exf6 Bd7 is Black’s idea) 26...Qa4!? Black can fight for the full point. My analysis goes for many more moves, but nowhere do I see a forced draw or Black becoming worse.   25.Red1? 25.Nf6† Bxf6 26.exf6 Qxd4 27.Bxc4 dxc4 28.Rxc4 Qxf6 29.Rxb4 reaches a position where neither side has real winning chances.   25...Rea8 26.Bb1 Qd8µ

Black has regrouped perfectly. His pieces are excellent and he can target the two weaknesses on a2 and d4.   27.Rf1 Kh8 27...Bg5!? was also strong. After 28.Bxg5 Qxg5 29.Nf6† Kh8 the knight looks pretty on f6, but it threatens nothing while the b1-bishop looks pathetic. By contrast, the e6-bishop and a3-rook are dangerously posted for a kingside attack.   28.Ne3 Bg5 29.Bxg5 Qxg5 30.Nc2 Rc3 I am not sure what both players missed after 30...Rxh3!–+ but the game continuation also leaves White in trouble.   31.f4 Qg3 32.Nxb4 Rxc1 33.Rxc1 Qxf4 34.Rd1

34...Rb8 35.Nc2 Bf5 36.Rf1 Qd2 37.Qf3 Qd3 38.Qf4 Nd2 39.Re1 Rxb1 40.Rxb1 Nxb1 41.e6 Bxe6 42.Qe5† Kg8 43.Qb8† Kg7 44.Qe5† Kh6 After 45.Qf4† g5 46.Qf6† Kh5 47.g4† Kh4 the checks die out, and White must either allow a queen exchange or be mated. Therefore he resigned. 0–1   GAME 42 Axel Rombaldoni – Sabino Brunello Alghero 2011   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.d4 d6 9.c3 Bg4 10.Be3 exd4 11.cxd4 d5 12.e5 Ne4 13.h3 Bh5 By nudging the bishop, White gives himself the option of flicking in g2-g4 a few moves down the line.

14.Nc3 Nxc3 15.bxc3 Na5 16.Bc2 Nc4 17.g4 Bg6 18.Bf5 This is the other main tabiya we can expect to encounter when facing the 10.Be3 variation.   18...a5!

Black attacks on the queenside using the same plan as in the previous game. An added benefit, with the bishop on g6 instead of e6, is that ...Ra6 might be a useful way of defending along the sixth rank.   19.Qd3 Rombaldoni is usually well prepared, especially having faced Sabino in many previous games; I am sure he also knew about Sabino’s recommendations in Attacking the Spanish.   19.Nh2 b4! 20.cxb4 axb4 21.f4 Ra3 22.Nf1 c5 23.Kh2 was seen in Vocaturo – Brunello, Internet 2007, and now Black

could have obtained a winning position with:

23...Bxf5!N 24.gxf5 Bh4 25.Re2 Qc8 White’s position collapses. It is important to remember this idea of exchanging on f5 at an opportune moment. Optically White seems to get an impressive pawn mass, but in reality it often turns out to be weak and unwieldy.   19.Bf4 was given as the main line in Attacking the Spanish, and is the only move examined by Kaufman. After 19...b4! 20.h4 Black once again gets an excellent game with:  

20...Bxf5! (20...Bxh4 is given by Kaufman, but this needlessly complicates things without any special gain) 21.gxf5 Qc8 22.Qd3 f6! Black was better once again in Ronchetti – Brunello, Fiuggi 2007.   19...Bxf5 This is fine, but 19...b4! looks even better; Sabino suggested this move in his book but perhaps he forgot his analysis.

A high-level correspondence game continued: 20.Re2 Bxf5 21.gxf5 f6! 22.e6 Qe8

23.Bf4 Qh5 24.Kg2 Bd6 25.Bg3 bxc3 26.Bxd6 Nxd6 27.Rc1 Nxf5 28.Rxc3 Rfc8 29.a4 c6 30.Rc5 Ne7 31.Qc3 Qf5 32.Re3 Ng6 White was in trouble and only lasted another ten moves in Swan – Lybin, email 2007.   20.gxf5 f6! 21.Bf4 In the event of 21.e6 Qe8 Black will attack the f5-pawn with ...Qh5 and ...Nd6.

21...Qd7!? 21...b4 also seems strong, but Sabino is willing to lose a tempo to force White to play e5-e6.   22.e6 Qe8 23.Nh2 A logical manoeuvre. The c4-knight is strong, so White wants to exchange it by bringing his own horse to e3.  

23...c6 24.Nf1 b4 25.Ne3 Qh5

26.Nxc4?! This initiates a forced sequence which results in a difficult position for White.   26.Kg2 is better, although Black remains at least equal after 26...Rfd8!?, with ideas of attacking the soon-to-be-weak (after ...bxc3) d4-pawn.   26...dxc4 27.Qxc4 Qxh3 28.Qxc6 Qxf5 29.Bg3 Rad8 This is not bad, but 29...Qg5! intending ...f5-f4 was more accurate.

30.c4 Rxd4 31.Rad1 Rxd1 31...Rg4! with the idea to move the queen and push the f-pawn seems better for Black, as 32.Rd7? is refuted by 32...Rxg3† 33.fxg3 Bc5† and Black wins.

  32.Rxd1 Qc5? Sabino blunders away an otherwise well-played game.   32...Qg5! would have forced White to find 33.Kf1! (33.Qb7? f5–+) 33...f5 34.Bd6! when there are several playable moves, but all of them should result in a draw with best play.  

33.Qb7! Re8 34.Rd7 h5 35.Rc7 Suddenly it’s all over. Despite the eventual defeat, Sabino (and especially Lybin in the note to Black’s 19th move) made a convincing case for the way Black handled the opening. 1–0   Conclusion   In this chapter we dealt with an assortment of systems after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6. Beginning with Spassky’s 5.Nc3, we saw that Black is doing fine after 5...b5 6.Bb3 Be7 7.0-0 d6 8.Nd5 Na5 9.Nxe7 Qxe7. The simple plan is to castle, eliminate the b3-bishop and prepare the ...f5 advance.   We then moved on to the Worrall System, beginning with the 5.Qe2 version. I found the most accurate move order to be 5...Be7! 6.c3 b5 followed by 7...d5. We then studied the main line with 5.0-0 Be7 6.Qe2 b5 7.Bb3 0-0 8.c3 d5 9.d3. In this position the super-accurate (in my eyes anyway) 9...Be6! is not only easy to learn and play, but also excellent from a theoretical point of view.   The next topic was 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0, when White has a couple of ways to deviate from the main lines. 9.a4 Bd7 10.d4 Re8 11.Nbd2 h6 12.Nf1 Bf8 gives Black a comfortable version of the Smyslov System, the main plan being to play ...Na5 and ...c5. If White pre-empts this with 13.Bc2, Black gets a fine game with 13...exd4! 14.Nxd4 b4 15.Bd2 Qb8!, planning to exchange on d4 and play ...c5.   The most important part of the chapter was devoted to 9.d4 Bg4, when White has two main options.   In the event of 10.d5 Na5 11.Bc2 Qc8! Black intends to play ...c6 and recapture with the queen, and only then think about improving his queen’s knight with ...Nc4-b6 or ...Nb7-c5. An important idea occurs after 12.Nbd2 c6 13.dxc6

Qxc6 14.b4 Nb7, when I recommend playing ...a5 in order to prevent White from building an easy queenside initiative with the plan of Bd3, Rc1 and c3-c4.   The other main line is 10.Be3, when I recommend 10...exd4 11.cxd4 d5 12.e5 Ne4 13.Nc3 Nxc3 14.bxc3 Na5 15.Bc2 Nc4, when we looked at two games. After 16.Qd3 g6 17.Bh6 Re8 18.Qe2 a5! Black intends ...c6 and ...b4, with a great game. Finally, 16.h3 Bh5 17.g4 Bg6 18.Bf5 can once again be met by the thematic 18...a5 intending ...b4. Another important motif is to play ...Bxf5 followed by ...f6 at the right moment, turning White’s aggressive-looking pawns into weaknesses.

  1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6   5.d3 page 263 5...d6! 6.c4 page 263 6.c3 g6 7.Nbd2 Bg7 8.Nf1 0-0 9.Bg5 (9.Ng3 page 265) 9...d5! Game 43, page 265 Main Line 8.0-0 0-0 9.Re1 Re8! page 268   The Fashionable 6.d3 5.0-0 Be7 6.d3 b5 7.Bb3 d6 8.c3 Game 44, page 269 8.a4 (8.h3 page 274) 8...Bd7! 9.Be3!? Game 45, page 274 (9.c3 page 275; 9.Bd2!? page 275) Anand’s concept with 8.a3! page 278 8...0-0 9.Nc3 Na5 page 279 8...Na5! 9.Ba2 c5 Game 46, page 279  

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How should Black proceed? (page 264)  

Suggest a course of action for Black. (page 276)

Suggest a strong plan for Black. (page 266)  

The last moves were 12.Bxe6 fxe6.

How can White secure an edge? (page 279)  

A subtle move is required. (page 271)  

White has just retreated with 16.Nb1.

Find a strong reply. (page 280)  

White has just played 9.Bd2.

How can Black exploit this? Suggest a good plan for Black. (page 275) (page 281)   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 In this chapter we will deal with systems involving an early d2-d3. These can be categorized according to whether White plays it on the 5th or the 6th move:   5.d3 has the practical advantage of eliminating the Open Spanish (5.0-0 Nxe4), but the drawback is that it commits

White to a slower plan of development. This allows Black to modify his own plans with 5...d6 followed by ...g6. I will explain more about this in the theoretical section below.   5.0-0 Be7 6.d3 is more challenging, and extremely popular nowadays. Things are a little different here, as we don’t have the option of a kingside fianchetto. Instead we will go for normal development with 6...b5 7.Bb3 d6, reaching an important tabiya where 8.c3, 8.a4 and 8.a3 will all receive close attention.

5.d3 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d3 This has the potential to be an annoying move if you don’t like putting your bishop on c5.

5...d6! 5...b5 6.Bb3 gives Black a choice. 6...Bc5! is theoretically the most solid move, and I would recommend Mikhalevski’s book for expert coverage of it; unfortunately, it does not really resemble any other part of our repertoire, so I don’t consider it an ideal choice for this book. 6...Be7 is the move we would like to play, but 7.a4 is an annoying reply. Black can choose between 7...b4, 7...Rb8 or 7...Bb7, but none of them quite matches up with my repertoire choices in other variations involving an early a2-a4.   The text move prepares a kingside fianchetto, and has a couple of key advantages. Not only is Black’s set-up easy to learn and understand, but it also leads to many Breyer-type middlegames. In some cases we may even transpose to a Breyer-like position where Black is several tempos ahead, as in the Breyer main line we will often fianchetto this bishop anyway after ...Be7, ...Re8, ...Bf8, ...g6 and ...Bg7.   6.c4!? This is known as the Duras Variation. White sets up a clamp on the light squares and can look for a suitable opportunity to play d3-d4 later. The pawn structure is highly unusual for the Spanish, so we should take care when meeting it.

6...g6 7.Nc3 Bg7 8.0-0 The immediate central advance is playable, but nothing special: 8.d4 exd4 9.Nxd4 Bd7 10.Nxc6 White had better make this exchange, otherwise Black will just take on d4 and a4, forcing White’s knight to recapture onto a poor square. 10...Bxc6 11.0-0 0-0

12.Bxc6 White will have to make this exchange at some point, and he may as well do it now to take away the threat to the e4-pawn. 12...bxc6 13.Bg5 Black’s structure is slightly damaged, but he has enough queenside counterplay to keep the balance. A good

example continued: 13...Qb8 14.b3 Re8 15.Qc2 Nd7 16.f3 Nc5 17.Rad1 a5 18.Be3

18...Qb4 18...a4!?N 19.Bxc5 axb3 20.axb3 dxc5 is also fine for Black. The tripled pawns control several important squares, the b3-pawn is weak and the bishop has a great outpost on d4. 19.Bd2 Qb6 20.Be3 Qb4 21.Bd2 Qb6 22.Be3 ½–½ Istratescu – Kritz, Peristeri 2010.   8...0-0 9.h3 If White tries to do without this move, Black will take control over the d4-square with ...Bg4 followed by ...Nd7-c5e6.

9...Nd7! 10.Bg5

10.Bxc6 bxc6 11.d4 looks fairly sensible, but after 11...exd4 12.Nxd4 Bb7! Black’s bishops have the potential to become dangerous on the long diagonals. 13.Nde2 c5 14.Re1 was seen in Yegiazarian – L’Ami, Stepanakert 2005, when the Dutch Grandmaster could have played 14...Qh4!N with a growing initiative.   10...f6 11.Be3 f5 11...Nc5?! 12.Bxc6 bxc6 13.d4²   12.exf5 12.Bg5 was seen in Green – Greet, Hastings 2009, when I think my editor-to-be should have played 12...Bf6!, with similar ideas as in the main line.   12...gxf5 13.Bg5 Bf6 14.Bh6 Re8 14...Rf7!? followed by ...Bg7 and ...Qf8 would also have been fine for Black.

15.Qd2 Nc5 16.Bd1 Ne6 Black enjoyed slightly the better chances in Green – Greet, Hamilton 2010.   Let’s now move on to the more typical Spanish set-up involving a pawn on c3 instead of on c4.   GAME 43 Magnus Carlsen – Fabiano Caruana Wijk aan Zee 2010   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d3 d6 6.c3 g6

7.Nbd2 7.Bg5 Bg7 8.Nbd2 0-0 gives White nothing better than transposing to the main game with 9.Nf1.   7...Bg7 8.Nf1 According to Mihail Marin’s article in NIC Yearbook 107, where he examined the ins and outs of his own experience defending this fianchetto variation, this is White’s most dangerous continuation. That being said, I find it hard to believe that White has anything close to an advantage if Black follows Caruana’s logical approach in this game.   8.0-0 is the main line, which will be examined in the next main game.

8...0-0 9.Bg5 After studying Caruana’s reaction I find it hard to believe this is White’s best, despite the praise it gets from Marin. Still, is there a dangerous alternative?

  9.Ng3 d5 10.Qe2 10.Bxc6 bxc6 11.Nxe5 is met by a motif that will be well-known to fans of the King’s Indian Defence: 11...Nxe4! Black gets his pawn back with a fine position, as 12.Nxc6?! only leads to more trouble for White after 12...Nxc3! or 12...Nxf2!. 10...Re8 11.0-0 b5 12.Bc2 White’s set-up is not too threatening, but Black still has to decide what to do with his position. Mikhalevski comes up with an instructive plan.

12...a5! Black gains space and prepares to meet a2-a4 with ...b4, after which White will no longer have the option of a4a5. 13.Bg5 White wants to put some pressure on d5 before Black plays ...h6 and stops this move, but Black can play: 13...d4! 14.Bb3 dxc3! 15.bxc3 a4 16.Bc2 Mikhalevski thinks 16.Bd5 is an improvement, but 16...Qd6 is still at least equal for Black. 16...Qd6! Black had an excellent position in Kovalev – Mikhalevski, Israel 2008, the main plan being to play ...Nd8-e6 and exert pressure on the weak d-pawn.   9...d5! Marin’s article focuses on 9...h6 10.Bh4 Ne7 followed by ...c6 and ...Qc7, aiming for queenside counterplay with ...b5 and so on. However, Caruana’s move seems more logical and just stronger overall.  

10.Qe2 Carlsen’s idea is to keep the centre stable, à la Steinitz, and attack on the flank with h4-h5.   10.exd5 Qxd5 11.Ne3 Qd6!? leaves Black with an absolutely fine position.   10...Qd6! 11.Bxf6?! Carlsen comes up with an excessively provocative plan, but Black’s position was comfortable in any case.   11...Bxf6 12.Ne3 12.exd5 Qxd5 13.Ne3 Qd6 14.Nc4 Qe6 15.Bxc6 Qxc6 16.Ncxe5 Bxe5 17.Nxe5 Qxg2³ is line given by several commentators, illustrating that Black is fine both strategically and tactically.

12...Ne7! 13.Bb3 c6 14.h4 Be6 15.Ng5

This seems like a try to mix things up, but it actually has a defensive purpose above all. The point is seen after 15.h5 a5! 16.a4? Otherwise ...a4 will be unpleasant, but this is even worse! 16...dxe4! 17.Bxe6 exf3 Black wins a piece.   15...Bd7 16.Rd1 Rad8 Regarding our opening study, we can safely stop here and claim a success for Black. The rest of the game is a demonstration of high-quality chess by both players and especially of how resourceful the current World Champion can be when he gets into trouble! I will keep the remaining annotations light, as the game has been analysed in various places.

17.Qf3 h6 18.Nc4! dxc4 19.dxc4 Nd5! 20.Nh3 h5! 21.Qg3 Bg4 22.Rd2 Bxh4!? 23.Qxh4 Qf6 24.Qxf6 Nxf6 25.Ng5 c5 26.f3 Bc8 27.Ba4 Kg7 28.Rxd8 Rxd8 29.b4 Rd3 30.bxc5 Rxc3 31.Kd2 Rxc4 32.Bb3 Rxc5 33.Nxf7 a5 33...Ne8! was better, preventing Nd6, as some commentators have pointed out.  

34.Rc1 Rxc1 35.Kxc1 a4 36.Bc4 b5 37.Nd6! Carlsen saves the piece and escapes to a knight endgame where he has some drawing chances, although Black should still be winning with best play.   37...bxc4 38.Nxc8 h4! Caruana wants to put the knight on f4, take the g2-pawn and march his h-pawn to glory.

39.Nb6 Nh5 40.Nxc4 Kf6 40...Nf4! is the winning move according to Mikhalevski’s analysis on ChessPublishing: 41.Nxe5 (41.Ne3 is met by 41...Nxg2! 42.Nxg2 h3 and the h-pawn cannot be stopped) 41...Nxg2 42.Ng4 Ne1 Black is winning, although it takes several more computer-inspired moves to prove it – not at all easy to calculate or judge correctly by Caruana on his 40th move.   41.Kd2 Nf4 42.Ke3! Nxg2† 43.Kf2 Nf4 44.Nb2 a3 45.Nc4 By now White is more or less out of danger, and Magnus holds comfortably.

45...Nd3† 46.Kg2 Kg5 47.Nxa3 Kf4 48.Nc2 Nb2 49.Nb4 h3† 50.Kxh3 Kxf3 51.Kh4 Kf4 52.Nd5† Kxe4 53.Ne7 Kf3 54.Nxg6 e4 55.Ne5† Kf4 56.Ng4 Na4 The game was interesting and rich in ideas, but the opening variation should not concern us, as Caruana showed that 8.Nf1 0-0 9.Bg5 d5! gives Black pleasant equality. ½–½  

5.d3 – Main Line with 8.0-0   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d3 d6 6.c3 g6 7.Nbd2 Bg7 8.0-0 0-0 9.Re1 In this popular position, it is important for Black to select the right move order.  

9...Re8! As Marin points out, this enables Black to avoid certain difficulties. The main point is to wait for the knight to go to f1 before giving White a target on the queenside.   The value of this move order can be appreciated if we compare the alternative: 9...b5 10.Bc2! 10.Bb3 Na5 11.Bc2 c5 gives Black a healthy position which can be strengthened with moves like ...Re8, ...Bb7 and perhaps later the Breyer manoeuvre ...Nc6-b8-d7. The problem for White is that his main plan of d3-d4 is unlikely to be of much benefit to him, as Black will always exchange twice on d4 and take advantage of the already excellent position of the bishop on g7. The last detail worth mentioning is that is 12.Nf1 should be met by 12...h6! to stop Bg5 before continuing with development. 10...Nb8! Marin gives this as Black’s best, and I agree. If Black starts with a move like 10...Re8 (or 10...Bb7), White has 11.a4! Bb7 12.b4² as in Nepomniachtchi – Kamsky, Plovdiv 2010, with the idea of Nb3. 11.a4 Bb7 12.b4 Nbd7 13.Nb3! You will become more familiar with this manoeuvre during our study of the main lines of the Breyer. White has obtained a somewhat favourable version of this type of position, which is why I recommend a different move order for Black. However, I would like to show that, even here, Black’s position is solid enough to equalize with a few more accurate moves.

13...Rb8! We will follow a game which features model play for the next few moves. 14.axb5 axb5 15.Be3 Re8 16.h3 d5 17.Nfd2 Bf8 18.Qb1 Qc8 19.Qb2 dxe4 20.dxe4 20.Nxe4 would have been met by 20...Nd5. 20...c5! Black had equalized in Feuerstack – Markowski, Wroclaw 2010.   10.Nf1 This is virtually an automatic move for White, but now the knight is further away from the queenside.

10...b5 11.Bb3 Retreating to c2 does not make as much sense when the Nb3 manoeuvre is unavailable.

11...h6! Black is ready to play ...Na5 followed by ...c5 with an excellent game, just as described in the 9...b5 10.Bb3 line mentioned above.   Just remember that the immediate 11...Na5?! 12.Bc2 c5 allows 13.Bg5!² followed by Ne3. This Bg5 possibility is one of the few drawbacks of Black’s fianchetto system compared with having the bishop on e7, as in the latter scenario Black can generally meet Bg5 with ...Nd7 or ...Nh5. However, even taking into account the tempo spent on ...h6, the fianchetto still saves time for Black compared with lines involving ...Be7 followed by eventual ...Bf8 followed by ...g6 and ...Bg7.

The Fashionable 6.d3 GAME 44 Ilia Smirin – Alexander Grischuk New Delhi/Teheran (2.1) 2000   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d3 This innocent-looking but venomous little move has become extremely popular in recent years, as more and more super-GMs look for ways to avoid the ‘Marshall Defence’.

6...b5 7.Bb3 d6 8.c3 We start our investigation of the d2-d3 systems with this standard Spanish move.   Nowadays the two most popular options at high levels are 8.a4 and especially 8.a3. These moves will be examined in the next two games.   8...0-0 9.Nbd2 Na5 I recommend continuing in Chigorin style. The problem with the Breyeresque 9...Nb8?! 10.Re1 Nbd7 can be seen by comparing the position to the 10.d3 Breyer from Chapter 11. Here White has not had to spend time on h2-h3, and can instead play 11.Nf1 and Ng3 with virtually an extra tempo.   10.Bc2 c5 11.Re1 White’s set-up has a rich lineage, having been played by early world champions such as Steinitz and Capablanca and more recently being championed by Smirin and Nunn, with generally impressive results. Black has a few good plans, but I particularly like the one chosen by Grischuk in this game – which was also recommended by Nunn himself in

NCO.

11...Nc6! 11...Nd7 was proposed by Marin, intending to put the knight on b6, from where it guards d5 while leaving some space for Black’s bishops on the kingside. Although this is a valid way of playing, I would prefer to use the two tempos to improve my other pieces.   12.Nf1 Re8! A typical mistake for Black in this line is to continue in typical Closed Spanish style with 12...Qc7?! when White goes 13.Ne3! followed by a quick d3-d4. The advantage for White compared with the main lines of the Spanish is that he has avoided spending a tempo on h2-h3, and Black will constantly have to watch out for the possibility of the knight on e3 jumping to d5.   13.h3! One of the key points of the Nunn-Smirin system is that the f1-knight may go to either e3 or g3 depending on how Black arranges his pieces. For example, after 13...Bf8?! White plays 14.Bg5! followed by Ne3, clamping down on d5. 13.Bg5 makes no sense with the bishop still on e7. One of the many good ideas for Black is to go for Marin’s proposed manoeuvre with 13...Nd7, as the probable bishop exchange will leave Black with a comfortable game.

13...h6! Black wants to continue with the typical Spanish regrouping involving ...Bf8 and ...g6, after which he can gain space in the centre with ...d5. The text move also enables his bishop to come to e6 without fearing harassment from a knight on g5.   The immediate 13...d5 is playable but somewhat uncomfortable for Black after 14.exd5 Qxd5 15.Bb3 Qd6

16.Ng5, which is similar to the type of scenario we were trying to avoid in our Worrall investigations.   14.Ng3 Since Black is gearing up for ...d5, White’s best bet is to put the knight far away from the potentially annoying dpawn and prepare d3-d4. One instructive line given by Nunn in NCO is: 14.Ne3 Bf8 15.Nh2 (Also after 15.a4 Be6! 16.Bb3 Qd7! Black had

nothing to worry about in Almasi – Fressinet, Germany 2012) 15...d5! 16.Nhg4 Nxg4 17.hxg4 d4 18.Nf5

White’s position looks threatening, but after 18...g6! 19.Nxh6† Bxh6 20.Bxh6 g5! it is suddenly White who is in trouble due to his imprisoned bishop!

14...Be6! Black anticipates d3-d4 and frees the c8-square for his rook. 14...Bf8 15.d4 cxd4 16.cxd4 exd4 17.Nxd4 Nxd4 18.Qxd4 is given as the main line in NCO. This may also be okay but, after doing some analysis, I concluded that Black has to be accurate to avoid being worse. The text move is much less risky.   15.d4 This must be best; White has made all the obvious preparatory moves and there is no other sensible plan.  

15.Bb3 can be met by 15...Qd7! as in the Almasi – Fressinet game mentioned on the previous page.   15...cxd4 16.cxd4 exd4 The double exchange on d4 is a typical theme in the Spanish, which we will see more of later. For now, I will summarize the situation by saying that Black is opening the centre to activate his pieces. He is willing to pay the price of an isolated d-pawn, especially as the liberating ...d5-break is on the cards.   17.Nxd4 Here is another high-level example with considerable instructive value: 17.Ne2 Nb4! 18.Bb1 d3 19.Ned4 Bf8! 20.a3 Nc2 21.Nxc2 dxc2 22.Bxc2

22...Bc8! Unblocking the rook and preparing to deploy the bishop on b7, where it will be perfectly placed. 23.Nd4 Bb7 24.Nf5 Rc8 25.Ng3 d5 26.e5 Ne4 Black had taken over the initiative in Svidler – Leko, Monte Carlo 2007.   17...Nxd4 18.Qxd4 Rc8 In the Svidler – Leko game, Black’s decision to reroute his bishop via c8 and b7 was justified by the changes in the position and the additional time White spent rounding up the d-pawn. In the main game we see why Black moved his bishop to e6 in the first place: he wanted to free the c8-square for his rook.   19.Bb3 d5! This is an important theme in the structure that results from a double pawn exchange on d4.   20.e5 The database shows that Smirin has had this position three times, which tells you something about his experience and expertise in this system.

20...Nd7 20...Bc5!? could also be investigated, and was used by Ponomariov to defeat Smirin in 2013, but the text move is the most straightforward and reliable.   21.Nh5! Smirin has played this twice, although at Biel 2002 he immediately agreed a draw here with Tkachiev. The knight was doing nothing on g3, so White relocates it to f4.   An inferior alternative is: 21.Qd1 Nc5 (21...Nb6!? is also good according to my analysis) 22.Be3 Nxb3 23.axb3 d4! 24.Bxd4 Qd5

25.Ne4 Red8 26.Nd6 Bxd6 27.exd6 Qxd6 28.Bc3 Qxd1 29.Rexd1 Rxd1† 30.Rxd1 Bxb3 In Lutz – Kramnik, Brissago 2004, White managed to hold the endgame a pawn down, but it was not exactly a theoretical success for him.

  21...Bc5 21...Qb6!? 22.Be3 Qxd4 23.Bxd4 Nc5 has occurred in a couple of engine games in my database, and also looks perfectly fine for Black.   22.Qf4 If the queen goes to any other square then ...Qh4 will be highly problematic.

22...Qg5! This is the key idea that justifies Black’s earlier play. The resulting queenless position is at least equal for him.   23.Qxg5 hxg5 24.Bxg5 Bd4 25.Re2 25.Rad1!N 25...Bxb2 26.Nf4 This is the only remotely challenging idea I could find for White, but Black has more than one decent reply. 26...Nb6!? This is my personal preference. One of my training games (a nice way to examine an opening, which I wholeheartedly recommend to the reader!) continued:

27.Nxd5 27.Re2 might be better, but 27...Bc3 28.Nxd5 Nxd5 29.Bxd5 b4= is fine for Black. 27...Bxd5 28.Bxd5 Nxd5 29.Rxd5 f6 30.Bf4 Rc2

31.Rd6 a5 32.e6 a4! White has been playing ambitiously, but he is actually the one who has to be more careful to maintain the balance. The queenside pawns are quite dangerous, whereas the passed e-pawn is easier for Black to contain because his king is nearby.

25...Bxe5 26.Nf4 Bxf4 27.Bxf4 Nc5 28.Bc2 d4 29.Rd2 Red8 30.f3 d3 31.Bd1 Na4 At this point the players agreed to a draw. Black could have tried 31...Nb7!? with the idea to push the pawns on the queenside and go for either ...Na5-c4 or ...Nd6 according to circumstances. The position is close to equal, but there is a lot of play left and I prefer Black’s chances. ½–½   GAME 45 Eltaj Safarli – Gabriel Sargissian Khanty-Mansiysk (rapid) 2013   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d3 b5 7.Bb3 d6 8.a4 This continuation became popular around five or six years ago, and only stopped being played at the top level around the year 2013 once Anand’s incredible new concept of 8.a3! (as featured in the next game) took off. I feel fortunate to have discussed the present line quite extensively since 2011 with the Danish GM Sune Berg Hansen, one of the top experts from White’s point of view. It is important to know that analysing with an engine will not do you much good – the machine will tell you that Black is equal almost everywhere until it is rather too late, allowing a White player equipped with some basic ideas and plans to win game after game after game. The fact that White wins both this and the following illustrative game highlights the practical problems Black may face – although in both cases he obtained a fine position from the opening.   In the recent Baku World Cup, Bartel played 8.h3 three times against Sargissian, presumably to take the experienced Lopez player out of his comfort zone. In his third attempt, the Armenian GM finally came up with what I believe to be the best reaction: 8...Na5 9.Nc3 0-0 10.Be3 Bb7 11.Re1 In Bartel – Sargissian, Baku (1.9 – Armageddon) 2015, Black exchanged on b3. Although this is not bad at all, I prefer the more flexible continuation:

11...c6!?N My idea is to play ...Nd7, then exchange on b3 and play ...f5. Crucially, by leaving the bishop on b3 for the moment, we discourage White from playing d3-d4, as he will have to worry about ...exd4 followed by ...c5-c4 ideas to trap the bishop. (The immediate 11...Nd7 allows 12.Nd5; true, Black is fine after 12...Nxb3 followed by ...f5, but preventing the Nd5 move is more ambitious.)   8...Bd7! The other way to handle the position is with 8...b4 but several Spanish experts, including Gustafsson, Marin and Svidler, seem to concur that, if Black can maintain his b5-pawn while playing a natural developing move, he should not face any problems.

9.Be3!? This move became popular in 2011 after Sutovsky used it to get an edge against Adams. White plays a flexible developing move, waiting for Black to commit himself before deciding on a further course of action.

  9.c3 This looks like a more typical Spanish move, but it presents no problems for Black. An old game continued: 9...0-0 9...Na5!? 10.Bc2 c5 11.d4 Qb8 Navara – Caruana, Rhodes 2013, is a good alternative for Black, but I find it more natural to castle first. 10.Nbd2 Na5 11.Bc2 11.Ba2 c5 12.Re1 Qc7 is also fine for Black, who can consider ...c4 at some point to block out the bishop.

11...c5 12.Re1 Qc7 Black has made all the obvious developing moves, and now comes up with a nice way to put pressure on White’s queenside. 13.Nf1 Rfb8 14.Bg5 b4 Black was at least equal in Pilnik – Gligoric, Buenos Aires 1955.   9.Bd2!? This move should not be underestimated, as it comes with a sneaky positional threat. The correct response is: 9...b4! 9...0-0?! 10.axb5 axb5 11.Nc3 Na5 12.Nd5!² is what Black should avoid. The text move is perfectly timed; White would love to transfer the b1-knight to c4, but his last move obstructs that manoeuvre. Meanwhile Black’s plan involves ...Rb8 to defend b4, followed by ...Qc8 and ...Be6, and finally ...d5 (possibly with ...Rd8 included). 10.c3 0-0 11.h3 Rb8 12.Bc4 Qc8 13.Re1 Be6 We are following the stem game in which 9...b4! was played. Here White decided to advance in the centre before Black could carry out ...d5, but Black was well placed to handle that too. 14.Bxe6 Qxe6 15.d4 exd4 16.Nxd4 Nxd4 17.cxd4

17...d5! 18.exd5 18.e5 Nd7 followed by ...c5 is more than okay for Black. 18...Qd7 19.Bg5 In this equal position the players agreed a draw in Anand – Leko, Wijk aan Zee 2008. Black can either take on d5 immediately, or after 19...h6 and ...Rfe8.

9...0-0 9...Na5!? 10.Ba2 bxa4 11.Nc3 Rb8 12.Bb1 c5 13.Nxa4 Qc7 14.Bc1 c4 also proved fine in Petruzzelli – Leenders, corr. 2012, but it can hardly be bad for Black to castle and wait for h2-h3 before taking action on the queenside.   10.h3 10.axb5 axb5 11.Nc3 prevents the plan used by Black in the game but White loses some flexibility on the queenside, and after 11...Nb4!? Black was fine in Kamsky – Caruana, Zug 2013.

  10...Na5 11.Ba2

11...bxa4! This excellent move reveals the drawback of Sutovsky’s line. Normally such a capture would be considered antipositional, but Black can take advantage of the unprotected b-pawn to obtain a good game.   11...c5 is not so good, and after 12.axb5 axb5 13.Nc3 b4 14.Nd5 White obtained some positional pressure in Sutovsky – Adams, Porto Carras 2011.   12.Nc3 c5 Black has good prospects on the queenside. Apart from the obvious ideas of ...Rab8 and ...Qb6 to put pressure on b2, he can also consider a timely ...c4.   13.Bd2 Rb8 14.Bb1 This looks a bit awkward, but it is hard to suggest another way for White to recover his pawn.

14...Rxb2 14...c4!?N is a good alternative, intending to exchange on d3 and occupy the weakened b3-square.   15.Nxa4 Rb5 16.c4 16.Nc3 Rb8 is fine for Black, for instance: 17.Ne2 Nc6! 18.Rxa6 Ra8 19.Rxa8 Qxa8 Followed by ...Rb8, and Black is the first to take control of the open files on the queenside.   The text move looks strong, as White intends to play Nxc5 after the rook retreats, but Black has prepared a nice idea.

16...Rb4! 17.Bxb4 cxb4 Black has a pawn for the exchange, with two bishops plus a mobile pawn duo on the queenside.   18.Nb2 Qc7 19.d4!

This weakens the c4-pawn but frees White’s pieces.   In the event of 19.Bc2 Rb8 20.Ba4 Be6 Black is ready to play ...Nb7-c5 and ...a5, with perfect coordination.   19...exd4 20.Nxd4 Rb8 21.Bc2

21...Qb6 This is not bad, but 21...g6!N would have been more accurate, taking the f5-square under control and preparing to transfer the bishop to g7 at some point in the future. If White tries 22.Ba4 then 22...Bc8! is a good reply.   22.Ba4! Now if the bishop retreats to c8 to avoid the exchange, White will land a knight on f5. We have reached a critical position, as Black may lose his compensation for the exchange if he is not alert.   22...Bxa4?! As we saw in the note to move 19, when White puts his bishop on a4 it gives us a suitable moment to play 22...Nb7!, intending ...a5 and ...Nc5, preserving full compensation for the exchange.   23.Nxa4 Qc7 24.Nb2 Nb7!? Black gives up his extra pawn in order to improve the offside knight – a reasonable practical decision, although he still has a worse position and has to fight for a draw.

25.Rxa6 Nc5 26.Nb5 26.Rc6 Qb7 27.Na4 Ncd7! keeps Black in the game.   26...Qd8 27.Ra7 Nfxe4? 27...Ncxe4! would have given decent drawing chances, as 28.Nd3 can be met by 28...d5!.

28.Nd3!+– Attacking the b4-pawn and forcing simplifications. The game is almost over, and Safarli converts his advantage comfortably.   28...Nc3 29.Nxc3 bxc3 30.Nf4 Bf6 31.Nd5 h6 32.Qf3 c2 33.Qf5 Rb2 34.Rc1 Qb8 35.Nxf6† gxf6 36.Ra3 d5 37.Rg3† Kf8 38.cxd5 Qe5 39.Qc8†

1–0

Anand’s concept with 8.a3! 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d3 b5 7.Bb3 d6 8.a3! This innocent-looking move creates a retreat square for the bishop and prepares Nc3. This became popular in 2013 and has become one of the most important main lines in the Spanish, with players like Anand, Caruana, Svidler, Leko and Dominguez all using it as their main weapon every time they face 3...a6. This is clearly a top-level system, mostly suitable for positionally-oriented players, as it contains a lot of subtle nuances and demands careful handling from both players.

8...0-0 Black has several set-ups that offer close-to-equal chances, but most of them have some kind of drawback, making them somewhat uninspiring to play over the board. We will soon study a game between Firman and Mista from the 2014 Bundesliga, but I will present the game with a different move order, beginning with 8...Na5!, which I believe to be more accurate for Black. In this short section I will show the drawbacks of Black’s other main options.   9.Nc3 Na5 This was Mista’s choice.   It is useful to understand why 9...Be6 is not really ideal for Black either. White plays 10.Nd5 and if 10...Na5 11.Nxf6† Bxf6 12.Bxe6 fxe6 the star move is 13.b4!, when 13...Nc6 favours White. As Svidler points out, if the pawn was on c5 then Black would be fine – and he proved it by playing just such a position against Topalov at the 2015 World Cup, as shown in the notes to Firman – Mista below. As things stand though, after 14.c3 White had an annoying edge in Leko – Kasimdzhanov, Zug 2013. White’s ideas include Be3 followed by a3-a4, or Qc2 and c3-c4, while it is hard to see how Black can create any counterplay.   The text move is a solid, thematic choice in the Spanish, but unfortunately it gives White the chance to demonstrate one of the important ideas associated with the 8.a3 system.

10.Ba2 Be6 10...c5 would be fine if Black had time for ...Be6, but instead 11.b4! gives White a pleasant advantage. After the text move, the Firman – Mista game continued 11.Nd5 Nxd5 12.Bxd5 Bxd5 13.exd5 c5. This can be considered our target position, and we will discuss it below under my recommended move order.   However, White has a couple of other possibilities: both 11.Bxe6 fxe6 12.b4 Nc6 13.Bd2 and 11.b4 Bxa2 12.Rxa2 Nc6 13.Bg5 give him chances for a nagging edge. Neither of these lines is a theoretical disaster, but I would prefer not to present White with these options. As Svidler explains, Black generally has an easier time if he gets to play ...c5 before White has time for b2-b4. The solution is simple: we will gain time for ...c5 by postponing castling!   GAME 46 Nazar Firman – Aleksander Mista Germany 2014   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d3 b5 7.Bb3 d6 8.a3 Na5! My proposed move order was rather rare for a long time but, after Svidler made a strong case for it in his video lectures at Chess24 and Adams won with it at the 2014 London Classic, it experienced a big boost in popularity. With several recent outings for it at the Baku World Cup, this trend shows no sign of slowing.   9.Ba2 c5 As mentioned earlier, getting the pawn here before White can play b2-b4 is a definite achievement. Black is ready to play ...Be6 next, and the fact that he has yet to castle is of little consequence.

10.Nc3 10.Nh4?! is unlikely to attract many followers. After 10...0-0 11.Nf5 Bxf5 12.exf5 c4 13.dxc4 Nxc4 14.Nc3 d5µ Black was already dominating in Giri – Adams, London 2014.   After 10.b4 Nc6! the b4-pawn is en prise, so White doesn’t have time for the desired Nc3-d5.   10.Nbd2 is not dangerous either. Black castles, plays his knight back to c6 and his bishop to e6, and generally enjoys a fine game.   10...Be6! I must attribute this plan to Svidler in his excellent Chess24 video lectures. It is fitting that Svidler himself was the first to play it with Black, against Topalov at the recent World Cup in Baku.   11.Nd5 This was the only move mentioned by Svidler in his video lecture, and it keeps us on track to transpose to Firman – Mista. There are two other important options:   11.Bxe6 fxe6 This position was tested twice in the Baku World Cup, with Black obtaining fine play in both cases. 12.b4 This was Wei Yi’s attempted improvement. 12.Ne2 Nc6 13.Ng3 0-0 14.c3 would have been excellent for White with the pawn on c7, but as things stands now Black is fine. 14...Qd7 15.Qe2 Rad8 16.Rd1 d5= Topalov – Svidler, Baku (4.2) 2015. 12...Nc6 13.bxc5 dxc5 14.Bb2 14.a4 b4 15.Nb1 c4! is similar to the game.

14...0-0 15.a4 b4 16.Nb1 c4! 17.dxc4 17.Nxe5 Nxe5 18.Bxe5 c3!? is dangerous for White, as his queenside pieces are trapped. Black can play ...Nd7c5 and either consolidate on the queenside or prepare a breakthrough with a timely ...b3. 17...Qxd1 18.Rxd1 Nxe4 19.Re1 We have been following the tiebreak encounter of Wei Yi – Ding Liren, Baku (rapid – 4.3) 2015. At this point Ding Liren missed a strong idea:

19...Rf4!N 20.Bc1 Bc5 Threatening f2 and preparing to bring the other rook to d8. A sample line is: 21.Be3 Rd8 22.Nbd2 Nxd2 23.Nxd2 Rxd2! 24.Bxd2 Bxf2† 25.Kf1 Rf5 26.Rad1 Bxe1† 27.Kxe1 Nd4 Black’s active pieces are a source of concern for White, and having an extra pawn doesn’t hurt Black either.   Another important option is:

11.Bg5!?N 11...0-0 Suddenly we have transposed to a position that has been tested in several high-level games.

12.b4 12.Bxf6 Bxf6 13.Nd5 Nc6 looks nice for White, as he has won full control over the d5-square. However, as Svidler explains, White doesn’t have a clear way to take advantage of this pretty-looking knight, so Black can play around it with moves like ...g6, ...Bg7 (or ...Bg5), ...Bh6 and eventually ...Kh8 and ...f5. I decided to explore the position with some deep computer analysis and my conclusion was, unsurprisingly, that Svidler was correct. 12...Nc6 13.Nd5 This position has only occurred in one game so far, but I think it is more challenging for Black than the line in the note above. I recommend:

13...Rc8!N I find this the simplest continuation to understand, although 13...Bxd5!? 14.exd5 Nd4 led to an interesting

struggle and an impressive win for Black in Dominguez – Aronian, Wijk aan Zee 2014, so this could certainly be considered as well. The idea of my suggested move is to force White to take action against the threat of ...Bxd5 followed by ...Nxd5, winning a pawn. 14.Nxf6† The computer suggests 14.Bxf6 Bxf6, when we reach a similar type of position to that discussed in the earlier note with 12.Bxf6. 14...Bxf6 15.Bxe6 fxe6 16.Bxf6 Qxf6 17.bxc5 dxc5 An instructive situation has arisen: Black’s pawn structure looks a bit ugly, but his pawns are not weak and he has the strong plan of ...Ne7-g6 to put White’s kingside under pressure.

11...Nxd5 12.Bxd5 I find this to be the most logical continuation, and it was the choice of Anand against Adams in a similar position, in a game to which we will soon transpose.   12.exd5 This is the only move mentioned by Svidler in his online lecture, but it is probably the weaker recapture as White risks being stuck with a bad bishop on a2. 12...Bd7 12...Bg4!? is a move which I dare to say may be even stronger than what Svidler gives. The main strategic idea is to exchange on f3 and then play ...Bg5, leaving Black with the more useful minor piece. Still, it is worth checking Svidler’s line and instructive comments. 13.b4 Nb7 Svidler points out that Black has a potentially dangerous plan with ...Qc7, ...Rae8 and ...f5, so White may want to try a forcing continuation: 14.bxc5 Nxc5 15.d4 exd4 16.Nxd4 0-0 17.Nc6 Bxc6 18.dxc6 And now the last detail is:

18...Ne4! 18...Bf6 19.Rb1 Ne4 allows 20.Bd2 when the c6-pawn may cause Black some problems. 19.Bd2 d5   Black is fine according to Svidler. I checked this carefully and I found no reason to disagree with the Russian superGM.   12...Bxd5 13.exd5 0-0 Finally we have transposed to the Firman – Mista game that we are supposed to be studying! Black has a simple plan with ...f5 on the cards next.

14.b4 Nb7 Although the knight is badly placed on b7, it can look forward to a brighter future on f7. At the risk of overloading

you with transpositions, we have now converged with Anand – Adams, Baden-Baden 2013 – a landmark game, as it was the one in which Anand introduced the entire system with 8.a3 followed by Nc3.   15.Re1 A natural move, but hardly an improvement over Anand’s play.   The aforementioned game continued 15.c3, when I like 15...f5 for Black. (Adams played 15...Qc7 followed by ...f5, which was not bad either, but it allowed White to exchange his inactive bishop with Bg5) An illustrative line is: 16.Re1 h6!  

This move prevents the Bg5 idea which was played by Anand. Black will continue with ...Bf6 and ...Qd7, enabling the knight to go to d8 and f7; this manoeuvre can be prefaced by ...Rac8 so that bxc5 will not be a problem. Black may also consider putting the queen on f7, since the knight can no longer go to g5.   15...Qc7 15...h6!?N merits serious consideration here as well.

16.Bb2?! It is not easy to decide where this bishop should go. It looks active on b2, but in the game it proves to be tactically exposed, so White probably should have preferred the d2-square.   16.Bg5!? also makes some sense. A logical sequence is 16...f6 17.Bd2 f5 18.Bg5 Bd8!?. Now 19.Bxd8 Nxd8 is fine for Black, and if White tries something else then Black can improve his position with moves like ...Rac8 and/or ...c4 before forcing the bishop exchange with ...h6.   16...f5 17.Qe2 Bf6 18.bxc5? This impatient move was probably based on a miscalculation. However, the computer-suggested 18.c4 is also not great for White, especially after the smart reply: 18...a5!„   18...Nxc5 White is strategically in trouble; the d5-pawn is weak, and Black has two significant threats in ...Qf7 and ...Na4.

19.d4 This must have been White’s idea, but it is tactically refuted.   19...e4! 20.dxc5 Bxb2 21.Ng5 Bxa1 22.Ne6 Qe7 23.Rxa1 Rfc8 White has no real compensation for his exchange; the knight looks nice on e6, but it does not cause any real trouble.   24.Re1 24.c6? Rxc6! may well be the detail that White missed in his earlier calculations.   24...dxc5 25.f3 Ra7 26.fxe4 fxe4 27.Qxe4 Rd7 Black has played well, and he wraps up the win with a few more accurate moves.

28.Rf1 b4 29.Rf3 Qd6 30.Rf5 bxa3 31.c4 a2 32.Rf1 Now Black found a nice way to end the game:   32...Rb8 33.Qc2 Rf7! White threw in the towel; the threat is ...Qe5, and 34.Re1 is met anyway by 34...Qe5! 35.Rxe5 a1=Q† with mate in two more moves. 0–1   Conclusion   Against 5.d3 I recommend 5...d6, intending a kingside fianchetto. Then 6.c4!? leads to an unusual structure for the Spanish, but Black should get a pretty comfortable version of a King’s Indian set-up. The main continuation is 6.c3 g6 7.Nbd2 Bg7 8.0-0 (8.Nf1 0-0 9.Bg5 is met by Caruana’s natural and strong 9...d5!) 8...0-0 9.Re1. In this position I agree with Marin that the flexible 9...Re8! is Black’s best, waiting for 10.Nf1 before playing 10...b5. In the event of 11.Bc2, Black goes for the Breyer regrouping with ...Nb8-d7 and ...Bb7, while the sneaky 11.Bb3 should be met by 11...h6!, preventing any Bg5 ideas before continuing with ...Na5 and ...c5, when Black obtains an excellent Closed Spanish set-up.   The majority of the chapter was devoted to 5.0-0 Be7 6.d3, when Black no longer has the option of a kingside fianchetto. Instead we play 6...b5 7.Bb3 d6, when White has three main options.   8.c3 is an older move which has been used successfully by Nunn and Smirin, but Black is doing fine after my recommended plan involving ...Nc6, ...Re8, ...h6 and ...Be6, intending to meet the d3-d4 advance with multiple exchanges and the active ...d5.   I recommend meeting 8.a4 with 8...Bd7, when neither 9.Bd2 nor 9.c3 seems especially threatening for Black. Sutovsky’s 9.Be3!? is interesting, but Black should be fine after my suggested 9...Na5 10.Ba2 bxa4!, followed by ...00, ...c5 and ...Rb8.   8.a3 is a major battlefield for the world’s top players, but 8...Na5! 9.Ba2 c5! is a clever move order which avoids most of the problems Black has encountered in other lines. A key point to remember is that 10.Nc3 Be6 11.Bg5 0-0 12.Bxf6 Bxf6 13.Nd5 is met simply with 13...Nc6, when Black plays around the strong d5-knight with a view to playing ...f5 in the future. I believe Black is fine in all variations, and in many lines he has decent prospects to take over the initiative.

  1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8 page 287 15.Nh2 Game 47, page 288 (15.a4 page 288) 15.b4! Ncd7! 16.a4 Game 48, page 291 16.Bb3 Game 49, page 292 16.d4 Game 50, page 294   Introduction to 10.d4 page 298 10...Nbd7 11.Nh4 Game 51, page 299 (11.Nbd2 page 299) 11.c4 c6 12.c5 Game 52, page 302 (12.Qc2 page 303; 12.Bg5 page 303; 12.cxb5 page 303; 12.Nc3 page 303) 12.a3 Game 53, page 305

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following

chapter.   Black is to move unless otherwise indicated.  

How should Black meet 15.a4? (page 288)  

How did Black meet Tal’s aggression? (page 293/4)

Fischer has played this with White.

How would you proceed? (page 301)  

Once again, you must play against Fischer! (page 303)  

Should Black gobble the c3-pawn? (page 295)  

Show that Black’s activity is more

important than his doubled pawns. (page 304)  

A critical theoretical position. (page 295) How did Ivanchuk establish full control? (page 306)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 I have already mentioned that the c6-knight tends to be Black’s problem piece in the Closed Spanish. With the Breyer System we invest two tempos to reroute the knight to the more stable and flexible d7-square.

10.d3 Not at all a bad move! White avoids an immediate clash and defends the e4-pawn securely, with the idea of completing development by transferring the b1-knight to g3. Once that has been accomplished, he can decide whether to play in the centre with d3-d4, on the kingside with Nh2, Qf3 and Ng4, or on the queenside with b2-b4 and a2-a4.   10.d4 is the main move of course. Coverage of this begins in the current chapter, and continues in the final two chapters.   10...Nbd7 Black continues with the Breyer regrouping. The knight is heading for c5 in order to drive the Spanish bishop to a worse position.   Another decent continuation is: 10...c5 11.Nbd2 (If 11.a4 then 11...Bb7, as proposed by Eljanov, with the idea that after 12.axb5 axb5 13.Rxa8 Bxa8 Black will obtain a harmonious position with ...Bc6, ...Nbd7 and ...Qb6-b7.) 11...h6!? Black intends ...Nc6, ...Bd7, ...Re8 and ...Bf8. This is another acceptable way to interpret the position, but I prefer the clarity of the main line where Black kicks the b3-bishop off the dangerous diagonal which it currently occupies.   11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 Black’s next step will be to drop his bishop back to f8, at which point everything will be ready for the liberating ...d5 break.

14.Ng3 Bf8 After several natural moves, we reach our first major branching point since 10.d3 was played. White’s most significant options are as follows:   15.Nh2 will be covered in our first illustrative game of this chapter, which is one of Portisch’s many positional masterpieces. (15.a4 is well met by 15...Nxa4! followed by a quick ...d5, as you will see in the notes.) 15...d5 is a logical reply, playing in the centre. Portisch gives a perfect demonstration of how to restrain White’s kingside ambitions while improving Black’s position in the centre.   15.b4! was introduced by Pachman in 1957, and was soon adopted by several top Soviet players including Tolush, Vasiukov and Petrosian. White’s idea is easy to understand: the c5-knight has forced the bishop to vacate the ideal b3square, so he chases it away with gain of tempo. That being said, Black has also gained something: the ...a5 lever is an important source of counterplay. 15...Ncd7! is the best retreat, when we will cover three main moves:   16.a4 is a logical follow-up to White’s last move, hoping to achieve something on the queenside. This will be covered in Game 48.   16.Bb3 was played several times by Tal, but in Game 49 we will see the attacking wizard being convincingly defeated by Spassky.   Finally, 16.d4 is White’s most popular move, and it will be examined in Game 50.   GAME 47 Walter Browne – Lajos Portisch Wijk aan Zee 1972   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8

15.Nh2 This move signifies that White intends to build up a kingside attack while keeping the centre closed.   15.a4 is convincingly met by: 15...Nxa4! 16.Bxa4 bxa4 17.Qxa4 (17.Rxa4 can be dealt with in the same way) 17...d5! Black has the slightly worse structure, but this is outweighed by the excellent b7-bishop, which comes to life after this thematic break. A model game continued: 18.Bg5 h6 19.Qb3! This little trick is White’s only challenging idea, but Black can liquidate to a good position with:

19...dxe4! 20.Bxf6 Qxf6 21.Nxe5 Qxe5 22.Qxb7 Reb8! 23.Qc6 Qb5! This all happened in Barta – Kostal, corr. 2003. The last move ensures that White will lose his b2-pawn after he takes on e4, thus leaving Black with a potentially strong passed a-pawn. The game is probably still dynamically balanced, but the a-pawn supported by a rook and the better minor piece gives Black chances to play for the win, and in the aforementioned correspondence game he was successful.

  15...d5 15...g6 is a reasonable alternative which is advocated by Kaufman, but I like the directness of Portisch’s play, so I won’t overload the reader by analysing a secondary option in a non-critical line.   16.Qf3 Ne6 17.Nf5 Kh8! This excellent prophylactic move takes the sting out of Qg3, which can now be answered by ...Nh5 as there is no check on h6. Black’s argument is that if nothing serious happens in his kingside, he will have excellent chances to take control in the centre and on the queenside.

18.h4 This is a logical attempt to bring more wood into the attack, but Black is ready for everything. Other moves are not too worrying either, for instance:   18.exd5 can be met by the Marshall-like 18...Nxd5!N 19.Rxe5 Qd7! when Black’s compensation is more than enough for a mere pawn. His main threat is ...Nc5 and the logical 20.d4 is met by:

20...f6! 21.Re4 (forced, as 21.Re1?? Nxd4! and 21.Re2? Ndf4! are both disastrous) 21...c5 (21...g6 also looks decent) Black has a strong initiative.   18.Ng4 is the other logical choice but after 18...Nxg4 19.Qxg4 (19.hxg4 c5„) 19...Qf6!? Black intends ...g6, and is more than okay. A good example continued:

20.Qg3 dxe4 21.dxe4 c5 22.Be3 c4 23.Rad1 Rad8 24.Rxd8 Rxd8 25.Bc1 g6 26.Ne3 Black had a nice positional edge in Tringov – Geller, Amsterdam 1970, which could have best been emphasized with 26...h5!N.   18...c5 19.Bg5 Nxg5 20.hxg5 Ng8 21.Qg3 g6 22.Ne3 d4

23.Neg4? Browne has the idea of exchanging the g5-pawn for the e5-one, but the Hungarian legend proves that there is a flaw in his idea.   23.Nef1 looks passive but is objectively best. Still, Black obviously has no problems and can continue to strengthen his position with 23...Rc8 or 23...Bd6.   23...Qxg5 24.Nf3 Qf4! 25.Ngxe5 Qxg3 26.fxg3 dxc3 27.bxc3 Nh6µ Black has a pair of powerful bishops and the better pawn structure. Browne fights hard, but Portisch stays in control and continues to improve his position.

28.Rad1 Rad8 29.Bb3 Kg7 White’s position is already on the verge of collapse, as he needs to find something against ...f6.

  30.g4!? Browne prepares to meet 30...f6 with 31.g5!, when he would still be much worse but would at least have a chance to complicate the game.   30...g5! Portisch is merciless! This move prevents g4-g5 and fixes the weakness on g4. The weakness on f5 is of no consequence, as White’s pieces are unable to occupy it.

31.Bd5 White is losing material anyway, so this is as good a try as any.   31...Bxd5 32.exd5 f6 33.c4!? Browne takes the smart practical decision to give up a piece and try to cause trouble by getting his remaining knight to e6.   33...fxe5 34.Nxg5 Kh8 34...Nxg4! 35.Ne6† Rxe6 36.dxe6 Kf6 would have won more easily.   35.Ne6 Rd7 36.Rxe5 Nxg4 37.Re4 Nh6 38.Rf1 Bd6 Portisch wraps up the endgame without any major problems, although Browne defends tenaciously and gets pretty close to his dream of drawing with a lone king against an h-pawn assisted by the wrong bishop.

39.Rf6 Ng8 40.Rf5 bxc4 41.dxc4 Nh6 42.Rf6 Nf7 43.Ng5 Rxe4 44.Nxe4 Kg7 45.Re6 a5 46.Kf1 h6 47.Kg1 Rd8 48.a4 Rd7 49.g4 Be5 50.Kg2 Bd6 51.Kh1 Kf8 52.Kg2 Re7 53.Kf1 Bf4 54.Nxc5 Rc7 55.Nb3 Rxc4 56.Nxa5 Rxa4 57.Nc6 Bd6 58.Rg6 Rc4 59.Ke2 Rc5 60.Nd4 Rxd5 61.Nf5 Bf4 62.Ra6 Bg5 63.Kf3 Re5 64.Ra8† Re8 65.Ra6 Kg8 66.Kg3 Rb8 67.Rg6† Kh8 68.Ra6 Rb3† 69.Kg2 Rb7 70.Kg3 Kh7 71.Re6 Nh8 72.Ra6 Ng6 73.Ra5 Nf8 74.Ra4 Nd7 75.Nd4 Ne5 76.Nf5 Rc7 77.Rb4 Rc4 78.Rb7† Kg6 79.Rb6† Rc6 80.Rxc6† Nxc6 81.Nd6 Kf6 82.Ne4† Ke5 83.Nf2 Kd4 84.Kf3 Ne5† 85.Kg3 Be3 86.Nh3 Ke4 87.Kh4 Nf3† 88.Kh5 Ng5 Browne decided it was not worth testing his opponent’s technique with bishop and knight against a lone king. 0–1     GAME 48 Thomas Rattinger – Olli Ylonen Correspondence 2013   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Re8 13.Ng3 Nc5 14.Bc2 Bf8 15.b4! Ncd7! This is the usual move and is widely regarded as Black’s best, but it is worth considering why this retreat, which appears to lose time, should be preferable to retreating the knight to e6.   The following game shows the potential troubles Black can get into: 15...Ne6 16.d4 g6 (16...a5!? is safer) 17.Bd2 Nd7 18.Bb3 Qf6 19.d5 Nf4  

20.c4! Nb6 21.Rc1 bxc4 22.Bxc4 Nxc4 23.Rxc4 Rac8 24.Qc2 Be7 25.Bxf4! exf4 26.Ne2± Black was already positionally busted in Short – Mamedyarov, Gibraltar 2012, as the e2-knight was heading for c6. The conclusion is that the knight on e6 merely gives White a target for d4-d5, as ...Nf4 does not really achieve anything.   16.a4 This move is directed against the ...a5 break. However, with the b5-pawn already defended, Black can choose between two other pawn breaks.

16...c5! 16...d5!? was actually the move I initially wanted to recommend, but after 17.exd5! (otherwise Black plays ...c5 next, with a good game) I was not completely satisfied with either recapture:   a) 17...Bxd5 runs into the annoying 18.Bg5!N 18...h6 (18...c5 19.bxc5 Bxc5 20.Nxe5!) 19.Bh4! when 19...g5 20.Nxg5!

sees White sacrifice a piece for excellent compensation – and if ...g5 leads to a good game for White, then it will not be easy to find a comfortable way to continue for Black.   b) 17...Nxd5 18.Bd2 N5f6 is the other option, when 19.Rb1! h6 was mysteriously agreed drawn in Minelga – Korosec, corr. 2008. However, White can initiate interesting complications with 20.c4!N when, after analysing this position for a while, I couldn’t find a clear-cut solution.   17.bxc5 Nxc5 18.axb5 axb5 Initially the engines give White a tiny edge here, but I don’t think he even has that.

19.Rb1 White has also tried: 19.Bg5 h6 20.Bxf6 (20.Bh4 can be met by either 20...Ncd7 intending ...Qc7 if Black wants to play safe, or 20...Rxa1 21.Qxa1 g5 if he wants to grab material) 20...Qxf6 21.d4 Ne6 Black was fine in Makovsky – Romanov, email 2010. For example, after 22.Qd3N 22...Bc6 followed by ...Ng5 I don’t see a convincing way forward for White.   19...Bc6 20.d4 exd4 21.Nxd4 Bxe4 Black has to take with the bishop because it was attacked, but losing the bishop pair is not such a problem here.   22.Nxe4 Ncxe4 23.Nxb5 Rc8= Black is well coordinated and his knights are no worse than White’s bishops.

24.Bd3 d5 25.Bf4 Bc5 26.Re2 g6 27.Bh2 Re6 As with many equal positions, you could play this on for a long time over the board, but the correspondence game soon peters out to a draw.   28.Qf1 Rb6 29.Be5 Qe8 30.Bd4 Bxd4 31.cxd4 Qf8 32.Qe1 Qh6 ½–½   GAME 49 Mikhail Tal – Boris Spassky Tbilisi (11) 1965   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.h3 Bb7 9.d3 d6 10.c3 Nb8 11.Nbd2 Nbd7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8 15.b4 Ncd7 16.Bb3 This was Tal’s pet line in this system, but both Smejkal and Spassky showed an excellent way for Black to defuse it.

16...a5! 16...h6 is offered by Kaufman but there is no need to weaken our kingside just yet, as we will see in the next note.   17.a3 17.Ng5 is met by 17...d5! 18.exd5 (or 18.bxa5?! h6) 18...a4! 19.Bc2 h6 followed by taking back on d5 with at least equal chances.   17...axb4 For some reason, in a later game Tukmakov went for 17...c5?! instead, but after 18.Ng5! he had some problems and Tal went on to beat him at Sochi 1970.   18.cxb4 h6! Having clarified the queenside structure in his favour, Black prevents Ng5 and gets ready to play either ...c5 or ...d5.   19.Nf5 Tal ignores the centre and plays for an attack. I should put this into context by mentioning that he was two points down in the match with two games remaining, so he had to play for a win at all costs. Although White’s subsequent attack is rather reckless, the way Spassky refutes it is highly instructive.   Later Tal switched to 19.d4, when the critical line continues: 19...c5! 20.bxc5 dxc5 21.dxe5 Nxe5 22.Nxe5 Rxe5 23.Bb2 Re8 Once again Black has nothing to fear, although it is worth looking a bit deeper to see where his play can be improved. 24.Qf3 c4 25.Bc2 Interestingly, Smejkal reached this position twice in 1971. In both games he played 25...Nd7, first losing to Tal and then drawing with Liberzon. However, Black has at least two ways to get an advantage:

25...Qd2N 26.Re2 Qg5 is one promising continuation; 25...Re6!?N intending ...Rea6 is another.

19...d5! 20.N3h4 c5! 21.Re3 c4! 22.Rg3 Kh7µ It must have required a lot of courage to let Tal amass so many pieces on the kingside, but Spassky correctly judged that he was not in any real danger.   23.Bc2 d4 24.Qf3 Black has several strong moves here, but I especially like Spassky’s choice for the way it combines active and defensive functions.

24...Ra6! 25.Nxh6 Something had to be done, otherwise Black could have played ...Qa8 next, with huge pressure.   25...gxh6 26.Nf5 Qa8 27.Nxh6 Bxh6 28.Qf5† Kh8 29.Bxh6 Rg8 Black is winning. Tal tries his best to conjure something, but Spassky does everything right.   30.Bg5 Qe8 31.f4 Nh7 32.Bh4 Rxg3 33.Bxg3 Rf6 34.Qg4 Rg6 35.Qh4 exf4 36.Qxf4 f6 37.Bf2 cxd3 38.Bxd3 Ne5 39.Bf1 Ng5 Black’s pieces have beautifully come into play, and the end is near.   40.Kh1 Nxe4 41.Rc1 Nxf2† 0–1   GAME 50 Evgeni Vasiukov – Yuri Averbakh Moscow 1964   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8 15.b4 Ncd7 16.d4 This central advance has been White’s most popular choice in the position.

16...a5! 16...g6 is the most common reply, but chipping away at White’s queenside strikes me as the most natural continuation, as well as one that takes less effort to learn.   17.a3 White may also allow ...axb4 and reply with the inevitable cxb4, and this is what happens in most games. The main question is how he should look to arrange his pieces in the resulting position.   17.bxa5 Rxa5 fails to put Black under any pressure. The most reasonable example I found continued: 18.Bd2 Ra3 (18...Qa8 19.a4 bxa4 20.Bxa4 Bc6 is also equal) 19.Qb1 c6 20.Bc1 In Westerinen – Razuvaev, Solingen 1976, Black retreated the rook and soon agreed a draw in an equal position. A far more interesting continuation, and possibly just a stronger move overall, would have been:

20...Rxc3!N 21.Bb2 Now Black has a choice: both 21...exd4 and 21...Rxf3 22.gxf3 exd4 23.Bxd4 Ne5 offer him full compensation for the exchange.   17.Bd2 This is the theoretically approved continuation, and it requires accurate handling. 17...axb4 18.cxb4

18...Ra3!N My analysis indicates that this is best, but it is worth pausing to see what other sources recommend.   a) 18...exd4 19.Nxd4 d5 is ECO’s suggestion, based on Tal – Timman, Niksic 1983, but after 20.Nxb5!N I found no easy way for Black to equalize completely. b) 18...d5 is recommended by Kaufman, but he does not consider the critical line: 19.dxe5! Nxe5 20.Nxe5 Rxe5 21.Bc3 Re8 22.e5 d4! This is necessary to avoid a serious disadvantage; still, after 23.Bxd4 Bxb4 24.Re3 White had some initiative in Berggren – Edwards, corr. 2012. Black may be able to equalize if he defends accurately, but this is not the kind of position I would want to get in a practical game. Let’s return to 18...Ra3!:

Black’s last move stops the idea of Bc3 followed by dxe5, and at the same time plans to improve the coordination of his pieces with ...Qa8, which may be followed by ...g6 at some point. Whenever the rook gets attacked it can retreat to a6. I don’t see any dangerous ideas for White, so I will just include the main line of my analysis. 19.dxe5 dxe5! Often Black is advised to exchange a pair of knights on e5, but right now I like my knight on d7, as it can support the ...c5-break, especially when White’s bishop goes to b3 or d3. 20.Qb1 Qa8 21.Bc1  

21...Ra7! This is slightly better than 21...Ra6, when White can annoy us a bit with 22.Bd3. 22.Bd3 Another nice line is 22.Be3 c5! 23.bxc5 Nxc5 24.Bxc5 Bxc5 25.Qxb5 Ra5 when Black’s compensation for the

pawn is clear: he has pressure against the a2-pawn and his mighty bishops practically cut the chessboard in half. 22...c5! 23.bxc5 b4! The c5-pawn isn’t going anywhere, so we should make it harder for White to take the b-pawn.

24.Bb2 After 24.Qxb4 Nxc5 the pawns on a2 and e4 are under heavy fire, and something is going to fall. 24.Be3 is possible, although Black is fine after either recapture on c5. 24...Qb8! Black secures the e5-pawn and is finally ready to take on c5, with at least equal chances. This variation contains a lot of subtleties, and I would advise the reader to try and memorize it, or at least play through it on a board to take in some of the nuances.

17...axb4 18.cxb4 c5 Kaufman assesses the position as equal, which seems about right. However, considering that Averbakh went on to

lose as Black, it is worth looking more closely at how to handle positions of this type. Later we will see Black struggling with a weak queenside majority against White’s stronger kingside majority, which is quite a common scenario in the Spanish, so it is worth being familiar with the practical dangers we might face.   19.bxc5 dxc5 20.dxe5 Nxe5 21.Nxe5 Rxe5 22.Bb2 Qxd1 Averbakh tries to keep it simple, but one can make the argument for avoiding the structure that occurs in the game – although objectively Black is still not worse.   An interesting alternative is: 22...Re6!?N 23.e5

23...Qd5! 24.f3 Rae8! My idea is to sacrifice the exchange for a pawn with ...Rxe5!, let’s say after 25.Bf5 or 25.Re2. This is a suitable time for it, as Black’s queenside pawns supported by the bishop pair should offer full compensation.   23.Rexd1 Re6?! Averbakh doesn’t want to see his pawn structure damaged, but the text move affects Black’s coordination and leads to more significant problems. It would have been much better to play 23...Re8 (or even 23...Re7!? 24.Nf5 Ree8) for reasons explained in the next note. Black has no reason to fear Bxf6 in either case, as the loss of White’s strong bishop is more relevant than the mostly cosmetic damage to Black’s kingside.

24.e5 Ne8 With the rook on e7 or e6, Black could have comfortably played 24...Nd5, but here it would have allowed 25.Be4 Rd8 26.Rd2. The machine points out that Black can still play this position and not be too much worse, but it is not nice having a piece pinned along two different lines.   25.a4! White aims for the favourable pawn structure discussed earlier.   25...bxa4 25...b4 is met by 26.Bb3 and, although I imagine that someone like Petrosian would play 26...Nc7, White can politely refuse the offer and play 27.Nf5! with some initiative.   26.Rxa4 Rxa4 27.Bxa4 Nc7 This is the structure I was talking about in the comments to Black’s 18th move. Similar situations (sometimes with a pawn on the b- instead of the c-file) can occur quite often in the Breyer and in the Spanish complex in general, as a result of the thematic pawn breaks and exchanges in the centre. White tends to get the better of it, although it all depends on the relative activity of the pieces of course. In the present case Black is within the drawing margin, but defending such a prospectless position is not the most joyful thing in chess, and often leads to further mistakes.

28.Rd7 Bc6?! The cool-headed 28...Re7 29.Rd8 Ne6 30.Rb8 g6!, followed by unpinning the bishop with ...Kg7, would have been acceptable for Black.   29.Bxc6 Rxc6 30.f4 Rb6 31.Rxc7 Rxb2 32.Rc8 Rb4 33.f5 Just six moves after I discussed the dangers of this structure, Black’s position has become critical. I won’t attempt to prove whether the position is winning or drawing at this stage – the important thing is simply to avoid such situations where at all possible.

33...g6 34.f6 Rf4 35.Nf1! Re4 36.Nh2 Rxe5 36...h5 might have offered more hope, but after 37.Nf3 White remains much better.  

37.Ng4 Rh5 38.Re8 Rh4 38...h6 would have allowed 39.Ne5 followed by Nd7.   39.g3! Rxh3 40.Ne5 Rxg3† 41.Kf2 Rg5 42.Nd7 Now it’s all over, as White wins the bishop while keeping the f6-pawn.   42...Rf5† 43.Ke3 h5 44.Rxf8† Kh7 45.Rxf7† Kh6 46.Re7 Averbakh resigned as the pawn promotes. 1–0

Introduction to 10.d4 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7

The remainder of the chapter will deal with a number of sidelines from this important tabiya.   11.Nh4 was played several times by Fischer. Game 51 is an impressive display of Spassky’s understanding of the Breyer, as he effortlessly equalizes against Robert Byrne.   11.c4 will be featured in two games, with Reshevsky and Ivanchuk as our role models. The former game is just as relevant, despite having been played more than forty years before the latter!   Before we examine those two moves, let me mention an important point about the main line:   11.Nbd2 By the way, 11.Bc2 Bb7 will most probably transpose to the next chapter. For instance, after 12.a4 Re8 13.Bd3 Bf8 White can grab the b5-pawn but he will lose the one on e4, which isn’t a great trade for him. Therefore he may as well play 14.Nbd2, reaching a variation covered in the next chapter.   Please look closely at the following position, as I cannot emphasize the correct move order strongly enough.

11...Bb7! That’s it – just develop the last of the minor pieces. 11...Re8?? is a move that appears in more than thirty games in my database. Surprisingly, the latest example is also the one that features the highest-rated player to have been punished. In Zatonskih – Paehtz, Berlin 2011, White of course played 12.Bxf7† and Black was already lost, as 12...Kxf7 is met by 13.Ng5† followed by Ne6xd8. That being said, in the famous encounter J. Polgar – Spassky, Budapest (10) 1993 – where Spassky won a great game by sacrificing a piece for two central pawns (this critical and still topical line is examined later in Game 64) – the exWorld Champion also played 11...Re8??, but Polgar didn’t notice the blunder and continued with 12.Nf1??. Polgar gives the full story, as well as expert analysis of their many Breyer encounters, in From GM to Top Ten.   After the text move the main line is 12.Bc2, protecting the e4-pawn in order to continue with the thematic Spanish manoeuvre of Nf1-g3. In some variations the bishop will go to d3 and attack the b5-pawn (a typical target for White in the Spanish) with the help of a2-a4. Still, the bishop is actively placed on the a2-g8 diagonal, and some players have tried to maintain it there. We will look at two such attempts, 12.a3 and 12.a4, in the comments to the following game.   GAME 51 Robert Byrne – Boris Spassky Moscow 1971   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nh4 A curious-looking idea, but any move to have been used repeatedly by Fischer deserves respect. White’s idea is to meet 11...Bb7 with 12.Nf5 and Bg5, and only then Nd2, when his pieces will exert strong pressure on Black’s kingside.   11.Nbd2 Bb7 One of the nice features of the Breyer is that, if White wants to carry out the thematic Nf1-g3 manoeuvre, he will

have to retreat his bishop from its excellent outpost on b3 to the more passive c2-square. Let’s see what happens if White tries to keep his bishop on its preferred diagonal. 12.a3 12.a4 can be met by the typical response of 12...c5, intending 13.d5 c4!. I will talk more about this structure in the final chapter. Of course Black can also play a non-committal move like 12...Re8, intending to transpose to one of the main variations, but this gives White the option of the ‘Zaitsev drawing manoeuvre’ with 13.Ng5. 12...c5 13.Ba2 13.d5 c4! gives Black the usual counterplay that comes when White has weakened his queenside by moving the apawn. 13.Bc2 seems inconsistent with White’s previous move, but has been the most popular choice. Still, Black can proceed with the same idea as shown below.

13...cxd4!? Black could also have waited a bit by moving a rook to e8 or c8 first. 14.cxd4 exd4 15.Nxd4 Re8 This position, from Nevostrujev – Jakovenko, Olginka 2011, resembles the situation in our main game below. Generally Black can be happy in these positions when the knight is on d2 rather than c3.

11...exd4! I like this move, as it gives Black a good game with chances to outplay his opponent without taking excessive risks. Studying this line also brings wider benefits, as it leads to a structure which every Ruy Lopez player ought to understand and be ready to play.   11...Re8!? 12.Nf5 Bf8 is recommended by Kaufman and leads to more complicated positions.   11...Nxe4!? is another obvious move, but White’s idea is that 12.Nf5 Ndf6 13.Qf3 gives him good play for the pawn he has temporarily sacrificed. Old theory proceeds with 13...Bb7, but Eljanov came up with 13...Bxf5 14.Qxf5 d5!N which essentially leads to a loss of two pawns for Black! His analysis continues 15.dxe5 Bc5 16.Be3 Bxe3 17.Rxe3 g6 18.Qf3 Nh5 19.Bxd5 Qxd5 20.Qxe4 Rad8 when it is now Black who has decent compensation for the missing pawn.   12.cxd4 Nb6! Normally it would be risky to give White such a strong central pawn duo, but here it is justified by the odd placement of the knight on h4. Black intends to play ...c5 next (almost no matter what!) with equal play.

13.Nd2 13.Nf5?! Bxf5 14.exf5 c5 is positionally risky for White; the machine claims he is still not worse, but in practice he has scored miserably from this position.   13.Nf3 Fischer played this a couple of times. The move has the clear-cut idea of improving the knight, so it fits Fischer’s style perfectly, but Black is not worse in the middlegame that ensues. 13...c5 14.Bf4 Bb7

15.dxc5 This move also matches Fischer’s direct style of play. 15.d5N should be considered, but a good reply is 15...Nfd7! intending ...f5. 15...dxc5 16.Qxd8 Bxd8 17.Bd6 Re8

17...Nxe4!?N 18.Bxf8 Kxf8 gives Black real compensation for the exchange, although I am not sure if he can claim full equality. 18.Bxc5 In Fischer – Robatsch, Vinkovci 1968, Black should have played:

18...Rc8!N 19.Bd4 Rxe4 20.Nc3 Rxe1† 21.Rxe1 Bxf3 22.gxf3 Kf8! The last move covers the e7-square, enabling Black to move his bishop. White’s kingside structure has been damaged but he has the two bishops, and these two factors should give a roughly balanced game. An illustrative line is:

23.Ne4 Nxe4 24.fxe4 Rc6! Having in mind both ...Nc4 and/or ...Bf6. The best my engine has to offer is: 25.Rd1 Bc7 26.Kg2 a5 27.Rd3 f6 Black has absolutely no problems.

  13...c5 Although this game took place in 1971, Spassky’s position has a modern feel to it, especially in the way he organizes his pieces over the next few moves.

14.Bc2 cxd4! It was not until the mid-80s, when Romanishin introduced this plan in the variation that bears his name (after 9...Na5 10.Bc2 c5 11.d4, the move 11...Bb7!? is known as the Romanishin Variation), when he waited until the white knight went to d2 before performing the double exchange on d4, that the wider chess public began to respect Black’s chances in this structure. Until then, thanks to the influence of some great games of Geller and Keres, it was thought that the correct strategy for Black was to exchange once on d4 and prepare the ...d5 break. Here though, with the knight on b6 instead of a5, the double exchange seems more justified. Once Black has put his bishop on b7 and his rooks on the cand e-files, the ...d5 break will come with a lot of energy.   15.Nhf3 Re8 16.Nxd4 Bf8 17.b3 Bb7 18.Bb2 Nbd7 19.Qf3 19.Nf5 g6 20.Ng3 Bg7 21.Rb1 Rc8 22.Nf3 Qc7 was also excellent for Black in a more recent game, R. Jones – Smeets, Novi Sad 2009.

19...Rc8 19...g6!?N could be a slight improvement.   20.Rad1 d5 After liquidating his isolated pawn Black has no problems whatsoever. The players did not take much longer to agree a draw, but the engines slightly prefer Black at this point.   21.Bb1 dxe4 Once again 21...g6!?N comes into consideration.   22.Nxe4 Nxe4 23.Bxe4 Rxe4 24.Rxe4 Ne5 25.Qe2 Bxe4 ½–½   GAME 52 Zadok Domnitz – Samuel Reshevsky Netanya 1973   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.c4 This move has been around since the 1950s and has been used on and off by many top players over the years. Recently it has been tried at the top level, but I think that this and the following game might help to relegate it back to obscurity for a while at least!

11...c6 12.c5 12.a3 is the currently fashionable line, which is covered in the next game.   12.Qc2 is not to be feared. According to Kasparov, both 12...Bb7 and 12...a5 are fine for Black.   12.Bg5 is met by 12...h6 13.Bh4 and now 13...Nh5! gives Black easy play, with White scoring just half a point from five games in my database. See Tukmakov – Karpov, Leningrad 1973, for an excellent illustration of Black’s chances.   12.cxb5 axb5 13.Nc3 has been tried by several strong players, most notably Fischer in Game 29 of his 1992 match against Spassky. Interestingly, the best reply was shown as far back as 1967:

13...Ba6! Black wants to play ...c5, which might be further prepared by ...Qb8. This move was first played by Portisch and was later given the stamp of approval by three Breyer stars in Karpov, Jansa and Unzicker.  

12.Nc3 b4! closes the position favourably for Black. Kasparov states that Black is already equal, but it is worth taking a closer look to understand why. 13.Na4 c5 14.d5  

14...Re8!? (14...Ne8 15.Bc2 g6 16.Bh6 Ng7 followed by the typical Czech Benoni regrouping of ...Kh8, ...Nf6-g8 and eventually ...f5 also gave Black good play in Thomassen – Avrukh, Reykjavik 2012) 15.Bc2 Nf8 16.a3 a5 17.b3 Ng6 18.Nb2 Bd7 Black followed up with ...h6, ...Nh7 and (when the knight had moved away from f3) ...Bg5, obtaining somewhat better chances in Nakamura – Carlsen, Medias 2011.

12...Qc7! 13.cxd6 Bxd6 14.Bg5 exd4! 15.Bxf6 gxf6 It has been known for many decades that this position offers Black excellent chances, even though a certain amount of care is required in order prevent White from building a kingside attack.   16.Nxd4

16.Qxd4?! Ne5 17.Nbd2 (17.Qe3 is better, but Black has at least a small edge after 17...Qa7!)

17...c5! 18.Qe3 (18.Qc3 is no better after 18...c4 19.Bc2 Qb6!Nµ when Black takes control of the important a7-g1 diagonal) 18...c4 19.Bc2 Bc5 20.Qh6 Qb6! Black had a clear advantage in Lepsenyi – Jenik, Hungary 1967. His last move defended f6 while attacking f2, thus gaining time for ...Re8, which in turn allows ...Bf8 whenever Black feels the need to expel the queen.   16...Nc5 This is the main line, but there is also something to be said for 16...Ne5!? 17.Nc3 Bc5 18.Nf5 Kh8 19.Qe2 a5 20.Bc2 Rg8 21.Kh1 Be6 when Black had slightly the better of a complex game in De Meye – Ghysens, corr. 2012.

17.Nf5? This has been the most popular choice in the position, but Reshevsky refutes it beautifully.  

17.Bc2 is rather slow, and 17...Be5! is a good response: 18.Qh5 (18.f4 was played in Simonsen – Almeida, Brazil 1981, when it is hard to guess why Black avoided 18...Bxf4, as 19.e5 fxe5 20.Qh5 f6 refutes the attack) White’s last move was played in Mayer – Kostic, USA 1964, with the point that capturing the knight would allow e4-e5. However, Black can keep a clear advantage with:

18...Ne6! 19.Nf5 Kh8µ Intending ...Rg8.   17.Nd2 is probably White’s best try. Interestingly, Black’s best reply has only been tested in a couple of correspondence games so far. 17...Nxb3! (17...Rd8 18.N2f3 Nxb3 19.Qxb3 c5 20.Nf5 Be6 21.Qc2 is the main line which is considered equal) 18.Qxb3 c5 19.Nf5 Here I found an improvement over Benovic – Meye, email 2006.

19...Bxf5!N 20.exf5 c4 21.Qc2 Rfe8 Compared with the note at move 17, Black can arrange his rooks optimally on e8 and d8, without wasting a tempo moving the f8-rook to d8. Black has the more mobile pawn majority and the stronger minor piece, so overall he has the better chances.

  17...Bxf5 18.exf5 Rad8 19.Qh5

19...Bh2†! The first in a sequence of simple but powerful moves, which sees Black’s pieces take up dominating posts in the centre.   20.Kh1 Nd3 21.Rf1 Be5 22.Nc3 Bxc3 23.bxc3 Qf4 24.Kg1 c5 25.Bc2 c4 26.a4 Rd5 27.axb5 axb5µ

White is in serious trouble. Domnitz decides to exchange queens and hope for the best, but Reshevsky never lets him off the hook.   28.Qg4† Qxg4 29.hxg4 Nf4 30.Rfe1 Rd2 31.Bd1 Nd5 32.Rc1 Rd3 33.g5 fxg5 34.Re5 f6 35.Re6 Nxc3 Black is two pawns up, and the c-pawn is close to promotion. The rest is easy.

36.Bh5 Nd5 37.g3 Rc8 38.Ra1 c3 39.Be2 Rd2 40.Bxb5 c2 41.Ree1 Nc3 White resigned. What a wonderful display of technique by Reshevsky! 0–1     GAME 53 Sergei Zhigalko – Vassily Ivanchuk Riga (rapid) 2014   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.c4 c6 12.a3 As I mentioned earlier, this has been tested by quite a few strong players in recent years, but Black has nothing to fear if he follows Ivanchuk’s example.  

12...bxc4 13.Bxc4 Nxe4 14.dxe5 14.Rxe4 d5 15.Re1 dxc4 has occurred in a bunch of games, but White is equal at best after losing the bishop pair.   14...d5 15.Bd3 15.Ba2 is possible, but Black can play in a similar style to Ivanchuk with 15...Ndc5!?N planning ...a5 and ...Ba6. (15...Kh8 is given by Eljanov to support his opinion that White can get an edge, but I think Black can do better.) A logical continuation is:

16.b4 (16.Qc2 is given as best by the engines, but it’s nothing – Black just continues with 16...a5 and stands excellently) 16...Ne6 Black has an excellent game. A good plan is ...Bb7, overprotecting the d5-pawn in order to play ...a5 and follow up with ...c5 in the event that the b4-pawn disappears.   15...Ndc5 16.Bc2

16...f5! 17.Nd4 Qb6 18.b4 Ne6µ Ivanchuk’s powerful play has given him a great position, and White has problems in all areas of the board.   19.Bb2?! 19.Be3 is a better try, although Black still does well after 19...f4 20.Nf5 Qb7 21.Nxe7† Qxe7 22.Bd4 and now both 22...f3!? and 22...Qh4!? give him a powerful attack.

19...a5! 20.Re2 axb4 21.axb4 Rxa1 22.Bxa1 Bxb4 23.Nxe6 Bxe6 24.Bxe4 fxe4 25.Nc3 Bc5 White is not only a pawn down, but also faces a possible disaster on the kingside. Zhigalko manages to defend there, but leaves the queenside and the centre at Black’s mercy.

26.Qe1 Qb3 27.Nd1 Ra8 28.Bb2 Ra2 29.Qd2 Qd3 30.Nc3 Qxd2 31.Rxd2 e3 32.fxe3 Bxe3† 33.Kf1 Ra8 34.Rd1 d4 35.Ne4 Bc4† 36.Ke1 Rf8 Black had other winning options along the way, but one cannot fault the way Ivanchuk beautifully constructed a mating net with just a few pieces. White could have resigned here, but instead he gives up a piece to prolong the game for a few more moves.

37.Nf6† gxf6 38.Bxd4 Bxd4 39.Rxd4 Bd5 40.Rg4† Kf7 41.exf6 Rg8 42.Rxg8 Kxg8 43.g4 Kf7 44.g5 Kg6 45.h4 Kh5 0–1 Conclusion   The 10.d3 system may not look threatening at first glance, but it is traditionally one of the most popular ways to meet the Breyer. White can play his next four or five moves on autopilot and then choose between several plans. After

10...Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5! 13.Bc2 Re8 14.Ng3 Bf8 White has three main options.   15.a4 allows 15...Nxa4! 16.Bxa4 bxa4 followed by ...d5, with easy play for Black.   15.Nh2 d5 16.Qf3 Ne6 17.Nf5 Kh8! leaves Black solid on the kingside, with good prospects to take over in the centre with ...c5 and perhaps ...d4, as Portisch has shown.   15.b4! Ncd7! is the critical continuation, when White has another choice. 16.a4 contains a lot of hidden resources, but 16...c5! offers Black fully adequate play. 16.Bb3 eyes up the f7-square, but after 16...a5 17.a3 axb4 18.cxb4 h6! Black is safe on the kingside and is poised to gain space in the centre with ...c5 and/or ...d5, as shown in the Tal – Spassky game. Finally, 16.d4 a5 17.a3 axb4 18.cxb4 c5 is fine for Black as long as he follows one of the recommended improvements over Vasiukov – Averbakh. It is also worth remembering 17.Bd2 axb4 18.cxb4, when 18...Ra3!N followed by ...Qa8 is the most convincing equalizer.   We finished the chapter by analysing some sidelines after 10.d4 Nbd7. Let me repeat the crucial point that, after the usual 11.Nbd2, Black must avoid 11...Re8?? on account of 12.Bxf7† when White has a winning position. Even the almighty Spassky, our greatest Breyer role model, once committed this amazing blunder against Judit Polgar, so please remain vigilant and play 11...Bb7 instead. Moving on, we saw that 11.Nh4 should be met by 11...exd4 12.cxd4 Nb6 followed by ...c5. If White puts the knight on d2 then Spassky’s double exchange on d4 makes a lot of sense, while Fischer’s 13.Nf3 is also not dangerous after my recommended novelty a few moves down the line.   Finally, 11.c4 c6! is nothing much to worry about. 12.c5 Qc7! 13.cxd6 Bxd6 14.Bg5 exd4! was shown to be fine for Black in the Reshevsky game. 12.a3 has been tried by some strong players recently but, after 12...bxc4 13.Bxc4 Nxe4!, Ivanchuk showed that Black has every reason to be happy.

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8   13.b4 Bf8 14.a4 Nb6 15.a5 Nbd7 16.Bb2 Rb8 17.Qb1 Game 54, page 312 17.Rb1 Ba8! 18.Ba1 g6 19.Bd3?! Game 55, page 314 19.c4 Game 56, page 317   Introduction to 13.a4 page 318 13...Bf8 14.Bd3 c6 15.b4 Game 57, page 319 15.Qc2 Game 58, page 323 (15.Nf1 page 323) 15.b3 g6 16.Bb2 Bg7 Game 59, page 326 (16...Nh5! page 326) 14.b3! g6 15.Bb2 (15.Ba3 page 330; 15.Bd3 c6 16.Ba3 page 330) 15...Bg7! (15...Nh5!? page 331) 16.Bd3 c6 17.Qc2 Rc8 18.Bf1! Nh5 19.dxe5! dxe5! 20.Ba3?! Game 60, page 329 The Super-Critical 20.b4! page 336  

Diagram Preview

On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How should Black continue? (page 313)  

White has just played 22.c4.

How would you respond? (page 328)

Can you spot Smejkal’s great positional idea? (page 315)

White has just played 20.e5.

How does Black fight for counterplay? (page 330)  

Suggest a plan for Black. (page 322)  

Find the best way to deal with the attack

on the b7-bishop. (page 338)  

How should Black play on the queenside? (page 325) Suggest a good plan for Black. (page 339)   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8

Now that the bishop no longer ‘salutes’ our f7-pawn, the time has come to carry out the typical Spanish regrouping with ...Re8 and ...Bf8, and possibly ...g6 and ...Bg7. One of White’s major concerns is the pressure on his e-pawn, so he has to make a decision regarding the future of his knight on d2. The traditional Spanish continuation is 13.Nf1 followed by Ng3, and this will receive dedicated coverage in the next and final chapter.   In the present chapter we will examine White’s other possible attempts. If the knight is not going to g3, it will have to remain on d2 for the time being, in order to maintain the defence of the e4-pawn. In that case, White’s general objective will be to play on the queenside. A typical plan involves a timely exchange on e5, ensuring that the e8-rook will not breathe fire on e4, followed by Nb3 – or possibly Nc4 if the b5-pawn has been removed.   From the tabiya shown in the diagram, there are two main ways for White to play on the queenside:   13.b4 Bf8 14.a4 is a rare choice nowadays, but it was a hot line in the 1970s. The reason why its popularity declined will be shown in Game 56, where the mighty Karpov turns in a model performance against the late Walter Browne, who was the main supporter of White’s cause. In later years Black was found to have adequate resources in other lines, so from a theoretical point of view we can consider the case closed.   13.a4 is more challenging; I suspect this line will become more popular in the near future, as analysis engines tend to like White’s chances. Fortunately, I had to face this move in a correspondence game, and I discovered an effective way for Black to solve his opening problems, even in the event that White improves on my opponent’s play.   13.b3 also deserves a brief mention. 13...Bf8 14.d5 c6 15.c4 is a logical continuation, when Black should play:

15...Qc7! This introduces a thematic Breyer plan, which we will encounter again in the next chapter. Black intends to play ...Rec8 (this is generally the correct rook, as it had no real function on the e-file and the other rook belongs on b8) followed by ...bxc4, ...Nc5 (which may be supported by ...Nfd7), ...a5 and ...Ba6, with pressure against the c4-pawn and on the queenside generally. If White wants to block the centre, he has to be more subtle about it (once again, I refer you to the next chapter for a more challenging version of this plan).     GAME 54 Karel Mokry – Jan Smejkal Bratislava 1983   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Bb7 10.d4 Re8 11.Nbd2 Bf8 12.Bc2 Nb8 13.a4 Nbd7 14.b4

14...Nb6! “Other plans do not promise Black full equality” – Kasparov.   15.a5 Nbd7 Having provoked White into releasing the tension on the queenside, Black can think about counterattacking with ...d5 (or possibly ...c5), so White’s next move is obligatory if he wants to play for any kind of edge.   16.Bb2! 16.d5 is of course met by 16...c6!, breaking up the centre to Black’s advantage.   It may seem paradoxical to develop one’s bishop on a blocked diagonal, but the text move is the only effective way to counter the ...d5 break. Now it is Black’s turn to search for a good move, as 17.c4 is a positional threat which could lead to unpleasant pressure.

16...Rb8! This excellent prophylactic move effectively rules out c3-c4, as the b4-pawn would become a serious weakness.   16...d5?! is met by 17.dxe5 Nxe5 18.Nxe5 Rxe5 19.c4 with good prospects for White.   16...c5?! is no better as 17.d5 leaves Black unable to challenge White’s centre with ...c6, so he will remain slightly worse with no visible counterplay.   16...exd4?! 17.cxd4 c5 is an attempt to improve on the ...c5 idea shown above, but this time White has 18.e5! with the initiative – and of course if 18...dxe5? 19.dxc5! Black is clearly worse.   16...Qb8?! has similar intentions to our main line, and was played by Spassky against Fischer in Game 10 of their 1972 match in Reykjavik. Unfortunately White can play 17.c4! anyway. (Fischer chose 17.Rb1 and got no advantage, although he eventually won the game.) 17...bxc4 18.Ba4! Black has some problems to solve. Anand suggested 18...Rd8 (after 18...exd4 19.Nxd4 Black cannot take on e4 because the d7-knight hangs) but after 19.d5 Rc8 20.Bc6 Black is cramped, as Kasparov points out.   17.Qb1 The idea of this move is to play Qa2 and Bb3, with pressure against f7.   17.Rb1 is the main line and is covered in the next two games.   17.c4? bxc4! is great for Black, who intends ...exd4 – and if 18.d5? then of course 18...c6! is extremely strong.

17...Nh5! Black finds an excellent way to exploit the white queen’s absence from the kingside! The knight is heading for f4 and perhaps later e6, while also making room for the queen to come to f6, defending f7 while exerting her influence on the kingside and in the centre. A secondary purpose of the knight could be revealed if Black plays ...c5 and White answers with d4-d5: with the centre blocked, Black can play ...g6 and ...Ng7, and possibly ...Be7 and ...Rf8, in order to prepare counterplay with ...f5.   18.Qa2 18.c4 is another idea, but Black did well after18...bxc4 19.Nxc4 exd4 20.Bxd4 c5!? (20...Nf4N and 20...Nhf6N are also sensible) 21.Be3? (21.bxc5 Nxc5=) 21...cxb4 22.Nb6 Nhf6 23.Qxb4 d5!µ This was Ljubojevic – Karpov, Manila 1976, another of Karpov’s technical masterpieces in the Breyer.   18...exd4 19.Bb3 Qf6 20.cxd4 Nf4 21.Re3

21...d5! An important move, shutting White’s bishops out of play while fixing d4 as a future target.   22.e5 Qg6 23.g3 Bxb4 24.Kh2 Ne6 25.Bxd5 Bxd5 26.Qxd5 Bxd2 27.Nxd2 Qc2 This puts pressure on White’s position, but 27...Rbd8!?N is an even more promising alternative according to the engine.   28.Bc3 b4 Once again Black has an improvement: 28...c5!N 29.dxc5 Ndxc5 Now ...b4 and ...Red8 are strong threats. A semi-forcing line continues: 30.Ra2 Qg6 Renewing both threats, so White has to offer to simplify. 31.Ne4 Nxe4 32.Qxe4

32...Qxe4 33.Rxe4 Rbc8! Advancing the b-pawn is no longer so strong, so Black should keep is as an asset for later. 34.Bd2 Red8³ Black is better, as both his rooks are on open files and he has a protected, outside passed pawn.   29.Ra2 Here the game was rather prematurely agreed drawn. Interestingly, the game Knaak – Jansa, Southport 1983, ended at an identical moment, but I was unable to ascertain which game came first. ½–½   Let’s now return to late in the opening and consider White’s attempt to prepare the c3-c4 break with 17.Rb1.

GAME 55 Wolfgang Unzicker – Jan Smejkal Lucerne (ol) 1982   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Bb7 10.d4 Re8 11.Nbd2 Bf8 12.Bc2 Nb8 13.b4 Nbd7 14.a4 Nb6 15.a5 Nbd7 16.Bb2 Rb8 17.Rb1 This is the best way for White to try and carry out his plan of c3-c4.

17...Ba8! A logical follow-up to Black’s last move.   18.Ba1 For the second turn in a row, White goes for the mirror image of his opponent’s move. Now he is finally ready for c3c4, but Black is well prepared to meet it and has more than one decent way to continue.   18...g6 This was Karpov’s choice, and I like it the most because of its flexibility.   18...h6 is also playable, although it allows White the interesting option of: 19.Re3!? (19.c4 leads nowhere for White; for further details, check out the impressive game Hjartarson – Beliavsky, Belfort 1988) 19...g6 20.Qe2 c6 21.c4 Bg7? (21...exd4! was subsequently proposed as an improvement) 22.dxe5 dxe5 23.c5!± In Shirov – Leko, Ljubljana 1995, White had the easy plan of Rd3, Rbd1 and Nf1, and he went on to suffocate his opponent. If you wish, you can look up the remainder of the game for an example of what Black must avoid when playing the Breyer.  

19.Bd3?! At first it is hard to see why this should be a mistake, but Smejkal, one of the great Breyer specialists, shows that the bishop has made itself a target for ...c5-c4. 19.c4 is the consistent follow-up to White’s last few moves, and we will examine it in the next game.   19...exd4! 20.cxd4 c5! Black is threatening ...cxb4 when the a5-pawn will hang, so White must either venture a questionable pawn sacrifice or continue as Unzicker did.   21.Bc3 Now the queenside pawns are secure, but Black gets to carry out his plan of gaining time against the d3-bishop.

21...d5! 21...c4 22.Bc2 d5! leads to the same thing, but Smejkal’s move order is flashier.   22.e5 c4 23.Bc2 Ne4! Smejkal sacrifices a pawn to clear the ideal d5-square for his knight. This is a vital element of Black’s plan; if he simply moved the knight away he would stand worse, as the c4-pawn would be firmly blockaded and Black’s pieces stymied.   24.Nxe4 dxe4 25.Bxe4 Bxe4 26.Rxe4 Nf6 Thanks to the undefended rook, the knight gets to the perfect d5-square.

27.Re1 Nd5 28.Qc2 Qd7 29.Nd2 The machine evaluates the position as equal, but to a human player it is obvious that Black is in the driving seat. Still, converting his positional advantage into a win is not at all easy. Smejkal hits on the right plan of building a kingside attack, helped by the fact that White has weakened his castled position with h2-h3.   29...f5! 30.Nf1 Rbd8 31.Red1 f4! 32.Nh2 Qf5 33.Qb2 Rd7! Preparing to swing the rook to the kingside.  

34.Nf3 g5 35.Be1 Rg7 36.Nh2 h5 37.f3 Otherwise ...g4 will be nasty.   37...Rd8 38.Rbc1 Rgd7 39.Nf1 Rc8 40.Bc3 Rg7 41.Nh2 Be7 There is still no clear way through, but Black can take his time to shuffle his pieces around and wear down his opponent, while White has no active ideas at all. For instance, exchanging queens on c2 would drop the b4-pawn.

42.Qd2 Rc6 43.Qe1 Rcg6 44.Bd2 Kf7 45.Rb1 Ke6 46.Rbc1 Rh6 47.Qe4 Rg8 Black doesn’t mind exchanging queens as White’s queenside would collapse in the endgame.

48.Be1 Rc8 49.Qe2 Rhh8 50.Bc3 Rh6 51.Rc2 Rg8 52.Rcc1 Qf8 53.Nf1? White was in trouble, as Smejkal was poised to capture the b4-pawn. Perhaps Unzicker became distracted by the queenside and dreamed of getting his knight into play, but momentarily forgot about the breakthrough Black had been planning all along.

53...g4 At last!   54.fxg4 hxg4 55.hxg4 Rh4 56.Nd2 Qh6 57.Qe4 Qh7 This was the last move according to the database, but I wonder if a data error occurred, as Black has an elementary win with 57...Rh1† 58.Kf2 Qh4† 59.Ke2 Qxg4† etc. In any case, Smejkal’s earlier handling of the game, with the positional pawn sacrifice and gradual improvement of his position, without allowing White a trace of counterplay, was

highly instructive. 0–1   The next game shows how Black should react when White executes his plan of c3-c4.   GAME 56 Walter Browne – Anatoly Karpov Amsterdam 1976   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.c3 d6 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.b4 Bf8 14.a4 Nb6 15.a5 Nbd7 16.Bb2 Rb8 17.Rb1 Ba8 18.Ba1 g6 19.c4 This is White’s most critical attempt, but Karpov shows that Black has nothing to fear.

19...bxc4 20.dxe5 20.d5 is met by the typical 20...c6 of course, while 20.Nxc4 d5! gives Black good prospects – that is why Browne exchanges on e5 before recapturing. Previously he had scored 2/2 against Ljubojevic and Smejkal with this variation, before running into Karpov. 20...Nxe5 20...dxe5!?N has never been tried, but the engines indicate that the position is equal after 21.Bc3 c5 22.bxc5 Rb5. More often than not though, it tends to be in Black’s interests to trade a pair of knights in the event of an exchange on e5, so I will tend to stick to that principle unless there is a specific reason not to.   21.Nxe5 dxe5 22.Bc3

22...Bc6 Hjartarson’s 22...c5 is fine as well. After 23.bxc5, even the novelty 23...Rc8!?N offers Black equal chances.   23.Qe2 Bb5 24.Nxc4 c5 25.bxc5 Rc8! This was Karpov’s improvement over 25...Bxc5 which had been played previously, for instance by Smejkal against Browne, four months before the present game took place. Black is in no way worse, and it is instructive to see how Karpov outplays a strong opponent from here.   26.Bb3 Browne supports the knight and plans Qa2 to put pressure on f7.   26...Rxc5 27.Bb4 Rc6 28.Bxf8 Rxf8 29.Qa2

29...Bxc4! 30.Bxc4 Qc7 Karpov goes on to show that his knight is not worse than White’s bishop, and that the pawn on a5 is just as weak as the one on a6.   31.Bf1 Rd8 32.Rb6 Rdd6 33.Rxc6 Rxc6 34.Bd3 Nh5 35.g3 Ng7 36.Rb1 Ne6³ I have already touched on the plan of regrouping the knight via h5, g7 (or f4) and e6, and now we see Karpov using the same idea to excellent effect.   37.Kg2 Nc5 38.Bc2 Rd6 39.Qc4 Qxa5 40.Rb8† Kg7 Karpov goes on to convert the extra pawn with his usual exemplary technique.

41.Rc8 Ne6 42.Qb3 Rd8 43.Qc3 Qxc3 44.Rxc3 Rd7 45.Ra3 Nc5 46.Kf3 Kf6 47.Ke3 Ke7 48.Ra5 Kd6 49.Bd3 Ra7 50.Bc4 Kc6 51.f4 f6 52.fxe5 fxe5 53.Bd5† Kb6 54.Ra1 a5 55.Rb1† Kc7 56.Rb5 Nd7 57.Kd3 a4 58.Kc2 Ra6 59.Rb7† Kd6 60.Rb1 a3 61.Bc4 Ra7 62.Rd1† Kc5 63.Ba2 Nf6 64.Re1 Kd4 65.g4 Nxe4 66.Rd1† Ke3 67.Re1† Kf4 68.Rf1† Kg3 69.Re1 Ng5 70.Re3† Kh4 71.Rxe5 Nxh3 72.Kb1 Nf2 73.Be6 Nxg4 0–1   So far we have seen that Black has fully adequate resources to deal with 13.b4. The remainder of the chapter will deal with a trickier way for White to play on the queenside.

Introduction to 13.a4   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 This is an important move, and leads to some of the most complex positions in the entire book. There are many possible move orders, which can make things confusing, so I will summarize the main options before we start analysing anything in detail.

13...Bf8 14.Bd3 14.b3! may transpose in a few moves but is a slightly more accurate choice, for reasons that will be discussed later. The present move order has been much more popular though, and it also allows me to introduce a couple of deviations which will be covered in the next few illustrative games.   14...c6 15.b3 The alternatives 15.b4, 15.Qc2 and 15.Nf1 will be discussed in Games 57 and 58.   15...g6 16.Bb2 Bg7 17.Qc2 Rc8

This is the position I really wanted to show you. The previous four moves can be shuffled in various orders, but it is good to keep the above position in mind as a reference point. The various plans from this point will be considered in Games 59 and 60, as well as in the analysis section at the end of the chapter.     GAME 57 Hagen Tiemann – Einar Laane email 2010   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 d6 7.c3 b5 8.Bb3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.Bd3 c6 15.b4 This is a straightforward attempt by White to gain space on the queenside. It was popular four or five years ago but today it is rarely seen at the top level.

15...Nb6 This is a typical move in Breyer lines involving the a4-b4 structure on the queenside; keep in mind that the first three games of the chapter all featured the near-identical position with White’s bishop back on c2 and Black’s pawn on c7. Black can also delay the knight move with 15...Rc8 (15...Qc7 also has the same general idea). This way Black prepares ...Nb6, after which axb5 will be met by ...cxb5 with an immediate attack on the c3-pawn. 16.Bb2 is a logical reply, when 16...Nb6 17.axb5 cxb5 18.d5 transposes to the main game. Alternatively, Black can keep things flexible with a move like 16...g6!?.   16.axb5! It is also vital to consider the alternative: 16.a5 Nbd7 17.Bb2 This keeps the game in parallel with Games 54-56, the only difference being the placement of the bishop on d3 and pawn on c6. Black can exploit this slight change in the position with:

17...exd4! Several strong players have missed this opportunity and played 17...g6, but Black’s results have not been encouraging at all. 18.cxd4 c5! It looks as though Black has lost a tempo with ...c6-c5, but White’s ‘free move’ of Bd3 significantly harms his chances. 19.bxc5 19.e5 dxe5 20.dxc5? (20.dxe5 is better, but Black is more than okay after 20...Nd5³) worked well for White in the analogous position, as noted in the 16...exd4?! line in Game 54 , but here it is easily refuted by 20...e4 (or 20...Nxc5), precisely because the bishop stands on d3 rather than c2! 19...dxc5 20.e5 Nd5 Black has excellent prospects, but the position is quite sharp so it is worth giving a few more moves.

21.h4!? Threatening a sacrifice on h7. 21.Qc2 h6 22.Bf5 c4 was better for Black in De Meye – Mignon, corr. 2014. 21...h6 22.Bb1 Now White wants to put the queen on c2, so... 22...Nb4! The whole attacking operation is defused. A correspondence game continued:

23.Bf5 c4 24.Ba3 Nd5 25.Bxf8 Nxf8 26.Ne4 Ne7 27.Nd6 Nxf5 28.Nxf5 Bd5 29.Nd2 Ng6 Black was in control and went on to win in Morcin – Lizorkina, email 2011.   16...cxb5! 16...axb5 is playable but rather dull. 17.Rxa8 Bxa8 (17...Qxa8 is similar; note that the inclusion of the ...c6 move helps White as it blocks Black’s bishop and queen) Black is slightly worse; I spent some time analysing this position and was unable to find either a forcing way for Black to equalize, or a clear way for White to build on his slight edge. A relevant practical example is:

18.Nb3 Bb7 19.dxe5 dxe5 20.Be3 Bc8 21.Qc2 Qc7 22.Na5 Bd7 23.Rc1 h6 24.Nd2 Na4 25.Ndb3 Rb8 26.Ra1 c5 27.bxc5 Nxc5 28.Nxc5 ½–½ Anand – Carlsen, Kristiansund (rapid) 2010. If you do not feel like taking any risks, then recapturing with the a-pawn is a valid choice. The game should end in a draw if Black does nothing weird, but he must be ready to defend for a while without much of a chance to take over the initiative.   Recapturing with the c-pawn keeps the position a bit more lively. I cannot see any theoretical problems for Black, and his position continues to hold up well in high-level correspondence games.   17.d5 This move gains space and blocks the bishop on b7; so far, nothing else has been played. Should White try some other move, Black can even consider ...Nc4 on the next turn, exploiting the vulnerability of the e4-pawn.   17...Rc8 18.Bb2 18.Qb3?! misplaces the queen, as will become clear a few moves down the line: 18...Qc7! Black starts by improving his queen with tempo. 19.Bb2 Nh5! (19...g6 transposes to our main game but the text move is stronger) 20.Rac1 In Shirov – Baramidze, Germany 2010, Black could and should have destroyed his opponent’s centre with:

20...Nf4!N 21.Bf1 f5 22.exf5 Bxd5³ Black gains a tempo against the queen and will strengthen his position with ...Qb7 next.   18.Ra3 is a fairly sensible way to defend the c3-pawn. It does not change the position a great deal though, and I suggest continuing with 18...g6, with similar ideas as in the main game.

18...g6 This move has many uses. The f8-bishop gains some prospects, and the g-pawn supports a future ...Nh5 and ...f5. If I had a couple of spare moves in this position, I would play ...Bg7 and ...Qd7 (or perhaps ...Rf8) before committing to ...Nh5 and ...f5. 18...Nh5 was Carlsen’s choice against Shirov. It may not be a bad move objectively, but at this stage of the game it is a bit too committal for my liking. 18...Qd7!? 19.Qe2 Nh5 20.Qe3 Nf4 21.Bc2 Nc4 seems perfectly reliable though, as some correspondence games

have demonstrated.   19.Qb3 White threatens to liberate his position with c3-c4.   19.Nf1 gives Black a pleasant choice. 19...Nc4 is one good option but 19...Nh5 also makes a lot of sense, as the bishop has been deprived of its natural retreat square on f1. A possible continuation is 20.Ne3 f5! 21.Nd2 fxe4 and the d5pawn becomes a major weakness.

19...Qc7! 20.Rec1 White has tried 20.Rac1, but the same plan works just as well for Black against that too.   20...Bh6 21.Rc2 Nxe4! 22.Nxe4 Bg7 Black threatens both ...Bxd5 and ...f5.   23.c4! Without this White would be in trouble. The following sequence is close to being forced for both sides.

23...bxc4 24.Bxc4 Nxc4 25.Rxc4 Qxc4 26.Qxc4 Rxc4 27.Nxd6 Rxb4 28.Nxe8 Rxb2 29.Nxg7 Kxg7 30.Rd1 Rb5 31.d6 Bxf3 32.d7 Bxd1 33.d8=Q Bb3

34.Qd6 a5 35.Qd3 At this point the correspondence players agreed a draw, as Black lacks the necessary coordination to keep all of his pawns. In a practical game, Black could still keep some winning chances alive with 35...Rb4 36.Qd6 (or 36.Qc3) 36...f6 37.Qc7† Bf7 38.Qxa5 Rd4, followed by ...g5, ...h5 and so on. ½–½ We now begin our coverage of White’s set-ups involving a pawn on b3 rather than b4.     GAME 58

Alexei Bezgodov – Yuri Balashov Perm 1997   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.Bd3 c6

15.Qc2 This has been the choice of several leading players, but I think it is slightly inaccurate. At first this seems like a flexible move, as White improves his queen before deciding whether to play b2-b3 or b2-b4 next. However, the queen on c2 is not so well suited to the b4 structure, especially if Black gets his rook to the open c-file like in the previous game. Therefore, in the present game White plays b2-b3 anyway, but this gives Black a slightly improved version of Games 59 and 60, for reasons explained at move 17.   Another idea in the same spirit is: 15.Nf1 This move keeps things flexible on the queenside, but White loses certain options such as Nb3. 15...exd4!? 15...d5 is also perfectly viable. True, 16.Bg5 puts a certain amount of pressure on Black, and has scored well for White in practice. However, the improvement 16...Qc7!N enables Black to equalize with a bit of care: 17.Nxe5 (both 17.dxe5 Nxe4 and 17.Ng3 Nxe4 18.Nxe4 dxe4 19.Bxe4 exd4 are solid enough for Black) 17...Nxe5 18.Bf4 Bd6 19.dxe5 Bxe5 20.Bxe5 Qxe5 21.exd5 Qxd5 22.Ne3 Qd7 Black intends ...Rad8 and does not stand worse. 16.cxd4 c5 17.Ng3

17...c4! 18.Bb1 Qa5 Black intends ...b4. So far, this strong plan has netted him 1½/2 in high-level correspondence games. 19.Qc2!?N Both 19.Ra2 b4, in Ivanov – Schweer, email 2009, and 19.Re2 b4, in Ohtake – Noble, corr. 2013, were more than okay for Black. The text move is mentioned by Eljanov, and is directed against Black’s plan. 19...b4!? White’s idea is to meet 19...Rac8 with 20.b4!, with complicated play – albeit not worse for Black, according to Eljanov’s analysis. While I was analysing Eljanov’s idea, however, I realized that Black could play ...b4 anyway as a sacrifice. 20.Qxc4 d5 21.Qb3 21.exd5? Rxe1† 22.Nxe1 Rac8! will leave Black with two bishops against a rook (as White will take on h7 after Black captures on c1).

21...Nxe4 22.Nxe4 dxe4 23.Ba2 Qf5 24.Nh4 Qe6 25.Qxe6 fxe6 26.Nf5 a5 27.Ng3 27.Bf4 Nb6! maintains the balance (but not 27...Bd5 28.Bxd5 exd5 29.Ne3 when the d5-pawn is a chronic weakness). 27...Bd5 28.Nxe4 Rac8 Black’s activity provides full compensation for the sacrificed pawn.

15...g6! This flexible move is useful in all scenarios.   15...Rc8 has been tested in some high-level games, but there is no need to commit the rook at this stage. An important point is that 16.axb5 cxb5?! does not work so well with the white pawn on b2 instead of b4, so Black would have to recapture with the a-pawn, leaving his rook misplaced. The position is still close to equal, but I prefer the continuation in the main game.   16.b3 16.b4 Rc8!? is fine for Black; he intends ...Nb6, and is ready to recapture with ...cxb5 now that White’s queenside structure has been fixed.   16.dxe5 dxe5! 17.Nb3 Rc8 18.Bf1 Nb6 19.Na5 Ba8 20.b3 occurred in Kryvoruchko – Baramidze, Antalya 2013. Here it would have been good for Black to play:

20...Qc7!?N 21.Bd2 Nbd7 22.b4 c5! Black has no problems at all. The same plan would also have worked just as well against practically anything else White might have tried over the past few moves.   16...Rc8 Now that White has played b2-b3, this move makes more sense than on the previous turn.   16...Nh5!? is a good alternative, and may transpose to the game after a subsequent ...Rc8.   17.Bb2 17.axb5 is not terribly effective here. 17...axb5 is a solid reply which gives Black a slightly improved version of the 15...Rc8 line referred to in the notes above. 17...cxb5!? is also perfectly fine, now that b2-b3 has been played.   17.Ba3 Nh5 is similar to the game continuation.

17...Nh5! An important moment! 17...Bg7 would take us to the ‘target position’ mentioned in the introductory section on page 319. We will return to that position in Games 59 and 60, and I will go on to show that it is satisfactory for Black. However, the particular move order chosen by White in this game allows us to benefit from keeping the bishop on f8 for a little longer. One reason for doing this is that it helps to control the c5-square (one of White’s main plans is to exchange on e5 and advance with c4-c5); another is that the thematic ...Nh5-f4-e6 manoeuvre might be interrupted by g2-g3, in which case the knight can still go to e6 via g7.   18.Bf1 exd4! 18...Bg7 transposes to Game 60 but, once again, Black can do better.   18...Nf4 is a decent enough move, continuing the knight’s journey towards e6, but the game continuation is simply stronger.   19.cxd4 d5! 20.e5 Blocking the centre is a natural reaction, but Black has an excellent reply.   20.g3 can be met by the flexible 20...Ng7, reminding us once again of the benefits of leaving the bishop on f8. I found a few correspondence games in which Black followed up with ...b4, ...Ne6 and ...Bg7, with good prospects.

20...b4!!³ This splendid positional move seizes space on the queenside and prepares ...c5 without allowing any counterplay with axb5.   21.Nh2 c5 From here on Black plays a series of logical moves and White gets into trouble without doing anything obviously wrong.   22.Qd3 cxd4 23.Ndf3 Nc5 24.Qd1 Ne4 25.Bxd4 Nf4 26.Ng4 Ne6 27.Bb2 Bg7 28.Rc1 Nc3 29.Qd2 d4µ The c3-knight smothers White’s position and the rest of Black’s pieces coordinate perfectly. Balashov remains in control for the rest of the game.

30.Ngh2 Rc7 31.h4 h5 32.Bc4 Rd7 33.Qd3 Kh8 34.Bxe6 Rxe6 35.Qc4 Bh6 36.Qxb4 a5 37.Qa3 Bxc1 38.Bxc1 d3 39.Bg5 Qc8 40.Bf6† Rxf6 41.exf6 d2 42.Nxd2 Rxd2 Black has won a piece, and a few more accurate moves are enough to persuade White to give up.

43.Qe7 Rd7 44.Qe5 Kh7 45.Qxa5 Nd5 46.Nf3 Nxf6 47.Ng5† Kg8 48.Qe5 Qc6 49.f3 Qb6† 50.Kh2 Qf2 51.Qg3 Qd4 52.Re5 Bd5 53.Re2 Kg7 0–1 GAME 59  Maria Sebag – Humpy Koneru

Doha 2011   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.Bd3 14.b3! g6 15.Bb2 Bg7 16.Bd3 c6 is a slightly more accurate way to reach the same position as in the game. This move order will be featured in the next illustrative game.   14...c6 15.b3 The alternatives 15.b4, 15.Qc2 and 15.Nf1 have been discussed in the previous two games.

15...g6 16.Bb2 16.Qc2 transposes to Bezgodov – Balashov.   16.Ba3 has also been tried; this is covered later, under 15.Bd3 c6 16.Ba3 in the notes to Game 60.   16...Bg7 16...Nh5! would have been a touch more accurate. The knight move almost always features in Black’s plans at some point, and I have already mentioned that there are certain advantages to leaving the bishop on f8: Black has better control over the c5-square, and the knight has the option of using g7 as a route towards e6. This explains why 14.b3! is a more accurate move order; as you will see in the notes to the next game, the early ...Nh5 loses some of its effectiveness when the bishop has not gone to d3, although it is still playable even then.   Despite Black’s minor inaccuracy on the last move, we have reached a standard position which will also feature in the next game.

17.Qc2 This multipurpose move has almost always been played. It is not worth discussing alternatives in detail, as Black’s plans will remain pretty similar.   17...Rc8! 17...Qc7 has been played, but at this stage it is not clear if Black’s queen will be best placed on c7, b6 or even f6. Therefore it makes sense to place the rook opposite the white queen before deciding what to do next.   18.Rad1 This move looks sensible enough, but 18.Bf1! is more flexible, as it saves time while keeping the option of utilizing the rook along the a-file. This move will be discussed in the next main game and the analysis section that follows it.   Having mobilized the last of her pieces, White intends to drop the bishop back to f1 to clear some space in the centre. At some point she will advance with c3-c4, possibly after exchanging on e5. However, when we look at the specifics, there is always something Black can gain in return. For example, if c3-c4 is played without dxe5, Black can take control of the dark squares by playing ...exd4, ...b4 and ...c5. If White exchanges on e5 before playing c3-c4, Black can exploit the fixed central structure, as shown in the note to White’s 21st move below.

18...Nh5 19.Bf1 Nf4 Having made all the obvious moves, it is time for White to make an important decision: to exchange or not to exchange?   20.dxe5 20.c4 is the other choice, when Black targets the dark squares with 20...exd4 21.Nxd4 (21.Bxd4 Bxd4 22.Nxd4 c5 is also nice for Black) and now either 21...Qb6, as in Korchagina – Kashlinskaya, Yekaterinburg 2013, or 21...c5!?N, with better chances for Black in both cases.   20...dxe5 21.b4 21.c4N is an obvious alternative. This introduces the positional threat of c4-c5, so Black needs to stop it with 21...Qe7!.   21.Ba3N stops the queen from going to e7 and thus threatens c4-c5, but Black has a powerful antidote which needs to be remembered: 21...Qc7 22.c4

22...b4!! (22...bxa4 23.c5! is a thematic sacrifice, which has been played by Anand in a similar situation, which Black needs to know to avoid) 23.Bxb4 c5© Black follows up with ...Ne6-d4, with excellent compensation for the pawn. Both of White’s bishops have become bad, the knight will be fantastic on d4 and there will be strong counterplay along the b-file.

21...Qe7 22.Nb3 Rc7! This instructive move by Koneru has several ideas. Black overprotects the knight on d7, ensuring that the queen will not be tied to its defence. The rook also protects the bishop and gives it the option of retreating to c8 if needed.   23.Ba3 23.Na5 gives Black a choice between 23...Nb6 and 23...Bc8, with fine play in either case.   23...Qf6!

23...Bc8!? also seems fine, but Koneru’s move introduces the idea of ...Nxh3†.   24.Nfd2 Ne6 25.Na5 Bf8!? 25...Ba8 and 25...Bc8 also make a lot of sense.   26.Nxb7 Rxb7 27.Nb3 Rc7! 28.Rd2 28.Re3 is offered by Mikhalevski as an improvement, but Black is fine after 28...Nb6, with similar ideas as in the game.

28...Nb6! 29.a5 Nc4 The engine points out that 29...Na4 30.c4 c5! was a good alternative.   30.Bxc4 bxc4 31.Nc5 Bxc5 31...Nxc5!? 32.bxc5 Rb8 followed by doubling on the b-file was also slightly better for Black, but one can understand Koneru’s decision to play with a knight against a bad bishop.   32.bxc5 Nf4 33.Re3 Nd3 White seems to be in trouble, but Sebag correctly judges that she can create a positional fortress.

34.Rexd3! cxd3 35.Qxd3 Rb8 36.Bb4 With the b-file closed and White controlling the only other open file, Black is unable to exploit her material advantage.   36...Ra8 37.Qg3 Re7 38.Rd6 Re6 39.Qd3 h5 40.Rd7 Qg5

41.c4 Qc1† 42.Kh2 Qf4† 43.Kg1 Qc1† ½–½   The next game features a slightly more accurate plan from White. My opponent is a master-level correspondence player with a respectable peak rating of 2340.    

GAME 60 Aleksandr Istomin – Nikolaos Ntirlis Correspondence 2014   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.b3! As I mentioned earlier, I consider the lines where White postpones Bd3 to be more critical. See Black’s alternatives in the notes to move 15 for some concrete lines where the difference can be felt.

14...g6! 14...b4?! is a new idea I investigated, but White is better after 15.cxb4 exd4 16.Nxd4 d5 17.exd5, as he has a healthier queenside.   14...c6 has been played, but there is no need to rush with this move before White forces it with Bd3, so I prefer to improve Black’s position in other ways.   15.Bb2 15.Ba3 has been tried a couple of times, but 15...exd4!N 16.cxd4 c5 17.Bb2 Bg7 equalizes cleanly, for instance:

18.e5 dxe5 19.dxe5 Nd5 Black is fine, as the e5-pawn is a serious target.   A better attempt involving the bishop going to a3 is: 15.Bd3 c6 16.Ba3 This position can of course be reached via the 14.Bd3 move order, as mentioned in the note to White’s 16th move in the previous game. With the moves Bd3 and ...c6 included, Black does not have the option of ...exd4 and c5, as the b5-pawn would hang. However, he can utilize the position of the bishop on d3 with:

16...Nh5! Black will carry out the thematic ...Nf4-e6 manoeuvre, gaining a tempo against the bishop along the way. 17.Bf1 17.b4 an be compared with 15...c6?!, as mentioned in the notes to Black’s next move in the main game. The difference here is that White has already played Bd3, which has not only cost him a tempo, but also enables

Black to gain additional time with his knight. My analysis continues 17...Nf4 18.Bf1 exd4 19.cxd4 d5 20.e5 a5! with strong counterplay for Black. Remember, this was only possible because of the time Black saved by waiting for the bishop to come to d3 before playing ...c6. 17...Nf4 18.Qc2 Ne6 Now that he has full control over the c5-square, Black is fine. An instructive example continued:

19.Rad1 Qc7 20.Bb2 Bg7 Preparing to liquidate the centre. Black can also play more ambitiously with a move like 20...Rac8!?N. 21.Ba1 bxa4 22.bxa4 d5 23.c4 exd4 24.cxd5 cxd5 25.Qxc7 Nxc7 26.exd5 Bxd5 27.Bxd4 Rxe1 28.Rxe1 Ne6 29.Bxg7 The mass exchanges have resulted in a level endgame, and a draw was agreed in Vaibhav – Kunte, New Delhi 2012.  

15...Bg7! As mentioned earlier, there are certain lines in this system where Black can benefit from keeping his bishop on f8 for

longer. In the present position, however, I believe that putting it on g7 is the most constructive move available.   Kaufman and Mikhalevski have both suggested 15...c6?! intending ...Nh5, but I would argue that this move is a concession when White has not provoked it with Bd3. If you wish to see concrete evidence, 16.b4! Nb6 17.c4! exd4 18.cxb5 axb5 19.a5 Nbd7 20.Nxd4± is clearly better for White, as demonstrated by a couple of correspondence games.   A more serious alternative is: 15...Nh5!? This is an extremely rare continuation, but it was recommended by Eljanov. Black’s main idea is to play ...exd4 followed by ...d5, intending to meet e4-e5 with ...c5. I spent quite a lot of time analysing this, and will just give a brief summary of my analysis to serve as a basis for future research: 16.g3 16.c4 b4! is mentioned by Eljanov; Black is doing well, and the position can be compared with the Leko – Wells game as referenced a little further. 16.Bd3 was seen in Hazai – Lukacs, Hungary 2001, when Black missed the chance to demonstrate one of the main ideas connected with his last move: 16...exd4!N 17.cxd4 d5 18.e5 Nf4 19.Bf1 b4! 20.g3 Ne6 and Black follows up with ...c5. 16.d5 c6 17.c4 b4 (17...bxc4 18.bxc4 Qc7 followed by ...a5 is a viable alternative) 18.a5 Nc5 and Black is fine, as the pressure against b3 and e4 stops White from building up on the kingside. Finally, 16.Nf1N is sensible, but 16...exd4 17.cxd4 Bg7! 18.Qd2 (18.Qc1!? b4!) 18...Nc5! 19.g4 Qf6! 20.N1h2 Nxe4! turns out okay for Black.

16...exd4! 16...c6!? could also be investigated – the move is not so bad with ...Nh5 and g2-g3 included. I like the directness of the text move though. 17.cxd4 d5! 18.e5 bxa4! 19.bxa4 Ng7! I analysed this out to eventual equality. My overall conclusion is that 15...Nh5!? is a decent move, but it gives White a lot of options and the route to equality is far from simple. Another slightly annoying point is that White can alter his move order by playing an earlier Bd3 to provoke ...c6. True, doing this has some minor disadvantages for White, but it also takes away some of Black’s possibilities, such as the improvement over the Hazai – Lukacs game.

16.Bd3 16.d5 c6 17.c4 b4! 18.a5 was played in Tseshkovsky – Timoscenko, Tashkent 1987, at which point Black closed the position with 18...c5. I think it would have been more prudent to play 18...Nc5!N, when the pressure on the b3- and e4pawns makes it hard for White to transfer his pieces to the kingside. Most engines already prefer Black’s position at this point.   16.b4N This move has not been played in this exact position, but it transposes to an existing game. In any case, it is worth comparing it to the 15...c6?! line mentioned above; here Black gets a good game precisely because the b7-bishop has not been blocked in. 16...d5! 16...exd4 17.cxd4 d5? is not the same thing due to 18.e5!±.

17.Nxe5 After 17.dxe5 Nxe4 Black picks up the e5-pawn with fine play. 17...Nxe5!N 17...Nxe4 was Black’s choice in Grigoriants – Inarkiev, Irkutsk 2010, but the text move is more convincing. 18.dxe5 Nxe4 19.f4 This is the only challenging move, but Black is well placed to meet it.

19...Nxd2! 20.Qxd2 f6 Black has no problems, as 21.e6 can be met by 21...f5!.

16...c6 We have now converged on the same position as Sebag – Koneru.   17.Qc2 Rc8 18.Bf1!

As I mentioned previously, this is more accurate than Sebag’s 18.Rad1.   18...Nh5

19.dxe5! At the time of writing, this strong move has only been tested in this and one other correspondence game.   The fourteen-year old Peter Leko went for a thematic plan against GM Wells, but his execution was not perfect: 19.c4?! exd4 20.Nxd4 b4!

Well played! White will suffer on the dark squares and the young Hungarian prodigy is already in some trouble. The pawn structure resembles certain variations of the King’s Indian Defence where Black has a lot of activity in return for a backward d6-pawn, which is hard to attack anyway. The difference in the relative activity of the pieces makes this an especially favourable version for Black.

21.Rad1 c5 22.Ne2 Ne5 My analysis engine loves: 22...Bxb2N 23.Qxb2 f5! 24.exf5 gxf5µ Just look at White’s pieces! 23.f4 Nd3! 24.Qxd3 Bxb2 25.g4 Ng7 Black has ideas such as ...Qh4 and ...Ne6 in mind. 26.Qc2 Bf6 27.g5

27...Bd4†N Black played 27...Bxg5!? and got long-term compensation for the piece in Leko – Wells, Budapest 1993, but he eventually went wrong and lost. Although the sacrifice is interesting, it is not necessary and the text move is just a better continuation overall. 28.Nxd4 cxd4 White has overextended himself on the kingside, and Black can target the weak e-pawn with ...Ne6-c5.   19...dxe5! It is sometimes useful to exchange a pair of knights, but 19...Nxe5 20.Nxe5 dxe5 (20...Bxe5 21.Rad1²) leads to problems after:  

21.b4!N White intends Nb3, when Black may come to miss his queenside knight. (21.c4 was a reasonable move in Shields – Di Peri, corr. 2012, but the text is even stronger.)

20.Ba3?! This was played after long consideration, so I guess my opponent was trying to figure out what would happen in the complications after 20.b4!. I think his chosen move was a mistake, as it allowed me to put my knight on f4 without sweating. 20.b4!N will be studied separately after this game.   20...Nf4! Now Black is fine; the knight is heading for e6, and he is ready to meet c3-c4 with the positional pawn sacrifice mentioned in the notes to the previous game.  

21.Rab1 In the event of 21.c4 b4! 22.Bxb4 c5 Black follows up with ...Ne6-d4, with great compensation.   21...Ne6 22.Red1 Qc7 23.Qd3 Rcd8 24.Qe3 Even though the position is equal, it still contains a great deal of complexity. This is one reason why the Spanish remains so popular at all levels: both sides can play for a win, even at correspondence level. I could criticize my opponent’s moves on the grounds that they allowed me to execute a nice plan which put him under a lot of pressure, but the position is so complex that these things can happen, even with the help of strong analytical engines. I think that it is better not to overload you with trees of variations, but simply to present the rest of the game with light annotations as an example of how Black can get the upper hand in such positions.

24...Ba8 25.Ra1 Qb7 26.Bb4 White’s idea is to put the bishop on a5 in order to take control over the d-file. He avoids exchanging on b5 for the moment, as he can do that under his own terms when it suits him. This all sounds good, but Black is perfectly placed to launch a counteroffensive on the kingside.   26...c5 27.Ba5 Rb8 28.g3

28...f5!ƒ 29.axb5 axb5 30.Bg2 f4 31.Qd3 Ndf8 32.h4 Qe7 33.Qe2 Bc6 34.b4 c4 35.g4 h6 36.Ra2 Qb7 37.Re1 Re7 At this point I hit upon the idea of sacrificing my knight with ...Nd4. However, since I have all the time in the world, I can prepare it by doubling on the closed e-file.

38.Raa1 Rbe8 39.Kf1 Nh7 40.Rab1 Rd7! I analysed the ...Nd4 sacrifice but found that White was could return the piece with Nxc4 in many lines, so I decided to go for the c3-pawn after all.   41.Kg1 Rd3 42.Rbc1 g5! This is the last nail in White’s coffin. With the g4-weakness immobilized, Black can target it with ...Qc8 and ...Bd7.

43.hxg5 Nexg5 44.Nxg5 Nxg5 45.Rc2 Qf7! 46.f3 h5 47.gxh5 Qxh5 48.Bb6 Bd7 49.Kf2 Ra8 50.Qf1 Ra6 51.Bd8 Qh4† 52.Kg1 Rg6 White’s pieces are almost paralysed and his king has no escape, so the end is near.

53.Bxg5 Rxg5 54.Qf2 Qh6 55.Re2 Bh3 56.Qc5 Qg6 57.Qf2 Rxg2† 58.Qxg2 Bxg2 59.Rxg2 Qb6† 60.Rf2 Bf6 61.Kf1 Bh4 62.Rg2† Bg3 63.Re2 Rd7 Having played on for a few moves while a queen for a rook down, White decided to call it a day with checkmate coming. 0–1

The Super-Critical 20.b4! 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.b3 g6 15.Bb2 Bg7 16.Bd3 c6 17.Qc2 Rc8 18.Bf1 Nh5 19.dxe5! dxe5! 20.b4!N As mentioned in the notes to the last game, this is the most challenging way for White to handle the position.

20...Bf8! Both 20...Nf4 and 20...Nb6 fall short of equality (trust me!) so the text move is the only one that remains. The idea is to stop c3-c4 by putting pressure on b4, which explains the next few moves as well.   21.Qb3 21.Nb3 Nb6 is fine for Black.   21...Qe7 22.Ba3! White can also insert 22.axb5 cxb5 (22...axb5?! 23.Ra7! gains a tempo on the b7-bishop in order to double rooks on the a-file) and only now play 23.Ba3. This move order trick was pointed out to me by my friend Nikos Sarakenidis (a strong correspondence player himself!) after studying my game against Istomin. However, Black has a nice way to meet it:

23...Bc6! (23...Qe6!? is another idea, but it is not so convincing after 24.c4!?) With the text move Black utilizes the newly vacated c6-square and prepares ...Nb8 in order to guard the a6-pawn and later reposition the bishop to e6 via d7. There are many possible lines but – trust me (I checked!) – Black has nothing to fear if he follows this plan.

22...Nf4! This is an absolutely crucial moment.   22...Rb8 is initially one of the top engine choices, but it is a weird move to play, and my analysis indicates that it falls short of equality.   The most important alternative is: 22...Qe6 I spent a lot of effort analysing this move, but eventually had to discard it due to the following continuation.

23.c4! 23.axb5 Qxb3 24.Nxb3 cxb5 25.Na5 Bc6 (25...Ba8?! 26.c4) 26.Nxc6 Rxc6 is evaluated by the engines as better for White, but if Black plays his d7-knight to b8 and the h5-knight to d7, he has nothing to fear. (I could go into much more detail about this, but I do not want to get too sidetracked as this is not where the real problem lies.) 23...bxa4 24.Qxa4 c5 25.b5! This idea was found by GM Sune Berg Hansen during a training camp with the Danish national team prior to the European Team Championship in Reykjavik. The position is extremely complicated but, alas, favourable for White. One direction which I analysed quite deeply is:

25...Ra8 26.Qd1 Nf4 27.Bb2 f6 Things do not look so bad for Black on the surface; given time, he will move his queen and bring his knight to e6 and d4. Unfortunately, White has the following annoying plan: 28.Re3! Followed by doubling rooks on the a-file. The longer you analyse with an engine, the more the evaluation creeps up in White’s favour.   Even though the immediate 22...Qe6 does not quite work, the idea should be remembered, as Black might still play it on the next move under more favourable circumstances. Therefore we will focus on the following attempt by White to force the play:

23.axb5! This time 23.c4?! gives White a significantly worse version of the previous note: 23...bxa4 24.Qxa4 c5 25.b5 Ra8 26.Qd1 Ne6 The knight is ready to land on d4, so Black stands better.   23...cxb5 24.c4 a5! It is vital to hit back to avoid being steamrolled on the queenside.

25.Rab1 25.bxa5 b4 26.Bb2 Ba6 gives Black excellent positional compensation. A sample line is 27.Qc2 Red8 28.Red1 f6 29.Nh2 h5!„ followed by ...Ne6 and ...Ndc5.   25...a4 26.Qb2

I also analysed the alternative queen move: 26.Qe3 Bh6 I took this move as my main line as it seems like the most natural choice, but 26...f5!? and 26...Rc7!? also lead to rich, complex positions. 27.Qa7 27.Qc3 bxc4 28.Nxc4 Qf6 is at least equal for Black.

27...Nf6! 28.cxb5 28.c5?! gives Black a pleasant choice between 28...Ne6 and 28...Ra8 29.Qb6 Reb8, both of which are annoying for White. 28...Rc3! 28...Rc2?! 29.Rbd1 favours White. 29.Bb2 29.Bc1 Ra8 30.Qb6 Rc7 threatens ...N4d5!. (30...Bd5 is offered initially by the engines; after 31.exd5 N4xd5 32.Qa5 Rxa5 33.bxa5 the position is unclear.) 31.Nxe5 Qxe5 32.Bb2 Qg5 Black has enough counterplay to draw, but nothing more. My main line continues:  

33.Nf3 (33.Qxc7? Nxh3† 34.Kh1 Nxf2† 35.Kg1 Nh3† 36.Kh1 Ng4 gives Black a dangerous attack) 33...Nxh3† 34.Kh1 Nxf2† 35.Kg1 Nh3†=

29...Rc2 30.Qxa4 Rxd2! 31.Nxd2 Nxh3† 32.gxh3 Bxd2 Black is an exchange down, but White’s king is exposed and ...Nxe4 is a nasty threat. My analysis continues:

33.Bg2 Nh5! 34.Re2 Nf4 35.Rxd2 Qg5 36.f3 Nxh3† 37.Kh1 Qxd2 38.Bxh3 Bxe4 38...Qh6 also leads to a draw.

39.fxe4 Qd3 40.Rf1 Qxh3† 41.Kg1 Qe3†= With a perpetual.   With the queen on b2 there are many ways Black can play the position, but I like the following direct idea.

26...g5! Black launches a kingside attack, exploiting the distant location of the enemy queen.   27.Rbd1 27.c5 Bc6 is about equal, although White should be slightly careful to avoid trouble on the kingside. 28.g3 (28.Rbd1 Qe6! threatens ...g4, and after 29.Nxg5 Qg6 the bishop on c6 proves its worth, as the d7-knight is defended. 30.h4 h6 Black regains his pawn with the better game.) 28...Ne6 29.Rbd1 Bg7

Black is fine; his ideas include ...Ndf8, ...Rcd8 and, at a suitable moment, ...Nd4.   27...g4! 28.hxg4 Qe6 29.g3 Qxg4 30.cxb5

30...Re6 30...Nh3†!? 31.Bxh3 Qxh3 32.Rc1 Qe6 also gives Black adequate compensation according to my analysis.   The text move gives Black significant attacking chances. Obviously we are well past the point of normal opening preparation, but I will give the rest of my analysis to illustrate some of the resources available to both sides.   31.Rc1 Rh6 32.Nh2 Nh3† 33.Bxh3 Qxh3 34.Ndf3 Rh5 35.Rxc8 Bxc8 36.g4 Rh6 37.Qc2

37...Nf6! 37...Nb6!? 38.Bc1 Rg6 39.Nxe5 Bd6 40.Qd3 Qxd3 41.Nxd3 Bd7 leads to an unclear endgame. Despite the initial pessimism of the engines, I found that Black’s active pieces and passed a-pawn give him enough counterplay to draw. The text move is much simpler though.  

38.Qxc8 Rg6 39.Nxe5 Rh6 40.Nef3 Rg6= Neither side has a convincing way to avoid the repetition.   Conclusion   Our tabiya for this chapter occurs after the opening moves 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8. We analysed two main branches, one of which was popular forty years ago, whereas the other rose to prominence much more recently. Both options share the common plan of putting pressure on Black’s queenside, especially the b5-pawn, which is a target in many variations of the Spanish complex.   The first branch was 13.b4 Bf8 14.a4 Nb6 15.a5 Nbd7 16.Bb2, when Black needs to remember to play 16...Rb8! to restrain White’s queenside ambitions. Then 17.Qb1 Nh5 worked well for Black in our first illustrative game, so 17.Rb1 Ba8 18.Ba1 g6 is more critical. After 19.Bd3 we saw Smejkal beautifully exploit the position of the bishop by arranging a timely ...c5-c4. White’s most logical continuation is 19.c4, but Karpov showed us how Black can get a good game by exchanging on c4 and later planting his bishop on b5.   13.a4 Bf8 is the other main option, where White keeps things more flexible on the queenside. Our first two illustrative games covered 14.Bd3 c6 followed by 15.b4, 15.Qc2 and 15.Nf1, none of which is too terrifying.   White’s most challenging set-up involves b2-b3 at some point. 14.Bd3 c6 15.b3 g6 16.Bb2 is one possible move order but, as I mentioned in the notes to Sebag – Koneru, Black can take the opportunity to play 16...Nh5! to get some extra mileage out of the bishop on f8. That is why I consider 14.b3! g6 15.Bb2 Bg7 16.Bd3 c6 17.Qc2 Rc8! to be White’s most accurate move order to reach the ‘target position’ described on page 319. From there, 18.Rad1 was dealt with handily by Koneru in Game 59, although you need to remember some key concepts such as the positional pawn sacrifice with ...b4!!, as described on page 328.   The most critical continuation is 18.Bf1 Nh5 19.dxe5 dxe5. Then 20.Ba3?! allowed me to get a pleasant position and eventually a nice win in Game 60, but 20.b4!N is the really critical idea which awaits a practical test. It took an enormous amount of work to determine what was really going on in the complex variations that follow. I may have gone a bit too far with my analysis, and you should not even think about trying to memorize all of it – I just included the long lines to demonstrate that Black’s position can ultimately withstand whatever White may throw at it.

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8   14.Bg5 Game 61, page 343   14.Ng3 g6 15.Bg5 Game 62, page 345 (15.Bd2 page 345) 15.b3 Bg7! page 349 16.a4?! Game 63, page 349 16.d5! Rc8! 17.c4 c6 18.Be3 Game 64, page 354 18.Bd2 Game 65, page 358 The Sophisticated 18.Bg5 page 360 15.a4 page 362 15...c5! 16.d5! (16.b3 page 363; 16.b4 page 363; 16.Be3 page 363) 16...c4! 17.Be3 Game 66, page 362 17.Bg5! Game 67, page 368   

Diagram Preview On this page you will find eight diagrams with critical moments from the coming chapter. I recommend that you take up

to ten minutes to think about each of them (though much less in some cases). The solutions are found in the following chapter.   Black is to move unless otherwise indicated.  

How should Black address the pin

on the f6-knight? (page 343)  

How should Black proceed? (The pawn

being on h6 instead of h7 can make a difference!) (page 361)  

Find a suitable way to counter

White’s queenside play. (page 348/9)  

Find a strong regrouping plan for Black. (page 364)  

White has just played 16.a4?!.

Can you find Spassky’s strong retort? (page 350)  

White has played 18.Qe2 (instead of 18.Qd2).

How does Black react? (page 371)  

How should Black improve his position? (page 373)

How should Black fight against

White’s strong pawn centre? (page 356)   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8

14.Ng3 Of course this is the expected move, but 14.Bg5 had a brief surge in popularity, at least until it became clear that 14...h6 15.Bh4 Qc8! was a good solution; see Game 61 for more details.   14...g6 From this point there are three main lines for us to study.   15.Bg5 is a plan devised by Karpov, who is also a Breyer expert with the black pieces. This move became popular again quite recently, but Black’s best continuation has been worked out and can be found in Game 62.   15.b3 introduces the idea of d4-d5 followed by c3-c4, while also remaining flexible enough to choose another plan. This is examined in three illustrative games and a final analysis section; I would especially like to draw your attention to the Fenwick – Andriuschenko encounter and the following analysis section, as they cover the most theoretically critical lines in the 15.b3 complex. 15.a4 is the traditional Breyer main line. I had to withstand some pressure in one of my correspondence games in order to fully appreciate the optimal way for Black to handle the position. This encounter can be found in the notes to Game 67, with the main game featuring a model demonstration of Black’s chances by GM Giri.

14.Bg5 GAME 61 Leinier Dominguez Perez – Sergey Karjakin Tashkent 2012   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.h3 d6 9.c3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Bg5 h6 15.Bh4

15...Qc8! Black should resist any urge to make ...g5 work. The text move unpins the knight and threatens to win the e4-pawn after ...exd4.   16.dxe5 Protecting e4 with moves like 16.Qb1 or 16.Nd2 can be met by a strong ...c5 break, either at once or after exchanging on d4.   16...dxe5 17.N3h2 17.N3d2 was met by the same plan of 17...a5 18.Ne3 Ra6 in Safarli – Amin, Tromso 2013.   17...a5! 18.Qf3 Ra6 This clever developing scheme promised Black a good game. The rook is useful on the 6th rank and the queen will go to a8.

19.Ng4 19.Ne3 Qa8 20.Rad1 Nc5 21.Bxf6 Rxf6 22.Nf5 b4„ was a similar story in Navara – Avrukh, Sibenik 2012.   19...Nxg4 20.Qxg4 Nc5 21.Qf3 21.Qxc8 would have been safer and equal, but also an admission that White has nothing.   21...Qa8 22.Rad1 b4! White has no real attacking chances, and Karjakin goes on to prove that Black has significant counterplay on the queenside.   23.Nd2 Rd6 This makes room for the queen or the bishop to go to a6 if and when needed.   24.Qe3 Ne6 25.Nc4 Rxd1 26.Rxd1 bxc3 27.bxc3 f6 28.Rb1 a4! The pawn is heading for a3 in order to fix the weakness on a2. Both of White’s queenside pawns are more vulnerable than their black counterparts, and a few more inaccuracies are all it takes for White’s position to become hopeless.

29.Kh1 Ba6 30.Nb2 a3 31.Nd3 Bc4 32.Bb3 This is a mistake according to the analysis engines, but I don’t see much point in analysing the ‘best’ move 32.Nc1, which is as pathetic as it looks.   32...Bxb3 33.Rxb3 Rd8 34.Rb1 Qa4 35.Nb4

35...Nf4 36.Bg3 c5 37.Nd5 Nxd5 38.exd5 Qc2 By now it is clear that a2 is a goner, and Black’s a-pawn will decide the game.   39.Rb7 Qxa2 40.Ra7 Qb1† 0–1   Black’s play in the above game was smooth and relatively easy to understand. Now we will turn our attention to the

14.Ng3 variations, starting with a game where White develops his queen’s bishop before doing anything on the queenside.

15.Bg5 GAME 62 Antonios Pavlidis – Levente Vajda Pardubice (rapid) 2014   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.c3 d6 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.Bg5 This is another big main line of the Breyer. White’s idea is to provoke a slight weakening of the black kingside. If Black leaves the pawn on h7 then White will play Qd2, when the g5-bishop becomes annoying and White can choose the right time to play Bh6.   15.Bd2 Bg7 16.Qc1 is another way to go for the same plan, but this way Black feels no pressure on his f6-knight, which enables him to play 16...d5! with full equality. This was demonstrated long ago in games such as Unzicker – Portisch, Santa Monica 1966, and Balashov – Spassky, Sochi 1973.  

15...h6 Black had better play this, otherwise Qd2 will establish an unpleasant bind on the kingside.   16.Bd2 Now Qc1 will come with tempo, so Black will not have time to play an effective ...d5 like he did in the note to White’s previous move.   16...Bg7! After 16...exd4!? 17.cxd4 c5 “Black’s counterplay is much more lively” according to Kasparov, but 18.d5 Nb6 19.Ba5 Nfd7 20.b3 Bg7 21.Rc1 Qf6 was slightly better for White in Fischer – Spassky, St Stefan/Belgrade (5) 1992.

Black can try to improve by postponing ...Nb6 (Eljanov, for example, gives 18...Bg7 as Black’s best), but overall I believe the main line presented here to be the most accurate defence.

17.a4! Karpov introduced this move in 1971, in a training match against Korchnoi. It turns out that this is the optimal moment to begin operations on the queenside.   17.Qc1 Kh7 18.h4 has the potential to lead to tremendous complications after 18...d5!?; see Smirin – Beliavsky, Odessa 1989, for a nice example involving a stunning tactical combination by Black. However, a much simpler route to a good position is:

18...exd4! 19.cxd4 (19.h5 c5!) 19...c5 20.d5 Ng4 21.h5 Nde5³ Tunyik – Andriuschenko, corr. 2011.   In his instructive book The Ruy Lopez: Move by Move, GM McDonald gives the following game by Adams as a model

example showing what can go wrong for White: 17.dxe5?! dxe5 18.Be3 Qe7! 19.Nd2 h5! 20.Ne2 Red8 21.Nc1 Nc5 22.f3 Ne6 23.Qe2 c5

24.Ncb3 Rac8! 25.Na5 Ba8 26.a4 c4 27.Qf2 Nd7! 28.axb5 axb5 29.Reb1 Ndc5 30.b3 and now 30...Nd4!?µ was seen in Noa. Davies – Adams, Edmonton 2009, the last move echoing the classic Fischer – Kholmov, Havana 1965, although 30...cxb3N followed by ...Na4 was also extremely strong.

17...Nb6! Korchnoi’s move stops Bd3 and tries to provoke the positionally undesirable a4-a5.   It is worth taking the time to see why the following alternative does not work as well: 17...c5 This is a typical reaction to a2-a4; see for instance the traditional main line as covered in Games 66 and 67. The reasoning is that, with the pawn on a2, White can always meet ...c5 with d4-d5, when Black has little counterplay

as he cannot challenge the d5-pawn with ...c6. With the pawn on a3 or a4, the situation is different as Black can play ...c4, establishing an outpost on c5 which can be occupied by his knight. 18.d5 c4 19.b4! Despite the logic behind Black’s play, this move leads to some pressure for White in all variations.  

19...cxb3 In Fischer – Spassky, St Stefan/Belgrade (1) 1992, Black left the queenside structure intact and Fischer went on to win an excellent game. In subsequent games Black has learned the lesson from Spassky’s defeat, and I cannot describe the situation any better than Mikhalevski on Chesspublishing, who comments: “Nowadays this is an automatic reaction to 19.b4. If Black doesn’t capture the b-pawn White will double, or even triple, his pieces along the a-file and threaten to open the a-file at the appropriate moment by means of axb5.” 20.Bxb3

White intends to open things up with c3-c4, so Black has two main ideas:   a) 20...Qc7 is recommended by Kaufman as “an excellent untried alternative to avoid 21.c4.” Kaufman’s book was published in 2012, and I see at least a dozen correspondence games played before that, so the move has been tried plenty of times; what’s more, it doesn’t stop White’s plan at all: 21.c4! bxc4 22.Rc1 Nc5 23.Bxc4 Reb8 24.Bb4 Nfd7

25.h4! Provoking a kingside weakening before switching back to the other flank. 25...h5 26.Ng5 Qd8 27.Re3 Bc8 28.Ba3 Rb7 29.Ba2 Rab8 30.Nf1 Nf8 31.Rec3 Black was unable to keep his position together in Romanov – Wegelin, corr. 2012.   b) 20...Nc5 is more reliable, albeit not particularly inspiring for Black: 21.c4 bxa4 (21...Qd7 is the modern main line, but I don’t think Black has solved all his problems here either) 22.Bxa4

22...Rf8!N The exclamation mark is mine, but the move is suggested by Kasparov. 23.Be3 Qc7 “...followed by ...Rfb8, ...Nfd7 and, at a convenient moment, ...Bc8”, according to the former World Champion. Also ...Bf6-d8 can be

considered. Black is a bit worse but he should be able to get half a point with a bit of care.

18.b3 This is the most topical continuation, but Black seems to have found the correct way to counter it.   18.a5?! is met by 18...Nc4 19.Bc1 d5!³ as in Kruszynski – Manasterski, Gdynia 1973.   18.Qc1 Kh7 19.axb5 axb5 20.b3

The engines offer this sequence as White’s best. However, this plan does not seem to be considered particularly strong by modern grandmasters as it hasn’t become popular since its first appearance in Karpov – Korchnoi, Leningrad 1971 (in which the moves Qc1 and ...Kh7 had not been played). A good reply is: 20...Nfd7N

A novelty here, although Korchnoi used the same idea in the similar position against Karpov. 20...c5 21.d5 Qc7 was also decent in Zacharov – Perez, Internet 2010. 21.Bd3 Rxa1 22.Qxa1 d5! This seems best, although 22...b4!? is also sufficient for equality, for instance: 23.d5 bxc3 24.Bxc3 c6 25.dxc6 Bxc6 26.b4 d5=

23.Bxb5 exd4 24.Nxd4 dxe4 25.Qa7 Against most other moves Black will strengthen his position with ...f5. 25...Qc8 Black has no problems whatsoever.

18...bxa4 19.bxa4 a5! 20.Bd3 Ba6 This is a simple way to equalize.  

20...Qc8!? is more ambitious. The main idea is 21.d5 c6 22.Rb1 Nbd7!, showing why the queen went to c8 rather than to d7.

23.c4 Nc5 24.Be3 Nfd7 25.Bf1 Qc7 In Ni Hua – Torre, Chiva Guanzhou 2010, Black followed with ...Ba6 with a fine position.   21.Bxa6 Rxa6 22.Qb3 22.d5 is met by 22...c6 23.Qe2 Qa8 24.c4 Nbd7 followed by ...Rb8.

22...Nbd7 The players agreed a draw here, which seems fair, albeit premature. ½–½

15.b3 This is one of White’s most important and topical variations against the Breyer. We will start with a substandard plan involving premature queenside activity from White, before working up to the more critical lines.   GAME 63 Robert Byrne – Boris Spassky Candidates Match, San Juan 1974   This fantastic game has been annotated by such outstanding players as Euwe, Kasparov and of course Spassky himself. That is why I am not going to examine every aspect of it as deeply as previous annotators did, but will instead emphasize the parts which are most relevant to our opening discussion.   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.b3

This is Unzicker’s move, which was taken up by Judit Polgar in the 1990s, and later by a bunch of other strong grandmasters. The idea behind it is simple: White wants to play d4-d5 to nullify the bishop on b7, but Black will almost always respond with ...c6, so White prepares to reinforce the central pawn wedge with c3-c4. Polgar writes quite a lot about her favourite way to deal with the Breyer in From GM to Top Ten, the second volume in her trilogy.   15...Bg7! 15...c5? 16.d5± will be torture for Black, who has no counterplay at all. Black needs to be able to meet the d4-d5 push with ...c6.   15...c6?! is passive according to Kasparov and I agree. True, it has been tested by some top players, who obviously think there is something to be said for preventing the d4-d5 advance. However, I prefer not to obstruct the b7-bishop

when I am not forced to, especially when there are other useful moves to play.   15...d5?! is met by 16.Bg5! h6 17.Bh4. Lev Psakhis attributes White’s idea to Klovans, but it only became popular after the Israeli grandmaster’s analytical refinements. The critical continuation is 17...g5 18.Nxg5! hxg5 19.Bxg5, when Black’s best seems to be:

19...exd4! 20.e5 Rxe5 (20...Nxe5? runs into 21.Nh5) 21.Rxe5 Nxe5 22.cxd4 Nc6! (22...Ng6? 23.Qf3!+–) 23.Qd3!? Nb4 24.Qf3 Be7

From this position 25.Bb1 led to an eventual draw in the high-profile game Grischuk – Mamedyarov, Moscow 2006, but 25.Bf5! is more dangerous and has achieved a near-perfect score in my database. Maybe deep analysis will reveal that Black can survive, but for practical purposes my advice would be to steer well clear of this. It is worth mentioning that Eljanov’s suggestion of 15...a5!? could be a viable alternative, but I see no reason for Black to deviate from Spassky’s choice.  

16.a4?! This is inconsistent with White’s plan, and gives Black the chance for an improved version of the familiar central break.   16.Bg5 would have been a good move after 15...c6?!, and it has been played several times in that position by Parimarjan Negi, among others. Here, however, it is ineffective due to 16...h6 17.Bd2 d5!, showing perfectly why it is worth giving the b7-bishop a clear view.   16.d5! is best, and we will examine it in the next game.   16...d5! 17.dxe5 17.Nxe5 is the alternative, but after 17...Nxe5 18.dxe5 Nxe4 19.Nxe4 dxe4 Black has no problems. Let’s continue this line a bit further: 20.Qxd8 Raxd8 21.Bg5

Kasparov gives 21...Rd5 and concludes that the position is equal. 21...Rb8!? could be a slight improvement, when I would take Black’s position if given the choice, although 22.Bf6 Bxf6 23.exf6 Re6 still leaves the position close to equal.   17.Bg5 was a strong move against 15...d5?!, but it has never been tried in this position. It is obvious that the piece sacrifice should not work with Black having played a useful move (...Bg7) and White an irrelevant one (a2-a4) but chess is a concrete game, so let’s see exactly where the difference lies: 17...h6 18.Bh4 g5 19.Nxg5 hxg5 20.Bxg5 exd4 21.e5

21...Rxe5 (21...Nxe5 is also strong, as 22.Nh5 doesn’t threaten the knight, which gives Black time to play 22...Qd6) 22.Rxe5 Nxe5 23.cxd4 Ng6 The fact that the knight on f6 is securely defended means that White is unable to develop much of an initiative, so his compensation for the piece is questionable to say the least.   17...Nxe4

18.Bxe4 18.Nxe4 seems more natural but after 18...dxe4 19.Bxe4 Bxe4 20.Rxe4 Nxe5 White faces some problems, for example: 21.Qxd8 Raxd8 22.Bg5 (22.Nxe5? is met by 22...Rxe5! 23.Rxe5 Rd1† 24.Kh2 Bxe5† 25.g3 Bxc3 26.Rb1 Bd2 and Black wins) 22...Nxf3† 23.gxf3

In Sanchez – Vishnu, Montcada 2014, the most accurate continuation would have been 23...Ra8!?N to make it harder for White to use the a-file. The endgame should be a draw of course, but Black can keep playing with no risk thanks to his better structure.   18...dxe4 19.Bg5 Byrne was relying on this intermediate move, but he probably did not anticipate Spassky’s reply.

19...exf3!? Sacrificing the queen!   19...Qc8 leads to equality according to Kasparov’s analysis.   Several commentators, including Kasparov, have pointed out that 19...f6!N is objectively best. 20.exf6 Nxf6 21.Nd2

(21.Qxd8?! Raxd8 22.Nd4 is equal according to Kasparov, but my electronic analysis partner screams out that 22...b4! clearly favours Black)

21...e3! (This seems stronger than Kasparov’s 21...Qd6!?) 22.Bxe3 Nd5³/= White still has to work to demonstrate full equality against the strong bishops.   20.Bxd8 Raxd8

21.axb5?! 21.Nf1 was suggested by Spassky, among others, as White’s best defence. 21...Nxe5 22.Qc2 fxg2 23.Nh2 Here the engines like 23...b4!³ in order to open the long diagonal. Spassky’s 23...g5!?÷ also leads to fascinating play; Black intends ...Ng6-f4, threatening to deliver mate on h3!   Bondarevsky’s 21.e6! seems best. My main line continues: 21...Ne5! 22.exf7† Kxf7 23.Qc1 fxg2 24.Qf4† Kg8

25.Ne4 (25.Red1 Nf3† [25...Rc8!?] 26.Kxg2 Nd2† is a draw) 25...Rf8 26.Nf6† Bxf6 27.Rxe5 Bxe5 28.Qxe5 Rd3!= It should be a draw, although mistakes could still happen on either side.   21...Nxe5

22.bxa6? This further mistake soon leads to a hopeless position.   22.Qc2 is better according to several analysts, but I am not sure if it can save White: 22...fxg2 23.Rxe5 Rxe5 24.bxa6 Bf3 Black has the terrible threat of ...Rde8 and ...Re1†. 25.a7 (25.Kh2 h5 26.h4 Bf6 is also difficult for White) 25...Rde8

26.a8=Q (26.Qd2 Bh6! is a nice tactic pointed out by Euwe and Median) At this point in the game the g2-pawn is worth much more than an exchange, so Black should play 26...Rxa8! 27.Rxa8† Bxa8 with a winning position.   Other options do exist for White, like 22.Qc1 for example, but the general conclusion is the same: White is struggling to defend.   22...Rxd1 23.Rexd1 Ba8 24.gxf3 Nxf3† 25.Kf1 Bxc3 White’s rook is no match for Black’s powerful pair of bishops. Spassky provides a textbook example of how to convert this type of advantage.

26.Rac1 Nd2† 27.Kg1 Ba5 28.b4 Nf3† 29.Kf1 Nh2† 30.Kg1 Nf3† 31.Kf1 Bb6 32.Rc2 Nh2† 33.Kg1 Nf3† 34.Kf1 Kf8 35.Ne2 Nh2† 36.Kg1 Nf3† 37.Kf1 Nh4 38.Nf4 Bf3 39.Rd3 g5 40.Ne2 Bg2† 41.Ke1 Nf3† 42.Kd1 Ne5 43.Rdc3 Bd5 44.Rd2 Bc4 45.Ra3 Ra8 46.f4 gxf4 47.Nxf4 Rxa6 48.Rxa6 Bxa6 49.Nd5 Bc4 50.Nxb6 cxb6 51.Rd6 b5

52.Kd2 Ng6 53.Ke3 h5 54.Rd7 h4 Black is in complete control and has just fixed the h3-pawn as a weakness, so White decided he’d had enough. 0–1   The next game shows a better way for White to handle the 15.b3 variation.   GAME 64 Mark Tseitlin – Boris Avrukh Haifa 2008   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.b3 Bg7 16.d5! This is more challenging than 16.a4?!, and it may easily cause Black problems if he doesn’t know the correct reaction. The first step is to prepare ...c6 without losing the d6-pawn.

16...Rc8! This is the modern trend. Obviously Black is preparing ...c6, but he should also think about what will happen after c3c4. One potential scenario involves ...bxc4 leading to an open b-file, so it might seem tempting to avoid committing the rook to c8 in case it is needed on b8 in the future. However, concrete analysis shows that the text move is the most accurate.   16...Nb6 is the other obvious way to prepare ...c6, but 17.Rb1! (17.Be3 Rc8 18.Qe2 c6 19.c4 transposes to the main game) creates an unpleasant X-ray effect on Black’s minor pieces on the b-file. Unzicker introduced this move in 1960 but it did not become popular until the 90s. An instructive example continued: 17...c6 18.c4 Qc7 19.Bd3 Nfd7 20.Be3 Nc5 21.Bf1 bxc4 22.bxc4 Nbd7 23.Qd2 Rab8

Black appears to have defended well on the queenside, but 24.Bh6! gave him problems on the other wing in Leko – Beliavsky, Bled (ol) 2002. Ambitious readers are encouraged to search for and study the rest of this masterpiece.   16...Qe7 and 16...Bf8 have also been played, but neither move is as strong as our main continuation. Here is an especially instructive game after the latter option: 17.Bg5 h6 18.Be3 c6 19.c4 a5 20.Qd2 Kh7 21.Nh2 b4 22.Ng4 Nxg4 23.hxg4 Qh4

24.g5! c5 25.Nf1! f6 26.g3 Qh3 27.f3! White was able to cause serious problems with Re2-h2 in Polgar – Spassky, Budapest (2) 1993. Once again, we can see how dangerous it is for Black to close the queenside completely. For a detailed analysis of this game, I once again refer you to Polgar’s book.   17.c4 c6 18.Be3 18.Bd2 and 18.Bg5!? are important alternatives which are covered, respectively, in the next illustrative game and the

analysis section that follows it.

18...Nb6! Kaufman cuts off his analysis at this point with the ‘=’ sign, but there is a lot more to it than that.   19.Qe2 This move defends c4 and creates the threat of a2-a4, but it is hit by a powerful piece sacrifice. The same idea works fine for Black against most other sensible moves, with one major exception: 19.Rb1!? White indirectly defends c4 by creating an X-ray along the b-file. 19...Qc7! The piece sac with 19...cxd5 20.cxd5 Nbxd5 21.exd5 Nxd5 gives Black a hard time after 22.Bd2 Qd7 23.Be4!. The text move threatens to take on d5, while also preparing ...bxc4 which would force White to exchange on b6. 20.dxc6 20.Bd3 allows Black to revert to 20...cxd5 followed by the typical knight sacrifice on d5. 20...Bxc6!? 20...Qxc6 21.Bd3 bxc4 22.bxc4 Nbd7 has proved to be fine for Black in a couple of high-level correspondence games, but the positional pawn sacrifice also makes a lot of sense.

21.Bxb6 Qxb6 22.Qxd6 White wins a pawn but the loss of his bishop makes his dark squares chronically weak. 22...Rcd8 23.Qb4 Bf8 24.Qc3

24...a5!?N I found a few games in which Black played 24...Bc5, then closed the queenside with ...b4 and manoeuvred the knight to e6. However, the text move seems simpler to me. 25.Re2 Bb4 26.Qa1 Qc5 Black will take on c4 next, maintaining excellent compensation.

19...cxd5! 20.cxd5 Nbxd5! 21.exd5 Nxd5 Black has two mobile central pawns, supported by two wonderful fianchettoed bishops, as compensation for a knight. Many experts, including Nunn, Kasparov and Mikhalevski, are of the opinion that Black’s compensation is at least enough for dynamic equality, and that in a practical game it is White who faces more problems. As you can probably guess, I do not disagree with them!

22.Bg5 22.Be4?! Nc3 23.Bxb7 Nxe2† 24.Nxe2 d5! was excellent for Black in Kamsky – Van der Sterren, Wijk aan Zee 1994.   22.Qd2N was suggested by Lars Bo Hansen and Mikhalchishin, but the idea has not been tested. Indeed, after: 22...Nxe3! 23.fxe3 Bxf3 24.gxf3 e4 25.Bxe4

25...Qg5 Black enjoys a big initiative.   22.Rac1 is the most popular move at correspondence level, but Black has done fine in that arena too. Here is a good example: 22...Nxe3 23.fxe3 (23.Qxe3 d5 is also excellent for Black) White hopes that the change in his pawn structure will help to restrain Black’s central mass of foot soldiers. 23...d5 24.Bb1 Rxc1 25.Rxc1 e4 26.Nd4 Qg5

27.Qe1 f5 28.Rc7 f4 29.Rxb7 Bxd4 30.exd4 fxg3 31.Rc7 Qf4 32.Rc2 b4 The engines call this equal, but the reality is that White can only wait for Black to improve his position step by step, which is exactly what happened in the game we are following.

33.Kh1 Rf8 34.Kg1 Kg7 35.Qc1 Qf6 36.Rd2 Rf7 37.Qe1 Qf4 38.Rc2 g5 Black prevailed eventually in Sukhodolsky – Makovsky, email 2008.   The above lines are not intended for you to memorize of course; they are merely included as instructive examples which show the enduring nature of Black’s initiative.

22...Qc7 23.Bd3? White attempts a tactical solution, but it is quickly refuted. 23.Be4 is better, but still less than comfortable for White after 23...Nc3 24.Qd2 d5 25.Bc2 f5.   23...Nc3 24.Qe3 e4 25.Rac1

25...Qd7! Black can also get a big advantage with 25...f5 or even 25...h6!? 26.Bxh6 exf3, but Avrukh’s move is even stronger.   26.Bxe4 Bxe4 27.Nxe4 Rxe4 Black is a pawn up with the more active pieces.   28.Qb6 d5! 28...Ne2† 29.Kf1 Nxc1 30.Rxe4 keeps White in the game.

29.Kh2 This avoids the fork on e2, but it allows Black to simplify to a winning endgame. The rest of the game contains some inaccuracies on both sides, perhaps due to time trouble, but Avrukh eventually manages to bring home the full point.

  29...Qc7† 30.Qxc7 Rxc7 31.Rc2 f5 32.g3 d4 33.Kg2 b4 34.Rd1 h6 35.Bf4 Rc5 36.Rd3 g5 37.Bd6 Rb5 38.Rcd2 a5 39.Nxd4 Rxd4 40.Rxd4 Bxd4 41.Rxd4 Nxa2 42.Rc4 Nc3 43.Bc5 Kf7 44.Kf3 Ne4 45.Be3 Rb7 46.Rc1 a4 47.bxa4 b3 48.Bd4 b2 49.Bxb2 Rb3† 50.Ke2 Rxb2† 51.Ke3 Ke6 52.Rc6† Kd5 53.Rxh6 Rxf2 54.Ra6 f4† 0–1   The next game follows the same path for 17 moves, but then White goes for a sophisticated Bd2-e3 manoeuvre, which forces Black to modify his plans.     GAME 65 John-Paul Fenwick – Andriuschenko Correspondence 2010   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.c3 d6 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 A similar idea for White is seen after 13.b3 c6 14.c4 Bf8 15.d5 a5 16.Nf1, as in Wan Yunguo – Wen Yang, Jiangsu Xinghua 2015. However, this move order is inaccurate due to:

16...Nc5! And after something 17.N3d2 Rc8 Black has an improved version of the type of position we will see in the main game.   13...Bf8 14.Ng3 g6 15.b3 Bg7 16.d5 Rc8 17.c4 c6 18.Bd2 This move receives a ‘?!’ from Kaufman but it has been used by some world-class players, and I did not find it easy to work out a clear way for Black to equalize. White’s main idea is to avoid the piece sacrifice seen in the previous game.

18...a5! White’s idea is to meet 18...Nb6 with 19.Ba5!. Black’s last move avoids the potential pin, but this is not the only idea behind it. In fact Black would prefer to put his knight on c5 instead of b6, but he does not want to allow Be3xc5. The text move enables him to play ...Nc5 and meet Be3 with ...Nfd7, without allowing the unpleasant b3-b4.   19.Be3! White renders the ...Nc5 plan less attractive and threatens to win a pawn with dxc6.   19...Ba6! 19...Qc7 can also be considered, but the text move is Black’s easiest equalizer. Black simply reinforces the b5-pawn and puts some pressure on c4 as well.

20.Bd3 Kaufman only mentions 20.dxc6N 20...Rxc6 21.cxb5 Bxb5 22.Bd3 Bxd3 23.Qxd3 Nc5 24.Bxc5 Rxc5 25.Rad1 d5 26.exd5 Rxd5 when Black’s extra central pawn gives him the upper hand. This seems correct, but it is not the most challenging way for White to play.   The other alternative to consider is: 20.Nd2 Nc5! 21.Bxc5 dxc5 This is a case where the exchange on c5 is less troublesome, mainly because there is no pressure on e5. Black can take advantage of that with the following manoeuvre:

22.Qf3 (22.Rb1 Re7! 23.Qf3 Ne8 was also preferable for Black in Demetrio – Marczell, corr. 2012) 22...cxd5 23.exd5 A draw was agreed here in Isaev – Telepnev, corr. 2011, but Black could certainly have played for more with 23...bxc4 24.bxc4 Rf8! intending ...Ne8-d6 and ...f5. Of course if 25.Rxe5? Nd7 White loses an exchange.   20...bxc4 21.Bxc4 21.bxc4 would only have left the c4-pawn under pressure.   21...Bxc4 22.bxc4

22...cxd5 This move gives Black an equal game without many complications.   22...h5 23.Rc1 Rc7 was an interesting idea seen in Cheparinov – Adams, Zafra 2009, but it seems to me that 24.Qd3!?N 24...Qa8 25.Nd2² favours White.   23.cxd5 Nc5! This knight is heading for f4.   24.Nd2 Nd3 25.Rf1 h5 26.Rb1 White takes control over an open file while removing his rook from the long diagonal.   26...Nf4 27.Ne2 Bh6 28.Nxf4 Bxf4 29.Bxf4 exf4 Black has an excellent outpost on e5, and the pawn on h3 gives Black some chances to open lines of attack on the kingside.   30.f3 Nd7 31.Qa4 Ne5 Given a few moves, Black will open the kingside with ...g5-g4, which may be further strengthened by a rook on g8.

32.Nb3 White intends to counter Black’s plan by putting his knight on f5 (after ...g5). The attack on the a5-pawn gives Black something else to think about.   32...Rc2 This is good enough, but in a practical game I would be tempted to try 32...g5!?N 33.Nd4 (33.Qxa5 Qf6 gives Black excellent compensation for a pawn, with ...g4 coming next) 33...g4 34.hxg4 hxg4 Now ...g3 is a terrible threat, so White’s next move is forced.

35.fxg4 Rc4 (35...Nxg4 36.Nf5 Qg5 also seems fine) 36.Qd1 Otherwise the knight is a goner. 36...Qf6 37.Nf5 Rxe4 The position remains equal but far from dead.   33.Rbc1

A draw was agreed at this point, probably in anticipation of something like 33...Rxc1 34.Rxc1 Qb6† 35.Kh1 Ra8= followed by ...Qb4. ½–½

The Sophisticated 18.Bg5 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.b3 Bg7 16.d5 Rc8 17.c4 c6 18.Bg5 This move brings further nuances to the position. White provokes ...h6, hoping to profit from the inclusion of the pawn move. The fact that it has been played by Parimarjan Negi, my esteemed foreword-writer, gives us a hint that it should be taken seriously.

18...h6 A few strong players (including Negi’s opponent, Zherebukh, in their 2013 game) have preferred 18...Qc7, but I would prefer not to commit the queen so early. The key question now is how the ...h6 move will affect the evaluation of the lines we have seen previously, so please pay close attention to what follows.   19.Be3 After 19.Bd2 a5 I found no real difference from the 18.Bd2 line.   19...Nb6 For the moment we will follow the same path as in the analogous 18.Be3 line.   20.Rb1 20.Qd2?!N 20...bxc4 21.dxc6 c3 22.Qc1 Bxc6 23.Bxh6 Bb7 is promising for Black.   20.Qe2 cxd5 21.cxd5 The thematic piece sacrifice is still possible here, although the pawn on h6 will change some of the finer details. 21...Nbxd5! 22.exd5 Nxd5 23.Rac1! Other moves can be handled in the same way as shown in the Tseitlin – Avrukh game and accompanying notes. The text is White’s best, and was played in T. Kosintseva – Hammer, Cap d’Agde 2010, when Black should have continued: 23...Nxe3N 24.fxe3

24...e4! This is where Black should deviate from the analogous variation considered a few pages earlier. 24...d5 25.e4! is good for White. With the pawn on h7 Black could play either ...Bh6 (not possible here) or 25...f5 26.exf5 e4, but here after 27.fxg6 Black is unable to recapture with his pawn. 25.Nd4 Qg5 Black’s ideas include ...d5 and ...Be5, as well as ...b4, ...a5 and eventually ...Ba6. If he is allowed to carry out these ideas he will have excellent play, but what if White tries to exchange queens?

26.Qg4 Bxd4! Finally we see some ‘poetic justice’, as the presence of the pawn on h6 helps Black by keeping his queen defended. 27.exd4 b4 28.Qxg5 hxg5 Black will follow up with ...d5, ...f5 and ...Kf7-e6, with superb positional compensation for the missing piece.

20...Qc7 21.dxc6 Qxc6! Once again we diverge from the analogous position covered earlier.   21...Bxc6 is possible, but the positioning of the pawn on h6 leads to subtle differences. For instance, 22.cxb5 axb5 23.Qd2 Kh7 24.Rec1!?N is interesting, with the point that 24...d5 25.Nf5! leads to wild complications.   22.Bd3 bxc4 23.bxc4 Nbd7 A similar position was noted on page 355 as being fine for Black, and the placement of the h-pawn does nothing to alter the evaluation. A logical continuation is:

24.Nd2!?N 24.Qd2 h5 gave Black good prospects in Zarnescu – Kozlowski, corr. 2012.  

24...Nc5 25.Bc2 Ba8 26.f3 Qc7 27.Ne2 Nh5 28.Nc3 Bf6! 29.Nf1 29.Bxh6 Bh4 30.Rf1 Ne6 gives Black too much counterplay for a mere pawn.   29...Bh4 30.Re2 Kg7 The position is complicated, but I like Black’s kingside prospects with ...Rf8 and ...f5.  

15.a4 The final two games feature a move which has traditionally been regarded as the Breyer’s absolute main line, even though recent fashion has swung in other directions.   GAME 66 Alin-Mile Berescu – Peter Acs Rimavska Sobota 1996   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.a4 In Dynamics of Chess Strategy, Vlastimil Jansa recalls ‘inventing’ this move during a long walk through the city of Trinec, the night before his clash with Smejkal at the 1972 Czechoslovakian Championship. The database shows that it had actually been played as early as 1955, and had been repeated several times, but evidently it was still not commonly known in the early 70s. All that changed though, and it would later come to be regarded as the ‘absolute’ main line of the Breyer.

15...c5! This is the best reaction to White’s last move. Timing is everything: if Black were to play this move with the pawn still on a2, he would be left with a serious spatial disadvantage after d4-d5, as the ...c6 break would no longer be available. The difference here is that Black can play ...c4 to secure the c5-square for his knight. This plan would not work properly with the pawn on a2, as White could simply play b2-b4 at a convenient moment, meeting ...cxb3 with axb3 and then playing b3-b4 again to rob the black knight of its outpost.   By the way, a non-committal move like 15...Bg7 would be met by 16.Bd3! c6, leaving Black with a slightly worse version of a structure we have already encountered.

  16.d5! The expected reply, gaining space and blocking the b7-bishop.   16.b3 d5! and 16.b4 cxd4 17.cxd4 d5! are given as good for Black in my old ECO, and modern engines confirm this evaluation.   16.Be3 has been played a few times by Sax and Kotronias, but Black has more than one good response. 16...exd4 17.cxd4 Bxe4 (17...cxd4!? 18.Bxd4 occurred in Kotronias – Hjartarson, Reykjavik 1992, and now 18...bxa4!N, intending ...Nc5 without allowing axb5, seems fine for Black) 18.Nxe4 Nxe4 19.Bxe4 Rxe4 20.dxc5 Nxc5

21.axb5 (21.Qd5? bxa4 22.Ng5 Re7³ Sax – Ivanchuk, Manila 1990) 21...axb5 22.Qd5 Re8 This is a known equalizing line, featured in ECO among other sources, based on Sax – Beliavsky, Dresden 1988.

16...c4! Let me reiterate that this plan only really works when a2-a4 has been played. If the pawn was on a2, White would be able to play b2-b3 (or b2-b4, depending on circumstances), meeting ...cxb3 with axb3.   17.Be3 White has a few ways to approach this position. The most blunt is to play for a kingside attack with moves like Qd2, Bh6, Nh2-g4 and eventually f2-f4. A more sophisticated strategy is to combine play on both sides with Qe2 followed by doubling rooks on the a-file; the idea is that, when Black transfers his pieces across to the queenside, White will switch back to the kingside and exert pressure with h4-h5, Nh2 and so on. Both of these plans work better after Black has weakened his kingside with ...h6, which is why 17.Bg5! is widely regarded as the best move; this will be dealt with in Game 67 below. The text move enables Black to achieve an optimal set-up, which was first demonstrated in a game between Karpov and Portisch, as referenced in the note to move 19 below.

17...Nc5! 18.Qe2! When there is no tempo to be gained against a pawn on h6, this is the best square for the queen, as I came to understand through personal experience. The queen is flexibly placed on e2, with potential to influence the play on both flanks.   18.Bxc5 dxc5 19.Qd2 can be compared with the 17...h6 line given in the notes to the next main game. In that line we will see Black being put under pressure by the annoying plan of Reb1 and b2-b3, but here Black has not had to waste a tempo dealing with an attack on the h6-pawn. This gives him time to carry out an excellent regrouping manoeuvre:

19...Re7!N 20.Reb1 Ne8! We have seen a similar idea in the note to move 20 in the Fenwick – Andruischenko game. 21.b3 cxb3 22.Rxb3 Nd6 Black stands well, as his knight has reached the perfect blockading square.

18...Qc7 19.Ra3 White can also start with 19.Nd2 Nfd7. Several moves have been tried from here and many transpositions are possible, so I will focus on a couple of noteworthy games with considerable instructive value:   a) 20.f3? This was played in one of the pivotal games of the 2015 Women’s World Championship. Mariya Muzychuk conceived the idea of putting her g3-knight on f2 (via h1) and then playing for a kingside attack with f3-f4, but she later admitted that it was simply a bad plan. 20...Nb6 21.a5 Muzychuk also criticized this move, although at least it is consistent with the aforementioned plan. However, it

turns out that Black is the one who can get better chances on the kingside now that she doesn’t have to worry about the queenside. 21...Nbd7 22.Nh1  

22...Be7 22...Bc8!? followed by ...Nb7 to win the a5-pawn is possible, but in that case White may obtain serious attacking chances. Muzychuk recommended 22...Bg7!N 23.Nf2 Rf8 intending ...f5, noting that the bishop should be on g7 rather than e7 to prevent Bh6. This indeed looks like the best plan for Black, but it is worth following the game for a few more moves to consider the interesting position that arises below. 23.g4 Qd8 24.Qf2 Bh4 25.Ng3 Rc8 26.Kg2 Nf8 27.Rf1

In M. Muzychuk – Pogonina, Sochi (6.2) 2015, Black played the natural 27...Bg5?! but 28.f4! gave White good

prospects on the kingside, and she went on to win in fine style. A better option for Black would have been: 27...g5! As recommended by Mauricio Flores Rios on his blog. Black intends to exchange on g3 as soon as White unpins the knight, and will get fine prospects on the kingside with ...Ng6, ...Kg7 and eventually ...h5.   b) 20.Ra3 Nb6

I will call this method of deploying the knights the Portisch Plan although, to be fair, it had already been used in similar positions by Robatsch and Furman. Another approach is to abandon the a-file with ...Rab8 and then redirect the bishop to c8. This plan is not bad, and has been played a couple of times by Karpov, but it is less ambitious and I regard it as slightly suboptimal. The point of the ...Nd7-b6 plan is to force a clarification of the queenside structure before White manages to create any opportunities on the kingside. 21.axb5 axb5 22.Rea1 Rxa3 23.Rxa3 Bc8! 23...Ra8 was also possible of course, but it turns out that White cannot do much along the a-file. Therefore it makes sense to carry out the thematic plan of improving the bishop by putting it on d7.

24.Qd1 Bd7 25.Qa1 Nba4 A draw was agreed in Karpov – Portisch, Milan 1975. If Black wanted to play on, a sensible plan would be to prepare ...f5 with moves like ...Be7, ...Qc8, ...Rf8, and possibly ...Bh4.

19...Nfd7 20.Rea1 Bg7 Black has a healthy position. His kingside is secure and he is poised to play ...Nb6, which will force White to make a decision regarding the queenside.

21.Nd2 Nb6 21...Rf8!? is not bad, and it actually sets a trap of sorts. Black intends to play ...Nb6 next, leading to something similar to the game continuation. If White tries to take advantage of the delay by opening the queenside with 22.b3?, there follows 22...cxb3 23.Bxb3 (23.Nxb3 Nxa4µ) 23...Nxb3 24.Nxb3 f5! 25.exf5 gxf5 with a great position for Black, as seen in Okos – Palecek, Slovakia 2000. Of course White should do something else on move 22, but it is hard to suggest something constructive, and ...Nb6 comes next anyway.

22.axb5 After 22.a5?! Nbd7 Black will never have to worry about the queenside again; what’s more, he can try to win the apawn with ...Bc8 and ...Nb7, similar to what we will see in the next main game.   White can also try and call Black’s bluff with a move like 22.h4. Obviously Black can reply with a noncommittal move like 22...h5, but he can also carry out his ‘threat’ on the queenside with: 22...Ncxa4 23.Bxa4 Nxa4 24.Nxc4

24...Bc8! Preparing to put the bishop on d7. There is no need to fear 25.Rxa4 bxa4 26.Rxa4, as 26...Rb8 leaves Black slightly better. By the way, Black can also threaten to win a pawn by playing ...Rec8 before ...Nxa4, but I generally prefer to keep this square clear for my bishop.   22...axb5 23.b4 cxb3 24.Rxa8 Rxa8 25.Rxa8† Bxa8 26.Nxb3 The queenside simplifications have led to a typical situation where Black is fine and the game is generally equal, but a lot of play remains.

26...Nc4 Black has no problem allowing an exchange on c5, as his active pieces and queenside majority balance out White’s protected passed pawn, which will be securely blockaded. The arising structure often occurs in Spanish lines where the white pawn has advanced to d5, so we should have at least one example of it in our mental database.  

27.Bxc5?! White should have preferred 27.Nxc5 dxc5 28.Bc1, preserving the bishop. After 28...Qa5 29.Qe1 Qa1!? it looks as though Black has some initiative, but the ruthless analysis engines insist that it’s just equal. Berescu probably thought the b3-knight could at least go to a5 and maybe c6 at some point, but this proves not to be the case.   27...dxc5 28.Bd3 Nd6 29.Qa2 Bb7 30.Na5 c4 31.Bc2 Bf6!³ Black will transfer his bishop to the a7-g1 diagonal before carefully preparing the ...f5 break. The rest of the game holds no great interest for our opening study, but it is worth paying attention to the way Black increases his advantage, as you may well have to play with a similar structure at some point.

32.Ne2 Bd8 33.Nxb7 Qxb7 34.Qa3 Qb6 35.Qa8 Kg7 36.g3 Bc7 37.Qa3 Bb8 38.Qc1 Ba7 39.Qe1 Qc5 40.Kg2 f5µ A glance at the position is enough to show why White should have kept his dark-squared bishop.

41.g4 f4!? 41...Nxe4 wins the d5-pawn but allows some simplifications. Acs prefers to avoid exchanges and continue torturing his opponent.   42.Ng1 g5 43.Nf3 Nf7 44.Qd2 Kf6 45.Bb1 Nd6 46.Qb2 Bb6 47.Kf1 Nc8 48.Kg2 Ba7 49.Qa2 Qb6 50.Kf1 Nd6 51.Kg2 Nb7 52.Qd2 Qa6 53.h4 h6 54.hxg5† hxg5 55.Qd1 Qb6 56.Qd2 Qc5 57.Qa2 Bb6 58.Qb2 Nd6

59.Ng1? After a period of stubborn resistance, White finally slips up.   59...Qc8! 60.Kh3 60.f3 Qc5 61.Nh3 Qe3 wins, and 60.Qe2 Ba5 61.Qf3 Nf7 also leaves White without a good defence against threats like ...Nh6 and ...Qc5-a3.   60...Qh8† 61.Kg2 Qh4 62.Qe2 Ke7 63.Qf3 Ne8! The knight is heading for f6, where it hits the weaknesses on e4 and g4. White’s queen is overloaded by defending not only those two pawns, but also the f2-square.  

64.Bc2 Nf6 65.Bd1 Nxe4 66.Nh3 Nd6 67.Bc2 Kf6 68.Bb1 Qh8 Having won a pawn, Black reroutes his queen to support the advance of the e-pawn.   69.Qe2 Qe8 70.f3 Be3 70...e4! 71.fxe4 Qe5 would have won more quickly, as White will have to lose either the e4- or the c3-pawn, leading to the swift collapse of his position. In the game Black goes for the same idea under less favourable circumstances, but he soon gets the desired result all the same.   71.Qc2 e4 72.fxe4 Qe5 73.Kf3 Ke7 74.Qb2 Qh8 75.Qg2 Qxc3 76.Qc2 Qh8 77.Qg2 Nf7 78.Ng1 Qc3 0–1   Our final game will deal with the most critical line of the 15.a4 variation.   GAME 67 Aleksander Mista – Anish Giri Doha 2014   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.a4 c5 16.d5 c4 17.Bg5! A subtle improvement over 17.Be3 as featured in the previous game. White is trying to provoke the weakening ...h6.

17...Nc5! For decades this move was believed to be inaccurate, as it allows White to put his queen on d2, making it impossible for Black to oust the bishop with ...h6. However, as you will see if you study the lines presented in the notes, Black’s idea is just that: to induce White to put his queen on d2, where it lacks some of the power it has on e2. It took me a lot of time and effort to appreciate this deep idea, but the fact that it was used by Giri, one of the strongest and bestprepared players of modern times, is a good indicator of its worth.   Up until quite recently, the main line was considered to be 17...h6 18.Be3 Nc5, but the following game cast a dark shadow over it: 19.Qd2 h5 20.Bxc5! dxc5 The computer evaluation is in White’s favour, but that is not the real issue; the position is complex and Black has done okay in high-level correspondence games. More often than not, the exchange of the bishop for the c5-knight is welcomed by Black, as it leaves him with the two bishops and a chance to put the knight on the ideal d6-square. Unfortunately, in this particular position White can develop some pressure on the queenside with:

21.Reb1! Nd7 (21...Qc7!? improves, but White keeps an edge with accurate play, starting with 22.Nf1! Kh7 23.Qd1) 22.b3! Nb6 23.Qe2! White’s play is annoyingly simple and strong, and he had the upper hand in Adams – Baramidze, Dortmund 2014.   17...Be7!? This move was proposed by Kaufman. I think it is okay for Black, but in the following correspondence game I came under a bit of pressure. 18.Be3 18.Bh6!? is not mentioned by Kaufman; I cannot see how Black can sensibly avoid transposing to our main game. 18...Qc7 19.Qe2! Nc5 20.Ra3 Reb8 It is hard to say if this is objectively best, but it is certainly consistent with Black’s general strategy. 21.Rea1 Here I came up with rather a subtle move.

21...Kg7!? Black would like to play 21...Nfd7, but 22.h4!? Nb6 23.axb5 axb5 24.h5 gives White some pressure on the kingside. The text move strengthens the kingside slightly and effectively stops h3-h4 for the moment, as ...Ng4 would be slightly annoying for White. 22.Nd2! Another high-level correspondence game continued 22.Nh2 h5 23.Nf3 Nfd7 24.Ng5 Qd8 and Black proceeded to play ...Qe8, ...Bd8, ...f6 and ...Qf7, obtaining some initiative in Kazoks – Ebert, corr. 2014. I considered this a model game for Black and was eager to follow it. 22...Nfd7 23.Qg4

This was the annoying point behind my opponent’s last move. The threat of Nf5† forces Black to retreat his king. 23...Kh8 24.Bh6 Black has to be careful, as White has various attacking ideas including h4-h5 and Nf1-e3.

24...Nb6 25.Qf3 Bf8 26.Qf6† Kg8 27.Bxf8 Rxf8 28.Nh5 Although in Koo – Ntirlis, corr. 2014, we agreed the draw, I do not believe this line is the best choice for a practical player, as things can easily go wrong when the kingside is left with so few defenders. This game was also useful in demonstrating why the white queen is much better placed on e2 than on d2.   Before moving on, let me mention that 17...Rb8 followed by ...Bc8 has been used by Karpov and, more recently, Baramidze and Dominguez. Another idea is Petrosian’s 17...Bg7 followed by ...Qe7-f8. Both of these continuations are solid but slightly passive. White’s normal plan is to double on the a-file and prepare b2-b3, with no more than a tiny edge.   18.Qd2 This has generally been considered the right way to ‘punish’ Black for not playing ...h6. White takes control of the kingside dark squares and hopes to build an attack with moves like Bh6, Nh2-g4 and f2-f4, but it turns out that Black can defend easily enough in such a scenario.   After I played through the main game, I wondered what would happen after: 18.Qe2 As we have seen, there are some advantages to putting the queen on this square, but Black can revert back to the traditional scheme with: 18...h6 19.Be3 Nfd7 Black has an improved version of the old main line, as White will lose a tempo if he puts the queen on d2.

20.h4 20.Ra3 Qc7 21.Rea1 Bg7 followed by ...Nb6 is fine for Black. The position is the same as the previous game, except for the position of the pawn on h6. 20.Nh2 makes no sense, as after 20...h5 White cannot follow up with Ng5 (unless he wants to lose another tempo) or f2-f4, as the presence of his queen opposite the black rook will allow a ...Bxd5 tactic. 20...h5 21.Ng5 Qc7 22.Ra3 In this position a draw was agreed between two correspondence GMs in Freeman – Telepnev, corr. 2014, as Black is fine after 22...f6 followed by ...Nb6. By the way, the Russian Igor Viktorovich Telepnev is a player whose Breyer games are worth studying closely, as he has employed it against strong opposition with excellent results.

18...Be7! Black is ready to continue with the ‘Portisch Plan’ of ...Nfd7-b6.  

19.Bh6 An exchange of dark-squared bishops will not suit White, so he may as well avoid it now.   19.Be3 Nfd7 20.Rf1 Qc7 21.Nh2 is given by Khalifman as slightly better for White.

However, 21...h5! is absolutely fine for Black, as demonstrated by a highly instructive Telepnev game: 22.Bg5 Bxg5 23.Qxg5 Qd8 24.Qh6 Qf6 25.Nf3 Qg7 26.Qe3 Nb6 Black is at least equal, and now White commits the positional crime of closing the queenside.  

27.a5? Nbd7 Now Black is free to attack on the kingside. 28.Ng5 h4 29.Ne2 f6 30.Nf3 g5 31.Nh2 Nf8 Black brought his knight to f4 and prepared ...f5, eventually breaking through on the kingside in Cvetnic – Telepnev, corr. 2012.

19...Nfd7 20.a5 This move feels strategically wrong to me, as White is not yet in a position to profit from brutally attacking on the kingside with Nh2 and f2-f4, as Black is not obliged to gather his forces on the queenside. My guess is that Mista, who undoubtedly possesses deep understanding of the Spanish, either knew or anticipated at the board the following improvement over old theory.   20.Nh2 This used to be considered good for White, based on the continuation of Kavalek – Spassky, Montreal 1979, where the former World Champion played 20...Rb8. The game was eventually drawn, but only after White missed a promising attacking continuation, which is pointed out by Jansa among others. However, in 2011 a major improvement was revealed. 20...Nxa4! 21.Bxa4 bxa4 This leads to a more interesting and double-edged game, and is surely what Giri had in mind. Telepnev has been the hero of Black’s cause, scoring +1 =5 –0 from this position in correspondence games. Naturally I will show a few examples from his praxis. 22.Rxa4 22.Be3 a5 leads to similar play; Black’s eventual defeat in Megaranto – Leitao, Tromso (ol) 2014, was caused by subsequent errors rather than the opening. 22...a5

Black will get excellent counterplay after putting the bishop on a6 and bringing the knight to c5, and most probably to d3. Obviously White will try to do something on the kingside, but Telepnev has shown that a timely ...f6 keeps things solid and enables Black to bring enough pieces into the defence. 23.Be3 23.Ng4 Ba6 24.Ra2 Nc5 25.f4 Bc8 26.f5 Bh4 27.Qe3 f6 28.Kh2 Qe7 29.Rf1 Kh8 30.Qf3 Rg8 In Manduch – Telepnev, corr. 2012, Black continued by exchanging on f5 and putting pressure on d5. 23...Ba6 24.Ng4 Bf8 25.Ra2 Black was also fine after 25.Ra3 Bb5 26.Bg5 f6 27.Be3 a4 28.Qe2 Re7 29.Nh2 Rf7 30.Qg4 Rb8 31.h4 Kh8 in Boldysh – Telepnev, corr. 2013.

25...Bb5 26.Bh6 f6 27.Rf1 Kh8 28.f4 Bxh6 29.Nxh6 Rf8 30.fxe5 dxe5 31.Qe3 a4 32.Ng4 Qe7 33.Rf3 h5 34.Nf2 Kh7 35.Nf1 Nc5 36.h4 Nb7 Black continued with ...Nd6 and was in good shape in Dolgov – Telepnev, corr. 2011.

20...Bc8! I like the way Giri handles the position. The moment the bishop is no longer needed on b7 to defend against axb5, he repositions it to a better diagonal, while also creating the threat of ...Nb7xa5.   21.Ra3 21.Nh2 with the idea of f2-f4 is possible, but Black is doing fine after 21...Nb7 22.f4 exf4 23.Qxf4 Ne5.   21...Nb7 22.Rea1 Ndc5 23.Nh2 Bd7 24.Be3 24.f4?! Bh4! would be annoying for White.

24...Bh4 25.Nf3 Bf6 26.Ne2 Qc8 27.Kh2 Bg7 Black has regrouped successfully and is ready for ...f5.

  28.g4 Na4! 28...f6!? 29.Ng3 Nd8 followed by ...Nf7 is a manoeuvre which is worth knowing about, but Giri’s move is stronger – Black simply intends to gobble up the a5-pawn.   29.Bxa4 bxa4 30.Ng3 30.Rxa4?! Bxa4 31.Rxa4 hardly gives White enough for the exchange. 31...f5!? is possible, but 31...Bf6! followed by ...Bd8! is even stronger.   30...Nxa5 31.Rg1 Nb3 32.Qe2 Rb8 Simple chess. The rook goes to the open file and, if needed, can defend the 7th rank via b7. Black might intensify the pressure on b2 with ...Qb8 at some point in the future.   33.Nh4 Kh8 34.Qf3

34...Qd8!? Defending the pawn with 34...f6 or 34...Rf8 is possible, but Giri decides to set a trap.   35.Qxf7? Surprisingly, White falls for it! Perhaps time trouble was a factor.   35.Ng2 had to be played, but Black can continue improving his position with 35...Rb7, while White has to find a role for his g2-knight.   35...Rf8 36.Nxg6† hxg6 37.Qxg6 White threatens Nf5 but there is a straightforward refutation, so it is not clear what Mista overlooked in his calculations.   37...Be8 38.Qe6 Rf6 The queen is trapped, so White resigned. 0–1  

Conclusion   Our tabiya for this final chapter occurs after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8. After taking a short look at the not-so-critical 14.Bg5 in our first game, we moved on to the traditional Breyer main line with 14.Ng3 g6. Game 61 covered 15.Bg5 h6 16.Bd2 Bg7, by which time Black is ready for ...exd4 followed by ...c5. Karpov’s 17.a4! is critical, but Korchnoi’s 17...Nb6! is a good reply. For instance, if 18.b3 bxa4 19.bxa4 a5! Black can think about putting his bishop on a6.   We then moved on to 15.b3 Bg7, when the main line continues 16.d5! Rc8! 17.c4 c6. Then 18.Be3 Nb6 19.Qe2 (19.Rb1 Qc7! leads to a different type of position with equal chances) 19...cxd5 20.cxd5 Nbxd5! 21.exd5 Nxd5 gives Black amazing compensation for the piece. 18.Bd2 a5 19.Be3 forces Black to find a different plan, but he can utilize the extra ...a5 move with 19...Ba6!, planning ...Nc5 at a suitable moment. Lastly we considered 18.Bg5 h6 19.Be3, when Black should follow roughly the same path as after 18.Be3, but with some small modifications due to the extra ...h6 move.   The final section of the chapter dealt with 15.a4 c5 16.d5 c4, when it was explained that 17.Bg5! is strongest. My advice is to ignore the bishop for the time being and play 17...Nc5! in order to lure the white queen to d2. After 18.Qd2 (18.Qe2 h6! 19.Be3 Nfd7 is fine for Black, as White can no longer win a tempo against the h6-pawn with Qd2) 18...Be7! Black is ready to continue regrouping with ...Nfd7. The illustrative games involving Giri, and especially Telepnev in the notes, show that Black is not worse at all.

Index of Main Games   Chapter 1 – Early Deviations and Gambits Game 1: Dimitrios Mastrovasilis – Zurab Sturua, Istanbul 2003 22 Game 2: Ian Nepomniachtchi – Francisco Vallejo Pons, Moscow 2007 28 Game 3: Robert Rabiega – Andrei Maksimenko, Berlin 2009 31 Game 4: Mauro Marchisotti – Alena Lukasova, Correspondence 2013 42   Chapter 2 – Bishop’s Opening and Vienna Game Game 5: Pierangelo Turati – Marek Sadowski, email 2009 48 Game 6: Stefano Rosselli del Turco – Akiba Rubinstein, Baden-Baden 1925 53 Game 7: Andrei Istratescu – Anatoly Karpov, Bucharest (rapid – 5) 2005 58   Chapter 3 – Four Knights – Introduction Game 8: Gennady Obukhov – Steff en Schubert, Email 2009 65 Game 9: Cees Raijmaekers – Guntis Gerhards, Correspondence 2014 69 Game 10: N. Kosolapov – Rashid Nezhmetdinov, Kazan 1936 71 Game 11: Lu Shanglei – Evgeniy Najer, China 2013 80 Game 12: Shakhriyar Mamedyarov – Dmitrij Jakovenko, Tbilisi 2015 83   Chapter 4 – Four Knights – 4.d4 and 4.Bb5 Game 13: Rok Hrzica – Tamas Banusz, Trieste 2008 90 Game 14: David Navara – Chanda Sandipan, Gibraltar 2013 96 Game 15: Hikaru Nakamura – Magnus Carlsen, Zurich (blitz – 5.3) 2014 98 Game 16: Hou Yifan – Anatoly Karpov, Sanya City (1) 2010 106   Chapter 5 – Scotch Game Game 17: John van der Wiel – Anastasia Savina, Antwerp 2011 117 Game 18: Vassily Ivanchuk – Zoltan Almasi, Havana 2014 120 Game 19: Wang Hao – Rustam Kasimdzhanov, Tashkent 2012 128 Game 20: Alexander Morozevich – Jeroen Piket, Internet (blitz) 2000 131 Game 21: Dusko Pavasovic – Dimitrios Mastrovasilis, Topola 2004 132 Game 22: Sergey Karjakin – Rustam Kasimdzhanov, Zug 2013 135 Game 23: Yu Yangyi – Ruslan Ponomariov, Danzhou 2014 139   Chapter 6 – Two Knights – 4.d4 and 4.d3 Game 24: John Menke – Fred Kahl, Correspondence 2014 153 Game 25: Edward Barr – Robert Zsifkovits, email 2007 157   Chapter 7 – Two Knights with 4.Ng5 Game 26: Marcel Griesmann – Robert Dubois, Correspondence 1990 170 Game 27: Sergey Turko – Nikolay Trofimov, Correspondence 2012 177 Game 28: Torgny Skogfeldt – Kari Kuosmanen, Correspondence 2013 180 Game 29: Vassily Ivanchuk – Alexander Beliavsky, Dortmund 1998 185 Game 30: Nigel Short – Garry Kasparov, Leuven (blitz – 8) 2011 192  

Chapter 8 – Exchange Variations Game 31: Peter Romanovsky – Mikhail Botvinnik, Moscow 1935 206 Game 32: Konstantinos Ignatiadis – Thomas Mihos, Poros 1998 211 Game 33: Bartlomiej Heberla – Mart Kraemer, Muelheim 2014 216 Game 34: Hendrik Sarink – Soren Larsen, email 2001 219 Game 35: Igor Bondarevsky – Vassily Smyslov, Moscow 1946 225 Game 36: Mateusz Bartel – Evgeny Tomashevsky, Eilat 2012 229 Game 37: Michele Godena – Vassily Ivanchuk, Reggio Emilia 2011 233   Chapter 9 – On the Road to the Main Line Game 38: Judit Polgar – Michael Adams, Las Palmas 1994 242 Game 39: Ole Norregaard Hansen – Mircea Bucur, email 2008 248 Game 40: Vugar Gashimov – Peter Svidler, Nice (blindfold) 2010 250 Game 41: Li Chao – Evgeny Tomashevsky, Ningpo 2008 255 Game 42: Axel Rombaldoni – Sabino Brunello, Alghero 2011 257   Chapter 10 – The Trendy d2-d3 Game 43: Magnus Carlsen – Fabiano Caruana, Wijk aan Zee 2010 265 Game 44: Ilia Smirin – Alexander Grischuk, New Delhi/Teheran (2.1) 2000 269 Game 45: Eltaj Safarli – Gabriel Sargissian, Khanty-Mansiysk (rapid – 4.9) 2013 274 Game 46: Nazar Firman – Aleksander Mista, Germany 2014 279   Chapter 11 – Breyer – 10.d3 and Sidelines after 10.d4 Game 47: Walter Browne – Lajos Portisch, Wijk aan Zee 1972 288 Game 48: Thomas Rattinger – Olli Ylonen, Correspondence 2013 291 Game 49: Mikhail Tal – Boris Spassky, Tbilisi (11) 1965 292 Game 50: Evgeni Vasiukov – Yuri Averbakh, Moscow 1964 294 Game 51: Robert Byrne – Boris Spassky, Moscow 1971 299 Game 52: Zadok Domnitz – Samuel Reshevsky, Netanya 1973 302 Game 53: Sergei Zhigalko – Vassily Ivanchuk, Riga (rapid) 2014 305   Chapter 12 – Breyer – Alternatives to 13.Nf1 Game 54: Karel Mokry – Jan Smejkal, Bratislava 1983 312 Game 55: Wolfgang Unzicker – Jan Smejkal, Lucerne (ol) 1982 314 Game 56: Walter Browne – Anatoly Karpov, Amsterdam 1976 317 Game 57: Hagen Tiemann – Einar Laane, email 2010 319 Game 58: Alexei Bezgodov – Yuri Balashov, Perm 1997 323 Game 59: Maria Sebag – Humpy Koneru, Doha 2011 326 Game 60: Aleksandr Istomin – Nikolaos Ntirlis, Correspondence 2014 329   Chapter 13 – Breyer Main Lines Game 61: Leinier Dominguez Perez – Sergey Karjakin, Tashkent 2012 343 Game 62: Antonios Pavlidis – Levente Vajda, Pardubice (rapid) 2014 345 Game 63: Robert Byrne – Boris Spassky, Candidates Match, San Juan 1974 349 Game 64: Mark Tseitlin – Boris Avrukh, Haifa 2008 354 Game 65: John-Paul Fenwick – Andriuschenko, Correspondence 2010 358 Game 66: Alin-Mile Berescu – Peter Acs, Rimavska Sobota 1996 362 Game 67: Aleksander Mista – Anish Giri, Doha 2014 369

Variation Index   Chapter 1 – Early Deviations and Gambits 1.e4 e5   The Portuguese Opening 2.Bb5 page 15   The Nakamura Variation 2.Qh5 page 16   The Ponziani Opening 2.Nf3 Nc6 3.c3 page 18   The Göring Gambit 2.Nf3 Nc6 3.d4 exd4 4.c3 page 20 4...dxc3 page 20 4...d5 Game 1, page 22   The Centre Game 2.d4 exd4 3.Qxd4 (3.c3 page 26) 3...Nc6 4.Qe3 Nf6 5.Nc3 (5.e5? page 31) 5...Bb4 6.Bd2 0-0 7.0-00 Re8 page 27 8.Qg3 Game 2, page 28 8.Bc4 Game 3, page 31   The King’s Gambit 2.f4 exf4 page 33 A) 3.Bc4 Qh4†! 4.Kf1 d6! page 34 A1) 5.Nc3 page 34 A2) 5.Nf3 page 36 B) 3.Nf3 Nf6! page 38 B1) 4.Nc3 page 38 B2) 4.e5 page 40 & Game 4, page 42   Chapter 2 – Bishop’s Opening and Vienna Game 1.e4 e5   The Bishop’s Opening 2.Bc4 Nf6 3.d3 (3.d4 page 47) 3...c6 page 47 4.Nf3 (4.Qe2 page 47) 4...Be7 5.0-0 d6 6.h3 Game 5, page 48 (6.Re1 page 48; 6.c3 page 47 & 49)   The Vienna Game 2.Nc3 Nf6 page 51

  A) 3.g3 page 51   B) 3.Bc4 Nxe4! 4.Qh5 (4.Nf3?! page 52) 4...Nd6 page 52 5.Qxe5† (5.Bb3 page 52) 5...Qe7 6.Qxe7† Bxe7 7.Bb3 Nf5! Game 6, page 53   C) 3.f4 d5! 4.fxe5 (4.exd5?! page 57) 4...Nxe4 page 57 5.Nf3 (5.d3 page 57; 5.Qf3 page 57) 5...Be7 6.Qe2 (6.d4 page 58) 6...Nxc3 7.dxc3 c5! 8.Bf4 Nc6 9.Qf2 Game 7, page 58 (9.0-0-0 page 60)   Chapter 3 – Four Knights – Introduction 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6   The Halloween Gambit 4.Nxe5 Nxe5 5.d4 Game 8, page 65   4.a3 d5! 5.Bb5 (5.exd5 page 69) Nxe4 page 69 6.Nxe5 (6.Qe2 page 69) 6...Qf6! Game 9, page 69   4.d3 Game 10, page 71   4.h3 page 73   4.Be2 page 74   The Italian Four Knights 4.Bc4 Nxe4! page 75   The Glek System 4.g3 d5! 5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3 Bc5! page 77 8.0-0 0-0 9.Re1 Qf6! 10.d3 (10.Qe2 pages 79 & 80; 10.Rb1 page 80) 10...Bb6! 11.Bg5 Game 11, page 80 (11.a4!? page 81) 11.Be3 Game 12, page 83   Chapter 4 – Four Knights – 4.d4 and 4.Bb5 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6   The Belgrade Gambit 4.d4 exd4 5.Nd5 Nb4! page 89 6.Bc4 (6.Nxf6†? page 90; 6.Nxd4?! page 90) 6...Nbxd5 7.exd5 Bb4† 8.Bd2 Game 13, page 90 (8.Kf1!? page 90; 8.c3 page 91)   The Scotch Four Knights 4.d4 exd4 5.Nxd4 page 93 5...Bb4! 6.Nxc6 bxc6 7.Bd3 (7.Bg5?! page 93; 7.Qd4 page 93) 7...0-0 8.0-0 d5 (8...Re8 page 94) 9.exd5 cxd5

page 94 10.h3!? Game 14, page 96 10.Bg5 c6 11.Na4 Game 15, page 98 (11.Qf3 page 95; 11.Ne2 page 98)   The Spanish Four Knights 4.Bb5 Bb4 5.0-0 0-0 page 102 6.d3 (6.Bxc6 page 103) 6...Bxc3! (6...d6 page 104) 7.bxc3 d6 8.Bg5! Qe7! 9.Re1 a6! (9...Nd8 page 105) 10.Bxc6 Game 16, page 106 (10.Ba4 page 106; 10.Bc4 page 106)   Chapter 5 – Scotch Game 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4   On the Way to the Main Line 4...Nf6 5.Nxc6 bxc6 Vallejo’s 6.Qe2!? page 113 (6.Bd3 page 113) Sidelines after 6.e5 6.e5 Qe7 7.Qe2 Nd5 page 116 8.Nd2 Game 17, page 117 8.c4 Nb6 9.b3 Game 18, page 120 (9.Nd2 page 120)   The Main Line 8.c4 Nb6 9.Nc3 Qe6 10.Qe4 Bb4 11.Bd2 Ba6 12.b3 Bxc3 13.Bxc3 d5 page 124 14.Qf3 dxc4 15.Rc1?! Game 19, page 128 (15.Be2 page 128) 14.Qh4 dxc4 page 126 15.Be2 Nd5 16.Bd4?! Game 20, page 131 16.Bxc4 Game 21, page 132 15.Rc1 Qg6! 16.Bb4 Game 22, page 135 16.g3!? Game 23, page 139   Chapter 6 – Two Knights – 4.d4 and 4.d3 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6   The Original Morphy Attack 4.d4 exd4 5.Ng5?! page 145 The Improved Morphy Attack 5.0-0 Nxe4 6.Re1 d5 7.Bxd5 page 146 (7.Nc3?! page 146) The Modern Variation – Introduction 4.d4 exd4 5.e5 d5! 6.Bb5 Ne4 7.Nxd4 Bc5! 8.0-0?! page 151 (8.Nxc6? page 151) 8.Be3 Game 24, page 153  

4.d3 – Introduction and 5.Nc3: 4.d3 Be7 5.Nc3 page 155 Main Lines after 5.0-0 0-0 page 157 The c2-c3 Plan 6.Bb3 (6.c3 page 157) 6...d6 7.c3 Game 25, page 157 The a2-a4 Plan 6.a4 page 160 The Right Move Order 6.Re1 d6 7.a4 page 161 The Critical Na3 Plan 7...Na5 8.Ba2 c5 9.Na3! page 163   Chapter 7 – Two Knights with 4.Ng5 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5   4...d5 5.exd5 Na5   The Kieseritzky Variation 6.d3 Game 26, page 170   6.Bb5† c6 7.dxc6 bxc6 8.Qf3 Be7! 9.Bxc6†?! Game 27, page 177 9.Bd3! Game 28, page 180 8.Be2 h6 Steinitz’s 9.Nh3 Game 29, page 185 9.Nf3 – The Main Line page 188 8.Bd3! – The Modern Trend 8...Nd5! page 191 9.Ne4!? page 191 9.Nf3 Game 30, page 192 The Critical 9.h4! page 197   Chapter 8 – Exchange Variations 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6   The Exchange Variation 4.Bxc6 dxc6 page 205 5.Nc3 f6 6.d3?! Game 31, page 206 6.d4 page 210 The Simple 5.d4 page 211 5.d4 exd4 6.Qxd4 (6.Nxd4?! page 211) 6...Qxd4 7.Nxd4 Bd6 8.Nc3 Ne7

9.Be3 Game 32, page 211 Capablanca’s Method 9.0-0 0-0 page 214 The Main Line 5.0-0 page 216 5...Bg4 6.h3 h5! 7.d3 (7.c3 page 216) 7...Qf6 8.Be3 Game 33, page 216 8.Nbd2 Game 34, page 219   The Delayed Exchange Variation 4.Ba4 Nf6 5.0-0 (5.Bxc6 page 225) 5...Be7 6.Bxc6 dxc6 page 225 7.Re1 Game 35, page 225 (7.Qe1 page 226; 7.Qe2 page 226; 7.Nc3 page 226) 7.d3 Game 36, page 229   5.d4 – The Central Attack 4.Ba4 Nf6 5.d4 exd4 6.0-0 Be7 page 231 7.Re1 page 232 7.e5 Game 37, page 233   Chapter 9 – On the Road to the Main Line   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6   Spassky’s 5.Nc3 page 240   The Worrall System 5.Qe2 page 241 5.0-0 Be7 6.Qe2 Game 38, page 242   Just Before the Main Lines 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 page 246 8.c3 (8.h3 page 246) 8...0-0 page 246 9.a4 Game 39, page 248 9.d4 Bg4 10.d5 Game 40, page 250 10.Be3 exd4 11.cxd4 d5! 12.e5 Ne4 13.Nc3 Game 41, page 255 13.h3 Game 42, page 257   Chapter 10 – The Trendy d2-d3   1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6  

5.d3 page 263 5...d6! 6.c4 page 263 6.c3 g6 7.Nbd2 Bg7 8.Nf1 0-0 9.Bg5 (9.Ng3 page 265) 9...d5! Game 43, page 265 Main Line 8.0-0 0-0 9.Re1 Re8! page 268   The Fashionable 6.d3 5.0-0 Be7 6.d3 b5 7.Bb3 d6 8.c3 Game 44, page 269 8.a4 (8.h3 page 274) 8...Bd7! 9.Be3!? Game 45, page 274 (9.c3 page 275; 9.Bd2!? page 275) Anand’s concept with 8.a3! page 278 8...0-0 9.Nc3 Na5 page 279 8...Na5! 9.Ba2 c5 Game 46, page 279   Chapter 11 – Breyer – 10.d3 and Sidelines after 10.d4 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d3 Nbd7 11.Nbd2 Bb7 12.Nf1 Nc5 13.Bc2 Re8 14.Ng3 Bf8 page 287 15.Nh2 Game 47, page 288 (15.a4 page 288) 15.b4! Ncd7! 16.a4 Game 48, page 291 16.Bb3 Game 49, page 292 16.d4 Game 50, page 294   Introduction to 10.d4 page 298 10...Nbd7 11.Nh4 Game 51, page 299 (11.Nbd2 page 299) 11.c4 c6 12.c5 Game 52, page 302 (12.Qc2 page 303; 12.Bg5 page 303; 12.cxb5 page 303; 12.Nc3 page 303) 12.a3 Game 53, page 305   Chapter 12 – Breyer – Alternatives to 13.Nf1 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8   13.b4 Bf8 14.a4 Nb6 15.a5 Nbd7 16.Bb2 Rb8 17.Qb1 Game 54, page 312 17.Rb1 Ba8! 18.Ba1 g6

19.Bd3?! Game 55, page 314 19.c4 Game 56, page 317   Introduction to 13.a4 page 318 13...Bf8 14.Bd3 c6 15.b4 Game 57, page 319 15.Qc2 Game 58, page 323 (15.Nf1 page 323) 15.b3 g6 16.Bb2 Bg7 Game 59, page 326 (16...Nh5! page 326) 14.b3! g6 15.Bb2 (15.Ba3 page 330; 15.Bd3 c6 16.Ba3 page 330) 15...Bg7! (15...Nh5!? page 331) 16.Bd3 c6 17.Qc2 Rc8 18.Bf1! Nh5 19.dxe5! dxe5! 20.Ba3?! Game 60, page 329 The Super-Critical 20.b4! page 336   Chapter 13 – Breyer Main Lines 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.Nf1 Bf8   14.Bg5 Game 61, page 343 14.Ng3 g6 15.Bg5 Game 62, page 345 (15.Bd2 page 345) 15.b3 Bg7! page 349 16.a4?! Game 63, page 349 16.d5! Rc8! 17.c4 c6 18.Be3 Game 64, page 354 18.Bd2 Game 65, page 358 The Sophisticated 18.Bg5 page 360 15.a4 page 362 15...c5! 16.d5! (16.b3 page 363; 16.b4 page 363; 16.Be3 page 363) 16...c4! 17.Be3 Game 66, page 362 17.Bg5! Game 67, page 368