Opening Repertoire: The Sveshnikov 1781945632, 9781781945636

Citation preview

First published in 2020 by Gloucester Publishers Limited, London. Copyright © 2020 Cyrus Lakdawala The right of Cyrus Lakdawala to be identified as the author of this work has been asserted in accordance with the Copyrights, Designs and Patents Act 1988. 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. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: 978 1 78194 563 6 Distributed in North America by National Book Network, 15200 NBN Way, Blue Ridge Summit, PA 17214. Ph: 717.794.3800. Distributed in Europe by Central Books Ltd., Central Books Ltd, 50 Freshwater Road, Chadwell Heath, London, RM8 1RX. All other sales enquiries should be directed to Everyman Chess. email: [email protected]; website: www.everymanchess.com Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from Random House Inc.

Everyman Chess Series Commissioning editor and advisor: Byron Jacobs Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Printed by TJ International Limited, Padstow, Cornwall.

About the Author Cyrus Lakdawala is an International Master, a former National Open and American Open Champion, and a six-time State Champion. He has been teaching chess for over 30 years, and coaches some of the top junior players in the U.S. Also by the Author: 1 … b6: Move by Move 1 … d6: Move by Move A Ferocious Opening Repertoire Anti-Sicilians: Move by Move Bird’s Opening: Move by Move Botvinnik: Move by Move Capablanca: Move by Move Carlsen: Move by Move Caruana: Move by Move First Steps: the Modern Fischer: Move by Move Korchnoi: Move by Move Kramnik: Move by Move Larsen: Move by Move Opening Repertoire: … c6 Opening Repertoire: Modern Defence Petroff Defence: Move by Move Play the London System The Alekhine Defence: Move by Move The Caro-Kann: Move by Move The Classical French: Move by Move The Colle: Move by Move The Four Knights: Move by Move The Modern Defence: Move by Move The Nimzo-Larsen Attack: Move by Move The Scandinavian: Move by Move The Slav: Move by Move The Trompowsky Attack: Move by Move

Contents About the Author Bibliography Introduction 1 9 Nd5: The Strategic Line with 11 c3 2 9 Nd5: 11th Move Alternatives for White 3 9 Bxf6 gxf6: The Tactical Line 4 7 Nd5: The Modern Approach 5 White’s Sixth Move Alternatives 6 3 Nc3: The Anti-Sveshnikov 7 6 … Bc5: The Mamba Index of Complete Games

Bibliography Books Anti-Sicilians: Move by Move (Everyman Chess 2016) New Ideas in the Sveshnikov Sicilian, Valery Neverov and Peter Marusenko (Batsford 1996) Secrets of Opening Surprises, Jeroen Bosch (New in Chess 2010) Starting Out Sicilian Sveshnikov, John Cox (Everyman Chess 2007) The Sicilian Sveshnikov, Vasilios Kotronias (Quality Chess 2014) The Sveshnikov Reloaded, Dorian Rogozenko (Quality Chess 2005) The Sveshnikov Sicilian, Mikhail Krasenkov (Cadogan Chess 1996) Online and Electronic Sources ChessBase 13 with Komodo ChessBase annotations by Ernesto Inarkiev, Ivan Ivanisevic Chess Today with annotations from Alex Baburin.

Introduction

Do you sometimes wonder about our long distant Devonian fish ancestor who decided to emerge from the water and wobble on to land, becoming the world’s first amphibian? What prompted them to do it? Then there is the question of the Sveshnikov Sicilian. What urge compelled Jacques Mieses to push his e-pawn to the fifth rank against Siegbert Tarrasch in the very first Sveshnikov Sicilian in 1888? Wasn’t he afraid of that giant hole created on d5, or the newly created backward dpawn? Then Jorge Pelikan began playing the line in the 1950s. However, it wasn’t until GM Evgeny Sveshnikov honed it in the early 1970s that it became a real weapon. Then Peter Leko and Vladimir Kramnik began playing it in the late 1990s and early 2000s, re-energizing the opening. In the present day, it is world champion Magnus Carlsen, who has embraced the Sveshnikov. So from a biblical sense, Mieses begat Pelikan, who begat Sveshnikov, who begat Leko,

Kramnik and Carlsen. And as Carlsen goes, so goes our book. I used Carlsen as our Sveshnikov role model for this book, so whenever possible, we cover the lines he plays. Unlike other black openings, with the Sveshnikov, we don’t seek bland equality. Instead, we crave imbalance and the possibility of swift retaliation. Choosing our opening is similar to choosing our religion. To practice, we must first generate faith. Mine was instant with the Sveshnikov. A couple of decades ago my friend, National Master David Hart, was over at my house and we were playing online blitz. I got a challenge by a GM, whose name I don’t even recall. As a joke, David goaded me into meeting the GM’s 1 e2-e4 with 1 … c7-c5. Then David re-goaded me into playing a Sveshnikov, a position which previously only triggered fear, since I had never played or studied it in my life. In the past, I had experimented with the Najdorf and the Dragon, who both made no secret of their utter contempt of me and I abandoned them both in disgrace. To our shock, I won the blitz game against the GM easily. I had found my Sicilian and have never looked back. Since then I have played the Sveshnikov with doglike faith. Do you remember that Seinfeld episode where George Costanza does everything exactly contrary to his natural inclination and then he is handsomely rewarded for it? Well, the Sveshnikov is the same for me. The opening goes completely against my natural strategic stylistic grain, yet for some reason, it works for me. The difference between study and over the board battles is the difference between receiving a post card from Venice and actually going there on vacation. Our opening choices are mirrors of who we are, rather than molders of us. We strive for a level of mastery in our openings where our effort is effortless. The only way to get there is via diligent, repetitive study. The Sveshnikov is not an easy repertoire to master. Choosing our opening repertoire – much the same as our vote – is a way of declaring our beliefs to the world. The goal of this book is to navigate the wilderness into easy-to-understand ideas, civilizing that which was once untamed within our mind. The Sveshnikov is a High-end Aggression Opening When a gorilla seeks to take control and goes alpha, he thumps

his chest and bellows warnings to potential rivals. Primatologists refer to this behaviour as “threat display dominance”. This also describes the Sveshnikov Sicilian, which can be a scary place. The time has come for me to issue to the readers, warnings, encouraging slogans and homilies. My student, professor Joel Sneed, once told me, “I love the Sveshnikov but can’t play it, since the time I did, I thought I was going to have a heart attack!” So here is what you get when you play the Sveshnikov. * High blood pressure. * A dangerously elevated heart rate. * The sweats, including heightened body temperature. This sounds bad, until we consider that we also get. * Heightened alertness via intensified, adrenalin-fuelled focus. * Intensified killer instinct. Our normal instinct for selfpreservation is magically transformed to rage. The Svesh is deeply entrenched in the Old Testament eye-foran-eye ethos. The hell with turning the other cheek. Basically our universal answer to everything is violence. Trust in the cosmic balance. We give White a d5-hole to aim at and take on a backward d-pawn. In exchange we receive an endless supply of energy and initiative. The Sveshnikov is a Theoretically Dense Opening The Sveshnikov is an exceptionally difficult opening to master and sometimes it may feel as if the variations retreat from us, as if in a dream we are unable to remember. When it comes to chess talent, there are the impoverished and the obscenely wealthy, yet when it comes to natural talent, I’m guessing that most of us are born in the middle class. If you are an average level club player, this opening is a tough climb. But if you are willing to put in the effort, the Sveshnikov can be yours. Those who love a clean, orderly world, shouldn’t ever play the Sveshnikov, which is a painting created by an evil Jackson Pollock.

Some openings are so complex and so theoretically daunting, that they may feel suitable only for titled players. My feeling is there is no such thing as an opening so complex, that it is unfit for public consumption – as long as we, the public, are willing to put in the hard work to study the opening’s deepest secrets. If you do, the opening will no longer feel like an ocean of confusion. At first glance, the Sveshnikov starting position may feel as if Black is determined to violate as many principles as possible in a single chess game. The reality is that our opening is 100% sound and one of Black’s highest scoring Sicilians, where Black is actually the one with the plus score (50.9% at the time of this writing, in my database). The fact that Carlsen took it up and scores well with it, should reassure us further. The Svesh levies upon us the following tributes. The Sveshnikov is All About Piece Activity and the Initiative When we hand over control of d5 with … e7-e5!?, we must endure a backward and potentially weak d6-pawn. We also hand over (temporary) control of the d5-square. Now when it comes to d5, the psychological jurisdictions of both sides seem to overlap, in that both players feel that they – and not the other guy – deserve eventual control over the square. In exchange for these concessions. a) We receive our fair share of the centre. b) We receive enhanced piece activity, compared with other Open Sicilian lines. c) Any Sicilian is an unbalancing proposition. With the Sveshnikov, we enhance the asymmetry to an even greater degree. We are pledged excellent return on our investment. d) Does it hint at an imperious nature when we brazenly attempt to seize the initiative, right from the opening and with the black pieces? Just as strategists are the natural constituents for the Caro-Kann or Petroff – tacticians, lovers of open games and players who love a heavy theoretical lift tend to feel at home within the Sveshnikov. The vast majority of positions we reach are high-end dynamic. You are not normally going to get such positions from the Caro-Kann and the Petroff, at least not on a routine basis. The nature of some openings is that they are moody and irritable. Sometimes the pawn structures you get will

give you the creeps. We must keep in mind that we always get piece activity in return. Eventually, we learn to trust it. Cross the opening with an inaccurate move order and there is no forgiveness on the Sveshnikov’s side. The scale of complexity nearly always surpasses either side’s ability to fully comprehend. e) The Sveshnikov is a line well suited for lovers of instant liberty. Yet that liberty comes at the cost of a hole on d5 and backward d-pawn. Are we busted? No. You will find that Black generates compensating piece activity and often creates a pawn break on d5, even though, in the starting position of the Sveshnikov, it feels inconceivable. When this pawn break is achieved, we begin to understand how the middlegame develops. Black, much later on, either clogs d5 with a capture, where White recaptures with the e4-pawn. Alternatively, Black simply seizes control over the d5-break and plays the freeing … d6-d5. Black can sometimes just play around the hole. White will have control over it, but we just go about our business and generate counterplay elsewhere on the board. In essence, the d5hole can be ineffective window dressing for White. Having played the Sveshnikov for two decades, I don’t ever remember losing a game due to White’s control over d5 or the fact that my backward d6-pawn was fatally weak. The “advantage” of White’s (temporary) control over d5 and the “weakness” of our backward d6-pawn is often illusory. In many lines we break on f5 (sometimes twice, since in certain lines our f-pawns are doubled). The following two diagrams are typical … f5 breaks for our side.

Black sometimes generates queenside play with a pawn break of … b5-b4. Here is an example:

In some lines White hands us both the bishop pair and dark-square control. Winning with both is a recurring theme for Black. In many lines Black generates a direct assault on White’s king, who often feels a sense of desertion. Now we may be material behind and our structure torn to shreds but we get our shot at checkmate! Just remember that a subtle mind (White) is sometimes unable to withstand the violence of a simpleminded barbarian (us!). It’s just that our simple, big idea is actually easier to implement than White’s multiple incremental changes. On the downside, if we allow our attack to run

its course, our weaknesses still remain and we have nothing tangible to show for them. Whichever way the game turns out, it makes for exciting chess. In essence, when we take up the Sveshnikov, we are given permission to give in to delirious abandon – in fact, it’s mandatory! The dynamism of the Sveshnikov is clear from the diagram below (MacKinnon-Shabalov). White owns all the strategic pluses (with control over d5, Black’s backward and isolated d6pawn and ragged structure). Yet White is busted, since his king is desperately unsafe, wherever he places it.

I won’t lie to you: Learning this repertoire requires an enormous level of toil on your part. Not only to memorize the moves, but also to understand deeply the sometimes baffling ideas and temporary breaks from reality. Some people

mistakenly believe that playing the Sveshnikov requires the talent of a prodigious memory, a flair for the initiative and a strong tactical eye. But then I ask: did reading just come naturally to you? Or did someone first teach your how to read? In the same way, the more we study and play it, the less and less baffling the Sveshnikov’s ideas seem to be.

It’s a dangerous proposition to invest in a guess, yet that is exactly what we do on a routine basis in the Sveshnikov. The above diagram is from Karjakin-Carlsen, Shamkir 2019. On his last move Carlsen calmly castled, leaving his h5-pawn en prise. Did this decision really receive the Mandate from Heaven? Well, I don’t know. For the pawn, he got e5 for his knight and play on the kingside light squares. Such a speculative sacrifice remains gauzy for a long time. Only after many moves can we say definitively whether it worked or not. In the Sveshnikov, we are confronted by such agonizing decisions, virtually every game we play. Take a look at the following diagram.

Have you ever heard a high school band attempt to play a symphony, just slightly, annoyingly off key? In the above diagram it’s move 22 and we are just starting with a tabiya position! The theory reconfigures the position to resemble an LSD trip from the 1960s. This is not a position we jump into lightly. To play Sveshnikov well, we should have an aptitude for open positions, since our goal is to open with the breaks … d6d5 and … f7-f5. When we don’t understand an advanced technology, it can first appear to us as magic. The more we study and play it, the less and less mysterious the Sveshnikov becomes. The next three diagrams are the most important in the entire book. Each contains multiple tributaries. Be patient with yourself and you will eventually master all three, along with their subsets.

In the third position above, I learned through social media that lies travel faster than truth. In the past, 8 … Nb8 (the line covered in Sveshnikov: Move by Move) was considered the sound line, while 8 … Ne7!? was relegated to the maybe playable category. This all changed when Carlsen essentially retained his world champion’s crown with this line against Fabiano Caruana. Subsequent attempts to bust him in the line have ended in terrible failure for all those who tried. So in this book we trust in Carlsen and look at his line against White’s 7 Nd5 variation.

The Mamba: 6 … Bc5

I’m pushing your boundaries of both loyalty and trust by proposing this crazy line. It is a close relative of the Cobra (1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e6 6 Ndb5 Bc5!?). As it possesses similar venomous qualities I have decided to christen it the Mamba. Some of us don’t feel comfortable playing non-international brand name openings. Despite appearances, Black’s last move 6 … Bc5!?, which ignored White’s threat to plonk a knight down on d6, is free from contradiction. We care about our results and are nervous about any risky, unproven lines. I just want to calm you down by telling you this line is purely optional, suited only for the adventurers among us. Is the individual thinker driven to extinction in the opening stage of the game? If Big Brother Theory’s job is to indoctrinate us with correct versus incorrect, it is a chess writer’s job to put out lesser played options, which tend to confuse and devastate booked-up opposition. If you are tired of playing well-known opening lines which feel worn out via overplay or overanalysis,

to the point where it seems as if all our opponents play the line error free, then maybe give this one a try. The 6 … Bc5 line and the Sveshnikov proper are different openings, yet not so far apart. Black’s last move, the 6 … Bc5!? provocation, is an attempt to step outside the normal Sveshnikov framework. One look at the previous diagram and you believe this is a good place to bring up the universal mom admonition: “If Cyrus jumped off a bridge, would you do it too?” Of course, we chess players are far wiser than Mom and we play all sorts of dangerous openings which skirt the periphery of respectability. Why? Because others do it as well! I thought my quirky Sveshnikov sideline would sell better if I named it the 6 … Bc5 variation, rather than Cy’s Folly, which is perhaps the more accurate name. I started playing around with this rare move in the early 2000s in online blitz and, to my complete shock, found the line is actually sound, despite poor statistics in the database. A good salesperson believes in the product being sold, so I tried the line 11 times in my own tournament games, against opposition ranging from 2024 to 2375, with an average rating of 2193. My results so far: 10½-½ – a 91.7% score. In my very first online blitz game the 6 … Bc5 line took out a GM in 12 moves. The first time I tried it in a tournament game, I beat a 2350 in 14 moves. This opening clearly contains ambush value and I encourage you to try it as a surprise line against unsuspecting opponents. I’m confident I won’t regret placing this line in the book. I encourage you to rise above partisan beliefs and try this line out. I think you will be pleasantly surprised by the results. Many thanks as always to Nancy for proofreading. We have a lot of ground to cover, so let’s get started on the book. Cyrus Lakdawala, San Diego, March 2020

Chapter One 9 Nd5: The Strategic Line with 11 c3

The move 9 Nd5 is the most popular from White’s perspective, for several reasons: 1. Unlike many other Sveshnikov lines, this one can be (somewhat) winged, theoretically speaking. 2. White feels comforted by the fact that s/he dominates d5. This costs White both the bishop pair and dark square control, so we are fully compensated. 3. White’s position is safe and considered difficult to lose. This book is an ambitious attempt to cover both Sveshnikov and

Anti-Sicilians in a single volume, as a complete repertoire, but I simply don’t have the room to cover all lines with a full game. The following game (which I spent two days annotating!) covers many lines which didn’t make it to full game status in the book. It’s kind of a codex of many of the lesser lines of our variation. Hopefully, the knowledge contained within the notes will help you sleep a little better at night. Game 1 A.Volokitin-A.Shirov Poikovsky 2008 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 We of the Sveshnikov clan are not fussily neat about our structure and the holes we voluntarily create, as long as we receive compensating piece activity. I have played the Sveshnikov for over two decades and don’t recall a single game I ever lost due solely to White’s control over the d5-hole, or my backward d-pawn. Believe in the magic of our piece activity. It will never let you down. 6 Ndb5

Now we enter the “real” Sveshnikov lines. 6 … d6 The assumption is that Black should prevent 7 Nd6+, which is advice I have ignored for 20+ years! I specialize in two odd side lines: a) 6 … h6 – the Ulfie, which I named after GM Ulf Andersson, whose online blitz games drew me to the variation decades ago. This line is covered in the final chapter of Sveshnikov: Move by Move. b) 6 … Bc5!? – the Mamba, with which I score around 90%, despite its sleazy outer shell. We cover this radical line in the final chapter of this book. 7 Bg5 White logically goes after Black’s main defender of the d5hole. This move is overwhelming White’s most popular choice. Let’s look at the numerous side lines, only the first of which is dangerous for Black:

a) Caruana tried 7 Nd5 twice against Carlsen in their World Championship match. We devote an entire chapter to this line later in the book (see Chapter Four). b) 7 a4 is played with the intention to suppress Black’s … a7-a6 and … b7-b5 break. Play can continue 7 … a6 8 Na3 Be6 9 Bc4 Rc8 10 0-0 Be7 11 Be3 0-0 12 Re1 Nb4! (b4 is an artificial hole for Black’s knight, since White is in no position to engineer c2-c3) 13 Nd5 Nbxd5 14 exd5 (mission accomplished; the d5-square is now sealed with a white pawn – one of Black’s key goals in the Sveshnikov) 14 … Bd7 15 Bb3 Ng4 16 Bd2 f5 and Black already looks better, since his majority is more fluid and therefore more dangerous than White’s on the queenside, B.Escalante Ramirez-E.Cordova, Montevideo 2018. c) 7 Be3 a6 8 Na3 b5 (threat: … b5-b4) 9 Nd5 (threat: Bb6 and Nc7+) 9 … Rb8 10 Nxf6+ Qxf6 11 c4 b4 12 Nc2 Qg6 13 Qd5 (13 f3 b3 offers Black decent play) 13 … Bb7 14 Qd3 Be7 15 f3 0-0 16 Qd2 a5 17 b3, L.Aguero Jimenez-S.Himanshu, Badalona 2016. Chances are equal after 17 … Nd8 intending … Ne6. d) 7 Na3 Be7 8 Bg5 0-0 9 Bxf6 Bxf6 10 Nc4 Be6! 11 Qxd6 Qc8 and Black had excellent compensation for the pawn, with the bishop pair, dark square control and a development lead, C.Bauer-J.Lautier, French League 2002. d) 7 Bd3 a6 8 Na3 Bg4! 9 f3 Be6, when both … d6-d5 and … b7-b5 are in the air and I already prefer Black. e) 7 Be2 a6 8 Na3 b5 9 Nd5 Nxe4 10 Bf3 Nc5 11 Nf6+ (after 11 0-0 Rb8 I don’t see enough compensation for White for the sacrificed pawn) 11 … gxf6 12 Bxc6+ Bd7 13 Bxa8 Qxa8, Z.Daleczko-K.Klim, Barlinek 2007. For the exchange, Black has huge compensation and stands better. He has a pawn, controls the centre, leads in development and may later be able to make use of the open g- and c-files. 7 … a6 8 Na3 b5 9 Nd5

This is White’s strategic line and, from my observation at club level, White’s most popular choice, since it’s relatively safe and easy to play. I also think this line is the most dangerous challenge to the Sveshnikov. We look at White’s major alternative 9 Bxf6 gxf6 10 Nd5 f5 in Chapter Three. 9 … Be7 10 Bxf6 10 Nxe7 Nxe7 11 Bxf6 (11 Bd3 Nd7 12 c4 b4 13 Nc2 a5 is about equal, A.Morozevich-M.Carlsen, Moscow 2006) 11 … gxf6 12 c4 Qa5+ 13 Qd2 Qxd2+ 14 Kxd2 b4 15 Nc2 Rb8 and White had nothing from the opening, C.Bauer-P.Tregubov, Belfort 2003. 10 … Bxf6 11 c3 Other 11th moves are examined in the next chapter. One minor exception is 11 Nxf6+ (White is blatantly playing to simplify and draw; in general, this exchange nearly always benefits Black) 11 … Qxf6 12 c3 0-0 13 Be2 Bb7 14 0-0 Rad8 15 Bf3, D.Larino Nieto-V.Kotronias, Budva 2009. I already prefer Black’s position after 15 … Ne7 16 Qe2 d5.

11 … Bg5 12 Nc2 The once offside knight hopes to shift back into play, via either e3 or b4. White can also play 12 Be2 first, and then 12 … 0-0 13 Nc2. We’ll look at this later in the chapter. 12 … 0-0 13 a4

This is by far White’s most important option. It breaks up Black’s queenside and saddles him with an isolated a-pawn. Others: a) 13 Be2 transposes to 12 Be2 0-0 13 Nc2, mentioned above. b) 13 h4 Bh6 (let’s hope nobody falls for 13 … Bxh4?? 14 Qh5, when Black can resign) 14 g4 f6 (safer than the older try of moving the bishop to f4) 15 Rh3 Ne7 16 Ncb4 Be6 (remember to constantly challenge d5) 17 Rd3 Nxd5 18 Nxd5 Kh8 19 Bh3 Bg8 20 a4 bxa4 21 Qxa4 a5 22 Rad1 Rb8 23 b3 Rf7! 24 b4 axb4 25 Nxb4 Qc7 and White’s position is overextended, E.Sutovsky-L.Lenic, Austrian League 2017.

c) 13 Bd3 Ne7 14 Nce3 Be6 15 Bc2 Nxd5 16 Nxd5 Rc8 17 0-0 was V.Ocnarescu-J.Konnyu, Eger 2005. Black stands no worse after 17 … a5. d) 13 Nce3 Be6 14 Be2 Ne7 15 0-0 Nxd5 16 Nxd5 is similar to the previous note. Black is just fine. 13 … bxa4 14 Rxa4 a5 15 Bc4 15 Bb5 has been played but is illogical to my mind for two reasons: 1. The fight is for d5, so White’s light-squared bishop belongs on c4, not b5. 2. With 15 Bb5, White chases the knight to exactly where it wants to go (e7) to challenge White’s d5-knight. Play can continue 15 … Ne7 16 Ncb4 (White takes advantage of the pinned a8-rook) 16 … Bh3 17 Nxe7+ Qxe7 18 Bc6 (after 18 Nd5 Qb7 19 Bc4 Bd7 is equal), O.KorneevA.Shirov, Pamplona 2006. The game is dynamically balanced after 18 … Bxg2 19 Rg1 Rac8 20 Rxa5 Bh3 21 Qh5 Bh4! 22 Nd5 Qd8 23 Qh6 g6 24 Ra6 Be6. 15 … Rb8 16 b3 White secures the b-pawn and the c4-bishop. Later on we look at 16 Ra2. 16 … Kh8 This move is in preparation for our only break, with … f7f5. The major decision we must make is: 1. Do we want to back it up with … g7-g6 and then … f7-f5 -? 2. Do we want to play f5-f5 without … g7-g6 - ? If so, then White will play e4xf5 and may try and utilize the newly-created hole on e4. 17 Nce3 Or: a) 17 0-0 f5 18 exf5 Bxf5 19 Nce3 Bg6 20 Bd3 Bxd3 21 Qxd3 Bxe3 22 fxe3 Rxf1+ 23 Qxf1 Rxb3 24 Rc4 Rb5 25 Nc7 Rb6 26 Nd5 Rb5 27 e4 Rc5 28 Rxc5 dxc5 29 Qf7, when White

had full compensation for the pawn but I would rather play Black, R.Kasimdzhanov-B.Gelfand, Tashkent 2012. b) 17 h4 Bh6 18 g4 Bf4! (this time 18 … f6 is slightly inaccurate; White stands better after 19 Nce3 Bxe3 20 Nxe3) 19 Nce3 Ne7 20 Nxe7 Qxe7 21 Nd5 Qd8 and the game is dynamically balanced, J.Overgaard-T.Klausen, correspondence 2007. 17 … g6!?

This is one of the most important tabiyas in the book. The plan is to play … f7-f5 and, if White plays e4xf5, Black recaptures with the g6-pawn. This is Black’s main line, which allows White a dangerous and completely sound pawn sacrifice. The positions go haywire and are probably the most baffling in the entire book. If the thought of entry into such craziness gives you pause, then try the far safer idea 17 … Be6, which is covered later in the chapter. The move avoids the weakening (or dynamic) … g7-g6 and sidesteps White’s h2-h4 pawn sacrifice. 18 Qe2

This move is not as dangerous for Black as the immediate push of the h-pawn to the fourth rank with the critical 18 h4!, which we cover in the next three games. A quieter option, 18 00, eschewing the idea of h2-h4 altogether, is examined in Game 5. 18 … f5 19 h4 The only move. As already emphasized, the h2-h4 pawn sacrifices are one of Black’s biggest challenges. 19 exf5?! has been tried twice, with two losses for White. After 19 … gxf5 20 0-0!? (I would play 20 f3) 20 … f4 21 Nc2 f3! White’s king is under assault and unlikely to survive, S.Novikov-R.Pommrich, correspondence 2010. 19 … Bxe3 20 Qxe3 f4! This way Black clogs lines and his king is far safer than in the 18 h2-h4 lines. 21 Qd2 Bd7 22 Ra1 White is rightfully anxious about knight discoveries. 22 h5 is well met by 22 … g5. 22 … Be6 White is confronted with a problem – where he should place his king. GM Volokitin decides to walk across to the queenside. I already think Black is doing well, with equality at a minimum. 23 Kd1!? The king will not find safety on the queenside. However, 23 h5 is no improvement as after 23 … g5 24 Qd3 g4 White’s artificially isolated h-pawn may later be picked off if pieces begin to come off the board, N.Pascual Perez-E.Kopasov, correspondence 2009. 23 … Ne7! Black’s strategy remains the same as always – challenge d5. I already prefer Black, whose king safety supersedes his wobbly structure. 24 Kc2

After 24 Nxe7? Qxe7 White has trouble defending b3. For example, 25 Bd5 Rfc8 26 Kc2 Bxd5 27 Qxd5 Rc5 and White’s king is in danger. 24 … Nxd5 25 Bxd5 Bxd5 26 Qxd5 Qb6 Black eyes up b2 and f2. Notice how Black isn’t tied down to defence of either the a5- or d6-pawns, since his threats against White’s king takes precedence. We see this phenomenon over and over in the Sveshnikov, where our pawns are weak/not weak, since White lacks the time and resources to actually capture them. 27 Rhf1 Rfc8

28 Ra4 When on defence the natural tendency is to flinch at every shadow. Passive defence is not White’s best path. White may just barely hold the game with the radical 28 Rfd1! Qxf2+ 29 Rd2 Qe3 30 Rd3 Qb6 31 Rxa5 Qf2+ 32 Rd2 Qe3 33 Rd3 with a likely repetition draw. 28 … Rc5

28 … Kg7 is slightly more accurate. 29 Qe6 White should hold the game due to Black’s exposed king. 29 … Rc7 30 Qd5?! Principle: Passivity is death in major piece endings. White should activate the sleeping f1-rook with 30 Rd1! Qxf2+ 31 Rd2 Qe3 32 Qf6+ Kg8 33 Qe6+ Kg7 34 Rc4! Rxc4 35 Qe7+, when White’s queen delivers perpetual check. 30 … Rbc8! Now White is in deep trouble. 31 Qxa5?! This compounds the problem. 31 c4 is better, though after 31 … Rb8 32 Qd3 Rcb7 33 Rd1 Qxf2+ 34 Rd2 Qb6 35 Ra3 Rd7 Black has winning chances.

31 … Qxa5 Also strong is 31 … Rxc3+! 32 Qxc3 Rxc3+ 33 Kxc3 Qb5 34 Rc4 Qa5+ 35 Kb2 Qd2+ 36 Rc2 Qd4+ 37 Kb1 Qxe4 38 Rd1 d5. It’s one of the great joys to push the d-pawn to the fifth rank. White is unable to hold the game due to Black’s surging central pawns. 32 Rxa5 Rxc3+ 33 Kb2 Kg7 34 Rd1 After 34 Ra7+ Kh6 35 g4 fxg3 36 fxg3 Rxg3 37 Rff7 Rh8 38 b4 Rg4 39 Kc3 Rxe4 40 b5 d5 41 b6 Rc4+ 42 Kd3 Rb4 43 Rab7 Rxh4 44 Rfe7 Rd4+ 45 Ke3 Re4+ 46 Kd3 Rc8 47 Rxh7+ Kg5 White’s b-pawn isn’t getting through, since his rook is in front of the pawn. 34 … Rc2+ 35 Ka3 Rxf2 36 Rxd6 Rc7 37 Rxe5 Rxg2 Our destination seems far away when we are exhausted. White’s rooks are out of alignment to halt the push of the f-pawn and everything about Black’s position suggests imminent death. 38 Ree6 0-1 After 38 … Rf7 39 e5 Re2 40 Rd3 Re3! 41 Rd1 f3 42 Rf1 Rf4 43 Kb2 Rxh4 the g-pawn rolls forward and Black wins. 38 Rd3 didn’t work either, as after 38 … Rf7 39 Rf3 h6 40 Ka4 Rg3 41 Rf1 f3 42 Rc5 Rg4 43 Kb5 Rxh4 44 e5 Re4 45 b4 g5 Black’s pawns are too fast. Game 2 D.Jakovenko-A.Shirov Foros 2007 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 This game is the classical strategist versus tactician matchup, the maker versus the breaker. My friend Alexei Shirov’s entire career was in defiance of statistical improbability. So it’s no surprise that the Sveshnikov is one of his favourite lines. 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Nc2 0-0 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 White covers the tender b-pawn, while keeping open the option of a rook swing to the kingside or centre later after Ra2. White can also play 16 Ra2 immediately (as we’ll see in Games

4 and 9), though as b2-b3 generally follows at some point, a transposition to 16 b3 lines is most likely. 16 … Kh8 17 Nce3 g6 Just a reminder: if the coming chaos raises your blood pressure, and causes ragged breathing and unpleasantly elevated heart rate, then you can always bail out with the more mellow option 17 … Be6 which is examined in Games 6-9. 18 h4! Oh Lord, give us this day our daily mate! Despite appearances, this pawn sacrifice is absolutely sound. This is one of our greatest – if not the greatest – challenge for Black in this variation. It is essential to study this game and the coming ones with great care. 18 … Bxh4! Black should accept the gift. Others are worse: a) 18 … Bxe3?! 19 Nxe3 scores poorly for Black. Play can continue 19 … Ne7 20 h5 g5 21 0-0 Bb7 22 Nd5 Bxd5 23 Bxd5 Rc8 (after 23 … Nxd5 24 Qxd5 Black’s a and d-pawns look desperately weak) 24 Bc4 f5 25 Be6! Rxc3, S.SjugirovA.Krapivin, St Petersburg 2009. Here White can endanger the black rook with 26 Bc4! Qb6 27 Qd2 Rg3 28 Rxa5 fxe4 29 Ra6 Qc5 30 Rxd6 and Black is strategically busted. b) 18 … Bh6?! 19 h5 Kg7 20 Qd2! (threat: h5xg6, followed by Rxh6 and then horrible discovered checks on Black’s helpless king) 20 … Rh8 21 g3 f5 22 f4 fxe4, W.PrzybylskiM.Swicarz, Rewal 2008. White has a promising attack, as well as control over the central light squares. 19 g3 White wants to play f2-f4 and then transfer the a4-rook all the way over to h2. 19 … Bg5! The comp tells me that 19 … Bf6 is playable, but in my database Black scores miserably with the move. An example is 20 Ra2 Bg7 21 f4 exf4 22 gxf4 Re8 23 Rah2 h6 with a promising attacking position for White, H.Ollmann-J.Barrios Troncoso, correspondence 2007.

20 f4 exf4 It isn’t easy to seal a fire hydrant, once it breaks and sprays water everywhere. Now all hell breaks loose. 21 gxf4 Bh4+ Disrupting White’s castling. 22 Kd2

A wild position which, as you may have guessed, is prone to erratic fluctuations of fortune. White’s ambitious plan is to tuck his king out of the way on the queenside and then play Ra2 and Rah2, attacking down the h-file. Of course this takes a lot of time, so it is Black’s job to generate play in the middle before White’s attack becomes overwhelming. The comps say Black’s position is playable. In the next game we consider 22 Kf1. 22 … Ne7 As always, it is essential to chip away at White’s control over d5. 23 Kc1

An alternative is 23 Qg1 Nxd5 24 Nxd5 Bf6! (Black scores well after this move) 25 Qh2 h5 26 Nxf6 Qxf6 27 Qg3 d5! 28 Bxd5 (after 28 exd5 Bf5 29 Rxa5 Ra8 30 Rxa8 Rxa8 Black’s attacking chances offer sufficient compensation for White’s three passed pawns; the comp assesses it dead even here) 28 … Rd8 29 Qg5 Qxg5 and a draw was agreed in H.Walsh-A.Bubir, correspondence 2008. Otherwise play might have continued 29 … Qxg5 30 fxg5 Rxb3 31 Rxa5 with equal chances in the ending. 23 … Nxd5 24 Nxd5 Be6 Or 24 … Bf6 25 Ra2 Rg8 26 Kb1 Bb7 27 Qf3 Rg7 (a fianchettoed rook makes an odd impression; the rook is out of play, yet it also keeps the black king safe) 28 Qe3 Bxd5 29 Bxd5 Qc7 30 Rh3 Kg8 31 Ra4 h5 32 Ka2 Rb4! (the rook can’t be touched) 33 Ra3 Rb5 with equal chances, J.TrusewiczM.Glazman, correspondence 2005. 25 Qd4+ After 25 Ra2 Bxd5 26 Qxd5 Qf6 27 Qd4?! (White should keep queens on the board with 27 Qd2) 27 … Qxd4 28 cxd4 Bd8 Black is a pawn up in the ending, A.Mista-A.Shirov, Warsaw (rapid) 2008. 25 … Kg8 26 Kb1 If 26 Ra2 Bxd5 27 Rah2 Bxc4 28 Rxh4 Qb6! 29 Qxb6 Rxb6 30 bxc4? (30 Rxh7! is correct; e.g. 30 … Rxb3! 31 Rh8+ Kg7 32 R1h7+ Kf6 33 Rxf8 Rxc3+ with winning chances for Black) 30 … h5 and White is busted in the ending, down a pawn and with the inferior structure, I.Aliaga Fernandez-V.Feliciano Ebert, Araruama 2011. 26 … Bxd5 27 Qxd5 One feature of this line is that it often evolves into an opposite-coloured bishops situation. Now Black’s extra pawn isn’t all that valuable. The deciding factor will be king safety, enhanced by the opposing bishops. My slight preference is for Black. 27 … Bf6 28 Kc2

This move was new. Yes, it’s depressing that theory must be learned until almost move 30 these days, but this is the world we live in. 28 Rxa5 was seen in K.Lagno-M.Voiska, Turin Olympiad 2006. I slightly prefer Black’s attacking chances after 28 … Bxc3 29 Ra6 Ra8! 30 Rxd6 Qb8. 28 … Qb6 Threatening to infiltrate on f2. 29 Rf1 Qe3 30 Qd3 Qc5 31 Rd1!? An optimistic frame of mind makes everything look ten times better than it actually is. Also, when we have an iron will to win to the point of addiction, quitting cold turkey is actually easier than cutting back. Is Jakovenko playing for the full point? Maybe White should repeat with 31 Qd5. 31 … Rb6?! More accurate was 31 … Qf2+ 32 Qd2 Qf3. 32 Qg3?!

After 32 Qf3! Kg7 33 Rd5 Qg1 34 Raxa5 White stands slightly better. 32 … Ra8 If 32 … Qc7 then 33 Rd5 Qe7 34 Qe3 Rb7 35 Rdxa5 and I prefer White. 33 Rd5 Qc7 34 Rf5 34 Rdxa5? Rxa5 35 Rxa5 Rxb3 36 Bxb3 Qxa5 leaves White a pawn down with an exposed king. 34 … Qd8 35 Rd5 Qc7 Getting out of the way of e4-e5. 36 Qh3 Ra7 White has full compensation for the missing pawn, but no advantage. 37 Qe3 Rb8 38 Qd3 Rd8 39 Qd2?! The balance is maintained with 39 Ra1. 39 … Kg7 Also strong is the immediate 39 … h5! 40 Qe3 h4 with an edge for Black. 40 Rd3 h5! 41 e5!? Jakovenko doesn’t want to just sit around and wait with 41 Rd5 and see how Black proceeds. 41 … dxe5 42 fxe5 Be7 43 Qf4 Attacking f7. 43 … Rxd3! 44 Qxf7+ Kh6 45 Bxd3

Exercise (combination alert): Find Shirov’s tactic and Black gets a winning position. 45 … Qxc3+! Answer: Queen sacrifice/discovered attack. 46 Kxc3 Bb4+ 47 Kd4 Rxf7 Black’s two connected passed pawns are decisive, since White’s e-pawn is blockaded. 48 e6 Rg7 49 Ra1 g5 50 Rf1 g4 51 Ke4 Be7! 52 Be2 h4 53 Rf5 g3 0-1

There is nothing to be done about the coming … g3-g2. If 54 Rh5+ Kg6 55 Bf3 then 55 … g2! overloads White’s bishop. Game 3 Z.Hracek-A.Mista Czech League 2006 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 Remember: if you don’t want to enter the position we reach below, I have a safe out for you on the 17th move with 17 … Be6, which we look at later in the chapter. 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 Nce3 g6 18 h4! Unfortunately for us, this gambit hasn’t been consigned to history’s dumpster of unsound lines. White’s sacrifice is both sound and dangerous. 18 … Bxh4 19 g3 Bg5

19 … Bf6 is playable and will be analysed (by transposition) in the next game. 20 f4 exf4 21 gxf4 Bh4+ 22 Kf1 This move is a paradigm-altering idea. This position may be the single most difficult challenge for the Sveshnikov side in the entire book. What is the difference between placing the king on f1, or on d2, as we saw last game? 1. On d2, White’s king is considerably safer, since it will scurry away to the safety of the queenside. 2. The d2 plan is slower, since White’s plan of shifting the king across requires a huge investment of time. 3. On d2 the king obstructs White’s a4-rook, which wants to swing over to h2, via a2. 4. White’s attack is faster if his king moves to f1. White’s only major problem is that Black can easily pry open the centre and kingside with a push of the f-pawn. This means that White’s king may end up just as unsafe as Black’s.

22 … f5 23 Ra2 White’s intent becomes obvious: he ignores the opening of the f-file and plans to stick this rook on h2. 23 exf5 is illogical, even though Komodo thinks it’s playable. 23 … Bxf5 24 Nxf5 Rxf5 25 Qg4 Bg5 26 Kg2?? (26 Bd3! Rxd5 27 Bxg6 Rb7 is just an unclear mess, rated at dead even by Komodo) 26 … Bxf4! 27 Nxf4 Rg5 (there goes White’s queen) 28 Nxg6+ Kg7 0-1 was A.Delchev-T.Kotanjian, European Championship, Kusadasi 2006.

23 … fxe4! 23 … g5?! scores horribly for Black. For example, 24 Rah2 (24 Qh5 fxe4 25 Rah2 Ne5 26 Ke2 Nxc4 27 fxg5! Rxb3 28 Rxh4 Rb2+ 29 Nc2! Bf5 30 g6 and White had a winning attack, E.Lobanov-W.Schneider, correspondence 2015) 24 … fxe4 25 Qh5 Rb7, V.Grigoryev-R.Nickel, correspondence 2014. We shouldn’t follow an IM’s or GM’s suggestion, simply because they have a couple of letters before their name. If you feel that Black’s defence is stretched too thin, you may be close

to being right! Objectively, White stands better in this irrational mess. The more booked-up side is going to win this one, so comp this position to death before you enter it. 24 Rah2 g5 This is an amazing move and I use the word “amazing” without any positive connotation! Black must reinforce the h4bishop. This is obviously a highly double-edged position, where either side can go down in an instant. 25 Ng2 White breaks down the clogging of the h-file. Komodo assesses the position as dead even. Be careful though. The game may be objectively even but I get the feeling that it is easier to play White than Black. Instead, 25 Qh5! Rb7 transposes to the note above. I will repeat my earlier warning: with the comp’s help, master every nuance of this position before agreeing to enter it. 25 … Rb7 26 Nxh4 gxh4 27 Rxh4 Rg7 28 Qh5?! This natural move was played in both games to reach this position, yet it may not be best. The untried 28 Ke1! may be White’s only path to equality. After 28 … Bf5 29 Rh6 it’s anybody’s game. 28 … Bf5!

Black’s has turned the position from hell into the position from not-hell. After this move the assessment turns in Black’s favour. Once again, theory ends near move 30. As indicated earlier: this is a theoretically difficult repertoire to master. Now Black gets a Transformer-like deconstruction/reconstruction favourable for his side. Instead, 28 … Be6 allows 29 Qxh7+! Rxh7 30 Rxh7+ Kg8 31 Rh8+ Kg7 32 R8h7+ Kg8 33 Rh8+ Kg7 with perpetual check, J.Radulski-V.Spasov, Panagyurishte 2012. 29 Ne3?! White’s best chance lies in 29 Qh6! a4! 30 bxa4 Qb8 31 Bb5 Rff7! 32 Rg1! (32 Bxc6?? Qb1+ 33 Kf2 Qc2+ 34 Ke1 Rg2 and White is mated) 32 … Rxg1+ 33 Kxg1 Ne7 34 Nxe7 Qa7+ 35 Kh2 Qxe7 36 Qg5 Qxg5 37 fxg5 Rc7 with the slightly better chances for Black. 29 … Bg6 30 Qd5 Ne5! Remember this trick to activate the knight. 31 Ke2 Nxc4

White should be okay after this exchange. Black can perhaps try 31 … Nf3 32 Rg4 Qf6! 33 Qxa5 Ne5! 34 Nd5 Qe6 35 Ne3 d5! with some advantage in the wild complications ahead. 32 Qxc4 Qf6 33 Kd2 d5! Of course White doesn’t dare take the pawn and open lines for the black rooks. 34 Qd4!? The decision to enter the ending is understandable yet probably not best, since White will be handicapped by broken pawns. Furthermore, Black’s bishop may turn out to be worth more than White’s knight. Instead, 34 Qc5 Rd8 35 f5 d4 36 cxd4 Qxd4+ 37 Qxd4 Rxd4+ 38 Ke2 Be8 39 Ng4 Bd7 40 Nf6 Bxf5 41 Rh5 Bg6 42 Rh6 should hold the draw.

34 … Qxd4+ As you wish. The ending is not going to be all that easy for White to hold.

35 cxd4 Rb7 36 Rh6? Threatening 37 Rxg6, which is easily sidestepped. Correct was 36 f5! Bxf5 37 Rf4 Rbf7 38 Ra1 and White should still hold the game with accurate play. 36 … Kg8 37 Rf1 This makes matters worse. After 37 Nxd5 Rf5 38 Ne3 Rxf4 White won’t hold the game either. He should have tried 37 f5! Bxf5 38 Ra6 Bg6 39 Rxa5 Rxb3 40 Nxd5 Rf2+ 41 Ke1 Rfb2. White is still losing here, but it’s his best chance. 37 … Rxb3 38 f5 Rd3+ 39 Ke2

Exercise (combination alert): What is Black’s strongest continuation? 39 … Rb8!

Answer: Ignore the hanging bishop on g6 and play for a deadly rook check on b2. 40 Nd1 It isn’t easy to maintain our self-esteem when we discover we are the butt of the position’s joke. 40 fxg6 is met by 40 … Rb2+ 41 Nc2 Rxc2+ 42 Ke1 Rc1+ 43 Ke2 Rxf1 44 gxh7+ Kh8 45 Kxf1 Rxd4 which is an easy win for Black. 40 … Bf7 Not the most accurate move. 40 … Kg7! 41 Rhh1 Be8 42 f6+ Kg8 is easy for Black. 41 Rg1+ Kh8 42 Rgh1 Bg8 43 Ra6 Rf3 43 … Rxd4 44 f6 Rf8 wins. 44 f6 Rf8 45 Rh6 a4! 46 Rxa4 R8xf6 47 Rxf6 Rxf6 The rest is simple, with two extra pawns. 48 Ne3 h5 49 Ra5 h4 50 Ra1 Be6 51 Rh1 h3 0-1 Game 4 R.Ponomariov-V.Kramnik Wijk aan Zee 2005 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Nc2 0-0 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 Ra2

This time White replaces b2-b3 with Ra2; again he plans eventually to swing the rook over to h2. Usually it makes no difference and the two lines transpose, as indeed occurs in this game. 16 … Kh8 Black’s main counterplay lies in an … f7-f5 break, so his king makes way for it. 17 Nce3 g6 18 h4! Bxh4 19 g3 Bf6!? This time Black’s bishop moves to f6, rather than g5, as we saw in the previous game, since on g5 we invite White to play f2-f4. After 19 … Bg5 20 f4 exf4 21 gxf4 Bh4+ 22 Kf1 f5 23 b4! fxe4 24 Rah2 g5 25 b5 Ne5 26 Qd4 Rb7 27 Rxh4 gxh4 28 Ke2 Kramnik went on to win brilliantly, although the computer at this point calls it equal, V.Kramnik-L.Van Wely, Monte Carlo (rapid) 2005. This game is annotated in Kramnik: Move by Move. 20 b3

Instead, 20 f4 exf4 21 gxf4 Bg7 22 b3 comes to the same thing; and here with 20 b3 we reach the same position as after 16 b3 Kh8 17 Nce3 g6 18 h4 Bxh4 19 g3 Bf6 20 Ra2. 20 … Bg7 21 f4 exf4 22 gxf4 Re8 23 Qf3 The e4-point is supported, while White prepares for Rah2. 23 … Kg8 24 Rah2 h5 25 Rxh5!?

From an objective standpoint, the comp calls the game dead even, so it is open for either side to win or lose. 25 Qg2 is a more sober path, as in K.Bulski-A.Leniart, Polish Championship, Warsaw 2010. Komodo suggests 25 … Ne7 26 Nxe7+ Qxe7 27 Qxg6 Kf8 28 Qxh5 Be6 29 Nd5 Bxd5 30 Bxd5 Rxb3! 31 Bxb3 Qxe4+ 32 Kf1 Qe1+ 33 Kg2 Qd2+ 34 Kg3 Qxc3+ 35 Kg4 Qxb3 36 Qxa5. I don’t believe White stands better, despite the comp’s favourable assessment. 25 … gxh5 26 Qxh5 Re6! 27 Qh7+ Kf8 28 Rg1 Rg6 With this key move Kramnik saves himself. So is it time for White to resign? No, not quite. The computer thinks White still

has full compensation for the sacrificed rook. 29 Rxg6 fxg6 30 f5!? Safer is to bail out to an immediate draw with 30 Nc7! Ne7 31 Ned5 Ng8 32 f5 Qg5 33 Nf4 Qg3+ 34 Ke2 Qg4+ 35 Ke3 Qg3+ 36 Ke2 and perpetual check. 30 … Qg5

31 Nf4! Threatening 32 Qg8+. 31 … Ne7 31 … Bxc3+!? is insanely risky. The comp works it out to a probable draw after 32 Ke2 Ke8 33 Qf7+ Kd8 34 Qg8+ Kc7 35 Nfd5+ Kb7 36 Qf7+ Ka8 37 Nc7+ Kb7 38 Bd5 Kb6 39 Bxc6 Kxc6 40 Qc4+ Kb6 41 Qxc3 Qh5+ 42 Kd3 gxf5 43 Ncd5+ Ka7 44 Qc7+ Rb7 45 Qxc8 fxe4+ 46 Kxe4 Qh7+. 32 Nxg6+ Ke8!

I strongly urge you to get a pet rat and study its movements if you decide to play this line for Black, since we must master a rodent-like scurrying quality of how to defend and run, without getting mated! After 32 … Nxg6?? 33 Qg8+ Ke7 34 Nd5+ Kd7 35 Qxg7+ Kd8 36 Qc7+ Ke8 37 fxg6 Qg3+ 38 Kd2 Qg5+ 39 Kd3 Qg3+ 40 Ne3 there is no perpetual check and Black loses. 33 Nd5 Qg1+ Or 33 … Qc1+ 34 Kf2 Qd2+ 35 Kf3 Bxf5! 36 Qxg7! Bxe4+ 37 Kxe4 Qg2+ 38 Kf4! Nxg6+ 39 Ke3 Qg5+ 40 Kf2 Qh4+ and perpetual check is the game’s natural result. 34 Kd2 Qg5+ 35 Ke1 Bf6!? Kramnik plays for the win. He has a draw with 35 … Qg3+. 36 Ngxe7 Bxe7 37 Qh8+ Kd7 38 f6 This is not the kind of position that forgives an accounting error, so if you play for the win, be certain you are getting what you think you are getting.

Exercise: (critical decision): Black can take perpetual check with either a queen check on c1, g1 or g3. But is there a win? Examine 38 … Bd8 and 38 … Bxf6. Does one or both of these lines win for Black? What is your decision? 38 … Qg3+! Answer #1: Taking an immediate perpetual check is Black’s wisest course. a) 38 … Bd8?? loses to the study-like shot 39 Bb5+!! (we catch a glimpse of the creature and see that it isn’t entirely human; note that the immediate 39 Qg7+? doesn’t work for White due to 39 … Qxg7 40 fxg7 Bb7 41 g8Q Bh4+ 42 Kd2 Rxg8 43 Nb6+ Kc7 44 Bxg8 Kxb6 and White must resign) 39 … Rxb5 40 Qg7+! (how bizarre; White is down a rook and a piece, yet forces a queen swap, which wins!) 40 … Qxg7 41 fxg7 and White promotes and wins. Answer #2: b) 38 … Bxf6 also leads to a draw, but here it is Black who must be careful to avoid tricks. So from a practical standpoint this line is inferior to taking immediate perpetual check, as Kramnik did in the game. After 39 Nxf6+ Kc6 (not 39 … Kc7?? 40 Nd5+ Kd7 – moving to c6 is met by a deadly queen check on e8 – 41 Qh7+ Ke6 42 Ne3+ Ke5 43 Qh2+ Kxe4 44 Bd5+!, Black must hand over the queen or get mated) 40 Bd5+ Kc7 41 Ne8+ Kd7 (not 41 … Kb6?? 42 Qd4+ Ka6 43 Nc7 mate) 42 Nf6+ with perpetual check. 39 Ke2 Qg2+ 40 Ke3 Qg3+ 41 Ke2 Qg4+ 42 Kf2 Bd8 43 Bf1 Kc6 44 Qxd8 Qh4+ 45 Kg1 Qg3+ ½-½ Game 5 T.Wilczek-D.Lafarga Santorroman Correspondence 2012 Is there such a thing as an error-free game, where one side – or

even both sides – play perfectly? Correspondence games come the closest, since there is plenty of time for each move and we are allowed a sidekick 3000+ rated computer, running day and night. 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 Nce3 g6 18 0-0

Sometimes White is just not interested and wants to play it safe against the Sveshnikov. Here White decides against crazy h2-h4 sacrifices. He just castles and hopes to make use of the d5-square and Black’s potentially weak d6-pawn. Black’s counterplay lies in the standard … f7-f5 break and going after the white king, in King’s Indian style, just as Black does in this game. 18 … f5 19 Qd3 Others:

a) 19 exf5 (I love it when White trades, since it hands us the centre and play along the g-file) 19 … gxf5 20 Qh5 (or 20 f4 exf4!, when White must avoid 21 Nxf4?? Qb6 22 Qe2 Bxf4 23 Rxf4 Re8 24 Rf3 Ne5 and wins) 20 … e4 21 f4 exf3 22 Rxf3 Ne5, A.Gypser-A.Flumbort, German League 2005. White scores poorly here, since Black’s bishops come alive. b) 19 Qf3 Be6 20 Rd1 Qd7 21 h3 was E.Riccio-D.Lafarga Santorroman, correspondence 2009. Black’s chances look no worse after 21 … Qf7. 19 … Be6 20 Rd1 Qd7 21 f3 Qg7 Another option is 21 … f4 22 Nc2 Bd8!, covering a5 and b6. If White’s d5-knight ever moves, Black’s dark-squared bishop can emerge on b6. The chances are dynamically balanced, S.Glukhovtsev-R.Starke, correspondence 2015. 22 Kh1 Bd7 23 Ra2 f4

Black releases the tension and plays directly for mate in King’s Indian fashion. But unlike a King’s Indian Black’s dark-

squared bishop is unopposed. For example, it can later try and slip into play via d8. 24 Nc2 Bd8 There are two ideas behind this move: 1. The bishop clears the way for a kingside pawn storm with … g6-g5-g4. 2. If White’s d5-knight ever finds a spot to relocate then Black’s dark-squared bishop can occupy a dangerous diagonal from b6. 25 Na3 g5 26 Nb5 Nobody can claim that Black’s d-pawn is an attractive child. Note how White is unable to apply real pressure against d6, which is weak only in name and can be easily defended. As always with the Sveshnikov, Black is willing to sacrifice comfort in the present, by taking on a d5-hole and a backward dpawn, to secure the welfare of a happy future, based around compensating piece activity. This line has yet to fall into theoretical disrepute and I predict it never will. After 26 h3 Qh6 (target: h3; … g5-g4 is in the air) 27 Kg1 Na7! 28 b4 axb4 29 cxb4 Qg7 30 Kf2 h5 31 b5 Nc8 chances are balanced between White on the queenside and Black on the kingside, K.Koegler-C.Jaulneau, correspondence 2009. 26 … Qh6 Covering d6, while preparing … g5-g4. 27 Rad2

A novelty, but no improvement. He should stick with 27 Qe2 g4 28 fxg4 Qh4 29 Qf2 Bxg4 30 Qxh4 Bxh4 31 Rf1 Rbd8 32 Ndc7 Rf6 33 h3 Bc8, I.De la Pena Balbuena-A.Albano, correspondence 2011. The players agreed to a premature draw here, where I slightly prefer White’s chances, since d6 is of concern. 27 … g4 Black achieves his thematic break. 28 Qf1 g3 Black decides to resolve the pawn tension rather than just wait. 29 h3 Rb7 If Black gets an eternity, then he plays … Bc8, … Rg7-g5h5, and … Rfg8, followed by … Bxh3. 30 Ra2 Bc8 31 Kg1 White’s king gets nervous and plans to run via f1.

31 … Rg7 32 Rc2 White finds that his king is unable to run, since h3 requires perpetual defence by the white queen on f1, which blocks the king’s route to freedom. 32 Qd3 Rg6 33 Kf1? fails to 33 … Bxh3 and if the sacrifice is accepted with 34 gxh3?? then 34 … g2+ wins. 32 … Bd7 33 Ra2 Rfg8! 34 Qe2? Our inaccuracies, errors and blunders can never be transcended. They can only be reduced. Even Capablanca and Fischer blundered – just not very much! Outwardly, White’s configuration looks healthy, while below lies the inner rot. Instead, 34 Rdd2! discourages Black from the following piece sacrifice. 34 … Bxh3!

One cannot accurately describe Black’s not-so-difficult-tosee sacrifice as one of unsurpassable beauty, yet nobody can deny its effectiveness. 35 Kf1!

Acceptance is out of the question. After 35 gxh3?? g2 nothing can be done against the coming … Qxh3 and … Qh1+. 35 … Rg5 36 Qd3 a4! This game is a constantly shifting entity, without even a single permanent element. This unexpected move from the other wing is based on the following ideas: 1. If White’s rook takes the pawn, then g2 is left insufficiently protected. 2. If the b-pawn captures, then Black has … Na5!, followed by … Nxc4, which destroys White’s light-square power. 3. If White bypasses with b3-b4 then, weirdly enough, Black gets a passed a-pawn, which appears to be meaningless. Yet White simply doesn’t have time to take it. This passed a-pawn won Black the game later on. 37 Ke1?! Instead: a) 37 b4 is more accurate, when 37 … Bxg2+! 38 Rxg2 Rh5 39 Ke1 transposes to the game. b) 37 bxa4? Na5! White is busted with the removal of his light squared bishop. c) 37 Rxa4?? (it takes hours to play a game and a single careless second to blow it to smithereens; this rook is needed to cover g2) 37 … Bxg2+! 38 Ke1 Bxf3! 39 Qxf3 g2 wins. 37 … Rh5 Here 37 … Na5! or 37 … axb3 38 Bxb3 Na5! is also an idea, since it removes a lot of White’s power on the light squares. 38 b4

Exercise (critical decision): How should Black continue? 38 … Bxg2! Answer: Annihilation of defensive barrier. 39 Rxg2 Rh2! Black’s threats clarify. Not 39 … Rh1+? 40 Kd2 Qh3 41 Qe2 White survives. 40 Qc2 The only move. a) 40 Qe2?? is met by 40 … Qh3! 41 Rd2 Rh1+, winning the queen. b) 40 Qf1?? Rh1 41 Rg1 g2 42 Qf2 Bh4 pins and wins White’s queen again. 40 … Rh1+ 41 Bf1

Not 41 Kd2?? Qh3, which is similar to what occurs in the game. 41 … Rxf1+! Black’s attack rages. 42 Kxf1 Qh1+ 43 Rg1 Qxf3+ 44 Ke1 Qh5 Black’s deeply passed f- and g-pawns are worth more than White’s rook.

45 Kd2 White’s king continues to run, as if in an opiate-induced haze. 45 … f3 It isn’t so difficult for Black to pilot his f- and g-pawns up the board. 46 Kd3 f2 47 Qe2 White has no choice but to hand over a full rook. 47 Rgf1 Qf3+ 48 Ne3 d5! 49 exd5 e4+ 50 Kd2 Bg5 is game over.

47 … Qxe2+ 48 Kxe2 fxg1Q 49 Rxg1 h5 Black’s simple plan is to push his passed pawns. Nevertheless, 49 … Ne7! is more accurate, and if 50 Nxd6? then 50 … Nxd5 51 exd5 Bb6 52 Rg2 a3! wins. 50 Nxd6 h4 51 Kf3 Ne7! 52 Nxe7 Bxe7 53 Nf5 Bd8 54 c4

Exercise (combination alert): White’s passed pawns are too slow. Prove why. 54 … g2! Answer: Sacrifice the g-pawn, after which Black’s a-pawn is too fast. Whereas 54 … h3?? allows 55 Rh1! and it is Black who is busted. 55 Rxg2 A slow, lingering death is a mournful process. There is no choice, since … h4-h3 can’t be allowed.

55 … Rxg2 56 Kxg2 Be7!

56 … a3 also wins; e.g. 57 Ne3 a2 58 Nc2 Bg5 59 b5 Be3! etc. 57 Kh3 Or: a) 57 Nxe7 a3 and the passed pawn slips outside the knight’s reach. b) 57 c5 a3 58 Ne3 a2 59 Nc2 Kg7 60 Kh3 Kg6 61 Kg4 Kf7 62 Kh3 Ke6 63 Kg4 Kd7 64 Kh3 Kc6 65 Kg4 Kb5 66 Kh3 Kc4 67 c6 Bd8 68 Kg4 Kc3 wins. 57 … Bxb4 58 Ne3 a3 0-1 After 59 Nc2 a2 60 Kxh4 Kg7 61 Kg5 Bc3 62 c5 Kf7 63 c6 Ke6 White’s passed pawn isn’t going anywhere, while Black’s will promote, costing White his knight. Game 6 A.Pauli-S.Kalinitschew

Fürth 1998 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 This is the Kalashnikov move order. We soon transpose to Sveshnikov with the inclusion of … Nf6 and Nc3. 5 Nb5 d6 6 N1c3 a6 7 Na3 b5 8 Nd5 Nf6 9 Bg5 Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Nc2 0-0 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 Nce3 Be6

If you weren’t crazy about Black’s position in the previous few games of this chapter, I’m going to show you a way to dodge it with a more sanitized version, where we insist on keeping things simple. The idea behind this move is to avoid … g7-g6, until White commits to castling. Then White no longer has access to scary h2-h4 ideas. Another option is to play 17 … Ne7. For example: 18 Nxe7 (here 18 h4 Bxe3 19 Nxe3 Bb7 already seems better for Black) 18 … Qxe7 19 Nd5 Qd8 20 0-0 (20 Qa1 f5 21 exf5 Bxf5 22 Rxa5 Bc2! 23 Ra3 Be4 24 0-0 Bd2 seems to give Black

sufficient counterplay) 20 … Bd7 21 Ra2 f5 22 exf5 Bxf5 23 Qe2 Rc8 24 Rfa1 Rc5 25 b4 axb4 26 cxb4 Rc8 27 b5 Be6 28 Ra6 Rc5 29 Nb6 d5 30 Ra8 dxc4! 31 Rxd8 Rxd8 32 Na4 Rxb5, J.Jwan-R.Mamedov, Adana 2006. The confused comp once again misassesses. White’s game looks a lot better than it actually is. Only Black can win from this position (he did!). 18 0-0 White plays strategically, seeing the h2-h4 idea doesn’t come with anywhere near the punch it did as when Black tossed in an early … g6-g6, as without that White’s h4-h5 lacks a contact point on the g6-square. Nevertheless, 18 h4 is still possible and is considered in the next two games. 18 … g6 Good things come to those who wait. Now the move is without danger for Black, since White does not have the prying h2-h4 mechanism. 19 Kh1 This way White can meet Black’s coming … f7-f5 with f2f3, without immediate fears on the a7-g1 diagonal. Also possible is 19 Qd3 f5 20 Rd1 Qd7. One of the curious features of our opening is that while, visually, d6 looks weak, in reality it is fully defendable. For example, here White’s d5knight, which looks so pretty, actually obstructs pressure on d6. Now 21 f3 is similar to the main game. 19 … Bh6 Black wants to avoid 19 … f5 20 exf5 gxf5 21 f4. 20 Qd3 f5 21 f3 After 21 exf5 gxf5 22 f4 Qd7 23 Ra3 Qf7 24 g3, S.BogicN.Vajdic, Banja Junakovic 2009, the game is sharp and probably dynamically balanced, despite the Sveshnikov-hating comp’s slightly favourable assessment for White. 21 … Bd7 22 Raa1 Qh4 This move threatens nothing for now, yet it’s psychologically disturbing for the defence to have a queen in the vicinity of the king. 22 … Be6 is a decent alternative.

23 b4 White wants to create a passed b-pawn which, however, isn’t really going anywhere. If instead 23 Rad1 fxe4 24 fxe4 then 24 … Ne7! and, amazingly, Black’s d6- and a5-pawns remain relatively safe and the game is dynamically balanced. 23 … axb4 24 g3?

As mentioned previously, it can drive the opponent crazy when our queen hovers near their king. White should have refrained from this weakening move, which doesn’t even chase the queen away. 24 cxb4 Nd4 25 Nc2 Nxc2 26 Qxc2 Be6 is equal. 24 … Qh3 25 Nxb4 Nxb4 26 cxb4 Rxb4 Black wins a pawn – but 26 … f4! is even stronger. After 27 gxf4 Bxf4 28 Rf2 Bxe3 29 Qxe3 Rxb4 30 Bf1 Qh4 White is busted. 27 Rfb1

Not 27 Qxd6?? Rb2 and White resigns. 27 … Rxb1+ 28 Rxb1 f4 Also strong is 28 … fxe4 29 Qxe4 Bxe3 30 Qxe3 Bc6 31 Be2 Qd7 with excellent chances for Black to convert.

Exercise (critical decision): It is plain that White is in deep trouble. Should he retreat with 29 Nf1, or should he go for it with 29 Qxd6 ? Pick the line which maximizes White’s survival chances. 29 Qxd6? Principle: don’t “solve” a problem when, in doing so, you create another greater problem. All openings have their sets of laws and by-laws. You can violate the latter – sometimes. Violate the former and the result is usually fatal. In this instance

White violates a key law when facing the Sveshnikov: Don’t take liberties with king safety. His counter-attack is about to end, before it even starts. Answer: White’s most stubborn path of resistance lies in 29 Nf1!. After 29 … fxg3 30 Nxg3 Be6 31 Bxe6 Qxe6 32 Rb6 Rd8 Black is still a long way from converting the extra pawn. 29 … fxe3!

Now White’s king is thrown into a windowless, lightless cell. Whereas 29 … fxg3?? throws the win away. After 30 Qxe5+ Bg7 31 Qxg3 Qxg3 32 hxg3 Rxf3 33 Kg2! White’s hanging knight isn’t really hanging and the game should end in a draw. 30 Rb7 After 30 Rb8 Bc8! 31 Qxe5+ Bg7 Black will convert the extra piece. 30 … Rxf3! 31 Qb8+ Unfortunately, 31 Qxd7 Rf1+ 32 Bxf1 Qxf1 is mate. 31 … Rf8 32 Qxe5+ Bg7

White’s queen hangs, as does a mating threat on f1. 33 Rb8 White had no better option. We feel an ironic sense of relief when our position hits rock bottom, since we realize it can’t get any lower. The worst has already occurred and things can either remain the same or, happily, get better. In this case, there is no “better” coming White’s way. 33 … Bxe5 34 Rxf8+ Kg7 0-1 Game 7 S.Barrientos Chavarriaga-J.Fluvia Poyatos Badalona 2005 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Nc2 0-0 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 Nce3 Be6 18 h4

This move hits with far less effect if Black withholds … g7g6, since then h4-h5 hits empty air, rather than a contact point. 18 … Bxe3 19 Nxe3 Qb6!? Warning: Those who love a clean, orderly world should avoid playing the Sveshnikov. This pawn offer intrigued me since Black won both games in my database. You can also play it safe and go with 19 … Ne7 20 0-0 Bc8! 21 Qd2 f5 22 exf5 Nxf5 23 Nxf5 Bxf5 24 Rfa1 Be4! 25 Be6 (after 25 Rxa5 Qxh4 26 Bd5 Bxd5 27 Rxd5 Rxb3 28 Rxd6 h6, the game is heading for a draw) 25 … Bb7 26 h5 Rf6 and Black stood only slightly worse, N.Latronico-A.Hebels, correspondence 2011.

20 0-0 White plays it safe and refuses the offer of the d6-pawn. In the next game we examine the critical 20 Qxd6. 20 … Bxc4 Undermining the defender of b3. 21 Rxc4 Both 21 Nxc4 Qxb3 22 Nxd6 Kg8 and 21 bxc4 Nd8! 22 Qd5 Nb7 are also okay for Black. 21 … Ne7 Black covers against both Nd5 and Nf5. 22 h5 22 b4 is met by 22 … Rbd8. 22 … f5 Something with an unintimidating outer appearance doesn’t mean it lacks power. Don’t underestimate the counterplay that

inevitably arises from the thematic … f7-f5 break. 23 h6!? This is risky, albeit playable. Instead, 23 exf5 Nxf5 24 Nxf5 Rxf5 25 b4 Rf6 leads to an equal position. 23 … gxh6!

The Sveshnikov often rewards “unclassical” play. Taking the pawn is correct since Black gets counterplay along the g-file. Whereas 23 … g6?! weakens Black’s king and back rank. White stands better after 24 b4. 24 Qh5 fxe4 25 Rxe4 Ng6 26 Rg4 Now Rxg6 is in the air, though it is easy for Black to defend against this. 26 … Rg8 27 Nf5 White decides to abandon his queenside pawns to lunge at Black’s king. The comp calls it even. 27 … Rbf8

Alternatively, 27 … Qxb3 28 Qxh6 Qe6 29 Rg5 Nf4 30 Rxg8+ Rxg8 31 Qxe6 Nxe6 32 Nxd6 is a likely draw. 28 Nxh6 Rg7 29 g3 d5?! Black undoubtedly wanted to meet Rxg6 ideas with … Qxg6, but it may not be wise to loosen the centre like this. Instead, 29 … Rc7! maintains the balance after 30 b4 axb4 31 cxb4 Rxf2! 32 Rxf2 Rc1+ 33 Kg2 Qc6+ 34 Re4! Qxe4+ 35 Qf3 Qxf3+ 36 Rxf3 Rb1 37 Nf7+ Kg7 38 Nxd6 Rxb4 39 Rf7+ Kh6. Black won’t be able to convert the extra pawn. 30 Kg2?! 30 Rg5! e4, Intending … e4-e3, is met by 31 Rf5 with pressure for White. 30 … Qf6 31 f3? 31 Kg1! e4 32 Rg5 (intending Rf5) 32 … Ne7 33 Rxg7 Qxg7 34 Kh2 e3 35 Ng4 Rf5 36 Qe8+ Rf8 37 Qh5 Rf5 is a repetition draw. 31 … e4! 32 f4 e3?! Correct was 32 … Qxc3! 33 Nf5 Qd2+ 34 Rf2 Qd3! 35 Nxg7 e3 36 Rf1 Qd2+ 37 Kh3 e2 38 Rf2 Qe1! 39 Rxe2 (forced) 39 … Qxe2 and White loses his knight, since 40 Nf5?? is met by 40 … Qf3! (threatening mate on h1) 41 Kh2 Qf2+ 42 Kh3 Qf1+ 43 Kh2 Ne5! wins; e.g. 44 fxe5 Rxf5 45 Qe8+ Rf8 46 Qh5 Rf2 mate.

Exercise (critical decision): Black’s last move sets up a sneaky trap. White’s choice is between initiative or material. Should he play 33 Qxd5, or push for the initiative with 33 f5 - ? One line should hold the game, while the other loses. Which one do you pick? 33 Qxd5?? Okay then, let’s just stick a finger into the live light socket. White misinterprets the poisoned d-pawn as a shining symbol of a second chance. Answer: 33 f5! should save the game for White. After 33 … Qxc3 (33 … Ne5 34 Rxg7 Qxg7 35 f6! Qc7 36 Ng4 is also okay for White, while after 34 … Kxg7 35 Ng4 Nxg4 36 Qxg4+ Kh8 37 Qd4 Qxd4 38 cxd4 the resulting ending should be a draw) 34

Rxg6 Qd2+ 35 Kh3 e2 36 Rf2 e1Q 37 Rxd2 Qxd2 38 Rxg7 Kxg7 39 f6+! Rxf6 40 Nf5+ Kf8 41 Qxh7 Qg5 42 Qe7+ Kg8 43 Qd8+ Kh7 44 Qe7+ Kg6 45 g4! Rxf5 46 Qe6+ Kg7 47 gxf5 with a drawn queen ending. 33 … Qxc3 The Promised Land is nearby for Black’s e-pawn. The threat is 34 … e2, to which, surprisingly, there is no remedy. 34 Rxg6 If 34 Qf3 then 34 … Qd2+ 35 Kh3 e2 and the e-pawn pushes through. 34 … Qc2+ 0-1 Game 8 V.V.Palciauskas-J.Bennborn Correspondence 2007 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 Nce3 Be6 18 h4 Bxe3 In the next game we’ll analyse a variation where Black preserves this bishop by retreating it to h6. 19 Nxe3 Qb6 20 Qxd6

This is the crucial line. White grabs the pawn and says “show me”. This is one of those lines where the comp gets fooled. At first it has White up by a solid margin. Then as the game moves along, White’s advantage continues to diminish. 20 … Rfd8 In return for the pawn Black has: 1. A development lead. 2. The possibility to undermine support for b2 with … Bxc4. 3. White’s back rank is weak, as we will soon see. 21 Qa3 Bxc4! 22 Rxc4 The only serious try, since 22 Nxc4 Qxb3 is fine for Black, and 22 bxc4?? loses a rook to 22 … Qb1+ etc. 22 … Qxb3!

Zwischenzug. The knight isn’t really hanging on c6, as White’s back rank is weak. 23 Qxb3 Rxb3 24 0-0 24 Rxc6?? Rb1+ pops the rook in the corner. 24 … Rd6 25 Nf5 Rf6 26 Rd1 h5 27 Rd6 Rb6! 28 Rxf6 gxf6 29 f3 White doesn’t have time to chase after the h5-pawn. For example, 29 Ng3 Ra6! (principle: place your rook behind your passed pawn) 30 Ra4 Ne7 31 Nxh5 Ng6 32 Ng3 (or 32 g3 Kh7 33 f4 Kh6 34 g4 exf4 35 Kf2 Nxh4 36 Nxf4 Kg5 with an almost certain draw) 32 … Nxh4 33 c4 Kg7 34 c5 Kf8 and Black stands no worse. 29 … Kh7 30 Kf2 Ra6 31 Ra4 Nd8 ½-½ Intending … Nb7-c5, breaking the blockade. They could have played on, but I would be happy to take Black. The comps finally realize that White has no advantage. The database version actually has Black winning here, for no good reason, other than that White abandoned the game. I’m going to

adjudicate the final position as a draw. Game 9 E.Bronnikova-T.Kosintseva Russian Women’s Team Championship 2006 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 Ra2 16 b3 Kh8 17 Nce3 Be6 18 Ra2 would be another route to the position below. 16 … Kh8 17 Nce3 Be6 18 b3 Ne7 Remember our rule: no … g7-g6, until White commits to kingside castling. This way the force of h2-h4 is ruled out. 19 h4 If 19 Nxe7 Qxe7 20 Nd5 Qd8 21 0-0, R.Freydl-S.Peschardt, correspondence 2005, then 21 … f5 22 exf5 Bxf5 and Black looks okay. 19 … Bh6!?

Black decides to retain the dark-squared bishop. If you are okay with a draw, you can exploit White’s “wasted” tempo on Ra2 and go for the liquidating 19 … Bxe3 20 Nxe3 Bxc4 21 Nxc4 d5 22 exd5 Qxd5 23 0-0 Qxd1 24 Rxd1 Rxb3 25 Nxe5 Rxc3 26 Rxa5 with total equality. 20 g3 Nxd5 21 Nxd5 f5 Black generates sufficient play down the f-file. 22 exf5 Bxf5 23 0-0 Rc8 As is usual in the Sveshnikov, Black is not busted, despite all our strategic liberties and the computer’s usual misassessment of an advantage for White. This way, Black accomplishes the following: 1. The advance b3-b4 is discouraged, since White’s bishop hangs on c4. 2. Black plans to play … Rc5, next, which adds protection to the a5-isolani. 3. With a rook on c5, Black adds heat to the d5-knight.

24 Qe2 Rc5 25 b4 Yes, White creates a passed b-pawn. In exchange, Black rid himself of his isolani. 25 … axb4 26 cxb4 Rc8 27 Rd1 Be6 As always in the Sveshnikov, Black’s pieces make their pilgrimage to our Mecca, which is always d5. This way Black can possibly bail out to an opposite-coloured bishops ending with a future … Bxd5. 28 b5 Rc5 The rook jumps right back into c5. 29 Bb3!? White is willing to hand over the b-pawn, if in return Black must play … Bxd5, which gives White the better of the remaining opposite-coloured bishops. Nonetheless, I would prefer 29 Rb2, following the principle: place your rook behind your passed pawn. 29 … Qb8! 30 b6 Not 30 Ra5? Qb7! 31 Qe4? Qf7! and White is busted. 30 … Bxd5 31 Bxd5 Qxb6 Black has won a pawn, yet White is fully compensated as he has the stronger bishop. In fact he soon finds a way to win the d6-pawn and the game is quickly drawn. 32 Ra6 Qc7 32 … Qb5 33 Qxb5 Rxb5 34 Rxd6 Rb2 35 g4 Bd2 36 Be4 is an almost certain draw. 33 Be4 Rd8 34 Qd3!

The d6-pawn falls. 34 … d5!? The clearest path to the draw lies in 34 … g6! 35 Rxd6 Qxd6 36 Qxd6 Rxd6 37 Rxd6 Rc7 38 Re6 Bf8! 39 Rxe5 Re7 40 Rxe7 Bxe7. White’s extra pawn is useless due to the oppositecoloured bishops. 35 Bxd5 Rc1 36 Kg2 Rxd1 37 Qxd1 g6 38 Qf3 Rf8 39 Rf6 39 Qe2 is the only way for White to try and win. 39 … Qe7 40 Rxf8+ Bxf8 41 h5 gxh5?! ½-½ We experience the Lion King circle of life, natural conclusion. White stands slightly better since Black nurses two isolanis and his king is slightly unsafe. Still, from a practical standpoint, Black’s chance to draw is 99%, next to White’s 1% chance to win. So White didn’t bother trying. And 41 … Kg7! would have ensured the draw. Game 10 W.Braakhuis-Jor.Rodriguez

Correspondence 2001 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Rb8 16 b3 Kh8 17 0-0 Only fundamentalists – true believers – opt for the crazy theoretical nightmare h2-h4 lines we looked at previously. Safe castling is what you are far more likely to see at club level, since most C-, B- and A-rated players would be nervous plunging into one of those h2-h4 pawn sacrifices. By castling White plays for the win, in relative safety. 17 … g6 In the next game we look at 17 … f5. 18 Qd3 f5

19 f3 White bolsters his centre; … f5xe4 can now be met by f3xe4.

Others: a) 19 Rd1 Bh6 20 Nde3 Rf6 21 f3 Ne7 22 Rda1 Bb7 23 Rxa5 (is Black busted? no; he gets plenty of compensating play on the kingside and e4) 23 … fxe4 24 fxe4 Rf4 25 Bd5 Bxd5 26 Nxd5 Nxd5 27 Rxd5 Qb6+ 28 Kh1 Qxb3 29 Ne3 (29 Rxd6 Qb1+! 30 Qd1 – the only move – 30 … Qxd1+ 31 Rdxd1 Rxe4 with a likely draw, although I would take Black if given a choice) 29 … Rf2 30 Re1 Rbf8 31 h3 Bxe3 32 Qxe3 Qc2 33 Rg1 Qe2 ½-½ V.Bologan-J.Degraeve, Belfort 1998. b) 19 exf5 has not been tried. After 19 … gxf5 20 Rd1 f4 21 f3 Bf5 Black has compensating play down the g-file and on the kingside to offset the weakness of d6 and a5. 19 … Bh6 We have seen this precaution before. Black wants to make certain that, if and when White plays f3-f4, it will not be with tempo on the bishop. 20 Ra2 This move was new. The position is slightly different from the previous game, where Black’s bishop was on e6. 20 Rd1 (A.Bedoya-M.Uribe, Medellin 1999) can be met by 20 … Ne7 21 Nxe7 Qxe7 22 Rxa5 (after 22 Qxd6?! Qxd6 23 Rxd6 fxe4 24 fxe4 Bb7! White is the one in trouble; if 25 Bd3? then 25 … Rfd8 26 Rxd8+ Rxd8 27 Ne1 Bd2 wins material) 22 … fxe4 23 Qxe4 Bb7 24 Qe1 e4 25 fxe4 Qxe4 26 Qxe4 Bxe4 27 Nb4 Be3+ 28 Kh1 Bf4, when Black’s bishops and piece activity offer full compensation for the missing pawn. 20 … Be6 21 Nce3 Qh4 As mentioned previously, it’s not easy for White’s peace of mind to have this queen lurking about the white king. 22 g3 22 Nc7? fxe4 23 fxe4 Bxe3+ 24 Qxe3 Rxf1+ 25 Bxf1 Bxb3 26 Rd2 Qe7 27 Nb5 a4 leaves White in deep trouble. 22 … Qd8! 22 … Qh3?! 23 exf5 gxf5 24 f4 leaves Black’s queen out of play, rather than attacking.

23 Kg2 Qd7

24 exf5 This opens the game for Black’s bishops. But White didn’t like the idea of Black playing either … f5xe4 or … f5-f4 at his leisure. 24 … gxf5 25 f4 Bg7 26 Rd1 Bf7 The idea is to swing up to h5. 27 h3 Bh5 28 Rf1 Bg6 The game is dynamically equal, despite the computer’s standard misassessment in White’s favour. 29 Bb5 The bishop was better posted on c4. 29 … exf4! 30 gxf4 Others:

a) 30 Nxf4 Qb7! 31 Bxc6 Qxc6+ 32 Rf3 Rxb3 leaves White tangled up and in trouble from Black’s bishops. b) 30 Rxf4 Ne5! 31 Bxd7 Nxd3 32 Rf1 f4! 33 gxf4 Be4+ 34 Kg3 Nc5 35 Ba4 Nxb3 is at least equal for Black. 30 … Ne5! The pin which never was. By now you know that in the Sveshnikov violence is the norm.

31 Bxd7 Nxd3 32 Ba4 Bh5 32 … Nc5 also places White on the defensive. 33 c4 Rf7 34 Bb5 Nc5 35 Rxa5 Nxb3 With advantage to Black, since his pieces – especially his bishops – breathe fire. 36 Ra3 Nd2 37 Rf2 Rg8! A new factor is added. White’s king is insecure, despite the fact that queens have been removed from the board. 38 Kh1 Ne4 39 Rfa2 Bd4

Black’s pieces are generating more and more power. 40 Kh2

Exercise (planning): How should Black continue? 40 … Rfg7!! Answer: Shift the f7-rook to g7, after which Black has a winning attack. The f5-pawn is protected tactically. 41 Rg2! If your most paranoid suspicions are suddenly confirmed, then guess what? You weren’t paranoid to begin with. 41 Rg2 is the only move. If 41 Nxf5? (principle: our absolute conviction that our move is correct doesn’t necessarily make our move correct) then 41 … Bg1+! 42 Kh1 Ra7!! (threatening 43 …

Rxa3 44 Rxa3 Nf2 mate) and White is curiously helpless, since 43 Rxa7 Bf3+ 44 Rg2 Rxg2 also leads rapidly to mate. 41 … Ng3! 42 Rf2

42 … Ne2! It’s one thing to achieve a winning position and quite another to govern it properly, so it doesn’t degenerate to its previous non-winning state. My head is swimming with confusion from the complications created by the central cluster of pieces. The threat is 43 … Ng1! and … Nf3+, when White’s choice is to hand over the exchange or get mated. 43 Ra2 After 43 c5 Ng1 44 Nf6! Bxf6 45 Nxf5 Rc7 46 cxd6 Rc3 47 Rxc3 Bxc3 Black should convert the extra piece, despite the reduced number of pawns on the board. 43 … Ng1! 44 Rg2 Nf3+ 45 Kh1 h6 Here 45 … Nh4!! 46 Rxg7 Kxg7! leaves White in a strange zugzwang. For example, after a waiting move such as 47 Bd7

Black wins with 47 … Bf3+ 48 Kh2 Kf7! 49 Bb5 Bxd5 50 Nxd5 Nf3+ 51 Kh1 Rg1 mate. 46 Rxg7 Rxg7 Threatening 47 … Rg1 mate. 47 Rg2 Ra7! Threatening White’s eternally weak back rank. 48 Nc2 Ra2! 49 Re2 Not 49 Nxd4?? Ra1+ 50 Rg1 Rxg1 mate. 49 … Bc5! 50 Be8

Or: a) 50 Bc6 Ne5! 51 fxe5 Bf3+ 52 Kh2 Bxe2 53 Ncb4 Rb2 54 Na6 Bxc4+ wins. b) 50 Nc3 Rb2 51 Ba4 Nh4! 52 Re6 Ng6 53 Rf6 Nxf4 54 Rxf5 Ne2! 55 Nxe2 (55 Rxh5 Nxc3 wins a piece) 55 … Bxe2 and White is tangled up and busted.

Exercise (combination alert): The defence is worn to tatters. Black to play and win material. 50 … Rxc2! Answer: Overloaded defender. Black picks up two bishops for the rook, leading to an easy win. Actually 50 … Ng1! also wins material, no matter how White responds. 51 Rxc2 Bxe8 52 Rb2 Threatening 53 Rb8. 52 … Bf7 53 Kg2 Nd4 54 Kg3 Nc6 55 Ra2 Nb4! Black’s job is made easier if White’s knight is exchanged off. 56 Rd2 Nxd5 57 cxd5 Kg7 58 Rd3

Exercise (planning): White’s rook is tied up defending d5. What is Black’s winning plan? 58 … Kf6! Answer: Black’s king infiltrates on the queenside, via e7, d7, c7 and b6. Whereas 58 … Kg6 59 Kh2 Kh5 60 Kg3 doesn’t get Black anywhere. 59 Kg2 Ke7 60 Kf1 Kd7 61 Ke2 Kc7 62 Kd2 Bb4+ 63 Ke3 h5 64 Kd4 Bc5+ 65 Kc4 h4 White is plagued with multiple weaknesses, which cannot be simultaneously defended. Black threatens to invade via the queenside and White must also cover the d5-pawn, the f4-pawn and the h3-pawn. 66 Rd1 Bh5 67 Rd2 Bf3 68 Rc2 Be4 69 Re2 Bb6! 0-1 Zugzwang. After 70 Ra2 Be3 or 70 Re1 Bf2 71 Re2 Bg3 the f4-pawn falls. Game 11 Li Ruofan-Zhang Zhong Manila 2007 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c3 0-0 12 Nc2 Bg5 13 a4 bxa4 14 Rxa4 a5 15 Bc4 Kh8 16 0-0 Rb8 17 b3 f5

Black decides that the … g7-g6 and … f7-f5 plan is too ponderous and makes the f5-break without preparation. Black is compensated for handing White control over e4 (if White plays e4xf5), the weak d6-pawn as well as the d5-hole, with the following: 1. Enhanced piece activity. 2. The bishop pair and potential for dark-square activity later on. 3. Play along the f-file. 4. By removing White’s e-pawn, Black gets a central pawn majority, if it proves possible to break White’s blockade on d5. This may seem inconceivable when we look at this position, yet in this game Black managed to do just that. It happens far more often than we would believe in Sveshnikov, where White’s seemingly invincible grip on d5 is often worn away over time. 18 exf5 White should liquidate, since Black stands better if the centre is shored up with f2-f3, which weakens dark squares. For

example, 18 f3 Ne7 19 Nxe7 Bxe7 20 Qd3 (20 Ne3?? hung a piece to 20 … Qb6 21 Qe2 f4, T.Mak-M.Amanov, Doha 2006) 20 … fxe4 21 fxe4 Rxf1+ 22 Kxf1 U.Fataliyeva-S.Gaboyan, Porto Carras 2010. White is in danger after 22 … Qb6 23 Ne3 Bd7, intending to shift the rook over to f8. 18 … Bxf5 19 Nce3

19 … Bg6 I prefer this to playing it to e6, since here the bishop slices through the centre. 20 Qe2 White would be wise to challenge the bishop with 20 Bd3 Bxd3 21 Qxd3 Bxe3 22 fxe3 Rxf1+ 23 Qxf1 Rxb3 24 Rc4 Rb5 25 Nc7 Rb6 26 Nd5, when he can take a draw by repetition, as in V.Bologan-A.Timofeev, Sarajevo 2005. 20 … Qd7 21 Rfa1 The game is also balanced after 21 Rd1 Bd8. 21 … Qb7

We reach the end of theory. With his last move Black’s queen prevents Bb5, while eyeing both the b3-pawn and d5knight. The computer prefers 21 … Bd8, as in B.BaroinM.Versili, correspondence 2005. 22 b4!? White logically seeks to open the a-file and create a passed pawn. It’s hard to imagine White losing if she does nothing and continues to manoeuvre with something like 22 Qg4 Bh6 23 Rd1. 22 … axb4 Black is happy, since now there is no need for him to babysit the weak a-pawn. 23 Nxb4 23 cxb4 allows Black’s knight into d4. 23 … Nxb4 23 … Ne7 is playable as well.

24 cxb4

24 … Be8! A dual-purpose move: 1. Black fights for control over b5. 2. Black prepares to transfer the bishop to c6, after which it takes aim at the d5-square and g2-pawn. 25 R4a3? This move commits White’s greatest Sveshnikov sin: losing control over d5. The WGM should have played 25 Bd5 with an approximately even game. 25 … Bc6! 26 Rb3 It may have been preferable just to hand over a pawn with 26 b5 Bxb5 27 Bxb5 Bxe3 28 Qxe3 Qxb5, when White is a pawn down and fighting for her life. 26 … d5

One of the ABCs in the Sveshnikov is that White should not allow us control over d5. 27 Ba6 Qb6 28 b5 Ba8 Principle: when it comes to bishops, distance does not isolate our opponents from their wrath. In case you believe that Black’s a8-bishop is some Pete Best-like fifth Beatle, who lives in the shadows, then I urge you to disbelieve the notion. First, the bishop, although sidelined on a8, continues to cover the critical d5-pawn and square. Secondly, if Black plays … d5-d4, the bishop suddenly becomes a monster along the h1-a8 diagonal. 29 Rd1 Rbd8

Black has a winning position for the following reasons: 1. His centre is rolling and his d-pawn is set to advance. 2. He has a powerful bishop pair. 3. White’s passed b-pawn is blockaded.

4. White’s offside a6-bishop has been downgraded to spectator status. 5. The f2-square is tender, due to the open f-file and Black’s dark-squared bishop, in conjunction with the black queen. 30 Rbd3 Qc5 After 30 … d4 31 Nc4 Qf6! Black’s e5-pawn is taboo, since 32 Nxe5?? loses a piece to the simple pin 32 … Rde8 33 Re1 Rxe5! (overloaded defender). White’s queen is unable to recapture due to the mating threat on f2. 31 Qh5 e4 32 Rd4 Bf6 33 Rc4 Clever, but ineffective. 33 … Qa3 34 Rcc1 g6 34 … d4! 35 Ng4 e3! is even stronger. 35 Qe2 Intending to advance b5-b6. 35 … Bg7? Better is 35 … d4! 36 Nc4 Qe7 37 b6 d3 38 Qd2 Bd4 39 Rf1 Bd5 40 Ne3 Bxb6, when Black is a pawn up with a crushing bind.

Black’s lax move gives White a saving chance. Exercise (planning): The coroner isn’t quite ready to sign the white position’s death certificate. Black played too cautiously. Find a plan where White can save the game: 36 Qd2? Answer: White missed her golden chance to save the game with 36 b6! (threatening b6-b7). This is the butterfly effect, where a tiny alteration (Black’s indecision and failure to act on his last move) can create a wildly different outcome than the one previously envisioned. After 36 … Rb8 37 Rb1! (renewing the threat) 37 … d4 38 b7 dxe3 39 fxe3 Qe7 40 bxa8Q Rxa8 41 Bc4 the outcome is a near-certain draw. 36 … d4

Black once again has a winning position. 37 Nc4 Qe7 38 Nb6 Marginally better is 38 b6 d3 39 Rb1 Bd4! 40 Rf1 Rb8 41 Qa5 Bc6 42 Ne3 Qd8 43 Nc4 Rf5 44 Qb4 Qf6, although White is still busted. 38 … e3! 39 fxe3 dxe3 40 Qe1

Exercise (combination alert): How did Black force the win? 40 … e2! Answer: Clearance. There is no defence to a deadly bishop check on d4. 41 Rxd8 Qxd8 0-1 White resigned as 42 Nc4 Qg5! 43 g3 Bd4+ forces mate.

More elegant is 41 … Bd4+!! (attraction/weak back rank) 42 Rxd4 Qe3+ 43 Kh1 Qxc1! 44 Qxc1 Rf1+ with mate next move. Game 12 A.Tukhaev-M.Carlsen World Rapid Championship, St Petersburg 2018 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 5 Nb5 d6 6 N1c3 a6 7 Na3 b5 8 Nd5 Nf6 9 Bg5 Via a Kalashnikov (4 … e5), we have transposed to our desired Sveshnikov, made even more special by the fact that a world champion plays it, turning it into hallowed ground. 9 … Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Be2 Another theoretical dodge, which seems to annoy most Sveshnikov players. In this version White doesn’t rush into Nc2 and a2-a4. The line is too meek to give White an edge, since the bishop is better off on c4, where it fights for the d5-square. White generally plays this move when s/he is out to draw, which isn’t such a bad idea if your opponent is Carlsen. However, Carlsen gives a perfect display of how Black can thwart White’s aim and play for a win. The shocker of this game was that Carlsen, in a clearly superior position, gifted his young GM opponent a win on the clock.

12 … 0-0 13 Nc2 Ne7 Immediately fighting for control of d5. 14 Nxe7+ In the next game we look at 14 Ncb4. 14 … Qxe7 15 0-0 Bb7 16 Qd3?! White erroneously believes he has the d5-square covered. He doesn’t. Better is 16 Bf3 a5! (preventing Nb4-d5) 17 Qd3 Bc6 18 Rad1 Rfd8 19 Rfe1 Qe6 20 b3 Rac8, S.LattermannP.Vandevoort, German League 1996. The position is equal, with Black having enough leeway to play for the full point. 16 … d5! At first it feels as if Black fires indiscriminately at anything that moves. Carlsen doesn’t worry that the math on d5 is against him. White is unable to hang on to the extra pawn and Black takes over the initiative by opening the game for his bishop pair.

17 exd5 Rad8 18 c4 A new move. The game’s once-familiar shape begins to morph into something alien and unrecognizable. Others: a) 18 Bf3? just loses time after 18 … f5. b) 18 Nb4? is another time-wasting move. Black continues 18 … a5 19 Nc2 Rxd5. c) 18 Ne3? e4 19 Qd4 f5 20 Rfd1 was R.KilianV.Achenbach, Ruhrgebiet 1998, where 20 … f4 21 d6 Qe6 22 Bg4 Qg6 23 Nf1 e3! gives Black has a winning position. 18 … bxc4 19 Qa3 Qc7 20 Rfd1?! This wins a pawn, yet worsens White’s position. White can reduce his disadvantage by forcing Black to hand over the bishop pair with 20 Ne3 Bxe3 21 fxe3 Bxd5. 20 … Bxd5 21 Qxa6 Ra8

Carlsen has generated a kind of souped-up Benko Gambit from hell, with the following compensation for the missing pawn: 1. A powerful bishop pair in an open position. 2. Open a- and b-files to apply pressure on the queenside. 22 Qb5 Be6 Threatening 23 … Rfb8, trapping Black’s queen. Slightly more accurate is 22 … Ra5! 23 Qb4 Be6. 23 Qb4 Rab8?! A strange choice of rook. My guess is that he wanted to preserve his f8-rook for a kingside attack when he pushed his majority. Capablanca’s games demonstrate that “natural” usually equates with “strong”. Here 23 … Rfb8 24 Qc3 Bf6! was better, with a strategically won game for Black, since White is unable to withstand the heat of Black’s searing queenside attack. 24 Qc3 Bf6 This is scary for White’s queen.

25 Rab1 e4!

Opening the dark-squared bishop’s diagonal is more important than allowing White’s knight access to d4. 26 Nd4 Qa7! 27 Qe3 The queen must unpin. After 27 a3?? Rfd8 28 Nc6 Qb6 White loses either his queen or his knight. 27 … Bxd4!? 27 … Bd5 is a serious option. 28 Qxd4 Qxa2!? Also possible is 28 … Qxd4 29 Rxd4 c3 30 b4 Rfd8! 31 Rxd8+ Rxd8 32 Rc1 Rd2 33 Kf1 Rxa2 34 f3 exf3 35 gxf3 Rb2 36 Rxc3 Rxb4. Normally, I would say White has good drawing chances here, despite being a pawn down and having split pawns. However, with Carlsen playing Black, you’d expect an eventual 0-1. 29 Rbc1

On 29 Qxe4! Carlsen probably planned 29 … c3! 30 Bd3 f5 31 Qe5 cxb2 32 Rd2 Rfd8 33 Qe2 Kf8 34 Rdxb2 Rxb2 35 Rxb2 Qa3 36 Qxe6 Qa1+ 37 Bf1 Qxb2 38 Qxf5+ Kg8, though White should hold this ending. 29 … Rxb2 30 Bxc4 Bxc4 31 Qxc4 Rxf2 32 Qxe4

The position is a ghost town, except for Black’s extra pawn. Objectively, White has excellent drawing chances. The problem is that the guy across the board is one of the greatest technical endgame players of all time. 32 … Rf6 33 Rf1 Re6 34 Qc4 Qd2 35 Qc3 Qg5 The draw is becomes very much easier for White if queens are removed from the board. 36 Qc5 Re5 37 Qc4 g6 38 Rc3 Qd8 39 Rcf3 Qe7 40 Rf4 Kg7 41 Qc3 f6 42 Rc4 Threatening 43 Rc7. 42 … Rf7 43 Rc6 Re2 44 Qd4 h5!

In order to win, Black must accomplish the following: 1. Keep queens on the board. 2. Push the kingside pawns forward to create a puncture in White’s kingside. 3. Go after White’s king, without allowing his own king to come under attack or perpetual check. 45 Qd3 Qe5 46 Qf3 Qd4+ 47 Kh1 Re3! 48 Qd1 48 Qf2 was more accurate, since g2 soon requires protection. 48 … Qe4 49 Rd6 Re2 50 Rg1 Re3 51 Rd4 Qb7 52 Rd3 Re5 53 Qf3 Qe7 54 Rd5 Re3 55 Qd1 h4! Now White must watch out for … h4-h3 thrusts, denuding his king. 56 Qg4 g5 57 Qd4 Re4 58 Qc3 Kh6 59 Rf5 Qe6 60 Rgf1 Kg6 61 Qd3 Qxf5?!

You would think a world champion would never go wrong in a simplified position, but this just isn’t the case. Carlsen’s last move is incorrect, since it will now be virtually impossible for Black to make progress without submitting to perpetual check. He should just patiently manoeuvre with something like 61 … Kg7 62 Qb1 Re2 63 h3 Qe3 64 Qa1 Qg3 65 Rg1. White’s chances of holding the draw are probably just as high here as Black’s to win. 65 … g4?? is, for now, unplayable due to exploitation of the pin of the f6-pawn with 66 Rg5+! Kf8 67 Rxg4 and it is Black who has to play for the draw. 62 Rxf5 Kxf5 63 g4+!

63 … Ke5 Black is unable to take his king out of the narrative, so White should deliver perpetual check. For instance, 63 … Kf4 64 Qf1+! Ke3 (64 … Kxg4?? 65 h3+ Kh5 66 Qf3+ Kh6 67 Qxe4 is a technical win for White) 65 Qe1+ Kd4 66 Qb4+ Ke5 67 Qb5+! is perpetual check. 64 Qb5+ Kf4 65 Qf5+

65 Qf1+! transposes to the above line, which is an instant draw. 65 … Ke3 66 Qc5+ Rd4 67 Qc1+ Ke4 68 Qc2+ Ke5 69 Qf5+?! Again 69 Qc5+! Rd5 70 Qc3+ Kf4 71 Qc4+ Ke5 72 Qc3+ Kd6 73 Qb4+ is perpetual check. 69 … Kd6 70 Kg2 Rd5?! 70 … Rf4! 71 Qc8 Rd7 72 h3 Ke7! and suddenly Black can tuck his king away and generates winning chances. 71 Qg6! Ke6 72 Kh3?! Correct is 72 Qe4+ Re5 73 Qc4+ Ke7 74 Qb4+! (the rogue is reigned in; Black’s king must not be allowed to escape into the cubby hole on f8) 74 … Ke8 75 Qb8+ is perpetual check. 72 … Rfd7! 1-0

Here White lacks perpetual check and Black has winning chances, except for one problem: Carlsen flagged here and lost the game. Obviously, if anyone is going to win, it will be Black.

Now you know: even world champions are sometimes forgetful of their clocks, while young GM Tukhaev knows how it feels when Caissa basically gifts undeserving you. This miracle also happened to me in 1996, when I had the incredible good fortune to play then world champion Garry Kasparov three online blitz games. By a miracle of miracles, I drew the first game with Black. The second game I duly got outplayed and lost a tough ending. The third game, I was two pawns down in a rook ending when Kasparov flagged and I “won”. Since that time, I have told everyone who would listen that I drew a match with Kasparov, while spotting him odds of an extra White! Game 13 T.Pattard-S.Kalinitschew Guben 2003 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 5 Nb5 d6 6 N1c3 a6 7 Na3 b5 8 Nd5 Nf6 9 Bg5 Be7 10 Bxf6 Bxf6 11 c3 Bg5 12 Be2 0-0 13 Nc2 Ne7 14 Ncb4 This is a line devoid of charm, yet heavy on practicality if you are a lower-rated player attempting to score a draw against a GM. White backs up the d5-knight.

14 … Be6 Another option is simply 14 … a5 15 Nxe7+ Qxe7 16 Nd5 Qb7 17 Qd3 b4!, as in V.Anand-V.Kramnik, Dortmund 1997, and I slightly prefer Black’s game. 15 a4 Logical. White inflicts damage to Black’s queenside. Others: a) 15 0-0 a5 16 Nxe7+ Qxe7 17 Nd5 Qb7 18 Qd3 Bxd5 (it is always a glorious feeling for Black to seal the hateful d5-hole with a white pawn) 19 Qxd5 Qxd5 20 exd5 Rab8. Here I already prefer Black, who has the superior bishop and pawn majority. If you think that this line is drawish, then think again. White scores horribly from this position. Out of eight games in my database, White drew two and lost the remainder! b) 15 Qd3 Nxd5 16 Nxd5 Rb8 17 0-0, R.Moll-H.Neto, correspondence 2013. The game is balanced. Black can either play for an immediate … f7-f5, or first build with … g7-g6 and then … f7-f5.

c) 15 Nxe7+ Qxe7 16 Nd5 Qb7 17 0-0 Bxd5 is similar to line “a”. Again I already prefer Black. 15 … bxa4 16 Rxa4 a5 17 Nxe7+ Qxe7 18 Nd5 Qb7

Threatening b2. 19 b4 19 h4 Bd8 20 Qd2 was S.Filatov-J.Ruan, World U-12 Championship, Vung Tau 2008. Black already stands better after 20 … Rb8 21 b4 Bxd5 22 exd5 axb4 23 Rxb4 Qa7 with the superior pawn majority and bishop, which can slide over to a powerful diagonal on b6. 19 … Bxd5! Your writer loves to repeat an important point ad nauseam, so that it sinks into the deepest part of the reader’s subconscious. I already mentioned this previously but will repeat the Sveshnikov Principle: when you get a chance to plug the d5hole with a white pawn, do so, even when it costs the bishop pair.

20 exd5 Qc7? This inaccuracy allows White to pin Black’s a5-pawn. After 20 … axb4! 21 Rxb4 Qc7 22 Qd3 Ra1+ 23 Bd1 Rc1 24 Rc4 Qe7 25 0-0 f5 Black’s majority is rolling while White’s isn’t. 21 Qa1! e4 22 0-0 e3 23 fxe3? If 23 f4 Bf6 24 Rc1! Qb7 25 bxa5 Qxd5 26 Qa2, White’s passed a-pawn is more dangerous than Black’s blockaded epawn. 23 … Rae8!?

The GM playing Black is not interested in drawing with an expert-rated player, so he takes a calculated risk by offering White a passed a-pawn. Black has a draw if he wants one with 23 … Bxe3+ 24 Kh1 Bb6 25 Bb5 Ra7 26 Bc6 axb4 27 Rxa7 Bxa7 28 cxb4. 24 bxa5 Rxe3 Double attack. 25 Bf3?!

After 25 Qd1 Rxc3 26 Kh1 the chances are even. 25 … Rxc3 26 h4?! This lunge creates a caustic effect on White’s kingside dark squares. The tempo gained is likely outweighed by the selfinflicted weakening of White’s king. The move violates the principle: opposite coloured bishops favour the attacker. Therefore avoid creating weaknesses that favour the opponent’s bishop – especially around your king. 26 a6 is correct. 26 … Be3+ 27 Kh2 f5 27 … Qc5 maintains an edge. 28 Ra3?! White can keep equality by following the Principle: Passed pawns should be pushed, therefore 28 a6 is again correct. 28 … Bd4! 29 Rxc3 White must cough up his precious a-pawn. 29 Qa2? Rc2! 30 Qb3 Rc8 31 a6 Qe7 32 Kh3 g5! is a decisive attack for Black. 29 … Bxc3 30 Qa3 Bxa5 Black has a winning position since he is a pawn up and White’s king is unsafe due to the bishops of opposite colours. 31 Rc1 Qd8 32 Kh3 The king covers h3. 32 … Bb6 33 Rc6 Bc5 34 Qa6 g5! 35 hxg5 Qxg5 36 g3?? 36 Qe2 was the last chance.

Exercise (planning): What did White miss on his last move? 36 … Bf2! 0-1 Answer: The g3-pawn cannot be defended, so White is mated.

Chapter Two 9 Nd5: 11th Move Alternatives for White

In this chapter we look at alternatives to 11 c3, the most important of which is 11 c4, which we cover in the first four games. I tried hard not to repeat material from Sveshnikov: Move by Move, but there are some exceptions, such as the following. No important new games have been played in this line since Sveshnikov: MBM, so this game is necessary for this book, in case you don’t have my old one. In this game and the next, note that Black wins both games because of the oppositecoloured bishops, which often arise in this structure. Other 11th moves – 11 Be2, 11 Bd3 and 11 g3 – are covered in Games 18-20, while 11 Nxf6+ featured in the notes to Game

1 in the previous chapter. Game 14 J.Magem Badals-A.Moiseenko Spanish League 2010 1 e4 c5 2 Nf3 Nc6 3 Nc3 Nf6 4 d4 cxd4 5 Nxd4 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c4

The Sveshnikov isn’t always a harsh landscape which is strategist-hostile and tactician-friendly. At least 50% of the lines will be strategic battles, such as this one. With this move White turns the game into a battle of opposites: 1. With his last move White increases his power on the light squares, while also enhancing Black’s influence on the dark squares. 2. With c2-c4, White Maróczy binds d6, relegating it to inferior backward status for the remainder of the game, since

now a future … d6-d5 break is almost impossible to achieve. 3. By playing c2-c4, White’s remaining bishop is now downgraded to “bad” status, since two central pawns are fixed on the same colour. 4. By playing c2-c4, White must commit to guard the d4square, which is now an occupiable hole for Black. 11 … b4 12 Nc2 We love it if White is tempted into our strategic trap 12 Qa4?! “winning” our b-pawn. You would be well advised to memorize the forcing line: 12 … Bd7 13 Nxb4? (White can still escape with an inferior yet playable game after 13 Nb1) 13 … Nd4 14 Qd1 Rb8 15 Qd2 Bg5 16 Qc3 Qb6 17 Nd3 Qb7! 18 f3 f5! and White is busted. For example, 19 exf5 (or 19 Nf2 Qxb2 and White is crushed) 19 … Bh4+! 20 Kd1 0-0, when White is unable to unravel and Black has a winning attack. 12 … 0-0 13 Be2 White decides to bring out his pieces quickly. We’ll look at 13 g3 in Games 16-17. 13 … Bg5 Continuing to offer the b-pawn. 13 … a5 14 0-0 Bg5, as in the next game, is a safer move order. 14 0-0 Here White can also play 14 Ncxb4!?, bravely/recklessly accepting the challenge.

After 14 … Nxb4 15 Nxb4 Qb6 (attacking the knight and the b2-pawn behind it) 16 Nd5, White immediately returns the pawn with a dynamically balanced game, P.NegiY.Kryvoruchko, Tromsø 2013, which was covered in Sveshnikov: Move by Move. Note that 16 Nd3? is too ambitious as Black gets massive compensation after 16 … Bb7 17 f3 Rab8, when White is in terrible danger as his king is caught in the centre, his dark squares leak, and Black can apply pressure on the queenside and will soon break open the game further with … f7-f5. 14 … a5 At long last, Black protects his b-pawn. 15 Qd3 White prepares to place a rook on d1 and hammer away at the backward d6-pawn.

Instead 15 Bg4 is an attempt to leave Black with a bad bishop. It is met by 15 … Be6, when chopping on e6 kills White’s hole on d5 and also opens the f-file. 15 … Be6 16 Rfd1 In the next game we look at 16 Rad1. 16 … Ra7 The idea is to move the rook to d7 to protect d6. Another option is 16 … Rb8. After 17 b3 (17 Nde3 Bxe3 18 Qxe3 Qc7 is fine for Black) 17 … g6 18 h3 Rb7 19 a3 (if White avoids this move, he has no counterplay) 19 … bxa3 20 Nxa3 Nd4 21 Nb5 Bxd5 22 Nxd4 exd4 23 cxd5 Bf6, Black stood slightly better in D.Krivic-H.Neto, correspondence 2013, since he has a passed pawn and his dark square control is worth more than White’s on the light squares. 17 Bf3 17 a3 bxa3 18 Qxa3 Qb8 is equal, K.Toma-A.Suarez Real, British League 2018. 17 … Rd7 18 Qe2 White makes room for Rd3 and Rad1. 18 … g6 19 Rd3 Kh8 20 b3 Nb8!

The knight heads for a better post on the c5-hole. 21 a3! White breaks at the ideal moment, when Black’s knight is temporarily misplaced. 21 … bxa3 22 Nxa3 Na6 23 Rad1 After 23 Rdd1! Nc5 24 Qc2 f5 25 Nb5 Bh4 26 Ndc3 fxe4 27 Nxe4 Bf5 28 g3 Be7, I slightly prefer White’s position. 23 … Nc5 24 Rc3 f5 Black begins to generate play on the f-file, which offsets White’s pressure on d6. 25 Nb5

25 … Rdf7! Threatening 26 … fxe4. 26 exf5 Bxf5 Threatening 27 … e4, clearing the way for … Rxf2. Moiseenko is opting for piece play, whereas I would be more inclined to go for 26 … gxf5 27 Bh5 Rg7. 27 Rf1! This move eliminates Black’s threat to f2. 27 … Bh6 28 Qd1 Be6 29 Bg4 Qh4! 30 Rg3? After 30 Bf3! e4 31 g3 Qg5 32 Nxd6 Rxf3 33 Rxf3 Rxf3 34 Qd4+ Bg7 35 Qxc5 e3 36 fxe3 Rxf1+ 37 Kxf1 Qe5 the comp assesses the position as dead even, despite White’s two extra pawns.

Exercise: Combination Alert: The twilight deepens and it becomes more and more difficult to see. White just allowed a combination. What did he miss? 30 … Bxd5! Answer: Step 1: Attraction/pin. Chop the d5-knight. 31 Qxd5 Rxf2! Step 2: Attraction. Sacrifice a rook on f2. 32 Rxf2 Be3 Step 3: Pinned piece. White loses material. 33 Qxd6 Also losing is 33 Rgf3 Qxf2+! 34 Rxf2 Rxf2 (the threats are 35 … Rf3+, followed by a back rank mate on f1 and 35 …

Rd2+, winning the queen back) 35 g3 Rd2+ 36 Kf1 Rxd5 37 cxd5 Nxb3 38 Nxd6 a4 with a won ending for Black. 33 … Bxf2+ 34 Kh1 Qf6 The human move. The comp found the impossible-to-see line 34 … Re8!! 35 Rf3 Ne4! 36 Qd7 Rd8 37 Rh3 Rxd7 38 Rxh4 Rd2 39 g3 Bc5 40 Bf3 Nf2+ 41 Kg2 e4, winning a piece. 35 Rf3 Qxd6 36 Nxd6 Rxf3 37 Bxf3 Nxb3 38 Nf7+ Kg7 39 Nxe5

Material remains equal but Black has a winning position for the following reasons: 1. Black’s outside passed pawn rolls toward the promotion square. 2. White’s passed pawn is firmly blockaded and going nowhere. 3. Black’s king is active and ready to help, while White’s lags way behind in the corner.

4. The drawing power of the opposite-coloured bishops is somewhat negated by the fact that knights remain on the board. 39 … a4 40 Bd1 Nc5 41 Nc6 Be1! Black’s last move prevents Nb4 and thus wins a piece. 42 Bxa4 Well, why not? Don’t believe that this is some kind of ineffectual suicide attempt from White, since everything loses: a) 42 Bc2 a3 43 Bb1 Ne4 44 Ba2 Nc3 45 Bb3 a2 and White must hand over the piece in any case. b) 42 Nd4 a3 43 Bb3 Bf2! (overloading the defender of b3) 44 Nc2 Nxb3 45 Nxa3 is pretty much the same. 42 … Nxa4 43 Nd4 White plays on, perhaps with the slimmest of hopes to end up with a piece down wrong-coloured h-pawn for Black, which isn’t likely to happen. White is busted for the following reasons (besides the obvious one of his missing piece): 1. Black also has a g-pawn. 2. Knights remain on the board, so even if White achieves his dream, it still would be lost. 3. White’s c-pawn is certain to fall.

43 … Kf6 44 g3 Bf2 45 Ne2 45 Nb3 Nc5 is also hopeless for White. 45 … Nb2 There goes White’s c-pawn. 46 Nc3 Ke5 47 Kg2 Bd4 48 Nd5 Nxc4 49 Nf4 Kf5 50 h3 Be5 51 Nd3 Ke4 The winning path for Black is to force White’s pawns forward and overextend them. There isn’t much White can do to prevent this. 52 Nf2+ Ke3 53 h4 Ke2! 54 Ne4 h6 55 g4 Ne3+ 56 Kh3 Nd5! 0-1

Game 15 E.Pähtz-P.Eljanov Isle of Man 2016 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c4 b4 12 Nc2 0-0 13 Be2 a5 14 0-0 Bg5 This way White’s f2-f4 is made less desirable, since it would increase the power of Black’s dark-squared bishop. 15 Qd3 Be6 16 Rad1

Which rook White places on d1 is important and shapes the play for the remainder of the game. In Game 14 White put the f1-rook on d1, meaning that plans with a2-a3 and queenside play were possible. By playing the a1-rook to d1, White signals that she plans to ignore the queenside and go for a break with f2-f4. 16 … Ra7 As in the previous game Black prepares to defend the d6pawn with … Rd7, just in case White relocates her d5-knight. 17 b3 Qb8 17 … Rd7 is Black’s main move and also fine. 18 Kh1 Theory ends here. White prepares f2-f4, with or without g2g3 thrown in. 18 Nde3, as in M.Kobalia-M.Devereaux, Tromsø 2010, can be met either by chopping the e3-knight, or playing 18 … Rd8 straight away. 18 … Ne7

Black begins the plan of challenging d5. 18 … a4 is also reasonable. After 19 g3 axb3 20 axb3 Bh3 21 Rfe1 f5 Black stands no worse. 19 Nxe7+ Bxe7 20 Ne3 g6 21 Nd5 White can also play directly for mate with 21 g3!? a4 22 f4 exf4 23 gxf4 Bd7. Delivering mate is always possible, but it’s every bit as likely that White will simply overextend. 21 … Bxd5! From observation and experience, I have noticed that when we reach such bad bishop versus bad bishop structures, the black piece is better, since it has a nasty habit of emerging on the g1-a7 diagonal, via d8. 22 Qxd5 Others: a) 22 cxd5 Bd8! 23 Rc1 Bb6 24 Rc6 Bc5 25 g3 a4 26 Bd1 axb3 27 Bxb3 Rc8 28 f4 exf4 29 gxf4 Qb7 30 Qf3 Rxc6 31 dxc6 Qxc6 32 f5 Qe8 33 Bd5 Bd4 and it’s hard to see how White loses. Still, I would rather take Black, who has an extra pawn. b) 22 exd5? is dumb, since it activates Black’s kingside pawn majority, while White is stuck with a fixed one on the queenside. 22 … Rc8 Halting c4-c5 tricks, while preparing to eject White’s queen from her central post. 23 f4!?

This creates a caustic effect on White’s dark squares. It’s difficult to let such a move go with adverse comment, but for now I will not dub it an error. With this radical decision, White is willing to take on an isolated e-pawn and an e5-hole in order to pry open the f-file. I suppose true faith in the (future) attack requires acceptance of the unseen and unproven. A less radical agenda would go something like 23 Qd3 a4 24 g3 axb3 25 axb3 Qb6 with an equal position. 23 … Rc5 24 Qd2 exf4 25 Qxf4 Bf8 26 Rd5?! As mentioned many times previously, it usually favours Black when d5 is plugged with a white pawn. White was better off trying 26 Rd3. 26 … Rxd5 27 exd5 After 27 cxd5 Bg7 28 Bc4 Be5 White has no counterplay and is stuck with a bad bishop. Black can slowly try for play down the a-file with … a5-a4, in conjunction with a slow push on the kingside with black queen on e7 and … h7-h5. 27 … Qb6

Black is essentially a pawn up, since White’s majority is frozen. 28 Bg4 Maybe White should try and bail out to an ending at once with 28 Qf2. 28 … Bg7 29 Re1 Be5 Eljanov seals the e-file. 30 Qe3 The WGM decides that her king is in greater danger than Black’s, so she tries her luck in the ending, hoping the oppositecoloured bishops will save her. 30 … Qxe3 31 Rxe3

The ending will be a tough one for White to hold for the following reasons: 1. Black’s bishop is far superior to White’s.

2. Black is virtually a pawn up, since he has a mobile 3 to 2 kingside majority, while White’s 4 to 3 needs a c4-c5 break to create a passed pawn. This will be very difficult to achieve. 3. Black may generate play with … a5-a4 and open the afile. 31 … Kg7 32 Bd1 Halting … a5-a4. The “undermining” shot 32 c5?? fails miserably to 32 … Bd4. 32 … h5 33 g3 f5 34 Kg2 Kf6 35 Re1 Rc7 This prevents c4-c5 tricks. 36 Re2 Kg5 He wants either to provoke White into the weakening h2h4+, or else open the h-file with … h5-h4. 37 Bc2 h4 38 Bd3 Kf6 39 Rc2 After 39 gxh4?! Rh7 40 Kh3 g5 favourably regains the pawn. 39 … Rc8 40 Re2 Rh8 41 Re3 Rh7 42 Bc2 Rh6 43 Bd3 Rh8 White is unable to improve her position, so Black makes a few waiting moves. 44 Bc2

Exercise (planning): What is Black’s winning plan? 44 … h3+! Answer: Push the h-pawn to h3, fixing White’s h2-pawn as a permanent target. 45 Kf1 g5 Here comes the majority. 46 Rf3 f4 47 g4 This attempt to seal the position fails. White is also unlikely to save the game after 47 gxf4 Bxf4 48 Kg1 Ke5. 47 … Bd4 48 Bd3 Ke5 49 Bg6! Intending Bh5 and Rxh3. The problem is that perseverance doesn’t do a lot of good when labouring on a lost cause. White does indeed win the h-pawn, but at too high a cost, since her pieces get tangled in the process. Still, it’s the best try.

49 … Be3 50 Ke2 Kd4 Now that is what I call a king. 51 Bh5 Black’s h3-pawn falls, but White’s pieces had to go into a terrible tangle to win it. 51 … Kc3 52 c5 Desperation. After 52 Rxh3 Kb2 Black’s king mops up the queenside pawns. 52 … dxc5 53 Rxh3 Rh7 54 Rf3 Rd7 55 h4! We are all men and women of indomitable spirit when everything is going well in our lives. It’s not so easy to fight when everything is falling apart. When we are losing, the normal must bow to the whims of the insane. Everything loses at this stage, but this is White’s best shot at complicating, since it creates a passed g-pawn. 55 … gxh4 56 Rh3 Rxd5 57 Kf3 Kb2?! Complicating the win, which is easier after 57 … Rd2! 58 Rxh4 Rf2+ 59 Ke4 Rxa2 60 g5 a4 61 bxa4 Rxa4 62 g6 b3+ 63 Kf5 b2 64 Rh1 Ra1 65 g7 Rxh1 66 g8Q b1Q+. 58 Bf7! Rd2?! Another inaccuracy and another setback for Black. Here 58 … Rd3! 59 Ke4 Rc3 60 g5 f3! 61 Rxf3 Bxg5 62 Rf2+ Ka3 is an easy technical win. 59 g5 Rf2+ 60 Ke4 Rg2 61 g6 Kxa2 62 Rxh4 Ka3?! Black’s position becomes fractionally more difficult to win with each passing inaccuracy. Black has better winning chances with 62 … c4! 63 Bxc4 Rxg6. 63 Rh8! c4 64 Bxc4 Rxg6 65 Rh5! In this version Black only gets one pawn, not two, so possibly White should hold the draw. 65 … Bb6 66 Kxf4 Rc6 67 Rb5??

White’s rook clogs the b5-square, which had to be left open to prevent Black from playing … a5-a4. After 67 Ke4! White should draw, since if 67 … a4 68 Bb5 axb3 69 Bxc6 b2 70 Rh1 Ka2 71 Bd5+ b3 72 Rd1 b1Q+ 73 Rxb1 Kxb1 74 Bxb3, I’m almost certain this is a theoretical draw. 67 … a4 This is the pesky detail White forgot. Now she no longer has Bb5, since her rook occupies the square. 68 Bd5 Rh6 69 Kg5 Or 69 bxa4 Kxa4 70 Bc4 Rh4+ and the loose white bishop falls. 69 … Be3+! How convenient. Both the black rook and bishop get covered in a single move, and with tempo. 70 Kf5 axb3 71 Ra5+ Kb2 72 Ke4 Bb6 73 Ra8 Rh4+ 74 Kf3 74 Kd3?? hangs the bishop to 74 … Rd4+.

74 … Bd4 75 Rb8 Bc3 76 Rg8 Rd4! The rook will block annoying checks on d2. 77 Rg2+ Rd2 78 Rxd2+ Bxd2 79 Ke2 Kc2! 0-1

After 80 Bxb3+ Kxb3 81 Kxd2 Ka2 Black promotes. Game 16 V.Iordachescu-Yu Yangyi FIDE World Cup, Baku 1 e4 c5 2 Nf3 e6 This is the other move order to reach the Sveshnikov. The players get the same position, with each side expending one extra tempo so we reach our “11th Move Alternatives” position on move 12, rather than 11. 3 Nc3 Nc6 4 d4 There we go. Now we are back on track. 4 Bb5 is an AntiSicilian which we don’t cover in this book.

4 … cxd4 5 Nxd4 Nf6 6 Ndb5 d6 7 Bf4 e5 8 Bg5 a6 9 Na3 b5 10 Nd5 Be7 11 Bxf6 Bxf6 12 c4 b4 13 Nc2 0-0 14 g3 A fianchetto feels slightly more accurate than placing the bishop to e2, for the following reasons:

1. On g2, Black will be less enthusiastic about the freeing break … f7-f5, since after e4xf5, White’s formerly bad bishop is unleashed. 2. The advance g2-g3 is a useful move as it helps support f2f4. If Black trades, then White can play g3xf4, retaining control over e5. White might also decide to toss in h2-h4, if Black’s bishop remains on f6 for too long, depriving it of a superior diagonal on c1-h6. In the next game White opts to play h2-h4 before g2-g3, which may be more accurate, as it pre-empts … Bg5 as in the next note. 14 … a5

Black can also take the chance here to play 14 … Bg5. In reply 15 Ncxb4? is too risky; e.g. 15 … Nxb4 16 Nxb4 Qb6 17 a3 a5 18 Nd3 Be6 19 b3 Qd4 20 Be2 Bh3 21 f3 Qc3+ 22 Kf2 Qd4+ 23 Ke1 and after 23 … f5 White is busted. 15 h4 Be6 16 Qd2 We’ll look at 16 Bh3 (or more precisely, 15 Bh3) in the next game.

16 … a4!? This line-opening pawn sacrifice, made in the spirit of the Benko Gambit, is sound. Alternatively, 16 … Rb8 is safe and equal. 17 Ndxb4 Qb6 Or 17 … Nxb4 18 Nxb4 Qb6 19 Be2 Rfb8 20 a3 Qb7 21 Qd3 Bd8! (the b4-obstruction will soon be removed) 22 0-0 Ba5 23 Qxd6, R.Ponomariov-B.Gelfand, Hengshui (blitz) 2019, and after 23 … Bxb4 24 axb4 Qxb4 the position is level. 18 a3?!

Better is 18 Bd3!, though after 18 … Nxb4 19 Nxb4 Rab8 20 a3 Rfc8 21 0-0 Bd8 Black gets full compensation for the pawn. 18 … Na5! 19 Rd1 Qb7 20 Ne3 Nb3 White is struggling after 20 … Rfc8!, but not 20 … Qxe4?? which hangs the exchange to the simple 21 Bg2. 21 Qc3 Instead, d3 is a better square for the queen. 21 … Nd4! 22 Bd3 Bd8! The bishop heads for a5. As in the Benko Gambit, White stands worse if he loses his extra pawn. 23 0-0!? There is nothing better. With his last move White sacrificed the exchange in a bid to seize light square power. 23 … Bh3! 24 Ned5!

White realizes that he must bolster the b4-knight, since its removal means the collapse of his queenside. Others: a) 24 Ng2 Bb6 25 Kh1 Rfb8 26 f4 Bc5 27 Kh2 Bxg2 28 Kxg2 Bxb4 29 axb4 Qxb4 and White is strategically lost. b) 24 Rfe1 loses the exchange to 24 … Nf3+. 24 … Bb6 25 Kh2 Bxf1 26 Rxf1 Bc5 27 Kg2 Kh8 28 Qd2 Rab8 29 Rb1 Qd7 30 Qg5 Rb7 31 Qh5 f6 32 Qd1 Rfb8 33 h5! h6 White defends well, managing to fix every black pawn on the board on the same colour squares as his remaining bishop. He also bolstered his b4-knight. 34 Ra1 White should simply repeat with 34 Na6 Ra8 35 Nab4 and see how Black plans to make progress. 34 … Rf8

Black is better off giving the exchange back to reach a dominating position with 34 … Bxb4! 35 axb4 Rxb4! 36 Nxb4

Rxb4 37 Ra2 Qc6. Here White is busted since he is tied down to the defence of b2 and c4. Black’s knight also outguns White’s bad bishop. 35 Rb1 Ra8 36 Ra1 Rab8 37 Rb1 Rf8 Both sides continue to underestimate the plan 37 … Bxb4! 38 axb4 Rxb4!. 38 Ra1 f5! Lines must be opened for Black’s rooks. 39 exf5 Now White’s once bad bishop becomes un-redundant. 39 … Nxf5 40 f3?! This hangs a pawn. White had to try 40 Be4 Nd4 41 f3. 40 … Qf7! 41 Qe2 Others don’t help: 41 Qh1? Ne3+ 42 Nxe3 Qxf3+ wins, while 41 g4? loses to 41 … Nh4+. 41 … Qxh5 White faces the twin spectres of material and strategic poverty. The f3- and g3-squares are impossible to defend forever. 42 g4!? This fails, but nothing else worked either. 42 Rf1 would have lost in the long run. 42 … Nh4+ 0-1

After 43 Kg3 Qg5 44 Rf1 Rbf7 45 Be4 Nf5+! 46 Bxf5 Rxf5 47 Nd3 h5 White’s game collapses completely. Game 17 F.Caruana-S.Mamedyarov Shamkir 2016 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 Some people are born to fight. GM Shakhriyar Mamedyarov belongs in this category, so the Sveshnikov is a perfect fit for his confrontational style. 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 c4 b4 12 Nc2 0-0 13 h4

The sophisticated idea of this move is to pre-empt Black’s … Bg5. 13 … a5 Instead: a) 13 … Be6 doesn’t alter much. White should again refrain from 14 Ncxb4? Nxb4 15 Nxb4 Qb6 16 a3 a5, when Black has regained his sacrificed pawn with a big advantage. b) 13 … Bxh4?? loses on the spot to 14 Qh5. 14 g3 Logically covering h4, while preparing to fianchetto. 14 … Be6 15 Bh3 15 Bg2 a4 16 0-0 b3! 17 axb3 axb3 18 Rxa8 Qxa8 19 Nxf6+ gxf6 20 Ne3 Nd4 21 f4 was L.Dominguez Perez-B.Gelfand, World Team Championship, Tsaghkadzor 2015. Black’s chances are no worse after 21 … Qa6. 15 … Rb8

Another option is 15 … a4 16 Nce3 Nd4 17 Kf1 b3 18 axb3 (or 18 Kg2 Rb8 and Black stood no worse in M.RüfenachtV.Iotov, correspondence 2015) 18 … Nxb3 19 Ra2 Nc5 and Black looks comfortable, A.Motylev-B.Grachev, Sochi 2017. 16 Qd3 16 b3 Nd4!? 17 Nxd4 exd4 18 Bxe6 fxe6 19 Nf4 d3! 20 0-0 Bxa1 21 Nxe6 Qe7 22 Nxf8 was J.Stephan-L.Dos Reis, correspondence 2007. Black looks just a shade better after 22 … Bd4 23 Qxd3 Bc5 24 Nxh7 Kxh7. I would rather have Black’s piece over White’s three extra pawns. 16 … Qd7!? We of the Sveshnikov clan tend to be unafraid of doubled fpawns. After 16 … a4 17 b3 Black stands okay, F.CaruanaL.Van Wely, Wijk aan Zee 2014. 17 Nxf6+ gxf6

Okay, so our f-pawns are doubled, but it’s not so easy for White to exploit them.

18 Bf5 18 Bg2 was seen in W.So-H.Pascua, Philippines Championship, Manila 2008. Here Black can try 18 … b3! 19 axb3 Qb7 20 Ra3 Nb4 21 Qd2 Nxc2+ 22 Qxc2 Qb4+ 23 Qd2 a4! with pressure for the sacrificed pawn. The text is Caruana’s attempted improvement. He tries for a light square bind. Black looks okay in this version too. 18 … Qb7! Mamedyarov takes aim at e4 and the h1-a8 diagonal with his queen. 18 … Bxf5 19 exf5 is also playable. 19 Ne3 And certainly not the greedy 19 Qxd6??, when 19 … Rfd8 20 Qc5 Bxf5 21 exf5 Nd4 22 0-0-0 Rdc8 wins. 19 … Nd4 Black’s power on the dark squares nullifies White’s on the opposite colour. 20 0-0 b3! The idea here is to artificially isolate the c4-pawn. 21 a3 It’s logical to close queenside lines if White is playing for a win. Later on, Caruana may have regretted this decision, since the b2-pawn became a huge problem for him in the endgame. Instead, 21 axb3 Qxb3 22 Qd1 Qxd1 23 Rfxd1 Ra8 24 Ra3 Rfb8 25 Rda1 Nb3 26 R1a2 Nc1 27 Ra1 Nb3 is a repetition draw. 21 … Kh8 22 Rac1 Rbc8 23 Rfe1 Rc5 The game is balanced. 24 Qd1 Intending to swing the queen over to h5, which induces Black’s next move. 24 … Bxf5 25 Nxf5 Nxf5 26 exf5 Rd8 27 Qh5 Kg7 28 Rc3

Caruana wants to launch an attack, which is pretty much impossible against Black’s well-protected king. White must watch out for both his c4-pawn and also … d6-d5 breaks. 28 … h6 29 Ree3?! Caruana overestimates his attacking chances. He doesn’t have time for g2-g4, Rg3 and g4-g5. He should have moderated his ambitions. The game remains in balance after 29 Qg4+ Kh7 30 Rd1 a4 31 Qh5 Qc7 32 Rd5. 29 … a4 Protecting b3. 30 Qe2 30 g4? Rh8! negates g4-g5 ideas. 30 … d5 Stronger was 30 … Rdc8! 31 g4 d5! 32 Qf3 Qd7 33 g5 (or 33 cxd5 Rxc3 34 Rxc3 Rxc3 35 Qxc3 Qxd5, threatening … Qd1+ and … Qc2!, while 36 Qe1 Qd3 sees White overextended)

33 … d4 34 Qh5 Rh8! 35 gxh6+ Kf8 36 Rcd3 Rxc4 and White is strategically busted. 31 Qf3? 31 cxd5 Rcxd5 leaves White vulnerable on both the first and second ranks, but this is better than the game continuation. After 32 Qf3 Qa7 33 Rc4 Rd4 Black is in control but not fully winning, with White’s position a shadow, neither alive nor dead. 31 … Rdc8! 32 cxd5 Rxc3 33 Rxc3 Rxc3 34 Qxc3 Qxd5 The ending may look even, but White is completely busted for the following reasons: 1. Black’s king is completely safe from perpetual check attempts. 2. There is no remedy to Black’s threat … Qd1+ and … Qc2, after which White’s b2-pawn falls. 3. When b2 falls, White is too slow in the promotion race. 35 Qb4 h5 Reversing the order with 35 … Qd1+ 36 Kg2 h5! is slightly more accurate. 36 Qxa4 After 36 Qe1 Qf3 White can barely move.

Exercise (critical decision): Should Black play 36 … Qd1+, followed by 37 … Qc2 - ? Or should he play 36 … e4 - ? In one line the assessment is -10.36 – more than a full queen up for Black; while in the other the assessment plummets shockingly to 0.00. Which line should Black play? 36 … Qd3?? Both action and inaction produce consequences. This is a catastrophic error, after which White should hold the draw. Answer: Black wins easily with the more vigorous 36 … e4! 37 Qa7 Qd2 38 a4 Qxb2 and Black wins the promotion race by a mile. 37 g4! Qb1+ 38 Kg2 Qxb2 39 gxh5

White has just enough to hold the draw here. 39 … Qc2 40 Qg4+ Kh7

Exercise (critical decision): White can play 41 h5-h6, or he can play 41 Qa4, pinning Black’s b-pawn. One line holds the draw, while the other loses. Which one would you play? 41 h6?? This is where your insane dog decides to dig up your backyard, for absolutely no reason. Most of us find it comforting that super-GMs make double question marks on occasion, the same way we do. Answer: 41 Qa4! holds the draw after 41 … Qd1 42 Qc4! Kg8 43 h6! b2 44 Qb5! Qg4+ (not 44 … b1Q?? and it is Black’s

king who is mated after 45 Qe8+ Kh7 46 Qxf7+ Kh8 47 Qg7 mate) 45 Kf1 Qd1+ and Black must take perpetual check. 41 … Kxh6 42 Qg8 Qxf5 Now there is no perpetual check. 43 Qf8+ If 43 Qh8+ Qh7, the problem is that 44 Qxf6+ fails miserably to 44 … Qg6 check, when queens are removed from the board and Black’s b-pawn promotes. 43 … Kh5 44 f3 Kxh4 Easier is 44 … Qc2+! 45 Kh3 Qc4! 46 Qh8+ Kg6 47 Qg8+ Kf5. 45 Qb4+ Qf4!

Also winning is 45 … e4! 46 Qxb3 Qh3+! 47 Kf2 Qg3+ 48 Ke2 Qg2+ 49 Kd1 exf3 50 Qxf7 Qe2+ 51 Kc1 Qe5 52 Qa7 Qe1+ 53 Kc2 Qe2+ 54 Kb3 f2. Black promotes and again there is no perpetual check.

46 Qxb3 In a famine it can be a long time before our next meal. The presumptuous pipsqueak on b3, who earlier made so much trouble for White, just fell. On top of that, White is now the only one with a passed pawn. These factors are still not enough to save Caruana, since mating threats begin to arise around his king. 46 … Qd2+ 47 Kf1 Kg3! Threatening 48 … Qf2 mate. 48 f4+ There is no choice. 48 Qb6?? allows 48 … Qd1 mate. 48 … Kxf4 49 a4 f5 50 Qb5 Qd1+ 51 Kf2 Qc2+ 52 Kf1 f6! 53 Qb4+ After 53 a5 Kf3 54 Qd5+ Ke3 White doesn’t have a single reasonable check, and 55 Kg1 f4 56 a6 f3 forces mate. 53 … e4 There are way too many pawns sheltering Black’s king and there is no way White can force a perpetual. His passed a-pawn is slower than Black’s mating/promotion threats. 54 Qb5 Kg3 White’s endangered king shivers in his hovel. 55 Qe2 Or 55 Qxf5 Qd1 mate.

Exercise (critical decision): Work out the ramifications of a queen swap into a king and pawn ending. Does Black win? 55 … Qxe2+! Answer: Black is faster after the queen trade. 56 Kxe2 f4 57 a5 f3+ 58 Kf1 e3 0-1 Have you ever played a game where a week’s worth of energy has been drained in just a few hours’ time? This game surely must have felt this way to Caruana and Mamedyarov. After 59 a6 White is too slow; e.g. 59 … e2+ 60 Ke1 f2+ 61 Kxe2 Kg2 and Black promotes first. Game 18 G.Prassas-S.Halkias Ano Liosia 2000 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 Be2 This move may be safe, but it certainly won’t give White an edge. The general consensus is that White’s bishop is best

placed on c4.

11 … Bb7 As usual Black fights for control over d5. There is of course nothing wrong with castling first. For example, 11 … 0-0 12 0-0 Bg5 13 c3 Ne7 (our continual strategy in Sveshnikov is to challenge the occupant of d5) 14 Nc2 Nxd5 15 Qxd5 Be6 16 Qd3 and now Black can either play the freeing … d6-d5, or maybe try and keep it more complicated with 16 … a5!? 17 Rfd1 Rc8! 18 Ne3 (18 Qxd6? Qxd6 19 Rxd6 b4 favours Black) 18 … b4 19 cxb4 axb4 20 Nd5 Bxd5 21 Qxd5 Rc2, when I slightly prefer Black, G.Peli-G.Galliano, correspondence 2016. 12 c3 12 c4 is well met by 12 … 0-0 (12 … Nd4 is also fine for Black), since if White gets greedy with 13 cxb5 Nd4 14 bxa6?! then 14 … Bxd5 15 exd5 Qa5+ 16 Kf1 (forced) 16 … Rxa6! sees White behind in development and in deep trouble. Instead,

14 Nc3 axb5 15 Naxb5 d5!! 16 exd5 Bxd5 17 Nxd5 Qxd5 18 00 Rxa2 is equal. 12 … 0-0 13 Nc2 Bg5 14 0-0

14 … Nb8!? This move scores better than playing the knight to e7, which may be more thematic. The idea is to shift to c5. If you are not looking for drama, there is the normal 14 … Ne7 15 Ncb4 Ng6 16 g3, A.Kovchan-J.Fluvia Poyatos, Oropesa del Mar 1998, when Black looks just fine after 16 … a5 17 Nc2 Bc6. 15 a4 White breaks up Black’s queenside, which is logical, if a little late. Instead, after 15 Qd3 Nd7 16 Rfd1 Nc5 17 Qf3 g6 18 Nce3 Kh8 19 Bf1, chances were balanced in S.TiviakovA.Shirov, Oakham 1992. 15 … bxa4 16 Rxa4 Or 16 Na3!? Nd7 17 Nc4 Nc5 18 Ndb6 Rb8 19 Qxd6 Qxd6 20 Nxd6 Bc6 21 Nbc4, A.Stearns-C.Lakdawala, San Francisco

2000. Black stands slightly better due to the bishop pair and pressure down the b-file after 21 … Be7 22 Nf5 Bf6. 16 … Nd7 17 Ra3 a5 18 Bc4 It doesn’t make much sense that White plays the meek 11 Be2, only to move it to c4 later in the game. Black is okay here, despite the Sveshnikov-hating comp’s usual misassessment of an advantage for White. 18 … Nc5 19 Qe2 Kh8 20 Rfa1 Creating a passed b-pawn with 20 b4 is more thematic. All the same, Black looks okay after 20 … Ne6. 20 … a4!

The more unlikely our idea, the more we want to play it. Once we commit to a move like this, we can’t subsequently turn back and pretend we never made it. The advance b2-b4 is prevented and White’s rooks look odd, pressing against the a4pawn. Black’s move ups the aggression factor by several levels of magnitude. The obvious drawback is that it pretty much dooms Black’s a-pawn if White plays Bb5. GM Halkias

correctly foresees that Black gets sufficient counterplay on the kingside if White wastes time picking off the pawn. 21 Ncb4 If 21 Bb5?! then, as the saying goes, you get what you pay for. The move is well met by 21 … f5! 22 Ncb4 fxe4 23 Bxa4 Nd3! 24 Nxd3 Bxd5 25 Nb4 Bb7 26 c4 Qb6 with pressure on the dark squares and against f2. 21 … Rc8?! Here 21 … f5! is thematic and correct. 22 Bd3! Preventing … f7-f5. 22 … g6 23 Bc2 Black’s a-pawn is doomed, so he must generate quick play with a break on f5. 23 … f5 24 exf5 gxf5 25 Bxa4 This is not the kind of position which is willing to give us a straight answer to even one of our questions. White has won a pawn, but it is obviously not the end of the story, since, as so often happens, Black soon generates a compensating kingside attack. 25 … Rg8 26 Bc2 Qf8 The black forces creep closer to the white king. 27 Qh5 I would toss in 27 f3.

27 … e4! 28 c4 A dual-purpose move. White backs up the d5-knight, while enabling defensive ideas such as Rg3. If 28 f4, Black gets full compensation for the pawn with 28 … exf3 29 Qxf3 Qg7 30 Re1 (30 Bxf5?! is just asking for it after 30 … Rcf8) 30 … Rce8 31 Rxe8 Rxe8 32 Ra1 Re5 33 Rd1 Be3+ 34 Kh1 f4 with chances for both sides. 28 … Ne6 29 Ra7?! After 29 Rg3 Rg7 the position is unclear. 29 … Qg7! Discoveries are in the air with the g5-bishop, which would then uncover a mating threat on g2. 30 b3?? 30 g3 was necessary; even then Black has a dangerous attack after 30 … Nd4.

Exercise (combination alert): You know that feeling we get when we bang down our awful move and instantly realize it loses, and then the room begins to spin and we lose all sensation in the body? Well, I’m pretty certain White felt that here. White covered c4, but forgot about something else on his last move. Find Black’s next and White’s position frantically shouts for help: 30 … Bf6! Answer: Double attack. Black threatens mate on g2 and also … Bxa1. Cheapos such as this aren’t just things which happen to someone else. We all fall for them from time to time.

31 Qh3 White has nothing left to bargain with. After 31 g3 Bxa1 32 Qxf5 (a few pawns won’t make up for the rook he just lost) 32 … Ng5 33 h4 Nf3+ 34 Kh1 Rcf8 35 Qxe4 Bd4 Black wins, since f2 falls. 31 … Bxa1 0-1 Game 19 R.J.Savory-D.Gormally EU Championship, Liverpool 2006 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 Bd3 This is another insipid attempt by White to avoid theory.

11 … 0-0 12 0-0 Bg5 13 c3 Or 13 c4, when 13 … b4 14 Nc2 Rb8 is equal. 13 … Rb8!

Black’s idea is to meet Nc2 with … a6-a5! and play on the queenside with … b5-b4. Another option is 13 … Ne7 14 Nxe7+ Qxe7 15 Nc2 Bb7 16 a4 bxa4 17 Rxa4 a5 and chances are balanced, P.Vinogradov-Y.Solodovnichenko, Kharkov 2002. 14 Nc2 a5! Black has equalized. As usual, the stupid, ignoramus comp (no, I won’t apologize to it!) gives White an unjustified plus, since it seems to believe that every book Sveshnikov position is awful for Black! 15 Qe2 Or 15 a3 Be6 16 Nce3 Bxe3 (16 … Ne7 is also equal) 17 Nxe3 Ne7 18 Qe2 Qb6 and Black stood no worse, R.RameshZ.Rahman, Madras 1995. 15 … Na7!? This move hasn’t been repeated, probably because most of us subconsciously frown upon decentralization. The alternative 15 … b4 is fine for Black; e.g. 16 Bb5 Bd7 17 Rfd1 bxc3 18 Nxc3 (or 18 bxc3 Ne7), J.Wallner-A.Lanc, Austrian League 2001, and now 18 … Qc7 is approximately equal, as 19 Qd3 can be met by 19 … Nb4 20 Nxb4 axb4 21 Nd5 Bxb5 22 Nxc7 Bxd3 23 Rxd3 Rbc8 24 Nd5 Rc2 25 Nxb4 Rxb2 with a balanced ending.

16 Nce3 Be6 17 a4 bxa4 As usual in the Sveshnikov, Black’s play down the b-file compensates the weak a5-pawn. 18 Rxa4 Bxe3 19 fxe3!? I’m not a big fan of voluntary defacement of structure. White seeks to open the f-file. I would prefer the more sober 19 Nxe3 Qb6 20 Rb1.

19 … Nc6 19 … Bxd5!? is a speculative pawn sacrifice. After 20 exd5 Nc8 21 Rfa1 f5 22 Rxa5 Ne7 Black may have full attacking compensation. 20 Bb5 I would play 20 Bc4 to recapture on d5 with a piece. The game looks balanced here. 20 … Bxd5! Don’t forget our rule to plug d5 whenever possible. 21 exd5 Ne7 22 Bc4 Qb6 I like this better than the risky pawn sacrifice 22 … f5!? 23 Rfa1 Ng6 24 Rxa5 f4, although it feels like Black has enough for the pawn. 23 b4?! 23 b3 is met by 23 … e4, intending … Ng6-e5. After 24 Rfa1 Nf5 25 Rxa5 Qxe3+ 26 Qxe3 Nxe3 27 Kf2 Nxc4 28 bxc4

Rb2+ 29 Ke3 Rxg2 30 c5 dxc5 31 d6 f5, it’s anybody’s game. The comp calls it dead even. 23 … axb4 24 cxb4 Advantage Black, for several reasons: 1. White’s pawns are extended and a liability. 2. Black’s healthy kingside majority is ready to shoot forward, whereas White’s b-pawn isn’t particularly dangerous. 3. White is stuck with a bad bishop versus a potentially good black knight. 24 … Ra8 25 Rfa1 Rxa4 26 Rxa4 f5?! Black should play 26 … Nf5!, pressuring f2. After 27 Kf2 Nh6! 28 h3 f5 White’s king is looking unhealthy. 27 Qf2! Ng6 28 Bf1 e4 29 Qd2 29 b5 was a better move, since then Black’s e-pawn hangs after … f5-f4. 29 … Ne5 Black should pounce on the chance to play 29 … f4!. 30 Qd4!? White may survive with 30 Ra6 Qc7 31 h3. 30 … Qd8!? Black avoids an ending in view of the passed b-pawn and intends to swing the queen over to the other side to attack White’s king. Indeed, after 30 … Qxd4 31 exd4 Ng4 32 b5 Ne3 33 Kf2 f4 White’s advanced and passed b-pawn is a tad scary. 31 Ra7?! A temporary initiative can create an illusion of power. This is a rare case where placing a rook on the seventh rank is a waste of time! White should follow the principle: passed pawns should be pushed and play 31 b5. 31 … Qg5 Threatening 31 … Nf3+, forking king and queen. Sometimes we just sense a combination coming, even before it is fully formed.

32 Qc3? White had to try 32 Be2 Rc8 33 Qd2 Nf3+ 34 Bxf3 exf3 35 Ra1 with an inferior yet still about playable position. 32 … f4! 33 b5 If 33 Rc7, then 33 … Qf5! (preventing the simplifying Rc8) 34 exf4 Qxf4 35 Qe1 Ng4 36 g3 Qf6 37 h3 Qd4+ 38 Kh1 Ne3 and White is busted. 33 … fxe3 34 b6

Exercise (combination alert): How did Black force mate? 34 … Rxf1+! Answer: Step 1: Attraction. 35 Kxf1

It isn’t a “decision” if our opponent makes it on our behalf. White’s king is dragged to the centre. 35 … Qf4+ 0-1 Step 2: Give a queen check on the f-file. If 36 Ke1 then 36 … Qf2+ 37 Kd1 e2+ forces promotion and mate. Game 20 Bal.Szabo-L.Lenic Hungarian League 2018 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6 Bxf6 11 g3 A fianchetto makes sense for White if playing for a draw, since it: 1. Suppresses Black’s … d6-d5 break. 2. Makes Black’s … f7-f5 break less palatable, as e4xf5 empowers White’s fianchettoed bishop along the h1-a8 diagonal.

11 … Bg5 Before White thinks about stopping the bishop move by playing h2-h4. Otherwise 11 … 0-0 12 Bg2 Bg5 transposes. 12 Bg2 0-0 13 0-0 Now that White’s king is safe, he can push his c-pawn either one or two squares on his next move. 13 … Be6 Challenging for d5 as usual, which Black can also do with 13 … Ne7. For example, 14 c3 (or 14 Nxe7+ Qxe7 15 c3 Bb7 16 Nc2 a5 17 Qd3 Bc6 18 Rad1 Rfd8 19 a3 and Black stood no worse, T.Davis-C.Lakdawala, San Diego 2002) 14 … Nxd5 15 Qxd5 Be6 16 Qd3 Qb6 17 h4 Bh6 18 c4?! (White was better off playing the f-rook to d1 and doing nothing), M.LitvinovD.Lintchevski, Samara 2016, and after 18 … Rfb8, Black looks slightly better. 14 c4!?

This move may be the seed of White’s coming difficulties. The challenge to the queenside comes at a cost as White just lost control over d4. 14 c3 is probably the better path, though after 14 … Ne7 15 Nc2 Nxd5 16 exd5 Bd7 17 Nb4 f5 18 Nc6 Qb6, I.OhatC.Kemeny, Hungarian League 2009, Black can ignore White’s knight and play on the kingside with the superior chances. 14 … Nd4 15 cxb5 This leads to opposing wing majorities and oppositecoloured bishops, where Black is better on both counts. However, 15 Qd3?! Bxd5! 16 exd5 Qb6 17 cxb5 axb5 18 Kh1 f5 leads to a strategically won game for Black, J.Urbina Perez-M.Gutierrez Valero, Lucena 2012. 15 … axb5 Black can also play the immediate 15 … Bxd5 16 exd5 axb5 17 Nc2 Nxc2 18 Qxc2 Qb6 19 Qc6 Qb8!, when White can’t easily challenge the c-file and is in trouble. 16 Nc2 White is unable to dodge a coming … Bxd5. If 16 Nb4? then 16 … Qa5 17 Nac2 Nxc2 18 Nxc2 Rfc8 and White is strategically lost. 16 … Nxc2 17 Qxc2 Bxd5! 18 exd5 White probably hoped for a draw here due to the oppositecoloured bishops, whose trajectories never overlap. Yet he stands worse for the following reasons: 1. He is unable to challenge the c-file, unless he is willing to damage his structure by playing f2-f4. 2. Black’s kingside majority is mobile, while White’s on the queenside is dormant. 3. As mentioned earlier in the book, such positions with opposite bishops nearly always favour Black. 18 … Qa5 I would prefer 18 … Qb6! 19 Qc6 Qb8!. 19 h4 Bh6 20 Bh3

Instead, 20 Qc6! Rfd8 21 b4! Qxb4 22 Rfb1 Qd4 23 Qxb5 Ra3 is similar to the game continuation but a slightly better version for White. 20 … Ra7 21 Qc6 Ra6 22 Qb7 Ra7 23 Qc6 Rd8 No draw. 24 b4?! White decides he has endured the bitterness of captivity long enough and violently attempts to free himself from Black’s pressure. However, the timing is wrong for this move and now his options narrow. Perhaps he should try and defend a pawndown opposite-coloured bishops ending by allowing his queen to get “trapped” with 24 a3!? Rc7 25 b4 Rxc6 26 bxa5 Ra6 27 Rfb1 Rxa5 28 Bf1 Rb8 29 Rb4 Bd2 30 Rb3 and maybe White can still hold the game. 24 … Qxb4 25 Rab1 Qd4 26 Rxb5 White has paid off the loan but not the interest. 26 … g6! Black avoids the back rank trap 26 … Rxa2? 27 Qxd6!, when White is back in the game. 27 Rb7

If it’s not one damned thing then it’s another. Just when White thought he had fixed problem A, problem B comes along. Exercise (combination alert): Black found a way to win a pawn by force. How did he accomplish this? 27 … Ra3! Answer #1: The monumental can be reached via incremental steps. Step 1: play the rook to a3, threatening … Rxg3+. Answer #2: 27 … e4! is just as strong, according to the computer. 28 Kg2 Not 28 Rxf7? Kxf7 29 Qc7+ Ke8 30 Qxh7 Rxg3+ 31 Kh2 Rxh3+ 32 Kxh3 Qd3+ 33 Kg2 Bf8 and Black consolidates.

28 … Rc3! Step 2: chase off the defender of d5 by transferring the rook to c3. 29 Rd7?! White had to try 29 Qb5, though after 29 … Rc2 30 Qb3 Rd2 the d5-pawn is doomed. 29 … Rxd7! This is more accurate than 29 … Rxc6 30 Rxd8+ Kg7 31 dxc6 Qd5+ 32 Kg1 Qxc6. 30 Qa8+ Kg7 31 Bxd7 Qe4+ 32 Kh2

Exercise (combination alert): White’s king is vulnerable. How did Black cash in? 32 … Rxg3! 0-1

Answer: Annihilation of defensive barrier. One against three is unfair odds and, if we close our eyes, we can imagine the unpleasantness which is coming the white king’s way. For example, 33 fxg3 (or 33 Kxg3 Bf4+ 34 Kh3 Qf3 mate) 33 … Qe2+ 34 Rf2 (34 Kg1 is met by 34 … Be3+) 34 … Qxf2+ 35 Kh3 Qf1+ 36 Kh2 Be3! forces mate in four moves.

Chapter Three 9 Bxf6 gxf6: The Tactical Line

You better not be squeamish about pawn structures, since in this chapter our structure turns into the car with the words “wash me” written on the filthy rear windshield. Have you ever had the feeling that a game moves in a pre-set direction, yet we can’t quite figure out where it leads? Well, that place is this chapter, which can turn into a bewildering array of positions, with a wide array of complexity levels. So let’s study it well and be better prepared than our opposition. Game 21 M.Brodsky-V.Kramnik Kherson 1991

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 You may well ask why Botvinnik’s spiritual heir plays the Sveshnikov. It’s a mysterious opening. I consider myself a pure strategist, in the fashion of Capablanca and Ulf Andersson. Yet I too play the Sveshnikov. In fact, it’s the only Sicilian I am able to play with some degree of competence. The young Kramnik was a huge influence in the Sveshnikov’s revived popularity with his advocation during the 1990s. 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6

This is the Chelyabinsk line, a clear structural divergence from the Classical 9 Nd5. With 9 Bxf6 gxf6, the following occurs: 1. White gives up the bishop pair. This can be useful for Black, especially if the game opens later, which it almost always does. 2. White may later pay for abandonment of the dark squares. 3. Black gets doubled f-pawns.

4. The f-pawns may not be doubled for long, since Black plays for an … f6-f5 break, sometimes twice, since there is a spare f-pawn on f7. 5. With the structures shattered on the kingside, and the queenside aired out, Black’s king lacks a completely secure haven. 9 … Qxf6?! is unthematic, simply because Black loses two tempi after 10 Nd5 Qd8, when we reach a position similar to the previous chapter, except that Black lost two moves! 10 Nd5 White evades the … b5-b4 fork, while occupying the d5hole. 10 … f5 Everything about the position feels transitional. Black chips away at White’s centre, while preparing to undouble the fpawns. 10 … Bg7 is also played but we don’t have room in the book to cover both lines. 11 Bd3

This is White’s most logical response, which develops, while covering e4. In this chapter we’ll also examine 11 c3, 11 exf5, 11 Bxb5, 11 Nxb5, 11 c4 and 11 g3. 11 … Be6 This way White no longer threatens e4xf5, since then the d5knight hangs. 12 Qh5!? This hyper-aggressive line reminds us of that awful 1950s sci-fi movie Attack of the 50 Foot Woman. Later we’ll look at 12 0-0 and 12 c3 Bg7 13 Qh5, while 13 Nxb5!? is covered together with other sacrifices on b5 further on in the chapter (see Game 36). 12 … Rg8! 12 … Bg7 13 0-0 f4 14 c4 – as once seen in J.R.KochG.Kasparov, Evry (simul) 1988 – scores well for White. 13 0-0-0?!

The computer frowns upon this shady sacrifice, which offers the g2-pawn in exchange for even more development. Next game we look at 13 g3. Note that 13 0-0?! is not a great idea for White, who walks straight into Black’s attack after 13 … f4 14 h3 Rg6 15 c3 Rc8 16 Be2 Rh6 17 Qf3 Qg5 18 Kh2 Ne7 19 Nxe7 Bxe7 20 Nc2 d5!, C.Onyekwere-A.Shabalov, Las Vegas 2006, when Black had a decisive attack. If 21 exd5?? e4 22 Qxe4 Bxd5 and it’s time for White to resign. 13 … Rxg2 We all have to make a living, don’t we? Grabbing the pawn is actually Black’s best move. 14 f4 After 14 Qf3 Bxd5! 15 Qxg2 Bxa2 Black gets huge compensation for the exchange. G.Hof-S.Marez, correspondence 2011, continued 16 b3?? (a misguided attempt to trap the bishop; 16 Rhe1 is correct) 16 … Qa5 17 Kb2 d5!, with a winning attack for Black.

Principle: when leading in development, open the game and create confrontation. In Kramnik: Move by Move, I described this position as one with “no sky above and no ground below”. It does look scary for Black, whose king floats in an already opening centre. However, White has his own share of troubles: 1. Black has the bishop pair, which increases in potency as the position opens. 2. White’s offside knight on a3 remains a strategic burden. 3. White’s king has castled into a dangerous attack. 4. Black’s king is safer than it looks, dead centre of the board, mainly due to the grip on the dark squares. 14 … Nd4! Targeting c2. 15 Ne3?! Why retreat a centralized piece? Perhaps in order to evade Black’s coming … Bxd5. 15 Kb1 is best, but even there Black stands clearly better after 15 … Bxd5 16 exd5 Rg4! (threatening … Rh4, trapping White’s queen) 17 h4 Rxf4 18 Rdf1 Rxf1+ 19 Rxf1 b4 20 Nc4 Qf6 21 Bxf5 Nxf5 22 Rxf5 Qg6 and Black is a pawn up. 15 … Rf2! Kramnik insists on a resolution of central pawn tension. 16 exf5? White underestimates Black’s initiative. Better were: a) 16 Rhf1, eliminating Black’s potent rook. b) 16 fxe5, although 16 … Bxa2 17 Nxf5 dxe5 18 Nxd4 exd4 19 Nb1 Rc8 20 Rhf1 Rxf1 21 Rxf1 Be6 leaves Black in control. 16 … Bxa2! Threatening … Rc8 and … Nb3 mate. 17 fxe5 How irritating when our opponent relies upon a statistical improbability and the damned thing works! Any hope of

trapping the bishop is disastrous for White. 17 b3?? Qa5 18 Kb2 d5 19 Nb1 Bxb1 20 Kxb1 Nxb3! 21 cxb3 Qa2+ 22 Kc1 Qb2 mate. 17 … dxe5 Still winning. The show-offy comp finds 17 … Rc8!, threatening mate on b3; e.g. 18 Bxb5+ Nxb5 19 Nxb5 Qb6 20 Nd4 dxe5 21 Ng4 exd4! 22 Rhe1+ Kd7 23 Nxf2 Bh6+ 24 Rd2 Rxc2+! 25 Kxc2 Qb3+ 26 Kc1 Qc4+ 27 Kd1 Bb3+ 28 Rc2 Bxc2 and it’s mate. 18 Nxb5

Exercise (combination alert): Each new hill becomes harder to climb for the exhausted white defence. Black has multiple paths to the win. Find the strongest one.

18 … Bh6!! Answer #1: Zwischenzug. This is even stronger than taking White’s offered knight. Answer #2: 18 … Qb6! 19 Nxd4 Bh6! 20 Rhe1 exd4 was also winning. 19 Rhe1 In case you were wondering why White didn’t take the loose bishop, here is the answer: 19 Qxh6?? is met by the beautiful finish 19 … Rxc2+! (removal of the guard; this is the old sci-fi movie cliché where the alien accosts us and demands to be taken to our leader) 20 Bxc2 (or 20 Nxc2 Nb3 mate) 20 … Ne2 mate. 19 … axb5 20 Bxb5+ Ke7 21 Qh4+ Black’s bishop remains invulnerable. 21 Qxh6?? is met by the same theme: 21 … Rxc2+! (an eye for an eye and bomb for a bomb) 22 Nxc2 Nb3 mate. 21 … f6 22 Qxf2 Bf7! Clearance. Threatening 23 … Ra1+ followed by … Nxb5+. 23 Bd3 Qb6! There is no defence to … Ra2. 24 Be4 Ra2! 25 c4 After 25 c3 the reply 25 … Ra1+ 26 Bb1 Ba2! 27 Kd2 Qxb2+ 28 Kd3 Qxf2 wins. 25 … Bxc4 White’s unfortunate knight is pinned. 26 Kb1 Qa5 Faster is 26 … Ra1+! 27 Kxa1 Qa6+ 28 Kb1 Ba2+ 29 Kc1 Qc4+ 30 Kd2 (30 Bc2 Nb3 mate) 30 … Nb3 mate. 27 Nd5+ Super-desperation! 27 … Bxd5 Good enough, even if Kramnik misses an immediate mate with 27 … Kf7! (threatening 28 … Ra1 mate) 28 b3 Ra1+ 29

Kb2 Qa2+ 30 Kc3 Qxb3 mate. 28 Qxd4 Ra1+ 29 Kc2 Rxd1! 30 Qxd1

Exercise (combination alert): The position is the not-quite finished portrait painting, still missing the eyes. Black to play and force mate in four moves. Answer: 30 … Qa4+! 31 Kc3 0-1 The worst part of blundering is that we have no time to grieve. White walks into a mate in one with 31 … Qc4. However, 31 Kb1 Ba2+ 32 Ka1 Bb3+ 33 Kb1 Qa2 would also be mate. Game 22 M.Versili-W.Standke Correspondence 2007

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Be6 12 Qh5 Rg8 13 g3

Sometimes we play sedately, perhaps to deflect suspicion of our planned misconduct to follow. It’s far more sensible to protect the g-pawn, since we saw White sacrifice it last game with disastrous results. 13 … Rg5 14 Qd1 After 14 Qxh7 Nd4 15 c3 Nf3+ 16 Ke2 Bxd5 17 exd5 e4 18 Bc2 Rg6 19 Qh3 Qg5 20 Rad1 Bg7 21 Nb1 Ke7 22 Nd2 Rh8 23 Qg2 Qg4 24 Kf1 Rh3! 25 Bb3 White is virtually in zugzwang! Hence 0-1 D.Hagelstein-J.Banet, correspondence 2003. 14 … Bxd5 Black again follows the Sveshnikov principle of plugging up d5 with a white pawn when possible. 15 exd5 Ne7 16 c3

The flashy 16 Nxb5 is ineffective. After 16 … Qb6 17 Nc3 Qxb2 18 Na4 Qd4 19 c4 Bg7 Black has a pleasant position. 16 … Bh6 17 Be2 If 17 Nc2 Kf8 18 a4 e4 19 Bf1 bxa4 20 Rxa4 Rb8 21 b4 f4 22 c4 a5! 23 b5 (23 Rxa5? is met by the undermining shot 23 … Rxb4!), G.Evans-A.Bubir, correspondence 2003, then 23 … Ng6 and Black looks better. 17 … Qb6 18 Nc2 Kf8 Castling by hand gets the king to safety. The untried 18 … a5, intending … b5-b4, also looks playable. 19 a4 This move was played in all 25 games in my database. It breaks up Black’s queenside pawns. As usual, Black achieves more than sufficient play against b2. 19 … bxa4 20 Nb4 e4 Black may later transfer a piece – probably the knight – to e5. Note that 20 … a5 doesn’t eject the knight, since it is met by 21 Qxa4. 21 Qxa4 21 Rxa4 a5 22 Nc6 Qxb2 also looks good for Black. 21 … f4!

Suddenly, White’s king doesn’t look all that safe; I already prefer Black’s position. 22 Qa5 Qb7 23 Bxa6 Principle: If something looks too good to be true, it probably isn’t. The position is fun, in a completely unsettling way. White wins a not-so-important pawn, while Black finds compensation from the fact that White loses time untangling from the pin. The computer still favours Black here, despite White’s extra pawn. 23 … Qd7! The queen heads in the direction of the white king. After 23 … Qa7 24 Qa3 Qc5 25 Nc6 Qxd5 26 Nxe7 Kxe7 27 Rd1 Qc5 28 Qxc5 dxc5 29 Bc4 Rb8 I prefer Black, though the most likely result is a draw. 24 Qa4 Qf5! Oh, no you don’t. No queen trade. A first step tends to encourage a second. White’s king is about to feel the heat. 25 Qd1 Ng6

25 … Rg6, intending to swing the rook to f6, is a possible improvement. 26 Nc2 Ne5! 27 Nd4 Qg6 28 gxf4 Rg1+ 29 Rxg1 Qxg1+ 30 Ke2 Qg4+ 31 Kd2 Qxf4+ 32 Kc2 Qxf2+ 33 Qe2 Qxe2+! An exchange of queens should not be interpreted as a plummeting bravado level. 34 Nxe2 Ng4 White has managed to evade a direct attack on his king. The trouble is that Black looks better in the ending, since he has connected and passed e- and f-pawns which are about to race down the board. White will race his b-pawn down too, but one versus two is not a good omen in the coming ending. 35 b4 f5 36 b5 Ne3+ 37 Kb3 Rb8 38 c4

I have some awful news for you: we are still in theory! Now the game becomes a promotion race, for which Black is better equipped.

38 Ra4 was an attempted improvement, which turned out to be no improvement. After 38 … Bg7 39 Ng3 Kf7 40 Nxf5 Nxf5 41 Rxe4 Be5 White’s passed b-pawn wasn’t enough and Black converted his extra piece, J.Achermann-C.Gilbert, correspondence 2011. 38 … f4 39 Nd4 f3 40 Nc6 Rb6! Using a full rook to blockade a single passed pawn seems like a bad deal for Black. But not doing so, and allowing the bpawn to shoot forward, would have been a lot worse. The comps still prefer Black here. 41 Bc8 Intending Bh3, halting Black’s passed f-pawn. 41 … Bg5! It takes hours to build a winning position and a fraction of inattention to ruin it. 41 … f2?! allows White to defend after 42 Bh3 Bf4 43 Ra2! Ng4 44 Kc3 Be3 45 Ra8+ Kf7 46 Nd4 Nxh2 47 Ra6! Rb8 48 Rxd6 Re8. The comp assesses this mess as equal. 42 Bh3

Exercise (planning): It appears as if Black is in trouble. After all, his passed f-pawn is blocked from promotion on f1 and White may even toy with Ra6 ideas. Find the single reply which keeps the advantage in Black’s hands: 42 … Nxd5!! Answer: Clearance. Black sacrifices his knight to break up White’s pawns and clear the path for his e-pawn. 43 Bf5! The only move. After 43 cxd5? Rxb5+ 44 Kc4 Rc5+ 45 Kb3 Rxd5 Black’s e- and f-pawns are too deep and he will win. 43 … Nf6! Covering the e4- and h7-pawns.

44 Kc2 After 44 Ra6! Rb7 45 Kc2 Be3 46 Nd8 f2 47 Bh3 Rd7! White is busted. 44 … Kg7 45 Nd4 Rb7 Black covers his second rank. 46 Ra6 f2 47 Bh3 Kg6! 48 Kd1 If 48 Rxd6 Bf4 49 Re6 Ra7 50 c5 Ra4 51 Nc6 Rc4+, White’s c-pawn falls. 48 … Be3! 49 Ne6 Bb6!

How annoying for White. Black has managed to blockade his passed b-pawn. 50 Ke2 Rb8 51 Ra3 h5! 52 Rg3+ Ng4! 53 c5!? Others don’t help: a) 53 Nf4+ Kg5 54 Ne6+ Kf5 55 Ng7+ Ke5 56 Nxh5 Ne3! 57 Rg5+ Kd4 58 Kxf2 Rf8+ 59 Ke2 Ra8! 60 Kf2 Kxc4 61 Kg3 d5 and Black’s passed pawns will prevail.

b) 53 Bxg4 hxg4 and White loses a piece if his rook chops the g4-pawn. 53 … Bxc5 54 Nxc5 dxc5 55 Bxg4 hxg4 56 Kxf2 Kf5 57 Rb3 Rb6! The passed b-pawn is blockaded and the rook ending is lost for White. 58 Ke2

If 58 Ke3 then 58 … Rh6 (threatening … Rh3+ and … Rxb3) 59 Rb1 Rh3+ (White’s king is expelled from his blockading post) 60 Kf2 Rxh2+ wins. 58 … Kf4 White is only one pawn down, yet it feels as if his position hasn’t eaten in a week. The rook ending is winning for Black. 59 Rb1 c4 60 Rf1+ Ke5 61 Rc1 Kd4 62 Rd1+ Kc3! 63 Rb1 Kc2 64 Rb4 c3 65 Ke3 Kc1! 66 Rxe4 Rxb5 67 Rxg4 Kb2

It becomes clear that Black will win a rook. The only question is: can White hold a draw with his king and h-pawn? The answer is no. Black’s king gets back in time. 68 Rc4 c2 69 Kf4 Rb4!

Black also wins with the prosaic 69 … Kb3 70 Rc6 Rb4+ 71 Ke3 Kb2 72 Kd2 Rd4+ 73 Ke3 Rh4 74 Rb6+ Kc1 75 Ke2 Rxh2+ 76 Ke1 Rh5 77 Ke2 Re5+. We have reached a basic Lucena position, which is an easy win for Black. If you didn’t know this, then immediately order First Steps: Fundamental Endings. Modesty prevents me from naming the author. The remainder runs: 78 Kf3 (or 78 Kd3 Kd1 and Black’s king is shielded from rook checks; he promotes next move) 78 … Kd2 (this technique is called building a bridge) 79 Rd6+ Kc3 80 Rc6+ Kd3 81 Rd6+ Kc4 82 Rc6+ Rc5 wins. 70 Rxb4+ Kc3 71 Re4 c1Q+ 0-1 A human would play on with queen versus rook, whereas it’s a total waste of time against a computer, who will win with the queen 100% of the time.

Game 23 A.Volokitin-M.Oleksienko Ukrainian Championship, Rivne 2016 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Be6 12 c3

White hurries to bring the wayward a3-knight back into the game via c2. 12 … Bg7 13 Qh5 Externally, this is one of the scariest lines for Sveshnikov players, but not after the lines have been run through the computer. 13 … 0-0! I promise you. Black is okay here. Trust the comps! 14 exf5

This is the start of a scary-looking forcing line. We’ll examine the calmer 14 0-0 in two games time. 14 … Bxd5 15 f6

Black resigns? Not quite, since we have access to two adequate resources. If you are just starting with Sveshnikov and haven’t ever seen this position, be careful not to reject it out of pre-existing beliefs. Relax. Black’s position is fully playable. 15 … e4 It’s time for us to enter territory inhabited by dragons and other fearsome creatures. As Tal use to say, the opponent can only take one piece at a time. 15 … h6 is also okay for Black. 16 fxg7 (16 Qf5 e4 17 fxg7 Re8 is similar to the game continuation) 16 … Kxg7 17 Nc2 Qg5 18 Qxg5+ hxg5 and Black had reached a decent ending, R.Bes-V.Kurgansky, correspondence 2010. 16 fxg7 Re8 17 Be2

The only move. 17 Qxd5? is a bad idea. White’s lag in development is of far more concern than Black’s awful structure. Play can continue 17 … exd3+ 18 Kf1 Ne5 19 Qd4 Qg5 20 Rd1 Qh5! 21 f3 Nxf3! 22 Qxd3 Ne5 23 Qd5 Rac8! 24 h3 Rc5! 25 Qd4 Re6 26 Kf2 Qf5+ 27 Kg3 (27 Kg1 Nf3+! 28 gxf3 Qxf3 forces mate) 27 … Qg5+! 0-1, since Black forces mate, no matter which way White’s king moves, K.SpraggettB.Amos, Calgary 1975. 17 … Ne5 The knight is powerfully posted on e5. It may even later hop into d3. Next game we look at Black’s main move 17 … Re5. 18 0-0 Qf6 19 Rfd1 Aimed at discouraging … Nd3. Instead, 19 Qh3 Bc4 20 Nxc4 bxc4 21 Qe3 d5 (Black has equalized) 22 f3 exf3 23 Bxf3 Nxf3+ 24 Qxf3 Qxf3 25 Rxf3 Rab8 26 Rf2 Kxg7 27 Rd1 f6 28 g3 ½-½ was Salcedo P.Mederos-M.De Heer, correspondence 2014. 19 … Nd3!

Maybe … Nd3 isn’t discouraged after all! Normally, an imagination-less rationalist like me loathes speculation, yet I fully endorse this pawn sacrifice, after which it is White who struggles to maintain equality. It’s shocking to have played an opening for decades and then come to the realization that there are still some secrets left unresolved. The Sveshnikov is not a rigid system. It generally flows and shifts along with our ambitions. Black gets full compensation for the pawn. 20 Bxd3 Not 20 Qxd5?? Qxf2+ 21 Kh1 Qxe2 22 Rf1 Ra7 23 Qb3 (23 Qxd6 Qxb2 is hopeless for White) 23 … e3 and White is paralysed. 20 … Re5! Black defends his bishop with tempo. 21 Qh3 exd3 22 Rxd3 Rae8 Threatening mate on the back rank. 23 Nc2? White maintains the balance only by returning the pawn immediately: 23 Rf1! (this is the magic protection spell that keeps White safe) 23 … Bxa2 is equal.

Exercise (combination alert): White’s last move loses material. Demonstrate how. 23 … Re2 Answer: Step 1: Start with a double attack on f2 and c2. 24 Qg3 The f2-square must be protected. 24 … Be4! Step 2: Skewer. White loses the exchange. 25 Ne1 Bxd3 26 Nxd3 Qg6 The ending is completely lost for White; the black rook invades his second rank. 27 Rd1 Qxg3 28 hxg3 Kxg7 29 Kf1 Kf8 30 g4 a5 31 g3 a4 32 a3 Rc2 33 Kg2 Ree2 34 Kf3 Red2! 35 Rxd2 Rxd2 36 Ke3

Rc2 White’s knight is eternally pinned down to defence of b2. 37 Ke4 h6 38 Ke3 White can do nothing but mark time and hope to set up a fortress. If 38 Kd5 Rd2 39 Kd4 Ke7 40 c4? then 40 … b4!, when 41 axb4 is met by 41 … Rxb2! 42 Nxb2 a3 43 Kc3 a2 and Black promotes. 38 … Ke7 39 Ke4 Ke6 40 Ke3

Exercise (planning): There remained a measure of vagueness to Black’s motives, until now. Let’s create the scaffolding for us to get from point A to point B. Work out Black’s multi-step winning plan. 40 … Kf6!

Answer: Black’s winning plan: Step 1. Play the king to g5, provoking White into weakening with f2-f3. Step 2. Play … Rg2 and … Rxg3. Step 3. Play … f7-f5, inducing g4xf4. Step 4. Push the newly passed h-pawn down the board. 41 Kf4 Kg6 42 Ke3 Kg5 43 f3 Step 1 completed. 43 Kf3? f6! is zugzwang. 43 … d5! Zugzwang! 44 Ne5 Or 44 Kd4 Rd2 45 Ke3 Rg2 46 Kd4 Rxg3 (step 2 completed) 47 Ke3 f5! (step 3 activated) 48 gxf5 Rg2 and all that remains is step 4: Black’s h-pawn wins the game. 44 … Rxb2 45 Nxf7+ Kf6 46 Nxh6 White’s kingside pawns are too slow. 46 … Rb3 47 Kd2 If 47 Kf4 Rxc3 48 g5+ Kg6 49 Kg4 Rxa3 50 f4 Rc3 51 f5+ Kh7 52 Kh5 Rxg3 53 g6+ Kg7, White’s promotion attempt fails. 47 … d4 47 … Rxa3 also wins. 48 Nf5 Or 48 cxd4 b4! 49 axb4 a3 50 Kc2 Rxb4 and Black promotes. 48 … dxc3+ 49 Kc2 Rxa3 50 Ne3 Ra2+ 51 Kd3 After 51 Kxc3 Ra3+ 52 Kd4 Rxe3 53 Kxe3 a3 White’s king falls outside the square of Black’s passed pawn. 51 … Ke6 51 … b4? complicates Black’s win by hanging the b-pawn after 52 Nd5+. 52 Nc2

Exercise (planning): What is Black’s clearest winning plan? 52 … Rxc2! Answer: Simplification. Black promotes first in the king and pawn ending. 53 Kxc2 b4 54 g5 Kf5 55 f4 Kg6 56 g4 Kf7 57 f5 Kg7 58 g6 Kf6 59 g5+ Kg7 0-1

Zugzwang! White’s pawns are frozen, while Black’s will move forward; e.g. 60 Kb1 b3 61 Kc1 a3 62 Kb1 a2+ 63 Ka1 c2 64 Kb2 and now Black has a simple deflection with 64 … a1Q+, winning. Game 24 J.Wang-Le Quang Liem Burlingame 2019 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Note that this line can also be reached via 11 c3 Bg7 (11 … fxe4? 12 Bxb5 axb5 13 Nxb5 is in White’s favour) 12 Bd3 Be6. 11 … Be6 12 c3 Bg7 13 Qh5 0-0 14 exf5 Bxd5 15 f6 e4 16 fxg7 Re8 17 Be2 Re5

Rather than 17 … Ne5 as seen in the previous game. Moving the rook into e5, with tempo, is Black’s main line. 18 Qh6 Or 18 Qh3 Rg5 19 Nc2 Ne5 20 Nb4 Be6 21 Qe3 a5 22 Nc2 Nd3+! (by now this pawn sacrifice should be a familiar theme) 23 Bxd3 exd3 24 Qxd3 Bf5 25 Qd2 Bxc2 26 Qxc2, E.TeodoroB.Selitsky, Dearborn 1992. Here Black stands clearly better if he finds 26 … Qe8+! as 27 Kf1 Qc6 28 f3 Qc4+ 29 Qe2 (29 Kf2? Qc5+ 30 Kf1 Re8 and White is busted) 29 … Qxe2+ 30 Kxe2 Rxg2+ 31 Kd3 Rxb2 and White is struggling in the ending. 18 … b4! This logical and thematic break either disrupts White’s pawns or allows Black’s knight a free hop into b4.

19 Nc2! I believe this is White’s best chance to extract something. Getting his knight into play is more important than structural integrity. If 19 cxb4 Nxb4 20 0-0 Qg5 21 Qxg5 Rxg5, Black stands no worse in the ending. 19 … bxc3 20 bxc3 Qa5!? The computer doesn’t much like this move, but as we know by now it often gets fooled in its evaluation of Sveshnikov positions. 20 … Be6 is the usual move, when White is unable to extract an advantage; e.g. 21 0-0 Qg5 22 Qxg5 Rxg5 23 Rfd1 Rd8! 24 Bxa6 Rc5 25 c4 Ne5 26 Ne3 Bxc4 27 Nxc4 Nxc4 28 a4 d5 29 Bxc4 Rxc4 30 a5 d4, ½-½ Salcedo P.Mederos-A.Solari, correspondence 2015. 21 Qd2 Or 21 0-0 Be6 22 c4 (after 22 Ne3! Rg5 23 h4 Rxg7 24 Qf4 Qe5 25 Qxe5 Nxe5 the ending looks only slightly in White’s favour) 22 … Rg5 23 Qh4 Re8! 24 Qxe4? (the pawn is

poisoned; White should have been satisfied with an unclear position after 24 Ne3) 24 … Bd7 25 Qf3 was J.HackerJ.Ramsdal, Ballerup 2019. Now 25 … Qd2! 26 Bd1 Ne5 27 Qe2 Bc6! 28 Qxd2 Rxg2+ 29 Kh1 Rxf2+ 30 Kg1 Rxd2 gives Black a winning position. 21 … Be6 Or 21 … Rc8 22 0-0, J.Nunn-J.Tisdall, Hastings 1978/78, and chances are balanced after 22 … Ne7. 22 0-0 Rd5! A new move, which first kicks White’s queen off the d-file before shifting the rook over to c5. The move is an attempt to improve upon the also playable 22 … Rc5 23 Nb4, W.Drygalski-H.Lew, correspondence 1991, where Black is okay after 23 … Qa3.

23 Qe3 Rc5

White’s c-pawn is doomed but he should get enough play for it to remain level. 24 Nb4 Nxb4 25 cxb4 Qxb4 26 Rab1 Qa4 27 Rb6 Rd5 Black wants to hang on to his hard won extra pawn. 28 Rfb1 Rc8 28 … Qd4 is also equal. 29 h4 After 29 Qf4! a5 30 h3 Qxa2 31 Ba6 Re8 32 Qg3 Re5 it’s anybody’s game. 29 … Qd4! Centralization ensures that Black’s king stays relatively safe. 30 Qf4?

The players’ conclusions radically differ. This psychological mistake is a form of confirmation bias. White wants to attack so he believes there is one, even where it’s not there. White

attempts to grab hold of an apparition and misses the fact that he cannot retain queens on the board after Black’s next move. Correct was 30 Bxa6! (principle: it’s better to avoid a problem rather than force one) 30 … Qxe3 31 fxe3 Rc3 32 Bb7 Rd2 33 R6b2 Rxb2 34 Rxb2 Rxe3 35 Kf2 Ra3 36 Bxe4 Rxa2 37 Rxa2 Bxa2 38 Ke3 Kxg7. White’s pawn is gone, but is it going to be missed? White will suffer a pawn down in the bishop ending, yet has excellent chances to hold the draw. 30 … Qd2! White’s attack is killed in its crib. Queens come off anyway, with White losing material in this version. 31 Qe3 If 31 Qxd2 Rxd2 32 Bxa6 (32 R1b2 Rxb2 33 Rxb2 a5 is a winning ending for Black) 32 … Rcc2 33 Rf1 Rxa2 34 Bb7 d5 (threatening … e4-e3) 35 Rb3 Kxg7, White is a pawn down and completely tied up. 31 … Qxe3 32 fxe3 a5 33 R1b2 Kxg7

Black is a pawn up with the more active pieces and a technically won ending. 34 Kf2 Rc3 35 g4 a4 36 R2b5 Rxb5 37 Bxb5 37 Rxb5 Rc2 38 Kf1 Rxa2! is dead lost for White. Not 38 … Rxe2??, falling for White’s trap: 39 Rg5+! is a powerful zwischenzug as there is no hanging rook on b5 and White is the one who will be up the exchange and winning. 37 … Rc2+ 38 Kg3 Rb2! White is totally tangled up and can resign. 39 Kf4 a3 40 Rxd6 Desperation. If we have one giant sin on our conscience, another one fails to increase the weight, since we are condemned to hell anyway. 40 … Rxb5 41 Ra6 Rb2 42 Rxa3 Rh2! 43 h5 Rxa2 44 Rc3 Rg2 45 Kxe4 Rxg4+ White’s chances to hold the draw against a Fide 2700+ player are zero and he could have saved energy by resigning.

46 Kf3 Rg5 47 Kf4 Rxh5 48 Rc1 Rh4+ 49 Kg5 Rg4+ 50 Kh5 The king is left to sleep it off on the park bench. 50 … f5 51 Rc7+ Bf7+ 52 Rxf7+ Kxf7 53 e4 Wow. This is what I call an optimist. It’s not easy to combine defiance with humour. I’m assuming White is impervious to pain, since he refuses to resign a full rook down. 53 … Kf6 54 exf5 Rf4 0-1 Game 25 V.Kramnik-L.Van Wely Wijk aan Zee 2005 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Be6 12 c3 Bg7 13 Qh5 0-0 14 0-0

Since the comps worked out 14 exf5 to the point where Black is guaranteed a decent position, White postpones the

move and simply castles, leaving Black to deal with the threat in the future. This is a calmer version of the line from the last two games. 14 … fxe4 Fluid. 14 … f4 releases the central tension and is also playable, but I’m not convinced that Black reaches full equality after 15 Rfd1 Rb8 16 Nc2 Qd7 17 h3 (covering against … Bg4) 17 … f5 18 Ncb4 Nxb4 19 Nxb4 a5 20 exf5 Bf7 (after 20 … Bxf5 21 Nc6! Rbe8 22 Bxb5 Black’s compensation is unclear) 21 Qh4 axb4 22 f6 Bg6 23 fxg7 Qxg7 24 Bxg6 Qxg6 25 cxb4 and White has the better chances, Joh.Anderson-S.Kerr, correspondence 2006. 15 Bxe4 f5 16 Nf4! We must be ready for the forcing sequence 16 Nf6+?? Rxf6 17 Bxc6 Rb8, threatening to trap the bishop with … d6-d5 and … Rb6. If White plays 18 Bf3?? the pain is increased further after 18 … Rh6, as the white queen is trapped. 16 … exf4 Are we in a battle between a player of the 21st century and one from the 19th century? It certainly looks that way. At this point, some readers may be freaking out about Black’s pawn formation, with four isolanis, two of them doubled and with more to come!. On the surface it may seem that Black’s resolve to destroy his structure swells in inverse proportion to White’s strategic wealth. Now I’m going to tell you the truth, the whole truth and nothing but the truth. This is not so. Black’s piece activity easily compensates. Komodo 13 assesses it at dead even. Though Black must avoid the following traps: a) 16 … Bd7?? 17 Bd5+ Kh8 18 Ng6 mate. I’m afraid if you fall for this one, your license to play the Sveshnikov is revoked! b) 16 … Bf7 17 Qxf5 Qh4 18 Bxc6 and Black can resign. Two unfortunates in my database fell for this one. 17 Bxc6 Rc8!

18 Qe2 If 18 Bb7 Rb8 19 Qe2 (not 19 Bxa6? Qa5 20 Bxb5 Rxb5 21 Nxb5 Qxb5 and Black’s bishops are worth more than White’s rook and pawns), A.Mista-O.Pucher, Plancoet 2012, then 19 … Bf7 20 Bf3 b4 21 cxb4 Rxb4 gives Black good play. 18 … Be5 Or 18 … Qf6 19 Bb7 Rce8 20 Qf3 b4 21 cxb4 Qxb2 22 Rab1 Qxa2 23 Bc6 Rc8 24 Rfe1 Be5!? (a speculative queen sacrifice; 24 … Bb2 25 Rxe6 Rxc6! 26 Qxc6 Qxe6 27 Qxa6 Be5 is only slightly in White’s favour) 25 Re2! Qxe2 26 Qxe2 Rxc6 27 g3? (correct is 27 Nc2 f3 28 gxf3 Rc4, when Black gets compensation for the sacrificed queen; maybe not full theoretical compensation but excellent practical chances) 27 … fxg3 28 hxg3 f4 29 b5 Rc3 30 b6?? (this looks like a timepressure error; 30 gxf4 Rxf4 31 b6 Rg4+ 32 Kf1 Rh4 is rated as even by the comp) 30 … fxg3 0-1 as the white king is shredded of cover, A.Ramaswamy-K.Haznedaroglu, Abu Dhabi 2004. 19 Qf3

19 Bf3 is also possible. 19 … b4!? In the Sveshnikov, we must pick between the following choices: 1. Wait patiently for something to happen. 2. Take risky action and make it happen. Our choice nearly always veers to number 2 on the list. This move leads to yet another isolani for Black. In return, the game is opened for his bishops and his activity level increases even further. 20 cxb4 Bxb2 One side seeks to build something, while the other side wants to tear it down. 21 Bd5! Based on the following principles:

1. When your opponent has the bishop pair, swap one of them off. 2. The structurally superior side always benefits from exchanges, since this in turn makes the position more predictable and stable, disallowing the structurally inferior side to complicate as much. After 21 Rab1 Qf6 22 b5 Bxa2 23 Bd5+ Bxd5 24 Qxd5+ Kh8 25 Nc4 Qd4 26 Qxd4+ Bxd4 27 Nxd6 Rcd8 28 Nc4 axb5 29 Rxb5 Kg7, White was unable to exploit Black’s doubled fpawns and the game was drawn, R.V.Hall-P.Svacek, correspondence 2006. 21 … Qf6 22 Rad1 Rc3 23 Rd3 Bxd5 24 Qxd5+ Qf7 25 Qf3 We love our relatives because we have to. With chess openings, we are free to pick and choose. The Sveshnikov is not for everyone. To play it, we agree to two things: 1. A massive theoretical lift, where theory goes on and on, twisting and turning in labyrinthine detail.

2. We cannot simultaneously desire a perfect structure and play the Sveshnikov. Part of the deal is that we often deliberately take on compromised pawns, in return for ferocious piece activity. Theory ends here. In 2005, Kramnik and Van Wely were two of the premier Sveshnikov players in the world, and I bet this position was still in, or close to, their pre-game analysis. Now you may look at Black’s all-isolani cast and have second thoughts. Once again, we must rely on Black’s activity to compensate for the poor structure. The comp gives White only the tiniest of edges, and I’m not even sure I agree with that.

Black’s activity may provide full compensation for his wretched structural defects. Also, Black’s bishop will be superior to White’s remaining knight. The supposition is theory, and practice should be exactly the same – but they are not! We can learn the

moves to get to this position but we still must develop a feel for Black’s coming plans. Van Wely’s play in this game is a perfect example of how to handle a position, weak in structure, yet abundant in activity. In Sanchez J.Rodenas-Joh.Anderson, correspondence 2008, White tried 25 Nb1 but was unable to squeeze anything out of it after 25 … Qxd5 26 Rxd5 Rc6 27 Rfd1 Rb8 28 a3 Be5 29 Nd2 Rb5 30 Rxb5 axb5 31 Nf3 Rc3 32 Nxe5 dxe5 33 h4 Rxa3 ½-½. 25 … Rfc8 26 Nb1 Rc1 27 Nd2 Rxf1+ 28 Nxf1 Be5 29 Ra3 Qc4 30 Qh5 The white attackers lack the numbers to cause Black’s king any trouble. 30 … Qf7 31 Qf3 White’s attack ends before it even begins. After 31 Qxf7+ Kxf7 32 Rxa6 Rc2 (threatening … Bd4, targeting f2, similar to the game) 33 Ra3 Rb2 34 Rb3 Rxa2 35 b5 Bd4 36 Rf3 Ke6 37 Rxf4 Bc5 White cannot hold on to the extra pawn and the game is a likely draw. 31 … Qc4 32 Qb7 Okay, let’s try the other side. 32 … Rc7 33 Qa8+ Rc8 34 Qxa6 Qxa6 35 Rxa6 Rc2 Threatening 36 … Bd4. 36 Ra3 Rb2 37 Rb3 Rxa2 38 g3 Here 38 b5 Bd4 39 Rf3 (forced) 39 … Be5 40 Rb3 Bd4 is a repetition draw, since 41 b6?? Rxf2 42 Nd2 Rf3+ 43 Kh1 Rxb3 44 Nxb3 Bxb6 is suicidal. White is two pawns down with his king cut off in a hopelessly lost ending.

38 … fxg3 39 hxg3 Bd4 40 Ne3 Kf7 Reinforcements are on their way. Black’s king will help stop White’s passed pawn. 41 b5 Bb6 42 Kf1 The cost for removing the b6-blockade is the f-pawn. Instead, 42 Nd5 Bxf2+ 43 Kf1 Ke6 44 Nf4+ Kd7 45 b6 Kc8 46 b7+ Kb8 47 Nd5 Bc5 48 Ne7 Ra6 (preventing Nc6+) 49 Ke2 Rb6 50 Rxb6 Bxb6 51 Nxf5 Bc5 52 Ne7 Kxb7 leaves White a pawn down, although he should hold a draw. 42 … Ke6 43 Nc4 Bxf2 44 b6 d5! 45 b7 Ba7 46 Rb6+!

Flashy but ineffective from the standpoint of playing for the win, since there is no requirement for Black to take the rook. Exercise (critical decision): Black’s king can move to e7 or d7. Only one of the squares saves the game. Which one? 46 … Kd7! Answer: Only the d7-square draws. Not 46 … Ke7?? which is met by 47 Ne5! and White wins decisive material with the threat of Nc6, taking the bishop and then promoting the passed b-pawn. 47 Rd6+ Kc7 ½-½ After 48 Rxd5 Ra4 49 Ne5 Kxb7 50 Rd7+ Ka6 51 Rxh7 Ra2 the position is a certain draw. Game 26

E.Inarkiev-B.Gelfand Nazran (6th matchgame) 2017 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Be6 12 0-0

This sensible move is the main line. White simply castles and continues to develop. 12 … Bxd5! Remember our Sveshnikov rule of thumb: given a chance, plug d5 with a white pawn. 13 exd5 Ne7 At first glance it may seem that Black just allowed a combination with 14 Nxb5, but Black does not need to fear this possibility. 14 c3 The ineffective “combination” 14 Nxb5 is seen in the next game.

14 … Bg7 Now negating the Nxb5 trick. 15 Qh5 Here 15 Nc2 0-0 16 a4 e4 17 Be2 bxa4 18 Rxa4 Qb6 19 Rb4 Qc5 20 Ne3 Rab8 21 Qb3 Rxb4 22 cxb4 Qc8 23 g3 Ng6 24 Qc4 f4 25 Qxc8 Rxc8 26 gxf4 Rb8 27 Bxa6 ½-½ was D.SadvakasovA.Khalifman, Russian Team Championship 2005. I would decline a draw with Black, who looks slightly better after 27 … Rxb4. 15 … e4 The g7-bishop’s power just increased tenfold. 16 Be2 Instead 16 Bc2 is illogical, since it hinders the a3-knight’s re-entry into society. 16 … 0-0 16 … b4?! is mistimed. After 17 cxb4! Bxb2 18 Rab1! Bxa3 19 Rb3 White regains the temporarily sacrificed piece with advantage. 17 Nc2 Re8 Others: a) 17 … Rc8 18 Ne3 f4 19 Nf5 Re8 20 a4 Ng6 21 axb5 axb5 22 Ra6 Bf8 23 Bxb5 Re5 24 Qh3 Rb8 25 Nd4 Qf6 and Black’s chances on the kingside and centre fully compensate for White’s extra pawn, V.Turkov-F.Finocchiaro, correspondence 2009. b) 17 … f4 18 Qg5 f5 19 Qxf4 Nxd5 20 Qd2 Nb6 21 Rad1 d5 22 Nb4 Qd6 23 Nxd5 Rfd8 24 c4 Nxd5! (White’s powerful knight must be eliminated; 24 … Nxc4? 25 Bxc4 bxc4 26 Qg5! is heavily in White’s favour) 25 cxd5 Rac8 and chances are balanced, despite White’s extra pawn, E.Calhau-G.Poli, correspondence 2008. 18 g3 Or 18 Rad1 f4 19 Nb4 Qb6 20 Qg4 a5 21 Nc6 Nxc6 22 dxc6 Qxc6 23 Qg5 Re5 24 Qxf4 d5 and Black stood no worse, J.Vosahlik-V.Iotov, correspondence 2007.

18 … f4!

A good new move. Gelfand wants to go after his opponent’s king. This is one of those positions that fools the computer, who calls it even. If you play the logical progression out about ten moves, then the evaluation always seems to swing in Black’s favour, no matter how White plays it, as is the case so often with the comp’s (mis)assessments in the Sveshnikov. Instead after 18 … Nxd5 19 Qxf5 Re5 20 Qh3 Qb6 21 Rfd1 Ne7 Black can play for a break on b4 with … a6-a5 and … b5b4, and chances look even, E.Van Leeuwen-M.De Oliveira, correspondence 2012. 19 gxf4?! With this move White forges the future chains which bind him. Inarkiev doesn’t like the idea of retaining tension and keeping Black’s central pawns fluid for too long. Perhaps he thinks he will be doing the attacking down the g-file, after Kh1 and Rg1.

The correct move was 19 Bg4 to seize control over f5. After 19 … Ng6 20 Bf5 Re5 21 Rad1 Qf6 22 Nd4 Rxd5 23 Rde1 Re8 24 Qh3 Rxd4! 25 cxd4 d5, the comp calls it dead even, while stylistically I prefer Black. 19 … Ng6 20 f5 Nf4! Black’s remaining bishop proves to be superior to White’s knight. 21 Qg4 Nxe2+ 22 Qxe2 Qh4 23 Kh1 Kh8 Both kings hurry off the open g-file. The game soon reveals that Black is the one doing the attacking. 24 Rg1! This is White’s best path. His game degrades if he attempts to eliminate Black’s e-pawn with 24 f3?! exf3 25 Qxf3 Be5 26 Rf2 Rg8 27 Raf1 Rg4 28 Nb4 Rag8 29 Nd3 Qg5 (threatening mate on g1) 30 Qd1 Bf6 31 Qc1 Qh5 32 Qd1 Qh6 and White is in trouble. If 33 Qc1?? then 33 … Bg5 34 Qe1 Be3 wins, since White is unable to move his rook away from f2; i.e. 35 Re2? Qh3 forces mate in three moves. 24 … Be5 Threatening mate on h2. 25 Rg2 Rg8 26 Rag1 Rxg2 27 Rxg2 Rg8 28 Qf1 Qh5 29 Ne3?!

White has more hope of surviving in the line 29 Rxg8+! (principle: when the opponent is attacking, exchange pieces) 29 … Kxg8 30 Qg2+ Kf8 31 f6! Qd1+ 32 Qg1 Qf3+ 33 Qg2 Qxf6 34 Kg1 Qh6 35 Qxe4 Qxh2+ 36 Kf1, although Black still has all the winning chances. 29 … Rg5! 30 c4

White was perhaps possessed by an irresistible urge to lash out at something – anything. 30 … Qf3! Black’s pressure grows to unbearable levels. 31 cxb5 axb5 32 b3 White is in no position to go pawn grabbing with 32 Qxb5?? as then 32 … Rxg2 33 Nxg2 Qd1+ wins. 32 … Bf4

Threatening 33 … Bxe3. The calm 32 … b4 was also strong. 33 Kg1 Not 33 Nc2?? Rxg2 34 Qxg2 Qd1+ and wins. 33 … Bxe3 34 fxe3 Qxe3+ 35 Kh1 Rxg2 36 Kxg2

Or 36 Qxg2 Qd3 37 h4 f6 and White is busted. 36 … Qd2+ 37 Kh3 Qxd5 Here 37 … b4! 38 Kg4 e3 39 Kf3 Qxa2 is easier for Black. 38 Qf4! f6! 39 a4 White desperately attempts to create play with a passed apawn. 39 … Qd3+ Euclid never said that the shortest distance between two points is a curved trajectory! Simpler is 39 … Qxb3+!, after which White’s position faces economic excommunication; e.g. 40 Kh4 bxa4 41 Qxe4 a3 42 Qe8+ Kg7 43 Qe7+ Qf7, when there is no perpetual check and Black’s a-pawn wins the game.

40 Kg2 e3! 41 Qf3! 41 Kf3?? is a fatal misstep. After 41 … e2+ 42 Kf2 Qd1 43 Qb4 Qf1+ Black promotes. 41 … bxa4 42 bxa4 Qc2+ 43 Kg3 Qxa4 44 Qxe3

Black’s extra pawn is decisive, yet there is one loose end, so he can’t just relax and declare victory. He has to be certain that White doesn’t have access to perpetual check. 44 … Qb5 45 Kg4 Qe5 46 Qf3 d5 47 Kh3 Qe4! 48 Qh5 48 Qxe4? dxe4 49 Kg4 Kg7 50 Kf4 Kh6 51 Kxe4 Kg5 52 h3 h6 is zugzwang. Black wins the king and pawn ending. 48 … Qe3+ 49 Kh4 d4! 50 Qd1 After 50 Qf7 Qg5+ 51 Kh3 Qxf5+ 52 Kh4 Qf4+ 53 Kh3 Qh6+ 54 Kg3 Qg7+ the queens are removed from the board and White must resign. 50 … d3 51 Qb3 Qf4+ 52 Kh3 Qxf5+ Now Black’s win is clear.

53 Kh4 d2 54 Qd1

Exercise (planning): Find Black’s most efficient route to victory. 54 … h5! 0-1 Answer: There is no defence to the coming … Qg4+, removing queens. The problem with 55 h3 is 55 … Qg5 mate. Game 27 Liu Guanchu-Yu Yangyi Chinese Team Championship 2015 1 e4 c5 2 Nf3 Nc6 3 Nc3 e6 4 d4 cxd4 5 Nxd4 Nf6 6 Ndb5 d6 7 Bf4 e5 8 Bg5 a6 9 Na3 b5 10 Bxf6 gxf6 11 Nd5 f5 12 Bd3 Be6 13 0-0 Bxd5 14 exd5 Ne7 15 Nxb5

Did Black just walk into a combination? Yes and no. White does win a pawn, but GM Yu Yangyi has deliberately allowed it. Black gets excellent play for the pawn. 15 … Bg7 16 Nc3 e4 17 Bc4 This way White: 1. Reinforces the d5-pawn. 2. Retains pressure on the a6-pawn. 3. Prepares to cover the b2-pawn with Bb3, if necessary. Instead, 17 Be2 Ng6 18 Re1 0-0 19 Bf1 Re8 20 Qd2 Rb8 21 Rab1 can be met by 21 … Ne5!?, offering a second pawn to flame his initiative, which Black justified by winning in Zhao Jun-Wei Yi, Chinese Team Championship 2015. Otherwise, the safer 21 … Qc8 22 Ne2 f4 23 Nd4 Qc5 24 Nf5 Rxb2 25 Rxb2 Bxb2 26 c4 Bg7 is about even. 17 … 0-0 18 Qh5 After 18 Qd2 Ng6 19 Bb3 Re8 20 Ba4 Re5 21 Rae1 Qh4 22 Qd1 Rb8 23 Bb3 Rbe8 24 Re3 f4, Black received massive

attacking compensation for the sacrificed pawn and won both games from this position in my database. 18 … Qc7 18 … Qc8 is also seen here. The idea is to play … Ng6 without fear of Qxf5. 19 Bb3 Bxc3! 20 bxc3 Now White’s bad bishop feels as if it just received a sentence of 25 years to life. 20 … f4

Black has full compensation for the pawn for the following reasons: 1. He has the superior minor piece. 2. He has the superior pawn majority. 3. White’s extra pawn is a not so valuable, doubled c-pawn. 21 Ba4

White returns his useless extra pawn to try and activate his bad bishop. After 21 c4 f5 22 Kh1 Ng6 the comp calls it even, while I prefer Black’s chances, S.Grishin-S.Abramov, correspondence 2015. 21 … Qxc3 22 Qg5+ Ng6 23 h4?!

A position can go from clarity to illegibility in the space of a single move. This is a new move, though not a very good one since it is Black, not White, who will be doing the attacking. White’s threat of h4-h5 is easily dealt with. Previous games had seen 23 Rae1 f5 24 Bd7 Qf6, as in M.Zarnescu-A.Tucci, correspondence 2012. Well, here is another massive comp misevaluation: Komodo 13 calls it about even, whereas in my database Black scores 7½/8 from this position. 23 … Kh8 24 Rae1 24 h5? is strongly met by 24 … Qe5!. We never know when our dirty trick is going to come in handy. This move shortcircuits White’s plan to win the f4-pawn. After 25 Qxe5+ (not

25 Qh6? Nh4 26 Bd7 Rg8 27 Bh3 Rg5 and White is crushed) Black replies 25 … Nxe5, when White is left in an awful ending, since Black has both superior minor piece and superior majority. 24 … f6! 25 Qf5 Rg8 0-1

White, who reached the dreaded “How-the-hell-did-I-gethere?” point, is helpless. For example, 26 Bd7 (or 26 Qh5 Ne5 27 Kh1 Rxg2! 28 Kxg2 Rg8+ 29 Kh2 Ng4+ and White is mated) 26 … Nxh4 27 Qh3 Qxh3! 28 Bxh3 f5 29 Kh1 Rg5 30 Rg1 Rag8 31 Rb1 Rh5 32 Rb3 Nf3! 33 Rd1 Ng5 34 Rdb1 f3 wins. Game 28 V.Bucinskas-V.Ivanchuk European Cup, Fügen 2006 1 e4 c5 2 Nf3 e6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 Nc6 6 Ndb5 d6 7 Bf4 e5 8 Bg5 a6 9 Na3 b5 10 Bxf6 gxf6 11 Nd5 Bg7

Our move order is 11 … f5 12 c3 Bg7, which comes to the same thing. 12 c3 f5 13 exf5

This is another crucial line. At first it feels as if White loses a tempo by allowing … Bxf5. White’s idea is that he later regains the tempo with Nc2 and Ne3. 13 … Bxf5 14 Nc2 Be6 The fight for d5 begins. The alternative is simply to castle first; we’ll look at this in Games 30-32. On the one hand, it rules out an early g2-g3 (as in the next game) as that would now be met by … Be4. On the other, after Nce3, Black can’t reply with Ivanchuk’s immediate … Ne7. Instead, 14 … Bg6?! is considered dubious for two reasons: 1. By moving the bishop to g6, Black violates our Magna Carta rule of always fighting for control over d5. 2. The bishop on g6 is vulnerable to h2-h4! pushes from White.

15 Nce3 In the next game we look at 15 g3 (except that it is on the 14th move due to the 1 … e6 and 2 … Nc6 move order number deviation in the current game). 15 … Ne7 Ivanchuk thematically fights for d5. 15 … 0-0 would just transpose to 14 … 0-0 lines.

16 g3 Others: a) 16 Nxe7 Qxe7 17 Bd3 d5 18 0-0 Rd8 19 Qh5 e4 20 Bc2 Qf6 21 Rad1 Qg6 was I.Ivanisevic-B.Damljanovic, Yugoslav Championship, Subotica 2000. The game is balanced after 22 Qh4. b) 16 a4 Nxd5 17 Nxd5 0-0! (Black’s b5-pawn is tactically protected) 18 Be2 Rc8! (the rook heads for c5, challenging Black’s knight) 19 axb5 axb5 20 0-0 (20 Bxb5?? Rc5 21 c4 Bxd5 White loses a piece) 20 … Rc5! was K.Sorri-

M.Manninen, Finnish Championship, Helsinki 1992. Again the game is balanced. c) 16 Be2 Nxd5 17 Nxd5 0-0 18 0-0 Rb8 was seen in Z.Vancsura-M.Krasenkow, Balatonbereny 1988. Black soon generates counterplay with … a6-a5 and … b5-b4. 16 … Nxd5 17 Nxd5 0-0 18 Bg2 18 a4 bxa4 19 Rxa4 Rb8 20 Ra2 a5 21 Bg2 f5 is fine for Black. 18 … a5 Black will generate play with a break on b4. 18 … f5 is an alternative. 19 0-0 Rb8

20 Qd2 Others: a) 20 Qh5 Qd7 21 Qh4 Bxd5 22 Bxd5 Kh8 23 g4 (intending Be4 next) 23 … f5! 24 gxf5 Qxf5 25 Be4 Bf6! (if it were not for

this move, Black would be forced to resign) 26 Bxf5 Bxh4 27 Be4 Rf4 ½-½ N.Short-V.Kramnik, Dortmund 1997. b) 20 a3 (White prevents … b5-b4) 20 … f5 21 Qh5 e4 22 Rad1, B.Blankenberg-H.Eshuis, correspondence 2008. Chances are even after 22 … Be5 23 Kh1 Bxd5 24 Rxd5 Qf6. 20 … Qd7 21 f4 Intending f4-f5 next. If 21 Rad1 then 21 … f5 22 Rfe1 Kh8 23 a3 Qf7 24 Qe3 e4! (the idea is … Be5) 25 Nf4 Be5 26 Nxe6 Qxe6 27 Rd2 b4 28 cxb4 axb4 29 a4 Rbc8 and Black is fine, H.Gromotka-P.Walczak, correspondence 2011. 21 … f5 22 Rac1 The computer doesn’t like this. After 22 a3 the game is equal. However, 22 fxe5?! isn’t a great idea since it opens the game for Black’s bishops. After 22 … dxe5 23 Rad1 Kh8 (evading discovered checks) 24 Qf2 Qf7 25 b3 a4 Black already stands better. 22 … Kh8 23 Rfd1 Rfc8 Komodo suggests the completely inhuman 23 … Qb7!. 24 b4?! White can stay close to even after 24 a3 e4 25 Ne3 Qa7 26 Kh1. 24 … e4! 25 g4!? All-or-nothing. White senses that his position is sinking and, with no clear plan in mind, attempts to confuse the issue. On the plus side for White, after his move the position suddenly begins to lose its strategic markers and reference points, which can be helpful to the inferior side. White must have rejected 25 bxa5 and probably correctly as after 25 … Qa7+ 26 Kh1 Rc5 27 Ne3 d5 the position is in Black’s favour.

25 … axb4 26 Nxb4 26 cxb4 fxg4 27 Rxc8+ Rxc8 28 Bxe4 Rc4 also looks unpleasant for Black. 26 … Bg8! Intending … d6-d5. 27 Kh1 After 27 Qxd6 Qxd6 28 Rxd6 Rxc3 Black’s bishops are monsters and White is unlikely to save the ending. 27 … d5 28 gxf5 Qxf5 29 Qe3 Rc4! Threatening … d5-d4. White is strategically lost. 30 Bh3 30 a3 Ra8 is also awful for White. 30 … Qh5 31 a3?

Exercise (combination inattention is natural,

alert):

In

losing

positions,

because we feel better when we ignore our pain. The menace to White’s position feels simultaneously invisible and omnipresent. White has just blundered in an already strategically lost position. What did he miss? 31 … Rxc3! Answer: Overloaded defender. 32 Rxc3 Qxd1+ 33 Kg2 Rb6! Ivanchuk is after his opponent’s king. I like his move better than the greedy – but also winning – comp suggestion: 33 …

Bxc3 34 Qxc3+ d4 35 Qc7 Qf3+ 36 Kg1 Re8 37 Qd6 Qd1+ 38 Kg2 Kg7 39 Qd7+ Bf7 and White has run out of checks. 34 Bf5

Exercise (combination alert): What does a person of power want most? More of it. Finish White off. 34 … Rf6! Answer: Overloaded defender. This combinational lock is child’s play for a safecracker of Ivanchuk’s ability. If White’s bishop has no urge to be a martyr and moves, then … Rg6+ devastates White. 35 Rc1 White dodges two traps, only to be swallowed up by another. Others:

a) 35 Qh3 fails miserably to the simple 35 … Qe2+ 36 Kh1 Qe1+ 37 Kg2 Rxf5!, leaving the c3-rook loose. b) 35 Bc8 Rg6+ and White loses his queen or his king. 35 … Qf3+! 0-1 One stone, two birds. Black wins a piece, while removing queens from the board. Game 29 J.O.Leconte-F.G.Balabaev Correspondence 2002 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 c3 Bg7 12 exf5 Bxf5 13 Nc2 Be6 14 g3

What is the difference between this game and the previous one? In this version White retains his knight on c2, since it impinges upon Black’s wish to generate queenside play with the plan … Rb8, … a6-a5 and … b5-b4.

14 … 0-0 15 Bg2 a5 Black still desires a break on b4. 16 0-0 f5 Instead, 16 … Rb8 17 Qh5 f5 18 Rad1 Kh8 19 Rd2 Ne7 20 Rfd1 was S.Goerlinger-J.Frijling, correspondence 1996. Play can reach a tense position with balanced chances after 20 … Bf7 21 Qe2 e4 22 Nxe7 Qxe7 23 Rxd6 b4. Here Black has full compensation for the sacrificed pawn. 17 Qe2 Rb8 18 Rfd1 Kh8 19 Rd2 e4!

Black gives up something and gets something in return. He creates a hole on f4, while opening a square on e5. With a subsequent … Be5, Black defends d6, his main strategic worry. 20 Nf4 Bc4 21 Qh5 Threatening 22 Ng6+, forking king and rook. 21 … Ne7 Covering against White’s threat, while reinforcing d5.

22 Nd4 White threatens a fork on e6, while increasing pressure on f5. The alternative is to fight for d5 with 22 Ne3 Bf7 23 Qg5 Bg8 24 Nh5 Be5 25 f4 exf3 26 Bxf3 Ng6 27 Qxd8 Rbxd8 28 Nd5 f4! 29 Ndxf4 b4 ½-½ S.Matyukhin-V.Pankratov, correspondence 2002. Black achieved full compensation for the pawn in this interesting ending, which was cut short with an agreed draw. 22 … Rf6 After 22 … Qe8! 23 Qxe8 Rfxe8 24 Bh3 b4! 25 cxb4 Rxb4 26 Nxf5 Nxf5 27 Bxf5 Bxb2 28 Rad1 Bxa2 29 Rxd6 a4, I like Black’s passed a-pawn more than White’s chances of generating play against Black’s king. 23 Re1 Or 23 b3 Rh6 24 Qg5 Qf8! 25 Nh5 (after 25 bxc4? Bf6 26 Nde6 Bxg5 27 Nxf8 Bxf4 28 gxf4 bxc4! 29 Nd7 Rb7 Black has the better ending after regaining White’s trapped knight) 25 … Bd3 26 Qxg7+ Qxg7 27 Nxg7 Kxg7 28 f3 d5 29 Bf1 Bxf1 30 Rxf1 Rd6, when chances look balanced in the ending. 23 … b4 Black achieves his thematic break and the game looks dynamically equal. 24 cxb4 axb4 After 24 … Rh6 25 Qg5 Qf8 26 Nxf5 Nxf5 27 Rxe4 Bg8 28 bxa5 Rb5 the game is an irrational mess, which the comp unhelpfully assesses at completely even. 25 Qd1?! I don’t see a good reason to retreat a queen which isn’t attacked. Correct was 25 Qg5. 25 … Bh6! Strange noises awaken the sleeping sentry on f4. Black threatens to damage White’s structure and king cover with … Bxf4.

26 Rc2 We can feel the energy ebb from White’s position. Others: a) 26 Nde2? Bxe2 27 Qxe2 Bxf4 28 gxf4 d5 is clearly in Black’s favour. b) 26 Nh5!? (White offers an exchange to escape the bind) 26 … Rg6 27 Bh3 Qa5 28 b3 Bxd2 29 Qxd2 Bf7 30 a4 Qe5 and Black has the better chances. 26 … Bxf4 27 Rxc4 d5

Advantage Black, whose massive centre looms over White’s position. One curious trait of the Sveshnikov is that, right out of the opening, it seems inconceivable that White will ever lose control over d5. And then later on, Black somehow manages to seize control of that very square. 28 Rc2 Bh6 28 … Bc7 intending … f5-f4 is a superior option.

29 Qh5 Bg7 30 Rd1 Ra8 White’s a2-pawn is a cause for concern. 31 b3 Qb6 Black targets the blockading knight on d4. 32 Ne2 32 Kh1 f4 also looks unsaveable for White. 32 … d4! Black finds a clever way to destroy White’s blockade of d4.

The patient and unflashy 32 … Rd8!, is also strong, intending … d5-d4 next, without giving White any options. 33 Nxd4 This move walks into a decisive pin. However, others don’t help:

a) 33 Rc4 is met by 33 … d3 34 Nf4 Raf8 and White is strategically busted. b) 33 Rxd4 Rh6 34 Qg5 Bxd4 35 Nxd4 Qf6 36 Qc1 Rg6 and White doesn’t have enough for the sacrificed exchange, though this was probably his best practical chance. 33 … Rd8! White’s knight is pinned, even though the d1-rook is protected. 34 Rc4 White will never untangle. If 34 Nxf5 Rxf5 35 Rxd8+ Qxd8 36 Qg4 then 36 … e3 and White is busted. 34 … Rfd6 35 Qf7 Qa7!

White’s position is plagued by disease and death. This wins a whole piece and is even stronger than 35 … Rxd4 36 Rdxd4 Bxd4 37 Rxd4 (not 37 Qxe7? Bxf2+ and wins, since 38 Kf1 Rd1+ 39 Ke2 Re1+ 40 Kd2 Qe3+ 41 Kc2 Qc1 is mate) 37 … Qxd4 38 Qxe7 Qd1+ 39 Bf1 Qd6 and Black consolidates.

36 Bf1 Rxd4 37 Rcxd4 Bxd4 38 Kh1 38 Rxd4 Rxd4 39 Qf6+ Kg8 40 Bc4+ Rxc4 is game over. 38 … Nc8?! This slight inaccuracy is perhaps based on the notion that a blackmailer should not be too greedy. If you ask for too much, too often, the person who you are blackmailing may have you killed. This concession isn’t necessary, as there was no hand over the f5-pawn. 38 … Qc5! hangs on to everything. 39 Qxf5 Nd6 40 Qg5 Re8 41 Qd2 Be5 0-1

Two pawns are not enough for being a piece down. Game 30 S.Kindermann-L.Van Wely German Bundesliga 1998 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 c3 Bg7 12 exf5 Bxf5 13 Nc2 0-0

Rather than begin the fight for d5 straight away, Black prefers to castle first. 14 Nce3 Be6 15 Bd3

In this version White deems a fianchetto too slow and his bishop takes direct aim at the black king. 15 … f5 16 Qh5 This is White’s most aggressive option. Later on we look at 16 0-0. 16 … e4 As pretty much always, we are prepared to give White control over f4, if in doing so we are handed control over e5. 17 Bc2 Ne7 Even though the dizzying array of options alter, our plans often remain the same. Black follows the standard plan of direct challenge to d5 whenever possible. 18 Rd1

Others: a) 18 Bb3 is considered in the next game. b) 18 Nf4 (White occupies the f4-hole with tempo) 18 … Bf7 (Black regains the tempo) 19 Bb3 d5 (now White’s queen really is attacked) 20 Qg5 Ng6! (Black is fine in the ending) 21 Qxd8 (after 21 Qxf5?? Nxf4 22 Qxf4 d4! 23 Nf5 Bxb3 24 axb3 Qf6 25 g4 dxc3 26 0-0 cxb2 White is busted) 21 … Raxd8 22 Nexd5 Nxf4 23 Nxf4 b4! (Black’s favourite counterplay mechanism again) 24 Ne6 Bxe6 25 Bxe6+ ½-½. Draw agreed as White has nothing, V.Anand-V.Kramnik, Frankfurt 2000. 18 … Rc8 There are two ideas behind this move: 1. Black may later play … Rc5, which not only pressures the d5-knight, it also indirectly protects Black’s slightly vulnerable f5-pawn. 2. In some cases Black may engineer a pawn break on b4, after which White’s c2-bishop is a problem if it’s loose. This means that White is prohibited from played c3xb4. The alternative 18 … b4 is the main line. Play can continue 19 cxb4 (after 19 Nxb4 a5 20 Nbd5 Nxd5 21 Nxd5 Rb8 22 Bb3 a4 23 Bxa4 Rxb2 Black gets more than enough compensation for the pawn) 19 … Rc8 (19 … Nxd5 20 Nxd5 Bxb2 21 0-0 Bxd5 22 Rxd5 Be5 23 f4 exf3 24 Rxf3 Qb6+ 25 Kh1 Qxb4 26 Rd1 Qg4 27 Rxf5 Qxh5 28 Rxh5 with an even ending, M.Erdogdu-M.Zaslavsky, Rijeka 2010) 20 Bb3 Ng6 21 0-0, C.Zeng-M.Tabatabaei, Tianjin 2018. Black looks okay after 21 … Kh8 with the dual threats of … Bxb2 and … f5-f4. 19 Bb3 White reinforces d5. 19 … Ng6 Black seizes control over f4 and gets White nervous about the push … f5-f4. 20 g3

This move hasn’t been repeated, although it’s Komodo 13’s top choice. Instead, 20 0-0 Kh8 21 f4 a5 22 a3 was H.NamysloT.Runting, correspondence 2001. Here Black can play the forcing 22 … b4!? 23 axb4 axb4 24 cxb4 Bxd5 25 Nxd5 Ne7 with only an edge for White. 20 … Be5 21 f4 More accurate is 21 0-0!, but even there Black holds his own with careful play; e.g. 21 … Kh8 22 f4 exf3 23 Bc2! b4! 24 Nxf5 Bxd5 25 Rxd5 (threatening 26 Nh4!) 25 … Rf6! 26 Nh4 Kg7 27 Nxf3 bxc3 28 bxc3 Rxc3 29 Be4 Rc4 30 Ng5 Rxf1+ 31 Kxf1 Qf6+ 32 Kg2 h6 33 Qe2 Qxg5 34 Qxc4 Nf4+ 35 Kf1 Nxd5 with a likely draw. 21 … exf3 22 0-0 f4! As I have often told students: if you like the warmth of a soft, cosy, protected environment, then I’m afraid the Sveshnikov is not the right opening for you.

Black isn’t going to sit around and allow White to pile up pressure on the f5-pawn. So Van Wely dissolves it, while exposing White’s king a little. Black stands at least equal. 23 gxf4 Nxf4 24 Nxf4 Bxb3 25 axb3 Rxf4 26 Rxf3 Rh4!? This tempting move may not be best. 26 … Rxf3 27 Qxf3 Kh8 is equal. 27 Qf7+ After 27 Qf5! Black stands slightly worse since he must avoid 27 … Bxh2+?? (or 27 … Rxh2?? 28 Ng4) 28 Kf1! Be5 29 Ng4 Bg7 30 Rxd6! and Black is busted. 27 … Kh8 28 h3 b4 29 c4!? 29 Ng4! bxc3 30 bxc3! Bxc3 31 Qe6 Bg7 32 Rxd6 Qc7 33 Rd7 Qc5+ 34 Kg2 Rg8 should be okay for Black. 29 … Qg5+

Exercise (critical decision): White can play his king to f2, f1 or h1. He can also block the check with 30 Ng4. Two of the options lose, while two hold the game. Make certain you are in the favourable bracket: 30 Kh1?? Answer: White overestimates his own attacking prospects, while grossly underestimating Black’s. Moving the king to h1 or blocking with the knight both lose, while White holds on by moving his king to either f2 or f1. For example: a) 30 Kf2 Bd4 31 Qe6 Rg8 32 Ke2 Rh6 33 Qd5 Bxe3 34 Rxe3 and White isn’t worse. b) 30 Kf1 Rg8 31 Ng4 Bxb2 32 Qe6 a5 33 Rxd6 and White is fine. c) 30 Ng4?? is undercut by the undermining shot 30 … Rxh3!. 30 … Rg8 Just like that, Black has a wicked attack. 31 Qf5 Qg7 32 c5

Exercise (combination alert): Find one powerful shot and White must immediately resign. 32 … dxc5 Why be satisfied with pretty good, when great is available? Answer: The text still wins but is far less potent than 32 … Bd4! (overloaded defender; threatening … Bxe3, destroying the defender of the g2 mating threat) 33 Ng4 dxc5 and there is no defence to the threat of … h7-h5!. 33 Qd7 Qg6 34 Qf5 Bxb2 35 Qxg6 White’s only ambition for now is to outlast Black’s attack. The trouble is that Black is winning the ending. 35 … Rxg6 36 Nf5 Re4 37 Rd8+ Rg8 38 Rxg8+ Kxg8 39 Nd6 Re2 40 Nc4

This blockade won’t last forever. 40 … Bd4 41 Rf5 Ra2! 0-1 Black will play … a6-a5-a4, creating two connected passed pawns and ending White’s resistance. Game 31 K.MacKinnon-A.Shabalov Calgary 2007 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 Bg7 11 c3 f5 12 exf5 Bxf5 13 Nc2 0-0 14 Nce3 Be6 15 Bd3 f5 16 Qh5 e4 17 Bc2 Ne7 18 Bb3

The bishop wasn’t doing anything on c2, so White re-routes it to fight for the opening’s holy grail square which, naturally, is d5. 18 … Ng6

Seizing control over f4. After 18 … Nxd5 19 Nxd5 a5 20 Nf4 Bxb3 21 axb3 Qe7 22 0-0 b4 23 cxb4 (Black is also okay after 23 Nd5 Qf7 24 Qd1) 23 … Qe5 24 Qg5 axb4 Black stood no worse, M.Adams-A.Beliavsky, European Team Championship, Debrecen 1992. 19 f4?! White possibly underestimates the danger to his own king. Instead, 9 g4 fxg4 20 h3 g3! (after 20 … gxh3? 21 Rxh3! Rf7 22 Qxh7+ Kf8 23 0-0-0 Bxh3 24 Qxg6 White has a wicked attack for the exchange) 21 fxg3 Ne5 22 Nf4 Bxb3 23 axb3 Qd7 chances were even, B.Rimkus-V.Lyukmanov, correspondence 2006. 19 … exf3 20 Qxf3? Faith divides them. Both sides believe in their own positions. The trouble is that they can’t both be right. White stands only slightly worse after 20 0-0! fxg2 21 Rxf5! Ra7, but not 20 gxf3? f4 21 Ng4 Rf5! and, just like that, White is busted, M.Bruere-R.Jamieson, correspondence 1985. 20 … f4

A good rule of thumb for Sveshnikov players is to find the most rage-inciting move on the board and then play it. It will usually be the correct one. Black’s initiative is out of control. 21 Nc2 Or 21 Ng4 Kh8 (threatening 22 … Qg5) 22 h4 Qd7 23 Nb6 Qe7 24 Bxe6 Qxe6+ 25 Qe2 Rae8 26 Qxe6 Rxe6+ 27 Kf2 f3! 28 gxf3 Nf4 29 Rad1 Re2+ 30 Kf1 Rxb2 and White is busted. 21 … a5 Threatening … a5-a4. 22 a3 Kh8 This is a good precaution, preventing tricks on the a2-a8 diagonal. 23 Nd4? The best course was a strategic exchange sacrifice for the light squares by playing 23 0-0-0 Qg5 24 Kb1 Bg4 25 Qf2 Bxd1 26 Rxd1 with maybe some fishing chances for White.

23 … Bxd4! 24 cxd4 Qh4+ More accurate is 24 … Qg5! 25 0-0 Nh4 26 Qf2 Rg8 27 Nxf4 Bxb3, winning. 25 Kd2 After 25 g3 fxg3 26 hxg3 Rxf3 27 gxh4 Rxb3 28 Nc7 Rf8 29 Nxe6 Re3+ 30 Kd2 Rxe6 Black is a piece up. 25 … Bg4 26 Qf1 f3

The pawn cover around White’s king is ripped open. 27 g3 Qg5+ 28 Ne3 White is forced to self-pin. If 28 Kc2 then 28 … Bf5+ 29 Kd1 f2 30 Nc3 Qe3 31 Qe2 Bg4! (removal of the guard/pawn promotion) 32 Qxg4 f1Q+ and the game is over. 28 … Rae8 28 … Ne7! 29 Qd3 Nf5 30 Rae1 f2 wins. 29 Qf2 Re4

Hitting d4, while preparing to pile on to the e3 pin by transferring the f8-rook to the e-file. 30 Rad1 Rc8!

Principle: when hunting the opponent’s king, don’t chase him – cut off escape routes. The move is even stronger than chopping on d4. 31 h3 a4?! Black is still winning after this, but 31 … Rxd4+! 32 Ke1 Re4 33 Kd2 Qe5 is more to the point. Black threatens 34 … Qd4+ and also 34 … Qxb2+. 32 hxg4 axb3 Threatening 33 … Rc2+. However, 32 … Rxd4+! was still stronger: 33 Ke1 Re4 34 Kf1 Qxe3 35 Qxe3 Rxe3 36 Bf7 Ne5 37 Be6 Rc2 38 Bf5 Rce2 39 Rxh7+ Kg8 40 Rb7 Nc4 41 Bd3 Nd2+, winning. 33 Qxf3!

Exercise (critical decision): Black’s candidate moves are 33 … Rc2+ and 33 … Rxd4+. The assessment of one is -2.69 in Black’s favour, while the other assessment is the dreaded 0.00. Which move would you play? 33 … Rxd4+? The diminution of Black’s power increases with each passing move. White should be just fine after this mistake. Answer: 33 … Rc2+ is decisive after 34 Kd3 Rxe3+! 35 Qxe3 Qxg4 36 Qe8+ Kg7 37 Rde1 Qf3+ 38 Qe3 Ne5+! (clearance; this is a difficult move to see beforehand) 39 dxe5 Qd5+ 40 Qd4 Rd2+ (deflection/removal of the guard) 41 Kxd2 Qxd4+ 42 Kc1 Qf2 (threatening mate on c2) 43 Kb1 b4! 44 a4 dxe5 45 Rxe5 Qc2+ 46 Ka1 Qc6 47 Ree1 Qxa4+ 48 Kb1 Qa2+

49 Kc1 Qa1+ 50 Kd2 Qxb2+ and Black should win the technical ending. 34 Ke2 Rxd1 35 Qb7?! This is not the time to strive for utopian idealism. White should survive with the simple recapture 35 Rxd1.

This appears to be a powerful shot. It feels as if the queen is the lone, mysterious stranger-silhouetted-against-the-sunset character from a Sergio Leone spaghetti Western, who appears in town and singlehandedly massacres a gang of twelve bad hombres, thus freeing the good townsfolk of fear. In reality, the queen is disengaged from the proceedings on b7. Exercise (combination alert): With his last move, White threatens mate on h7, Black’s loose c8-rook, with check and Black’s hanging d1rook. Undaunted,

Shabalov finds a truly startling move. What did he play here? 35 … Re1+!! Shift the trajectory by a fraction of a degree and, in turn, this shifts the destination 180 degrees. Answer: Deflection. White’s rook is induced off the h-file. Not 35 … Rxh1? 36 Qxc8+ Nf8 (or 36 … Kg7 37 Nf5+ Kf7 38 Nxd6+ Kg7 39 Nf5+) 37 Qxf8+ Qg8 38 Qf6+ Qg7 39 Qd8+ and White escapes with perpetual check. 36 Rxe1 Re8 37 Kf2 Qf6+ 38 Kg1? Sometimes the exertion of staying alive, diverts us from opportunity. After 38 Nf5! Qxb2+ 39 Kf1 Rf8 White only stands a touch worse and may well save the game.

Exercise (combination alert): White’s last move loses. Prove how. 38 … Qd4! Answer: Pinned piece. 39 Kf2 After 39 Qf3 Qxb2 40 g5 Rf8 41 Qe2 Qe5 Black’s b-pawn is decisive. 39 … Ne5! 0-1 Threatening 40 … Nxg4+. This move also cuts off any Qf3 defence. After 40 Kg1 Qxb2 41 Qe4 Qc3 42 Rd1 b2 43 Kg2 Qc1 44 Kf2, Black has the beautiful shot 44 … Qxe3+!! 45 Kxe3 Nc4+ 46 Kf3 Rxe4 47 Kxe4 Nd2+! and wins. Game 32 N.De Firmian-A.Shabalov Virginia Beach 2007 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 c3 Bg7 12 exf5 Bxf5 13 Nc2 0-0 14 Nce3 Be6 15 Bd3 f5 16 0-0

White reasons that a queen on h5 is a vulnerability as well as an attacking unit. So he refrains from Qh5 and simply castles, leaving all options open. 16 … e4 This is the most committal of Black’s options as it weakens d4 and f4. In return Black gets the use of e5 for the bishop. Alternatively, 16 … Ra7 and 16 … Kh8 are also popular. 17 Nf4 Bf7 18 Bc2 Be5 19 Nh3 Or 19 Nfd5 Kh8 20 f4 exf3 21 Qxf3 Bxd5 22 Nxd5 Qh4?! (22 … Ne7 looks better) 23 g3 Rg8 24 Nf4 Qh6 25 Bxf5?! (White has the advantage after 25 Bb3!) 25 … Raf8 26 Ne6?? (better is 26 Ng2! Ne7 27 Qd3 Rxf5 28 Rxf5 Nxf5 29 Qxf5 b4, although Black has enough play for the pawn) 26 … Bxg3! (annihilation of defensive barrier/removal of the guard) 27 hxg3 Rxf5! 28 Qxf5 Rxg3+ 0-1 (since 29 Kf2 Qe3 mate is mate), V.Topalov-M.Carlsen, Abidjan (blitz) 2019. 19 … Kh8 Shabalov prepares to attack down the g-file.

19 … f4? is a mistake. After 20 Nd5 Bg6 21 Nhxf4! Bxf4 22 Nxf4 Rxf4 23 Qd5+ Bf7 24 Qxc6 d5 25 Rad1 Black is a pawn down and busted. 20 f4 Bf6!? This pawn sacrifice is adventurous and perhaps unnecessary. After 20 … Qb6! 21 fxe5 Qxe3+ 22 Kh1 Bc4 23 Re1 Qh6 24 Qxd6 Qxd6 25 exd6 Rad8 26 b3 Bg8 27 Rad1 Rf6 Black regains the pawn and the game is even.

21 Nxf5 d5 22 Nh6 Be8 23 Ng5!? As most of us understand from high school and college: our love isn’t always returned. As it turns out, White’s hoped-for attack is an object unworthy of his love. Safer and stronger is 23 Kh1 Ne7. 23 … Ne7 24 Qe1!? De Firmian dreams of an attack against Black’s welldefended king. Others:

a) 24 Qd2 is met by 24 … b4! with plenty of compensation for the pawn. b) 24 Ne6?? is a fake fork. After 24 … Qb6+ 25 Nd4 Bg7 26 Ng4 Bh5 27 h3 Nf5 28 Qd2 Nxd4 29 cxd4 Qxd4+ 30 Qxd4 Bxd4+ 31 Kh2 Bxb2 White is a pawn down, facing two connected central pawns and Black’s bishop pair. He can resign. 24 … Qb6+ 25 Kh1 Bg7 26 Ng4 h6 27 Nxe4?!

Both sides are trying hard to out-crazy each other. White’s thoughts flow from a desire not to lose, to a desire to win. De Firmian decides to sacrifice a piece. He is already a pawn up so he gets three pawns for it in total. As it turns it, this isn’t enough. And White doesn’t get the attack he hoped for either. De Firmian probably sensed the initiative slipping away if he went passive with 27 Nh3. A common fallacy in chess is that courage is always a virtue. It isn’t. 27 Nh3, which goes passive, isn’t all that bad for White. After 27 … b4! 28 c4! Rc8 29 Ne5 Ng6 30 Qg3 the comp assesses it as dead even.

27 … dxe4 28 Bxe4 Rc8 A reasonable alternative is to place the rook on the open dfile. 29 f5 Ng8! White’s attack looks stalled after this powerful Steinitzian retreat, which disallows f5-f6 ideas, while reinforcing h6. 30 Qh4 Nf6! Clever. Shabalov realizes that his h6-pawn isn’t really hanging. 31 Nxf6 After 31 Nxh6 Nxe4 White lacks a useful discovery with his knight. If 32 Ng4+ Kg8 33 Rae1 Qd8! 34 Qxd8 Rxd8 35 Rxe4, then 35 … Rxf5! exploits White’s loose back rank. White is busted. 31 … Qxf6 32 Qg3 Bf7

White’s three pawns are not enough for the piece, since they completely lack fluidity. 33 a3 The trouble with a move like this is that White essentially freezes his queenside pawn majority. 33 … Rc4 Here 33 … Rcd8 is slightly more accurate, intending to swoop down to the seventh rank, or if White challenges the file with 34 Rad1 then Black seizes control over it again with 34 … Bb3. 34 Rae1 Qg5 34 … Rd8 is also strong. 35 Qf3 If 35 Qxg5 hxg5 36 f6 Bh6 37 h3, Shabalov foresaw that the f6-pawn isn’t going anywhere and is eventually doomed. 35 … Bf6 36 Bd5 Bxd5 37 Qxd5 Rh4!? Shabalov is after the white king. 37 … Rf4! is more accurate. 38 Qf3! Rg8 39 g3?! He had to try 39 Re2 Bd8! 40 h3 Bc7 41 Qe3 Rd8 42 Qxg5 hxg5. Even here White is busted.

Exercise (combination alert): Why confess to a sin to which you are not guilty of committing? On the last move of the time control, White blundered. What did he miss? 39 … Rg4? Time pressure turns even strong GMs into a temporary state of patzerhood. Answer: Both GMs missed the simple 39 … Qxg3!. Pinned piece, with a technical win for Black in the ending. 40 Re6 h5! 41 Rxa6? White missed his chance for counterplay with 41 Rd1! h4 42 Rdd6! Qc1+ 43 Qd1 Qxd1+ 44 Rxd1 Rf8 45 Kg2, when he may save the game.

41 … h4 42 Ra8 Not 42 gxh4?? Rg1+ and mate next move. 42 … hxg3 43 Rg1?? After 43 Rxg8+ Qxg8 44 hxg3 Rxg3 45 Qh5+ Kg7 46 Rg1 White can fight on.

Exercise (combination alert): White hallucinated on his last move. Black to play and win. 43 … g2+! Answer: The g2-square is the epicentre of the coming explosion. Simply push the g-pawn. 44 Qxg2 Or 44 Rxg2 Qc1+ and mate in two moves.

44 … Rxg2 45 Rxg8+ Qxg8! 0-1 I wasn’t there, but I’m almost certain Nick must have raised his eyes towards the ceiling and implored it for an answer to an unspoken question: “How the hell did I miss such an obvious move?” Black doesn’t lose his queen with this recapture. I have a feeling Nick expected 45 … Kxg8?? 46 Rxg2 Kf7 47 Rxg5 Bxg5 48 b3!, when c3-c4 is coming, Black’s final pawn is removed from the board, and the game is drawn. Game 33 T.Kantans-M.Matlakov Gibraltar 2017 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e6 6 Ndb5 d6 7 Bf4 e5 8 Bg5 a6 9 Na3 b5 10 Bxf6 gxf6 11 Nd5 f5 12 g3

White’s e-pawn isn’t really hanging, so he fianchettoes his bishop to fight for control over d5. But this is less accurate than the sequence 12 c3 Bg7 13 exf5 Bxf5 14 Nc2 Be6 15 g3 of Game 29, because of Black’s reply.

12 … fxe4 As just mentioned, Black doesn’t get to hang on to this pawn. The capture is simply to destroy White’s pawn centre. 13 Bg2 Bg7 Others: a) 13 … Bf5 is a waste of time, since White wins the pawn back anyway. The light-squared bishop belongs on e6. After 14 0-0 Bg7 15 Re1 0-0 16 c3 Ne7 17 Nc2 Rc8 18 Nxe7+ Qxe7 19 a4 Qb7 20 axb5 Qxb5 21 Nb4 a5 22 Nd5 (threatening a fork on e7) 22 … Rfe8 23 Bxe4 Be6 24 Qd2 White’s grip over d5 gives him an edge, M.Adams-L.Van Wely, Tilburg 1998. b) 13 … f5?? is a catastrophically stupid act; 14 Qh5+ Kd7 15 Qxf5+ Ke8 16 Qh5+ Kd7 17 Bh3 is mate. 14 Bxe4 Black will gain a future tempo on this bishop with an … f7f5 push. 14 … Be6! Black should not be afraid of sacrificing the exchange for White’s light-squared bishop. 15 0-0 Others: a) 15 Qh5 prevents castling. Then 15 … Rc8 16 0-0 Ne7 17 Nxe7 Qxe7 18 Rad1 b4 19 Nb1 was J.Tomczak-M.Krasenkow, Trzcianka 2018. I slightly prefer Black’s position after 19 … Rc5!, intending … d6-d5 next. b) 15 Nf6+?! (Black should be delighted if White takes this greedy route) 15 … Bxf6 16 Bxc6+ Ke7 17 Bxa8 Qxa8 18 0-0 h5! (even stronger than winning the exchange back with 18 … Bh3) 19 Qd3 h4 20 Rad1 Qc6 and Black has a raging attack for the sacrificed exchange. I would not want to take on the white pieces here. 15 … 0-0 16 c3 White prepares to bring the a3-knight back into play. 16 … f5 17 Bg2 Rb8

As always against any fianchetto variation from White, Black’s counterplay relies on a break on b4. 18 Nc2 a5 We have now reached virtually the same position as in Game 29, except that here the sequence Bg2, Bxe4 and Bg2 has cost White a tempo, so Black has the very useful … Rb8 as a completely extra move. 19 Qe2 Alternatives: a) 19 Qd2 Qd7 20 f4 b4! 21 cxb4?! (21 Rad1 e4 is about even) 21 … e4 22 Kh1 was A.Neverov-O.Gladyszev, Russian Team Championship 2008. After 22 … Nxb4 23 Ncxb4 axb4, Black gets great pressure for the pawn, and if 24 Nxb4? d5 White is strategically busted. b) 19 a3 was seen in F.Caruana-V.Ivanchuk, Reggio Emilia 2011/12. White squelches a … b5-b4 break. In doing so he must keep his c2-knight where it is, on eternal babysitting duty. The game continued 19 … Kh8 (this way he doesn’t need to calculate Nf6+ and Bxc6 ramifications on every move) 20 Qe2 (or 20 Qd2 e4 21 Rad1 Ne5 22 Nce3 is even, whereas 22 f4? Nd3 23 Nce3 Qc8!, threatening … Qc5, going after White’s rootless d5-knight, saw White already in deep trouble, A.Raetsky-P.Kukula, Wroclaw 2014) 20 … Qd7 21 Rad1 e4 (as always in fianchetto lines, Black is happy to hand White control over f4, if he gets e5 in return) 22 f3 (chipping away at Black’s centre) 22 … Nd4! (interference; this startling move is easy to overlook) 23 Nxd4 Bxd5 (threatening 24 … Bc4) 24 Qd2 b4. Ivanchuk has achieved the thematic break on b4 and I already prefer Black due to his dark-square power. This game is annotated in Sveshnikov: Move by Move. 19 … Kh8

Black proceeds in the style of Game 29, first safeguarding his king before advancing the e-pawn. An earlier game, Sh.Matthews-Wa.Elliott, Jamaican Championship, Kingston 2010, saw the immediate 19 … e4 20 Nce3 Ne7 21 Rad1, when 21 … Nxd5 22 Nxd5 b4 offers equal chances. 20 Rfd1 Qd7 21 Rd2 White plans to double rooks and pressure d6. 21 … e4 This is played with two ideas in mind: 1. A future … Be5 will defend the d6-pawn. 2. To harass White’s d2-rook with … Ne5 and … Nc4. Also possible is 21 … Bh6 22 f4 Bxd5 23 Rxd5 exf4 24 Rad1 Rbe8 25 Qxb5 Qa7+ 26 Kh1 Ne5 27 Rf1 Ng4! 28 Nd4 Ne3 29 Re1 Qf7 30 Rxd6 Bg7, when Black stands better since White’s exposed king is of greater importance than his extra pawns. 22 Nce3 Ne5

In the future the knight may sink into c4, or d3, or even f3. 23 Nf4 Slightly inaccurate. White remains equal with 23 a3. 23 … Bh6! 24 Nxe6 24 Ned5? doesn’t help, as after 24 … Bxd5 25 Rxd5 Bxf4 26 gxf4 Ng6 White’s kingside has been shattered and he is in deep trouble. 24 … Qxe6 25 Rd4 b4

26 cxb4?! White may have been better off heading for an inferior opposite-coloured bishop ending with 26 Nc4 bxc3 27 bxc3 Nxc4 28 Qxc4 Qxc4 29 Rxc4. 26 … Rxb4 27 Rxb4 axb4 28 a4 After 28 Qb5 Bxe3 29 fxe3 Qh6 30 Qxb4 Qxe3+ 31 Kh1 Nd3 32 Qc3+ Kg8 33 Bh3 Qf3+ 34 Bg2 Qf2, White is in trouble

since Black’s e- and f-pawns will surge, either exposing White’s king or threatening to promote. 28 … bxa3 It’s tempting to eliminate the passed a-pawn, but this actually helps White, since it activates his rook. Instead, 28 … Nd3!? leads to great complications; e.g. 29 a5 f4 30 Nc2 d5 31 a6 Ra8 32 Nd4 Qe5 33 Nb3 Qf6 (threatening the a6-pawn, as well as … f4-f3) 34 Bxe4! dxe4 35 Qxe4 Rd8 36 Qf3! (White must navigate the position the way a lumberjack walks over floating logs, as his path is wobbly and fraught with danger; not 36 a7?? fxg3! 37 fxg3 Qf2+ 38 Kh1 Qxa7! 39 Rxa7 Nf2+ 40 Kg2 Nxe4 41 Rb7 Bf8 and Black consolidates) 36 … Ne5 37 Qe4 Rf8 38 g4 Nxg4 39 a7 Qxb2 40 Qf3 Ne5 41 Qd5 Ng4 42 Qg2 f3! 43 Qxg4 Ra8 44 Qxf3 Qg7+ 45 Qg2 Rxa7 46 Rxa7 Qxa7 47 Qd5 and the game will likely end in a draw. 29 Rxa3 Nd3 30 Ra5 f4

Exercise (critical decision): White has only one move to remain even. What should he play? 31 gxf4? White’s king’s cover is torn to shreds after this obliging move. Answer: White saves himself if he finds 31 Nf5! Rxf5 32 Rxf5 Qxf5 33 Bxe4 Nc1 34 Qc4 Qe5 35 Kg2 Nb3 36 Qc8+ Kg7 37 Qb7+ Kf6 38 Qxh7 Qg5 39 Qh8+ Qg7 40 Qd8+ Qe7 41 Qh8+ with perpetual check. 31 … Nxf4 32 Qg4 Qe8 33 Qd1?! Nd3?! Black misses 33 … d5!, when 34 Nxd5 is met by 34 … Qe5! and there is no good answer to the coming … Rd8. 34 Qh5 Or 34 b4, and if 34 … Rxf2 35 Nf5! Qg8 36 Qa1+ Rb2 37 Ra8!, Black has to deal with two annoying pins and 37 … Bf8 38 Qf1 Rf2 39 Qa1+ Rb2 leads to a draw by repetition. 34 … Bxe3 Or 34 … Qe6 35 Qd5 Qc8 36 Qxe4 Nxf2 37 Qd4+ Bg7 38 Qd2 Qe6 39 Ra8 and White should hold the draw. 35 fxe3 Qc6 36 Ra1?! White shouldn’t allow Black to play … d6-d5. He can hold after 36 h3 Qc1+ 37 Kh2 Qxe3 38 Rf5 Rg8 39 Qh4 (threatening 40 Qf6+) 39 … Qd4 40 Rf7 Rg7 41 Rf8+ Rg8 42 Rf7 with a draw by repetition. 36 … Qc2 The simple 36 … d5 is also strong. 37 Qh6?? The attempted salvation becomes the instrument of his future damnation. The question marks become plural. “There are two ways to be fooled. One is to believe what isn’t true. The other is to refuse to accept what is true,” wrote Søren

Kierkegaard. This move, a case of the former on the list, is undoubtedly a time-pressure blunder. White cannot allow Black’s queen entry to f2. He had to try playing a pawn down with 37 Qh4 Qxb2 38 Rf1 Rxf1+ 39 Bxf1 Qe5. 37 … Qf2+ White is helpless to halt the coming storm. 38 Kh1 Rg8 39 Qh3

Exercise (combination alert): This is the final chapter of the murder mystery novel, where the detective gathers the suspects into a single room to reveal the killer’s identity. In this case there are three killers. Black has multiple ways to win. Find one of them.

39 … Ne1! Answer #1: Playing the knight to e1 wins a clear piece. There were two other winning options. Answer #2: 39 … Qxb2! (with dual threats of 40 … Qxa1+ and 40 … Nf2+, forking) 40 Rf1 Nf2+ 41 Rxf2 Qxf2 and there is no reasonable remedy to the coming queen check on e1. Answer #3: 39 … Qf3! (this startling move also works; e.g. 40 Qxf3 (or 40 Bxf3 Nf2 mate) 40 … exf3 and White is unable to move his bishop due to the knight mate on f2. 40 Ra7 Well, this mate threat on h7 doesn’t cut it, since Black can simply remove queens next move while winning a piece. 40 … Qxg2+ 41 Qxg2 Nxg2 0-1 Black threatens the e3-pawn as well as … Nh4-f3 and … Rg1 mate, which eternally ties White’s rook to the defence. Game 34 A.Ruiz Symmes-A.Borowiec Correspondence 2003 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Bxf6 gxf6 9 Na3 b5 10 Nd5 f5 11 c4

This position is similar to, but not the same as, the ones involving c3-c4 we examined in the previous chapter. In this version two factors are altered: 1. Black’s f-pawns are doubled. 2. White’s e-pawn is under attack. Black actually proceeds exactly as we did last chapter by pushing the b-pawn with … b5-b4 and not fearing Qa4. 11 … b4 12 Qa4!? This adventurous move is inherently in Black’s favour, although care is needed, since great complications arise. Instead, after 12 Nc2 fxe4 13 g3 Bg7 14 Bg2 0-0 15 Bxe4 Rb8 16 b3 f5 17 Bg2 e4 18 Rb1 Qa5 19 0-0 Qxa2 20 Nde3 Qa5 21 Qxd6 Rf6 22 Qf4 Qe5 23 Qxe5 Nxe5 24 Nd5 Rf7 25 Ncxb4 Black got full compensation for the sacrificed pawn, H.Nakamura-B.Gelfand, Moscow 2013. This game is annotated in Sveshnikov: Move by Move. 12 … Bd7

Let’s try not to fall for the boneheaded trap 12 … bxa3?? 13 Qxc6+ Bd7 14 Qxa8!, when Black can resign. 13 Nb5! The only move, as can be seen: a) 13 Nxb4?? Nd4 14 Qd1 Rb8 15 Nac2 Ba4! 16 b3 Qa5 17 Nxd4 Qxb4+ 18 Qd2 Qxd2+ 19 Kxd2 exd4 20 bxa4 fxe4 and White is completely busted in the ending. b) 13 Nb1? Nd4 14 Qd1 fxe4 15 Nd2 Bg7 16 Nxe4 0-0 17 Be2 a5 18 Nd2 f5, when White lost tempi regaining the e4-pawn and has already landed in a rotten position. 13 … axb5! Black accepts the challenge. 14 Qxa8 Qxa8 15 Nc7+ Kd8 16 Nxa8

Black has entered an ending a full exchange down but has plenty of compensation; in fact I slightly prefer Black’s chances. In the Sveshnikov Black will constantly be either down material

or in possession of a wretched structure. In both cases ferocious piece activity usually compensates, as it does here. 16 … Nd4 17 Rd1?! Here 17 Kd1?! is also poor; e.g. 17 … fxe4 (threatening … Bg4+ and … Bh6+, followed by … Nc2+, picking off the a1rook) 18 h3! (V.Cobic-D.Stankovic, Bela Crkva 1987). Black is close to winning after 18 … Be6 19 Nb6 Bh6! 20 a4 bxa3 21 Rxa3 Kc7 22 Nd5+ Kb7 23 Ne3 Rc8, as White is fighting for his life, despite the extra exchange. However, White has an improvement in 17 Bd3. Everyone seems to (falsely) believe this move is unplayable. It isn’t and White may actually save the game with it. I hate to offer suggestions to help our opponents, yet our goal in chess is to follow knowledge, wherever it takes us. For example: 17 Bd3! fxe4 18 Bxe4 bxc4 19 Nb6 f5 20 Bd5 (not 20 Bb1?? Bc6 21 0-0 Rg8 22 f3 Nxf3+ 23 Kf2 Nxh2 24 Rg1 d5, when … Bc5+ is coming and White can resign) 20 … c3 21 bxc3 bxc3 22 Nxd7 Kxd7 23 Rc1 c2 24 Bb3 Bh6 25 Rxc2 Nxc2+ 26 Bxc2 Ke6. Black stands a touch better but the ending should be drawn with correct play. 17 … fxe4 18 cxb5 d5 19 Nb6 Be6 20 h3? This is an act of omission White can ill afford. He had to try 20 Na4 Rg8 21 h3 Bh6 22 Be2 Rxg2, when White is struggling but has more chances than in the game.

In such dire positions, even a single inaccuracy may be the last. Black is still down a full exchange, yet he stands clearly better for many reasons: 1. A monster knight on d4. 2. The bishop pair in an open position. 3. Dark-square domination. 4. A massive pawn centre which will later roll forward. 5. White’s knight is out of play. 6. White’s kingside pieces are in desperate need of employment. 7. Believe it or not, White’s king is unsafe, as we discover in the coming combination alert. Exercise (combination alert): White’s misery index is way too high. How did

Black convert his overwhelming strategic advantages into a combination? 20 … b3! Answer: Step 1: Clearance. Black threatens a devastating bishop check on b4. 21 a3 Nc2+ Black finally takes the fight to White’s shores. 22 Ke2 22 Kd2? merely hands Black a free tempo. After 22 … Bh6+ White’s king must head for e2 all the same, since 23 Kc3?? allows 23 … d4+ 24 Rxd4+ exd4 mate. 22 … d4 Step 2: Clearance once again. Black opens the c4-square and creates the threat of … Bc5, followed by … Bc4+. The second point is that by pushing the pawn to d4, Black traps the white knight, as it can no longer escape via the route a4 and c3. 23 f3 Or 23 Rd2 Kc7 24 Na4 Bd6 25 g4 Ra8 26 b6+ Kb8 and White’s knight is trapped. 23 … Bc5 24 Na4 d3+ 0-1

If 25 Kd2 Be3+ 26 Kc3 Bd4+ 27 Kd2 e3+! 28 Kxd3 Ke7 29 Nc3 Rd8, Black threatens … Bxc3+ and … Rxd1, while 30 Be2 (30 Rc1 Bb6+ also forces mate) 30 … Bb6+ 31 Ke4 f5+ 32 Kxe5 Bc7+ 33 Rd6 Bxd6 is mate. Game 35 I.Efimov-P.Eljanov European Championship, Ohrid 2001 1 e4 c5 2 Nf3 Nc6 3 Nc3 Nf6 4 d4 cxd4 5 Nxd4 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Nxb5!?

White’s army marches into battle, pennants aloft and fluttering in the wind. The Great Romantics would approve of White’s piece sacrifice; the comps don’t think much of it. White is lucky to remain even after this exuberance. 11 … axb5 12 Bxb5 Bb7! It seems odd to not break a pin, yet the bishop is better placed on b7, where it takes aim at d5, and indirectly at e4 and g2. For the nervous among us 12 … Bd7 is also playable, when 13 exf5 Rb8 14 a4 leads to equal chances. White scores a lot better from this position in my database than from the 12 … Bb7! lines. 13 exf5 Ra5 Others: a) 13 … Bg7 14 0-0 (14 f6! is a stronger move) 14 … 0-0 15 Bc4 Qg5 16 a4 Kh8 Black had the better chances, S.SulskisL.Van Wely, Moscow 2004. This game is covered in Sveshnikov: Move by Move.

b) 13 … Bh6 is untried and may be an improvement. I can’t find equality for White after 14 f6 Kf8!; e.g. 15 Qh5 Bf4 16 Nxf4 Qa5+ 17 c3 Qxb5 and White is in some danger. 14 Qd3 14 a4 is met by 14 … Rxb5! 15 axb5 Nd4 16 c4 Bxd5 17 cxd5 Bg7. The comp calls it even, while I prefer Black’s minor pieces over White’s rook and wobbly pawns. 14 … Bg7 15 Qc4 Alternatives: a) 15 Nb4 is ineffective due to 15 … Rxb5 16 Qxb5 Qa8 17 Nd5 0-0 with advantage for Black. b) 15 b4 is met by 15 … Ra8 16 Qe4 Kf8! 17 Bxc6 Bxc6 18 0-0 Ra4 and Black’s extra piece is worth more than White’s pawns, U.Drössler-K.Norchenko, correspondence 2013. 15 … Rxb5 16 Qxb5 Ba8!

Eljanov’s move improves upon 16 … Qa5+?! 17 Qxa5 Nxa5 18 Rd1, when White stood at least equal, J.Hrdina-J.Markauss,

correspondence 1992. Even after 16 … Ba8, it may appear on the surface as if White, with rook and three pawns against Black’s minor pieces, is winning but this is not the case. Firstly, Black’s minor pieces are two bishops in an open position. Secondly, White’s pawns are ineffective and pushing them takes a long time. Black will easily generate counterplay before that happens. 17 Qb6? Efimov misassesses the ending and this move instantly gets him in trouble. He had to try 17 c3 0-0 18 Rd1 Qg5 19 f6 Bxf6 20 h4 Qh6 21 Nc7 Ne7 22 Nxa8 Rxa8 23 Rxd6 Rxa2 24 Rd8+ Kg7 25 0-0 Qxh4 albeit with the better chances for Black. 17 … Nd4 Now the ending is heavily in Black’s favour. 18 Qxd8+ Kxd8

19 Ne3 Covering against the dual threats … Bxd5 and … Nxc2+.

19 … Bh6! Threatening 20 … Bxe3, after which both c2 and g2 hang. 20 0-0-0?! White had more chances to save the game after 20 Rd1 Rg8 21 Rf1 Bxg2 22 Nxg2 Rxg2 23 c3 Nxf5. 20 … Bxg2 Exploiting the pin on the e3-knight. 21 Rhe1 21 Rhg1? loses a piece to 21 … Bxe3+ 22 fxe3 Ne2+. 21 … Be4! 22 c3 Kd7! This nonchalant non-response must have come as a shock for White, who is oddly paralysed by Black’s minor pieces. 23 b4 Not 23 cxd4??, which allows 23 … Rc8+ 24 Kd2 Rc2 mate.

Exercise (combination alert): Find a forcing line which wins material for Black. 23 … Nc2! Answer: Step 1: Chase White’s e1-rook to e2. 24 Re2 Others don’t help: a) 24 Rg1 Nxe3 25 fxe3 Bxe3+ and the g1-rook falls. b) 24 Rf1 Bxe3+ 25 fxe3 Nxe3 and one of White’s rooks will be picked off. 24 … Nxe3 25 fxe3 Bf3 Step 2: Bishop skewer. 26 Rdd2 Bxe2 27 Rxe2 White’s two pawns for the piece are not enough. 27 … Rc8 28 Kc2 d5 29 Kb3 Rc4! Black’s king can deal with White’s two connected queenside pawns, while the rook can maraud along the fourth rank, going after the e3-straggler. 30 a4 Or 30 e4 Rxe4 31 Rxe4 dxe4 32 Kc2 (White’s three connected passed pawns can easily be halted) 32 … Kc6 33 c4 e3 34 Kd3 e4+ 35 Ke2 Bf4 36 h3 Kb6 37 Ke1 Be5 38 Ke2 Bd4 39 Kd1 Bc3 and wins.

30 … Re4 31 a5 Kc6 32 a6 Bxe3 33 Rg2 The rook plans to raid the seventh rank. After 33 b5+ Kc7! 34 Re1 Ba7 35 Rd1 Kd6 White’s pawns are blockaded. 33 … Kb6 33 … Rf4 leads to a quicker win. 34 Rg7 Kxa6 35 Rxf7 Rf4 36 f6 Rf5 37 c4 dxc4+ 38 Kxc4 Bg5!? He is cutting it close, since if every black pawn is traded away, then it’s a theoretical draw. 38 … h6 is easier. 39 Rxh7 Rxf6 40 Kd5 Bf4 41 h4 Kb5 42 Rb7+ Ka4 43 b5 Kb4 44 Rh7 44 h5 is met by 44 … Rh6. 44 … Kxb5 45 h5 Kb4 46 h6 Rb6 47 Rh8? 47 Ke4 was necessary, when Black still has work to do.

Exercise (combination alert): White just blundered in an already lost position. How did Black exploit this lapse? 47 … Bxh6! 0-1 Answer: Black’s e-pawn is immune due to … Bg7+, which would win White’s rook. Game 36 A.Tari-S.Zhigalko European Rapid Championship, Katowice 2017 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bd3 Be6 12 c3 Bg7 13 Nxb5!?

The timing of this sacrifice is designed so that Black is unable to play … Bb7 as in the previous game. The line is similar – but not the same. White reasons that Bd3 will come for free, since Black must move his light-squared bishop twice. 13 … axb5 14 Bxb5 Bd7 This allows Black more flexibility than 14 … Rc8 15 Qa4 Bd7 16 exf5. 15 exf5 So White gets his third pawn for the piece. Nonetheless, I would once again prefer to have Black’s piece than White’s pawns. 15 … 0-0 16 0-0 e4 This allows Black’s pieces access to e5. 17 a4 Re8 18 Qg4 18 Re1 Re5 19 Ne3 Ne7 20 Bxd7 Qxd7 21 f6 Bxf6 22 Qg4+ Qxg4 23 Nxg4 Re6 24 Nxf6+ Rxf6 25 Rxe4 Kf8 has been seen in five correspondence games, in all of which White managed to

hold the draw. 25 … Re6, M.Yeo-F.Kwiatkowski, British League 2009, also leaves chances equal. 18 … Kh8 It isn’t a bad idea to slip the king off the g-file. 19 Qf4 19 Rad1 Rb8 20 Rfe1 Ne5 21 Qxe4 Bxb5 22 axb5 Rxb5 23 f4 Rxb2 24 Kh1 Rg8 25 f6 Bxf6 26 Nxf6 Qxf6 27 fxe5 Qxe5 28 Qxe5+ dxe5 29 Rxe5 Rgxg2 30 Re8+ Kg7 31 Rg1 was Wei YiYu Yangyi, Danzhou 2014. This game too soon petered out to a draw, 19 … Be5 20 Qh6

Of course 20 Qxe4?? hangs the queen to the simple discovery 20 … Bxh2+. 20 … Rc8 Or 20 … Rg8 21 Rfe1 Rc8 22 f6 Rg6 23 Qe3 Bxf6 24 Qxe4 Bg7 25 Rad1 Re6 and a draw was prematurely agreed in V.Borges-E.Choma, Curitiba 2016. I slightly prefer Black.

21 f6 Rg8 22 Rae1 22 Rfe1 would transpose to the previous note. 22 … Rg6 23 Qh4 Rg4 24 Qh5? White is in “it’s-him-or-me” mode. He should have offered a draw by repetition with 24 Qh6, when Black can either take the draw or play on with 24 … Rg6 25 Qh4 Bf5. 24 … Bxf6 25 Qxf7 Be5 Black has a strong attack against White’s under-defended king. 26 f4? White had to try 26 g3 Rg7 27 Qh5 Be6 28 Ne3 Na7 29 Ba6 Rc5 30 Qe2 Nc6, with an inferior yet still playable game.

Exercise (combination alert): White missed something on his last move. How should Black continue?

26 … exf3?! Black allows the opportunity to sneak by, not sensing his position’s concealed power. Answer: 26 … Rg7! is very strong; e.g. 27 Qh5 Bg4 28 Qh6 Rg6 and White’s queen is trapped. He is hopelessly lost after 29 Qxg6 hxg6 30 fxe5 dxe5. 27 Rxf3? 27 g3 Rg7 28 Qxf3 was forced, but even here a dark shadow hangs over White’s king. 27 … Rg7! Now this idea is even stronger. 28 Qh5 Bg4 0-1 Since 29 Qh6 Bxf3 picks off an entire rook. Game 37 T.Luther-P.Leko Essen 2002 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bxb5!?

This line – the equivalent of discharging a 12-gauge shotgun at point blank range – is a line eminently suited for the natural anarchist. A single move has the power to throw an otherwise stable position into utter chaos. The bluntness of White’s intent is jolting. The sacrifice is barely sound. The only danger for Black is forgetting the analysis. For the piece White gets two pawns and is pretty much guaranteed an exchange as well. So from an arithmetical standpoint, this isn’t even a sacrifice. Yet it is, since in a middlegame Black’s minor pieces and powerful pawn centre will outgun White’s rook and two pawns. As in the last two games, I believe it is White who fights for equality in this line. Nonetheless we must be heavily prepared for it, or White will win by simply outbooking us. 11 … axb5 12 Nxb5 Threatening to fork on c7. 12 … Ra4! This is universally thought to be Black’s best response.

13 b4!? Even worse than the opponent defeating us is when we hang ourselves. This is White’s most radical option. Such “power” moves, while seizing the initiative, also bring with them a curse, since after Black’s next move White’s game is strategically beyond repair. He must whip up a winning attack, or die trying. In the next game we look at 13 Nbc7+. 13 … Rxb4 14 Nbc7+ In such positions even strong players may come to starkly opposite conclusions as to who is winning and who is losing. The comp slightly favours Black. GM Luther feels this is not a time for decorum or proper manners, so he escalates, ignoring Black’s hanging rook and goes all out for mate. 14 Nxb4 Nxb4 15 0-0 was S.Sulskis-V.Cmilyte, Lithuanian Championship, Siauliai 2005. Here Black has access to a powerful and untried move in 15 … Qd7!, when 16 Qe2 fxe4 17 Rfd1 Rg8 18 Rab1 Nc6 19 Qxe4 Rg6 20 c4 Bb7 21 Nc3 f5 provides a strong initiative. 14 … Kd7 Welcome to the Sveshnikov. Believe it or not, this is a book position, and one which favours Black. 15 0-0 Rg8 16 g3 After 16 Nxb4 Nxb4 17 Nd5 Nxd5 18 Qxd5 Ke7, A.Naranjo Moreno-D.Alsina Leal, Linares 2017, Black’s two bishops are obviously worth more than White’s rook and pawn. Also, Black’s king looks perfectly safe in the centre. 16 … Rb7 17 Qh5 Threatening both f7 and f5.

17 … Qg5! This move can be regarded as a refutation. Instead: a) 17 … Rxc7?? allows 18 Qxf7+ Ne7 19 Nf6+ Kc6 20 Qb3!! Rb7 21 Qa4+ Kc7 22 Qa5+ Rb6 23 Rab1 Kb7 24 Nd5 Nxd5 25 Qxd5+ and Black is busted as the g8-rook hangs. b) 17 … Ne7?! 18 Qxf7 Rg7 can be met by 19 Nf6+! Kxc7 20 Ne8+ Kb8 21 Nxg7 Ng6 22 Rab1! fxe4 23 Rxb7+ Bxb7 and it’s anybody’s game. 18 Qxf7+ A queen exchange would be fatal for White. 18 … Be7 19 Nxe7 Nxe7 20 Ne6 Qg6! 21 Nf8+ White’s initial fury is now subdued by the pragmatic desire to survive. 21 … Rxf8 22 Qxf8 fxe4 Black is winning with two minor pieces against White’s rook and pawn. The light squares around White’s king are

punctured beyond repair. 23 Rfb1 Rc7 24 a4 e3! Even chopping on c2 was winning. 25 fxe3 25 Qf3 was necessary, although Black is still winning after 25 … exf2+ 26 Qxf2 Rxc2 27 Qa7+ Rc7.

Exercise (planning): What is Black’s decisive idea? 25 … Qe4! 0-1 Answer: Seize control over and set up a queen and bishop battery on the h1-a8 diagonal. White is helpless after 26 Rb2 Bb7 27 Rxb7 Rxb7 28 a5 Rc7, as … Rxc2 is threatened and the white a-pawn is going nowhere. Game 38

S.S.Ganguly-D.V.Prasad Hyderabad 2000 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Bxf6 gxf6 10 Nd5 f5 11 Bxb5!? axb5 12 Nxb5 Ra4! This is the same move I recommended in Sveshnikov: Move by Move, and nothing has altered my opinion. White is the one struggling for equality. All the same, White is very likely a specialist in this position, who is accustomed to confusing Sveshnikov players with the sacrifice. So Black must be thoroughly familiar with the defensive ideas in order to survive and win. 13 Nbc7+ This is White’s most dangerous continuation for two reasons: 1. The move immediately displaces Black’s king to the insecure d7-square. If White waits too long, then Black’s king may find a way to slide into a safer, less congested home such as d8 or f8. 2. By playing the immediate 13 Nbc7+ White clears the path for the a- and b-pawns to push forward, without a blocking knight on b5. 13 … Kd7 From a superficial glance, Black’s position gives off the smell of unwashed, homeless dishevelment. The computer isn’t fooled and slightly prefers Black, whose king is a lot safer than it looks.

14 0-0 Once again the best move. White offers another pawn, picking up an important tempo for his attack. It’s much too late for safety measures. a) 14 c4? Rxc4 15 0-0 Nd4! 16 Nb6+ (or 16 Na8 Ke8 17 Nab6 Rc2 18 Re1 Bh6 19 exf5 Bd2! 20 Qg4 Kf8! 21 Nxc8 Qxc8 22 Qh4 Qxf5 23 Ne3 Qg6 0-1 Z.Tsydypov-D.Lintchevski, Moscow 2017) 16 … Kxc7 17 Nxc4 fxe4 18 Qa4, C.Lhagvasuren-G.Tunik, Ulaanbaatar 1984, is best met by 18 … Kb8, when Black consolidates. b) 14 Qh5? is just dumb, since it allows … Rxe4 with check, ruining White’s castling, while keeping the h1-rook out of play. After 14 … Rxe4+ 15 Kf1 Ne7 16 Qxf7 Kc6 White is just losing. c) 14 Nb6+? Kxc7 15 Nxa4 is met by 15 … Rg8 16 0-0 fxe4, threatening … Bh3, and White is completely busted. 14 … Rxe4

Black eliminates an important central pawn at the cost of a tempo. Instead, GM Kotronias gives 14 … Rg8 an exclam, though I’m not convinced it is any better or worse than the move GM Prasad played in the game. After 15 Qh5 Qg5! 16 Qxf7+ Be7 17 g3 Qg6 18 Qxf5+ Qxf5 19 exf5 Nd4, R.Zelcic-I.Jelen, Bled 2005, and now 20 Nb6+ Kxc7 21 Nxa4 Bxf5 Black’s minor pieces and light-squared attack is worth more than White’s rook and three pawns. 15 Qh5 Ne7 This move is the crucial alternative to moving the knight to d4. 15 … Nd4 was covered in Sveshnikov: Move by Move. 16 Qxf7 Kc6!

There is a Polish proverb that says: “Not my circus; not my monkeys.” We should remember this when the opponent attempts to draw us into their crazy world. But what if it’s our crazy world too? I have probably been in this position more than 50 times with students in analysis sessions, while covering the

Sveshnikov. Having been here so often the themes have become very familiar. I suggest that you too reach such a level of familiarity. The black king is safe, mainly since White’s own queenside pawns shelter him to a great degree. The comps have worked it all out and call it dead even here. 17 c4 After 17 Rfd1 Qd7 18 Qh5, in every game in my database Black played the inferior 18 … Rg4, but 18 … Kb7! is an untried improvement, when 19 Rd3 Kb8 20 Rb3+ Bb7 21 Qf7 Rd4 22 f3 Qc6 23 Rb6 Qd7 24 Rb3 Qc6 25 Rb6 leads to a draw by repetition. 17 … Qd7 17 … Kb7 18 Nb5 Kb8 19 c5 was R.Wagner-F.Vidalina, correspondence 2005. White attempts to violently open central lines but Black stands better after 19 … dxc5 20 Nf6 Nc6! 21 Nxe4 fxe4 as the minor pieces are worth more than White’s rook and pawn. 18 Na8 Qe6 Prasad goes for the full point. 18 … Ng6 is a key alternative, which leads to a forced draw after 19 Nb4+ Kb7 20 Qd5+ Kb8 21 Nc6+ Kxa8 22 Qb5 (threatening mate on the move) 22 … Qb7 23 Qa5+ Qa6 24 Qc7 Qb7 25 Qa5+ etc. 19 Qe8+ Qd7 20 Nxe7+ Bxe7 21 Qxh8 Don’t panic. We are still okay here, since White’s knight in the corner is trapped. 21 … Kb7! 22 Rfd1 The first new move of the game. B.Langhein-G.Traut, correspondence 1983, had seen 22 Rac1 Kxa8 23 Rc3, when 23 … Qe6! 24 Rfc1 (or 24 Qxh7 Rh4) 24 … Rg4 25 Qxh7 (or 25 Rb3 Qg8) 25 … Bb7 looks like the best way to defend.

22 … Qe6 22 … Kxa8 23 c5 d5 24 Qxh7 Qe6 is fine for Black too. 23 Rac1 Kxa8 24 Qe8 Rd4!? This self-damage is unnecessary. After 24 … Bg5! 25 Qa4+ Kb8 26 Rc3 Bb7 Black stands no worse. 25 b4?! In this position patience is required. The problem of course is that patience is not the strong suit of most attacking GMs. White should take the opportunity to damage Black’s pawns with 25 Rxd4! exd4 26 Kf1, when he stands no worse. 25 … Ka7?! Better is 25 … Bg5! 26 Qa4+ Kb8 27 Qb5+ Bb7, which is in Black’s favour. 26 Qa4+ Kb8 27 Rxd4! exd4 28 c5?! Ganguly raises the ante, not wishing to back down. He is determined to open lines against Black’s king, but in doing so he

risks overpressing. White should take perpetual check with 28 Qb5+ Bb7 29 Qe8+ Ka7 30 Qa4+. 28 … dxc5 29 bxc5 Bxc5! 30 h3 After 30 Qb5+ Bb6 31 a4 Ka8 32 h3 Ba7 33 Rc6 Qd7 Black’s king is safe and White is in deep trouble due to Black’s passed d-pawn and bishops. 30 … Bb6 31 Rc6?! Inattention is the thief who steals our dreams. The only way to try and resist was 31 Qb4 Bb7 32 Rd1 Qc6 33 f3 Ba6, though White is still in deep trouble.

Exercise (combination alert): White’s last move was a blunder. Black to play and force the win of material. 31 … Qe1+

Answer: Step 1: Give check on e1, forcing the white king to h2. 32 Kh2 Bc7+ Step 2: Give check on c7, forcing White to hand over the exchange. 33 Rxc7 Forced, as 33 g3 Qxf2+ 34 Kh1 Bb7 wins at once. 33 … Qe5+! Step 3: Give check with the queen, defending the d4-pawn, before capturing the rook. After 33 … Kxc7 34 Qc4+ Kb8 35 Qxd4 White has chances to survive. 34 g3 Kxc7 35 Qc4+ Kb8 36 Qb4+ Bb7 37 Qf8+ Kc7 38 Qf7+ Kb6 39 Qb3+ Qb5 40 Qe6+ Qc6 Black lines up a queen and bishop battery on the h1-a8 diagonal. When his checks run dry White is doomed. 41 Qb3+ Kc5 42 Qa3+ Kd5 0-1 After 43 Qb3+ Ke5 44 f4+ Kf6 the checks duly run out.

Chapter Four 7 Nd5: The Modern Approach

When White plays 7 Nd5, Black will exchange knights. White then recaptures with the e4-pawn, which immediately creates an imbalance of pawn majorities. In this book we concentrate on the 8 … Ne7!? line, recently popularized by Carlsen in his 2018 World Championship battle with Fabiano Caruana. In the match Carlsen held off Caruana’s Rossolimo (3 Bb5) with ease. Then in the final game at a classical time control, Caruana risked an Open Sicilian, which was met by the Sveshnikov. Carlsen got a promising position, yet shockingly chose to agree to a draw to push the match into a rapidplay playoff, his realm of superiority. In his only rapid game as Black, Carlsen dismantled his challenger with the Sveshnikov.

So if a world champion is willing to stake his world title upon the Sveshnikov, perhaps we too should generate faith in the opening. Game 39 F.Caruana-M.Carlsen World Championship (Game 12), London 2018 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5

7 … Nxd5 8 exd5 Let’s try and discover meaning in the new geometric constellations: 1. An imbalance arises. White has a 4 to 3 queenside pawn majority, while Black has the same on the kingside. 2. Black’s goal is to hinder White’s majority, while activating and pushing forward on the kingside.

3. The d5-square has been plugged with a white pawn which, as I have mentioned several times in the book, favours Black. On the other hand, after 8 exd5, White gains a tempo on the c6-knight, which justifies the concession. 8 Qxd5? keeps d5 unblocked but loses too much time. Black seizes an advantage with 8 … a6 9 Nc3 Nd4! 10 Kd1 (not 10 Bd3??, since 10 … Rb8! threatens to trap White’s queen with … Be6, while after 11 Qc4 b5 White must resign since the c3knight must be given up) 10 … Rb8! 11 Qc4 Be6, R.TeschnerG.Kieninger, Augsburg 1951, when White is strategically busted, down in development and with the king stuck in the centre. 8 … Ne7 The main move here is to retreat the knight to b8, after which it regroups to d7, fighting for control over c5, while bolstering e5. This line is covered extensively in Sveshnikov: Move by Move. Carlsen had in fact drawn his previous two match games as Black with 8 … Nb8. This time he comes armed with a lesser played move. 8 … Ne7 clogs up the black kingside, without visible benefit. However, the idea of the move is that Black generally goes after the white king, so it’s logical to transfer the knight in that direction with a future … Ng6 or … Nf5. 9 c4

We’ll examine 9 c3 at the end of the chapter. 9 … Ng6 As Carlsen goes, so goes our book. I have used Carlsen as our Sveshnikov role model, so we only cover the lines he plays. We don’t have room to look at the major alternative 9 … Nf5. 10 Qa4 Threatening awful discoveries with the b5-knight. The recent trend is that all the world’s top players seem to be focused on this move (although Carlsen is currently holding his own easily), so we’ll devote our attention to it. Later on we’ll take a look at 10 Bd3. 10 … Bd7 11 Qb4

Evading the pin, while attacking d6. 11 … Bf5 The idea is to prevent Bd3 from White. Carlsen also plays 11 … Qb8, as we’ll see in Games 42 and 43. 12 h4 White’s main move. The idea is to displace Black’s knight with h4-h5. The downside of the move is that White may lose his option to castle short, since the inclusion of h2-h4 places his king at risk if he later castles kingside. Others: a) 12 Qa4 repeats moves after 12 … Bd7 13 Qb4. Now Black can either return the bishop to f5 or, if a potential draw is undesirable, avoid a further repeat by opting for 13 … Qb8 this time, transposing to 11 … Qb8 lines. b) 12 Be3 attacks a7 but can be met by 12 … Be7! (offering the a-pawn to gain time) 13 Nxa7 0-0 14 Be2 Nf4 15 Bxf4 exf4 16 Nb5, as in V.Gonchar-A.Emelyanenko, correspondence 2001. After 16 … Bf6 17 0-0 Be5 Black has the bishop pair and darksquare power, leading to full compensation. 12 … h5!

Carlsen comes prepared with a risky new move, an improvement over the older 12 … a6. Black prevents h4-h5 at the cost of weakening his own h-pawn and making it vulnerable to ideas such as Be2 and Qd1. Carlsen factored in the following: 1. The plan of Be2, Qd1 and finally Bxh5 takes White an eternity. So, even if the pawn is lost, it costs White a lot of time, virtually ensuring that Black gets good play for the sacrificed pawn. 2. Black may later relocate the knight to either e7 or f8 and then play … g7-g6, securing the h5-pawn. So the fact that Black has played an early … h7-h5, doesn’t necessarily mean he is obliged to sacrifice. Instead, after 12 … a6 13 h5 axb5 14 hxg6 fxg6 15 Qxb5+ Qd7 16 Qxd7+ Bxd7, it’s difficult to see how White can ever lose and he can perhaps play safely for the win, J.BrookesC.Pauwels, correspondence 2008. 13 Qa4

Faced with a novelty Caruana opts to repeat moves to gain time on the clock. 13 … Bd7

14 Qb4 14 Qb3 was seen in B.Dastan-T.Kraus, Teplice 2019. I ran the position on Komodo and it went for explosive complications with 14 … Nxh4!? 15 Rxh4 a6! 16 Bg5 Qxg5 17 Nc7+ Kd8 18 Nxa8 Qxh4 19 Qb6+ Ke8 20 Qxb7 Be7 21 Nb6 Bf5 22 Nc8 Bxc8 23 Qxc8+ Bd8 24 Qxa6 Qe4+ 25 Be2 Qxg2 26 Qxd6 Qh1+ 27 Bf1 Qe4+ 28 Be2, when Black can either take the repetition draw, or go for the full point with 28 … Rh6. 14 … Bf5 As Black in a world championship match, Carlsen is happy to repeat. Otherwise 14 … Qb8 transposes to 11 … Qb8 lines again. 15 Be3

15 Bg5 doesn’t really gain a tempo, as after 15 … Qb8 “you have to decide how to fight against the idea of … a7-a6 followed by … Be7,” writes GM Inarkiev. We’ll look at this line in the next two games. 15 … a6 16 Nc3 Qc7 Carlsen suppresses White’s intended c4-c5, while provoking Bb6. 17 g3 After 17 Bb6 Qd7 18 g3 Be7 19 Be2 0-0! 20 Bxh5 Bd3 21 0-0-0 e4 22 Bxg6 fxg6 23 Rhe1 Rae8! 24 Re3 Bf6 Black has full compensation for the sacrificed pawn. 17 … Be7

18 f3!? This move is dual purpose: 1. It restricts Black from playing … e5-e4, followed by … Ne5.

2. If Black castles kingside, then White may castle queenside and try and open lines with a g3-g4 break. 18 Be2!? is also interesting, because it is not clear how profitable 18 … e4 is for Black. 18 … Nf8! A wonderfully flexible move. Carlsen realizes that Black’s knight has served its purpose on g6 and is restricted. So he transfers it to d7, where it: 1. Suppresses White’s c4-c5 break. 2. Protects b6 from invasion by Bb6 or Na4-b6. 19 Ne4 The idea is to play Bd3. Inarkiev suggests 19 Be2!? Nd7 20 0-0, but Black looks okay to me after 20 … 0-0 21 Ne4 Bg6. 19 … Nd7 20 Bd3 0-0!? 21 Rh2?!

Caruana certainly did not want to face down Carlsen in a rapidplay tiebreak to decide the World Championship. So he is

desperate for a win in this final classical game of the match. The plan is to castle long and then go crazy on the kingside. The trouble is that Black will be faster. White should have just castled kingside and taken his chances. Going after the bishop pair with 21 Nf6+?! is a bad idea. After 21 … Bxf6 22 Bxf5 g6 23 Bd3 Nc5 24 Bc2 b5! Black seizes the initiative. 21 … Rac8 22 0-0-0 Bg6! Brilliant strategic judgment. Carlsen realizes that Black needs to play … f7-f5, even if it means handing White a hole on g5. After 22 … b5 23 Kb1 bxc4 24 Rc2 the comp calls it even, while I slightly prefer White.

23 Rc2! White’s king requires help. 23 Kb1?! f5! 24 Ng5 Bxg5 25 hxg5 e4! 26 fxe4 Ne5 leaves him in deep trouble. 23 … f5! 24 Nf2!

The fact that the knight had a juicy hole on g5, and deliberately avoided it, is a deeply troubling sign for White’s position. Caruana avoids 24 Ng5?! because of 24 … Bxg5 25 hxg5 e4 26 fxe4 fxe4 27 Be2 Nc5 (intending … Nd3+) 28 Bxc5 dxc5 29 Qb3 e3 30 Bd3 Qxg3! 31 Bxg6 Qxg5 (with dual threats of … Qxg6 and … e3-e2+) 32 Re2 Qxg6 33 Rxe3 Rf2 and Black is in control of the position. 24 … Nc5 25 f4 Damned if you do and damned if you don’t. White, faced with only unpleasant choices, seeks to resolve the central tension. In doing so he allows Black’s dark-squared bishop into the coming attack. I can understand why Caruana didn’t want to grovel for a draw in an ending after the objectively correct 25 Bxc5 Qxc5 26 Qxc5 Rxc5 27 b4 Rc7 28 c5 Be8! 29 Rdd2 g6 30 Bc4 a5! 31 a3 Kg7 32 Nd3 Ba4 33 Rc3 Rfc8. Here White’s queenside is under intense pressure and, even worse, Caruana would be defending it against one of the greatest endgame players in history. 25 … a5?!

Not the best choice. Even the most confident player in the world may succumb to a case of the jitters when the world championship is on the line in a single game. Carlsen missed a pair of potent continuations, either of which might have dramatically ended the match in his favour: a) 25 … exf4! 26 Bxf4 (after 26 gxf4 Bxh4 White has nothing for the pawn) 26 … b5! 27 Qd2 (27 cxb5?? loses on the spot to 27 … Nxd3+ as White’s defender of c2 is eliminated) 27 … Bf6 28 cxb5 axb5 29 Bxb5 Rb8 30 Bc4 Rfc8 with strong pressure against the white king, who is unlikely to survive. b) 25 … b5! at once is also promising, when Black not only threatens to win the c4-pawn, but is also ready to play … a6-a5 and … b5-b4. Here the lines are much more complex. After 26 Bf1 it is important to respond 26 … exf4! 27 gxf4 Bxh4 28 cxb5 axb5 29 Kb1 (29 Bxb5 Na6! 30 Qa4 Qxc2+ 31 Qxc2 Nb4 32 Qxc8 Rxc8+ 33 Kb1 Be8! transposes) 29 … Na6! 30 Rxc7 Nxb4 31 Rxc8 Rxc8 32 Bxb5 Be8! 33 Be2 Rc2 34 Rd2 Rxd2 35 Bxd2 Nxd5 36 Nh3 Kf8. White is a pawn down in a difficult

ending. It’s not so easy to mobilize his queenside passed pawns, so Black has all the winning chances. 26 Qd2 e4 27 Be2 Be8 The bishop was doing nothing on g6 and is transferred to the queenside in preparation for an attack. 28 Kb1 Bf6 Not yet 28 … Ba4, as 29 b3 Qb6 30 Rb2 doesn’t bother White. 29 Re1?!

Let’s take stock: 1. Black’s passed e-pawn juts into White’s position and keeps him nervous about invasions on d3. 2. Black’s c5-knight is well posted. 3. The centre is blocked. This helps White, since Black has no easy way to open the queenside.

Exercise (critical decision): Black has two candidate moves: a) 29 … a4, expanding on the queenside; b) 29 … Ba4, provoking White into weakening with b2-b3. One leads to an advantage, while White hangs on in the other. Which one would you play? 29 … a4?! Carlsen - understandably – plays it too safely. Answer: He may have won the game had he chosen 29 … Ba4!, which is very strong. It was the great Mikhail Tal who taught us that sometimes we solve the problem not by what we see, but by that we are unable to see. White falls into a geometric paradox, no matter how he responds: a) 30 b3 Bxb3 31 axb3 Nxb3 32 Qd1 a4 (threatening … Qa5) 33 Rg1 b5 is winning for Black, since if 34 cxb5? then 34 … Qa5 35 Ka2 Rc3 and White’s position collapses. b) 30 Bxh5?! Bxc2+ 31 Qxc2 b5 32 cxb5 Qb7 and White is busted. c) 30 Rcc1 b5! 31 cxb5 Qb6 32 Bxh5 Qxb5 33 Bd1 Rb8 34 Bd4 Bxd4 35 Qxd4 Bxd1 36 Rexd1 Rfe8! 37 h5 e3 38 h6 Rb7 and again Black is winning. 30 Qb4 He should energise his knight with 30 Nh3!. 30 … g6 After 30 … Na6! 31 Qb6 Qxb6 32 Bxb6 g6 33 Bf1 Bd7 34 Nd1 Nc5 35 Ba5 Rfe8 Black stands slightly better. 31 Rd1 Ra8 ½-½

Carlsen held a slight edge in a sharp position, with a 38 to 15 minute advantage on the clock, but shocked everyone by offering a draw which, from a sporting aspect, is a giant anticlimax. The critics on the internet went nuts about this apparently preposterously wasteful decision, accusing Carlsen of losing his nerve, which is of course nonsense. I was one of the few who argued that his decision to draw was a smart one, for the following reasons: 1. Sure, Black stands slightly better but if he goes all out for the win, he can easily lose, since there is no risk-free way to play for the full point. In such a crazy complicated position, even a world champion may falter. 2. Caruana is behind on the clock, yet in such a situation he will do nothing and invite Carlsen to come and get him. I actually don’t believe White’s game is all that difficult to play, at least for now. 3. Carlsen was clearly the heavy favourite at the faster time controls, so why not enter the rapidplay phase.

Game 40 F.Caruana-M.Carlsen Grenke Classic, Karlsruhe/Baden 2019 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Qa4 Bd7 11 Qb4 Bf5 12 h4 h5 13 Bg5 Qb8 14 Qa4 This time White isn’t just repeating moves. He lures the black bishop back off the b1-h7 diagonal in order to place his queen there. In the next game we look at 14 Be2. White has also tried 14 g3!? a6 15 Nc3 Be7 (“The exchange of the dark-squared bishops is profitable for Black,” writes GM Inarkiev) 16 Be3 0-0! (this is essentially a declaration: “Go ahead. I dare you to waste time grabbing my h-pawn.”) 17 Be2 e4 18 Bxh5 Ne5 19 Be2 Qc8 20 Na4 Bd8 21 Qxd6 Nxc4 22 Qf4? (22 Qc5 Nxe3 23 Qxe3 gives Black excellent compensation for the pawn but this is how White should have continued) 22 … Bc7 23 Qg5 f6 24 Qh5 Ne5! 25 Rc1 Bg6, and a dismayed Karjakin, Carlsen’s former challenger, found that his queen was trapped, S.Karjakin-E.Inarkiev, Nazran 2019. 14 … Bd7 15 Qc2!?

There would be no point in going back to b4, as it no longer threatens anything, so Black could just play 15 … a6. Caruana’s move carries high-risk strategic freight. He invites the imbalancing 15 … Bxb5, gambling that his bishops and light square power are enough to offset his soon-to-be inferior queenside pawn majority and isolated d5-pawn. 15 … Bxb5 I like the decision to chop and don’t believe White now stands better. 16 cxb5 Be7 Besides developing, Black seeks to eliminate White’s bishop pair and perhaps turn the game into a bishop versus knight contest. 17 Bd3 Nf8!? Caruana’s opening experiment, while not inherently strong, had the effect of taking Carlsen out of his home preparation. Correct was 17 … Bxg5! 18 hxg5 (note that 18 Bxg6? is a strategic error: after 18 … Bd8! 19 Bd3 Ba5+ 20 Kf1 Bb6

Black’s opposite-coloured bishop is overwhelmingly superior to White’s clunker on d3) 18 … Ne7 19 g6 Qc8! 20 gxf7+ Kxf7 21 Qd2 Qg4 22 0-0 Qf4! and Black is just fine in the ending. 18 b6 White’s advantage is more substantial if he seizes the c-file at once with 18 Rc1!. 18 … Nd7 More accurate than 18 … axb6?! 19 Bb5+ Nd7 20 Bxd7+ Kxd7 21 0-0 with some advantage for White. 19 bxa7 This way Caruana’s queenside pawn majority is fluid once again. 19 Bb5 would be met by 19 … Qd8. 19 … Qxa7 20 0-0 Bxg5 21 hxg5 0-0 22 Bf5 Caruana wants to seize control over c8, so that Black is unable to oppose the file. 22 … Rad8 Now White’s bishop proves to be superior to the black knight. Carlsen should have played 22 … Nc5! 23 b4 Na4 24 Rfd1 Qb6 25 a3 Qd8 26 Qd2 g6 27 Bh3 Re8, when Black stands at least equal. 23 b4 The idea is to cut off … Nc5. 23 … Qd4! 24 Qc7 Nb6 25 Rad1 Qf4 26 Bh3 Trading down with 26 Qxb6!? Qxf5 27 Qxb7 Qxg5 was an interesting option, when a battle between White’s two connected queenside passers and Black’s direct attack is in process. The comp unhelpfully evaluates it at dead even. To a human, it feels as if one of the two must have the advantage. Which one is the unanswered question. 26 … Na4

27 g6! After 27 Qxb7? Nb2! 28 Rb1 Nc4 (threatening a fork on d2) 29 Rbc1 Nd2 30 Rfe1 Qxg5, Black’s chances to get at White’s king are superior to White’s chances of promoting one of his queenside pawns, which look too slow. 27 … fxg6! 28 Qc2! Double attack. Caruana avoids Carlsen’s clever strategic trap 28 Qxb7? Nc3 29 Rde1 Kh7 30 a3 Qc4 31 Be6 Ne2+, when White must hand over the exchange, since 32 Kh2 Qh4+ 33 Bh3 Rxf2 34 Rxf2 Qg3+ 35 Kh1 Qxf2 is hopeless. 28 … b5 29 Qxg6 Rf6! 29 … h4 30 Be6+ is in White’s favour. 30 Qxh5 White has won a pawn, yet Black’s superior piece activity easily compensates. 30 … g6! 31 Qg4 Qxg4 32 Bxg4 Rf4 33 Be6+ Kg7 34 Rc1!

Caruana correctly returns his extra pawn to seize the seventh rank for his rook. After 34 a3? Nc3 35 Rd2? Ra8 White is in deep trouble. 34 … Rxb4 35 Rc7+ Kh6 36 g3

Carlsen is a scary guy in any simplified position or ending as he plays them virtually error free. 36 Re1! Nc5 37 Re3 Rf8 is also slightly in Black’s favour. 36 … Nc5 37 f4! Endgame principle: the side in trouble should aim to exchange pawns. 37 … exf4 38 Rxf4 Rxf4 39 gxf4 Ra8 The a2-pawn is doomed, so Caruana goes after d6. 40 Rc6 Ne4 41 Bh3?! This inaccuracy leads to more suffering. After 41 Rb6! Rxa2 42 Rxb5 Rf2 43 f5 gxf5 44 Rb4 Rf4 45 Bxf5 Rxf5 46 Rxe4 Rxd5 47 Kf2 the ending is drawn.

41 … Rxa2 42 Bg2 Ng3! 43 Rxd6 Ne2+ 44 Kf1 Nxf4 Now White finds himself a pawn down, yet his chances to hold the draw are excellent, since Black has only two pawns left. 45 Be4 Ra4

Exercise (combination alert): Usually if one of Carlsen’s opponents is a pawn down in an ending, their task is hopeless. But not this time. How did Caruana force a clear theoretical draw from this position? 46 Bxg6! Answer: The transformative, light bulb moment. This is not just sour-graped petulance in a lost position. Caruana finds a creative solution – giving up a piece on g6. Black cannot hang

on to his final pawn, so the game is drawn. Despite Carlsen’s various attempts to decoy, sidetrack and distort, Caruana keeps his bearings. 46 … Nxg6 47 Rb6 Rf4+ 48 Ke1 b4 Carlsen cannot hang on to his final pawn. 49 d6! Rd4 49 … Kg7 50 d7 Rd4 51 Rxb4! Rxd7 is similar. 50 d7! Rxd7 51 Rxb4 This position is theoretically drawn. Black has 50 moves to probe before White can claim the draw. Maybe Carlsen can beat an IM here. But Caruana? No way. 51 … Kg5 52 Ke2 Re7+ 53 Kd2 Kf5 54 Rb5+ Ne5 55 Rb4 Rd7+ 56 Ke2 Nd3 57 Ra4 Nf4+ 58 Ke1 Kg4 59 Ra3 Cutting off Black’s king. 59 … Rd8 60 Rc3 Rh8 61 Ra3 Rh2 62 Kd1 Kf5 63 Re3 Nd5 64 Re8 Kf4 65 Kc1 Nc3 66 Re7 Ne4 67 Re8 Ke3 68 Re7 Rh6 69 Rc7 Kd3 70 Rd7+ Rd6 Carlsen agrees to the draw with a rook swap. 71 Rxd6+ Nxd6 ½-½ Game 41 S.Karjakin-M.Carlsen Shamkir 2019 In their 2016 world championship match, Carlsen was unable to defeat Sergey Karjakin a single game with the black pieces, at the classical time control. In 2019, Carlsen took up Sveshnikov and – bam! – a win with Black! 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Qa4 Bd7 11 Qb4 Bf5 12 Qa4 Bd7 13 Qb4 Bf5 Once more the players insert a couple of repeats before getting down to business. 14 h4 h5 15 Bg5 Qb8 16 Be2

In this version White keeps his queen on b4 and prepares to complete development. 16 … a6 17 Nc3 Qc7 18 g3 Be7 19 Be3 e4!? Carlsen clears e5 for his knight, at the risk of artificially isolating his e-pawn. A few months later an improvement was found with 19 … Nf8. K.Alekseenko-Y.Kuzubov, Konya 2019, continued 20 a4 Nd7 21 a5, when Black sacrificed his h-pawn by castling kingside, which I believe is unnecessary. Black looks fine after the simple 21 … Bg6. 20 0-0 0-0!? Carlsen takes the high-risk route which, by the way, is actually approved by the comp. White will waste time if he chops the h5-pawn. Carlsen gets away with such risky ideas, but I’m not sure everyone else would. The safer option is 20 … Ne5 21 Bb6 Qd7 22 Bd4 Rc8, when Black looks okay.

21 Bxh5 Karjakin risks losing time to pick up a pawn. 21 … Ne5 22 Be2 Qd7!? Carlsen takes aim at White’s weak kingside light squares. Objectively, 22 … Rac8! was stronger; e.g. 23 Rac1 Nd3 (not 23 … Nxc4?? which walks straight into a pin after 24 Nxe4 b5 25 Nd2 and Black is busted) 24 Bxd3 exd3 25 Na4 Rfe8, when Black’s light-squared power and passed d-pawn offer full compensation for the missing pawn. 23 Qa4 Principle: the player with extra material can bully via the threat to swap pieces. 23 … Qc8

24 c5?! The once-solid boundaries become fluid. The advantage soon swings to Black after this natural move. Not 24 b3?? either, as 24 … Bg4! 25 Kh2 Bxh4! gives Black a winning attack, since 26 gxh4? is crushed by 26 … Bxe2 27 Nxe2 Qg4 28 Ng3 f5!. Instead, 24 Bd4! is indicated and after 24 … Nf3+ 25 Bxf3 exf3 26 Qd1 Bh3 27 Qxf3 Bxf1 28 Rxf1 Qxc4 29 Rd1 the comp gives White full compensation for the exchange and pawn. 24 … dxc5 25 Nxe4 c4! A dual-purpose move: 1. Now … b7-b5 is coming. 2. White’s knight suddenly hangs, since it got cut off from protection by the queen. It’s clear that Black has full light-square compensation for the sacrificed pawn. 26 Nc3?!

Natural doesn’t always equate to best and White loses further time after this move. Karjakin had to endure the (admittedly inhuman) self-pin with 26 Qc2!, intending 26 … Re8 27 Rad1 Bf8 28 Rd4, though White’s e4-knight still dangles uncomfortably. 26 … b5 Tempo number two. 27 Qd1 b4 Tempo number three. 28 Na4 Threatening Nb6. 28 … Be4! Now … Qh3 is a scary threat. 29 Qd4 Qf5! Threatening 29 … Nf3+.

Exercise (critical decision): White’s kingside light squares are under siege. He has three options: a) Give a pawn back with 30 f3. b) Play 30 Bf4. c) Boot the knight away with 30 f4. Only one option allows White to survive. Which one do you want to play? 30 f4? Answer: a) 30 f3! was forced. After 30 … Nxf3+ 31 Bxf3 Bxf3 32 Nb6 Rad8 33 Rf2 Qh5 34 Qxc4 Bxd5 35 Nxd5 Rxd5 White’s position is a tad loose, yet still quite playable. b) 30 Bf4?? runs into 30 … Bf6 (threatening a discovery against White’s queen) 31 Bxe5 Bxe5 32 Qe3 Rae8 33 Rac1 Bd6 34 Rxc4 Bc2 35 Bg4 Qg6 36 h5 Qf6 and White loses a piece. 30 … Qg6! 31 Bf2 Not 31 fxe5?? Qxg3 mate. 31 … Nd3! Do you remember that amazing world championship game where Kasparov defeated Karpov with an octopus-like knight on d3? Well, this looks like a repeat. 32 h5 After 32 Qxc4 Nxf2 33 Kxf2 Rac8 34 Qd4 Bf6 35 Qe3 Bxd5 36 Nb6 Rfe8, White can resign. 32 … Qf5 33 Bg4 Or 33 Bxd3 cxd3 34 Rfe1 Bf3 35 Re3 Qxh5 36 Rxf3 Qxf3 and wins. 33 … Qxg4 34 Qxe4 Bd6! 35 Qg2 Or 35 Qxc4 Nxf4 36 Nb6 Rab8 37 Rfe1 f5! 38 Qd4 Nh3+ 39 Kg2 f4 and White is crushed. 35 … Rae8 36 Bd4 Qxh5 37 Qf3 Qg6

Threatening 38 … Bxf4. Of course Black is not about to swap queens. 38 Kh1 If 38 Nb6 Bxf4! 39 Nd7 Bxg3 40 Nxf8 Kxf8 41 Qg2 then 41 … Re2!! 42 Qxe2 Bf2+! forces mate: 43 Kh1 Qh6+ 44 Kg2 Qg5+ 45 Kf3 (or 45 Kh1 Qh4+ 46 Kg2 Qg3+ 47 Kh1 Qh3 mate) 45 … Qg3+ 46 Ke4 Qf4 mate. 38 … Re4! 39 Bf2 Rfe8 0-1

Carlsen threatens nothing for the moment, yet White is in zugzwang. For example, 40 Rad1 (or 40 Qg2 Re2 and White is paralysed; 40 Nb6 Re2 is also crushing) 40 … Nxf2+ 41 Qxf2 Re2 42 Rde1 (42 Qf3 Qh6+ mates in two more moves) 42 … Qh5+ 43 Kg1 R8e3! is too awful to contemplate. Game 42 F.Caruana-M.Carlsen World Championship (rapid playoff), London 2018

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Qa4 Bd7 11 Qb4 Qb8

This is Black’s main move, covering d6. Carlsen dispenses with 11 … Bf5, as he played in the previous three games. 12 h4 The idea is to harass Black’s knight. This is all the rage among top players attempting to bust Black’s set-up. 12 Be3 is more restrained. Play can continue 12 … a6 13 Nc3 a5 (Black gains a tempo, at the cost of weakening b5) 14 Qa3 (this way the queen keeps an eye on the c5-break as well as the d6-pawn) 14 … Be7 15 0-0-0! (this is actually logical, since Black has a harder time engineering a … b7-b5 break) 15 … 0-0 16 Kb1 Qc7 17 f3 a4 18 g3 (preventing … Nh4 and … Nf5) 18 … f5 19 Be2 Rfc8, as in L.H.B.Hansen-G.R.Arnold, correspondence 2016. Here the comp gives White an edge, while I think Black is just fine. 12 … h5!?

This is a high-risk proposition from Carlsen. It may be the height of arrogance when a puny IM (your writer) contradicts a reigning world champion, but I will make a suggestion here. The committal 12 … h5 gets me nervous. So I simply dispense with it. In two training games against my 2150-rated student Aaron Householder, I ignored White’s threat to push the h-pawn and opted for 12 … a6. Play continued 13 Nc3 Be7 14 h5 Nh4 15 h6 (a new move and an improvement over 15 Na4?!, which makes no sense as Black gains time with the simple 15 … b5) 15 … g6 16 Ne4 Nf5 17 Bg5 Bxg5 18 Nxg5 b5 19 Rc1 0-0, A.Householder-C.Lakdawala, training game, San Diego 2019. We reached this position twice on our analysis board in our lesson. I already prefer Black, for the following reasons: 1. Black’s king looks safe. 2. White’s h-pawn can turn out to be a weakness. 3. White’s king lacks a safe haven. 13 Be3!? This had only been played before in correspondence chess. As we’ll see, it was repeated (unsuccessfully) by Van Foreest in the next game in this book. 13 Bg5 had been tried previously, the idea being to discourage … Be7, or provoke Black into the weakening … f7f6. After 13 … a6 14 Nc3, B.Galstian-V.Galyamov, Artek 2000, I don’t think Black should be afraid of playing 14 … Be7!. For example, 15 Bxe7 Kxe7 16 Ne4 Bf5 17 Ng3 (not 17 Bd3?? Nf4 when White’s pieces on e4, d4 and the g2-pawn are overloaded) 17 … Bd7 18 Rc1 b6 19 Qd2 (19 Be2?! is met by 19 … Nf4) 19 … Kf8! and Black looks fine to me. One unravelling idea is to play … Rh6 and … Nf4. Note that 20 Be2 is met by 20 … Nf4. 13 … a6 14 Nc3 a5!? Once again, Carlsen is not afraid to weaken b5, if in turn he gains a tempo and gets … a6-a5 for free. In the next game he opts for 14 … f5 instead. 15 Qb3 Instead, 15 Qa3 keeps the pressure up on d6, but Black can reply 15 … f5 (this move halts Ne4 ideas) 16 0-0-0 Be7, when

the computer likes White, while I think it’s misassessing again and that Black is okay. Caruana was perhaps afraid that his queen would be out of play on a3, so he re-routes it to d1, where it attacks the h5-pawn. 15 … a4 16 Qd1 Be7 17 g3

Defending h4, while preventing future knight incursions on f4. Both sides clearly had this position set up in home analysis, since up to this point the moves were rapidly banged out with pre-game preparation speed. Now Carlsen sunk into a long thought. 17 … Qc8 A multi-purpose move: 1. Black prevents Bh3. 2. The fianchetto Bg2 is also prevented, since the c4-pawn would hang.

3. White’s c4-c5 break is made more difficult. 18 Be2 Bg4 19 Rc1 Caruana plays for a c4-c5 break anyway. The downside is that he can no longer castle queenside. 19 … Bxe2 20 Qxe2 Qf5 Now Black can castle short, since h5 is protected. 21 c5?! This thematic break is mistimed. White should castle into it with 21 0-0!, even though it is a psychologically difficult move to make, since it almost begs Black to slide the queen into h3 and sacrifice on h4. For example, 21 … 0-0! (this move, declining to attack, is equally difficult for Black; however, after 21 … Qh3 22 c5 Black’s sacrifices fail; e.g. 22 … Bxh4? 23 Ne4! Be7 24 Qb5+ Qd7 25 c6 bxc6 26 dxc6 Qc7 27 Bb6 and Black is busted) 22 Bd2 with an edge for White. 21 … 0-0! Carlsen’s counterattacking magic is such that it turns his opponent’s weapons into flowers. It requires amazing strategic judgment to understand that White’s coming c5-c6 is actually not a threat, for the following reasons: 1. After c5-c6, Black will exchange, giving himself a passed d-pawn. 2. White’s pawn on c6 is actually vulnerable to … Rfc8, … Bd8! and … Ne7.

3. White is going to have a hard time making something of the c6-pawn, since his h1-rook is out of the battle for an undetermined period of time (it is too risky for White to castle). 22 c6 Consistent. If you say “A” – you have got to say “B”. 22 … bxc6 23 dxc6 Rfc8! 24 Qc4 Bd8!

Threatening 24 … Ne7! and then chopping White’s c-pawn. 25 Nd5 e4! Not a single player in chess history has been unequivocally error-free, yet sometimes, watching the best games of the great players, we do get that impression. So inconceivably strong is Carlsen in simplified positions, that he has reached a point where his endgame power is never to be questioned. His opponents play under the assumption that they will be outplayed.

Some of us tend to forget that Carlsen is also equipped with a Tal chip and, to my mind, no player in the world imitates AlphaZero as closely as he does. In crazy positions, it is no easy task to cut through all the sudden computer-like geometric anomalies, and continual attempts at obfuscation. Here the e5square is opened for Black’s knight, and White begins to feel the pain on the light squares around his king. 26 c7?

Exercise (combination alert): What did Caruana miss with his last move? 26 … Bxc7! Answer: The pawn can be taken with impunity, and it becomes clear that White was cheated blind in the deal. 27 Nxc7 Ne5 White’s central and kingside light squares are horribly vulnerable. 28 Nd5? Caruana’s attempt bounces off Carlsen’s position like bullets off Superman’s chest. The only chance to play on for a while is 28 Qd5. Even then, White is dead lost after 28 … Rab8! 29 Kf1 (29 0-0 Rxc7! wins a piece, since recapture loses the white queen to … Nf3+) 29 … Qf3 30 Kg1 Nd3 31 Rc6 Rxb2 and White is unable to deal with Black’s myriad threats.

28 … Kh7! 0-1 Since 29 Qe2 Nd3+ 30 Kf1 Nxc1 is curtains for White. After heroically fighting Carlsen to a 6-6 tie at the classical time control, Caruana got annihilated in the rapidplay, with a reverse 0-3 hattrick. To his great credit, Caruana has obviously been working on his rapid chess such that, just a month before this book was written, he destroyed Kasparov in a blitz match. I have a feeling Carlsen won’t roll him over so easily next time, if they face off in a rapidplay playoff in the future. Game 43 J.Van Foreest-M.Carlsen Wijk aan Zee 2019 1 e4 c5 2 Nf3 Nc6 3 d4 It takes courage to face down a reigning world champion in his pet line, yet Van Foreest does just that. He knows the Sveshnikov is coming. 3 … cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Qa4 Bd7 11 Qb4 Qb8 12 h4 h5 13 Be3 a6 14 Nc3 f5!?

Theoretical novelty. Van Foreest undoubtedly came prepared for 14 … a5 which Carlsen played against Caruana in the previous game. 15 0-0-0 As we saw in Game 39, castling queenside isn’t very safe for White, especially if Black has access to a … b7-b5 break. Instead: a) GM Ivan Ivanisevic suggested 15 Na4!? without further analysis. Black can continue 15 … f4 16 Bb6 Be7 17 0-0-0 0-0, but I slightly prefer White’s chances. b) 15 g3 is also playable. After 15 … Be7 16 Be2 the comp wants to lash out with 16 … b5!? 17 cxb5 axb5 18 a3 (18 Nxb5?! 0-0 favours Black) 18 … Bd8, when the game is an unclear mess. 15 … Be7 16 g3 This is more to discourage … f5-f4 than actually to protect White’s h-pawn.

16 … 0-0!? When I watched this game live online, I thought: “What about your h5-pawn?” Carlsen is willing to sacrifice it to go after the white king. The Benko Gambit-like line opening 16 … b5! is promising: 17 cxb5 axb5 18 Kb1 0-0 and I like Black’s chances. Of course, with the arrogance of an annotator’s 100% hindsight, it’s easy to make such suggestions, and a lot more difficult to take such a chance over the board. 17 Be2

Attacking the black h-pawn. 17 … e4! 18 Bd4 If 18 Bxh5, Black gets full compensation after 18 … Ne5 19 Be2 b5! 20 cxb5 axb5. 18 … Bf6 Eliminating White’s most active piece.

19 Bxf6 Others: a) 19 c5 Bxd4 20 Qxd4 Ne5 21 Kb1 Rc8 and I like Black’s game. b) 19 Bxh5?! is time consuming. After 19 … Ne5 20 Be2 Nd3+! 21 Rxd3 exd3 22 Bxf6 dxe2 23 Bd4 b5! White is in deep trouble. 19 … Rxf6 20 Qb6! White must not allow Black to play … b7-b5. 20 … Ne5 21 Kb1 Be8!

A dual-purpose move: 1. The bishop protects the loose h5-pawn. 2. The bishop clears d7 for … Nd7, which ejects the white queen’s blockade of b6, making way for a … b7-b5 break. 22 Rd2 Nd7 23 Qd4 Qc7 24 Nd1 Ne5

The knight returns, now that Black managed to seize control over the b6-square. 25 Ne3?! Each side moves in the direction of their prescribed intent, but not at the same pace. This virtually begs Black to play his next move. 25 Rc2 is necessary. 25 … f4!

Another excellent strategic call by Carlsen, who realizes that the f2- and h4-pawn weaknesses in White’s camp override the fact that White gets an open g-file to attack Black’s king. 26 gxf4 Rxf4 Suddenly, White must defend the targets on h4 and f2. 27 Rg1 White attempts to get aggressive – the problem is that his pieces are too clunky to organize an attack down the g-file. 27 … Bg6

Now White must factor in future … Nd3 ideas. 27 … Rxf2 is also possible; for example, after 28 Qxe4 Rf6 29 Rdd1 Bg6 30 Rxg6 Rxg6 31 Bxh5 Rf6 32 Ng4 Nxg4 33 Bxg4 Raf8 34 Be6+ Kh8 White doesn’t have enough for the exchange.

28 Ka1?! White runs from troubles, yet they soon overtake him. It was time to go desperado and sacrifice the exchange for light squares with 28 Rxg6! Nxg6 29 Ng2 Rf6 30 Qxe4 Raf8 31 Bxh5 Ne5 32 f4 Qxc4 33 Rd4 Qf1+ 34 Rd1 Nc4! 35 Qc2 Rxf4! 36 Nxf4 Qxf4 with a clear advantage for Black, but this is still a better chance for White than the game continuation. 28 … Raf8 28 … Rxf2! also wins; but not 28 … Rxh4? which loses material to 29 Ng2!. 29 c5 Desperation.

29 … Rxf2 Or 29 … dxc5 30 Qc3 Qd6. 30 Qc3 If 30 c6 bxc6 31 dxc6 Qxc6 32 Nd5? Carlsen undoubtedly planned the cute shot 32 … Rf1+! (weak back rank) 33 Rd1 (33 Rxf1? Rxf1+ 34 Rd1 Qc1+! is mate next move) 33 … Rxd1+ 34 Rxd1 Kh7 35 Ne7 Qc2 36 Nxg6 Nxg6 37 Bxh5 Nf4 38 Bg4 d5 and White is busted. 30 … Qxc5 White’s game is collapsing. 31 Qxc5 dxc5 32 d6 Kh7! When Carlsen is at his best, as he was in this tournament, it feels as if he is several steps ahead of his opponents. This move sets up the coming combination. 33 d7

Exercise (combination alert): We sense that White’s position is about to know hunger. Find one strong move and White’s position collapses. 33 … Nf3! 0-1 Answer: Double attack/pinned piece. White has been double crossed and loses material, even though he is about to promote. After 34 d8Q we understand the reason for Carlsen’s 32 … Kh7! – White’s promotion doesn’t come with check, so 34 … Rxd8 35 Rxd8 Rxe2! (35 … Nxg1 is also winning) 36 Rc1 Rxe3 wins, as the black e-pawn will eventually cost White a rook. Game 44 K.Leenhouts-D.Hausrath Dutch League 2004 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Bd3

No early queen sorties this time. White simply brings out a piece. The key question is whether d3 is a better square than e2 for the bishop. On d3, White risks tempo loss in the future if Black plays … f7-f5 and … e5-e4. Others: a) 10 Be3 a6 11 Nc3 Be7 12 Bd3 0-0 13 0-0 transposes to the main game. b) 10 Be2 Be7 11 0-0 0-0 12 Be3 a6 13 Nc3 Bg5 14 Bxg5 Qxg5 15 Ne4?! (this move actually loses, rather than gains time, since Black gets … f7-f5 with tempo) 15 … Qd8 16 b4 was N.Nikolovski-Ki.Georgiev, Macedonian Championship, Skopje 2019. Black stands clearly better after 16 … f5 17 Nd2 a5 and if 18 a3 Nf4, then … Qg5 is coming and White is in danger. c) 10 h4 isn’t as effective if it is played without the inclusion of Qa4. Play can continue 10 … a6 11 h5!? (after 11 Nc3 it’s better not to accept the offer of White’s h-pawn, since Black will have a hard time castling with White in charge of an open h-file, so 11 … Be7 12 h5 Nh4 and Black is just fine as it is hard to see what White achieved with the mad pushes of the h-pawn) 11 … Nh4 (taking on b5 is also fine for Black according to Komodo;

11 … axb5 12 hxg6 is slightly in Black’s favour) 12 Nc3 Be7 13 a4 0-0 14 a5 Nf5 15 Bd3 Bg5 and Black stood no worse, A.Rodriguez Vila-A.Shirov, Arica 2018. d) 10 g3 Be7 11 h4 e4! is fine for Black. We learned from Carlsen that Black should not be afraid to push the e-pawn to clear e5 for the g6-knight. e) 10 c5?! a6 11 Nxd6+ Bxd6 12 cxd6 Qxd6, D.JanosevicB.Filipovic, Pula 1988, saw White pick up the bishop pair, but he gave away too much for it, falling behind in development and taking on an isolated d-pawn. Black’s kingside pawn majority will also later roll. 10 … Be7 The immediate 10 … a6 doesn’t alter much. 11 0-0 After 11 c5 a6 12 Nxd6+ Bxd6 13 cxd6 Qxd6 14 0-0, Black can draw (if he wants) with 14 … Qxd5 15 Bxg6 Qxd1 16 Bxf7+ Kxf7 17 Rxd1 Be6. I can’t see either side winning this. 11 … 0-0 12 Be3 a6 13 Nc3 f5 14 f3 14 Qb3, intending c4-c5 tricks, can be met by 14 … e4. G.Fuchs-R.Hanel, Austrian League 2012, continued 15 Be2?! (White should have played 15 Bb6 Qe8, which is about even), when Black could have seized a strong initiative with 15 … b5! and if 16 cxb5 then 16 … f4 17 Bd4 f3 18 Bd1 fxg2 19 Re1 Bg5 with a strong attack against White’s vulnerable king. 14 … Bg5 An exchange of dark-squared bishops would allow Black’s queen to swoop into the kingside. 15 Bf2 15 Qd2 Bxe3+ 16 Qxe3 Nf4 17 Ne2 (17 Rfd1 Qg5 18 Qd2 Bd7 also looks promising for Black) 17 … Nxd3 18 Qxd3 was E.Meduna-D.Brock, Bayerisch Eisenstein 2012. I prefer Black’s chances after 18 … b5!. Black has the superior majority and the bishop looks better than White’s knight. 15 … Nf4 16 c5 Rf6!

The position is now similar to a Classical Dutch. This means the rook can be lifted into the attack along the third rank. 17 b4 If 17 Bc4, V.Rohackova-A.Stojanovic, Mureck 1998, then 17 … dxc5 18 Bxc5 Kh8 Black stands better. 17 … Rh6 Intending … Qe8-h5. 17 … Rg6 is an option too, but White looks okay after 18 Ne2 Nxd3 19 Qxd3 Bd7 20 Rfe1. 18 Bc4?! White doesn’t have time to hang on to his bishop. It was imperative to remove or chase away Black’s powerful knight with 18 Ne2. Then after 18 … Nxd3 19 Qxd3 Bd7 20 Rfe1 Black only stands slightly better. 18 … Qe8 19 Na4? White’s knight is clearly not X-Men material. Yes, Black’s queenside may be falling apart. But it won’t do White a damned bit of good if his king is mated.

19 Kh1 is a better try. After 19 … Qh5 20 Bg1 Nxg2! 21 Kxg2 Bf4! 22 Kf2 Bxh2 23 Ke3 Qh4 24 Be2 Qxb4 White is in trouble, but not as much as in the game. 19 … Qh5 Threatening mate on the move. 20 Bg3

Exercise (combination overpowered. Continue

alert):

White’s

position

Black’s attack. 20 … Nh3+! Answer: Annihilation of defensive barrier. Step 1: Sacrifice the knight on h3. Acceptance is forced.

is

The sacrifice on h3 is stronger than 20 … Nxg2?!, when after 21 f4! Nxf4 22 Qxh5 Nxh5 23 Nb6 Rb8 24 Nxc8 Rxc8 25 Rxf5 White has chances to fight on. 21 gxh3 If 21 Kh1 f4, White can just resign. 21 … Be3+ Step 2: Give a bishop check on e3 and drive White’s king into the corner. 22 Kh1 f4 Step 3: Push to f4, with tempo. 23 Be1 Bxh3 Step 4: Chop on h3 with the bishop. The computer already announces mate (in twelve) for Black, which is no surprise. 24 Qc2 The queen’s situation is the story of the lonely widow who waits by the window for her husband each evening at 6:00 p.m., who will never return home. If 24 Rg1, Black forces mate with the pretty clearance trick 24 … Bf1!. 24 … Rg6! Threatening … Bxf1 and … Rg1 mate. 25 Bc3 Bxf1 26 Rxf1

Exercise (combination alert): I forgot just how many stages of grief there are, but I’m certain this is the final one. Black to play and force mate in two. 26 … Qxf3+! Oh, the glory. Answer: Queen sacrifice/attraction/removal of the guard. 27 Rxf3 Well, why not die rich? 27 … Rg1 mate Game 45 M.A.Brooks-H.Nakamura US Championship, St Louis 2009

I used to play blitz with Nakamura on a daily basis when he was a 13-year-old IM (even though I was the stronger player at the time, Hikaru considered this as an impertinence on my part!). 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c4 Ng6 10 Bd3 Be7 11 0-0 0-0 12 Bxg6!?

With this exchange both sides gain something and lose something. At first glance it may look illogical, since White gives up the bishop pair. Here is the hidden logic behind the move: 1. Black’s main asset in the position is the potential for a kingside attack. 2. Without a knight, Black’s kingside attacking chances are reduced. 3. White has pawns fixed on the light squares d5 and c4, which means his remaining bishop will be a good one, posted on e3, in conjunction with f2-f3.

4. Black’s kingside pawn majority slightly loses its flexibility and ability to push forward, no matter which way Black recaptures. On the debit side of the balance sheet, White gives away the following: 1. The bishop pair. 2. Later on, White may be in deep mourning of the voluntary loss of influence on the light squares. 3. White’s natural plan is to play for an eventual c4-c5. Doing so automatically opens the game, which may end up helping Black, who now has the bishop pair. In conclusion. The capture gives White no advantage, yet it is intended to throw the Sveshnikov player out of our comfort zone. 12 … hxg6 Principle: capture towards the centre. This also keeps Black’s kingside pawn majority fluid. Later Black may even be able to play … Kh7, … Rh8 and … Kg8, transferring his rook to the h-file in order to pursue an attack. 12 … fxg6?! seems illogical, since Black is not going to mate White any time soon, just because the f-file is open. Then 13 Be3 a6 14 Nc3 Bf5 was H.Lopez Martin-J.Grinwis, correspondence 2005. I slightly prefer White after 15 Qe2 b6 16 f3 Bg5 17 Bf2. The comp calls this even, but I don’t see a plan for Black, other than to wait. 13 Be3 f5 Nakamura activates his kingside majority. Instead, 13 … a6 14 Nc3 was seen in L.Bednarski-M.Wisniewski, Polish League 1992. Here Black can try the adventurous 14 … b5! with excellent Benko Gambit-like compensation if White accepts, while on 15 b3 Rb8 16 Rc1 f5, I slightly prefer Black. 14 f3 Preventing … f5-f4-f3 ideas. Others: a) 14 Nxa7?! f4! 15 Nxc8 Qxc8 16 Bd2 Qxc4 17 Qb3 Qxb3 18 axb3 gives Black the superior ending, since White is

essentially a pawn down. b) 14 Bxa7?! f4! 15 a4 (not 15 f3?? Rxa7! 16 Nxa7 Qb6+ 17 Kh1 Qxa7 and White is busted) 15 … f3 16 gxf3 Rf4 gives Black a monster attack for the two pawns sacrificed. 14 … b6 Nakamura wants to slow down White’s queenside pawn majority. Also tempting is 14 … a6 15 Na3 b5!?. 15 f4?!

There are three problems with this move: 1. White has just played f2-f3, so the push to f4 constitutes a waste of a tempo. 2. Black can eventually take on f4 and then activate his fluid kingside pawn majority with … g6-g5. 3. After a future … e5xf4, Black’s dark-squared bishop’s activity is greatly enhanced via a transfer to the a1-h8-diagonal

with … Bf6. Instead, 15 a4 is about even after 15 … a6 16 Nc3 Rb8. 15 … a6 It’s important to play this before … e5xf4, since it chases the knight away from immediate access to the d4-square. 16 Nc3 exf4! 17 Bxf4 Bf6 18 Be3 White wants to play for an eventual break on c5, and also gives himself the option of challenging Black’s dark-squared bishop with Bd4. 18 … Bd7 19 Bd4 Rc8 20 Qd3 Qc7! Black again attacks the c4-pawn, provoking White into b2b3, weakening protection of the c3-knight. 21 Bxf6 Rxf6 This may not be the most accurate way to recapture. Black should play 21 … gxf6! 22 Qg3 Kf7 23 Rac1 Rce8 24 b4 Re5 with a clear edge. 22 b3? This hangs a pawn to Black’s … b6-b5 idea, which exploits the fact that White’s c3-knight is insufficiently protected (preventing c4xb5). A counterattack on b6 with 22 Qd4! was necessary, after which Black only holds a slight edge. 22 … Qc5+ 23 Kh1 b5 24 Rae1? Brooks must have misevaluated, thinking he may have some counterplay for the pawn. He doesn’t. 24 Qg3 was necessary. 24 … bxc4 25 Qg3 Since 25 bxc4 Qxc4 is a hopelessly lost ending for White. 25 … cxb3 26 Ne4

Exercise (critical decision): “Discouragement” is not a word which comes up often in the mind of an optimist. Calculate the ramifications of 26 … fxe4. White plans to play Rxf6 and then come crashing through on g6. Should Black accept, or should he sacrifice the exchange? 26 … fxe4! Answer: Step 1: Accept the first piece and chop e4. Nakamura understands that he is being baited but has analysed deeper. 27 Rxf6! When it comes to swindles, we chess players are devoid of conscience. Has White’s conspiracy born fruit? The answer is

no. There is an absence of a key component in White’s idea: Black is not obliged to accept the sacrifice. 27 … Bf5! Nakamura is a genius in irrational positions that require pure calculation. In such situations it will be him who does the tricking, not his opponents. Step 2: White’s once tidy conclusion falls into disarray. Black doesn’t have to recapture the rook. Obviously Black must avoid 27 … gxf6?? as it’s perpetual check after 28 Qxg6+ Kf8 29 Qxf6+ Kg8 30 Qg6+; but 27 … b2! is also crushing: 28 Qxg6 Qxd5! (preventing both Qf7+ and Rf7 ideas) 29 Ref1 Rb8 30 Rf7 Qe5! 31 Rxd7 b1Q 32 Rd8+! Rxd8 33 Qf7+ Kh7 34 Rxb1 e3 with an easy win for Black. 28 Rxd6 Or 28 Rxg6 Bxg6 29 Qxg6 Qxd5 30 axb3 Qe5! and the e4pawn can’t be touched due to White’s weak back rank, so Black consolidates. 28 … b2! Black’s passed pawn is too deep. 29 Rc6 After 29 h4 e3 the b-pawn promotes. 29 … Rxc6 30 dxc6 Qc1! 31 Qb3+ Kh7 32 Qd1 Qxd1 Simplest. Black promotes, while White lacks the time to do the same. 33 Rxd1

Exercise (critical decision): Calculate the ramifications of 33 … e3. Does it win? Factor in that White, too, owns a deeply passed c-pawn. 33 … e3 Answer: The pawn push works, as the piteously frail white rook is no match against the surging black b- and e-pawns. 34 c7 e2! 0-1 The difficulty level of this exercise cannot be accurately described as inexpressibly hard. In fact, the solution is rather easy. Black’s b-pawn will win the white rook, while his e-pawn will be the one to promote and back rank mate White. After 35 Re1 b1Q 36 Rxb1 Bxb1, the trouble is that 37 c8Q is met by 37 … e1R mate. Game 46

J.G.Gallagher-V.Ivanchuk European Cup, Saint Vincent 2005 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5 8 exd5 Ne7 9 c3

At first sight White’s move may make no sense. After all, if White’s goal is to advance the queenside pawn majority then why not push the pawn two squares, rather than meekly one? The answer is that in this version White isn’t playing the pushmy-majority game. Instead, he has the following ideas: 1. With a white pawn on c4, Black sometimes is allowed undermining tricks with … b7-b5. In this version Black is deprived of this possibility. 2. White seeks a set-up involving a2-a4, Na3, Nc4 and Be3, targeting both d6 and the potentially sensitive b6-square. 3. The move involves a trap if Black gets suckered into … a7-a6?? on the next move. 9 … Nf5

In all the games of this chapter so far, Black’s knight has moved to g6. In this version he is better off with … Nf5. The reasons are: 1. White is discouraged from Be3, so his dark-squared bishop is thus deprived of its natural square. White can eliminate the f5-knight in two ways but both have problems. The plan with Bd3 and Bxf5 costs White the bishop pair and weakens the light squares. Alternatively, White can play g2-g4, but this obviously weakens the kingside. 2. On f5, the d6-square is reinforced and Black no longer worries about Qa4 tricks from White. 3. On f5, Black can even consider playing … g7-g6, fianchettoing and then castling kingside, since the d6-pawn is reinforced. Let’s try not to fall for the trap 9 … a6?? 10 Qa4! with terrible threats on d6 and c7. The knight can’t be captured either. Tactically, Black’s only move would be the rather unpalatable 10 … Kd7 (I count 13 unfortunates in my database who had the bad luck to play Black in this position). 10 a4! With d6 protected 10 Qa4?! makes no sense. White just loses time after 10 … Bd7 11 Qb3 a6 12 Na3 b5 and Black already looks slightly better.

10 … g6 I prefer the fianchetto to the also playable 10 … Be7 11 Bd3 0-0 12 0-0 Nh4 13 f4 a6 14 Na3 exf4 15 Bxf4 Ng6, as in R.Zelcic-M.Illescas Cordoba, European Team Championship, Plovdiv 2003. I slightly prefer White after 16 Be3 Ne5 17 Nc4 Nxc4 18 Bxc4 Bd7. 11 Na3 11 Bd3 Bg7 12 0-0 0-0 13 Qb3 Re8 14 Re1 was L.YudasinV.Kramnik, Candidates match (game 3), Wijk aan Zee 1994. Chances look balanced after 14 … Bd7. 11 … Bg7 12 Bb5+ As played in all four games in my database. White deems it beneficial to trade light-squared bishops, since Black’s plan to attack the white king depletes somewhat with an exchange of pieces. If 12 Nc4 0-0 13 Be2 then 13 … Ne7 and Black begins to push on the kingside. 12 … Bd7 13 Bxd7+ Qxd7 14 0-0 0-0 15 Nc4 Rac8 16 b3 h5 17 Ra2 e4!?

Following the principle: advance pawns on the wing where you have a majority. Of course, this move comes with some risk, since on e4 Black’s e-pawn is artificially isolated. A few years later, in P.Rossi-C.Humeau, Grosseto 2009, Black tried the less committal 17 … Rfe8 18 Qc2 Rc5 19 Rd1 e4 20 Ba3 Rc7 21 Re1 Qd8!, when Black looks slightly better. Now White blundered with 22 Rxe4?? Rxe4 23 Qxe4 Re7 24 Qb1 Qe8 25 Kf1 Bxc3 (threatening mate once again on e1) 26 Nd2 Re2 27 Nf3 Rxa2 28 Qxa2 Qe4 and White can resign. 18 Re1 Rfe8 19 Qc2?! Lethargy is often the precursor of a slow decline. This is inaccurate, since it shuts the a2-rook out. White should temporarily sacrifice a pawn to generate play with 19 Rae2! Bxc3 20 Bd2 Bf6 21 Rxe4 Rxe4 22 Rxe4 b5, though I slightly prefer Black’s chances. 19 … Rc5! 20 Rxe4 Rxd5 21 f3?! We are all forced to perform penance in the future for our lack of understanding in the present. White minimizes his disadvantage with 21 Bd2! Rxe4 22 Qxe4 Rc5. 21 … Rc8! Black’s pressure against c3 mounts. 22 Ba3 b5! 23 axb5 Rxb5 Threatening … d6-d5. 24 Re1 d5 25 Ne5 Qc7! With a double attack against e5 and c3. 26 f4 Qb6+! Even stronger than 26 … Qxc3. 27 Qf2 Rxb3 To the criminal mind, theft is only reprehensible when the perpetrator is caught. 28 Qxb6 axb6 29 Bb2 This sad-looking move is forced. After 29 Bb4? d4 Black wins.

29 … Nd6 30 Re2 Bxe5 31 fxe5 Nc4 The b2-bishop is displaced further. Black is a pawn up with dominant pieces and pressure against White’s c3-pawn and e5pawns. The game is over. 32 Kf2 32 Bc1 saves the e5-pawn, at the cost of the c3-pawn after 32 … Rxc3. 32 … Re8 33 Bc1 Nxe5 Ivanchuk picks up pawn number two. 34 Rad2 Rb1 35 Ba3 Re6 36 Kg3 36 Rxd5?? loses a rook to 36 … Ng4+ 37 Kf3 Rf1+ 38 Kg3 Rxe2. 36 … Nc4! 37 Rxe6 fxe6 38 Ra2 Nxa3 You may be wondering why such a dominant knight is willing to trade itself for an inferior bishop. Ivanchuk decides that simplification to a pure rook ending is the easiest win for Black. 39 Rxa3 Kf7 0-1

Chapter Five White’s Sixth Move Alternatives The Sveshnikov, much like social media, has a combative culture. Not in this chapter, though. In this chapter we examine a series of not-Sveshnikovs, where White attempts to dodge the main line 6 Nbd5 with some not-so-scary alternatives. The good news is that these sidelines are pretty much worthless for White who, in general, will be fighting for equality. To my mind White embraces a heretical doctrine when dodging the fight with any move other than 6 Ndb5. White has five secondary, yet playable knight moves – 6 Nb3, 6 Nf3, 6 Nxc6, 6 Nde2, and 6 Nf5 – and we will look at each of them, in that order. Game 47 O.Hadid-L.A.Liascovich Pereyra Puebla 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 This is a Kalashnikov move order. Our Sveshnikov move order runs 4 … Nf6 5 Nc3 e5 6 Nb3?!, transposing below. 5 Nb3 Nf6 6 Nc3 There is no opening system designed to generate a theoryfree zone, but is this really the best a player can do when given the white pieces? Okay, White dodges sharp theory, but at what cost? White is already fighting for equality. After the passive knight retreat to b3, we instantly seize a minor initiative.

6 … Bb4 We pin White’s c3-knight, while worrying our opponent about the e4-pawn as well as … d7-d5 freeing breaks. 7 Bd3 Others: a) 7 Bg5?! h6 8 Bxf6 (otherwise White loses a pawn to … g7-g5) 8 … Bxc3+ 9 bxc3 Qxf6 10 Bc4 d6 11 Qd3 0-0 12 0-0-0 Rd8 13 Kb1 Be6. The d6-pawn isn’t even weak and White had zero compensation for his wrecked structure, D.MajigsurenE.Sevillano, Reno 2004. b) 7 Bc4?! 0-0! (7 … Nxe4 allows 8 Bxf7+ Kxf7 9 Qd5+, regaining the piece; Black is okay here too, but why voluntarily give up castling privileges?) 8 0-0 Bxc3 9 bxc3 Nxe4 10 Ba3 d6 11 Qf3 Nf6 12 h3 (in order to prevent … Bg4) 12 … Be6 and White found himself a pawn down with a broken structure, G.Thorhallsson-A.Shirov, Reykjavik 2017. c) 7 f3?! defends e4 without even developing a piece. After 7 … d5 8 exd5 Nxd5 9 Bd2 Nxc3 10 bxc3 Bd6! Black halts f2-

f4 ideas and stands clearly better. d) 7 Bd2 can be met, if you seek to complicate, with 7 … Bxc3! (the safe line is 7 … 0-0 8 Bd3 d5 9 exd5 Nxd5 10 Nxd5 Qxd5 11 0-0 e4 12 Bxb4 Nxb4 13 Be2 Qg5 14 Qc1 Qg6 and Black already stands slightly better, A.Cilloniz RazzetoC.Martin Sanchez, correspondence 2012) 8 Bxc3 0-0! (8 … Nxe4 9 Qg4 is messy and equal) 9 Bd3 d5 10 exd5 Nxd5, and if White attempts to preserve the bishop pair with 11 Bd2 e4 12 Be2 then 12 … Qf6 sees b2 attacked and White in deep trouble. He lacks the time for 13 c3? as 13 … Qg6 14 Kf1 Rd8 leads to a wretched game for White. 7 … d5 We easily achieve our freeing break and it is White who fights for equality. 8 exd5 Nxd5 9 Bd2 Nxc3 Black inflicts damage to White’s structure, free of charge. 10 bxc3 White gains a tempo, which is not enough to justify the structural damage. 10 … Bd6!

The d6-square is the correct spot for the bishop. The idea is that if and when White plays f2-f4, Black will chop on f4, exchanging dark-squared bishops and thus reducing White’s kingside attacking possibilities. 11 0-0 0-0 12 Re1 Others: a) 12 Qh5 f5 13 Bc4+ Kh8 14 Rad1 Qc7 15 Nc1?! (15 Rfe1 is better but even here White’s game stinks) 15 … b6 16 Bb3 Bb7 17 Ne2?! f4! 18 f3 Rae8 19 c4 was J.Devecchi- G.Sanchez Rojo, Tournefeuille 2008. Black already has a winning position, which can be converted after 19 … Bc5+ 20 Kh1 Re6 21 Qh4 Rh6 22 Qe1 Qf7 (threat: … Rxh2+ and … Qh5 mate) 23 h3 Bc8! (targeting h3) 24 Ng1 Qh5 (threat: … Bxg1 and … Bxh3) and Black has a decisive attack. b) 12 f4 f5! 13 fxe5 Bxe5 14 Qf3 Be6 15 Rae1 Bd5 16 Qh3 g6 17 c4 Bf7 18 Kh1 Qd6 19 Be3 Rfe8 20 Bg1 Bg7 21 Rd1 Qc7 22 Bf2 was I.Brkic-A.Shariyazdanov, Zadar 1999, where 22 …

Ne5 gives Black a strategically won position. This game is covered in Sveshnikov: Move by Move. 12 … Be6 13 Rb1 f5?! This move is an inaccuracy which should reduce Black’s advantage. 13 … Qc7! 14 Qe2 f5 is the correct move order. Now if White plays 15 Nd4?? Nxd4 16 cxd4 e4 17 c4 Rae8, he loses material since the bishop is loose, as will be the c4- and h2pawns. 14 Be3?! White misses his chance for 14 Nd4! Nxd4 15 cxd4 e4 16 c4 b6 17 Bf1 Qc7, when he stands only slightly worse. 14 … Qe7 15 Qh5 Our pretend threats sometimes feed on the opponent’s paranoia. Not here though. No attack will arise from this queen sally, since Black is the one who controls the centre. On the very next move White pulls an about-face on the issue. 15 … g6 16 Qg5!?

Principle: One sure sign that our position is declining is that we find ourselves “agreeing” to one concession after another. White, whose “attack” is spent before it even begins, faces a painful impasse. 1. Fear of a direct attack on his king. 2. Offering to enter an ending where he nurses three isolanis, two of them doubled and on an open c-file. The cost of entry into either option is prohibitive. 16 Qe2 e4 17 Bc4 Rf7 is also miserable for White. 16 … Qxg5 17 Bxg5 e4 18 Bb5 Be5! Targeting c3. 19 Bxc6?! White makes a major concession. He gives up the bishop pair to seal the c-file from Black’s rooks. He was probably better off with 19 c4 Nd4 20 Nxd4 Bxd4 21 Red1 Be5 22 Bd7 Bxc4 23 Rxb7 Rfb8 24 Rxb8+ Rxb8 25 a4 a6. 19 … bxc6 20 Nd4 This is the only way to defend c3. 20 … Bxa2 Black creates a nasty passed a-pawn. 21 Rb7 White puts up more resistance with 21 Ra1 Bd5 22 Ra5, although even here his position is way past any attempt to repair the damage. 21 … Rf7 22 Rxf7 Bxf7 23 Nxc6 Bxc3 24 Rd1 a5

What do we do when every logical try is out of the question? When we don’t have much to begin with and then lose even that, we derive a bitter satisfaction from the irony. White is dead lost. 1. He is a pawn down. 2. Black’s bishops rule the board. 3. Black’s a-pawn, backed up by a rook, is ready to push its way down to the promotion square. 25 Bc1 Ra6 26 Ne7+ White’s rootless knight floats, devoid of stable posts. How long can it persevere in the desert, alone and without provisions? a) 26 Nd4 Rd6 27 Be3 f4 wins. b) 26 Nb8 Rb6 27 Bf4 a4 28 Nd7 Rb5 29 Bd6 Bb2 30 Rb1 Rd5! wins a piece due to White’s weak back rank. 26 … Kf8 27 Rd7

27 Ba3? fails to 27 … Bb4 28 Bxb4 axb4 29 Nc8 (or 29 Nd5 Bxd5 and White is unable to recapture due to his weak back rank) 29 … Rc6 30 Rd8+ Kg7 31 g3 Rxc2 32 Nd6 Be6 33 Rb8 b3 34 Kf1 b2 (threat: … Rc1+ and … b2-b1Q) 35 Kg2 e3 and wins. 27 … Rb6! There is no defence to the coming … Rb1. 28 g3 Rb1 29 Rd1

Exercise (combination alert): What is Black’s simple win here? 29 … Rxc1! 0-1 Answer: Simplification. White is crushed after 30 Nxg6+ hxg6 31 Rxc1 a4, when the a-pawn costs him his rook.

But not 29 … Kxe7?? which falls for White’s trap after 30 Bg5+ (discovered attack) 30 … Ke6 31 Rxb1 and Black suddenly finds himself the exchange down and fighting for a draw. Game 48 J.Molinero Martin-M.Kuijf Sitges 2003 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 Dammit people, stop messing with my book. Once again Black plays a Kalashnikov move order. Ours runs 4 … Nf6 5 Nc3 e5 6 Nf3, transposing. 5 Nf3 Nf6 6 Nc3 This time White’s knight has retreated to f3, which isn’t much of an improvement over the previous game’s b3, since on f3 it also inhibits the natural f2-f4 push. 6 … Bb4

Threatening … Nxe4. White scores an insipid 30% from this position, so this isn’t exactly a time for us to quake in fear. 7 Bd3 Alternatives are: a) 7 Bd2 Bxc3 8 Bxc3 Nxe4 9 Nxe5 (9 Bxe5?? Nxe5 10 Nxe5 Qa5+ 11 c3 Qxe5 wins a piece) 9 … Nxc3 10 Nxc6 Nxd1 11 Nxd8 Nxb2 12 Nxb7 Bxb7 13 Rb1 Bxg2 14 Bxg2 Rb8 15 00 0-0 and Black was a pawn up, N.Erdem-G.Abramov, correspondence 2013. b) 7 Bc4 0-0 (7 … Nxe4 8 Qd5 Nd6 9 Bb3 leaves Black a bit tangled, so White gets full compensation for the pawn) 8 0-0 Bxc3 9 bxc3 Nxe4 10 Ba3 d6 11 Re1 Ng5! 12 Nxg5 Qxg5 13 Bxd6 Rd8 14 Qf3 Qf6 15 Qxf6 gxf6 16 Rad1 Bf5 and I slightly prefer Black’s superior structure to White’s bishop pair, A.Duffy-R.Dzenis, correspondence 2011. c) 7 Bg5 Bxc3+ 8 bxc3 h6 9 Bxf6 (9 Bh4 Qa5! favours Black) 9 … Qxf6 and White had no compensation for the broken queenside. 7 … d5 Once again we effortlessly achieve our thematic freeing break. 8 exd5 Nxd5 This is consistent with the previous game. A case can be made for 8 … Bxc3+ 9 bxc3 Qxd5 10 0-0 0-0 11 c4 Qa5 12 Bd2 Qc7, R.Svrsek-F.Büttner, correspondence 2012. Again I much prefer Black’s structure and centre over White’s bishop pair. 9 Bd2 9 0-0 Bxc3! avoids loss of time and is safer than pawn grabbing with the knight. After 10 bxc3 0-0 11 c4 Nde7 12 a4 Bf5 White didn’t get enough for the break-up of his queenside, V.Florea-M.Zlatariu, correspondence 2008. 9 … Nxc3 10 bxc3 Be7! In the previous game the bishop went to d6, in order to discourage White’s f2-f4 thrust. Here White’s knight is stationed on f3, which means that f2-f4 is not so easy to play. So the

bishop is probably better posted on e7, where it prevents Ng5 jumps. 11 0-0 0-0 12 Rb1 After 12 Re1 f6 13 Be4, B.Lusin-C.Amura, San Fernando 1993, and now 13 … Be6, I don’t see White’s compensation for his lack of centre and damaged pawns. 12 … f5 13 Bb5 Threatening Bxc6 and Nxe5. The computer prefers 13 Bc4+ Kh8 14 Bc1 Qc7, though Black looks clearly better here as well. 13 … Bf6 14 Be3 Qe7! Preventing Bc5 ideas. 15 Qd5+?! This is a loss of time. 15 Qe2 Be6 is unappetizing, yet White should play this way. 15 … Be6 16 Qc5 f4?! Black missed a stronger continuation in 16 … Qxc5! 17 Bxc5 Rfc8 18 a4 a6 19 Be2 Nd8! 20 Bd6 Nf7 which is a lost ending for White. 17 Bxc6?!

Better is 17 Qxe7! Nxe7 18 Bc5 Rfc8 19 Bxe7 Bxe7 20 Bd3! (stronger than 20 Nxe5?! Bxa2) 20 … Rc7 21 a4 and White only stands slightly worse. 17 … bxc6 18 Qxe7 Bxe7 19 Bc1 e4 19 … Bf6! is slightly more accurate. 20 Re1 Bxa2 21 Rb7 Bd6 Again 21 … Bf6! 22 Rxe4 Rfd8 23 Bxf4 Rd1+ 24 Re1 Rad8 25 Rb8 Rxe1+ 26 Nxe1 Rxb8 27 Bxb8 Bxc3 28 Kf1 a5 and White won’t save the ending due to Black’s outside passed apawn, coupled with the bishop pair. 22 Rxe4 Bd5?! Here Black should play 22 … Rfb8 23 Rd7 Bd5 24 Re1 Rd8 and if 25 Rb7?! then 25 … a5! and White is busted. 23 Ra4! Rfb8 24 Rxb8+ Rxb8 25 Nd2 Covering b1. Black has misplayed the ending and now only has a slight plus.

25 … Be5 26 c4 Bf7 27 Rxa7 White should hold this. 27 … Bc3 28 h3 Bg6 29 Kh2? White equalizes with 29 Ra3 Bf6 30 Nf1 Be5 31 Nd2!, as Black must avoid 31 … Bxc2? 32 Nf3 and if 32 … Bd6?! 33 c5! Bc7 34 Ra7, Black is the one fighting for the draw.

Exercise (combination alert): What did White overlook on his last move? 29 … Bd4! Answer: Double attack. When f2 falls, White’s king is boxed in and endangered. 30 Ra6 Bxf2 31 Rxc6 Bxc2 32 c5 Be3?! Black misses the winning plan 32 … Re8! (threatening … Re1 and … Bg3 mate) 33 Nf1 g5 34 Bd2 Bd3 and White can

resign. 33 Rb6 Re8!? Presumably Kuijf didn’t believe in his winning chances in the technical ending 33 … Rxb6 34 cxb6 Bxb6. 34 Rb2 Bd1 35 c6 Rc8

Exercise (combination alert): The c6-pawn is about to fall. White can still hold a draw if he finds a startling defensive idea found by the computer. Take your time on this one, since the defence resembles an endgame study, more than an actual game. What should White play here? 36 Nb3?

The human move loses. Answer: White holds the draw if he finds the deeply hidden computer shot 36 Nc4!! Bxc1 37 Rb1! (the miracle arrives – White’s lost piece is regained with this eerily calm move; this idea applies the same principle as a person who turns off a light in an unoccupied room to conserve energy) 37 … Kf8 (not 37 … Rxc6? 38 Rb8+ Kf7 39 Ne5+ Kf6 40 Nxc6 and it is White who has all the winning chances) 38 Ne5! Bd2 39 Rxd1 Bc3 40 Nd3 Rxc6 41 Nxf4 Be5 42 g3 Bxf4 43 gxf4 and the rook ending is drawn. 36 … Rxc6 37 Bxe3 fxe3 Now White is in deep trouble once again. 38 Nd4 Rc8 39 Ne2 I would play 39 Kg3!, based on the principle: Activate your king in an ending. 39 … h6 40 Kg1 Rf8?! Preventing Kf1 is logical, but the comp doesn’t like this move and prefers 40 … Kh7 41 Kf1 Bc2. 41 Nc3 Re8! 42 Kf1 White has drawing chances. Not 42 Nxd1?? e2 and White loses the exchange. 42 … Bh5 43 g4 Bf7 44 Rb4! Preventing … Bc4+. 44 … Rc8 45 Ne2?! After 45 Nd1! Bc4+ 46 Ke1 e2 47 Nc3 Ba6 48 Nxe2 Re8 49 Rb2 Re3 50 Kf2 Rxh3 51 Nf4 Rh2+ 52 Ng2 Rh1 53 Kg3 White should hold the draw. 45 … Ra8?! Both sides are sucked into the morass of the position’s subtleties. 45 … Bd5! gives Black a winning position. 46 Nd4 g6?! This gives White more chances to hold the game. Black should have tried 46 … Ra2 47 Nf5 Rh2 48 Nxe3 Rxh3.

47 Ke2 Ra3 48 h4! Rc3

Exercise (critical decision): Should White do nothing but give a rook check on b8, or should he get proactive with 49 g5 - ? In a situation of absolutes, our position cannot be bartered with. How can White achieve his goal without impediment? 49 g5? Missing our opponent’s combination is as annoying as forgetting our password when we desperately need it. This move overlooks a tactical idea from Black’s side. The preliminary 49 Rb8+! Kg7 50 Rb7 Kf6 is correct, and only then 51 g5+! (the correct timing for the push) 51 … hxg5

52 hxg5+ Kg7 (or 52 … Ke5 53 Nf3+ Ke6 54 Nd4+, drawing) 53 Ne6+ Kg8 54 Rc7! Ra3 55 Rc8+ Kh7 56 Rc7 Kg8 57 Rc8+ with a draw by repetition. 49 … hxg5 50 hxg5 0-1 I assume that White’s flag must have dropped here. After 50 … Bc4+ 51 Kf3 (not 51 Ke1?? Rc1 mate, or 51 Kd1?? Rd3+ 52 Ke2 Rxd4+ 53 Kxe3 Rg4 and wins) 51 … e2+ 52 Kf2 Rc1! 53 Nxe2 Rc2 54 Kf3 Black has cleverly won the exchange, but the reduced material means that White still has excellent chances to hold. Indeed, the endgame databases declare the position to be a theoretical draw. Game 49 Har.Schmitt-M.Pirs Correspondence 2013 1 Nf3 c5 2 e4 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Nxc6

You can’t see me now but I’m wagging my finger in admonition at White, in preparation for a “you-should-know-

better” sermon. We arrive at our third, completely illogical theoretical sidestep. White chops on c6, automatically giving Black potential central control, as well as an open b-file. 6 … bxc6 Principle: Pawn capture toward the centre. If you are a connoisseur of boredom, like me and GM Ulf Andersson, then opt for the dull, technical ending with 6 … dxc6!? 7 Qxd8+ Kxd8. This position, a powerful sedative, is equal and perfect for those who love instant endings. 7 Bg5 Others: a) 7 Bc4 Bb4 8 0-0 0-0 (8 … Bxc3 9 bxc3 Nxe4 10 Qg4 0-0 11 Qxe4 d5 12 Qxe5 Re8 13 Qg3 dxc4 is equal) 9 Bg5 h6 10 Bh4 d6 11 Qe2 Qe7 12 a3 (this is a waste of time since Black plans to chop the knight anyway; 12 Rfd1 is more useful but even there after 12 … Bxc3 13 bxc3 g5 14 Bg3 I slightly prefer Black) 12 … Bxc3 13 bxc3 Be6 14 Bd3 g5 15 Bg3 and I prefer Black’s structure to White’s not-so-scary bishops, R.SherfN.Birnboim, Netanya 1987. b) 7 Bd3 doesn’t try to halt … d7-d5 and instead reinforces e4. After 7 … Bb4 8 Bd2 0-0 9 0-0 d5 10 Nxd5 Nxd5 11 exd5 Bxd2 12 Qxd2 cxd5, Black stands slightly better due to the better centre, S.Mielke-K.Venske, German League 2005. 7 … Rb8 Black immediately utilizes the open file. 8 Rb1 Qa5 Creating threats of … Rxb2, followed by … Qxc3+ and then … Qxb2 and … Nxe4, exploiting the pin on the c3-knight. Alternatively, 8 … Bb4 9 Bd3 h6 10 Bd2 (after 10 Bh4 d6 11 0-0 g5 12 Bg3 Bxc3 13 bxc3 Be6 I prefer Black) 10 … 0-0 11 0-0 d5 also looks better for Black, again due to the greater central influence, J.Soriano Folch-L.Gomez Jurado, Mislata 2005. 9 Bd2 Qc7 10 Bd3 Be7 There is nothing wrong with the immediate 10 … d5.

11 0-0 0-0 12 f4 This a slight improvement on 12 Bc4?, M.Kocic-J.Kosa, Subotica 2007, when Black stands clearly better after 12 … Rd8. However, pushing the f-pawn is not that great either, since after Black’s reply neither f4xe5 nor f4-f5 has any potency. White is better off going passive. 12 … Bd6! With the strategic threat of … e5xf4. This artificial yet strong move induces White into clogging with f4-f5, allowing Black a coming pawn break on d5. Black has another decent option in 12 … exf4! (principle: counter in the centre when attacked on the wing) 13 Bxf4 Bd6 14 Bxd6 Qxd6 15 Qe2 Re8 and Black controls e5, which greatly reduces White’s attacking options. 13 f5 After 13 Qf3 exf4 14 Bxf4 Bxf4 15 Qxf4 d6! 16 Qh4 Qb6+ 17 Kh1 Qd4! Black disallows both exchange sacrifices on f6 and also e4-e5 pushes. 13 … Be7 Clearing the way for … d6-d5. 14 g4!?

Why are the seven deadly sins so much fun? It was Saint Augustine who prayed: “Make me chaste, oh Lord, but not yet!” White, refusing to grieve over his past or fret about the future, decides he will not go down passively, so he launches a kingside pawn storm. The danger of course is future overextension. 14 … d5 15 g5 Nxe4 16 Nxe4 dxe4 17 Bxe4 Qd8! A double attack. Black threatens … Qd4+, followed by … Qxe4, and also … Bxg5. 18 Bc3? White presses on with his own agenda but in the wrong way. Others: a) 18 f6?? loses material to 18 … Qd4+ 19 Rf2 Bc5!. b) After 18 Kh1! (principle: surrender is actually a superior path to persistence if you were on the wrong track to begin with) White may hold the draw; e.g. 18 … Bxg5 19 Bxg5 Qxg5 20 Qd6 Qh4! 21 Rbe1 Rxb2 22 Qxe5 Qf6 23 Qxf6 gxf6 24 Rg1+

Kh8 25 a3 c5 and Black won’t have an easy time converting his extra pawn, which is doubled, isolated and blockaded. In fact every pawn on the board is an isolani! 18 … Bc5+ 19 Kh1 Qxg5 20 Bxe5?! White should tone down his ambitions and play 20 Qf3.

Exercise (combination alert): The players exist in a state of mutual intolerance. Black has access to an amazing combination here. If you mistime the coming sequence by even a split second, it will fail. How should Black continue? 20 … Qe7! Answer: Step 1: Sacrifice the exchange, after which White’s king becomes seriously exposed on the light squares.

21 Bxb8 Qxe4+ 22 Qf3 Bxf5! Step 2: Exploit the pin of White’s queen by chopping f5. 23 Rbe1 23 Bg3 Qxf3+ 24 Rxf3 Be4 25 Kg2 Rd8 is a won ending for Black. 23 … Re8!!

Step 3: Zwischenzug/sham queen sacrifice. The rook shift to e8 is an act of generosity which is mostly symbolic, since Black immediately regains the investment, plus interest. This move is proof that hard variations function as the element that tests our abstract theory against reality. 24 Rxe4 Bxe4 25 Qxe4 Rxe4 Black has emerged a pawn ahead with the more active pieces. Also, his 3 to 1 majority on the kingside is fast, whereas White experiences almost impossible obstructions to create even a single passed pawn on the queenside.

26 Rd1 f5! A dual-purpose move. 1. Black creates luft for his king. 2. Principle: Passed pawns should be pushed. 27 Rd8+ Kf7 28 Rd7+ Ke6!? Also strong is the simple, drama-free move 28 … Re7 29 Rd2 g5 with a winning position for Black. 29 Rd2 Not 29 Rxg7??, which hangs the rook to 29 … Re1+ 30 Kg2 Rg1+. 29 … f4 30 Kg2 Kf5 31 h3 g5 32 a3 Re3 33 b4 The comp claims this move is a mistake in an already lost position. But is it a mistake if everything else loses? For example, 33 Rd7 Rg3+ 34 Kh2 Bg1+ 35 Kh1 h5 and White’s counterplay is way too slow. 33 … Rg3+ 34 Kf1 White’s king lives at the bottom of a well and is in grave danger, despite the reduced material on the board. 34 Kh2 Bg1+ 35 Kh1 Be3 is also an easy win for Black. 34 … Be3 35 Rd7 Ke4!

Threatening … Kf3 and … Rg1 mate. 36 Re7+ Kf3 37 Rxe3+ Black is obviously faster in the promotion race. 37 … Kxe3 This is winning, but 37 … fxe3! 38 Bxg3 Kxg3 was more efficient, with an easily won king and pawn ending for Black; e.g. 39 c4 Kxh3 40 b5 c5 41 a4 g4 42 a5 g3 43 b6 axb6 44 axb6 g2+ 45 Kg1 e2 46 Kf2 g1Q+ 47 Kxg1 e1Q mate. 38 Bxa7+ Ke4 39 a4 Rxh3 40 b5 cxb5 41 axb5 Ra3 42 b6 Ra1+ 43 Ke2 f3+ 44 Kf2 Rb1 Oh no you don’t. 45 c4 g4 46 c5 Rb2+ 0-1

Black forces mate after 47 Kg3 Rg2+ 48 Kh4 f2; e.g. 49 b7 f1Q 50 Bb8 Qf6+ 51 Kh5 Qg6+ 52 Kh4 Qh6 mate. Game 50 P.Nikac-D.Komarov Niksic 2000 1 e4 c5 2 Nf3 Nc6 3 Nc3 Nf6 4 d4 cxd4 5 Nxd4 e5 6 Nde2

This clunky retreat is yet another attempt to dodge main line theory. Here are the pluses and minuses of the move. The pluses: 1. White no longer fears … Bb4 and … Bxc3+, since the c3knight is covered its colleague on e2. 2. White’s knight may transfer to g3, where it may later hop into f5. This could induce the weakening … g7-g6 from Black. 3. White can play Nh5, exchanging off Black’s f6-knight, a key defender of d5. The minuses: 1. The e2-knight blocks the natural development of the f1bishop. 2. The f2-square now becomes difficult to guard, as shown in the game continuation. 3. The principle: Bring your pieces out early in the opening of the game, is not one that requires the wisdom of hindsight. White’s last move does indeed develop, but to a lousy square.

The move incurs a loss of time since the e2-knight must be moved twice in the opening. In conclusion the move contains more drawbacks than benefits. 6 … Bc5! Targeting f2. 7 Ng3 White must guard against … Ng4. Note that 7 h3?? is terrible. My wife was talking to me as I wrote this, except I didn’t hear a word of what she said, since I was so distracted by White’s last move, which totally ignored Black’s threat. I have found through experience that those who “surprise” us with little played sidelines, often don’t know the theory. With the simple double attack 7 … Qb6, Black simultaneously wins a pawn while destroying White’s structure. A.Silvestre-C.Lakdawala, San Diego (rapid) 2017, continued 8 Be3 Bxe3 9 fxe3 0-0! (principle: the threat is stronger than its execution; this move is an improvement over both 9 … Qxe3 and 9 … Qxb2) 10 Qd2 (White can save both pawns with 10 Qc1 but then 10 … Rd8!, intending … d7-d5, leaves him fatally behind in development) 10 … Qxb2 11 Rb1 Qa3 12 Rb3 Qc5 13 g4!? d6 14 Ng3 Be6! (the pawn is returned to seize the initiative) 15 Rxb7 Rab8 16 Rxb8 Rxb8 17 Bg2? (this fatally weakens c4; he had to try 17 Kf2) 17 … Na5 18 00?! Nc4 19 Qe1 Nxe3 20 Rf2 Nexg4! 21 hxg4 Nxg4 22 Bf3? (I analysed 22 Nh1 Nxf2 23 Nxf2 Rc8 24 Nd5 Bxd5 25 exd5 f5! and White has no chance to save the game) 22 … Nxf2 23 Qxf2 Qxc3 24 Qxa7 Qc5+ (I knew 24 … Rb1+ was stronger, but in rapid chess I always pick the simplest win if it’s 100% certain) 25 Qxc5 dxc5 0-1. 7 … d6 Threatening … Ng4. Instead 7 … Qb6!? is met by 8 Qd2 Ng4 9 Nd1, when White will regain the lost time. 8 Be2 Nd4!?

The GM is looking for some adventure. I would play 8 … a6 which secures Black’s dark-squared bishop from Na4 ideas. After 9 0-0 h6 10 Nh5 Nxh5 11 Bxh5 0-0 12 Nd5 Qh4 13 Bf3 Nd4 14 Ne3 b5 15 c3 Nxf3+ 16 Qxf3 Bb7 17 Nd5 Bxd5 18 exd5 Rac8 19 g3 Qe7 20 a4 f5, Black had the superior pawn majority and used it to generate a kingside attack, A.Rodriguez Vila-L.Van Wely, Minneapolis 2005. This game is annotated in Sveshnikov: Move by Move. 9 Bg5 Ne6 Attacking the g5-bishop with tempo. 10 Bb5+ Kf8! Black wants to retain his good light-squared bishop. Loss of castling won’t bother Black at all, since he can play … h7-h6, … g7-g6 and … Kg7, castling by hand.

11 Bxf6 White is willing to hand over dark squares to secure his grip on d5. The game is dynamically balanced. 11 … Qxf6 12 0-0 Nf4 13 Nd5 Nxd5 14 Qxd5 g6 Black’s king will be well hidden on g7. 15 Ne2 White seeks to open the f-file with a future f2-f4. 15 … Kg7 16 Kh1 Be6! Well calculated. Komarov sees that he doesn’t need to waste a tempo on … Rb8. 17 Qd2 After 17 Qxb7?? a6 18 Bd3 Rhc8! there is no answer to the coming … Ra7, trapping White’s queen. 17 … Rhd8 18 Nc3?

White gets distracted. He should have stuck to his plan with 18 f4. Even then, I prefer Black after 18 … d5!. 18 … Bd4! If White’s knight moves, then b2 hangs. 19 Bd3 If 19 Rab1 then 19 … Bxc3! 20 Qxc3 Bxa2 steals a clear pawn, which Black should convert. 19 … Rac8 20 Nb5 Bb6! I prefer keeping it simple, over the greedy 20 … Bxb2!? 21 Rab1 d5, although even here White experiences difficulties. 21 Nc3 21 a4 a6 22 Na3 d5! is awful for White. 21 … Ba5! Just as in some of the earlier games of the chapter, White is destined to sustain significant structural damage on the queenside. 22 Rab1 Bxc3 He won’t give White time for b2-b4. 23 bxc3 b6 White is busted since he has zero compensation for his awful queenside. 24 a4 d5! Black eliminates his single weakness of the backward dpawn, while threatening … d5xe4. 25 exd5 Bxd5 I would probably recapture with the rook. 26 f3 Bc6 27 Ra1 After 27 a5 bxa5 28 Qe3 Rc7 29 Rfe1 Re8 Black’s a5-pawn may be doubled, but it’s still extra and passed. 27 … Rd5 28 c4 Ra5!

Black’s contempt for White’s attempts to counterattack is clear. This move denies White any a4-a5 liquidation ideas, and White’s counterplay level falls to absolute zero. 29 Qb4 Rd8 30 Be4 Rd4 GM Komarov doesn’t want to take on e4 and open the f-file for White. 31 c3 Rd2 32 Rfd1 Desperation. This loses the a4-pawn, but I don’t have any great suggestions for White. 32 Bxc6 Qxc6 is just slow death as the c4-pawn will eventually be picked off. 32 … Rxd1+ 33 Rxd1 Rxa4 34 Qb3 Ra5 35 Bd5?

Principle: no piece in our position can act as if it has its own life to lead, with no connection to the remainder of our pieces. Exercise (combination understand: garbage in,

alert):

As

all

programmers

garbage out. This one is easy. White just blundered in an already losing position. Black to play and win. 35 … Ba4! 0-1 Answer: Bishop skewer. Game 51 R.Zelcic-Iv.Saric Stari Mikanovci (rapid) 2010 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Nf5

This is the soundest of all of White’s sixth move dodges. Nevertheless, Black easily equalizes in this one too. 6 … d5! Black takes advantage of the wobbly footing of the f5-knight to achieve the key Sveshnikov freeing break.

7 exd5 Others: a) 7 Ng3? d4 8 Nce2 h5! (threatening … h5-h4, followed by … g7-g6, trapping White’s knight) 9 h4. From here, out of seven games from the database, I see seven wins for Black, which is a pretty good indicator that White’s position sucks. b) 7 Nxd5? Nxe4 and White’s position is a mess. c) 7 Bg5? Bxf5 (7 … d4! is even stronger) 8 Bxf6 gxf6 9 exf5 Bb4 10 a3?! (Black already stood better but this makes matters worse) 10 … Bxc3+ 11 bxc3 Qa5 12 Qd2 and White has a miserable structure, while simultaneously lagging in development, J.Balares-C.Lakdawala, San Diego (rapid) 2009. This game is annotated in Sveshnikov: Move by Move. 7 … Bxf5 8 dxc6 bxc6 Don’t fear Black’s two queenside isolanis in this position. Analysis proves that it is White’s lack of development which makes him the one fighting for equality. 8 … Qxd1+ is a key alternative but I prefer to keep White’s knight on c3, where it does nothing, rather than trade queens and allow it to reroute to e3, where it regains a tempo on the f5bishop, while protecting the c2-pawn. 9 Qf3 White’s main move by a about a 6 to 1 margin. There is a double attack on f5 and c6. 9 … Qd7 This move is superior to 9 … Qc8, staying off the d-file, which has also been played. I’m not a fan since the move is artificial and clogs Black’s route to a future … Rad8. 10 Bc4! Zelcic wisely hands over a pawn to complete development. White gets into trouble by hanging on to it. Instead: a) 10 Bg5 e4 11 Qe2 Be7 12 Rd1 (12 Bxf6? Bxf6 13 Nxe4 0-0 14 Nxf6+ gxf6 15 Qd1 Rfe8+ 16 Be2 Qe6 17 Kf1 Rad8 18 Bd3 c5 19 b3 Qe5 and White, who is guilty of the mother of all chess sins – falling behind in development in the opening – is

unable to unravel) 12 … Qe6 13 Qc4 Rb8 14 Qxe6 fxe6! 15 b3, V.Akopian-Y.Yakovich, Rostov on Don 1993. Black stands better after 15 … Bb4 16 Bd2 Rd8 (threat: … Rxd2 and … Nd5) 17 Nb1 Bc5. b) 10 Bd3 e4 11 Qe2 Bb4 12 Bd2?! (the unplayed 12 Bc4 is better) 12 … 0-0 13 Bc4 Bg4 14 Qe3 Rad8 with a wretched game for White, who lags seriously in development, J.JavorskyV.Hybl, correspondence 2005. c) 10 Bb5?? e4 – Oops. White’s pin on the c6-pawn is broken and a piece is lost. In my database, nine unfortunates fell for this trap as White. 10 … Bxc2!?

Black takes the time out to snatch a pawn. Chances are dynamically balanced here. We can also meet the seductive c2offer with indifference and play 10 … Be7, which is about even after 11 0-0 0-0 12 Rd1 Qc8 13 Re1 e4. 11 0-0 Be7 12 Qe2

Double attack. White regains the sacrificed pawn but with a loss of time. Instead, 12 Re1 0-0 13 Rxe5 Bd6 looks even. 12 … Bg6 13 Qxe5 0-0 14 Rd1 Qb7 15 Qg3 The first new move. I give Black an edge, since White’s development lag is more significant than the fact that Black’s queenside pawns are broken. a) 15 Ba6?! (this exchange costs White time) 15 … Qxa6 16 Qxe7 Rfe8 17 Qc5? Bc2! 18 Rf1 Qxf1+! 19 Kxf1 Bd3+ 20 Ne2 Rxe2 21 Qc3 Rae8! 22 g3 was R.Bellin-V.Ikonnikov, Huy 1993. White would be crushed if Black found 22 … R2e3+! 23 Kg1 Re1+ 24 Kg2 Bf1+ 25 Kf3 Bh3!, threatening … Rg1! and … Bg2+, with a winning attack. b) 15 b3 is unplayed and may be an improvement. The game looks balanced after 15 … Rfe8 16 Qg3 Rad8. 15 … Rfd8 It doesn’t matter which rook Black places on the d-file, since White will swap. 16 Rxd8+ Rxd8 17 Bb3 White has no easy route to complete development. 17 … c5! Any attempt to develop the c1-bishop will be met by … c5c4 and … Qxb2. 18 h3!? Luft is a luxury. Perhaps he should try 18 Be3 Bd3! 19 Qf4 Rc8! (Black seizes control over c4 and … c5-c4 is coming) 20 Qf3! Qxf3 21 gxf3 c4, though White is the one fighting for the draw in the ending. 18 … Bd6 Black’s intent (if he has one, besides harassment of White’s queen) remains without form or shadow, yet we realize that White is in danger due to the fact that his queenside pieces remain asleep and undeveloped. This is a position where precise actions matter, not chance or intuition.

Better is 18 … Bd3!, intending … c5-c4. After 19 Bh6 Nh5 20 Qe5 c4! 21 Qxh5 cxb3 22 Qg4 Bg6 23 Bg5 f6 24 Be3 bxa2 25 Rxa2 a6! Black can press with his bishop pair in the open position; e.g. 26 Qc4+ Bf7 27 Qxa6 Qb8 and Black’s bishop pair outweighs White’s extra b-pawn. 19 Bf4?! White enters a risky calculation battle. The move violates the principle: If your goal is to extinguish, then don’t ignite. 19 Qh4! is the sober and better choice. 19 … Nh5 20 Qh4? White reduces his disadvantage to a minimum if he plays 20 Qg5!, attacking the d8-rook. After 20 … Be7 21 Qe5 Nxf4 22 Qxf4 the defence bends with the wind without snapping. White only stands slightly worse here. 20 … Be7! 21 Bg5 If 21 Qg4 then 21 … Rd4 22 Bd5 Rxd5 23 Nxd5 Qxd5 24 Rd1 Qa8 is a winning position for Black, who has picked up two full pieces for a rook. 21 … Rd4 Also strong is 21 … h6! 22 Bxe7 Rd4 23 g4 Nf4 24 Qg3 Qxe7 and Black dominates. 22 g4? White’s last prayer lay in 22 Re1! Rxh4 23 Rxe7 Qb8 24 Bxh4, though even here White should lose.

Exercise (combination alert): The GM playing White just gave in to delirious, time-pressure-fuelled abandon. Find Black’s next move and we have no doubt of the verdict. 22 … Nf4! Answer: Step 1: Black makes good use of the fact that the pinned g5-bishop is unable to capture the f4-knight, which is happy to fill the void. A not-so-veiled threat to deliver mate on g2 arises. 23 Nd5 Zelcic desperately attempts to confuse the issue. 23 Qg3 Bxg5 24 h4 Rd3 is hopeless. 23 … Bxg5!

Step 2: Simply chop on g5. White’s queen and king are forked if he recaptures. 24 Qxg5 In the disciplined manner of a professional, I resist the urge to place a double question mark after this move. First, it’s a rapid game and White may have had one second on his clock. Secondly, it’s not a blunder if everything else loses. White’s options have run out. 24 Re1, threatening a back rank cheapo, is met by the simple 24 … Re4. 24 … Nxh3+ 0-1

Chapter Six 3 Nc3: The Anti-Sveshnikov This book doesn’t cover the anti-Sicilian systems, but I made an exception and added this chapter. This particular AntiSicilian is specifically designed to mess up the Sveshnikov, so it justifies inclusion. I offer three Carlsen games which lead the game to a Closed Lopez character after 3 … e5. Black equalizes completely in this variation. Game 52 P.Svidler-M.Carlsen Grenke Classic, Karlsruhe/Baden 2019 1 e4 c5 2 Nf3 Nc6 3 Nc3

This annoying Anti-Sveshnikov line tends to lead to dry positions and that is exactly why it is chosen by White players of a certain disposition. Evading the Open Sicilian and Rossolimo isn’t the most dynamic colourful way for White to proceed and Black equalizes comfortably. I chose three recent Carlsen games to show that it isn’t as drawish as it may first appear. 3 … e5 This may be Black’s best – and most Sveshnikov-like – move. Black abandons control over d5 but in return receives more than a fair share of central control. We won’t look at Black’s main options: 3 … g6; 3 … d6; 3 … e6; and 3 … Nf6. 4 Bc4 White increases pressure on the d5-square. 4 Bb5 doesn’t make as much sense. After 4 … d6 5 d3 Nf6 6 h3 Be7 7 0-0 0-0 in E.Jensson-A.Shirov, Bilbao 2014, White saw nothing better than 8 Bc4, which merely lost a tempo over normal lines. 4 … Be7 After 4 … g6!? 5 d3 h6 6 h4 d6 7 h5! g5 8 Nh2! Bg7 9 Ng4 Nge7 10 Ne3, White stood better due to his grip on the light squares, I.Nepomniachtchi-M.Carlsen, Grand Chess Tour, Zagreb 2019. Carlsen managed to mix it up later and won anyway. 5 d3 d6 6 Nd2

Intending Nf1 and Ne3 to increase his grip over d5. An important point is that only one white piece can occupy d5 at a time even if White has three pieces trained on the square. Later on we look at 6 0-0. 6 … Nf6 7 Nf1 Nd7!? There are two ideas behind this move: 1. Black makes a way for a future … f7-f5. 2. Black may later activate or exchange off his potentially bad dark-squared bishop with … Bg5. 7 … Bg4 8 f3 Be6 is the main line. 8 Nd5 In the next game we examine 8 a3. 8 … Nb6 9 Nxb6 9 Nfe3 was seen in P.Harikrishna-S.Nihal, Malmö 2019. Here Black can try 9 … Bg5 with an even position. 9 … axb6

You may be wondering if the b5-hole is a concern for Black. For now it isn’t, since White’s knight is unable to reach it and Black doesn’t care if White’s bishop sits there. 10 c3 0-0 11 Ne3 Bg5 Carlsen activates his bad bishop. 12 0-0 Kh8 Now … f7-f5 is coming. Another possibility is 12 … Bxe3 (Black rids himself of the bad bishop and, at the same time, removes White’s powerful knight, which would have occupied d5) 13 Bxe3 Qe7 14 f4 exf4 15 Bxf4 Be6 16 Bxe6 Qxe6 17 Qb3 Qxb3 18 axb3 Rad8 19 Bg3 Rd7 20 Rf5 Rfd8 21 b4 Ne7 and the players agreed to the draw in this equal ending, P.Leko-V.Kramnik, Dortmund 2003. 13 a3 A new move. Svidler safeguards the c4-bishop and leaves open the possibility of expansion on the queenside with b2-b4. 13 Nd5 is White’s main move, when J.SindarovN.Yakubboev, Uzbeki Championship, Tashkent 2019, fizzled out with 13 … Bxc1 14 Qxc1 ½-½. After 14 … f5 Black looks fine. 13 … f5!?

This is certainly a radical approach. Carlsen is willing to hand over his light-squared bishop to open the f-file and gain a preponderance of central control. 14 Nxf5 Bxc1 15 Rxc1 Bxf5 16 exf5 d5 So White gets the perhaps superior minor piece and potential to cause trouble on the light squares, while Black gets the centre and an open f-file. 17 Ba2 Rxf5 18 Qg4 Svidler seeks to increase his influence on the light squares. The comp prefers 18 Qb3 Ne7 19 Rce1 Qd6 20 f3 Raf8, when I see no useful plan for White but to wait. 18 … Rf6 19 f4?! Here 19 Rce1 was correct.

This move 19 f4 demonstrates why chess so difficult and full of contradiction. White’s logical pawn push obeys two principles: 1. Chip away at your opponent’s centre from the wings. 2. When you have a bishop versus the opponent’s knight, open the game. Nevertheless, Carlsen proves that Svidler’s move is not the best and actually offers Black an edge. 19 … exf4 20 Qg5 It could be that Svidler originally planned 20 Rxf4?? but then saw 20 … Ne5 21 Qe2 (not 21 Qg3? Rxf4 22 Qxf4 Nxd3, forking queen and rook) 21 … Rxf4 22 Qxe5 Qf8 23 Qxd5, when White has a pawn for the exchange but his back rank is very weak; for example, 23 … Re8! 24 Qxb7 c4! and if 25 Bxc4 then 25 … Qc5+ 26 Kh1 Qe3! 27 Rg1 Qxg1+! 28 Kxg1 Re1 mate. 20 … Qf8 21 Qxd5

Svidler has regained his pawn, at the cost of handing Carlsen a ferocious initiative. 21 … Rd8 22 Qf3 White loses a tempo, one way or another. After 22 Qe4 Re8 23 Qd5 Re5 24 Qf3 Qe7 25 Bc4 Re3 26 Qd5 Rf8 Black is in control. 22 … Ne5! 23 Qe4! Not 23 Qxb7?? f3 24 g3 Nxd3 25 Rc2 c4! and White can resign. 23 … Ng4! Carlsen correctly opts for initiative over pawn grabbing. 23 … Nxd3? would allow White to escape with 24 Rcd1! Nxb2 25 Rxd8 Qxd8 26 Rf2 Nd3 27 Rd2 Rd6 28 Bb1 c4 29 Qxc4 Nc5 30 Rxd6 Qxd6 31 Qd4 and White should hold the game. 24 Rce1 Ne3 25 Rf2

Exercise (critical decision): Should Carlsen protect his loose b7-pawn with 25 … Rd7, or should he go for it with 25 … Re8, sacrificing b7 and staking everything on his attack? 25 … Re8!! Answer: Carlsen offers the b7-pawn for his attack, despite the reduced material on the board. Subsequent analysis proves it to be completely sound. 26 Qxb7 g5! Curiously, White is helpless against the push of Black’s gpawn. 27 Rfe2? White’s only chance was to apply the principle: meet the opponent’s wing attack with a central counter and play 27 d4!, although 27 … g4 still gives Black a strong attack. 27 … g4 Here 27 … f3! is even more powerful; e.g. 28 Rf2 Re7 29 Qa6 fxg2 30 Rxf6 Qxf6 (threatening … Qf1+) 31 d4 cxd4 32 cxd4 Qf1+ 33 Rxf1 gxf1Q+ 34 Qxf1 Nxf1 35 Kxf1 Re4 36 d5 Kg7 with a straightforward technical win for Black. 28 Rf2 Qh6! Threatening … g4-g3. 29 Qc7 Ref8! 30 h3 Nothing saves White at this point: a) 30 Qe5 g3! (clearance) 31 hxg3 Ng4! wins. b) 30 g3 fxg3 31 Rxf6 (31 Qxg3 Rxf2 wins) 31 … Qxh2 mate. c) 30 Rfe2 f3! 31 gxf3 gxf3 32 Rf2 (or 32 Rxe3 f2+ 33 Kg2 fxe1Q 34 Rxe1 Qd2+ with mate in three) 32 … Rg6+ 33 Kh1 Ng4 34 Qg3 Nxf2+ 35 Qxf2 Rg2 forces mate. 30 … gxh3 31 g3

Svidler hopes to shield his king behind the opposing pawns. Unsurprisingly, the plan doesn’t work since Carlsen will simply keep pushing them. But there was nothing better; e.g. 31 Qe5 Ng4 32 Qe2 hxg2 and White is unable to avoid mate. 31 … fxg3 32 Rxf6 h2+ 33 Kh1 It’s time for the king to pay his debt to society. ________W [WDWDW4Wi] [DW!WDWDp] [W0WDW$W1] [DW0WDWDW] [WDWDWDWD] [)W)PhW0W] [B)WDWDW0] [DWDW$WDK] W––—W Do you possess the skills to solve this unravelable mystery? GM Svidler proved he is a good sport by allowing Carlsen to deliver mate in one. 33 … g2 mate Game 53 S.Karjakin-M.Carlsen FIDE Grand Prix, Abidjan (rapid) 2019 1 e4 c5 2 Nf3 Nc6 3 Nc3 e5 4 Bc4 Be7 5 d3 d6 6 Nd2 Nf6 7 Nf1 Nd7 Carlsen repeats his idea from the previous game. 8 a3 Karjakin creates some air for his light-squared bishop. 8 … Nb6 9 Ba2 0-0 9 … Be6 would be met by 10 Nd5, when 10 … 0-0 11 Nfe3 transposes to the game 10 Ne3 Be6 Alternatively, 10 … Bg5 11 h4!? (again 11 0-0 Be6 12 Ncd5 transposes below) 11 … Bxe3 12 Bxe3 Nd4 13 Qd2 Be6 14

Bxe6 Nxe6 15 f4 exf4 16 Bxf4 d5 led to equal chances in A.Levitskiy-V.Baidetskyi, Vinnytsia 2018. 11 Ncd5 Bg5 12 0-0 Bxe3 13 fxe3

This move was actually a novelty. Karjakin opens the f-file for his rook, at the slight cost of hemming in his dark-squared bishop. Instead, after 13 Bxe3 Bxd5 14 exd5 Ne7 15 c4 Nf5 16 b4 Nd7 17 Bb3 Rc8 18 Qd2 Nxe3 19 fxe3 Qg5 20 Ra2 f5, V.Smirnov-S.Zhigalko, Belarus Championship, Minsk 2002, if anyone stands better it is Black. 13 … Ne7 Just as in the Sveshnikov, Black’s prime directive in this line is to constantly challenge White’s occupation of d5. 14 c4 The d5-square is reinforced. The cost is that that White clogs the position for his bishop pair. Carlsen has achieved dynamic equality.

On the other hand, if 14 Nxb6 Qxb6 15 Bxe6 fxe6 16 b3 Rxf1+ 17 Qxf1 Nc6, there is no advantage in White having bishop against knight. 14 … f5 15 Bd2 Bxd5!? Carlsen goes radical with his desire to extend the imbalances, handing over both his bishops for both white knights. For now the game is closed, yet his move is a highstakes gamble, since the game may open up later on to the benefit of White’s bishops. Instead: a) 15 … fxe4?! only helps White; e.g. 16 Rxf8+ Kxf8 (if the queen captures then 17 … Nc7 forks rook and bishop) 17 dxe4 and White’s doubled e-pawns, far from being a weakness, help him, since they cut off key squares d4 and f5. b) 15 … Nbxd5 is the safe route and the one which the vast majority of us would undoubtedly have taken. After 16 exd5 Bd7 17 b4 b6 18 bxc5 bxc5 19 d4 cxd4 20 exd4 Ng6 21 dxe5 Nxe5 22 Bc3 Rc8 23 Qd4 White only has a shade of an edge. 16 exd5 Nd7 With White’s b2-b4 is coming, Black must be ready to play … b7-b6. 17 b4 b6 18 Bb3 The bishop crawls its way back into play, heading for a4. 18 … Ng6 19 Ba4 Qe7 20 Bc6 White makes headway into Black’s queenside light squares, and it appears that Black’s kingside attack is way behind. However, the position is deceptive and Carlsen had foreseen that Black does indeed have enough play. 20 … Rad8 21 Qa4!? The problem with a move like this is that the white queen abandons the king. All the same, this is getting scary for Black as a7 hangs and if 21 … Nf6, then White has 22 Rxf5. Is Carlsen in trouble? The answer is no. It would take great willpower to avoid 21 Qa4!? and play a more sober move such as 21 e4!, which may in fact have been

White’s best choice. 21 … Nf6 The computer suggests 21 … e4!? with great complications; e.g. 22 dxe4 Nde5 23 exf5 Nxc4 24 Bc1 Qe5 25 Ra2 Rxf5 26 Rxf5 Qxf5 27 Qc2 (White’s queen is forced to fall back, as 27 Rf2? Qb1 28 Qc2 Nxa3! favours Black) 27 … Qg4 28 Qe2 Qe4 29 Qc2 and Black can either take the repetition draw, or play for the full point by moving the queen to h4. 22 bxc5 bxc5

Exercise (critical decision): Carlsen, refusing to bow obsequiously before his opponent’s threats, has just left his f5-pawn hanging. Should Karjakin’s rook take it, or should he refrain with a move such as 23 Rae1 - ?

23 Rxf5? Answer: Believe it or not, this pawn is poisoned and the evaluation now swings from even to -2.88, which means White is totally busted. 23 Rae1 was necessary. 23 … e4?! Black’s advantage diminishes a little after this inaccuracy. It was stronger to send in the cavalry with 23 … Nh4! 24 Rff1 Ng4! 25 Qd1 (25 h3 can be ignored with 25 … Qg5!) 25 … Rxf1+ 26 Qxf1 Qg5 27 Qe2 e4! (threatening 28 … Qe5 with a double attack on h2 and a1) 28 h3 Ne5! (even stronger than … Qe5 now) 29 Rf1 Nef3+ 30 Kh1 Qg3! 31 gxf3 Qxh3+ 32 Kg1 Qg3+ 33 Kh1 exf3 34 Qf2 Qg2+! and Black wins. 24 Qc2 Ng4 The alternate knight thrust is stronger; e.g. 24 … Nh4! 25 Rf4 g5! 26 Rff1 Ng4 27 h3 exd3 28 Qxd3 Nf2! 29 Rxf2 (29 Qe2 Qe5! is winning for Black) 29 … Rxf2 30 Kxf2 Qf6+ 31 Ke2 Qxa1, when White finds himself down an exchange with no compensation. 25 Raf1 Rxf5 26 Rxf5 Qh4?! Slightly inaccurate again. After 26 … Nh4! 27 Rf4 Qg5! (Black threatens to burn everything to the ground with 28 … Nxg2!) 28 Ba5! Nf3+! 29 gxf3 Nxe3+ 30 Rg4 Nxg4 31 fxg4! Rf8 32 Qe2 exd3 33 Qe4 Qc1+ 34 Qe1 Qxc4, White is busted due to Black’s two connected passed pawns and White’s airedout king. 27 h3 Nxe3 Here Black gets a greater advantage with 27 … exd3! 28 Qxd3 N4e5 29 Qb1 Nxc4. 28 Bxe3 Qe1+ 29 Kh2 Qxe3 30 dxe4 Nf4!? Carlsen wants to trap the wobbly rook on f5. The alternative is to fill the vacuum with 30 … Ne5 31 Rf1 g5! 32 Qf2 Qxf2 33 Rxf2 Nxc4 and White is the one fighting for the draw. 31 Qb2?! There is no reason to give Black the e4-pawn. White should have played 31 Qb1.

31 … Qxe4 32 Bd7?? 32 Rg5 was forced, though White is still in deep trouble after 32 … Ng6 33 Rg4 Qe3 due to the power discrepancy of the minor pieces.

Exercise: Just because truth is on our side doesn’t automatically mean we are destined to win. We still have to work out the details. White’s last move mistakenly attempts to return his bishop to play. How did Carlsen refute it? 32 … g6! Answer: Step 1: Lure White’s rook to f6 with tempo. Even at the end of a rapid game, the odds of Carlsen falling for 32 … Rxd7?? 33 Qb8+ Rd8 34 Qxd8+ Qe8 35 Qxe8 mate

are not great. 33 Rf6 Forced, since 33 Rg5? Rxd7 is a free piece. 33 … Nd3! Step 2: Transfer the knight to d3, attacking White’s queen, knocking it off track from the b-file, so it can no longer give check on b8. 34 Qc3 34 Qb1 drops the rook to 34 … Qe5+ etc, and 34 Be6+ doesn’t help because Black would just take it. 34 … Rxd7 This is possible since White no longer has Qb8+. 35 Re6 Qf4+ 0-1 Black ends up with an extra knight. Game 54 V.Sanal-Y.Kuzubov Turkish League 2019 1 e4 c5 2 Nf3 Nc6 3 Nc3 e5 4 Bc4 Be7 5 d3 d6 6 0-0

In this version White declines to resort to contortions based on Nd2 and Nf1, obsessing over the d5-square. Instead, he castles, preparing to launch the f-pawn forward with Ng5 and f2-f4. 6 … Nf6 6 … Bg4?! is tempting yet inaccurate. It is White who gains time after 7 h3 Bh5 8 Nd5 Nf6 9 Ne3 Nd4 10 g4 Bg6 11 c3 Nxf3+ 12 Qxf3 Qd7?! (Black reduces his disadvantage with 12 … 0-0 13 a4) 13 g5 Nh5 14 Nd5 Bd8 15 Qg4 and White had achieved close to a winning position in V.Kupreichik-A.Vaulin, Tula 2002. 7 Ng5 This way White attacks f7 and has time to unleash f2-f4. Instead, 7 h3 is feeble and allows Black an easy Closed Ruy Lopez-like equality after 7 … 0-0 8 Ne2 a6 9 a3 b5 10 Ba2 Be6, L.Aronian-M.Carlsen, World Blitz Championship, St Petersburg 2018. 7 … 0-0 8 f4 Bg4

This move disrupts the white queen’s protection of c2 and is considered a shade more accurate than 8 … exf4. 9 Qe1 exf4 At first it appears as if Black loses a tempo, opens the f-file for White and reduces central influence. In fact, the move follows the principle: open the centre when the opponent is building a wing attack. 10 Bxf4 Nd4

11 Qd2 Or 11 Qf2 h6 12 Nxf7 Rxf7 13 Bxf7+ Kxf7 14 e5 dxe5 15 Bxe5 Kg8 16 Ne4 Be2! 17 Nxf6+ Bxf6 18 Bxf6 Qxf6 19 Qxf6 gxf6 20 Rxf6 Nxc2 21 Rc1 Bxd3 22 Rf2 and here Black can either go for it after 22 … Nd4!?, or play it safe with 22 … b6 23 Rfxc2 Bxc2 24 Rxc2 with a likely drawn ending. 11 … Qd7 Another option is 11 … h6, despite the potential risk of giving White a sacrificial target on h6 and an attack on the g-

file; e.g. 12 Nf3 Nxf3+ 13 gxf3 Bh3 14 Rfe1 Nh5 15 Kh1 Nxf4 16 Qxf4 Bg5 17 Qg3 Be6, P.Harikrishna-T.Radjabov, European Cup, Rhodes 2013, and now 18 Rg1 Qf6 19 Nd5 Bxd5 20 Bxd5 Rab8 21 f4!? Bxf4 22 Qf3. All the same, Black should be able to defend, and he is a pawn up. 12 a4 If 12 Be3 Be6 13 Nxe6 fxe6 14 Bxd4 cxd4 15 Ne2 Kh8 16 Bb3 (not 16 Nxd4? d5 17 Bb5 Qc7, threatening … Qb6 and White is in trouble) 16 … d5 17 exd5, A.Volokitin-Yu Yangyi, World Rapid Championship, Doha 2016, then 17 … Nxd5 is equal. Black’s d-pawn doesn’t really hang since 18 Nxd4?? Bc5 19 c3 e5 wins a piece. 12 … h6 After 12 … a6 13 e5 dxe5 14 Bxe5 Bf5 15 Rae1 h6 16 Nf3 Rae8 (16 … Be6 17 Bxe6 fxe6 18 h3 looks a shade better for White) 17 h3 Bh7 18 Nxd4 cxd4 19 Bxf6 Bxf6 20 Rxe8 Rxe8 21 Nd5 Bg5 22 Qf2 Bg6, White didn’t get anywhere in R.Robson-D.Swiercz, St Louis 2019. 13 Nf3 Nxf3+ The position becomes tricky, since White may attack down the newly open g-file. On the other hand, White’s king is also now less secure. 14 gxf3 Be6 It is in Black’s best interest to exchange pieces. 15 Kh1 15 Bxh6? is premature and unsound. Black defends after 15 … gxh6 16 Qxh6 Nh7 17 Kh1 Kh8 18 Rg1 Rg8 19 Nd5 Bxd5 20 Bxd5 Bg5 21 Qh5 Rg6 22 Raf1 Rh6! as White’s attack is repelled, and if 23 Qxf7 Qxf7 24 Bxf7 Bf4 25 Rg2 Ng5 26 Bd5 Kg7!, then … Rah8 is coming. White is completely tied down to defence of h2 and f3, and his kingside pawns are blockaded. 15 … Kh7!

A dual-purpose move: 1. Black’s king helps with the defence, offering added protection to h6, a point where White hopes to sacrifice. 2. Black’s king makes way for … Rg8 which will reinforce g7. 16 Rae1 Or if 16 Qe2 Nh5 17 Bd2, M.Emelyanov-C.Cruzado Duenas, correspondence 2014, then 17 … Bf6 and the game is equal. 16 … Nh5 Gaining a tempo, while clearing f6 for the dark-squared bishop. 17 Bg3 After 17 Be3 Bh4 18 Rd1 Bf6 19 Nd5 Bd4! Black is just fine. If now 20 Bxd4?! cxd4 21 b3 Bxd5 22 Bxd5 Rac8 23 Qf2 Nf4, Black already stands slightly better.

17 … Bf6 18 Nd5!? Bxd5 19 Bxd5 Bxb2!? This move allows White, a lower rated GM, a draw. If Black is playing for the full point, he should try 19 … Be5 20 Kg2 g5. 20 Rb1?!

White either missed or rejected a chance at an upset draw. After 20 c3! Nxg3+ 21 hxg3 Qh3+! 22 Kg1 Qxg3+ 23 Kh1 Qh3+ the game ends in perpetual check. 20 … Bf6 21 a5 Not 21 Rxb7?? Nxg3+ 22 hxg3 Qh3+ 23 Kg1 Bd4+ and wins; while after 21 Bxb7 Rab8 22 Bd5 Nxg3+ 23 hxg3 Qxa4 Black wins a pawn, since White can’t afford 24 Rxb8? Rxb8 25 Bxf7?? Qd7! 26 Bd5 Qh3+ 27 Kg1 Bd4+, forcing mate. 21 … Rab8 22 f4 b5! Now White doesn’t have enough compensation for the pawn. 23 axb6 Rxb6 24 Kg2

A wise precaution, countering ideas of … Nxg3+ and … Qh3+. 24 … Rfb8 25 Rxb6 Rxb6 White must keep watch over Black’s passed a-pawn. 26 Qe2 g6! It’s critical for Black to avoid bishops of opposite colours as these may increase White’s chances to draw; e.g. 26 … Nxg3?! 27 hxg3 Bc3 28 Qh5 g6 29 Qd1 a5 30 f5 Bg7 31 Qg4 Qe7 32 Qf4 Be5 33 fxg6+ fxg6 34 Qf7+ Qxf7 35 Rxf7+ Bg7 36 Ra7 Rb2 37 e5! Rxc2+ 38 Kf3 dxe5 39 Rxa5 and White should easily hold the draw. 27 f5 Kg7 28 Be1!? The idea is to stall … a6-a5; the cost is an exchange of rooks, which greatly reduces White’s chances to go after the black king. 28 … Rb1! Black’s rook slips in and becomes even more active. 29 fxg6 fxg6 30 Qf3 Kh7 31 h3 Kg7 32 Bc4 Rc1 33 e5?! This is a misguided attempt at a combination. He should have marked time and asked Black how he planned to proceed. 33 … Rxc2+ 34 Kg1 dxe5 35 Qa8!

Exercise (critical decision): White threatens mate on the move on g8. Black can give a queen check on d4, or he can sacrifice the exchange on c4, nullifying the mating threat. One line gives Black a winning position, while the other allows White to escape. 35 … Rxc4! Answer: Now White’s “attack” is turned into the status of an imaginary friend. Black gives up the exchange, for which he gets three pawns and a winning position. 35 … Qd4+? would allow White to escape with a draw after 36 Bf2 Qd8 37 Qxa7+ Qe7 38 Qb8 Qf8 39 Qa7+ etc, since 39 … Be7?? even loses to 40 Bh4 Qd8 41 Rf7+ Kh8 42 Qxe7 Qd4+ 43 Bf2.

36 dxc4 Nf4 Attacking h3, which can’t be saved. Now it will be four pawns for the exchange. 37 Bg3 37 h4? fails to save the pawn after 37 … Bxh4!, since if 38 Bxh4? then 38 … Qg4+ 39 Kf2 Qe2+ 40 Kg1 Nh3+ 41 Kh1 Qxf1+ 42 Kh2 g5 wins. 37 … Nxh3+ 38 Kh1 Nf4 39 Rb1 Threatening 40 Rb7. 39 … Be7 Simpler is 39 … Qh3+! 40 Bh2 h5 41 Qxa7+ Kh6 42 Qa8 (covering the g2-mating threat) 42 … Nd3 43 Qg2 Qf5 and Black is winning. 40 Qe4 Qh3+ 41 Bh2 Kf7 42 Ra1 Qd3 A queen exchange leads to a technically lost ending for White. 43 Qb7 a5! 44 Bxf4 Not 44 Rxa5? e4! (interference; White’s queen is cut off from g2) 45 Bxf4 Qf1+ 46 Kh2 Qxf4+ 47 Kg2 Qf3+ 48 Kh2 Qe2+ 49 Kh3 Qe3+ 50 Kg2 (or 50 Kg4 h5 mate) 50 … Qd2+ and White’s rook falls. 44 … exf4 45 Re1 Qd6 46 Qa7 Qf6 47 Qxa5 Protecting the rook from … Qh4+ and … Qxe1. 47 … f3 White’s game reaches the final stages of its ebbing life. Black’s king is safe and his three connected passed pawns are way too much for White to handle. 48 Qd2 Qh4+ 49 Kg1

Exercise (combination alert): I have heard of an ancient, abstract lore know as mathematics, yet have never been able to understand it or apply it to my own games. How did Black force resignation through a forcing continuation? 49 … f2+! 0-1 Answer: Attraction/bishop skewer. After 50 Qxf2+ Qxf2+ 51 Kxf2 Bh4+ White will be two pawns down in the king and pawn ending.

Chapter Seven 6 … Bc5: The Mamba We end the book with a quirky chapter on an experimental variation that is a lot better than it looks – a Sveshnikov hybrid, which is simultaneously a Sveshnikov and not a Sveshnikov. This is a dangerous (but playable!) alternative to the standard 6 … d6. If you want to enter a position where you are booked up and your opponent isn’t, then take a look at the games in this final chapter. For an explanation of the derivation of the name “the Mamba”, see page 13. Here I’m attempting to resurrect a line which is considered “dead”. In fact, it’s so dead that very few people even know of its existence. So dissimilar are the plans in the Sveshnikov and the Mamba, that they cannot even be considered the same language, spoken with different accents. My not-so-easy goal is to prove to you that it is not shady in the least. I have tried the Mamba in eleven tournament games (so far) and my score is a shockingly high 10½-½, so maybe the system has some merit. Even if you believe it is a bit shady, it’s still a devastating surprise weapon. Two of my students play it with success, even against comped-up opposition, so my belief is that the line is sound. I encourage you to give it a try. One thing I can guarantee: you will know it better than your opponent, who will almost certainly have never seen it before. Game 55 B.Baker-C.Lakdawala San Diego (rapid) 2005 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5!?

I have a small confession to make. I rarely play the main lines of the Sveshnikov. Instead, my two true loves are both bizarre sidelines, one of which (the Ulfie) is featured in Sveshnikov: Move by Move, while the other, the Mamba, is covered in this book. In the early 2000s I started goofing around with a crazy notion: is White’s coming Nd6+ an actual threat? I saw one anonymous IM allow it in an online game, with success, so I followed suit in an online blitz game against GM Eric Lobron and won a miniature. This got my attention and I worked out a full repertoire for this offbeat Sicilian. This was my very first outing with the line, against a high-rated opponent. The swiftness of the result jolted me, as much as it did my opponent. 7 Nd6+

This is what almost everyone plays. We’ll look at a few other tries – 7 g4, 7 Bg5, 7 Be2, and 7 Be3 – later in the chapter.

7 … Ke7 This is crazy, but not by this opening’s standards. Have no fear. It’s all part of the master plan! Before you claim that I should be declared insane, let me point out that this is my highest scoring opening – with White or Black! 8 Nf5+ This natural move is most frequently played. Others: a) 8 Nc4 is covered in Game 62. b) 8 Nxc8+ picks off the bishop pair, at the cost of eliminating an active piece for an undeveloped one. Now 8 … Rxc8 9 Bc4 was seen in W.Browne-E.Liu, US Championship, San Diego 2006, where my student Elliot should have played 9 … Bb4! 10 0-0 Bxc3 11 bxc3 d6 12 Ba3 Qc7 13 Qd3 Rhd8 with a quite playable game for Black, whose superior structure makes up for White’s bishops. c) 8 Bg5?? is a blunder due to 8 … Bxd6! 9 Qf3 (White now realizes that if 9 Nd5+ Ke8! 10 Nxf6+ gxf6, he loses a piece, since after 11 Qxd6 his own bishop hangs with 11 … fxg5) 9 … Bb4 10 0-0-0 Bxc3 11 Qxc3 d6 12 Bb5 Qa5 13 Qd3 Rd8 14 f4 Nd4 15 Bc4 Be6 16 Bxe6 Nxe6 0-1 B.Barquin-C.Lakdawala, San Diego (rapid) 2012. 8 … Kf8

9 Be3 Alternatives: a) 9 Bc4 is examined in Game 61. b) 9 Bg5 d6 10 Ne3! (10 Nd5?? runs into 10 … Qa5+! 11 Bd2 Bxf2+! 12 Ke2 Qc5 13 Nxf6 Bxf5 14 exf5 gxf6 15 Bh6+ Ke7 and White is completely busted; the same cheapo works after 11 Qd2) 10 … Be6 11 Ncd5? (correct is 11 Bd3 h6 12 Bxf6 Qxf6, although Black looks just fine here) 11 … Qa5+! 12 c3 Nxe4 13 b4 Bxe3! 14 bxa5 Bxf2+ 15 Ke2 Bg4+, when Black regains the sacrificed queen with a completely winning position, U.Atakisi-B.Heberla, Turkish League 2017. c) 9 Bd3 was seen in T.Abergel-R.Gervasio, Paris 1998. Here Black played 9 … d6, which is fully viable. However, our thematic push 9 … d5! is stronger. Now after 10 exd5 Nxd5, White should just castle and accept the queenside damage after 11 … Nxc3. If he tries to evade that with 11 Nxd5?!, Black has 11 … Qxd5 with the dual threats of … e5-e4, winning a piece,

and also … Qxg2. Black stands slightly better after 12 f3 Be6 13 Qe2 Rd8, since White is unable to castle kingside easily. d) With 9 Ne3, as in D.White-J.Moeckel, correspondence 2008, White seeks complete domination of d5. The trouble is that only one white piece can occupy the square at a time and Black can play around the knight. Here 9 … d6 10 Bc4 Be6 11 Bb3 g6 12 0-0 Kg7 looks fine. White’s control over d5 isn’t worth much. 9 … Bb4

In my experience this position arises quite frequently in this variation. It’s fun (in a kind of unsettling way!) and I don’t believe White has even an iota of an edge, and neither does the computer. Incidentally, in case you prefer the Four Knights route to the Sveshnikov, this line can be reached that way too: 2 … e6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 Nc6 and now 6 Ndb5 Bc5!? 7 Nd6+ Ke7 8 Bf4 e5 9 Nf5+ Kf8 10 Be3 Bb4, as in fact occurred in Game 60.

10 Qf3? This natural move is a serious mistake. We’ll look at alternatives in Games 58-60. 10 … d5! Black plays the thematic break, which arrives with a coiled spring effect. White is already in trouble, since Black menaces … d5-d4, as well as … d5xe4 and also … Nxe4. 11 exd5 As we’ll see in Game 57, 11 0-0-0 is not the solution for White. 11 … Qxd5 Forced, since 11 … Nxd5?? walks into a fatal pin after 12 00-0. But now Black threatens 12 … Qxf3, followed by … Bxf5, winning a piece. 12 Qh3?? And this is just a blunder. The next game’s 12 Nh4? is also very bad. The only move is 12 Ng3, when Black either can trade queens on f3 and wreck White’s structure for the ending, or play the braver comp suggestion 12 … Qa5!, when it is White’s king who is in danger, not Black’s.

Exercise (combination alert): How did Black exploit White’s last move? 12 … Nd4! Answer: Step 1: Double attack. 12 … Nd4 threatens c2 and f5. 13 Bd3 After 13 0-0-0 Bxc3 14 bxc3 Bxf5 White can resign. 13 … e4! Step 2: Overloaded defender. I had comped this entire sequence at home in pre-game preparation and knew that Black was winning easily here. 14 Bxd4 Bxf5 0-1 Step 3: Chop on f5, with a triple attack on White’s queen and both his bishops. White loses a piece, no matter which way

he plays. Don’t underestimate the power of the Mamba. If we know it and our opponent doesn’t, we wield a terrible weapon. Game 56 K.Griffith-C.Lakdawala San Diego (rapid) 2012 Tournaments tend to be infested with gifted kids, such as Kyron Griffith, who was still a mid-teen in those days. Today, Kyron’s rating is close to 2500 and he has IM norms, so I’m glad I played him when he was still a kid! 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Nd6+ Ke7 Can you find a single vice which is missing from Black’s position? White’s knight has been allowed into the taboo d6square, and now Black loses castling privileges. 8 Nf5+ Kf8 9 Be3 Bb4 10 Qf3? White’s mistake is due mainly to having never seen this variation before. So another strong player gets befuddled by the 6 … Bc5 line and plays the same move Bruce Baker played in the first game of this chapter. Amazing. 10 … d5! Although this is a lazy slap, more than an actual blow, the move is still strong and it is White who fights from a position of inferiority from here on. 11 exd5 Qxd5 12 Nh4?

When we enter a sharp opening line without knowing it well – and I promise you that not a single one of your opponents will know this line well – we awaken a monster. We may kill the enemy but, far more likely, we may get killed. Kyron is worried about entering an ending where Black can inflict a split kingside with … Qxf3, so he completely misplaces his f5-knight and lands in even greater problems. As mentioned previously, 12 Ng3 is the only move. 12 … Bg4?! Lacking a killer instinct. White is flat out losing if Black had kept queens on and played 12 … Qa5! 13 Bd2 Bg4 14 Qg3 Rd8 15 f3 Rxd2! (I actually saw this line but grossly underestimated Black’s attack) 16 Kxd2 Be6 17 Bd3 Nd5 18 Rab1 f6 19 Rhc1 Bxc3+ 20 bxc3 Nxc3 and White can resign. 13 Qxd5 Nevertheless, this ending is also in Black’s favour as White’s queenside pawns are about to get smashed. 13 … Nxd5 14 Bd2 Nxc3 15 bxc3 Bc5

15 … Be7 gains a tempo. 16 Bd3 Rd8 17 Rb1 17 0-0?? hangs a piece to the simple 17 … e4. 17 … b6 Now 17 … e4? is overly hasty. After 18 Bxe4 Re8 19 f3 b6 20 Kf1 Be6 21 Nf5 Bc4+ 22 Ke1 g6 23 Bh6+ Kg8 24 Ne3 Bxe3 25 Bxe3 f5 26 Bxc6 Rxe3+ 27 Kd1 Rxc3 28 Kd2 Ra3 29 Rb4 Be6 30 Ra4 Rxa4 31 Bxa4 White should save the game. 18 0-0?! 18 Be3 was better.

Exercise (critical decision): White’s last move allows 18 … e4. Should Black play it, or should he unravel with 19 … f6 - ?

18 … f6?! There it is again, my obsession with remaining safe, when going for it would have ended the game. Answer: I should have taken the bait and played 18 … e4! 19 Bxe4 Rxd2 20 Bxc6 Rxc2 21 Nf3 g6 with a won ending for Black, who has: a rook on the seventh, pressuring both a2 and f2; the bishop pair in an open position; and the superior pawn structure. Furthermore, the sleeping h8-rook will soon be activated after … Kg7. 19 Rfe1 Kf7 20 Be3 Bxe3 Exchanges tend to help the side with the superior structure, which in this case is Black. 21 Rxe3 Be6 22 a4 After 22 c4 g6 23 c5? bxc5 24 Rb7+ Rd7 25 Rxd7+ Bxd7 26 Nf3 Rb8 and White is a pawn down and busted. 22 … Ne7! Threatening 23 … Nd5 with a double attack on e3 and c3. 23 Ree1 Nd5 24 c4 Nf4 25 a5 White is desperate to dissolve his weak queenside pawns, even at the cost of one of them. 25 … Nxd3 26 cxd3 bxa5 27 Rb7+ Rd7 28 Rxd7+ Bxd7 Not only is Black an outside passed pawn ahead, his bishop is also clearly superior to White’s remaining knight. 29 Ra1 29 Rb1 is met by 29 … Bc6, when White’s rook is denied entry. 29 … a4 30 Rb1 Bc6

Here I missed the stronger 30 … a3! 31 Kf1 Rb8! (removal of the guard) 32 Ra1 (taking the rook allows the a-pawn to promote) 32 … Rb3 33 Ke2 Rb2+ 34 Ke3 a2 and White is paralysed. 31 Kf1 Rd8 32 Ke2 Rd7! Planning to challenge White’s most active piece with … Rb7. 33 Nf3 Rb7 34 Rxb7+ 34 Ra1 Rb3 35 Nd2 Rc3 36 f3 Ke6 is also hopeless for White. 34 … Bxb7 35 Ne1 a3 36 f3 g5! Black fixes the white pawns on the same colour squares as his bishop, making them more vulnerable. 37 h3 a2 38 Nc2 h5 39 Kd2 h4! Same principle. White is vulnerable to a king invasion via e6, f5, f4 and g3.

40 Kc3

Exercise (combination alert): Low on the clock (as usual), I missed an immediate win for Black: 40 … f5 Still winning. Answer: Even stronger is 40 … g4! with a brute force breakthrough. I saw this move as well but incorrectly rejected it. After 41 fxg4 (41 hxg4 Bxf3! wins since the h-pawn will promote) 41 … Bxg2 42 Kb2 Bxh3 43 Ne3 Black has the trick 43 … Bg2! and wins, which I missed in my analysis. 41 Kb2 g4 42 Ne1 If 42 fxg4 Bxg2 43 Ne3 then 43 … fxg4! 44 Nxg4 Ke6 45 Nf2 Kf5 46 Kxa2 Kf4 47 c5 Ke3 48 Ng4+ Kd4 49 Nf2 Kxc5

wins easily. 42 … gxh3 43 gxh3 Kf6 44 Kxa2

Black’s king will waltz over and chop the h3-pawn and there is nothing White can do. 44 … Kg5 45 Kb3 Kf4 46 Kc3 Kg3 47 d4 exd4+ 48 Kxd4 Kxh3 49 Ke3 Kg3 50 Nd3 h3 51 Nf2 f4+ 52 Ke2 Bxf3+ 0-1 Game 57 B.Stamper-C.Lakdawala San Diego (rapid) 2012 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Nd6+ Ke7 8 Nf5+ Kf8 9 Be3 Bb4 10 Qf3? Hooray! Everyone seems to make the same opening error, no matter what their rating. 10 … d5! 11 0-0-0!?

My opponent, plainly a man of high spirits, wants to mix it up and declares all-out war with this tempting move, which seeks to take advantage of his lead in development. The trouble with this theory is that I had comped this position at home (it’s in Black’s favour) and knew exactly how to proceed. 11 … Bxf5 Getting rid of White’s most dangerous piece. I prefer this to the wilder variation 11 … Bxc3!? 12 exd5 (or 12 Bc5+ Kg8 13 exd5 Nd4 14 Ne7+ Qxe7! 15 Qxc3 Qd8 16 Bxd4 exd4 17 Qxd4 Qd6 and White doesn’t have enough for the piece) 12 … Nd4! 13 Bxd4 exd4! (not 13 … Bxd4??, which loses instantly to 14 Qa3+ Kg8 15 Ne7+ Kf8 16 Nc6+ etc) 14 bxc3 g6! 15 Nh6 (or 15 Nxd4 Bg4 16 Qf4 Bxd1 and White can resign) 15 … Kg7 16 g4 Kxh6 17 Qf4+ g5 18 Qxd4. According to the computer, White doesn’t have enough for the piece. Still, the position looks a bit scary for Black, since h2-h4 is coming. The game continuation is more practical. 12 Qxf5?!

White is trying for too much. After 12 exf5 d4 13 Bg5 Bxc3 14 bxc3 Qa5 Black stands better, but White is in the game. 12 … Bxc3 13 bxc3 Nxe4 14 Bc4 There is no choice. 14 Kb2?? Qa5 15 Rd3 Nd6! is a crushing attack for Black. 14 … Nd6! Double attack. 15 Bc5

Exercise (combination alert): Optically, White’s initiative looks scary. Black must search for the mysterious consolidation. What is it? 15 … Ne7!

agent

which

binds

his

Answer: Double attack again. Now White’s queen really is attacked (twice!), since the pin is broken. The trouble is that White’s bishop also hangs. 16 Qf3!? White continues to hand over material, with no evidence of regret or grief. He goes all in, yet fails to generate enough for the piece. All the same, this move may actually be the best practical chance, since it forces Black to find a series of onlymoves. Instead: a) 16 Qxe5?! Nxc4 17 Qe2 h5! 18 Rhe1 Nd6 19 Qe5 Rh6! sees Black consolidate. b) 16 Bxd6 Qxd6 17 Qd3 Rd8 18 Rhe1 g6 leaves White a pawn down and strategically lost. 16 … Nxc4 17 Rxd5 Qc7 18 Qe4

Exercise (combination alert): The position demands clarification. Find one

strong move and Black can kill White’s attack and consolidate. What is it? 18 … Ke8! Answer: Step 1: Defensive move/unpinning/simplification. By unpinning, White’s d5-rook is under attack. It can’t move, since it must protect the c5-bishop. 19 Qxc4 After 19 Bxe7 Kxe7 20 Rhd1 Rad8 21 Qh4+ f6 22 Rxd8 Rxd8 23 Rxd8 Kxd8 24 Qxh7 Qf7 Black will win easily. 19 … Nxd5 20 Qxd5 Rd8! This forces 21 Qc4, after which his pinned bishop enables a deadly simplification. 21 Qc4 b6 22 Qa4+ Qd7

23 Qxd7+ Kxd7

This move connects Black’s rooks and activates the king for the ending. 24 Be3 Ke6 0-1 Game 58 P.Graves-C.Lakdawala San Diego (rapid) 2008 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Nd6+ Ke7 8 Nf5+ This is second grade soccer, where all the kids chase the ball without any subsequent plan. The majority of my opponents have gone for this position, not realizing that it is just fine for Black. 8 … Kf8 9 Be3 Bb4 10 Ng3

White sees that his knight will be hanging on f5 after Black plays … d7-d5, so he retreats it at the cost of time and the initiative. Black already stands at least even.

Instead, 10 Nd6 was tried in “Gorgonibablo”-C.Lakdawala, Internet blitz 2019. I played my queen to c7 and won a clumsy game. Correct was 10 … Qe7 11 Nf5 Qe6! (otherwise 11 … Qd8 repeats the position) 12 Bd3 d5 13 exd5 Bxc3+ 14 bxc3 Qxd5 15 0-0 Bxf5 16 Bxf5 Qxd1 17 Rfxd1 g6 18 Bh6+ Ke7 19 Bd3 Nd5, reaching a dynamic ending where Black stands no worse. 10 … d5 This move screams out to be played. 11 exd5 Nxd5 12 Bc5+?! This is an example of the cure being worse than the disease. White gives away both the bishop pair and dark squares in an attempt to preserve his structure. He was better off meekly retreating with 12 Bd2 Nxc3 13 bxc3 Be7 14 Bc4 h5! 15 Qf3 (threatening mate in one; 15 Nxh5!? Qd6 offers Black excellent compensation for the pawn, since White’s queenside castling options evaporated when his structure was damaged) 15 … Qe8 16 Nf5 Bxf5 17 Qxf5 g6 18 Qf3 Kg7, when White’s bishop pair and Black’s superior structure balance each other out. 12 … Bxc5 13 Nxd5?! Falling behind in development. Correct was to offer an inferior ending with 13 Qxd5 Qxd5 14 Nxd5 f5, when Black has some advantage, yet not as much as in the game. 13 … Be6 14 Ne3 Qb6

With a double attack on e3 and b2 and a Komodo assessment of -1.79. White is a master who, at the time of the game, was rated 2239, yet he too falls victim to the befuddling power of the Mamba and is already in deep trouble with the white pieces on move 14! 15 Qc1 Forced. 15 … g6! A multi-purpose move: 1. Black’s king will find a relatively safe haven on g7. 2. Once Black’s king reaches g7, the rooks are connected, which is one of Black’s biggest issues in the 6 … Bc5 line. 3) Black plans to roll his kingside pawn majority with … f7f5 in the near future, whereas White’s majority is no threat at all. The computer’s violent choice is 15 … f5!?, but it isn’t necessary when Black already has a strategically won game. Then 16 Ngxf5 Bxf5 17 Nxf5 Bxf2+ 18 Ke2 h6 19 Ne3 Bh4 is

heavily in Black’s favour, but I actually prefer the position I got in the game to this one. 16 Be2 Kg7 17 0-0 Rad8 Here the hasty 17 … f5?? walks into White’s trap. After 18 Nexf5+! Bxf5 19 Nxf5+ gxf5 20 Qg5+ Kf7 21 Qxf5+ Ke7 22 Bg4! White has a decisive attack for the sacrificed piece. 18 c3 Preventing … Nd4 and perhaps preparing b2-b2 to nudge Black’s powerful dark-squared bishop. 18 … Rhf8 I wanted to play … f7-f5 without allowing a double knight sacrifice. 19 Nc4 I was alert to sleazy tricks like 19 Nef5+ Kh8! 20 b4 (20 Qh6 is met simply by 20 … Rg8) 20 … gxf5 21 bxc5 Qxc5 22 Qh6 Qe7 23 Nh5 f6, when Black should consolidate. 19 … Qc7 20 Nh5+!?

I saw this beforehand and wasn’t terribly concerned. White lacks the necessary firepower to deliver either mate or perpetual check, so this may be a waste of time. Then again, I don’t really know what to suggest here for White. For example, after 20 Ne4 Be7 21 Qe3 f5 22 Nc5 Bc8 Black’s centre rolls forward and White remains without targets and without a useful defensive plan. 20 … Kh8 20 … gxh5?? would be dumb since it allows White immediate perpetual check with 21 Qg5+ Kh8 22 Qf6+ Kg8 23 Qg5+ and so on. 21 b4!? Lashing out like this only weakens c3, yet who can blame White, since there are no good options: a) 21 Qh6 f5 22 b4 Bxc4 23 Bxc4 Bd6 24 Ng3 e4 and Black dominates. b) 21 Nf6 Be7 22 Ne4 (22 Qg5?? hangs a piece to 22 … Bxc4! 23 Bxc4 Qd6) 22 … f5 23 Ng5 Bg8 24 Re1 e4 25 Nh3 f4

and White is getting pushed off the board. 21 … Be7 This move conveniently covers against Nf6 ideas. 22 Ng3 f5 White is strategically lost for the following reasons: 1. Black’s kingside pawn majority is obviously more potent than White’s majority on the queenside. 2. Black’s kingside pawns may enable a decisive attack if the e- or f-pawns can push forward to f3 or e3. 3. Black has the bishop pair. 4. Black controls the dark squares. 5. White’s backward c3-pawn may turn into a target after … Bf6 and … e5-e4 ideas. 23 Re1 Bf6 24 Bf1 e4 25 a4?!

White blunders in an already lost position.

Exercise (planning): How does Black win material? 25 … Ne7 Answer: Retreat the c6-knight to e7, after which White’s c3pawn falls. 26 Na3 Qxc3 27 Rb1 Qxc1 28 Rexc1 28 Rbxc1 Bb2 wins more material.

Exercise (combination alert): This one is easy. Black to play and win material. 28 … Ba2 0-1 Answer: The b1-rook is trapped.

Game 59 J.Humphrey-C.Lakdawala San Diego (rapid) 2015 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Nd6+ Ke7 8 Nf5+ Kf8 9 Be3 Bb4 10 Bd3

This takes care of … Nxe4 ideas and reinforces the f5knight. The problem is that the bishop move is passive. Black easily equalizes with the usual thematic break. 10 … d5 There is no reason to hold back with 10 … d6, even if it is also fully playable. 11 exd5 Qxd5 I thought this move came with the dual threat of … Qxg2 and the disconnection shot … e5-e4. I was wrong about the second threat! After 11 … Nxd5 12 0-0 Nxe3 13 Nxe3 Bxc3 14 bxc3 g6 the comp rates the game as even with a balance between White’s

development and Black’s superior structure.

12 Ng3?! My opponent suffers the same hallucination as me and unnecessarily covers against Black’s … e5-e4 non-threat. The comp shows that it can be ignored with 12 0-0!!. The tactics are super-tricky, so a human is unlikely to see through them over the board. Play continues 12 … Bxc3 13 bxc3, when Black must avoid 13 … e4? due to 14 Bxe4! (this stunning shot plays upon Black’s weak back rank) 14 … Qxd1 (14 … Qxe4?? 15 Bc5+ Kg8 16 Ne7+! short circuits Black’s back rank) 15 Bc5+ Kg8 16 Raxd1 Bxf5 (now 16 … Nxe4?? loses to the destabilizing 17 Ne7+!) 17 Bxf5 and Black is in deep trouble, since White’s lead in development lead and bishops in the open position, far outweigh Black’s unblemished structure. Instead, Black should just chop the f5-knight with 13 … Bxf5, when 14 Bxf5 Qxd1 15 Rfxd1 g6 is about even. 12 … Qxg2

Not only does this win a pawn, it also accomplishes the following: 1. It destroys White’s option to castle kingside. 2. It weakens the light squares around White’s king, especially f3, which may be exploited with … Bg4, combined with … Nd4. This obvious move is a clear improvement over 12 … Ng4?, as in M.Baldauf-R.Vidonyak, Neuhausen 2004, where 13 0-0 Bxc3 14 bxc3 Nxe3?! 15 fxe3 g6 16 Rf6! Be6 17 Ne4 Kg7 18 Qf3 would have seen a dangerous attack brewing for White. 13 a3 Bg4 Gaining a tempo while preventing White from castling long. I prefer this to 13 … Bxc3+ 14 bxc3 Be6, even though Black’s extra pawn and superior structure outweigh White’s development lead. 14 Qd2?

The queen should have gone to b1.

Exercise (combination alert): White’s last move allowed Black a combination. How should he continue? 14 … Ba5?! This is still favourable for Black Answer: But I missed the far more favourable 14 … Nd4!, threatening the fork on f3. I saw this move and misassessed just how strong it was for Black. After 15 Bxd4 exd4 16 axb4 Re8+ 17 Nce2 Bxe2 18 Bxe2 Ne4 19 Qxd4, embarrassingly, I overlooked the über-simple fact that 19 … Nxg3 undermines White’s hanging h1-rook. 15 b4 The comp suggests the unnatural 15 Bf1 which no human is going to play. 15 … Nd4! Threatening 16 … Nf3+. Better late than never. It’s still strong, yet not as strong as on the previous move. 16 Bxd4 This concession is forced. 16 … exd4 17 bxa5 dxc3 18 Qxc3 h5! With two ideas: 1. Black threatens to undermine the h1-rook’s support with … h5-h4. 2. Black can later activate the sleeping h8-rook, via h6 or h5. I think this move is more accurate than chasing the white king to safety with 18 … Re8+ 19 Kd2 Qxf2+ 20 Kc1. 19 Qc5+! This way White can play Kd2, while covering both f2 and his h1-rook against … h5-h4. 19 … Kg8 20 Kd2! h4 21 Nf5 Bxf5

I’m happy to eliminate a potentially potent attacker. 22 Qxf5 Rh5!

The great issue with the Mamba is the buried h8-rook. With Black’s last move, the issue is resolved. 23 Qf4? 23 Rhg1 was necessary. 23 … Nd5 The knight is unleashed as an agent of coercion, directed at White’s queen and later (hopefully) his king. It becomes clear that it is White’s king who is in the greater danger, despite the open g-file. White simply doesn’t have the time to use it. 24 Qd4 Qg5+! A dual-purpose move: 1. White’s rooks get disconnected and Rg1 ideas are no longer available.

2. Black’s queen covers d8, in preparation for … Nf4 and … Rd8, going after White’s bishop. 25 Kd1 Nf4 26 Qe3 26 Rc1 Rd8 27 Qe4 Rh6 is also completely lost for White. 26 … Qf6 27 Rb1

Exercise (combination alert): This one is easy. Black to play and win the exchange. 27 … Nd5 This move is a prelude to the position’s coming Armageddon. Answer: Double attack. The human move, which is a simple double attack on the queen and on the fork square c3. Unbelievably, the comp doesn’t want the exchange and actually prefers either 27 … Nxd3 or 27 … Rd8.

28 Qd2 28 Qh3 was the only move, although the game is still hopelessly lost for White. 28 … Qf3+! 0-1 Double attack. A full rook is better than an exchange. Game 60 F.Sieber-Be.Krause Hanover 2015 1 e4 c5 2 Nf3 e6 Don’t get fooled by the 2 … e6 move order. We soon reach the same position. 3 d4 cxd4 4 Nxd4 Nc6 5 Nc3 Nf6 6 Ndb5 Bc5!? 7 Nd6+ Ke7 8 Bf4 e5 9 Nf5+ Kf8 10 Be3 Bb4 11 Bc4

This move makes sense. White develops while reinforcing d5.

11 … d5 Black plays the break anyway. This thrust is made possible by the insecure positioning of White’s f5-knight. Instead: a) 11 … d6 is playable, but unnecessarily passive. After 12 Ng3 Bxc3+ 13 bxc3 h6 14 Qd3 White stood somewhat better in L.Blandon-A.Guasch Figuerola, Sitges 2019 b) 11 … Nxe4?! is greedy and backfires after 12 Qd5 (double attack) 12 … Bxc3+ 13 bxc3 Qf6 14 Qxe4 d5 15 Qxd5 Bxf5 16 0-0, when Black lags in development. 12 Bxd5 12 exd5?? falls for Black’s trap. After 12 … Na5! (double attack; White’s c4-bishop and f5-knight are loose) 13 Bd3 e4! White loses a piece and resigned in E.Lobron-C.Lakdawala, Internet (blitz) 2005. The 6 … Bc5 line takes down a GM in 12 moves, with the black pieces – and to sweeten the deal, I didn’t make a single move on my own, since it was all pre-game computer analysis.

12 … Nxd5

12 … Bxf5 is also playable; for example, 13 Bxc6 Nxe4 14 Bxe4 Bxc3+ 15 bxc3 Qxd1+ 16 Rxd1 Bxe4 17 Bc5+ Ke8 18 00, F.Zulfic-Qing Aun, Adelaide 2017. At first glance it may seem as if Black is in deep trouble, due to the massive development lag, yet after 18 … f6 19 Rfe1 Bc6 20 f4 e4 21 Bf2 Kf7 the computer slightly prefers Black. White’s extra development proves to be useless. If anyone is going to win, it will be Black, due to White’s broken queenside structure. 13 Qxd5 Instead, 13 exd5 Bxf5 14 dxc6 Qxd1+ 15 Rxd1 bxc6 16 0-0 Ke7 was Zhang Jilin-B.Nagy, Budapest 2005. Here 17 f4 can be met by 17 … Bxc2 18 Rd2 Bf5 19 fxe5 Ke6! 20 a3 Bxc3 21 Rd6+ Kxe5! 22 Bf4+ Ke4. La, la la la la. No worries. Black’s king is completely safe, since White simply lacks material firepower to inflict any damage. After 23 bxc3 Rhd8 24 Rxc6 Rac8 25 Ra6 Be6 26 Ra5 Rd5 27 Rxa7 Rxc3 the game is even and likely to be a draw. 13 … Qxd5 14 exd5 Bxf5 15 dxc6 b6!?

This ambitious advance ends theory. The move has two ideas: 1. Black wants to win the c6-pawn, without damage to the queenside structure. 2. With … b7-b6 tossed in, Black will be able to unravel quickly with … Ke7, since White no longer has Bc5+ ideas. Earlier 15 … bxc6 16 0-0-0 f6 17 Na4 Kf7 18 Nc5 Bxc5 19 Bxc5 ½-½ was N.Gavrilakis-A.Mastrovasilis, Greek League 2009. The opposite-coloured bishops and low material count all but ensure a drawn result. 16 0-0-0 White increases her lead in development, while covering the c2-pawn. After 16 f4 Bxc3+ 17 bxc3 Ke7! 18 fxe5 Rhc8 19 0-00 Rxc6 White’s extra pawn is totally meaningless, due to the damage to White’s queenside structure. In fact, I slightly prefer Black. 16 … Bxc3 The knight cannot be allowed to jump into d5. 17 bxc3 Ke7 Here we see the benefit of … b7-b6. White is unable to give a bishop check on c5. 18 Rd5 18 f4 can be ignored with 18 … Rhc8 19 Rd5 Be6 20 Rxe5 f6 21 Re4 Kf7 22 Re1 Bf5 23 Rc4 Be6 24 Re4, when Black can take a repetition draw by moving back to f2, or go for the full point with 24 … Bxa2 25 Bd4 Rc7 26 c4 Bxc4 27 Bxb6 Rxc6 28 Re7+ Kg8 29 Bd4 a5. Black stands no worse. 18 … f6 19 Rhd1?! This move isn’t forceful enough and fails to generate threats. Instead, 19 f4 is met by 19 … Be6 20 Rd2 e4 21 g4 g6 22 g5 Rhc8 with even chances. 19 … Be6?! More accurate is 19 … Rac8! 20 f4 Bg4 21 Re1 Be6 22 Rd2 exf4 23 Bxf4 Rxc6. Black has serious winning chances, due to

White’s infirm structure. 20 Rd6?! White should offer the exchange with 20 c4!, since 20 … Bxd5!? 21 cxd5 Kd6 22 c4 is no worse for White, while 20 … Rac8 21 Rd6 Rc7 22 f4 Rhc8 23 fxe5 Rxc6 24 exf6+ gxf6 is a likely draw. 20 … Rhd8! Black gets serious chances to win, even with all the rooks coming off the board. 21 Rxd8 Rxd8 22 Rxd8 Kxd8 23 a4!? The pawn is more vulnerable on this square. 23 a3 would make it more difficult for Black to create a passed pawn on the queenside. 23 … Kc7 24 Kb2 After 24 a5 bxa5 25 Bxa7 Kxc6 Black may eventually be able to create two passed pawns, whereas White’s doubled cpawns are useless. 24 … Kxc6 25 Ka3 Bd5 26 f3

This is a tough decision. 26 g3 might later allow Black’s king to waltz into the kingside, via the light squares. The plan would be: 1. Provoke White into playing h2-h4. 2. Transfer the black king to g4. 3. Play … h7-h5 and … g7-g5, forcing White to exchange on g5. 4. Play … h5-h4, creating a passed pawn on the kingside, which won’t be easy to stop. 26 … Bc4 Black’s idea is to play … Bf1 and then … Be2, provoking f3-f4. Then Black can respond with … e5-e4, creating a passed pawn. 27 Kb4 Bf1 28 g3 Be2 29 c4? A pawn sacrifice is too aggressive a plan. White cannot survive a 4-to-2 kingside pawn minority.

29 f4 isn’t going to earn the Congressional Medal of Honour for bravery, but it’s still necessary. After 29 … Kd5 30 fxe5 fxe5 31 Bg1 It’s not clear if Black can force a win. 29 … Bxf3 30 c5? Premature. 30 … bxc5+ 31 Kb3 I guess White just realized that 31 Bxc5?? loses to 31 … a5+ 32 Kc4 Be2+, when her king is disconnected from her bishop. 31 … Bd1 32 Kc3 Be2 33 Kd2 Bc4 34 Kc3 Kd5 35 Bf2 a6 36 Be3 g5 Black begins to roll the kingside majority. 37 Bf2 h6 38 Be3 f5 39 Bc1 f4 40 gxf4 gxf4 This way Black immediately achieves two connected passed pawns. 41 Bd2 Bf1 42 Bc1 Bh3 43 Kd3 Bf5+ 44 Kc3 44 Kd2 e4 is equally hopeless. 44 … Ke4 The entry of the king ensures that one of Black’s passed pawns will promote. 45 Ba3 Or 45 Kc4 f3 and the pawn cannot be stopped. 45 … Ke3!

The c5-pawn is meaningless. 46 Bxc5+ Ke2 47 Bd6 f3 0-1 Game 61 C.Milton-C.Lakdawala San Diego (rapid) 2009 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5!? 7 Nd6+ Ke7 8 Nf5+ Kf8 9 Bc4

In this version White doesn’t play Be3. 9 … d6! This quiet move is a superior option here to the normal … d7-d5 theme. White must now factor in … Bxf5 and also … Be6 ideas. Instead, after 9 … d5 10 Bxd5 Bxf5 11 exf5 Nxd5 12 Nxd5 White stands slightly better due to the extra f5-pawn. 10 Bg5 This move was new. White intends to follow up with Nd5. It was too late for 10 Be3, since Black would just chop on f5 with a good position; while after 10 0-0, as in J.FoltysK.P.Richter, Munich 1941, I would opt for 10 … h6 11 Ne3 Be6 and wouldn’t mind playing this as Black. 10 … h6 The alternative was 10 … Bxf5 11 exf5 h6 12 Bh4 Bb4 13 Qd3 Ne7 14 0-0 Bxc3 15 bxc3 Rc8 16 Bxf6 gxf6 17 Bb3 d5 18 Rad1 h5 with roughly equal chances. 11 Bh4?!

This either loses the initiative or forces White to sacrifice unsoundly. Best was to destroy the defender of d5 with 11 Bxf6 Qxf6 12 Nd5 Qd8 13 Nfe3 a6 14 0-0 g6. Optically, White looks better due to the domination of d5, but Sveshnikov aficionados will know better. Black simply plays around the piece. Black has the bishop pair, dark-square control and his king is relatively safe after … Kg7. The position seems balanced. 11 … g5 12 Nxh6?

I suspect White knew that this sacrifice was unsound, but did it anyway in the hope of creating confusion in a rapid game. Objectively, White should resign himself to strategic inferiority by backing down with 12 Bg3 Bxf5 13 exf5 Bb4 14 0-0 Bxc3 15 bxc3 Ne7 16 Bd3 Kg7. 12 … Rxh6 12 … gxh4? is dumb and allows White right back into the game after 13 Nxf7 Bxf2+! 14 Kf1! Qb6 15 Nxh8 Qc5 16 Qd3 Kg7 17 Na4 Qd4 18 Qxd4 Bxd4 19 Nf7 Nxe4. This ending is an unclear mess.

13 Bxg5 Rg6 14 Qd2 14 h4 is met by 14 … Qb6 (with a double attack on f2 and b2) 15 0-0 (15 Qd2 Bd4! 16 Bb3 Bxc3 17 Qxc3 Nxe4 is hopeless for White) 15 … Qxb2 16 Nd5 Bg4 17 Qd3 Nxd5 18 Bxd5 Nb4 19 Qg3 Nxd5 20 Qxg4 Nf6 21 Qf5 Kg7 and Black consolidates. 14 … Bxf2+?! At the board I originally thought this shot ruins things for White. I failed to appreciate the potency of opening the f-file for his attack and also the loss of time incurred by the move. Black should still be winning, but it would have been wiser – at least from a practical perspective – to continue in consolidation mode with 14 … Bb4!, threatening 15 … Nxe4!. After 15 Bd3 Nd4 16 h4 Ne6 White’s attack sputters into non-existence.

15 Qxf2 Not 15 Kd1?? Bd4 16 Rf1 Bxc3! 17 bxc3 Nxe4! 18 Rxf7+ Ke8 19 Bxd8 Nxd2 and White is down material and busted.

15 … Rxg5 16 0-0-0 Opening the f-file for White’s attack was not a great idea. Black must play carefully now. 16 … Be6? Short on time, I predictably blunder and allow White a combination. The careful 16 … Rg6! 17 Rhf1 Nd4 18 h3 Be6 19 Bxe6 fxe6 20 g4 Kg7 was correct, when Black should consolidate. 17 Bxe6 fxe6

Exercise (combination alert): There is a shot which allows White back into the game. What should he play? 18 Rhf1? Oh, thank goodness.

Answer: By this point I was frozen with terror, having seen 18 Rxd6! Qe7 (forced) 19 Nd5! (19 Rd7?! fails to 19 … Rxg2!) 19 … Qxd6 20 Qxf6+ Ke8 21 Qxg5 exd5 22 Qg8+. For example, after 22 … Qf8 23 Qe6+ Kd8 (23 … Qe7 24 Qg8+ repeats) 24 exd5 Nd4 25 Qxe5 Qf2 (threatening mate on c2) 26 Qd6+ Ke8 27 Qg6+ Kd7 28 c3 Qf4+ 29 Kb1 Qf5+ 30 Qxf5+ Nxf5 31 Rf1 Nd6, White has four pawns for the piece and it’s anybody’s game to win or lose. The computer calls it even. 18 … Rg6 Now Black is winning again. 19 Rd3?! The initiative is not a renewable resource. Allow it to die and it likely will never recharge. The rook lift is too slow. The best option was to push on with 19 g4! Kg7 20 g5 Rxg5 21 Rxd6! Qe7 22 Nd5 Nxd5! 23 exd5 Nd4! 24 h4 Rc8 25 c3 Rf5 26 Qg2+ Kh8 27 Rxf5 Nxf5 (27 … exf5?? loses to 28 Rh6+) 28 Rxe6 Qxh4, when White is losing but still gets practical chances. 19 … Qe7 20 Qh4 White isn’t interested in winning a useless pawn at the cost of his attack. After 20 Nb5 Rd8 21 Nxa7 Nxa7 22 Qxa7 Kg7 Black should win the game. 20 … Kg7 21 Rh3!

Exercise (critical decision): Should Black play 21 … Rf8 0 ? If not, is there a stronger move? 21 … Rg8! Answer: White set up a trap on his last move and it’s critical to play the rook to g8, not f8. 21 … Rf8?? walks into 22 Rxf6! and, shockingly, Black has no good way to recapture and must resign in disgrace. 22 g4!? White’s war devours resources at a staggering rate. This is a desperate attempt to open lines. However, nothing works at this point. 22 Rxf6?? fails miserably with Black’s rook posted on g8. After 22 … Rxf6 23 Qh7+ Kf8 Black consolidates. 22 … Rxg4!

White can make no use of the open g-file. 23 Qh6+ Kf7 24 Qh5+ R8g6 25 Rg3 Or 25 Rhf3 Rf4! 26 Rxf4 exf4 27 Rxf4 Ne5 and White’s attack goes dead. 25 … Rxg3 26 hxg3 Kg7 27 Qh4 Ng4 28 Qh1 Nd4 29 Kb1 Qg5 30 a4 Rh6 31 Qg1 Qe3! 0-1 Game 62 A.Costello-C.Lakdawala San Diego (rapid) 2015 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Nd6+ Ke7 8 Nc4

A change from 8 Nf5+ in all the games so far. This move is Komodo’s top choice. 8 … d6 9 Be2 Or 9 Ne3 Kf8 (Black’s king moves, just to deprive White of the pleasure of saying “check” when a knight lands on d5) 10

Ned5 h6! (it’s crucial to prevent Bg5) 11 Bc4 (11 Be3 can be ignored with 11 … Be6! 12 Bxc5 dxc5 13 Bc4 g6 14 0-0 Kg7; from my experience in such structures, White’s “domination” of d5 is purely decorative and Black stands just fine) 11 … a6 12 Qf3!? (this feels artificial; more natural is 12 0-0 Be6 13 Be3 Bxe3! 14 fxe3 Ne7! 15 Qf3 Bxd5 16 Nxd5 Nexd5 17 Bxd5 Qe7 with dynamic equality; this is a rare case when Black can get away with opening the f-file for White) 12 … Nd4 13 Qd3 b5 14 Bb3, T.Rost-J.Pietrasanta, French Team Championship 1996. Here Black played his knight to g4 and won after a wild adventure. Simpler and stronger is 14 … b4 15 Ne2 Nxb3 16 axb3 Nxd5 17 exd5 (if 17 Qxd5 Be6 18 Qd3 a5 19 Be3 Qc7, I slightly prefer Black) 17 … g6 18 Be3 Bxe3 19 Qxe3 Kg7. It’s going to be a battle of majorities and I like Black’s chances in the coming fight, due to bishop over knight. 9 … h6 The idea is to prevent Bg5 to retain the f6-knight, one of the key stewards of the d5-square. 10 0-0 Be6 11 Ne3 Rc8 12 Ncd5+ Bxd5?

This is an exception to the Sveshnikov rule of plugging d5 with a white pawn if possible. The problem is that Black is unable to halt White’s queenside majority, which begins to surge. Black is fine after 12 … Kf8!. 13 exd5 Nb8 14 c4 Nbd7 15 Rb1 a5?! Black should have played 15 … Bxe3 16 Bxe3 Re8 17 b4 b6 with an inferior yet still playable position. 16 a3 g6 17 b4 axb4 18 axb4 Bd4 19 Nc2! Ba7? 19 … Bb6 was forced. 20 Na3!

The awful truth is I’ve been outplayed strategically by a 12year-old (he is an IM today!). I totally missed this idea. Black is busted. 20 … Bb6 21 Bd3?! This wastes time. 21 Nb5! Ne8 22 Bb2 is stronger. 21 … Kf8 22 Nb5 e4?

Clearing e5 for the knight, but too ambitious. 22 … Qe7 was forced. 23 Be2?! The comp bashes out 23 Nxd6! exd3 24 Nxc8 Qxc8 25 c5 and claims a decisive advantage for White. Certainly Black’s position looks terrible. 23 … Ne5 24 Qb3 Kg7 25 Bb2 Re8 26 Qg3!

Exercise (critical decision): Calculate 26 … Rxc4. Should Black play it? 26 … Qe7? This passive response is hopeless. Answer: Black can still make a fight of it by sacrificing a full rook with 26 … Rxc4!! (this is absolutely forced) 27 Bxe5 Rxe5 28 Bxc4 e3!; for example, 29 Rb2 Ne4 30 Qf3 exf2+ 31

Kh1 Qh4 32 g4 h5! and White’s position is not so easy to defend. The comp only gives White an edge, despite the extra rook! 27 Bxe5! White’s queenside pawn majority is about to surge. 27 … dxe5 28 d6 Qe6 29 c5 Bd8 30 Rfd1 Ra8 31 d7?! After 31 Qe3 Qf5 32 Nc3 Ra3 33 Qc1 Ra8 34 Bb5 Rf8 35 Re1 Black’s e4-pawn is a goner and it really is time to resign. 31 … Re7 32 Nd6!

Now 32 … Rxd7?? loses miserably to 33 Bc4, since 33 … Qe7 34 Nf5+ pops the queen. As I sat, contemplating resignation, a wonderful thing happened: ½-½ My opponent offered me a draw! In what should have been my only loss with this line, I escaped. Alex was only 12 at the time and had never drawn or beaten a titled player before. I quickly accepted from this wretched position.

Game 63 A.Witt-C.Lakdawala San Diego (rapid) 2015 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 We now move on seventh move options where White decides not to check on d6 with the knight. 7 g4? White dispenses with small talk and formalities and gets right into the battle stage. This Keres Attack thrust doesn’t work out well at all against the Mamba.

7 … d6 8 h3 8 g5? invites the black knight exactly where it wants to go. After 8 … Ng4 9 Nxd6+ Qxd6 10 Qxd6 Bxd6 11 h3 Nd4! 12 Bd3 Nh2! (White doesn’t regain the sacrificed piece after all) 13 Rxh2 Nf3+ 14 Kf1 Nxh2+ 15 Kg2 Nf3 16 Kxf3 Bxh3 Black is up a full exchange.

8 … a6 9 Na3 This knight is destined to remain on the periphery for the remainder of the game. 9 … b5 10 g5!? The nature of a wish is that it is generally for that which cannot be. This actually makes matters worse. I don’t blame White for rejecting the cringeworthy 10 Nab1. 10 … b4 11 gxf6 11 Na4 Bd4 12 c3 Nxe4 13 cxd4 bxa3 14 Bg2 d5 is also awful for White. 11 … Qxf6!

Oh, the bliss of threatening a vulgar mate in one! Zwischenzug. Black is in no hurry to regain the lost piece. 12 Qd2 The queen blocks the c1-bishop. He was afraid of future … Nd4 attacks if he placed the queen on e2.

12 … bxc3 12 … bxa3! is actually stronger, since it allows the knight access to d4. 13 bxc3 With a structure this awful, White’s position is a slave to time’s whip. Virtually all endings will result in disaster for him. 13 … Qf3 Double attack on h1 and e4. 14 Rg1 Qxe4+ 15 Be2 Be6 16 Qd3! White is willing to enter an ending a pawn down, but only if Black fixes his structure. After 16 Rxg7 Nb4! 17 Bb2 (17 cxb4 Bxb4 18 c3 Qh1+ 19 Bf1 Bxh3! wins) 17 … Qh1+ 18 Bf1 Nxa2!, … Rb8 is coming and White’s position collapses. 16 … Qxd3?!

As usual, I was too eager to enter a pawn-up ending, and allowed my opponent a concession by undoubling his once broken pawns. Stronger was 16 … Qh4! (attacking f2) 17 Qg3 Qxg3! (now is the right time to exchange) 18 Rxg3 with a hopeless ending for White, who is a pawn down and supports five isolanis. 17 cxd3 g6 18 Bg4 White has no useful plan and so hopes to attain future targets by provoking the black central pawns forward. 18 … f5 19 Bf3 d5 20 Nc2 e4!? Somewhat rushed. 20 … Kd7 is more useful since Black can play … e5-e4 at any time. 21 Be2 Kf7 21 … Bd6! prevents Bf4. 22 d4 Given the chance 22 Bf4!, seizing control over b8, is more accurate. 22 … Bd6 23 Ba3 Bxa3 24 Nxa3 f4 Now h3 is under attack. 25 h4 h5 26 f3 White wants to play Bd3. The high cost is to allow Black a choking, protected passed pawn on e3. 26 … e3 27 Bd3 Ne7 28 Rb1 Rhb8 29 Ke2

Exercise (planning): Come up with the fastest plan to convert. 29 … Nf5! Answer: Play … Nf5!, forcing White to exchange, so that Black seizes control over b1. 30 Bxf5 Bxf5 31 Rxb8 Or 31 Rb3 Kf6 32 Rc1 Rxb3 33 axb3 Rb8 34 b4 g5 35 hxg5+ Kxg5 36 Rg1+ Kf6 37 Rh1 Rc8 38 Rc1 Rg8!, when Black infiltrates and wins. 31 … Rxb8 0-1 Nothing can be done about the coming … Rb2+. Game 64 Ro.Bruno-C.Lakdawala Los Angeles 2005

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 Ron also lives in San Diego and we had played many times before this game, so I knew he would come heavily theoretically armed against my usual openings. Therefore in this game I tried the 6 … Bc5 line. I made a good choice since Ron’s eyes widened in shock after my sixth move. 7 Bg5

This is one of White’s better lines against 6 … Bc5. After 7 Nd5?! 0-0 I have caught more than one person in online blitz games with the trap 8 Bg5? (8 Nbc3 Nxd5 looks fine for Black, no matter how White recaptures on d5) 8 … Bxf2+!, and if 9 Kxf2?? then 9 … Nxe4+ etc. White has to play 9 Ke2 Bc5 10 Bxf6 gxf6 11 Qd2 Kg7 with some, albeit not enough, compensation for the missing pawn; e.g. 12 Kd1 a6 13 Nbc3 Rg8 14 Be2 Kh8. Instead, L.Liaskos-N.Galopoulos, Kallithea 2008, continued 10 Qe1?! Be7 11 Nxe7+ Qxe7 12 Qh4 Qc5! 13 Nc3 Ne8! and White had no compensation whatsoever.

7 … 0-0! 7 … h6 8 Be3! would give Black an extra … h7-h6 on the 7 Be3 of Game 66, but simply castling is better than that. 8 Qf3! White’s most logical move, pressuring f6 and intending Nd5, while preparing to castle queenside. Not 8 Nd5? Bxf2+!, which is just a permutation of the same trap. Instead, M.Veljkovic-G.Galiot, Belgrade 2010, saw 8 Qd2 a6 9 Bxf6 Qxf6 10 Nc7 Rb8 11 N7d5 Qd8 12 h4 b5 13 Ne3 Bd4 14 Ncd1?! (this is clunky; White should just go for it with 14 00-0 Ne7 15 Bd3 d6 and the comp says even, while I slightly prefer Black’s chances) 14 … Ne7 15 h5 Bb7 16 f3 d5, when Black already had a strong initiative, development lead, the bishop pair and dark-square control. 8 … Be7 8 … a6? is too slow. White ignores it with 9 Nd5!, since 9 … axb5 10 Nxf6+ gxf6 11 Bxf6 Qa5+ 12 c3 Rd8 13 Qg3+ Kf8 14 a4 is pretty much just winning, P.Lehtinen-A.Siikaluoma, correspondence 2007. 9 Bxf6 As in regular Sveshnikov lines, White eliminates the defender of d5, at the cost of weakening the dark squares. The comp gives White an edge, while I am comfortable with Black. Instead, 9 0-0-0 may be White’s best move, as in G.HalvaxK.Neumeier, Austrian Championship, Pinkafeld 2015, which continued 9 … a6 10 Bxf6 Bxf6 11 Nd6. Here Black should play 11 … Be7 (threatening … Nd4) 12 Nf5 Bc5 13 h4 g6 14 Ne3 d6 15 h5 Qg5, when Black has a good grip on the dark squares. 9 … Bxf6 10 Nd5 Be7 Covering the d6-square. Not 10 … Bg5?! 11 h4! and if Black tries to hang on to the dark-squared bishop with 11 … Bh6? then 12 Nd6 Nd4 13 Qd1 puts him in deep trouble, since c2- c3 follows and Black experiences serious difficulties in developing the queenside. 11 Nbc7!?

White gets tempted by the allure of the d5-square and begins to fall behind in development. Instead: a) 11 c3 d6 12 0-0-0 Be6 13 Kb1 a6 14 Nbc7 Rc8 15 Nxe7+ Qxe7 16 Nd5 Qd8 17 Qe3 Ne7 and Black stood at least equal, V.Ivanov-S.Gromov, correspondence 2013. b) 11 Nxe7+! Qxe7, G.Ruben-M.Ajoy, correspondence 2003, and now 12 0-0-0 may be White’s last hope for an edge. I intended 12 … d6! as I can’t allow his knight to roost there. When in a strategically inferior situation, we have little choice but to ignite a crisis. After 13 Nxd6 Be6, I hoped my open c-file and attacking chances would compensate for my lost pawn. 11 … Rb8 12 Nxe7+ Nxe7! The prime directive in the Sveshnikov: Black must fight for eventual control over d5. 13 Qc3!? White dithers when he should be completing development. He is probably still okay in the ending after 13 Nb5! d5! 14 exd5 Qxd5 (14 … Nxd5 15 0-0-0 Qg5+ 16 Kb1 Nf4 17 Qe3 is also fine for White) 15 Qxd5 Nxd5 16 0-0-0 Be6 17 Bc4 Ne3! 18 fxe3 Bxc4 19 Nd6 Be6 20 b3, as the entrenched knight on d6 is impressive. Personally, I still slightly prefer Black, but the comp calls the ending even. 13 … d5! 14 exd5?! White had to try 14 Qxe5! Nc6 15 Qg3 dxe4 16 c3 (preventing … Nb4) 16 … Bg4! 17 Bc4, when he stands only slightly worse. 14 … Nxd5 15 Nxd5 Qxd5

White is now in deep trouble for the following reasons: 1. He lags in development, with his king stuck in the middle of the board. 2. He can’t play Rd1, since this hangs his a2-pawn. 3. He can’t move his bishop, since this hangs g2. 4. He must watch out for a queen check on e4, which either wins the g2- or c2-pawn, or forces his king to move to the open d-file. 16 Qe3 After 16 Qf3 e4 17 Qf4 Be6 18 Be2 Rbc8 19 c3 Rfd8 20 0-0 Qd2 21 Qxe4 Qxb2 22 Rab1 Qxc3 23 Qxb7 Bxa2 24 Ra1 Rc7! Black hangs on to the extra pawn. 16 … Bf5 17 c3 This is a forced weakening of d3, which acquires great significance later on. 17 … Rbd8

18 Rg1! Freeing his bishop at the cost of eliminating kingside castling as an option. 18 f3 e4! 19 f4 Bg4 20 h3 Bd1! is also unpleasant. 18 … Bg6 Clearing the path for the f-pawn to move forward. 18 … Bc2 is perhaps even better. 19 g4! A strong defensive move, hoping to discourage … f7-f5, even at the cost of opening the f-file. Whereas after 19 b3 Rd7 20 Be2 Qc6 21 Rd1 Rxd1+ 22 Bxd1 Rd8 (threatening … Rd3) 23 Be2 Bb1!, White is in deep trouble. 19 … f5!? Opening the f-file has to be beneficial for Black. I also considered probing for weakness along the third rank with the rook lift 19 … Rd6. 20 gxf5 Bxf5?!

This automatic recapture reduces Black’s advantage. Instead, 20 … Bh5! threatens mate and after 21 Be2 Bxe2 22 Qxe2 Rxf5 the major piece ending is rough for White, since his king will never feel safe. 21 Be2 Kh8 22 h4 h6 23 h5 Bh7 24 b3 He is sick and tired of his a1-rook babysitting his a-pawn. The trouble is that with every pawn push, White’s queenside becomes more vulnerable. 24 … b6?! 24 … Qa5! leaves White struggling to unravel. 25 Rd1! Qe6 Now after 25 … Qa5 26 Rxd8 Rxd8 27 Qg3 Rd7 28 Kf1 and it isn’t so easy for Black to improve his position. 26 Qg3?! White unravels with 26 Rxd8! Rxd8 27 Rg3, when it’s getting harder and harder for Black to claim more than an edge. 26 … Rxd1+ 27 Bxd1 Qf6 28 Be2 Rd8 29 a4? Lashing out unnecessarily. White gets excellent drawing chances if he gives up his a-pawn with 29 Qe3! Bb1 30 Rg3! (30 a4? Bc2 31 c4 e4 leaves White in deep trouble) 30 … Bxa2 31 Rg6 Qf4 32 Qxf4 exf4 33 Bd1 Bb1 34 Rc6, when Black will have a very tough time converting his extra pawn. 29 … Bc2 30 Bc4?! Giving up a pawn with 30 b4 was necessary. 30 … e4! Black threatens … Bd3, cutting off the white queen’s protection of c3.

31 Ke2 Bd3+ With the elimination of bishops, Black’s rook infiltrates at d3. 32 Bxd3 Rxd3 33 Qg4 After 33 Qb8+ Kh7 Black threatens … Qf3+ and then mate on d1. If then 34 Rg3 Rf3! 35 Rxf3 Qxf3+ 36 Ke1 Qxc3+ 37 Kf1 Qh3+ 38 Ke1 Qh1+ 39 Kd2 Qxh5 40 Qxa7 Qd5+ Black will win the queen ending. 33 … Qxc3 34 Kf1 The e-pawn is taboo, as 34 Qxe4?? Qc2+ 35 Kf1 Rd1+ wins White’s queen.

Exercise (planning): We come to a crossroads. The plans are: a) playing directly for mate with 34 … e3; or b) removing queens from the board with 34 … Qc1+, followed by 34 … Qg5, entering a pawn-up rook ending. One wins easily, while the other allows White to resist. Which one would you play? 34 … Qc1+?! Of course I jump at the chance to simplify, when I could have knocked out my opponent. Answer: Plan a) 34 … e3! wins; e.g. 35 Qf3 (threatening perpetual check) 35 … Qxb3 (planning to meet a back rank check with … Qg8) 36 Rg3 Rd1+ 37 Kg2 exf2! and White is down too many pawns.

35 Kg2 Qg5? Black should reverse intent with 35 … Qc6! 36 Kh2 Qf6 37 Qxe4 Qxf2+ 38 Rg2 Qe3 39 Qxe3 Rxe3 40 Rb2 Re5, when h5 falls and White finds himself down two pawns in the rook ending. 36 Qxg5 hxg5 37 Kh2? 37 Re1! offers White chances to hold the game; e.g.37 … Rd4 38 a5 bxa5 39 Rc1 Kh7 40 Rc5 Kh6 41 Rxa5 Rd7 42 Re5 Rb7 43 Rxe4 Rxb3 44 Ra4 Rb7 45 Ra5 Kxh5 reaches one of those annoying rook endings where one side is up two pawns and it’s still a draw. 37 … Rxb3 Now Black is winning again. 38 Rxg5 Ra3 39 Re5 Rxa4 40 Kg3 Kh7 41 Re6 Rb4 42 Kf4 a5 The e-pawn remains safe, since a king and pawn ending is hopeless for White. 43 h6 e3+!? Simply 43 … g6! is the easiest path to the win. 44 Kxe3 gxh6 45 f4 h5 46 Re7+ Kg6 47 Re6+ Kg7 48 Re5 h4 49 Rh5 h3

50 f5 50 Rxh3?? hangs the rook to the obvious 50 … Rb3+. 50 … Kf6 51 Kd2 Rb3 52 Kc2 a4 0-1 Game 65 P.Hodges-C.Lakdawala San Diego (rapid) 2015 1 e4 c5 2 Nf3 Nc6 3 Nc3 Nf6 4 d4 cxd4 5 Nxd4 e5 6 Ndb5 Bc5 7 Be2

My opponent thought for an immense period of time for a G/45, using up around 10 minutes for this safe and apparently new move. He obviously wanted to hop the knight into d6, but feared preparation. This is another good reason to use the Mamba as an ambush weapon. Close to 100% of the time, we totally throw our opponents out of their sphere of opening preparation. His move is an attempt to return to civilization. 7 … d6 8 Be3 0-0 9 0-0!? Tempting Black to double the e-pawns and open the f-file. The d6-square would then be under attack and a retreat with … Ne8 is forced. Instead, 9 Bxc5 dxc5 10 Qxd8 Rxd8 11 Nc7 Rb8 is completely equal. After 12 Rd1, I was intending to play 12 … Be6!? 13 Nxe6 fxe6, leading to a structure which arises in the next game. Note that 12 N7d5? would allow an undermining combination: 12 … Nxe4! 13 Nxe4 Rxd5 14 c4 Rd8 15 Nxc5 b6 16 Ne4 Nd4 17 Rc1 Bb7 18 f3 Bxe4 19 fxe4 Ne6 20 b4 a5 21 a3 Rd4 and White is in trouble in the ending. .

9 … Bxe3!? Black is willing to go passive in turn for a superior structure. 10 fxe3 Ne8 11 Bc4? 11 Nd5 should be played, so that the b5-knight is allowed a vacancy on c3. Here the chances are approximately balanced. 11 … a6 12 Na3 Be6 This eliminates the pressure on f7. Also tempting is 12 … b5 13 Bd5 Bb7 14 Nab1 Qb6 and Black stands better. 13 Bd5 Rc8 14 Qf3 Swapping bishops with 14 Bxe6 fxe6 hardly improves; e.g. 15 Rxf8+ Kxf8 16 Qg4 Kf7! 17 Rf1+ Nf6 18 Qh3 (threatening 19 Qxh7) 18 … b5 (Black responds with an even bigger threat) 19 Nd1 Ke7 20 c3 Nb8 21 Nf2 Nbd7 22 Qg3 Qg8 and White lacks compensation for his lousy structure. Komodo already has Black up by over one point. 14 … Nf6 15 Rad1?

Exercise (combination alert): My opponent, by now in extreme time pressure, misses a simple tactic. How does Black win material? 15 … b5?? I would throw a fit if one of my students overlooked a simple win of the exchange. Grossly overconfident, I made my move instantly, even though I had plenty of time on my clock. Then an instant later, I groaned mentally and kicked myself. Answer: The obvious 15 … Bg4 16 Qf2 Bxd1 wins easily. 16 Ne2?! He had to try 16 Bxe6 fxe6 17 Nab1 Na5 18 Rc1 with a miserable, though still playable position. 16 … Bxd5?! I was flustered at missing 15 … Bg4 on the previous move and, as a result, compounded the mistake with another one. Stronger was 16 … Bg4 17 Qf2 Ne7 18 c3 Qb6 19 Rd3 Nexd5 20 exd5 Ne4 21 Qe1 f5 with a wretched position for White. 17 exd5 Ne7 18 e4?! This is suicide in the guise of an attack. Another timepressure error. 18 c3 was forced, although Black stands clearly better after 18 … Qb6 19 Rf2 e4 20 Qh3 b4. 18 … b4 19 Nb1 Rxc2 The queenside pawns begin to fall. 20 Ng3 The conspicuous absence of a white attack means there is no way to avert the inevitable. Similarly, after 20 Rd2 Qc7 21 Ng3 Rxd2 22 Nxd2 Qc2, more white pawns fall. 20 … Rxb2 21 Nf5 Nxf5 22 Qxf5 Qc8! 0-1 Game 66 B.Baker-C.Lakdawala

San Diego (rapid) 2009 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 Bc5 7 Be3!

My opponent, National Master Bruce Baker, fell victim to a miniature in the first game of the chapter. This time he comes heavily armed with the computer’s “refutation”. Here Komodo 13 has White up by almost a point, so this one is going to be tough. However, I was prepared for this line and had previously tested it in blitz, surviving and even winning multiple games against titled players. Black’s game is playable but only if played perfectly. In many lines, the comp assesses the situation favourably for White, but in practical play Black has excellent chances. My feeling is that the comp’s favourable assessment for White is something of an illusion and that in reality Black is fine – as long as you master the mechanics of the coming position. 7 … d6!

The thing about comping sharp opening lines is that they stop the guesswork. This is Black’s only move. Black must enter a slightly inferior but playable ending. Not 7 … Bxe3? 8 Nd6+ Kf8 9 fxe3 and Black’s dysfunctional position is already lost. 8 Bxc5 Declining the exchange of queens fails to test Black. For instance, if 8 Qd2 0-0 9 0-0-0, as in A.Suetin-V.Osnos, USSR Championship, Leningrad 1963, then 9 … Be6! 10 Nxd6 Bd4 looks to give Black excellent play for the pawn. 8 … dxc5 9 Qxd8+ Kxd8 10 0-0-0+ After 10 Bc4 Be6 11 Bxe6 fxe6 12 Nd6 Kc7 13 Ncb5+ Kb8 14 f3 a6 15 Na3, Black was at least equal in K.KahlH.Buchmann, correspondence 2013. 10 … Ke7! I always prefer this move to 10 … Nd4? 11 f4! Bg4 12 Rd2 exf4 13 Nxd4 cxd4 14 Rxd4+, as in M.Proof-M.Deren, correspondence 2006, as even after 14 … Ke7 15 e5 Nd7 16 Nd5+ Kf8 17 Nxf4 Nxe5 18 Nd3! Nxd3+ 19 Bxd3, I just don’t trust Black’s ability to survive this ending so far behind in development and with the h8-rook out of play. 11 Nc7 Other lines are: a) 11 Nd5+ Nxd5 12 exd5 Nd4 13 d6+ Kd8 14 Re1 f6 15 f4! Nc6 16 fxe5 Nxe5. I have won multiple times against titled players in blitz from this position. The idea is that Black will slowly complete development and, who knows, maybe later win the d6-pawn. b) 11 Bc4 Rd8 12 Rxd8 Nxd8 13 Rd1 Be6 14 Nd5+ Bxd5 15 exd5 a6 16 d6+ (I love it when White pushes the d-pawn to d6, since the majority of the time I win it later in the game) 16 … Kd7 17 Nc7 was A.Cioara-N.Giffard, Le Touquet 2002. I like Black’s game after 17 … Rc8 18 f3 Ne6 19 Nxe6 fxe6 20 a4 Rc6, as White’s d-pawn falls and if anyone is going to win, it will be Black. 11 … Rb8 12 N3d5+

We should also look at 12 N7d5+ Nxd5 13 exd5 (13 Nxd5 transposes below) 13 … Nd4 14 Ne4 Bf5! 15 d6+ (or 15 f3 Bxe4! 16 fxe4 f5! and if 17 c3 fxe4 18 cxd4 exd4 Black has plenty of compensation for the piece) 15 … Kd7 16 Nxc5+ Kc6 17 Nd3 Kxd6 18 c3 Nc6 19 Nb4+ Ke7 20 Nd5+ Ke6 21 Bc4 Na5 22 Be2 Rhd8 and Black is just fine. The practical problem is that you need to know this analysis. If you don’t, you will sink under White’s lead in development. 12 … Nxd5 13 Nxd5+ Or 13 exd5 Nd4 14 c3 Nf5 15 Nb5 Nd6 16 Nxd6 Kxd6 and Black already stands slightly better, since he is the only one with a mobile pawn majority. Again, I have won blitz games from this position. 13 … Kf8 14 Ne3 14 c3 Be6 is also fine for Black. 14 … Be6 15 Bc4 Ke7! A new move and an improvement over 15 … Bxc4? 16 Nxc4 Ke7 17 Rd5, when Black was already in deep trouble, D.Dochev-F.Lindgren, Umel 1997.

16 Bxe6 16 Nd5+ is met by the cool 16 … Kf8!, intending … g7-g6 and … Kg7, while after 17 Ne3 I planned to repeat moves with 17 … Ke7. 16 … fxe6! Principle: An visually ugly structure isn’t really weak, if our opponent’s pieces lack access to its underbelly. 17 c3 b5 I didn’t think Black was worse here and offered a draw. However, a draw would have bumped Bruce out of the money. So he declined and played on. 18 Kc2 Rhd8 19 Rxd8 Rxd8 20 a4! White’s only chance to play for the win is to retain one pair of rooks on the board and play on the open a-file. 20 … a6 21 axb5 axb5 22 Ra1 Kd6 23 Nf1!?

23 Rd1+ Kc7 24 Rxd8 Nxd8 is a near certain draw. White will not be able to exploit Black’s doubled e-pawns without rooks on the board. 23 … Kc7 24 Ng3 c4 25 Nh5 g6!? I didn’t want to let his rook into my camp after 25 … Rf8 26 f3 Rf7 27 Ra8, even though the comp now calls it dead level. After 27 … Kb6 28 Rg8 g6 29 Rg7 Rf8! 30 Ng3 Ra8! 31 Rxh7 Ra1 Black stands no worse, but I wanted no part of such a sharp line, when my goal was merely to evade a loss. 26 Ng7!? This is an insanely risky gamble to say the least, since he self-traps his knight. Objectively, this move must be good for Black. Yet White has suddenly created a position where Black can mess things up and lose, since the complexity level has risen considerably, with both clocks running low. Instead: a) 26 Nf6?? loses to 26 … Rf8! 27 Ng4 h5, when f2 falls and White is dead. b) 26 Ng3 is technically best, but after 26 … Kb6 a draw is the likely outcome. 26 … Kd7 Here I missed 26 … Kd6! 27 Ra6 Rf8 28 f3 h6! 29 Kc1 (if 29 Nxe6 Re8!, his knight must sacrifice itself on d4, since it gets trapped on g7 by … Re7!) 29 … Rc8! 30 Rb6 Rg8! 31 Nxe6 Re8 32 Rxb5! Rxe6 33 Rd5+ Kc7 34 Rc5 Kb6 35 Rxc4 with good chances to convert. 27 Ra6 Rc8 Not 27 … Rg8??, which is met by the overloading trick 28 Nxe6!. 28 h4 He desperately wants to free his knight. 28 … h5 29 f3?! He had to try 29 Rb6 Rb8 30 Ra6 Nd8 31 g4! hxg4 32 Kd2 Ke7 33 h5 gxh5 34 Nxh5 Nb7, but it leaves Black with all the winning chances.

29 … Ke7 30 Rb6 Kf6 Inaccurate, though not technically an error as Black can reverse course. After 30 … Kf7! 31 Rb7+ Ne7 32 Rxb5 (32 g3 Rg8 traps the knight) 32 … Kxg7 33 Rxe5 Kf6 Black is winning. 31 Rb7 He should try 31 Ne8+!, forcing Black to find 31 … Ke7 32 Ng7 Kf7! as above. 31 … Rg8 32 g4! hxg4 33 fxg4 Ne7! 34 Nxe6

At the board I was terrified of the line 34 g5+ Kf7 35 h5! (for a moment I thought maybe Black was busted, until I found 35 … gxh5! (35 … Rxg7? 36 h6 Rg8 37 Rxb5 is at least equal and White may even win) 36 g6+! Kxg7 37 Rxe7+ Kxg6 38 Rxe6+ Kg5! 39 Rxe5+ Kg4 40 Rxb5 h4. I calculated the line to this point; my clock was too low to invest any more time, but I felt intuitively that Black must win the race. And indeed he does: e.g. 41 b3 h3 42 bxc4 h2 43 Rb1 Kh3 44 Rh1 Rg1 45 Rxh2+ Kxh2 46 c5 Kg3 47 c6 Kf4 48 Kd3 Rg6 and so on.

34 … Kxe6 35 Rb6+ Kd7 36 Rxb5 Kd6 37 Rb4 Kc5-+ 38 Rb7 Nc6 39 Rh7 Rf8! 40 Rg7 Rf2+ 41 Kb1 Rh2 42 Rxg6 Rxh4 43 g5 Dropping his e-pawn, but 43 Rg7 Rh1+ 44 Kc2 Re1 45 g5 Rxe4 46 g6 Rg4 47 Kd2 Kd6 48 Kc1 Rg1+ 49 Kd2 Ne7 would win anyway. 43 … Rxe4 44 Rg8 Rg4 45 g6 Ne7 0-1

Index of Complete Games Baker.B-Lakdawala.C, San Diego (rapid) 2005 Baker.B-Lakdawala.C, San Diego (rapid) 2009 Barrientos Chavarriaga.S-Fluvia Poyatos.J, Badalona 2005 Braakhuis.W-Rodriguez.J, Correspondence 2001 Brodsky.M-Kramnik.V, Kherson 1991 Bronnikova.E-Kosintseva.T, Russian Women’s Team Championship 2006 Brooks.M-Nakamura.H, US Championship, St Louis 2009 Bruno.R-Lakdawala.C, Los Angeles 2005 Bucinskas.V-Ivanchuk.V, European Cup, Fügen 2006 Caruana.F-Carlsen.M, Grenke Classic, Karlsruhe/Baden 2019 Caruana.F-Carlsen.M, World Championship (Game 12), London 2018 Caruana.F-Carlsen.M, World Championship (rapid playoff), London 2018 Caruana.F-Mamedyarov.S, Shamkir 2016 Costello.A-Lakdawala.C, San Diego (rapid) 2015 De Firmian.N-Shabalov.A, Virginia Beach 2007 Efimov.I-Eljanov.P, European Championship, Ohrid 2001 Gallagher.J-Ivanchuk.V, European Cup, Saint Vincent 2005 Ganguly.S-Prasad.D, Hyderabad 2000 Graves.P-Lakdawala.C, San Diego (rapid) 2008 Griffith.K-Lakdawala.C, San Diego (rapid) 2012 Hadid.O-Liascovich.L, Pereyra Puebla 2006 Hodges.P-Lakdawala.C, San Diego (rapid) 2015 Hracek.Z-Mista.A, Czech League 2006 Humphrey.J-Lakdawala.C, San Diego (rapid) 2015 Inarkiev.E-Gelfand.B, Nazran (6th matchgame) 2017 Iordachescu.V-Yu Yangyi, FIDE World Cup, Baku 2015 Jakovenko.D-Shirov.A, Foros 2007 Kantans.T-Matlakov.M, Gibraltar 2017 Karjakin.S-Carlsen.M, FIDE Grand Prix, Abidjan (rapid) 2019

Karjakin.S-Carlsen.M, Shamkir 2019 Kindermann.S-Van Wely.L, German Bundesliga 1998 Kramnik.V-Van Wely.L, Wijk aan Zee 2005 Leconte.J-Balabaev.F, Correspondence 2002 Leenhouts.K-Hausrath.D, Dutch League 2004 Li Ruofan-Zhang Zhong, Manila 2007 Liu Guanchu-Yu Yangyi, Chinese Team Championship 2015 Luther.T-Leko.P, Essen 2002 MacKinnon.K-Shabalov.A, Calgary 2007 Magem Badals.J-Moiseenko.A, Spanish League 2010 Milton.C-Lakdawala.C, San Diego (rapid) 2009 Molinero Martin.J-Kuijf.M, Sitges 2003 Nikac.P-Komarov.D, Niksic 2000 Pähtz.E-Eljanov.P, Isle of Man 2016 Palciauskas.V-Bennborn.J, Correspondence 2007 Pattard.T-Kalinitschew.S, Guben 2003 Pauli.A-Kalinitschew.S, Fürth 1998 Ponomariov.R-Kramnik.V, Wijk aan Zee 2005 Prassas.G-Halkias.S, Ano Liosia 2000 Ruiz Symmes.A-Borowiec.A, Correspondence 2003 Sanal.V-Kuzubov.Y, Turkish League 2019 Savory.R-Gormally.D, EU Championship, Liverpool 2006 Schmitt.H-Pirs.M, Correspondence 2013 Sieber.F-Krause.B, Hanover 2015 Stamper.B-Lakdawala.C, San Diego (rapid) 2012 Svidler.P-Carlsen.M, Grenke Classic, Karlsruhe/Baden 2019 Szabo.B-Lenic.L, Hungarian League 2018 Tari.A-Zhigalko.S, European Rapid Championship, Katowice 2017 Tukhaev.A-Carlsen.M, World Rapid Championship, St Petersburg 2018 Van Foreest.J-Carlsen.M, Wijk aan Zee 2019 Versili.M-Standke.W, Correspondence 2007 Volokitin.A-Oleksienko.M, Ukrainian Championship, Rivne 2016 Volokitin.A-Shirov.A, Poikovsky 2008 Wang.J-Le Quang Liem, Burlingame 2019 Wilczek.T-Lafarga Santorroman.D, Correspondence 2012

Witt.A-Lakdawala.C, San Diego (rapid) 2015 Zelcic.R-Saric.I, Stari Mikanovci (rapid) 2010