The Wisest Things Ever Said About Chess E-Book

Andrew Soltis

• Herausgeber: Batsford
• Kategorie: Lebensstil
• Sprache: Englisch

Seitenzahl: 403

Veröffentlichungsjahr: 2013

Beschreibung

• 'The best opening is the opening your opponent doesn't know.' • 'The winner of the game is the player who makes the next-to-last mistake.' This fascinating book contains 300 of the most astute insights on chess ever uttered, culled from three centuries of great players. Each of these invaluable maxims is illustrated with an annotated chess position, making the book a short cut to learning from the masters. These snippets of wisdom are arranged into chapters for easy reference: Calculation, Intuition, Strategy, Position Evaluation, Openings, Sacrifices, Attitude, Endgames, Mistakes, Studying, Time Management and Tournament Tactics. This is a great book to dip in and out of – every page contains a nugget of wisdom that will help you hone your own chess skills and win your next game.

Leseprobe

First published in the United Kingdom in 2008 by

B T Batsford 10 Southcombe Street London W10 0RA

An imprint of Anova Books Company Ltd

Copyright © B T Batsford 2008

Text copyright © Andrew Soltis

The moral right of the author has been asserted.

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, mechanical, photocopying, recording or otherwise, without the prior written permission of the copyright owner.

First eBook publication 2013 eBook ISBN: 978 1 8499 4102 0

Also available in paperback Paperback ISBN: 978 19063 8800 3

This book can be ordered direct from the publisher at the websitewww.anovabooks.com

CONTENTS

Introduction

Chapter 1: Attack

Chapter 2: Calculation

Chapter 3: Defense

Chapter 4: Endgame

Chapter 5: Evaluation

Chapter 6: Mistakes

Chapter 7: Move Selection

Chapter 8: Openings

Chapter 9: Pawns

Chapter 10: Pieces

Chapter 11: Psychology

Chapter 12: Sacrifice

Chapter 13: Strategy

Chapter 14: Studying

Chapter 15: Tactics

Chapter 16: Technique

Chapter 17: Tournament Tactics

Index of Quoted Players

INTRODUCTION

Each new generation of chessplayers discovers ways to play the game better. What we don’t appreciate is how much of this is rediscovery. We rediscover what was simply forgotten. When Vladimir Kramnik said calculation is “far more important in defense than in attack” it was a profound insight. Kramnik indicated it was also original. It wasn’t. Rudolf Spielmann – and, most likely, others before him – expressed the same idea in similar words (as in # 39).

This is a book about ideas like that. Some are well remembered, others have been forgotten. These thoughts are expressed in maxims, aphorisms, wisecracks, kibitzes and proverbs that have been passed down to us. Some can be called “rules.” That’s a word that divides players. Amateurs love rules. Masters love to ridicule them. “A chessplayer’s talent is measured not by his knowledge of the rules but his ability to find exceptions to them,” wrote Viktor Korchnoi. “Only the exceptions matter,” said Alexei Shirov. But GMs follow rules like everyone else. The occasions when they do so vastly outnumber the times that they break rules, as a look at any grandmaster game plainly shows.

Rules and maxims should not be used as a substitute for thinking. But they are wonderful as prompters to thinking. Good rules point in the right direction. They “should be used only as mnemonics, reminders of what to think about,” as Kenneth Mark Colby, an avid amateur player and respected psychiatrist, wrote. One of the great chess teachers, Cecil Purdy, got it right. “The true art of the teacher lies in stating rules in memorable form,” he wrote.

I’ve tried to collect the wisest of sayings expressed in the most memorable forms. That doesn’t mean the earliest form, and this is not a book of forensic etymology. In many cases the collective wisdom of chess has come to us as folk wisdom, of unknown origin. In many other cases, the first way a thought was expressed was improved by others. Wilhelm Steinitz wrote in The Modern Chess Instructor, “In the ending…it is generally best to place the rook behind the pawns in order to not obstruct their advance. But when fighting against hostile pawns, it is most advisable to attack them in the rear or to stop the one furthest advanced in the same manner.” This is an extremely useful thought, and may very well have passed through other minds before Steinitz’s. But the version we remember is Siegbert Tarrasch’s succinct “A rook belongs behind a passed pawn.”

Many if not most of the great sayings like that are sweeping generalizations. That’s part of the reason we remember them. When Tarrasch wrote, “There is no move that does not weaken some part of the position,” what was remarkable is that he could preface it with a single exception – “Except for the mating move…” He admitted he was joking, several pages later in The Game of Chess, when he offered another rule without an exception: “Never move a pawn and you will never lose a game.”

Wit, like exaggeration, is a common feature of great sayings. But there are many familiar witticisms that you won’t find in these pages because they have little application that will help you improve your game. As much as I enjoy sayings like “A draw, like love, requires two participants” (Georg Marco) or “Every chess master was once a beginner” (Irving Chernev), to name a few, I didn’t feel they belonged in a book devoted primarily to advice.

The wise sayings I chose, and in some cases paraphrased, vary widely in form and content. Some just give you something to think about (“A castled rook is already developed,” #178 and “A knight endgame is a pawn endgame,” #68). Others provide a plan of action (“With a pawn on e5 White must attack the king,” #4 and “You’re never mated with a knight on f8,” #36).

Others tell us what not to do (“Never look for other possibilities when you have one satisfactory winning line,” #271). Still others try to puncture your prejudices (“There is no such thing as a winning move,” #97 and “A passed a-pawn looks more dangerous on the second rank than on the seventh,” #91).

One can only marvel at the richness of wisdom that has been handed down to us. Our task is to build on it. But first we have to remember it. Lev Polugayevsky said the benefit of “book” knowledge in the opening is to enable players to explore at move 15 without having to rediscover what to do at move two. “Most of all, it’s necessary to use book to avoid inventing the bicycle,” he wrote, according in 64 magazine, July 2001. That’s true of all aspects of chess skill. Remembering will save us from having to invent a lot of bicycles.

CHAPTER ONE:ATTACK

1

Modern chess is much too concerned with things like pawn structure.

Forget it. Checkmate ends the game.

Many wise sayings by masters apply only to master chess. But with these words Nigel Short was reminding players of all strengths of the object of the game. You don’t need to find 50 moves of finesse if you can end the game by the caveman method at move 25.

When mate is possible in the near future, it takes precedence over everything else. Both players follow that simple priority in the following game.

Anand – Ftacnik

Biel 1993

Black to Play

You could spend hours trying to find the best move in such a chaotic position if you consider criteria such as strategic goals, pawn structure and the esoteric stuff. But mate simplifies the move-selection process enormously. Natural moves that have nothing to do with delivering mate, such as 1…e5, must be viewed suspiciously – or not at all. (In this case 1…e5? 2 xd5! exd4 3 xd4 is needlessly dangerous.)

Black correctly played 1…xb2!. He would be the one with better mating chances after 2 xb2 a4+ 3 c1 c3 and 4 d3 e5!.

White hardly looked at 2 xb2 because he, too, understood that in this position only one thing mattered. “White should just ignore everything except mate itself and just hack away,” he wrote in his best-game collection. He won brilliantly after 2 fxe6!.

He spent only 30-35 minutes for the entire game. His task was aided by looking mainly at lines such as 2…xd1 and finding the way to mate after 3 exf7+ xf7 4 xd1, e.g. 4…e4 5 xe4 dxe4 6 c4+.

2

The first essential for an attack is – the will to attack!

So said Savielly Tartakower in 500 Master Games of Chess. Let’s face it, attacking is more fun than defending. But some players are reluctant to attack if it means burning bridges. At certain moments, bridges must be put to the flame or else the flame goes out.

Gongora – Perea

Las Tunas 2003

White to play

White has a clear positional advantage. But the center is blocked, his best piece is about to be kicked back with …e6 and the kingside, while porous, is not easily penetrated.

But White had the will to attack. He knew he had to search for a target. He found 1 b4! axb4 2 axb4. This may seem bizarre because he is ruining his own king’s pawn protection. Yet it is the only good plan because the queenside is the only area White can open quickly. He would be much better in the endgame that follows 2…c6 3 b3 – threatening c3 – 3…a4 4 b6.

Black tried to launch his own attack, 2…a7?. It can’t succeed because of White’s space edge – 3 b2! c6 4 a1 b8 5 b6 d8 6 b5 d7 7 d4.

The threats include 3a3-a8. Black tried the desperate 7…d5!? 8 xd5 e5+ 9 c3 e6 and resigned after 10 xb7! xh2 11 d3.

White’s use-it-or-lose-it decision follows another of chess’ fundamental rules:

3

The player with advantage must attack.

Wilhelm Steinitz was famous for making rules. Some became the foundation of how we play the game. Others couldn’t even be followed successfully by Steinitz himself.

One of his most famous rules – what we can call Steinitz’s Law – is based on the premise that sound attacks stem from positional advantages. Your attack cannot succeed if you don’t have an edge to justify it, he said.

This was brilliant and original. It replaced the older view that attacks succeeded or failed because of the attacker’s genius or lack of it. Steinitz could have stopped there. But he added a guide to action: The player who obtains that edge not only has the right but the obligation to attack. If he doesn’t, his advantage is bound to evaporate.

van Wely – Ki. Georgiev

Groningen 1997

Black to Play

Black has just opened the c-file with …dxc4? and that made White’s pieces much better than Black’s. On 1…b8 he obtains a powerful passed d-pawn with 2 d5! xc4 3 bxc4.

But what happens if Black plays 2…exd5 instead? Logic indicates White should still be better but there is no purely positional way to prove it. Steinitz’s answer is White must attack – 3 g4!.

Then 3…dxc4 4 xf6+ is deadly (4…gxf6 5 xf6 and g5+ or h6, and 4…h8 5 c2).

In the game, Black avoided all this with 1…e7 instead of 1…b8. But 2 d5! again converts White’s advantage into an attack. For instance, 2…xd5 3 g4 with ideas of xg7 or h6+, e.g. 3…h5 4 xg7 xg7 5 h6+ g8 6 xd5 and f6+.

Vladimir Kramnik said the only one who managed to violate Steinitz’s Law successfully was Anatoly Karpov. Karpov didn’t always attack when his position improved; he found other ways to enhance his chances, Kramnik said on e3e5.com. “In my opinion, there were no players before or after him who were able to do this,” he added.

4

With a pawn on e5

White must attack the king.

The Russians credit Alexander Alekhine with this rule: a pawn on e5 gives White not only an incentive but a duty to go for the king. The pawn denies Black the f6 square for a knight. Often that means f7 can be attacked by a knight on g5 and rook on the f-file or that h7 can come under fire from a bishop at d3 and a queen at h5. Most successful xh7+ sacrifices occurred with a White pawn on e5.

A pawn at e5 protected by another pawn also provides strategic benefits. Early writers used to advise White, when he has pushed his pawn to e5 in a French Defense, to maintain it there at all cost because of its positional plusses. Alexander Petroff – of the Petroff Defense – wrote in 1824, “When several pawns form a diagonal chain one must maintain the one that stands at the head.”

Alekhine – Euwe

Amsterdam 1936

White to play

Alekhine found himself forced to attack: Black has taken aim at his only targets, e4 and d4, and is threatening …dxe4. White defended with 1 e5 and that convinced him to go after the Black king.

Play went 1…h7 2 f5 f6 3 g4! fxe5 4 xe5 f6 5 d3 followed by g1 and f3-f4. Of course, the pawn on e5 didn’t guarantee the attack’s success. In fact, White lost. But 1 e5 made kingside attack the best plan.

5

The king makes his neighborhood vulnerable.

This may sound obvious today. But Evgeny Znosko-Borovsky was contradicting Steinitz when he wrote in The Middle Game in Chess:

“Wherever the king is placed, the position in his vicinity becomes automatically weaker, for his presence attracts enemy attacks.”

In contrast, Steinitz believed a king was capable of defending itself and needed no security blanket. Castling makes that wing stronger, not weaker, because the king is there, he wrote.

There are plenty of Steinitz losses to show that he really believed this nonsense. But on other occasions, Steinitz violated his own rules and won.

Steinitz – Tchigorin

World Championship 1892

White to play

Steinitz the theorist would say to himself, “White can’t attack because he has no advantage and because Black’s king makes the kingside stronger.”

Yet Steinitz the player chose 1 h4!, launching an attack that succeeded in spectacular fashion, after 1…e7 2 h5 d5 3 hxg6 fxg6 4 exd5 xd5 5 xd5 xd5 6 b3!.

The end was dramatic, 6…c6 7 e2 d7 8 e3 h8 9 0-0-0 ae8 10 f1 a5? 11 d4! exd4 12 xd4 xd4 13 xd4! xd4 14 xh7+! xh7 15 h1+ and mates (15…g7 16 h6+ f6 17 h4+).

In reality, the presence of the Black king on the kingside, plus the …g6 weakening justified 1 h4!, as Znosko-Borovsky would have said. Z-B’s principle can also be illustrated by the following game:

Svidler – Kramnik

World Championship 2007

White to play

White chose a natural plan, directed at h7 and f7 – 1 h4! d7 2 d5 c6 3 f5, with the idea of g5. Black in turn went after the queenside, (3…c4), and a sharp middlegame followed.

But let’s imagine that the king positions are reversed. We’ll replace 0-0-0 by White with 0-0. For Black, we’ll suppose he castled queenside (K on c8, Q on d7, QR on d8).

Then White would ignore the kingside and prefer the queenside attack with d5 and xa7 or a5. Black would not have attacking chances of his own and would have to find counterplay elsewhere.

Of course, pawn structure also plays a role in determining how successful an attack will be. But Philidor was overstating the case when he wrote, “Pawns alone form the basis of attack and defense” (#155). Pawns delineate the avenues and supply lines of attack. But the king’s location provides the incentive for attack.

Even if the king’s neighborhood seems rock solid, his presence makes it less so. Compare it with banks. They have more security than candy stores. But for some reason robbers target banks more often than candy stores.

6

True attacks are realized only with many coordinated pieces and rarely with one or two.

This should be apparent to anyone who has played chess for more than a few months. Yet it was considered a remarkable insight when Francois-Andre Philidor wrote it in L’analyse du jeu des Echecs in 1749.

His successors offered rules to ensure coordination of the attacker’s pieces. “Never commence your final attack until the queen’s rook is in play,” Joseph Blackburne wrote in his game collection. An illustration of complete coordination is:

Karpov – Polugayevsky

Candidates match 1974

White to play

The pressroom grandmasters were going into rapture over 1 xf6 gxf6 2 h6 and 3 e3, with the threat of g3+-g7 mate.

But White preferred to attack only after he trained all five of his pieces on the kingside, 1 f4! a8 2 f2 ad8 3 g3 c3 4 f3 c2 5 df1 d4 6 h6 c6 7 f5!.

By then Black was doomed, e.g. 7…e5 8 xg7! xg3 9 xg3 and mates.

This doesn’t mean you cannot attack unless you coordinate all of your pieces against the king, as Anatoly Karpov did here. His longtime rival Garry Kasparov attacked brilliantly when he could employ most, but not all, of his pieces. What both Karpov and Kasparov agreed on was that an attack was doomed to failure if only one or two pieces were enlisted in it.

7

When you have opposite attacks going, the quantity of pieces is often more important than their quality.

One of the apparent contradictions in chess is this: A successful attack means applying superior force at a target. But an attack may need to be launched by a sacrifice. By definition, a sacrifice reduces the size of the attacker’s force. How then can a smaller army defeat a greater army?

Garry Kasparov helped clear up the matter, at least in regard to Exchange sacrifices, with #7, which he said about this position.

Movsesian – Kasparov

Sarajevo 2000

Black to Play

Black is several moves away from castling, connecting rooks and coordinating all of his forces. Yet he played 1…xc3! 2 bxc3 c7.

The fact that the “quality” of his remaining pieces was less than White’s, to the tune of the Exchange, didn’t matter because Black soon had more in quantity in the target area: 3 e2 e7 4 g5 0-0 5 h4 a4 6 c1 e5 7 h5 d5.

White’s rooks would be bystanders to the winning Black attack after 8 exd5 xd5. Instead, White tried to use his heavy pieces with 8 h2 d6 9 h3.

But he lacked an open line there, 9…xd3 10 cxd3 b4! 11 cxb4 c8 12 a1 dxe4 and Black was winning. A cute finish would have been 13 fxe4 xe4! 14 dxe4? e5+ 15 d4 xd4+ 16 xd4 xc1+! and mates.

8

Three pieces is a mate.

This adage, which has been attributed to several authors, tries to quantify Philidor’s “true attacks” rule, #6. Three attacking pieces working in tandem usually succeed against a castled king.

Yes, it’s an exaggeration. Castled kings defended by three pieces can withstand a lengthy assault by three attackers. It’s a mismatch – more attackers than defenders – that makes an attack succeed. The greater the mismatch, the better the chance of success.

In the following position all but one of the heavy pieces are engaged on the queenside, where Black threatens to liquidate into a won endgame after …xc6. Yet White won in five moves by exploiting his opponent’s inability to shift pieces to the other wing:

Adams – Morozevich

Sarajevo 2000

White to play

When White pushed his pawn to c6 earlier, he foresaw that it would be lost. But he figured that when Black encircled it, his kingside would be ripe.

He looked at 1 g3 xc6. But he concluded that 2 xf7, with its threat of xg7 mate, wouldn’t be clear at all after 2…g5!.

Instead he chose 1 h1! xc6 2 xc6 xc6 3 g1! and it all became simple. With four attacking pieces, the kingside mismatch is bound to make an attack on g7 or f7 work.

The rest was 3…xb2 4 xg7! c3 5 xe8 Resigns. (No better was 4…xg7 5 xe6+ or 3…g6 4 e4! dxe4 5 h6.)

9

With bishops of opposite color, the player with the attack has an extra piece.

This maxim of the Soviet school makes two points. First, the presence of the bishops rewards the player who seizes the initiative. “Some like to attack, others to defend and there are some worshippers of maneuvering strategy,” Vladimir Simagin wrote in Shakhmaty v SSSR in 1967. “However, to achieve success in play with bishops of opposite color, a player is obligated to fight for the initiative.”

The second point is that such an initiative will be directed at squares that your bishop can attack and which the enemy bishop cannot defend. This is illustrated by the following game, which Russian annotators dubbed “The White Square Symphony.” White had a positional edge earlier and realized it could be exploited only by attack. He began with 17 consecutive moves on light squares. Then:

Karpov – Kasparov

World Championship 1985

White to play

As Gerald Abrahams said of opposite-colored Bs, “They operate each in a different diocese, and ignore each other.” Here White ignored Black’s bishop and began the final assault with more light-square moves, 1 c4+ f8 2 h7!.

To meet the threat of 3 g8 mate, Black played 2…f7. Then 3 e6! set up another light-square threat, e8 mate.

Black resigned after 3…d7 4 e5! because he couldn’t cover b8 as well as the light squares (4…e7 5 f4+ f7 6 b8+ e8 7 d6+ e7 8 xf7+ xf7 9 g6+ and mates).

10

A ‘bad’ bishop is much better in attack than in defense.

Philidor was the first to write about the significance of good and bad bishops. He appreciated that good bishops were excellent for attack. But “he held the incorrect opinion that from a purely defensive point of view the bad bishop could render better service than the good,” wrote Max Euwe.

Euwe took a more modern view that the real difference between good and bad bishops lay in attack. “A good bishop is one which is the same color as the squares on which the enemy pawns are placed,” wrote Ludek Pachman in The Middle Game in Chess, “whilst a ‘bad’ bishop is one which is restricted by its own pawns.”

This view assumes a bad B will be a bad attacker. But that’s not necessarily true if there are bishops of opposite color. When Boris Gulko wrote #10, in American Chess Journal (issue 2, 1993), he cited this case:

Kamsky – Kasparov

Manila 1992

Black to Play

Both sides have a slightly bad bishop. Black’s is limited by the pawn at d6 but otherwise enjoys great firepower. White is worse because his own bishop can’t attack anything, thanks to that pawn on d5.

If those pawns were moved back a rank, to d4 and d5, “Black’s bishop would become ‘good’ but useless,” Gulko said. White’s bishop would have something to attack but remains hopeless in defense.

But as it stands, Black’s bishop confirms #9. After 1…a7 and …a8, White’s bishop and heavy pieces cannot defend the key squares, a1 and b2.

White tried to create a light-square fortress with 2 c1 a8 3 b3. But Black shifted the assault to other dark squares, 3…f4 4 c2 e7! 5 d3 c5 6 b1 e3 7 d4 a2+ 8 d1 xf3! and won.

11

When the opponent has no counterplay, patience is the attacker’s greatest asset.

This comment, from John Nunn in Secrets of Grandmaster Chess, underlines the difference between quick and slow-developing attacks. Quick attacks are more like combinations. They achieve a knockout in a few moves and can often be calculated in advance. Slow-developing attacks occur when you can pin the enemy down but can’t break through immediately. Victory usually comes from shifting the attack from one point to another, while denying counterplay. That requires patience.

Svidler – Sakaev

St. Petersburg 1996

White to play

White is playing “with an extra piece” thanks to the opposite-colored bishops. But his natural target, g7, is well defended. White must rely on other winning ideas, such as getting the queen to g6 and perhaps h7.

At same time he should try to watch out for …e5! when it means that xe5 will permit …xc5+ and when fxe5 allows Black counterplay on the f-file.

White took simple precautions – 1 h3 d7 2 h2! e8 and then 3 e2! g8 4 g4 d7 5 e5!. Threats such as g6xh6 couldn’t be averted for long – 5…e7 6 f6 f7 7 g6 f8 8 e5.

The threat is h7 followed by d6+ or h8+. The rest was 8…g8 9 xh6 e7 10 g4 b5 11 g6 e8 12 g5! (idea: h5/h7+) xf4 12 xg7+ xg7 14 xg7+ f8 15 xb7 a4 and Black resigned.

12

Too many checks means there’s something wrong with the attack.

This comment from Viktor Korchnoi seems odd to young attackers. They love to check because the consequences are easier to foresee. They know the old chestnut – “Always check. It may be mate” – isn’t quite true. But they enjoy making forcing moves anyway.

However, there are natural limits to the number of desirable checks. “The king should not be checked to death or it may escape alive,” as W.H.K. Pollock said. Restricting the king serves better than chasing it.

Gelfand – Kramnik

Candidates match 1994

White to play

White sacrificed a bishop on h7 and Black managed to create an escape route for his king with his last move, …fe8. White responded with the obvious checks and added a pawn to his compensation, 1 h7+? f8 2 h8+ e7 3 xg7.

But something was clearly wrong with the attack because Black’s queen could have defended with 3…f4!, based on 4 f3 xf3! 5 gxf3 g8.

The correct way to finish off was 1 h5!, threatening mate on h8, and then 1…f8 2 g5! g8 3 g3!. Now none of Black’s pieces are allowed to defend the king. After 3…g6 4 h3 there is no way to stop mate on h8.

In that example White violated another cardinal rule of attack. He missed the knockout. As Leonid Stein put it:

13

The most important thing in our business is an ability to deliver the final blow.

It’s not enough to correctly identify when to attack, to choose the right target, to act with speed and accuracy and to limit the number of checks. You also have to finish your opponent off when you get the chance.

Anand – Kamsky

Dortmund 1992

Black to Play

Black had already missed three wins as he drove the king from g1 to c5.

Here he has another chance to put White away, 1…d7! and …c8+ since 2 xd7 xd7 is easy. But Black chose 1…d5+? and after 2 b4! a5+ 3 a3 d3 4 b2 b4 5 a1! White’s king is safe.

In fact, White even had winning chances after 5…a4 6 e2 e4 7 d1 bxc3? 8 xc4+! with the idea of g4+ before a draw was agreed. As Gerald Abrahams had said, “Good positions don’t win games. Good moves do.”

CHAPTER TWO:CALCULATION

14

The essence of chess is seeing the move after.

Good players are good predictors. They can often see the move after – that is, foretell how their opponent will respond to the move they are considering playing.

During Boris Spassky’s first World Championship match, former champion Mikhail Botvinnik asked him, “Do you always guess Petrosian’s moves?” “Not always,” Spassky conceded. He lost that match and realized one reason why was he was often surprised by Tigran Petrosian’s moves. In their second match, three years later, he said, “I guessed everything,” and won the match.

van der Wiel – Karpov

Amsterdam 1988

White to play

Black’s last move (…d7) made a threat of …a4, skewering queen and rook. But it also cut off a key escape route for Black’s queen. White felt confident when he chose 1 d3 because it attacks the queen and gains time for the 2 c2 defense.

But he didn’t see that the obvious reply, 1…a5, would prevent 2 c2 because 2…a4 3 b3 xc3! leads to a won endgame for Black.

Instead, White settled for 2 b4 xb4 3 cxb4. But his weakened queenside, 3…fd8 4 b1 a4 (5 f4 xb2 or 5 b3 c3), eventually cost him the game.

It was Gerald Abrahams, in Technique in Chess, who said #14. But good calculation requires efficiency, not just foresight:

15

Masters don’t calculate more. They calculate better.

“If anything, grandmasters often consider fewer alternatives” than non-grandmasters, said British GM David Norwood. Extensive research on how masters choose moves, by master/psychologist Adrianus de Groot and others, tends to support this.

The alternative moves we consider playing are called candidate moves. Grandmasters are more efficient calculators because they intuitively ignore candidates that lesser players agonize over. “Weak players spend more time considering weak moves, while strong players spend more time considering strong moves,” said Jonathan Levitt in Genius in Chess.

Topalov – Aronian

Wijk aan Zee 2006

White to play

Black would love to trade bishops and rooks after …xg2. A master sitting in White’s chair wouldn’t take long to consider any line that allows that. Instead he’d be attracted to 1 xe4.

Of course, to play that with confidence requires quite a bit of variation-crunching. But by spending “more time considering strong moves” White was able to see that the pawns cannot be stopped after 1 xe4!! xe4 2 d5.

The game ended with 2…ce8 3 d6 e1+ 4 h2 f5 5 g3 g6 6 g5! xg5 7 xg5 d1 8 c6 e2+ 9 g3 Resigns.

In the dark ages before chess clocks, masters routinely took hours over a single move. They justified this by exaggerating the number of candidates they needed to consider. “A fine player examines occasionally from five to twenty or more moves on each side,” William Lewis, one of the strongest masters of his era, declared, according to Howard Staunton’s Handbook. “Can this be done in a moment?”

The answer is “No – and it’s not necessary.” Masters calculate more efficiently today.

16

You’ll never become Alekhine if variations rule you, rather than the other way around.

This was Mikhail Botvinnik’s warning to a youngster named Garry Kasparov who very much wanted to be an Alekhine. “This pained me,” Kasparov recalled in the Russian magazine 64. “But the wise Botvinnik was right.”

What he meant is that if you seek complications because you are good at calculating them, you lose control of the move selection process. You are seeking complexity for the sake of complexity, not because it is right in this position.

“Do not calculate complicated lines before you are absolutely sure it is necessary” was the sound advice of another great teacher, Mark Dvoretsky. Another former Botvinnik student, Vladimir Kramnik, could have taken that to heart.

Karpov – Kramnik

Monaco 1997

Black to Play

Black’s pieces are much more aggressively placed and he got his internal calculating machine running at high speed when he considered 1…e4!?!.

He easily saw five moves ahead, 2 fxe4! xf2 3 xf2 xh2+ 4 e1 xg3+ 5 d2 d4. That’s impressive. But the variations had taken control of him. After 6 c2! his very strong position had become very difficult.

This wasn’t at all necessary. With the simple 1…f6! and …df8 or …g6 Black has a brutal attack which would likely have won faster and without risk.

There are also natural limits to how much you can calculate:

17

You can’t see 64 squares at the same time.

When psychologist Alfred Binet studied blindfold chess he conducted a survey of leading players. He asked whether they could mentally see the entire board at the same time. “Of course,” Siegbert Tarrasch responded. “But it is difficult.”

In truth, it is virtually impossible even with eyes wide open.

A player typically examines sectors of the board, perhaps as few as 12 or 16 squares at a time, while considering a tactic. After he’s seen what he wants to see, his eyes quickly shift to another sector.

But unless you’re looking at an endgame with only a few pieces, you can’t take in everything at once. That’s why even world-class players miss tactics that are staring them in the face in a distant sector.

Kasimdzhanov – Topalov

Leon 2007

White to play

Black’s queen had been shifting between a5, a2 and a3 when he found himself in trouble after 1 e5! and then 1…dxe5 2 d6 and 3 d7.

Annotator Maxim Notkin blamed Black’s problems on his inability to look at both the queenside and kingside at the same time. That’s why he missed 1…xh2! which would have given him good chances (2 xh2 c4! 3 xc4? g4+ 4 g1 g3).

“If our eyes had the capability of taking in various sectors, probably we’d play stronger chess,” Notkin wrote on e3e5.com. “However, an argument with the Creator is beyond the scope of this article.”

18

The hardest moves to see are very long moves backwards, and knight moves.

This is one of many perceptive observations from Cecil Purdy. We often overlook long retreats because we typically look for our opponents to make forcing moves, usually advances. We have a problem with knights because they don’t move in straight lines. Even masters overlook strong knight moves in time pressure.

If it’s a knight that is retreating, it can be doubly easy to miss. In the diagram below White appears to enjoy some superiority because of his better pawns and minor piece. It seems that Black’s knight would only be dangerous if it lands on d4 or f4 in some tactically dangerous way. However, the knight won the game in one move – by going backward:

Morozevich – Movsesian

Sarajevo 2007

White to play

White jumped at the chance to activate his queen, 1 e3, with a threat of 2 xa7.

However, 1 e3?? lost to the unlikely 1…g7!. There is nothing to be done about 2…f5 or 2…h5, followed by a killing check on g3.

The game ended with 2 g1 f5 3 e1 xg1+ 4 xg1 g8+ 5 h1 g3+ 6 g1 f1+ 7 h1 xh2! Resigns.

No better is 4 xg1 g8 5 g2 xg2 6 xg2 xg2+ 7 xg2 e3+ and …xc4.

19

Knowing when to calculate is just as important as knowing how to calculate.

Masters streamline their thought processes by choosing some moves based on general principles, intuition, positional understanding or some other criterion. They don’t spend time on calculating those moves, except for a “blunder check” – making sure they didn’t allow a simple enemy tactic.

How do you know when you need to calculate more? A good tipoff is when the first candidate you are attracted to doesn’t meet your expectations.

Gerard – Glek

Clermont-Ferrand 2003

White to play

White naturally looks at the forcing 1 xe6+. But after seeing how good 1…h8 and 2…f3! is for Black he turns to 1 xe6, which threatens a deadly discovered check.

This meant he has to calculate beyond two moves into the future. He sees that after 1…h8 2 e7! Black’s pieces would be severely restricted, e.g. 2…d4+ 3 h1 f6? 4 e4 and xb7.

He played 1 xe6!. Black also saw 1…h8 2 e7!. Like White he turned to a second candidate, even though it meant he, too, would have to calculate several moves ahead.

He checked out 1…h4 and found that the most dangerous-looking tactic, 2 e7+ and 3 xd7, would walk into 3…e1 mate.

There were more difficult lines to calculate but they all turned out well for him (2 xf4 xf4! 3 f6+ c4! and 2 e7+ h8 3 xb7 e5!).

So Black played 1…h4! and White found nothing better than 2 g3 h3! 3 xf4 h8, with advantage to Black.

“Knowing when…” has been attributed to several masters and appears in several forms. Garry Kasparov wrote in How Life Imitates Chess, “You have to realize when you are leaving the realm of what can be confirmed beyond a reasonable doubt.”

20

Long variation, wrong variation.

Bent Larsen said his aphorism applies to both long variations of published analysis, which he instinctively distrusted, and to over-the-board calculation. The further you try to see with certainty, the more likely you’re missing something.

David Bronstein agreed. “At first you see the position clearly,” he said in one of his last interviews. “Within three moves it becomes somewhat like in a fog. And within five moves you only see the contours of the position.”

The other problem with calculating a long variation is that you spend so much time on it that you tend to forget about the other alternatives at move two. “The things you overlook in your long calculation will seldom be as important as the move you overlooked at the start,” as Jacob Aagaard said in Excelling at Chess Calculation.

Fischer – Petrosian

Bled 1961

Black to Play

Black saw that 1…xe4 would likely draw but recognized that White could press him for some time because of his better rooks and passed c-pawn after 2 xe4.

He also noticed 1…d6. It seems to just lose a pawn after 2 xd6 xd6 3 xe6+ and 4 xa8. However, he correctly carried his analysis further and realized 3…fxe6! 4 xa8 c5 5 b3 d7 followed by …d4 and …c5 creates an absolute blockade.

Black rechecked this seven-move variation, found no flaw and played 1…d6?. He was “shattered,” according to White, by the reply 2 xa8!.

Normally, Black would have recognized that this should favor White on general principles, because of his passed pawn and better minor piece. But Black hardly looked at 2…xd1+ 3 c2 f1, apparently because he was wrapped up in his seven-move fantasy.

He lost soon after 4 xa5 xf2+ 5 b3 h2 6 c5 and that emphasizes the role of evaluation:

21

Calculation without evaluation is possible only in the endgame or studies.

Veteran trainer Adrian Mikhalchishin made this point in 64 (July 2005) when he advised his colleagues about how to teach young players:

A youngster should learn to evaluate the position after each half move. That is, when considering a candidate, he should determine who is better after each reasonable response by his opponent. Then he should look for his best reply to each enemy response and again evaluate the position. And so on.

Skipping the evaluation step is as bad as not looking for a tactical shot in response to it:

Yusupov – Speelman

Linares 1991

Black to Play

Black seems in trouble in view of h6 and ah1. That’s why he calculated 1…xf3 in considerable detail, e.g. 2 xf3d4+ or 2…f5+ followed by …xe4+ or …xg4(+).

But it didn’t matter which player had calculated 1…xf3 2 xf3 better. White won because only he correctly evaluated 2 h6+! and saw it leads to a winning endgame, 2…xh6 3 xh6 and 4 ah1.

Had Black seen how bad that was he would have pushed 1…xf3? out of his head. Then he might have found the best try, 1…h5! with the idea of 2 gf2 d5!.

22

As a rule, an experienced master never calculates all the variations of a combination.

After saying this in his best-game collection, Yuri Averbakh added, “Relying on his intuition, he works out only the necessary minimum.”

This may sound like a recipe for blunder. After all, not calculating every variation should increase the possibility of an oversight. But a master knows that you are more likely to err when relying on calculation than when relying on intuition.

Timman – Short

Candidates match 1993

Black to Play

White played the opening quickly and Black assumed he couldn’t play 1…xa1 because of a prepared refutation.

He chose 1…f3.

In fact, White didn’t have an answer prepared for 1…xa1. “I had intuitively relied on it that White, with three pawns for a rook and his good piece coordination, should be able to launch an assault on the unprotected enemy king,” he wrote in New In Chess.

In other words, White trusted his intuition about the position – and Black trusted White. Neither tried to determine the absolute “truth” by calculation. (Later analysis indicated that 1…xa1 2 e6 f8 3 e1 f5 4 c3 is far from certain.)

23

I don’t think like a tree?

Do you think like a tree?

What tree are we talking about? The one Alexander Kotov made his trademark in Think Like a Grandmaster.

Kotov endorsed a rigorous, if not rigid, procedure which called for analyzing each candidate move as if it were a tree branch, with subvariations sprouting from it. Analyze each branch and sub-branch once and only once, Kotov said, and when you’re done, play the candidate that results in the best position when met by the best defense.

But few people actually think this way. The most famous put-down of Kotov is #23, from another Soviet-trained GM, Anatoly Lein, (as quoted in Improve Your Chess Now). In practice, masters employ a much more chaotic method, as Veselin Topalov explained when annotating the game that made him famous.

Topalov – Kasparov

Moscow 1994

White to play

White felt he had to stop Black from castling or he would get the worst of it. He looked at 1 xd6 and concluded that only 1…xg4, threatening 2…xb2!, would be dangerous.

Then he found 2 a4 and concluded 2…e3 3 hxg4 was best and would favor him. So far, so good. He played 1 xd6.

But as Black was thinking, he suddenly remembered that during his tree-climbing he had found a flaw on the 2 a4 e3 limb – 3 hxg4? allows 3…xd2+ 4 xd2 d8!, which favors Black. But he forgot about that line when he took the d-pawn.

Fortunately by the time Black played 1…xg4, he had found an apparent refutation, 2 xg4 xb2 3 e5.

This is based on 3…xa1+ 4 f2 b2 5 b1, trapping the queen. Black jumped at the opportunity to play the 3…xe5 4 b1 xc3 endgame and eventually lost.

But it needn’t have lasted that long. Only after the game did White realize he had botched his analysis not once, but twice. After 2 xg4! xb2:

Instead of 3 e5?, he could simply play 3 ge2!. Black, a piece down, can resign because 3…xa1+ loses to the same idea White had seen in the 3 e5 branch, 4 f2 b2 5 b1,

White to play

– again winning the queen.

The truth is each player finds the method of move selection that is most comfortable to him. Psychologist Alexey Bartashnikov of the Lvov State Institute of Physical Culture reported in 64 (12/1997), “Research shows that masters calculate variations on an average of five-to-six moves deep.” But great calculators often reject a candidate move without looking one move deep. During his 2003 match with X3D Fritz, Garry Kasparov passed up the chance for a kingside pawn storm in one game. He said he relied on intuition rather than calculation. “Without a light-squared bishop, such attacks never work,” he explained.

And after Alexander Alekhine won a game with a rook sacrifice at Margate 1938, Miguel Najdorf expressed doubt that anyone could see all the head-spinning complications. Najdorf told the world champion he didn’t believe “you saw the entire combination.” Alekhine replied, “I also didn’t see it. I felt it.”

24

More often than not it’s difficult to calculate the consequences of the strongest move. It’s easier to convince yourself that all the others are worse.

Mark Dvoretsky was making two points with this comment in 64 (March 1998). First, in most cases you can’t easily prove a particular candidate works better than all the others. Many candidates result in fuzzy conclusions. But, second, you may be able to prove that all the alternatives are worse in some way or another.

The process of elimination works well when your opponent’s intentions are clear, such as when he is making threats. One by one, you can eliminate the candidate moves that fail to blunt those threats.

Svidler – Hansen

Groningen 1995

White to play

White didn’t like the messy 1 xa8 axb3, followed by …bxa2, …a5, or …bd7 and …xa8. He found 1 c5!.

Black shot back 1…b3! quickly. That surprised White. But it shouldn’t have if he had realized that Black was using the process of elimination. All the other candidates would have allowed White a powerful knight move, such as xa8, xe6 or cxa4.

For example, 1…dxc5? 2 xd8 xd8 3 xa8 and the knight is not trapped (3…bd7 4 g5 h5 5 b5!).

White replied quickly to 1…b3! even though the position was highly complex. He said later that it wasn’t difficult to determine 2 axb3! was the best move – because everything else would lose, e.g. 2 cxb3 xb6! 3 xe6 c6+! costs a piece, or 2 xe6 bxa2!.

After 2 axb3! play went 2…xb6 3 xe6 axb3! 4 c3!. White had the upper hand and won.

25

Chess isn’t checkers.

Benjamin Blumenfeld was a celebrated teacher of the early Soviet era, best known for the Blumenfeld Counter Gambit and for his advice on calculation. With #25 he was trying to correct a bad habit of young players, who assumed that if their opponent could capture a piece or pawn he would do so.

Nothing is automatic, Blumenfeld told them: “When you combine and give up a piece, look carefully then to see if your opponent is obligated to take it or not.” The opponent might be able to make a stronger move than the capture. Or he might play a zwischenzug, a forcing in-between-move that allows him to capture more favorably on the next move.

Kramnik – Ki. Georgiev

Moscow 1994

White to play

Black had just retreated the knight from a6 and threatens the queen. But White saw a mating pattern: If the knight is out of the picture and the a-file is open, White mates with a1.

With that in mind he looked at 1 a1. It is based on 1…xe6?? 2 b6+! axb6 3 axb6 mate.

But chess isn’t checkers, he added, and Black could reply 1…e8!. Then White has to force a draw with knight checks.

Instead, White found a more forcing order – 1 b6+! axb6 2 axb6 so that 2…xe6?? 3 a1 mate.

Black defended with 2…a6 but White continued 3 a1 d8 4 e7, threatening 5 xa6+ bxa6 6 a7 mate. This was decisive after 4…d6 5 xa6+! b8 6 xd6+.

26

Chess isn’t Tennis.

This old Marshall Chess Club kibitz points out the deceptive nature of the back and forth, serve and volley rhythm of a middlegame: White makes a threat. Black parries. White makes another threat. Black parries, and so on. In that way White always controls the cadence of play.

But the rhythm is broken if Black can reply with a move that both defends and attacks. Then he controls the rhythm. If his opponent can do the same – defending against the Black threat and making one of his own – he can regain control. But often a good calculator makes a gross error by losing control, as in the following example:

Gelfand – Anand

Biel 1993

Black to Play

Black, two pawns down and with a bad king, allowed this position to come about because he thought he could play 1…xd2.

He had seen 2 a4!, with its threat of h4+. But Black based his survival on 2…g5, which covers h4 and gains time by threatening his own mate on g2.

Some strong GM-spectators thought Black had turned the tables – until they saw the “only” move, 3 g3. This, too, defends against a threat and makes one of its own.

The volleying is over. Black has no adequate answer and he resigned soon after 3…e5 4 h4+ xh4 5 gxh4.

27

Wing attacks can be estimated before they’re calculated.

This is the essence of what Eduard Gufeld said in Chess Strategy