Your Kingdom for My Horse: When to Exchange in Chess E-Book

Andrew Soltis

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

Seitenzahl: 358

Veröffentlichungsjahr: 2015

Beschreibung

The great Mikhail Botvinnik, world champion said: 'The foundation of chess is exchanging/' Knowing when and what to exchange or trade in chess is essential to improve your game, and this is the only book to help you do that. In chess an exchange or trade of chessmen is a series of closely related moves, typically sequential, in which the two players capture each other's pieces. All chess pieces may be exchanged or captured in an exchange – apart from the king which however can capture an opponent's piece. Either the player of the white or the black pieces may make the first capture of the other player's piece in an exchange, followed by the other player capturing a piece of the first player, often referred to as a recapture. These maneouvres happen throughout chess, but understanding when and how to do this to your best advantage can improve your game significantly. Written by the hugely popular chess writer, Andrew Soltis, My Kingdom for a Horse tells you whether you should exchange your bishop for a knight, which pair of bishops you should exchange, when you should keep rooks on the board and when should you refuse to trade anything. This unique book will provide the answers on an important and integral part of chess strategy.

Leseprobe

YOUR KINGDOM FOR MY HORSE

WHEN TO EXCHANGE IN CHESS

Andrew Soltis

Contents

Introduction

Section One Simple Exchanges

Chapter 1 Queen Takes Queen

Chapter 2 Rook Takes Rook

Chapter 3 Knight Takes Knight

Section Two The Bishop Matrix

Chapter 4 Bishop Takes Bishop

Chapter 5 Bishop Versus Knight

Chapter 6 The Two Bishops

Chapter 7 Opposite Colored Bishops

Section Three Endgame Faceoffs

Chapter 8 Heavy Pieces

Chapter 9 Queen Versus Two Rooks

Chapter 10 The Fischer, Petrosian and Capablanca Endgames

Quiz Answers

Introduction

Almost every experienced player knows there are three basic ways to significantly improve a chess position. These are:

(a) Win material.

(b) Seize the initiative, such as by launching an attack on the enemy king.

(c) Secure a positional edge, such as by damaging enemy pawns.

But there is a fourth way: Make a favorable trade.

Even an exchange of identical pieces – a bishop for a bishop, a rook for a rook – can be very favorable. It can earn an advantage greater than what you might get from each of the other methods.

And this kind of edge often decides a game more quickly:

Nakamura – Kasimdzhanov, Baku 2014

1 e4 e5 2 ♘f3 ♘c6 3 ♗b5 ♘f6 4 d3 ♗c5 5 ♗xc6 dxc6 6 h3 ♘d7 7 ♗e3 ♗d6 8 ♘bd2 0-0 9 0-0 ♖e8 10 c3 c5 11 a3 a5 12 ♖e1 ♘f8 13 d4 cxd4 14 cxd4 exd4 15 ♘xd4 ♗e5 16 ♘2f3 ♗xd4 17 ♗xd4 b6 18 ♗e5! ♕xd1 19 ♖exd1 ♖a7 20 ♖ac1 c5 21 ♖d6 ♖b7 22 ♗g3 ♖xe4? 23 ♖d8 ♗d7

A quick look tells us White is a pawn down. A further examination shows that his pieces are more active than Black’s. But when we begin to calculate we realize he can’t immediately exploit his rook’s penetration to the eighth rank because 24 ♗d6 is blunted by 24 ... ♖e8.

But then we see 24 ♖e1!. Black’s essential defensive piece, the rook, must either be traded off (24 ... ♖xe1+ 25 ♘xe1) or go to a useless square like a4. In either case, White would carry out his winning threat of ♗d6.

In retrospect, 24 ♖e1! was a crushing move. Black resigned immediately. But because we are so used to thinking that strong moves can only be “loud” moves – captures, checks or advances – we fail to appreciate quiet piece exchanges.

World Champion Magnus Carlsen underlined this when he explained to surprised spectators how his 2014 challenger, Vishy Anand, found himself in a lost position with the White pieces without seeming to make a single error.

“It was one of those situations when you feel the position is equal,” Carlsen said. “But with every exchange your position might be a little bit worse.”

Swap-happy

Newcomers to chess often start with a very wrong idea about trading pieces. They see it as one sided.

“I like having a queen. I don’t want to give it up,” they say. So they blindly refuse to exchange queens even when it means reaching an easily won endgame.

But soon they acquire some experience – and go to the opposite extreme. Post-beginners have learned the “exchange value” of pieces. They understand that swapping a bishop for a knight is more or less neutral, from a material point of view. So they trade pieces indiscriminately.

What they don’t appreciate is that “exchange value” is chess shorthand. It allows you to make a quick evaluation of “I play bishop-takes-bishop and he replies king-takes-bishop.” But shorthand doesn’t take note of the superior qualities of supposedly equal pieces. One of the bishops may have much greater range or more potential targets than the other.

Simple arithmetic suggests another way pieces are different. If you spend one tempo to develop a knight and then trade it for a knight that your opponent invested two, three or four moves in, you are earning an advantage, an advantage in time.

That kind of edge may seem intangible. It often is. But in some cases it can be converted into tactics:

Bezgodov – Panchenko, Voronezh 2002

1 d4 f5 2 ♘h3 ♘f6 3 ♗g5 e6 4 e3 ♗e7 5 ♘f4 0-0 6 ♘d2 ♘e4? 7 ♗xe7 ♕xe7 8 ♗d3 ♘xd2? 9 ♕xd2 c6 10 g4! fxg4 11 h3 g3 12 0-0-0 g5 13 ♘h5 g2 14 ♖hg1 ♕f7 15 ♕e2 ♕xf2 16 ♖xg2 ♕xe2 17 ♖xg5+ ♔f7 18 ♗xe2 ♔e7 19 ♖f1! ♖xf1+ 20 ♗xf1

To avoid time-losing retreats, White traded bishops at move seven, Black swapped knights a move later, and Black later exchanged queens and rooks. But there is a difference between these trades.

Black spent two tempi on his knight (… ♘f6-e4) and White only one (♘d2), before a swap. Black lost a tempo. That’s not so bad. But then Black spent three tempi on his queen before trading it. Then he swapped his last developed piece, a rook.

The bottom line is that White still has three active pieces and Black has none. The rest was 20 ... d5 21 ♖g7+ ♔d8 22 ♖xh7 ♘d7 23 ♘f4 ♘f6 24 ♖f7 Resigns. Black had no pieces to stop the h-pawn’s advance, e.g. 24 ... ♘e8 25 h4!.

Four Golden Principles

In the simplest form, good trading comes down to the following four principles. These aren’t the only guidelines you can follow. Chess isn’t that easy. But these four are the most important.

(a) Exchange your pieces for better ones.

(b) Trade when you’re ahead in material.

(c) It’s not what is removed from the board that matters, but what remains.

And

(d) Exchanges increase in importance as the game goes on.

These sound simple, if not obvious. But being faithful to them is extraordinarily hard. If making good exchanges were easy, almost all serious players would be masters.

Take the first one. In a typical middlegame there are bound to be pieces that attack, defend or otherwise control more squares than other pieces. You will find yourself with good pieces, better pieces, mediocre pieces and, most likely, a poor piece or two.

This is not necessarily your fault. Some pieces are born “bad,” because of the opening. The problem child of the French Defense and Queen’s Gambit Declined, for example, is Black’s light-squared bishop. Once you play ... e6, a perfectly good move, you find yourself with a wayward child on c8.

The simplest way to improve a position is to swap your worst piece for a better enemy piece. For example, 1 e4 e6 2 d4 d5 3 e5 and now 3 ... b6 4 ♘f3 ♘e7 5 ♗d3 ♗a6! 6 0-0 ♗xd3 7 ♕xd3.

Black’s worst piece was his light-squared bishop, because of the pawn structure. White’s best piece in the French Defense is often his light-squared bishop, for the same reason.

So we can feel pretty certain that Black has made a good trade. You might think he must have better winning chances than White. Alas, chess isn’t that kind to Black.

He is only equal. But he avoided the horrible middlegame that Black sometimes gets in a French when his bishop remains stuck around c8. And he can get the upper hand after, say, 7 ... ♘f5 8 c3 ♗e7 9 ♗e3?! 0-0 10 ♖e1 c5! 11 ♘bd2 ♘c6.

The second principle – Trade when ahead in material – is familiar to even post-beginners. Again there is a foundation in basic arithmetic: When material is unequal, the ratio between the superior side and the inferior one becomes greater as pieces depart the board.

Let’s put that in practical terms. Being ahead a pawn in a complex middlegame may be a minor advantage. There are a lot of other pieces on the board to compensate for the pawn. But being a pawn ahead in a king-and-pawns endgame is typically decisive.

Material will be even at the start of most of the examples in this book. Nevertheless, exchanges matter, particularly towards the end of a game, as the fourth principle says.

If your bishop is better than your opponent’s, that superiority is likely to be magnified as queens, rooks, etc. are swapped off. As world champion Max Euwe observed, “A minor piece, representing about 10 per cent of the fighting forces on a full chessboard, obtains a higher percentage value after every exchange.”

“What Remains”

Siegbert Tarrasch wrote one of the most influential books of all time, The Game of Chess. It was subtitled, “A Systematic Textbook for Beginners and More Experienced Players” – yet it said virtually nothing about exchanging pieces.

However, Tarrasch left us one of the wisest bits of advice: It’s not what is removed from the board that matters, but what remains.

White has an extra pawn, and what’s more it’s an outside passed pawn. White’s pieces are also far more active than Black’s.

All the ingredients for victory are present. But is there a fast way to nail it down?

Yes, 1 ♗b5! was a paralyzing shot. The main point is that Black’s only useful move, 1 ... ♘c6, was met by 2 ♗xc6! ♖xc6.

The trade looks wildly unprofitable: Bishops are usually better than knights. Moreover, White gave up a piece with lots of range for a knight that was doing nothing on b8.

Yet the position was transformed after one more move, the precautionary 3 ♔g2.

Black can’t move his queen or king (3 ... ♕b8 4 ♕f6 ♕f8 5 ♖a8! and ♕g7 mate, or 3 ... ♔h8 4 ♕f6+ ♔g8 5 ♖a8!).

Black’s rook has several moves (3 ... ♖c4, 3 ... ♖b6, 3 ... ♖c2, etc.). But White will answer them with 4 ♕f6 and threaten a killing combination, 5 ♖a8! ♕xa8 6 ♕g7 mate. (White played 3 ♔g2 so that Black cannot meet 4 ♕f6 with ... ♕xh6 with check.)

The trade at move two means that after 3 ... ♖c8 the Black rook will be confined to the first rank. How does White win then? Simply by advancing his a-pawn and eventually playing ♖b7 followed by a6-a7, ♕f6 and ♖b8.

So why weren’t the benefits of 1 ♗b5! and 2 ♗xc6! obvious? Because we are all mentally hot-wired to focus on the pieces as they go off the board.

This blinds us to the new position. It has changed because captured pieces don’t matter. “What remains” matters.

This principle will sometimes conflict with one of the other three principles. In the last example, it conflicted with “exchange your pieces for better ones.”

And when it conflicts, it takes precedence. The reason is that “better ones” is a somewhat abstract term. “What remains” is concrete.

Black threatens ... ♘xb3. White stopped him with 1 ♗xd4 cxd4. He not only eliminated Black’s powerfully centralized knight but kept material equal. This must be a favorable swap, right?

Wrong. What remained was a position in which Black had a protected passed pawn, 2 ♔e2 ♔d7 3 ♔d3 c5, a huge asset. White should have appreciated how 1 ♗xd4?? burned his bridges. The fourth golden principle tells us that piece exchanges tend to increase in significance as the game goes on, and 1 ♗xd4?? was the last possible piece exchange.

White’s only hope was that Black’s king could not get past his third rank. This quickly proved false, 4 ♔e4 ♔e6 5 f4 gxf4 6 ♔xf4 d3! 7 ♔e3 ♔xe5 8 ♔xd3 ♔f4 9 ♔e2 ♔g3 10 ♔e3 ♔xh3 11 ♔f4 ♔h4 12 ♔f5 ♔g3 White resigns.

Was White simply lost in the diagram? A closer look tells us that six of Black’s pawns were on the same color as White’s bishop. The chances were high that there was a way to exploit the pawns.

If White, a world-class player, had looked for it, he would have found 1 ♔e3! ♘xb3 2 e6! fxe6 3 ♗g7 ♔e7 4 ♗xh6 ♔f6 5 f4, when the bishop is very much alive and the outcome is in doubt. The same goes for 1 … ♘e6 2 ♔e4 ♘f4? 3 h4!.

Even computers get confused when “What remains” conflicts with one of the other principles.

In a cramped position Black chose the tempting 1 ... ♘b5 and White traded, 2 ♘xb5 ♖xc2 3 ♖xc2 ♕xb5.

Engines praise Black’s 1 ... ♘b5? even after 4 ♖c7! reveals how White’s rook controls the only open file and much of the seventh rank. Both man and machine were misled by the apparently favorable swap of a passive Black knight at c7.

Black had no counterplay: 4 ... a5 5 ♕c2 h6 6 h4 ♕a6 7 b3 ♕b5 8 h5 ♖d8 9 ♖c6 ♖f8 10 f4 ♕a6 11 ♖c7 ♕a8 12 ♕c6 ♕a6 13 ♕c2 ♕a8 14 g4 ♔h8 15 a4 ♔g8 16 ♔f2 ♔h8 17 ♕c6 ♕b8 18 ♔g3 ♔g8 19 g5!.

White threatens g5-g6, to expose targets at f7 and g7 for his queen and rook. Black resigned after 19 ... hxg5 20 fxg5 ♕d8 21 ♔g4 ♔h8 22 ♕b7 ♔g8 23 g6! ♕e8 24 ♖e7. Black simply wasn’t in the game after 1 ... ♘b5? 2 ♘xb5!.

Other Advice

Some great players offered general, even sweeping, advice about exchanging. Mikhail Botvinnik, for example, said, “The foundation of chess is exchanges.” That sounds impressive but it’s not particularly useful.

Something else he wrote was: “Pieces should be exchanged only when it brings concrete advantages.”

Following Botvinnik’s rule may require a lot of calculation. That was the case with 1 ... ♗f4!. It follows a positional guideline we’ll examine in Chapter Six (Trade a pair of bishops when your opponent has the two-bishop advantage).

But 1 ... ♗f4 also appears to be a gross violation of “Exchange your pieces for better ones.”

Black foresaw, in calculations three, four and five moves into the future, that the ugly e3-bishop is not an inferior piece. It is the backbone of White’s position. It defends the d-pawn and watches the f4 and g5 squares that would be essential to a kingside attack by Black.

For example, 2 ♘g2 ♗xe3! 3 ♘xe3 ♕f4! threatens to mate with 4 ... ♕xf2+. That would win after 4 ♖c2 ♘g5! or 4 ♗xe4 ♖xe4.

Better is 3 ♖xe3 but Black has the better chances after 3 ... ♕a7 or 3 ... ♘g5.

The story is similar after 2 ♘f3 ♗xe3! 3 ♖xe3 ♕f4. So White tried to defend the key squares with 2 ♕f3.

His idea is 2 ... ♗xe3 3 fxe3!, a recapture that would have allowed a strong ... ♕g3+ in the lines we just looked at.

But Botvinnik foresaw 2 ... ♗g5!. Then 3 ♗xg5 hxg5 (also winning is 3 ... ♘xg5) 4 ♘g2 ♘xd4. Also bad is 3 ♗xe4 dxe4 4 ♕g3 ♕a5 or 4 ... ♕xg3+ 5 fxg3 ♘b4.

So White soldiered on with 3 ♘g2 and Black won with 3 ... ♗xe3!. White didn’t like 4 ♖xe3 ♘xd4 or 4 fxe3? ♘g5! so the game ended with 4 ♕xe3 ♖e7 (threat of 5 ... ♘g5) 5 ♕f4 ♘xd4 6 ♕xc7? ♘f3+ 7 ♔f1 ♘ed2 mate.

But Botvinnik’s rule doesn’t work consistently. More useful is a bit of advice from another world champion, Vladimir Kramnik.

Kramnik – Piket, Dortmund 1995

1 ♘f3 d5 2 d4 e6 3 g3 ♘f6 4 ♗g2 ♗e7 5 0-0 0-0 6 c4 dxc4 7 ♕c2 a6 8 a4 ♘c6 9 ♕xc4 ♕d5 10 ♘bd2 ♖d8 11 e3 ♕h5 12 e4 ♗d7 13 b3 b5 14 ♕c2 ♗e8 15 axb5 axb5 16 ♖xa8 ♖xa8 17 ♗b2 ♖a2

White saw that 18 ♖a1 would force 18 ... ♖xa1+ 19 ♗xa1. White would have traded off Black’s most active piece. Some computers say it is the only move to keep a White advantage.

But White chose 18 ♕c1!? instead because of what he called “an ancient rule” – Don’t trade pieces when you have an edge in space.

Kramnik’s thinking is more important than the consequences. In the postmortem, several improvements were found over what transpired, 18 ... ♕h6 19 h3 g6 20 ♖e1 ♕f8 21 ♕b1 ♘b4 22 ♘e5 ♘d7!? 23 ♗c3 ♖a6 24 ♘xd7 ♗xd7 25 ♘f3 and White soon had a significant, eventually winning, edge (25 ... ♕a8 26 ♘e5 ♗c6 27 ♕d1 ♘a2? 28 ♗b2 b4 29 d5! ♗e8 30 ♘g4!).

There are many other great players and teachers who left us guidelines for exchanges. One of the most perceptive was Aron Nimzovich.

Nimzo’s Rules

Nimzovich may have been the first great chess thinker to denounce bad trading. He complained about “indiscriminate bartering” – that is, random exchanges. In My System he gave his own rules. “If an exchange does not come under one or other of these, it is bad,” he wrote.

Nimzovich’s rules were few:

Exchange to gain or avoid losing a tempo. Nimzovich claimed that a gain or loss of time “plays an essential part of every exchange.” For example, when you trade a piece to avoid retreating it you are saving the tempo you would have lost.

Nimzovich enumerated only a few other conditions when a trade is permissible. “We exchange in order to seize (or open) a file without loss of time,” he wrote. And, he added, we exchange in order to “destroy a defender.”

Like many great teachers, Nimzovich often violated his own rules:

Nimzovich – Rubinstein, San Remo 1930

1 ♘f3 ♘f6 2 c4 c5 3 d4 cxd4 4 ♘xd4 e6 5 ♘c3 d5 6 cxd5 ♘xd5 7 ♗d2 ♗c5 8 ♘b3 ♗e7 9 ♘xd5 ♕xd5 10 ♗c3 ♕xd1+ 11 ♖xd1 f6 12 ♘a5 b5 13 g3 a6 14 ♗g2 ♖a7

White has a small edge, in part because Black swapped his developed queen for an undeveloped one. But he tried to prove he had a bigger one with 15 ♘c6? ♘xc6 16 ♗xc6+ ♔f7. He had given back his lead in development by trading an active knight for the one on b8.

The upshot was that after 17 ♗a5 ♗b7 18 ♗xb7 ♖xb7 and 19 ♖c1 b4:

Still deluded with the belief in his advantage, White played 20 ♖c6 ♖b5! 21 ♖xa6? (21 ♗b6 ♖b8 favors Black) 21 ... ♖c8! 22 0-0 ♖c5 and lost his bishop.

And when it comes to advice, we should mention what David Bronstein had to say. “Do you know my theory of how Capablanca played?” Bronstein asked playfully. He answered his own question:

“He always tried to exchange one bishop, so that he should have no problems about how to arrange his pawn chain.”

“Then,” Bronstein added, “he exchanged one rook, if possible, (and) he had no problems about which rook to place on an open file.”

“And it remained to exchange one knight so that the remaining knight knew which weak square to control in the center.”

This was an exaggeration, of course. But as we’ll see, there was some truth to it – and some good advice for us.

Exceptions

Each of the golden principles comes with the usual exception: Tactics trump every chess rule. Threats and checks will tell you when to violate even the four golden principles.

This was played in a tournament with co-ed teams, of two players each. Without talking to one another, the teammates took turns making moves: the female player made the odd-numbered move and her male partner made the even-numbered move.

When former world women’s champion Xie Jun chose 1 ... ♗xg3 both of the men at the table grimaced. Her opponent, Nigel Short, told her later that “he has to think for at least half an hour” before he can allow himself to make this kind of move “because it is against his basic rules of positional chess play – trading a strong bishop for a non-threatening knight.”

But 2 hxg3 ♕b5! revealed why the capture was strong. Black’s rook can invade at e2.

The game was over soon after 3 b3 ♖e2+ 4 ♔b1 ♖xf2! (threat of ... ♖f1) 5 c4 ♕a5 6 a3 ♘g2 7 ♗xg2 ♖xg2 8 ♖a2 ♖xg3.

Xie Jun was right: 1 ... ♗xg3! was an exception. And Short was right, too. It’s worth a half hour to prove to yourself that such an ugly-looking move works.

And now it’s time for some review. At the end of each chapter there will be quiz positions to see how much you’ve digested in the previous pages. The answers to the questions can be found in the back of this book.

Quiz

1. Cmilyte – Kosteniuk, Women’s World Championship 2006

1 d4 ♘f6 2 c4 e6 3 ♘c3 ♗b4 4 ♕c2 d5 5 a3 ♗xc3+ 6 ♕xc3 ♘e4 7 ♕c2 c5 8 dxc5 ♘c6 9 e3 ♕a5+ 10 ♗d2 ♕xc5 11 ♖c1 a5 12 ♘f3 ♘xd2 13 ♕xd2 dxc4 14 ♗xc4

White threatens a discovered attack on the queen with 15 ♗xe6. What do you think of 14 ... ♘e5 so that 15 ♘xe5 ♕xe5 or 15 ♗xe6? ♕xc1+ ?

Should Black trade minor pieces (1 ... ♘xe4) or rooks (1 ... ♖d4) or neither?

Black is offering a trade of knights. Which is White’s best option:

(a) keeping Black constricted with 1 ♘d2,

(b) trading twice on e5 and trying for e4-e5 or

(c) 1 ♘xe5 ♗xe5 2 ♗h6 followed by a kingside attack?

Black threatens … ♘xb4. White has two attractive paths, 1 ♗xd5 and 1 b5. Which is better?

SECTION 1: SIMPLE EXCHANGES

Chapter One: Queen Takes Queen

Pieces are often traded in inverse order of strength: The weakest go off the board first. A pair of minor pieces typically is traded in the opening or early middlegame, then another pair, or two of the rooks. Queens go late.

This isn’t by design. It just happens this way, partly due to the order in which pieces are developed and come into (capturable) conflict.

We will start with the strongest first because a trade of queens is usually the easiest of all exchanges to understand. (A swap of minor pieces, if it includes at least one bishop, is often the most difficult to evaluate.)

In addition, a trade of queens is the most consequential of exchanges. Trading the board’s most powerful pieces always changes the nature of the position to some significant degree. No other swap does that.

As beginners, we learn the most important consequence: ♕x♕ is a way to escape from a dangerous middlegame. If the resulting endgame is equal, that’s often a good transition. If it’s a favorable ending, all the better.

White’s king is under pressure. Black’s threats begin with 1 ... ♘f4+, which would assure at least a draw. Should White try to calculate 1 ♘xd6 ♘f4+ 2 ♔f2 (2 ♔f1? ♕f3+ wins) 2 ... ♘e2+ 3 ♔e1 ♘xg3 ?

He didn’t have to. As soon as Black made his previous move, White instantly replied 1 ♗xe6+! and 1 ... ♕xe6 2 ♕b3!.

This forced a trade of queens and – because Black’s knight at h3 was trapped – guaranteed that White would regain the sacrificed piece with a pawn profit. Black tried to defend a bad endgame after 2 ... ♘f4+ 3 gxf4 ♕xb3 4 axb3 ♗xf4 5 ♖xa6 but the outcome was obvious.

Merely the offer to play ♕x♕ can have a drastic impact on a middlegame. Or even on an opening. Several gambits that were once considered promising have been nudged toward retirement because of an offer to trade queens.

For example, the Evans Gambit (1 e4 e5 2 ♘f3 ♘c6 3 ♗c4 ♗c5 4 b4 ♗xb4 5 c3 ♗a5) with 6 0-0 was a terror until Emanuel Lasker popularized the defense of 6 ... d6 7 d4 ♗b6!?.

The point is that Black would be happy to see 8 dxe5 dxe5 9 ♕xd8+ because White’s initiative dies after 9 ... ♘xd8 10 ♘xe5 ♗e6. Then the superior pawn structure favors Black. Nowadays, Evans Gambiteers keep queens on the board with 9 ♕b3 or earlier with 8 ♕b3 and 6 d4 d6 7 ♕b3.

Or consider the trappy line that runs 1 e4 e5 2 ♘f3 ♘c6 3 ♗c4 ♘f6 4 d4 exd4 5 0-0 ♗c5 and then 6 c3.

There any many ways for Black to lose quickly and brilliantly. But he has nothing to fear after 6 ... ♘xe4! 7 cxd4 d5! and then 8 dxc5 dxc4.

After 9 ♕xd8+ ♔xd8 10 ♗e3 ♗e6 or 10 ♘g5 ♘xg5 11 ♗xg5+ f6 12 ♗f4 ♘b4 White is worse. His gambit is now virtually retired from tournament play.

The Price of Refusal

“Okay,” a novice may say, “I’ll agree that a queen exchange is bound to change a typical position of mine. But what if I offer to trade and my opponent refuses?”

The answer is that refusing often comes with a price: You typically offer to swap by advancing your queen so that it attacks the enemy queen. If your opponent refuses to trade, it usually means he shifts his queen to a less useful square and leaves your queen on a more powerful square.

For example, in the last diagram, 9 ♕e2 (or 9 ♕c2) would invite 9 ... ♕d3!. The belated trade, 10 ♕xd3 cxd3, is quite good for Black (11 ♖e1 f5 12 ♘fd2 ♘d4 with the idea of 13 ... ♘c2 or 13 ... ♘e2+).

And a full retreat, 10 ♕e1?!, would allow Black’s queen to keep her dominating position. In this case – and in many others – a retreat to avoid ♕x♕ is worse than an exchange.

Atkins – Capablanca, London 1922

1 e4 c6 2 d4 d5 3 e5 ♗f5 4 ♗d3 ♗xd3 5 ♕xd3 e6 6 ♘e2 ♕b6 7 0-0 ♕a6!

White gained an advantage in kingside space with 3 e5 that might promise him attacking chances. But his plans would have been dashed by 8 ♕xa6. Clearly 8 ♕g3?? ♕xe2 won’t do.

White tried 8 ♕d1 (more ambitious is 8 ♕f3) and then 8 ... c5 9 c3 ♘c6 10 ♘d2 cxd4 11 cxd4.

But before he could organize his pieces with 12 ♘f3, Black played 11 ... ♕d3!.

Now the price of refusal goes up. White had to acquiesce to a queen swap or the loss of a pawn (12 ... ♘xd4).

He chose 12 ♘b3 ♕xd1 13 ♖xd1, and eventually lost after 13 ... ♘ge7 14 ♗d2 a5 15 ♖ac1 b6 16 a4 ♔d7! 17 ♘c3 ♘a7 (to discourage ♘b5) 18 ♔f1 ♘ec6 19 ♔e2 ♖c8.

The price of refusal applies to other pieces as well, as we’ll see in later chapters. If your opponent controls an open file with his rook, you might challenge it by moving a protected rook to the same file. He can avoid ♖x♖ by abandoning the file. But that leaves you in control of it.

The same goes if he concedes control of the best diagonal on the board to avoid a trade of bishops or retreats a knight from a central outpost to avoid a knight swap.

Queens are unique in their ability to repeatedly renew an offer to trade:

Rosentalis – Kovalenko, Tromso 2014

1 e4 c6 2 d4 d5 3 ♘c3 dxe4 4 ♘xe4 ♘f6 5 ♘xf6+ gxf6 6 g3 ♕d5 7 ♘f3 ♗f5 8 c3 ♘d7 9 ♗g2 ♕c4 10 ♕b3!

Black found a way to stop White from castling, 9 ... ♕c4. But now 10 ... ♕xb3 11 axb3 would hand White a comfortable edge because of his superior center and control of the half-open a-file. He could pressure the queenside with b3-b4 and ♘d2-b3-a5.

Black declined with 10 ... ♕a6, which still bars 0-0. But White repeated the offer with 11 ♗f1 ♗d3 (11 ... ♕a5? 12 ♕xb7) 12 ♗xd3 ♕xd3 13 c4!.

Note that material is equal. What made these offered trades favorable to White was purely positional. The double b-pawns after 13 ... ♕xb3 14 axb3 are actually an asset.

Black turned the tables with 13 ... ♕e4+ 14 ♕e3 f5!. Then 15 ♕xe4? fxe4 would be excellent for him because of factors like the weakness of the d-pawn (16 ♘d2 ♗g7; 16 ♘e5 ♖d8 or 16 ... ♘xe5 15 dxe5 ♗g7).

But he failed to appreciate White’s reply, 15 0-0, and responded 15 ... ♗g7?.

Having made the switch from spurning a trade to offering one, Black was unprepared for 16 ♕b3!.

He suddenly saw that White was threatening ♖e1 followed by ♕xb7 and also had ♗g5 or ♗f4 and ♖ae1 coming up. e.g. 16 ... ♘b6 17 ♖e1 ♕g4 18 ♗f4.

Black made matters worse with 16 ... 0-0-0? 17 ♗f4 because his king position was highly vulnerable and his queen nearly trapped. (In fact, 17 ♘g5 would have won material immediately.) He didn’t last long after 17 ... c5 18 ♖ae1 ♕c6 19 d5 ♕f6 20 ♕a3!.

What Remains

So far, the ground rules for ♕x♕ seem simple: Seek a queen trade when you are ahead in material or at risk of an attack, or when the enemy queen is more active than yours.

But as we’ll find in each of the ensuing chapters, many good trades – of rooks, bishop and knights – are based on other factors. Chief among them is “what remains.”

The principle takes precedence when you have an initiative and your opponent’s passive queen holds his position together.

Kasparov – Petrosian, Bugojno 1982

1 d4 ♘f6 2 c4 e6 3 ♘f3 ♗b4+ 4 ♗d2 ♕e7 5 g3 ♗xd2+ 6 ♕xd2 0-0 7 ♗g2 d5 8 0-0 dxc4 9 ♘a3 c5 10 dxc5 ♕xc5 11 ♖ac1 ♘c6 12 ♘xc4 ♕e7? 13 ♘fe5 ♘xe5 14 ♘xe5 ♘d5 15 ♖fd1 ♘b6 16 ♕a5! g6 17 ♖d3 ♘d5 18 e4 ♘b6 19 ♗f1 ♖e8 20 ♖dd1 ♖f8 21 a3 ♔g7 22 b3 ♔g8 23 a4 ♖d8

White has an undoubted positional edge and should be looking for a knockout. A natural candidate is 24 ♕b5, since Black has no good answer to the threat of 25 a5. His position would be ugly after 24 ... f6 25 a5, for example.

But 24 ♕c5! prompted immediate resignation. He could have avoided major material loss with 24 ... ♕xc5 (or 24 ... ♕e8 25 ♘g4!) 25 ♖xd8+ ♕f8 26 ♖xf8+ ♔xf8. However, 26 ♖c7! would then threaten 27 ♖xf7+ as well as 27 a5, e.g. 26 ... f6 27 ♖f7+ ♔g8 28 ♖xf6 ♘d7 29 ♖xe6.

The key to this and so many other winning trade offers is realizing which of your opponent’s pieces are essential. And when you’re defending, it’s the key to knowing which of your own pieces is indispensible.

White should be able to convert his positional advantage into something concrete. But 1 ♘e7+, 1 ♗f5, 1 g4 and 1 ♘xe5 are disappointing.

White found 1 ♕e3! and Black had to figure out what White intended. Black’s queen plays no kingside role so White seemed to be looking in that direction.

Was he planning 2 ♕h6 followed by 3 ♘e7+ ♔h8 4 ♘xg6+ ? No, Black would have a simple defense in 2 ... ♗g7.

Was White’s idea 2 g4 so that 2 ... ♘g7 3 ♕h6 ? No, 2 ... ♘hf6 may hold.

White’s main threat turned out to be 2 ♕a7!. Black’s super-passive queen was performing a vital task, defending the queenside.

The best practical chance may have been, 1 ... ♖a8!?. It looks ridiculous. But it stops White’s threat and forces him to look for a way to win on the kingside, such as 2 g4 ♘hf6 3 ♕f2.

That may win, but after 1 ... ♘df6? 2 ♕a7! ♕xa7 (2 ... ♕c8?? 3 ♘e7+) 3 ♗xa7 White had an easier time.

The a-pawn cannot be defended and White’s two connected passed pawns are too strong. After 3 ... ♘xd5 4 ♗xa6 ♗c3, he could have finished off fastest with 5 b5. No better was 3 ... a5 4 b5 since the b-pawn marches on.

When You’re Down Material

“What remains” also applies to another kind of counter-intuitive trade. It arises when a player has given up material to seize the initiative. He wouldn’t offer to swap queens then, would he?

Kotov – Gligoric, Stockholm 1952

1 d4 ♘f6 2 c4 g6 3 ♘c3 ♗g7 4 g3 0-0 5 ♗g2 d6 6 ♘f3 c5 7 d5 ♘a6 8 0-0 ♘c7 9 e4 ♗g4 10 h3 ♗xf3 11 ♕xf3 a6 12 ♕e2 ♘d7 13 ♖d1 b5!? 14 cxb5 axb5 15 ♘xb5 ♘xb5 16 ♕xb5 ♕c7 17 ♕c4 ♖fb8 18 ♕c2 ♕a7 19 a3

Black made a Benko Gambit-like pawn sacrifice with 13 ... b5 (long before the Benko Gambit was known). The offer would fail if White could fortify his queenside with moves such as ♖b1, b2-b3, a3-a4, ♗f1 and ♗e3.

But before that happened Black got in 19 ... ♕a4!. If White declines to trade, say with 20 ♕e2, Black’s advanced queen gives him compensation. For example, 20 ... c4 followed by ... ♘c5-d3, ... c4-c3 or … ♖b3/… ♖ab8. Or 20 ♖d2 ♘e5 21 ♕xa4 ♖xa4 22 ♖c2 ♘d3.

So, White opted for 20 ♕xa4 ♖xa4. But Black’s position had improved because White no longer had a queen to guard the b-pawn and the invasion square b3. White would be worse after 21 ♖b1? ♖xa3 (22 bxa3 ♖xb1).

White defended with 21 ♖a2 and Black replied 21 ... c4 22 ♗f1 ♘c5.

White has anticipated ... ♘d3. But his e-pawn has become the newest target.

After 23 f3 Black went astray in complications, 23 ... ♘a6? 24 ♖d2 ♘b4 25 ♖a1 c3 26 ♖g2 ♘xd5 27 b3! ♖xb3 28 exd5, and lost.

However, after 23 ... ♖b3 Black would have had sufficient pressure to prevent White from untangling (24 ♗e2 ♔f8 25 ♔f2 ♔e8).

What convinced Black to play 1 ... ♕a4 was that he could see that the next three moves would be more or less forced. As a result, he could visualize the position after 21 ♖a2 c4 fairly clearly.

Judging what happens after a queen swap is more difficult when you can’t calculate more than a move or two ahead. Then you have to rely on evaluating other factors, such as relative development:

Lautier – M. Gurevich, Biel 1993

1 c4 e6 2 ♘c3 d5 3 d4 c6 4 e4 dxe4 5 ♘xe4 ♗b4+ 6 ♗d2 ♕xd4 7 ♗xb4 ♕xe4+ 8 ♗e2 ♘a6 9 ♗c3 ♘e7 10 ♗xg7 ♖g8 11 ♗f6 ♕f4 12 ♗c3 ♖xg2 13 ♘f3 f6

White explained why 14 ♕d2! ♕xd2+ 15 ♘xd2 made sense: A queen is a great attacking force but it is also “the most efficient defensive piece.” Without it, Black can be overwhelmed, e.g. 15 ... f5 16 ♗h5+ ♘g6 17 ♘f3 ♔e7 18 ♔f1 ♘f4 19 ♗e5.

Instead, he played 15 ... e5 16 ♘e4 ♔f7. But White’s initiative rolled on with 17 ♖d1! ♖g8 18 f4! ♘g6 (18 ... exf4 19 ♗h5+ ♘g6 20 ♘xf6) 19 ♖d6 ♔e7 20 ♖xf6 ♘xf4 21 ♗xe5 ♘xe2 22 ♗d6+ ♔e8 23 ♔xe2

and he quickly won (23 ... ♗h3 24 ♖h6 ♗g4+ 25 ♔e3 and ♘f6+).

When Not to Trade

Refusing an offer to swap queens is often just a matter of comparing your chances in an endgame with your chances if you continue the middlegame.

This is hard (for humans) to calculate. But we can draw some general conclusions: A trade of queens, such as 1 ♕f3+ ♕d5 2 ♕xd5+?? ♘xd5, is a death sentence for White. With no risk of being mated, Black would be able to advance his passed queenside pawns.

Instead, White should stay in the middlegame with 2 ♕f6! because Black’s king safety is very much an issue.

In fact, White tried 1 ♕h5 ♘d5 2 ♕xh6 and a fighting draw came about after 2 ... ♘f4! (3 ♕xg5?? ♘e2+ 4 ♔h1 ♕xh2+! 5 ♔xh2 ♖h8+ and mates).

White didn’t believe he was better in these complications. But he knew he would be lost if queens left the board.

When material is equal but one side has a positional edge, the defender may have to choose which heavy pieces to swap in order to reduce pressure. The first thing to remember is that ♕x♕ is likely to have greater consequences than ♖x♖. That often means greater chances of making your position worse.

White’s queen dominates from e4. Should Black try to trade it with 1 ... ♕a8?

Or should he try to improve his own queen’s range by trading rooks?

Or should he avoid all trades?

The best active policy is 1 ... ♖d8! because 2 ♖xd8+ ♕xd8 (or 2 ♖f1 ♖f8) removes a good defensive piece. Black’s queen would have chances to land with effect on d1 or (later) d3.

Black made a logical but faulty choice, 1 ... ♕a8? 2 ♕xa8 ♖xa8.

White’s minor pieces were superior after 3 ♘a5! ♗f8 4 ♘f4 and his king and rook were more active after 4 ... ♖e8 5 ♔f2.

White’s king will easily get to e4 and his rook will look for an opportunity to invade on one of the three files.

There is no quick route to victory but the outcome was finally clear after 5 ... g5 6 ♘h5 ♗e7 7 ♔e3 ♔f7 8 ♖f1+ ♔g6 9 g4 h6 10 ♖b1 ♗d8 11 ♔e4 ♗c7 12 ♖f1 ♗d8 13 ♖f3 ♗e7 14 ♖b3 ♗d8 15 ♘c6 ♗c7 16 ♖f3 ♘d7 17 a4

17 ... ♖f8 18 ♘e7+ ♔h7 19 ♖xf8 ♘xf8 20 ♘f6+ ♔g7 21 ♘e8+ ♔f7 22 ♘xc7 ♔xe7 23 ♘xa6.

Timing

As you approach the endgame, the opportunities to trade queens increase in frequency. And from our golden principles, we know that exchanges grow in significance as the game goes on.

What this should tell the player with advantage is: You don’t have to jump on the first train you see. There is bound to be another.

Black has just played ... ♕d7. He sees that White’s bishops will be less of a danger after 1 ♕xd7 ♘xd7 because 2 ♗d5 ♘c5 offers a drawish bishops of opposite color ending, 3 ♗xc5 ♗xc5 4 ♗xb7.

So what should White do in a position he knows should be superior? Not 1 ♗d6? ♗c7, since 2 ♗xc7 ♕xc7 only helps Black. So does 1 ♕d6 ♕xd6 2 ♗xd6 ♘xe4.

White found 1 ♕c3. The tactical point is that 1 ... ♘xe4? can be punished by 2 ♕xe5 or 2 ♕c2.

But the main idea was to keep queens on until he can liquidate into an ending that is clearer than 1 ♕xd7. Queens maximize White’s practical chances.

Those chances were shown by 1 ... ♕c7 2 ♕b4! with a threat of ♕f8+. Then 2 ... ♘d7 3 ♗d5 would begin to squeeze Black off the board (3 ... ♗c5? 4 ♕c4! wins a pawn).

There followed 2 ... ♕d8 3 ♗b3 (to stop ... ♘xe4) 3 ... ♕e8 4 ♕d6. Thanks to waiting with his queen, White got to occupy the powerful d6 square.

After 4 ... ♔g7 he made a useful “pass,” 5 ♔f3!. It was better to wait Black out than allow exchanges such as 5 ♗b2 ♘xe4 (6 ♕d5 ♘c5 7 ♗xe5+ f6 8 ♗d4 ♘xb3).

Black was running out of moves (5 ... ♕d8 6 ♕xe5; 5 ... ♕xb5 6 ♕f8+) and facing a threat of ♗b2 and ♗xe5.

He tried to get his knight into action with 5 ... ♘h7 6 ♗d5! ♘g5+ 7 ♔g2. But White was close to getting the won ending he’s been waiting for. For example, 7 ... ♘h7 8 ♗xb7 ♕xb5 9 ♗d5 (threat of ♕xe5+) 9 ... ♕e8 10 ♗c6 ♕e6 11 ♕xe6.

Black couldn’t take the pressure and lost after 7 ... ♕c8 8 ♕xe5+ ♔h6 9 ♕f4 (faster is 9 ♕f6) 9 ... ♕xh3+ 10 ♔g1 ♗d8 11 ♗xf7 ♕d7 12 ♗d5.

Double-edged

Finally, let’s examine what happens when trading queens comes with a positional cost, such as doubling your pawns. If you also receive some other benefit, such as the two bishops, the queen swap may make sense.

We see this in an opening variation that runs:

1 e4 e5 2 ♘f3 ♘c6 3 ♘c3 ♘f6 4 d4 exd4 5 ♘xd4 ♗b4 6 ♘xc6 bxc6 7 ♗d3 d5 8 exd5 cxd5 9 0-0 0-0 10 ♗g5 c6 11 ♕f3 h6!? 12 ♗xf6 ♕xf6 13 ♕xf6 gxf6

Black wouldn’t dare try this without a queen trade (12 ... gxf6?? 13 ♕g3+ ♔g7 14 ♕h4 threatens mate and ♕xb4).

But with queens off, it’s a different story. There is virtually no chance of a mating attack. Black can focus on exploiting the b- and e-files and his two bishops.

White’s best trading option is ♘e2-g3 followed by ♗f5 to eliminate Black’s bishop pair. For example, 14 ♘e2 ♖b8 15 ♘g3 ♖e8 16 ♗f5 ♗xf5 17 ♘xf5.

But the simple 16 ... ♗e6! defends, since 17 ♗xe6? fxe6 would correct Black’s pawn structure and deny White ♘f5. Black would be better.

What if White first plants his knight, rather than bishop, on f5? For example, e.g. 14 ♘e2 ♖b8 15 ♘g3 ♗a5 16 ♖ab1 ♖e8 17 ♘f5 threatens ♘xh6+.

Then 17 ... ♗xf5 18 ♗xf5 is a bishops-of-opposite-color endgame that seems to favor White because of the pawn structure. But with rooks on the board, Black can generate plenty of counterplay, e.g. 18 ... ♖e2! 19 g4 ♖b4 20 ♔g2 ♖f4!.

White cannot defend f2 (21 ♔g3 ♗c7). We are following Spielmann – Rubinstein, Berlin 1926, which was drawn after 21 ♖bd1 h5 22 h3 ♗b6 23 ♗d3 ♖exf2+ 24 ♖xf2 ♖xf2+.

There are many other changes that ♕x♕ can bring about in a chess position. It sharply increases the power of rooks. It boosts the value of passed pawns – and may double or triple their importance. These changes are better explored in books devoted to the endgame, so we’ll turn now to a series of quiz positions. Once again, you’ll find the answers at the back of this book.

Quiz

5. Hulak – Korchnoi, Sarajevo 1984

1 d4 ♘f6 2 c4 e6 3 g3 d5 4 ♘f3 dxc4 5 ♗g2 ♗d7 6 ♘e5 ♗c6 7 ♘xc6 ♘xc6 8 ♕a4 ♕d7 9 e3

What should Black do?

6. Eljanov – Kramnik, Moscow 2010

1 d4 ♘f6 2 c4 e6 3 ♘c3 ♗b4 4 ♕c2 d5 5 a3 ♗xc3+ 6 ♕xc3 ♘e4 7 ♕c2 c5 8 dxc5 ♘c6 9 e3 ♕a5+ 10 ♗d2 ♘xd2 11 ♕xd2 dxc4 12 ♗xc4 ♕xc5 13 ♖c1 ♕e7 14 ♘f3 0-0 15 b4 ♖d8 16. ♕b2 ♗d7 17 0-0

White controls more space with his superior pieces. What Black can do?

7. Radjabov – Dizdarevic, Sarajevo 2002

1 d4 d6 2 c4 e5 3 d5 f5 4 e4 ♘f6 5 exf5 ♗xf5 6 ♘c3 ♘a6 7 ♘ge2 ♘e4 8 ♘xe4 ♗xe4 9 ♘g3 ♗g6 10 ♗d3 ♗xd3 11 ♕xd3 ♘c5 12 ♕c2 ♗e7 13 ♗e3 ♘d7 14 ♕