Giddins - 101 Chess Endgame Tips

101 Chess
Endgame Tips
·Steve Giddins
First published in the UK by Gambit Publications Ltd 2007
Copyright © Steve Giddins 2007
The right of Steve Giddins to be identified as the author of this work has been asserted in accordance
with the Copyright, Designs and Paten!S Act 1988.
All rights reserved. This book is sold subject to the condition that it shall not, by way of trade or
otherwise, be lent, re-sold, hired out or otherwise circulated in any form of binding or cover other
than that in which it is published and without a similar condition including this condition being
imposed on the subsequent purchaser.
ISBN-13: 978-1-904600-66-4
ISBN-10: 1-904600-66-2
DISTRIBlJTION:
Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN, England.
Tel +44 (0)20 8986 4854 Fax +44 (0)20 8533 5821. E-mail: [email protected]
Gambit Publications Ltd, 99 Wallis Rd, London E9 5LN, England.

E-mail: [email protected]
Website (regularly updated): www.gambitbooks.com
Edited by Graham Burgess

Typeset by John Nunn
Cover image by Wolff Morrow
Printed in Great Britain by The Cromwell Press, Trowbridge, Wilts.
10 9 8 7 6 5 4 3 2 l
Gambit Publications Ltd

Managing Director: Murray Chandler GM
Chess Director: Dr John Nunn GM
Editorial Director: Graham Burgess FM
Gennan Editor: Petra Nunn WFM
Webmaster: Dr Helen Milligan WFM
Contents
Introduction 6
Symbols 7
Bibliography 7
Acknowledgements 7
Kings and Pawns
2-The Gentlemanly Art of Shoulder-Charging

Tip 1-Defying Euclid 8
Tip 9
Tip 3-A Royal Pas de Deux IO
Tip 4 -The Breakthrough 11
Tip 5 - More Pawn Breakthrough Ideas 12
Tip 6-Don't Forget about Stalemate Resources 13
Knights
Tip 7 - Simplifying into a King and Pawn Ending 14

Tip 8-Knights Hate Rook's Pawns 15
Tip 9-Deflections 16
Tip IO-Pawns on the Same Side 17
Tip 11 -Space is Important in Knight Endings 18
Tip 12-In the Footsteps of the Master 19
Tip 13-Exploiting a Positional Advantage 20
Tip 14-Activity is King 21
Bishops
Tip 15-Barrier Reefs (Part l) 22

Tip 16-Barrier Reefs (Part 2) 23
Tip 17-Bishop and Wrong Rook's Pawn 24
Tip 18-King at the Rear 25
Tip 19- Good Bishop versus Bad 26
Tip 20 - How to Manufacture a Passed Pawn 27
Tip 21-The Principle of Two Weaknesses 28
Tip 22-Not-so-Distant Passed Pawn 29
Tip 23-The Crippled Majority 30
Tip 24-Pawns Don't Move Backwards 31
Tip 25 - The Power of Two Passed Pawns 32
Tip 26 - The Power of Connected Passed Pawns 33
Tip 27 - Barrier Reefs (Part 3) 34
Tip 28-Pawn Placement 35
Knight vs Bishop
Tip 29-Knight vs Bad Bishop 36

4 JO] CHESS ENDGAME TIPS
Tip 3 J -Knights are Better than Bishops at Attacking Weak Pawns

Tip 30-The Agile Knight 37
Tip 32-Knights are Better than Bishops at Attacking Weak Pawns (Part 2)
38
39
Tip 33- Bishop vs Knight on an Open Board 40
Tip 34- Fischer's Classic 41
Tip 35-Knight on the Rim 42
Tip 36- The Triumph of the Two Bishops 43
Rooks
Tip 37 -Know the Basics 44

Tip 38-The hnportance of King-Shelter 45
Tip 39-Know the Basics, Part 2 46
Tip 40-Know the Basics, Part 3 47
Tip 41 - When Two Extra Pawns Don't Win 48
Tip 42-Pawns on the Same Side 50
Tip 43-Defence from the Side 52
Tip 44-The Strength of the Rook at the Side 53
Tip 45 - The Strength of the Rook at the Side (Part 2) 54
Tip 46 - Karpov's Masterclass 55
Tip 47 -Shoulder-Charging 56
Tip 48-Keep the Furthest Pawn 57
Tip 49-Passed Pawns Mean Counterplay 58
Tip 50 - The Exception that Proves the Rule 59
Tip 51-More Chances with a Knight's Pawn 60
Tip 52 -Never Forget about Stalemate 61
Tip 54 - New York J 924 Revisited

Tip 53 - Triangulation 62
63
Tip 55- Don't Get in a Huff 64
Tip 56- The Active Rook 65
Tip 57 - Another Active Rook 66
Tip 58-Yet Another Active Rook 67
Tip 59-W hen Similar is Not the Same 68
Tip 60-Seizing Space 69
Tip 61-The Weakness of Weak Pawns 70
Queens
Tip 62-Evading the Checks 71

Tip 63-King Safety is the Key 72
Tip 64-King Safety is the Key (Part 2) 73
Strategy and Technique
Tip 65 - Exchanges to Realize a Material Advantage 74

Tip 66- Rooks are Better than Knights - Usually! 75
Tip 67 -When Two Pieces are Better than One 76
Tip 68-The Value of Knowing Your Theory 77
Tip 69 - Drawing with Rook vs Rook and Bishop 78
Tip 70 -When Two Rooks Beat a Queen 79
Tip 71- Transforming an Advantage 80
CONTENTS 5
Tip 72 - Centralizing the King 81

'lip 73 - Transforming an Advantage 82
84
Tip 74 - The Perils of Passivity 83
Tip 75 - A Lesson in Not Hurrying
Tip 76 -Positional Alchemy 86
Tip 77 -The Importance of Calculation 87
Tip 78 -Concrete Play 88
'lip 80 - Bad Bishops are Often Good Defenders

Tip 79 Schematic Thinking 89
90
-
Tip 81 -The Principle of Two Weaknesses Revisited 91

Tip 82 -Pawn-Structure Subtleties 92
Tip 84 -The Master at Work

Tip 83 - More Breakthroughs 93
94
'lip 85 - Workmanlike Technique 95
Tip 86 -A Two-Bishops Masterclass 96
Tip 88 -·Overcoming the Blockade

Tip 87 -Appearances can be Deceptive 97
98
'lip 89 -Squeeze Play 99
'lip 90 - The Power of the King 100
'lip 91 -Calculate Your Way to Victory 101
Tip 93 -The Importance of Playing On

Tip 92 -Fighting for Space 102
103
Tip 94 -The Outside Pawn-Majority 104
Tip 95 - Too Brutal 105
'lip 96 - The Minority Attack 106
Tip 97 - Attacking the King 107
Tip 98 -The Power of Zugzwang 108
'lip 99 - Passed Pawns are the Key 109
Tip 100-Don't Forget Combinations! 110
Tip IOI - Positional Draw 111
I ntroduction
erage player. There are several reasons for this. In the first place, many players take the superficially
The endgame is probably the most neglected part of chess, especially from the point of view of the av
logical, but fallacious, view that it is better to study openings, since if one misplays the opening badly
enough, one will not even survive into an ending. This may be strictly true, but only of very bad open
once one has achieved this, there is no good argument for ignoring endgames any longer.
ing play. It does not need too much knowledge to enable one to play the opening reasonably well, and
Secondly, many players believe that endings are boring. I firmly believe that this is completely
untrue, and, on the contrary, the endgame is the best and most enjoyable part of chess. I hope in this
book to show why this is the case.
Thirdly, the traditional three-hour playing sessions in club and league chess have tended to militate
against reaching very many endgames, and when one did get one, the chance to play it out was usu
local chess which has changed for the better in recent years, and the replacement of adjournments and
ally lost, due to the intervention of that dread figure, the adjudicator. Thankfully, this is one aspect of
adjudication by quickplay finishes means that endgame technique is now more important than ever.
In writing this book, I hope to give the average player a good introduction to many important
ideas and techniques in endings. The positions given include many basic, theoretical endings, and
also a significant number of more complex positions, which illustrate more general points of tech
nique. I hope that this material will not only improve the reader's endgame play per se, but also
stimulate further study. The material I can cover in a book of this size is of necessity limited, but
there is a wealth of fine books on the endgame, and any player who wishes to study further has no
lack of opportunity to do so.
Important Endgame Principles
Many important endgame principles are illustrated in the ensuing examples, but it will be useful
here to summarise the main points of endgames:
• Material matters in endgames. This may sound trite, but it is an important point. Whereas in the
middlegame, sacrificing material to open lines and activate pieces is a standard device, it is much
less common in the endgame. While we shall see that tactics and combinations have their role in
the endgame, it is usually only in rook (and some queen) endings that piece activity is more im
portant than an extra pawn or two. So, within reason, it pays to be a miser in the endgame.
• In similar fashion, pawn-weaknesses tend to grow in importance in endings. In the middlegame,
it is frequently a good idea to accept an isolated or doubled pawn, in order to activate one's
pieces and/or open lines. In the endgame, the simplified positions and (normally) absence of
queens tend to make such dynamic play much more difficult to achieve, and consequently static
weaknesses tend to be more important.
• We shall see much in this book on the subject of the 'principle of two weaknesses'. One weak
ness is frequently not enough, and the key to winning many positions is to create a second weak
ness in the defender's position, so as to stretch the defence to breaking point.
• The other cardinal endgame principle which I shall emphasize time and agairi is 'do not hurry'.
The endgame usually has a somewhat slower and less dynamic tempo than most middlegames,
and this means that careful and slow manoeuvring is often the order of the day. Numerous good
positions are spoilt by the player rushing things, when a small piece of preliminary care would
have eliminated all of the opponent's counterplay.
• Finally, it is important to have the right attitude to the endgame. There is a rather dreadful song,
from the Hollywood musical Camelot, called How to Handle a Woman, the crux of the advice
being "love her, simply love her". I don't know about women, but this is certainly the right way
to approach the endgame. As I said above, the endgame is the best part of chess, containing a
INTRODUCTION 7
wealth of depth and beauty, and the more one studies it� the more apparent this becomes. Re
gardless of any specific knowledge it may convey, if this book helps the reader to appreciate and
develop a love for endgarnes, it will have done its job.
Steve Giddins
Rochester, November 2006
Symbols
+
++
check Echt European team championship
# IZ
double check Ct Candidates event
checkmate interzonaJ
!! brilliant move Z wnal
! good move ECC European Clubs Cup
!? interesting move OL olympiad
?!
tt
dubious move jr junior event
? bad move team event
?? blunder 1-0 the game ends in a win for White
Ch championship 1h-1h the game ends in a draw
Cht team championship 0- 1 the game ends in a win for Black
Wch world championship (n) nth match game
Wcht world team championship ( D) see next diagram
Ech European championship
Bibliography
Many of the positions used in this book were taken from my own endgame files, collected over the
past 20 or more years. Other specific references used include:
Fine:Basic Chess Endings
Beliavsky & Mikhalchishin: Winning Endgame Technique

Panchenko: Teoria i praktika shakhmatnykh okonchanii
Smyslov & Levenfish: Rook Endings

Averbakh et al.: Comprehensive Chess Endings
Shakhmatny Biulleten, 1955-91

Shakhmaty v SSSR, 1946-91
I also drew on many individual players' game collections, notably those of Botvinnik, Smyslov and
others.
Acknowledgements
Thanks are due to John Saunders, Editor of British Chess Magazine, for permission to reproduce
some material, in modified form, which originally appeared in that journal. Gerard Welling's
splendid chess library supplied at least one reference which would otherwise have eluded me, as
also did Yochanan Afek, while the Gambit production team did their usual first-rate job.
This book is dedicated to the memory of the late A.J. ('John') Smith, for his friendship, his incalcu
lable contribution to my chess development, and above all, for passing on his own deep love of the
endgame. Sorry I wasn't there at the end, John.
8 101 CHESS ENDGAME TIPS
_- - - - - -]
r :
�p-�--- Defying Euclid
We ate all taught at an early age that the short This same idea crops up in many practical
est distance between two points is a straight endings, such as the following:
line. However, there are some important differ
ences between chessboard geometry and its Eu
. - . .
� • • =
clidean equivalent.
w
• • . a �- - . - .
� • • • m .
�·- �-·�·
.
• • • •
w
·- � - . .
� . . •• -�- - .
• • • • D m B �,
� . . . . ·-- -
• • • • Em. Lasker - Tarrasch
�·='__L-_ St Petersburg 1914
R. Reti Tarrasch had exchanged into this position

Kagans Neueste Schachnachrichten, 1921 thinking that he was winning, since his king can
stop the h-pawn, while he stands to promote a
White draws this hopeless-looking position, pawn on the queenside. Indeed, after 40 h4
because the king march 'iPg7-f6-e5-f4-g3-h2 g4 41 f6 c4 42 bxc4 bxc4 43 e5 c3 44
does not take any more moves than the straight bxc3 a4 ( D) the c3-pawn blocks the white king,
route '.th7-h6-h5-h4-h3-h2. By using the for and there is no way to stop Black's a-pawn pro
mer route, however, the white king is able to moting.
threaten to queen his own pawn, and thereby
gain vital tempi.
1 'iPg7 h4 2 <t>f6 'i.irb6 ( D)
2... h3 3 e6 �b6 4 d6 draws.
. . - -�
� • • •
-
-�·
w
Wfi �
� � •
• • • •
. . . �,
� . . .
. . - .
However, after 40 h4 g4 Lasker played
41 g6! threatening 42 h5. After the forced
�- - . - 41 �4 the white king had gained a tempo to
.••
return via a different diagonal, one which is not

3 eS <h:c6 4 f4 obstructed by White's own pawns. He drew af
White draws because his indirect king march ter 42 fs g3 43 �e4 f2 44 �dS e3 45
has forced Black to spend two tempi on ...b6 xcS �d3 46 �xbS �c2 47 'it>xaS �xb3
and ...�xc6. l/2-1/2.
101 CHESS ENDGAME TIPS 9
[_ --- �- - -1
�p� The Gentlemanly Art of Shoulder-Charging
Another aspect of the chessboard geometry seen This same idea crops up in many endings,
in Tip 1 is the use of one's own king to obstruct including some spectacular endgame studies.
the enemy king's path. By marching on an indi Here is a typical example:
rect path to its destination, the king can, as it
were, 'shoulder-charge' the enemy king out of
the way.
N. Grigoriev
Shakhmatny Listok, 1931
Schlage - Ahues If White tries to bring his king to the queen

Berlin 1921 side directly, he is defeated by a shoulder
charge by the black king; e.g., I �g4? 'ii>c2 2
In order to win, White needs to take the �f4 (or 2 <ifi>f3 �d3!) 2 ...'ii>d3 3 'ii>e5 'ii>c4 4
black a-pawn, without allowing Black's king to 'ii>d6 'ii>b5 5 'ii>c7 'iii>a6 winning. White must be
get into the comer in time. In the game, he ready to meet ...'it>xb6 by 'ii>b4, without having
failed, and Black drew after 1 'it>e6 'it>c3 2 'ii>d6? his way blocked by the enemy king. There is
'ii>d4 3 'ii>c6 'ii>eS 4 �b7 �d6 5 'ii>xa7 'ii>c7, etc. only one narrow path to the draw, which in
However, by employing the shoulder-charg volves reaching b4 via e l.
ing technique, White could have won by 2 'it>dS! 1 �g3!! �c2 2 'ii>r2! 'ii>d.2 3 'ii>rl! �dl 4
( D). 'ii>f2 'ii>d.2 5 'ii>ft ( D)
. . . �
�·· • •
n
B
u • -
- • �
• • • •
. . . �
� . . .
. - - . �
�-- • • <it.
This prevents the black king from getting to 5 'itd3
•••
d4, and wins a vital tempo. 2 'it>b4 3 'it>c6 �c4 The only remaining try.
6 'ifi>el 'ifii c4 7 'itd2 �b5 8 <ifi>c3 �b6 9 'iii> b4
•••
4 <ifi>b7 'it>dS 5 'iha7 'iii'c6 6 �b8 and White

wins. and White has the opposition and draws.
10 101 CHESS ENDGAME 'TIPS
..
[TIP 3 A Royal Pas de Deux
Again, the only move to draw, keeping the

distant opposition. 63 <j;lc2? loses because the
black king can go back to the kingside and pen
etrate there. After 63...<j;le6 64 <j;lc3 (64 <j;ld2
c;i;>f6 65 'it>e2 <j;lg5 66 'ifi>f2 'iii>h4 wins for Black)
64 ...<M6 65 'iii>d4 ..tr>g5 66 'it>e5 'it>g4 Black wins.
63 <j;lcS (D)
•••
Timman - Yusupov
Amsterdam (Donner mem) 1 994
In many king and pawn endings, it is vital to

understand the distant opposition. Here, Jan
Timman shows how to use the technique to
draw a position a pawn down. White must main
tain the opposition, else the black king will pen 64 'iii>c l!
etrate and win the e3-pawn. Black can get in Again, keeping the distant opposition. 64
either on the kingside or queenside, and White 'iii>d2? loses to 64 ...<j;lb4, when once again, the
must cover both avenues of attack. white king is outflanked. 65 'it>c2 �c4 66 <j;ld2
58 <j;lg2! <j;lb3, etc.
The only move, taking the distant opposi 64 ..tbs 65 'it>bt ! ..taS (D)
•••
tion. 5 8 <j;lg3? loses to 58 ...<j;lg5 59 <j;lf2 <j;lh4 60

<j;lg2 <j;lg4 61 ..tr>f2 c;i;>h3, as Black gets in and
wins the e3-pawn.
58 <j;lg7 59 <j;lg 1
•••
59 <j;lg3 also draws.

59 w 60 c.t>n 'it>e6 61 �e2 <j;>dS 62 �d2
•••
�d6 (D)
Black tries his last trap.

66�cl!
Distant opposition is important, but so are el
ementary tactics! 66 'ito>a1 ? keeps the distant op
position, but loses to 66 .. .f4, promoting a pawn.
11z_112
After 66 ...'it>a4 67 �c2! White retains the
63 'it>dl! opposition.
The Breakthrough
Passed pawns can be manufactured in the most 42...f4! 43 'ifi>dS

surprising of circumstances. For example, ev Or 43 exf4 h4 44 gxh4 g3 45 fxg3 e3, win-
ery beginner soon becomes acquainted with the ning.
following device: 43 h4! 44 �e4
44 gxh4 g3 45 fxg3 fxe3 and again Black
•••
promotes a pawn.
44 f3 45 gxf3 h3 0-1
•••
In both of the preceding examples, the de

fender's pawns were on the second rank, but
breakthroughs also occur in the middle of the
board, as in the next case.
Despite the symmetrical pawn-structure,

White forces a winning passed pawn by 1 g6!
fxg6 (or l ...hxg6 2 f6 gxf6 3 h6) 2 h6 gxh6 3
f6, and wins.
Tiris is a very basic example, but such break
through tricks arise in many positions. Here are
two examples from grandmaster practice.
Kharlov - T. Ernst
Haninge 1992
The Swedish GM casually played an incau-

tious move:
47 gS??
•••
White's reply left him with a very red face:

48 g4!
White forces another passed pawn. White's
two widely-separated passed pawns are far
stronger than Black's three connected pawns,
which can be stopped by the white king. The
Pomar - Cuadras game ended:
Olot 1974 48 hxg4 49 hS rs so h6 r4+ s1 'ifi>f2 g3+ s2
•••
'ifi>g2 'ifi>e4 53 h7 1-0

Breakthrough tactics often arise in positions
with doubled pawns, as here. Black won as fol So, the golden rule is: never forget about
lows: breakthrough possibilities!
[TW 5 More Pawn Breakthrough Ideas
The pawn breakthrough is such an important Again, White appears to be losing, since 41
idea, and comes in so many forms, that it is g6? hxg6 42 hxg6 �f6 is hopeless. However, a
worth seeing some more examples. typical breakthrough tactic turns the tables:
41 h6! 1-0
1be threat of 42 g6 is much stronger than its
immediate execution. Black cannot stop a pawn
promoting after 42 g6; e.g., 41...f4 42 g6 Wf6
43 gxh7, etc. This structure is quite common
and should be remembered.
Averbakh - Bebchuk
Moscow Ch 1964
Were it not for an immediate breakthrough,

White would be lost due to Black's outside
passed pawn. As it is, though, he won after...
51 eS! fxeS
Forced, since after 51...�dS 52 e6, Whire V. Borisenko Belovba - Zvorykina
wins easily with his protected passed pawn. USSR w omen's Ch (Riga) 1962
52 g5 hxg5
Again, 52...Wd7 loses: 53 f6 'it>e8 54 fxg7 Here, a similar idea occurs. Black won by:
'1;n 55 gxh6 b5 56 �e4 b4 57 'it>d3 ! and the 39 a4!
.••
white king stops both black pawns. 39 ..b4? 40 a4 achieves nothing, but now
40...b4 is a winning threat.
.
53 f6! 1-0
The h-pawn will queen. 40 �e4 b4 41 �d4 bxa3 42 �c3 (D)
42 'iti>g5! 43 d4 �xg4 0-1

•••
Capablanca - Ed. Lasker White loses the d-pawn and then must move
London ( ojfhand game) 1913 her king, with fatal consequences.
101 CHESS ENDGAME nPs 13
�;P 6.] Don't Forget about Stalemate Resources
Stalemate ideas can often rescue an apparently 44 fi'c7+ 45 �??

•••
hopeless position, and should not be forgotten. 45 b6 wins.

.
45 �a8! ( D )
•••
Chigorin - Tarrasch
Ostend 1905 Oops ! As we shall see again in Tip 55, cap
turing is not compulsory i n chess. 46 \!i'xc7 is
In this famous example, Chigorin played 50 stalemate, and White has no way to untangle
gxf6? gxf6 51 �g4 @e4 52 �h3 'iti>f4 and then his pieces.
46 �a6 \!i'c8+ 47 �a5 'ilc7 1/2-1/2
ever, thanks to a stalemate idea, he could have
resigned, since he is losing his f5-pawn. How
drawn by 50 �g4 �e4 51 g6 h6 (5 l ...hxg6 52 In the next, more modem, example, Black
fxg6 f5+ 53 �g5 f4 54 h5 f3 55 h6 gxh6+ 56 proves equal to the task.
�xh6 is also drawing) 52 �h5! @xf5 stalemate.
Unfortunately, the great Russian champion

seems to have had something of a blind spot for
such stalemate tricks. Here he is, later in the
same tournament, donating another half-point
to the cause, this time by missing a stalemate
defence for the opponent:
Hellers - Eingom
Debrecen Echt 1992
The Ukrainian GM had previously been win

ning this game, but a time-trouble blunder had
left him facing defeat. However, he showed the
strength of character to find the only saving
line, which involved trapping his own king on
h4: 45 'it>g6 46 'it>xe5 'iii>g5 47 'it>e4 h5 48 'it>e5
�h4 49 @xf4 (stalemate) i1z_11z.
Chigorin - Schlechter .•.
Ostend 1 905
14 IOI CHESS ENDGAME TIPS
Simplifying into a King and Pawn Ending
Simplifying to a king and pawn ending can fre

quently be the easiest way to realize an advan
tage. But it is essential to calculate the ending
properly. w
Aronin - Smyslov
USSR Ch (M oscow ) 1951
White can win in many ways thanks to his

Alekhine - Yates powerful knight and Black's bad bishop, but
Hamburg 1910 chose:
42 :gS Wh7 43 :Xg7+? :Xg7 44 tlJxg7
White won by precisely calculating the pawn <J;xg7 45 g4 ( D)
ending:
40 .l:.e5! :xe5 41 fxe5 rJile7 42 <;i;>d3! ( D)
It is all about the opposition. Not 42 'iti>d4?
<;i;>e6, while 42 <;i;>b4 leads only to an unclear B
queen ending after 42 ...<;i;>e6 43 'it>xb5 'itxe5 44
'iti>xa4 <;i;>e4 45 b4 rJilxe3 46 b5 f4 47 b6 f3 48 b7
f2 49 b81Ifl11. It was certainly not for this that
White exchanged rooks at move 40!
Things look hopeless, but Smyslov found a

miraculous idea:
45 hxg3! 46 fxg3 g4! !
•••
Giving White a protected passed pawn!

47 h4 c5 48 'iti>e2 rJilh7 49 <;i;>d3 rJilh6
The point. If now 50 <;i;>c4?, then 50 ... f5! 51
exf5 e4 and White actually loses. He can there
50 c3 a5 51 cxb4 axb4 11z_11z

fore make no progress.
42 <;i;>d7 43 e4 f4 44 rJile2 �e6 45 <;i;>f2! 1-0
This dropped half-point cost Aronin an Inter
•••
45 ...<;i;>xe5 46 'it>f3 and White wins.

zonal qualification, and haunted him for the rest
But miscalculating the pawn ending can be of his career. He was never the same player
disastrous, as in the following tragic example: again.
JOI CHESS ENDGAME TIPS 15
I TIP 8 ] Knights Hate Rook's Pawns
The knight is a piece whose activity is always The knight is helpless against the a-pawn.
severely curtailed when it is near the edge of 44 c3 a3 0-1
the board. Because of the rook's pawn's prox
imity to the edge, knights have particular trou
ble fighting such pawns.
Kasparov - Adams
L inares 1999
A. Cheron (end of a study)
.:
Nouveau Traite Com p/et d 'Echecs, 1952 Here too, with his king so far away, the de
fender's lone knight proves unable to cope with
This position is a simple example. White the rook's pawn.
wins by 1 tllg7+! tllxgl 2 h6. The knight can 61 'it>d4 tlla6
not stop the h-pawn, and after 2 � 3 h7! nor After 61... tlla4 White wins in typical fash
ion: 62 tLlc3 tllb2 63 tlldl ! tllx dl 64 a4 tllb2 65
•••
can the black king.

a5 'it>g7 66 a6 tlla4 67 a7 tllb6 68 'it>c5 tlla8 69
Such tactical ideas frequently occur in prac 'it>c6, etc. Such knight sacrifices are a common
tice, as in this exceptionally nice example: feature of these endings, and this is a device
which is worth studying and remembering.
62 'it>d5 'it>g6 63 ttJd4 'it>f6 64 'it>d6! <J;f7 ( D)
64 ...tllb8 65 tLlc6 tlla6 66 tllb4 winning.
Bonner - Medina Garcia

Haifa OL 1976
Black forced a winning passed a-pawn by 65 tlle6 l-O

means of a spectacular tactical blow: After 65 ...tllb 8 (65 ...'iii>e8 66 tllc7+) 66 tllc5
41...tDcJ! 42 bxc3 a4 43 cxd4 cxd4 'it>e8 67 'it>c7 White wins.
·· ·-· 01 ·
TIP � Deflections
In knight endings, deflecting the enemy knight lbxd2 and 43...'0t>c4? 44 a4 �b3 45 lbb5 �xa4
by a sacrifice is a typical device. 46 lbcl4 only draw.
44 'Itel 'it>d4
Now 45 lbc6+ �e3 is no help to White.
45 c;i;id2 ltJe4+ 46 'it>cl ( D )
B 46c;i;idl �e3.
Barcza - Simagin
B udapest - M osc ow 1949
This was the position reached after White's

sealed move, 42 lik6xt:.a7. Despite White's 46 lbd6!
•••
extra pawn, Black's powerful passed d-pawn, Again the only move; for example, 46...'it>e3?
supported by king and knight, give him all the 47 lbb5 d2+ 48 c;i;ic2 'it>e2 49 lbd4+ draws. It is
winning chances. However, the path to victory essential to keep the white knight from the b5-
is not easy, since White only needs to sacrifice square.
his knight for the dangerous d-pawn and Black 47 'it>d2
would already be unable to win most posi 47 lbc6+ loses to 47...'0t>c3 48 lbe7 (48 tbe5
tions. Detailed adjournment analysis enabled d2+ 49c;i;id1 lDe4 50 lDf3 lDf2+ wins for Black)
Simagin to find a superb winning line, based 48 ...d2+ 49 �dl lDe4 50 lbd5+ 'Ot>c4 51 lbe3+
largely on knight deflections. 'lt>d3.
42 d3 43 <js>fl (D)
••• 47 lbc4+ 48 'ii;ic t d2+ 49 �c2 c;i;ie3 50 lbb5
.••
(D)
43.. M!
The only winning move, threatening 44...d2,
.
50 lba3+!! 0-1
•••
and forcing the reply. Lines such as 43...'it>d4? After 51 lbxa3 �e2 Black promotes his d
44 lbb5+ c;i;ie3 45 lba3! d2 46 lbc4+ 'it>d3 47 pawn.
Pawns on the Same Side
Many 4 vs 3 endings on the same side are drawn pawns Black can exchange, the better his draw
(see Tip 42 for a discussion of the equivalent ing chances.
rook ending), but with knights, the ending is 66 tbc6 67 tbg6 �e6 68 f4 �4 69 <M2
cilf.7 70 lbb4 tbc2 71 <oP e2 �e6 72 <iPd2 lbb4
•••
usually winning. Fine pointed this out many

years ago in Basic C hess Endings, but did not 73 'ii>c3 tbal+ 74 �d4 tbcl 75 lbfS W ( D)
have a practical example from GM play, and
was forced to illustrate his point with a hypo
thetical case. His conclusions have since been
confmned by GM practice, the following being
an example.
White has finished Stage 1, and now carries

out Stage 2 - creating the passed pawn. Note
how the excellently-placed knight on f5 ties
down the black king.
76es fxe5+ 77 fxe5 ltJe2+ 78 �d5 tbc3+ 79
Portisch - lvkov 'iii> d6 tbe4+ 80 �d7 tbc5+ 81 '>t>c6 tbe6 82 �d6
Sarajev o 1962 tbf4 83 ltJe7 ttJe6 84 tbg6 tbg5 85 �d7 ( D)
White's plan is simply to centralize his king

---- --- �
and then create a passed pawn on the e-file.
65 e4 ltJe7 66 tbf4! ( D) B � .W. � -
�-r<h···
• • • lD�
•
- •�- nu •
-�
. . -��
� . . .
. . . �
� . ..
85...tbe4
Black is in zugzwang and must lift the block
ade of the e-pawn, since 85 ...tbe6 loses to 86
lbh8+.
It is important to stop Black exchanging 86 e6+ �g8 87 e7 lD f6+ 88 �d8 <M7 89
pawns with ...g6. Generally speaking, the more tbe5+ 1-0
I T1P 11] Space is I mportant i n Knight Endings
Knight endings are often said to be the closest trick lLl xb5 (or lLl xd5), if his knight strays too
to pure pawn endings. One respect in which far away. However, he now finds an excellent
that is the case is the impoJ1ance of extra space, regrouping for his knight, bringing it to its ideal
and a more active king. square, after which the win is clear.
48 lLl g7! 49 lbd.1 lLlf5 ( D)
•••
Punnett - R. Bellin
British League (4NCL) 200516 From f5, the knight does everything - at
tacks e3 and d4, and remains in touch with the
Material is equaJ and the pawns symmetri squares e7 and d6, so as to cover the potential
cal, but White is lost, because he is too cramped passed white c-pawn.
and passive. Black won in systematic fashion, 50 lbc3 �h3! ( D)
as follows:
44 �c2
After 44 exf4 tbe6 45 lLle2 'it>f6, Black will
regain the pawn and leave White with a weak
ness at d4.
44 <M6 45 ..ti>d2 <Ms 46 lLle2 lLle6 47 M
•••
..ti> g4 48,�e2 ( D)
mm B �
B � - - -
-·-·- .
w • •
�•u•m m
•D D•m••
Black seizes the chance to penetrate even
further.
� � D m 51 f3?
Losing immediately, but 51 lbdl 'it>g2 is
• -
- ·�w
u • zugzwang; e.g., 52 lLlc3 lLlxe3! 53 fxe3 f3t.
� ----- - etc. This tactical threat of ...lLlxe3 is another
reason why the black knight stands so well on
White has guarded all the approaches for f5.
now, and Black must aJso watch for the tactical 51 �g2! 52 fxe4 f3+ 53 <Ji>d2 f2 0-1
•..
�p 12 I n the Footsteps of the Master
46-.WdS 47 h3 .i. d8 48 �e2 .i. c7 49 'iit d3

.i. g3
Black loses patience. He should keep the
w bishop on d8, stopping ll'lg5. 'The Capablanca
method has worked!" (Botvinnik).
50 llle4 Ji.el ( D)
50.. ..i. h4 51 lbc3+.
Botvinnik - Riumin
Moscow 1936
White has an extra pawn, and the c5-pawn is

weak. White should be technically winning, but
methodical, careful play is needed.
34 �2
Centralizing the king. 51 ll'l g5!
34-.�5 35 �f3 �eS 36 e3 .i. d6 37 ll'l b2 Forcing a second weakness on g6.
�d5 38 ttJc4 .i. c7 39 �e2 �e4 40 lDdl+ 'iti1'5 51... h6 52 llle4
41 'iUJWe5 ( D) Now White has ideas of <.te4-f5-g6.
52 .i. h4 53 ttJc3+ 'iii> c6
------ �
•.•
53...'iii> e6 54 ll'lb5.
54 <Jr>e4 .i. f6 55 ll'l bl 'iii> d6 56 tbaJ 'ifi>e6 ( D)
w �
,v, - ··�
� -·
- 56 ...'iii>c6 57 Wf5.
. . . �
� - m - m . �

.� . . �
�LS- • nr-o... n
w
dwd
LS�B�
� � -
� �-� �
--· -
Botvinnik writes: "Although White has the
upper hand, it is not so easy to win the position.
White therefore continues to wait, in the hope
of some sort of inaccuracy on Black's part - a
method of play in the endgame which I ob 57 ll'l b5 a5 58 ll'l c7+ �d7 59 ll'l d5 .i.b2 60
served in the play of Capablanca himself." ll'l b6+ 'iii>c6 1-0 -
42 ttJc4+ W d5 43 �e2 �e4 44 lDdl+ �5 After 61 lbc4 and 62 'itif5, the white king en
45��e5 46 g4 ters decisively on the weak light squares - an
Taking f5 from the black king. other triwnph for the two-weaknesses principle.
�P 13 I Exploiting a Positional Advantage

The threat was a6, followed by bxc6 and
lbb5xa7.
31 b6+
w This is better than 31 bxa6 ..tb8 32 lba2
lbd5, followed by .....ta7.
31 <Ji>b7 32 g3 hS?!
••.
In time-trouble, Black devalues his kingside

pawns. A waiting move, such as 32...lbd7, is a
better try, when White would still face signifi
cant hurdles in converting his advantage.
33 b4! liJd.7 34 f4! exf4
34 ...e4 35 ..td4 and White penetrates.
3S gxf4lbf8
Sveshnikov - A. Sokolov The only way to engineer some counter
Moscow 1991 play.
36 tbe2 lbg6 37 liJd.4 ( D )
White has a small advantage, in the form of
better queenside pawns and slightly more space.
. - -- .
• -
� .�· • -
�
It is not very much, but in the face of White's
fine play, a couple of inaccuracies are all it B
takes for Black to lose the position. iDi• ••�
21 O liJd.7 22 ..t>t2 r s 23 ..te3 ..te7 24 b4 eS
25 a4 ..t d6 26 ..t d3 lDr6 27 cS+ �e6 � un .,.,
• •
Sveshnikov recommends 27...'iic 7 28 ..tc4 �
• � u
� �" � �
(28 b5 lbd5! and the pawn ending is a draw)
28... a6 as a tougher defence.
• • • •
28 bS! �d7 29 �c4 'ii c7 ( D ) • • • •
29 ...e4? 30 fxe4 fxe4 31 'ifi>d4.
� -··
37 lbxh4
•.•
37...lbxf4? 38 lLlxf5 g6 39 lbe7 and Black is

w in zugzwang and must lose a pawn.
38 lbe6 ..t c8?
Parrying the threat of 39 lbd8+, but losing
prosaically. Black's best chance was in fact to
40 lbxc6 h4 offers good drawing chances thanks

allow White's idea: 38...lbg6 39 lbd8+?! 'ii>a8!
to the power of Black's h-pawn.

39 lbxg7 lbg6 40 lbxhs
The rest is easy.
40 ..t d7 41 'ii d3 'ii c8 42 <Ji>e3 tf'Je7 43 lbg7
.••
30 a5 a6 lbdS+ 44 'ittf3 lbe7 45 tf'Je6 lbdS 46 4J d4 1-0

14
[-TIP
----- -- ] Activity is King
Never forget about tactics! The knight is

corning to e8, winning a pawn.
42 lbcs
•••
42...gxf6 43 gxh6 is winning for White.

43 lbe8 hxg5 44 fxg5 lbe4 45 h6 gx h6 46
gx h6 lbg5 47 lbf6 lbf3 48 lbg4 lbg5 49 �d4
f5! (D)
Kamyshov - Zagoriansky
Leningrad ( Trade Uni ons) 1938
Although material is equal and the pawns

symmetrical, White's more active king and ex
tra space decide the game.
37 b4 lbb7 38 h5
This kingside pawn set-up, dubbed the quart Allowing his king to get across via d6.
grip by Hans Kmoch, often arises in the Caro Against other moves, 50 lbf6 and 51 h7 wins.
Kann Defence. As this game shows, it can con 50 exf6
fer a significant advantage on White, who has a 50 lbf6? lbf3+ and 51...lbxe5.
variety of line-opening and breakthrough ideas, 50 �d6 (D)
•••
such as g5 or f5.
38 lbcs 39 g5 lbb7 40 lbe4 c.t?b6 41 �c4
•••
The king is ideally placed here, defending

the weakness on b3 and constantly threatening
to attack Black's weakness on a5. By contrast.
the black king has to adopt a purely defensive
role.
41 �c6 (D)
•••
Black has managed to stop the kingside

pawns, but in a typical example of the two
weaknesses principle, White wins by turning his
attention back to the weakness on the queenside.
51 lbe5 lbh7 52 lbc4+ �d7 53 f7 �e7 54
lbxa5
Curtains.
54 �7 55 lbc6 �g6 56 a5
•••
42 lbf6! and White won.

�..,- 15 J Barrier Reefs (Part 1)
We have already encountered the concept of This preserves the possibility to go to either
'shoulder-charging', when one side's king gets d6 or a5, depending on which route the black
in the way of its opposite number. The creation king chooses.
of a barrier, to prevent the enemy king from ap 1 � 7 2 a4
...
proaching the crucial area of the board, is an The two variations then run as follows:
other common endgame device.
2 �e8 3 a5 �d8 4 i. d6 �c8 5 a6 ( DJ
.•.
0. Duras
Deutsche Schachzeitung, 1908
2 �e6 3 a5 �d5 4 a6 �c6 5 i. a5 ( DJ
•••
This is a simple but classic example. Clearly,

White's bishop needs to move out of the way of
the pawn, but where? In order to win, White
------ --�
must prevent the black king from reaching the B • ••• •
comer. Black has two possible routes to a8, ei � · ·· . �
', . . .
ther behind the d-pawn (f7-e8-d8-c8, etc.), or
in front of it (f7-e6-d5-c6). Analysis shows
that if he goes round the back, White can keep . . . �
the black king out of the comer by the con
struction i.d6/�a6, with the black king on c8.
� . . .
On the other hand, if the black king chooses . . . �
the frontal route, he can again be kept out of the
comer if White achieves the set-up i.a5/£:,a6
w� • 8_9<t>
vs <j;c6. White wins in both cases. If Black pushes the
These considerations then supply White's d-pawn, White simply brings his king across.
only correct first move: Once Black has to move his king, he allows the
1 J.b4!! a-pawn to queen.
Barrier Reefs (Part 2)
1be type o f barrier idea seen i n Tip 15 often This involves three steps: put the white king on
arises in more advanced form. f8 (shutting out the black kfug entirely), the
bishop on h5, and then bring the white king
back to f6.
74 <M7 ..tc4+ 75 <Ms ..td3 76 .ie8 ..te4 77
..th5 ..td3 78 w ..te4 79 <M6
Mission accomplished.
79 ..td3 80 .ig6 ..te2 81 ..tc2 ..th5
•••
The final step is to get the white bish�p to f7.

Since the immediate 82 ..tb3? �e8 draws, the
black king must first be driven away from e8.
82 ..ta4!
A nice echo of move 68, placing Black in
zugzwang.
82 �c7
.••
Penrose - Franklin Or 82.....te2 83 g6 ..tc4 84 g7 ..ta2 85 ..tc2

British Ch (Aberystwyth) 1961 ..tc4 86 ..tg6 and 87 ..tf7 winning.
83 .ib3 1-0
White needs to prevent Black from giving up 84 ..tf7 wins.
his bishop for the g-pawn, or getting his king in
front of it. The alternative defence 68 ..tcS fails to an
.••
68 .idl! other barrier: 69 ..tc2 �b6 70 ..tf5 ..ta6 71 g4

Preparing a barrier against the enemy king. 1;c7 721;e5 �d8 73 cM6 �e8 74 g5 rJ;f8 75
68 .i.d3
•• ..th7! ( D).
68.....tc8 will be examined below.
69 g4 �b6 70 �eS 'i;c7 71 <M6 �d8 72
. - - �
� • • •.t
.ia4! (D)
B
.t.• • = �
� • • D
. . . �
� . . .
. . . �
-- - -
The black bishop cannot get to the e8-h5 di
agonal in time: 75... .ic4 76 g6 .ib3 77 g7+
�e8 78 ..tg6+ �d7 79 ..tf7 and White wins.
The point: the black king is cut off. 75...�e8 is a tougher defence, as 76 g6?? al
72 i.e4 73 g5 ..td3
•.• lows 76....ic2, but White wins by, e.g., 77 rl;g7
Now White must drive the black bishop off ..te2 78 1;h6 .ic4 79 .ibl �e7 (79... .if7 80
the b l -h7 diagonal by playing his bishop to g6. ..tg6) 80 g6, etc.
14 l0 l CHESS ENDGAME TIPS
l!IP 17
r.-----J Bishop and Wrong Rook's Pawn
It is well-known that bishop + rook's pawn do

not win if the bishop does not control the
queening square and the black king can get in
the comer. This classic ending shows White
employing some fancy footwork to keep the
black king out.
Hebden - Mortazavi
British Ch ( Norwich) 1 994
53 ii.fl gs 54 .txh3 rs ss .tg2! fxg4

Else 56 h3 keeps the g-pawn.
56 ci;;d4 �6 57 .te4 'l;g7 58 .tc6 'l;g6 59
L. Paulsen - Metger J.e8+ wrs 60 �e3 �eS 1-0
Nuremberg I 888 White has a simple plan: win the e6-pawn
(by, for example, 61 .ta4 �5 62 .tc2+ 'ite5 63
1 'itd4! .te4 'WW6 64 'itd4 'itg7!? 65 .tc6! {65 �e5?
Not 1 'itc5? b6+, nor 1 'itc4? b5+!, as oc 'ith6 and ...�h5} 65...�f6 { 65...�h6 66 .te8}
curred in the game, after which the black king 66 'it>e4, etc.) and then get a position akin to
reaches b7. this:
1 'itc6
...
l...b6 2 a6.
2 .tb6 �d6 3 �c4 'itc6 4 �b4 'iti>d6 5 �b5
'itd7 6 �c5 �c8 7 .ta7 ci;;c7
7 . . b6+ 8 �xb6, or 7 ...b5 8 a6.
.
8 �b5 <i;d7 9 .tb8 'itc8 10 .tf4 �d7 11

'iti>b6 �c8 12 .tg3
and White wins.
Sometimes, even getting the king in the cor

ner does not draw (see following diagram).
Black had just swindled his opponent from a
clearly lost position and was now congratulat
ing himself on having apparently secured the This would be a draw without the second
draw. White cannot stop ... g5 and ...f5, leaving black pawn on g5, but here 1 ii.h7 g3 (forced) 2
him with the wrong rook's pawn, nor can he hxg3 g4 (no stalemate!) 3 .te4 wins, since
keep the black king from the comer. However, White has transformed his h-pawn into a g
play continued: pawn.
IOI CHESS ENDGAME TIPS 15
King at the Rear
1'2-1/2
87 ..i.c7!
••
White can make no progress.

However, this technique doesn't always draw.
Here is a classic example.
Taimanov - Fischer
Buenos Aires I960
It looks as though the remoteness of Black's

king will cost him the game because he cannot
get his king in front of the pawn. Indeed, against L. Centurini
Capablanca at New York 1916, Janowski re La Regence, 1856
signed a very similar position. However, Black
can hold by taldng his king to the rear of the White wins elegantly:
pawn, an important defensive technique in such 1 iLh4 �b5
positions. Else iLf2-a7.
81 cM4! 82 b5 �e4 83 iLd4 iLc7 84 �c5
••• 2 iLf2 �a6 3 iLc5!
'it>d3! ( D) A waiting move, but the only one. White
wants to force the black bishop from h2, but
must not allow it to d6.
3 ...i.g3
.
3 ... iLf4 and 3...iLe5 are the same.

4i..e7 �b5
He must stop iLd8-c7.
5iLd8 �c6
We have returned to the initial position, but
with Black's bishop on a different square.
6 iLh4!
Gaining a vital tempo, the point of White's
3rd move. Note that if Black's bishop were on
d6, 6 iLe7 iLh2 draws, as there is no 7 iLc5.
The crucial idea. The black king heads for 6 iLh2 7 iLf2
•••
c4. Now Black cannot stop the bishop reaching

85 �c6 �c4 86 iLb6 iLf4 87 iLa7 a7: 7...iLf4 8 iLa7 iLh2 9 iLb8 iLgl 10 iLg3
Or 87 iLd8 iLe3. iLa7 11 iLf2! and White wins.
Good Bishop versus Bad
If 51 <itc2, the black king gets in by 51 ... �a3

52 i.f2 'iitb4 53 ¢>d3 �b3, etc. The constant
use of zugzwang enables the black king to inch
B its way in.
51 �b3 52i.et �b2 53i.f2 <itct ( D)
•••
-- --- �
w � - - -
- - - �
� -·- -
m D -·�
� -�- D
m m m �
White has two pawns fixed on dark squares,
which is enough to cost him the game.
� - -----="'-
' ----'
--
4S <itd7!
••. 54i.e3+
A typical idea. Black cannot penetrate on the 54 �e2 'iir'c 2 is no better.
kingside, so he plans to widen the front by play 54 �dl SSi.f2
.••
ing his king to b5 and then advancing ...a5-a4. 55 i.f4 i.b4 56 i.d2! is a nice try, pointed
This will open a path into White's position via out by Panchenko, but it does not save the
the queenside. The drawback to the plan is that it game. Panchenko's analysis continues 56...i.e7
involves exchanging a pair of pawns, in a posi 57 i.f4 'iitel 58 �e3 i.b4 59 i.e5 i.d2+ 60
tion where they are already thin on the ground. �d3 i.g5 winning.
However, Kramnik has judged the position to a 55 i.aJ 56 ..t>e3 i.ct + 57 'it>d3 i.d2 58
•••
nicety, and realizes that he can still win, even i.e3i.el 59i.f4i.f2 60i.eS ct>et ( D)
with only two pawns each.
46 'Ot>e2 �c6 47 �d3 <itbS 48 <itc2 a5 49
�d3 a4 SO bxa4+ 'Ot>xa4 ( D)
The black king has come a long way over the

past 1 6 moves, but he has penetrated to the very
heart of White's position.
61 �c3 �e2 62 c;itb4 'it>f3 63 �c5 'iite4! 0-1
How to Manufacture a Passed Pawn
h-pawn will decide the game. This device is

well worth remembering. In the initial posi
tion, it seemed impossible for Black to create a
passed pawn, but he has found a way.
32 .id2 rs 33 .i.el .ie7 34 .td2
34 f3+ <ii;>h 3 does not help.
34.; .i. d6 3S .i. e l r 4! 36 f3+
•
36 gxf4 <ii;>xf4 37 f3 .i.e7 and the h-pawn ad

vances decisively.
36 'it>h3 37 gxf4 <ii;> g2! ( D)
•.•
Wijk aan 'Zee Ct (4) 1994

Khalifman - Salov
White stands worse, because the d4-pawn

blocks his bishop. Since the pawn-structure is
symmetrical and he has no other weaknesses,
he should still draw, but one error proves enough
to lose.
25 <ii;> h7 26 � �g6 27 'iii> e2 b4 28 g3 'M s
•••
2 9 .i. b2 g S ( D)
There is no defence to the plan of ....ie7 and
...h4. .
38 r s .i. e7 39 rfi .ixf6 40 .i.xb4 h4 41 .i.d6

b3 42b4(D)
30 .i. cl?
White's only clear error, but it costs him the
game. He presumably missed Black's 31st.
Salov gave 30 �d3 !, when 30 ... g4 is met by 31
h4. White would then have little to fear. 42 .i.e7!
••.
30 g4 31 hxg4+ A neat final touch, eliminating the b4-pawn.

Now 31 h4 'ifr>e4 is hopeless for White.
•••
Instead, after the hasty 42....i.h4? (do not hurry!)

31. <ii;> xg4!
•• 43 b5 .ig3? 44 .ixg3 'it>xg3 45 b6 h2 46 b7 h1 'ill
The key to Black's plan. Now a subsequent 47 b8'ii'+ White even wins .
...f5-f4 will force gxf4, when Black's passed 0-1
28 JOI CHESS ENDGAME nPs
[rrP 2 1J The Principle of Two Weaknesses
One weakness is often not enough to lose, and typical example of the kingside sacrifice, set up
opening up a second front is the key to winning by White's 40th move) 47 �d4, when Black is
·
many positions. in zugzwang. Notice how in these variations, it

is the need to defend not only the c5/d5/e5
squares, but also the second weakness on the
kingside, which overwhelms the black defences.
45 .ia4 .ifS 46 .ieS ( D)
Miles - Mariotti
Las Palmas 1978
Black has one weakness on d5, but by itself,

that is not enough. 47 d4+
Desperation. Passive defence with 46.. ..i. e6
•••
40 b6!
Creating a new weakness at h7. White now leads to another zugzwang after 47 a4 � d6 48
has ideas of getting his bishop round the back to �d4 �e7 49 .ic6 �d6 50 .ib7, and White
g8, or sacrificing on g6. Miles assesses the po wins.
sition as drawn if Black can play 40. .. gxh5 41 47 exd4+ �d5 48 .txf7+ <ot>e4 49 d5 �4
gxh5 h6, since Black only has one weakness, (D)
which he can defend adequately.
40 .tb5 41 f4
•••
Not 41 �c3? .ie2.

41 .Jld7 42 �c3 <ot>d6 43 g5 �c5 44 a3 (D)
••
-- - - �
B � B.t.B A B A
- - -·�
�
� �··
� � � n
a�.aRe• •
� � u .
B.tB B •
Black seems to have achieved some counter
play, but a pleasingly thematic tactical blow on
�-- _1L_ _ _ g6 shatters this illusion.

SO .ixg6! 1-0
44 .tcs Once again, the second weakness proves de
Or 44. . ..ib5 45 f5 .ie8 46 .id3 �d6 (the al cisive. After 50 ....txg6 5 1 d6 .ie8 52 g6 White
•••
ternative 46 . .. .i d7 47 fxg6 fxg6 48 .i xg6 is a wins.

� ?[] Not-so- Distant Passed Pawn
pawn ending after 68. .�d4 69 �d2 h4 70

.
g5!) 68 ... �e3 69 �g l + �xf3 70 �xd4, so

Black prepares the sacrifice by getting his h
pawn one square nearer to queening - another
case of 'do not hurry'.
68 �f2 �c3! (D)
68...�d4? is still premature: 69 �xh4 �e3
70 J.g5+ 'it>xf3 7 1 <Ji>xd2 �g3 72 'it>d3 J.g7 73
�e4 �xh3 74 � and draws.
Donner - Smyslov
Havana 1964
In normal circumstances, a same-coloured

bishop ending with an extra outside passed pawn
is a fairly simple win. Here, however, Black's
task is more difficult, because his passed pawn
is not far from the kingside and so does not de
flect the white pieces far enough away (with an
a- or b-pawn, for example, the win would be el 69 �gl �d4!
ementary). In order to win, Black needs to get Now White no longer has the move 70 �xh4,
his king to e3, which requires subtle play. this is decisive.
62 �f2+ �c3 63 J.b6 d2 64 �f2 �d3 65 70 �xd4
�b6 �f4 66 �r2 �es (DJ The main line is 70 �h2 'it>e3 7 1 �gl + �xf3
Preparing 67...�d4. 72 hd4 'it>g3 (D).
67 �gt! h4! Compared with the line beginning 67...�d4,

Black's plan is to sacrifice his bishop, so as Black's h-pawn is a vital extra step forward. 73
to get his king amongst the white pawns. How 'iit xd2 �xh3 74 �f2 g5 75 �e3 �g2 76 �xg5
ever, the immediate 67...�d4? does not win af h3 wins for Black.
ter 68 J.h2 (even 68 �xd4 leads to a drawn 70 'hd4 71 'iitxd2 'itte5 72 'it>e3 g5 0-1
...
30 JOI CHESS ENDGAME nPs
The Crippled Majority
The difference between having a healthy ma White's plan is to improve his bishop as
jority and one which is crippled, and cannot much as possible, and then play g4, creating a
produce a passed pawn, can be enough -to de passed pawn on the kingside.
cide the game. 40 i..e3 i..d 6 41 i..g5 i..eS 42 i..e7 i..d4 43
i..dS i..c3 44 i.b6 i..b4 45 g4 ( D)
. ---- �
B �. • • •
- · · •••
m • •
� - •.•.
� -fj,� ·/. - � �
� · · · � -�
• • • •
�1-- - . - .. •
45 .i..c 3
••
Sacrificing an unimportant pawn, rather than

This is a classic Exchange Lopez ending. exchanging pawns on g4, which would allow the
White is effectively a pawn up, because his white king into e4. However, even this proves
kingside majority is healthy, while Black's insufficient to hold the game.
queenside majority cannot produce a passed 46 i.xcS i..el 47 i..d6 i..c3 48 i..c7 ( D)
pawn.
34 b3!
- - . -.
. m • •
Keeping the enemy pawns doubled is the
simplest approach. Winning a pawn by taking B
twice on b5 would allow unnecessary counter • • • ··�
� • •.•.
play.
34 bxc4 35 bxc4 i..f6 36 i..d2 i..d8 37 i.c3
i.c7 38 lt>b3 W 39 i.d2 �g6 (D) � - � · - � �,
•..
. - - � -�
m ·--w a � . . . �
w �. - · • . ------
•m• ••·· �
- �•.•
•. .
White's final winning plan is to put the
bishop on d8, exchange on h5, and then bring
� - � · . �, his king to the centre. Black cannot both stop
�. B • � D� the h-pawn and defend his own a-pawn (two
• m • •
weaknesses).
48 fxg4+ 49 fxg4 i..el 50 i.d8 'iii>h 6 51
� - -- .L_
•••
i.e7 �g6 52 i.gS 1-0

There is no defence to the white king's march.
JO 1 CHESS ENDGAME TIPS 31
[r1P ?-4] Pawns Don't Move Backwards
Emanuel Lasker wrote in his Manual of Chess The key move. The advance ...b5 will leave
that one should "...distrust a pawn move, and White a very unpleasant choice.
examine its balance sheet carefully". Pawns do 26 'lii>e3 b5 27 axb5
not move backwards, so any mistaken pawn ad This grants Black a potential outside passed
vance is irreparable. pawn, but if.he allows Black to exchange on a4,
he will be unable to defend both the a4-pawn
and the c4-square. Tartakower points out that
Black also has the idea of playing ...b4, so White
must guard against the black king's entry on d4.
27 cxbS 28 g4 h6 29 h4 .t.d7 ( D)
•••
G. Thomas - Tartakower
Hastings 194516
It is hard at first sight to believe that White is

lost in this position, but he is. Black has two ad
vantages - the more active king, and weak 30 g5 hxgS 31 hxgS a4 32 'lii>r4 a3 33 .t.bl
nesses to attack at b3 and a4. If White's pawn (D)
were on a2, he would be drawing, but unfortu
nately, he has just disregarded Lasker's advice
and played the dreadful move 23 a2-a4?. It is
enough to seal his fate.
23 'itd6 24 f3 .
•••
24 .t.xh7 is impossible because of 24...g6.

24 �cS 25 'iP!2 c6! (D)
•••
33 b4! 0-1
•••
Black creates an opening on the queenside to

allow his king to penetrate to b2. An example of
how the smallest subtleties (here, the position
of White's a-pawn) can make the difference be
tween a win or a draw.
[.�--�--] The Power of Two Passed Pawns

�5
In opposite-coloured bishop endings, setting up Botvinnik sacrifices two pawns to get a sec
two passed pawns, at least two files apart, is of ond, distant passed pawn:
ten the key to winning. 59 gS!! 60 fxgS ( D)
60 hxg5 h4 6 1 iLd6 iLf5 62 g6 .llxg6 63 f5
•••
.i.xf5 64 'ifi>xb3 <t>g2 and Black wins.
In Basic Chess Endings, Fine quotes this po

sition as a win for White after l Ji.f3 and 2 'ifi>e6.
In such positions, Black has to use his bishop to 60 d4+! 61 exd4 'it>g3 62 iLaJ
•••
stop one pawn, and his king the other. White 62 iLe7 <t>xh4 63 g6+ 'ili>g4 wins for Black.
then simply brings his own king round to sup 62 'ifi>xh4 63 'ia>d3 'ia>xg5 64 'ili>e4 h4 65 'ili>f3
•••
port the pawn which is stopped by the enemy .i.dS+ 0-1

bishop. The black king marches to c2.
However, if one changes the position to put
the f5-pawn on e5 instead, it is a draw. With the And here is another world champion, using
pawns so close together, the black king is able the same idea:
to help his bishop in stopping the advance of
both pawns.
The classic practical example of creating a

second, distant passed pawn is the following
position:
Spassky - Jonsson
Stockholm 1 969
The two queenside passed pawns are too

close together, and another is needed, further
away. Spassky won by:
Kotov - Botvinnik 46 h4! gxh4 47 g5 h3 48 c5+iLxc5 49 g6 h2
USSR Ch (Moscow) 1955 SO gT e2 51 g8'ii' el'ii' 52 ii'd5+ 1-0
101 CHESS ENDGAME IlPS 33
I TIP 2 6- ] The Pow�r of Conn ected Passed Pawns
In the absence of two widely-separated passed Black does all he can to prevent White from
pawns, the other main weapon in opposite carrying out his plan. White's king is tied to f3,
coloured bishop endings is to create two con and he cannot play f4, because then he will
nected passed pawns. never be able to force the capture ... gxh5; Black
will simply leave his bishop on dl and answer
h5 with ....i. xh5. White therefore needs to sac
rifice the h-pawn, in order to force ... gxh5, but
first he regroups his pieces.
45i.d4 rJ;f8 46 �e3 rJ;f7 47i.e5 �e6 (D)
Smyslov - Stein
USSR Ch (Moscow) 1 969
White has an extra pawn, and his two queen

side pawns hold Black's three. His winning
plan is to set up connected passed pawns on the 48 .i.gJ! '1;f6
e- and f-files. If 48...�d7, Smyslov gives the variation 49
42 g4 hxg4 h5 ! gxh5 50 f4 c5 51 bxc5 �c6 52 �d4 �b5 53
'This is forced, else after gxh5 gxh5, White f5 winning.
has his two connected passed pawns anyway. 49 .i. f4 �e6 (D)
43 rhg4!
The point. He now intends to play h5, to
eliminate the g6-pawn, after which the pawns
on e4 and f3 will march home unopposed.
43....i. dl 44 coW4 rJ;{7 (D)
so h5! gxh5 51 .i. g3 1-0

There is no defence to the march of the con
nected passed pawns, while Black's h-pawn is
easily stopped by White's bishop.
-
� 27- J
TIP
---- - -
Barrier Reefs ( Part 3)
We have seen two examples of barriers already. 35 . .Wf5 would draw, since it keeps the white
.
Here, two more are involved. king back. Yanofsky then gives the illustrative
line 36 i.f8 g6 37 �d4 i.g2 38 �c5 �e6,
drawing easily.
36 �4! g6 37 g4! hxg4 38 �g4 (D)
Euwe - Yanofsky
Groningen 1946
28 i.c5!
Winning another pawn, since after 28 ...dxc5? The attempt to bring the black king to the
29 a6, the pawn queens. However, Black should queenside to relieve the bishop fails after
still draw. 38 ...'it>d? 39 'Ot>g5 i.e4 40 h4 'iitc 8 4 1 a811'+
28 i.dJ 29 i.xd6 i.xe4 30 a6 cS 31 i.xcS
••• i.xa8 42 'it>xg6, when the two distant passed
hS 32 �2 i.dJ 33 a7 i.e4 34 g3W e6 35 'ite3 pawns win easily.
(D) 39 'itgS rM7 40 i.d4! i.g2 41 h4 i.hl 42 b4
i.g2 43 bS i.hl 44 i.f6! i.g2 (D)
B
w
Despite his two-pawn deficit, Black is draw

ing. His bishop stops the queenside pawns, and White has achieved his desired formation
his king can stop the white king from getting to and now shows his hand.
the queenside to help. However, things are not 4S h5! gxhS 46 �5 1-0
so simple as they seem, and a casual move by The white king will reach c7, while his king
Black now proves the decisive mistake. and bishop also provide a barrier which Black's
35 i.g2?
.•• king cannot cross; e.g., 46 ...�e8 47 �e6, etc.
- --1
28
- ·-·
Pawn Placement
In his wonderful boo k Learn from the Leg

In most endings, one should place one's pawns 40 iJ..g 7?
on the opposite-coloured squares from one's
bishop, so as to avoid creating a bad bishop. ends, Mihai Marin shows that White could still
However, in opposite-coloured bishop endings, j ust draw by 40 �d2 �xb2 41 <ifi>xdl c4! 42
the defender should generally place his pawns .i.g7+ c3 43 �e2! , although very accurate play
on the same colour squares as the bishop, so is required. After the text-move, Karpov wins
that it can defend them, while the enemy bishop by a breakthrough on the queenside.
cannot atutck them. 40 �c2 41 .teS .tbs (D)
..•
Ljubojevic - Karpov The bishop is coming to b3.

Milan 1975 42 .i. f6 iJ..f 7 43 .te5 iJ..b3 44 .i. g7 bS 4S
.i.f8 c4 46 .i.g7 ( D)
3S hS?
Breaking the golden rule. 35 g5 and 36 f4 is
correct.
3S gxhs 36 gxh5 rs+ 37 �e3 �ds
••.
Now the white king is tied to defending his

pawns. By the time he gets them both onto dark
squares, the black king has penetrated.
38 h6 �c4 39 f4 �b3 (D)
46 b4! 47 �d4
•••
If 47 axb4, then 47 ...c3! 48 bxc3 (or 48

.i.xc3 a3) 48 ...iJ.. c4! ! , beautifully nailing the
c3-pawn to the spot, after which the a-pawn
queens.
47 c3 48 bxc3 bxa3 49 c4 a2 SO �cS �bl
Sl <ili>b4 al'ik 52 JJ..xal �al S3 cS '>ti>b2 54 c6
•.•
a3 S5 c7 .i. e6 S6 �cS a2 57 �d6 .i. c8 0-1

36 JOI CHESS ENDGAME TIPS
�TIP- 29
- -i]
- ·- · · - - ·
Knight vs Bad B ishop
since just pushing the a-pawn by itself is not ef

fective.
34 g4?
B This makes Black's task easier, since now
the g-pawn will be fixed as another weakness.
He should sit tight, although the position is still
lost.
34 <M7 35 iLd2 a4 36 �cl <M6 37 �a3
•••
g5! (D)
Pinter - Timman
Las Pa/mas IZ I 982
In blocked positions, bishops are usually a li

ability, and this is a drastic example. In a typical
Nimzo-Indian pawn-structure, White has seri
ous queenside pawn-weaknesses on a4, c4 and
c3, and his bishop is badly obstructed. Even
though W hite's bishop is theoretically 'good',
it is of little help. Fixing the g4-pawn.
29 aS!
••• 38 h5 <J;e7
Black fixes the a4-pawn, and there i s n o de Now the final winning plan is clear. The
fence to ... lllc8-b6xa4. White's bishop simply knight will move to f6, forcing the white king to
has no route to get behind the black pawns, for come to f3. Then Black's king will penetrate on
example to attack the base of Black's pawn the queenside.
chain at d6. He can only await events. 39 �cl llld7 40 'iii>e2 lllf6 41 <M3 �d7 (D)
30 �e2 lllc8 31 �d3 lllb6 32 h4 lllxa4 33
�cl lllb6 (D)
0-1
Black will play his king to a5 and then break
Black's only remaining problem is to find a with ... b5, winning easily. The white bishop re
way to penetrate W hite's camp with his king, mains helpless.
The Agile Knight
Although knights lack the range of bishops, 52 1'.xb4 axb4 53 a5 (D)

their agility is nevertheless considerable, and
should not be under-estimated.
. -- - �
B � • • ••
. . •.�,
� •.• •
- u
- � � -
�·�
�
· ·<it· -
. . . �
IL • -u · •-u
How does the knight get back?
53 lbhS!!
•••
Gheorghiu - Yusupov A paradoxical move, stepping further away

Lucerne Wcht 1 985 from the a-pawn.
54 'it>e3
Black has a classic knight vs bad bishop po After 54 a6 lZ'if4+ 55 �e3 lbe6 56 a7 lbc7
sition, but his king has no route into the white the knight gets back in time.
position. 54 lZ'if6 55 'it>f4 'it>xh6 56 a6 lZ'id7 57 a7
..•
45 f4! 46 'it>e2
••• lZ'ib6
46 gxf4 ltJ<i6 47 f3 gxf3 48 'it>e3 lZ'if5+ 49 Once again, the black knight arrives just in
'it>xf3 lbxd4+ 50 �g4 lZ'ixb3 and Black wins. time.
46 lZ'id6!
••• 58 �4 �g6 59 �4 'iii>r6 (D)
The knight is heading to f5, from where it
will attack all of White's weak pawns.
• • - -�
� • • •.
47 'iii>d3
47 f3 is trickier, but after 47...lZ'if5 48 fxg4 w
lbxd4+ 49 �d3 lZ'if3 50 �f2 lbe5+ 5 1 �e2 f3+
� .
. �
- �, � v
� ••• .
52 �fl lZ'ixg4 (Yusupov), Black should win.
47 lbr5 48 h5 fxg3 49 fxg3 'it>f6 SO h6
m D m �
•••
�g6!
Black's last problem is meeting the threat of
�xb4, creating a passed a-pawn (knights hate
·�· . .
rook's pawns!). Thus, not 50...lZ'ixh6?? 5 1 �xb4 . . . �
and it is White who wins.
51 �d2 lbxg3!
� . - .
A finely calculated move. White's a-pawn 60 �g4 lbas 0-1
appears unstoppable, but Yusupov has it all Both 6 1 'iiif4 h5 and 6 1 'it>h5 �f5 are easily
worked out. winning for Black.
[ TIP
-· ·__!!_J
� 1-1 Knights are Better than Bishops
at Attacking Weak Pawns
When there are weak pawns to attack, knights Putting another pawn on a light square and
can often be better than bishops, even on an weakening g5. 28... g5 is better.
open board. This is largely because they can at 29 b4!
tack squares of either colour, unlike a bishop. White immediately seizes the weakened g5-
Thus, the pawns cannot 'hide' from a knight, square.
whereas they can sometimes do so from a 29 �c6 30 ttJc3 'iite5 31 g3 h6 32 f4+
•••
bishop. Now there is another weakness on g6.

32 'iii>d6 (D)
•••
Gligoric - Trifunovic
Budapest 1 948 33 tbdl!
The knight is coming to c4, to drive back the
Black's isolated queenside pawns would be black king.
immune from attack by a white bishop on a4, 33 'ii>d5 34 'ii>d2 'it>d6
If J4.. �d4, the ting will be driven back by
•••
whereas there is no hiding place from the .
knight. lbe3-c2+, etc.

26 :d3! :xd3 35 ltJeJ .i. d5?! 36 b5! �f7 37 hxg6 �xg6
Gligoric claims this is forced, else the c5- Now f5 is another weakness.
pawn falls immediately. However, this does not 38 'iPc3 'iPc6 (D)
seem to be the case after 26 ...'iii>d6 27 :c3 :ds,
which may well therefore be a tougher defence.
27 cxd3 'ittd6 28 'it>e3 f5? (D)
w
White has a winning advantage, which he

converted instructively. For the remaining
moves, see Tip 71.
Knights are Better than Bishops

at Attacking Weak Pawns (Part 2)
We saw in Tip 3 1 that a knight is often better 31 lt:Jd4 c5
than a bishop at attacking weak pawns. This is Forced, in view of the threat of 32 e4, but
another, even clearer, example. now Black's pawns are weakened further.
33 ... aS 34 lbc3 i.c6 35 'ifi>c2, followed by

32 bxc5 bxc5 33 lbb5 a6
�d3-c4, wins for White.

34 lbc,7 i..c4 35 lt:Je8 rs (D)
Note the way the white knight is able to
chase Black's pawn-weaknesses from square to
square, a perfect illustration of the knight's su
periority in such positions.
Botvinnik - Tai
Moscow Wch (1) 1 961
23i.xf6! gxf6 24 b4i.rs

Botvinnik recommended 24 ...i..dS as better,
but Tai fails to understand that the knight is su-
•
perior to the bishop.

25 lbb3
White wants to exchange the rooks, leaving 36 h4!
a pure lb vsi. ending. Do not hurry! 36 lt:Jf6+? ! allows counterplay
25 i.d3+ 26 �el b6 27 l:r.acli.e4 28 f3
••• by 36 ... �g7 37 lbd7 i.fl.
l:txdl+ 29 l:r.xdl lhdl+ 30 'iit>xdli.dS? (D) 36 � 37 lbd6
•••
30...i.fS 3 l lbd4 i.d7 is better, but White re Not 37 lbf6? ! �e7 38 lbxh7?? f6 and Black
tains the advantage. wins.
37 i.fi 38 g3 (D)
•••
Despite the fact that the position is open and

there are pawns on both sides of the board, the See move 36 !
knight is much stronger than the bishop here, 38 'ifi>e7 39 lt:JxfS+ 'ifi>e6 40 e4 'iit>e5 41 �d2
••.
because it has weak pawns to attack. 1-0

Bishop vs Knight on an Open Board
Gligoric's 39 ...ltJh6 ! offers better drawing

chances after 40 "'d3 tiJf5 41 'iil>c4 ltJxh4 42
'iii>xc5 �e5.
w 40 b5 tiJb6 41 "'d3 �e5 42 .i. a8 (D)
i.- - • �
B '� . . .
. . .
� m m m
� -
, ,
- -�
. . . �
Spassky Fischer . ·�· .
�· . -��
-
Santa Monica 1966
In general, in an open position with pawns

� . . -
on both sides of the board, the bishop is supe 42 'ltd6
•••
rior to a knight, because of its longer range. Black can win the h-pawn by 42 ...'ii>f4 43
Here, White's task is complicated by the fact 'it>c4 'it>g4 44 .i.f3+ 'lth4 45 'it>xc5 g4 46 .i.e4
that there are relatively few pawns remaining ¢ixh5 47 a4, but then his forces are too far from
on the board. the queenside.
35 b4 43 �c4 g4 44a4 tiJg8 45 aS ttJb6 46 .i. e4 (D)
Fixing g6, and setting up a potential passed
pawn by a subsequent g4 and h5.
35 lbc4 36 �e2 ltJe5 37 �e3 'iPf6 38 �4
ttJf7 (D)
•••
Note how the bishop is able to influence

events on both wings, whereas the knight is tied
to blockading the h-pawn. This is the key to the
bishop's superiority in positions with pawns on
39 'it>e3? both wings.
This inaccuracy endangers the win. 39 .i. d5 46 g3 47 �b5 ltJg8 48 .i. bl lbb6 49 �a6
'itc6 50 .i.a2
••.
g5+ 40 hxg5+ ltJxg5 4 1 .i.c4 is better, when

Black is in zugzwang and cannot prevent the 50 �xa7? c4 allows unnecessary counter
white king from penetrating to the queenside. play (do not hurry !).
39 gS?
••• 1-0
, ..34
.. l
! TIP
L___ . -
Fischer's Classic
Note how Fischer uses a series of bishop

moves to lose tempi and gradually force Black
into zugzwang. His first aim is to get his king to
a6.
55 �b7 56 .if3+ (D)
•••
Fischer - Taimanov
Vancouver Ct (4) 1971
Fischer gives a classic demonstration of the

superiority of bishop over knight with pawns on
both sides of the board. The winning task is 56 �c7
.••
complicated by several factors, notably the sym After 56 ... �a7 any waiting move forces the
metrical pawn-structure and the fact that the white king's entry to c6.
kingside is blocked. On the other hand, Black's 57 'iii>a6 lbc8 58i.d5 lbe7 59i.c4 lbc6 60
kingside pawns are fixed on light squares, with .i.f7 lbe7 61 .i.e8
g6 being especially weak, and this eventually Zugzwang.
decides matters. 61 �dS (D)
•••
45 'iti>d3 li1e7 46i.e8

Tying down the knight.
46 �dS 47 .i.f7+ �d6 48 �c4 'iii>c6 49
•••
i.e8+ �b7 50 'iti>b5 lbc8 (D)
62i.xg6!
The crowning blow. White will win too many
pawns.
62 lbxg6 63 �xb6 'iii>d7 64 'ihc5 lbe7 65
b4 axb4 66cxb4 lbc8 67 a5 lbd6 68 b5 lbe4+
•••
With a small threat...

51 .i.c6+ 69 'it>b6 'iii>c8 70 �c6 'iii>b8 71 b6 1-0
51 iixg6?? lbd6# would be unfortunate. Immaculate play by Fischer, showing per
51 'it>c7 52i.d5 lbe7 53i.f7 �b7 54 .i.b3
.•• fectly how to exploit the advantage of bishop
<i;a7 55 .idl over knight in such a position.
Knight on the Rim
Probably the knight's biggest weakness is its Black now has to play 40...lbxd5 and hope
lack of mobility near the edge of the board. for the best. Instead...
40 JW.4? 41 lba4 lbxa4
••
The trap closes:

42 .td4!
w Now 42. . . e5 43 d6 queens the pawn, while
after 42 ... exd5 43 ..tr>g3, the white king simply
marches over and takes the trapped knight.
42 <l;g7
•..
This allows White to establish a decisive

passed d-pawn.
43 d6 � 44 i.xf6 lbb6 45 <l;g3 lbd7 46
i.b2 �g8 47 <l;g4 <i;b7 (D)
Fischer - Addison
USA Open (Cleveland) 1957
29 .teS!
Paralysing the knight. If it moves, the pawn
ending is lost after the exchange. Consequently,
Black is helpless, as White simply advances his
queenside pawn-majority, while Black's own
kingside majority is crippled by the doubled
pawn.
29 'it>bS 30 'it>d3 g4 31 b4 a6 32 a4 gxf3 33
••.
gxf3 �h4 34 b5 axb5 35 a5! �h3 36 c6 1-0 48 'it>h5 lbf8 49 g4 lbd7 50 g5 hxg5 51
�g5 <itg8 52 i.d4 <ifr>h7 53 h4 (D)
The same idea of stalemating the knight on
the edge of the board occurs in the next example.
53 lbf8?
•.•
This allows the white king in immediately,

Belavenets - llyin-Zhenevsky but against other moves, White wins·by h5-h6
USSR Ch (Tbilisi) 1937 and i.g7, followed by the transfer of the king
via b5 and c6, to promote the d-pawn.
40 d5! 54 �f6 1-0
[ TIP 3 6 J
-
The Triumph of the Two Bishops
40 �e3
Not 40 ..llxe5? g5+, when Black would draw
easily.
40 ..t>fs 41 'iPd4 liJn 42 i.as
•..
Do not hurry! Before pushing the pawns,

Ivanchuk improves his pieces to the maximum.
42 ltJe5 43 ..llb4+ <t>e8 44 <itcS liJg6 45
•..
i.d2 liJeS 46 'it>d6 ..lla4 (D)
lvanchuk - Kariakin
Wijk aan 'Zee 2006
The earlier moves can be seen in Tip 86. Al

though all the pawns are on one side, White's
two bishops are decisive.
33 h6 34 'it>g3 l:.c8 35 l:.d6 .Iles 36 ..lle2
.••
It is important to retain the bishop-pair. The

great strength of two bishops is that, between 47 f4
them, they can control every square on the Beginning the process of driving Black's
board, whereas one bishop on its own can only pieces out of the centre.
attack 32 out of the 64 squares. This is the key 47 liJf7+ 48 �c7 i.d7 49 i.b4 i.a4 50
•.•
as to why 2i.s vs i.+ltJ is frequently a much i.c4 liJh8 51 g3 liJf7 52 h4 liJh8 53 .i.d6 liJg6
bigger advantage than i. vs ltJ. It is also the rea 54 .i.e2 liJh8 55 .i.c4 ltJg6 (D)
son why the defender often tries to exchange
his remaining bishop for its opposite number.
White would have few winning chances after
36 ..ll xe8?.
36 ..l:.c6 37 lhc6 ltJxc6 38 �4 ltJes 39 ..llc7
•.
..lld7 (D)
56 e5
Finally creating a passed pawn.
56 fxe5 57 fxe5 ltJe7 58 e6 g6 59 ..llc5 liJfS
.••
60 ..lld3 liJe7 61 'ifi>d6 h5 62 'iti>e5 liJf5 63 'iti>f6

..llc6 64 e7 1-0
i.c4-f7+ follows.
Know the Basics
There are not many endgame positions which a

player needs to know by heart, but there are a
few, mostly in rook endings. Here is one of
them. w
56 'ifi>fs+ ! 'it>g6 57 f7 'it>f6 58 �g8 wins.

56 .l:a8!
•.
Rule 3. Now the position is a draw once

again.
Capablanca - Menchik 57 l:te7 l:ta6??
Hastings 1929130 57 ...l::tb8! (Rule 3) draws.
58 �+!
This position is a draw, but Black must know This time, Capablanca finds the correct
how to defend. There are three rules to bear in method of exploiting Black's mistake.
mind: 58...�g6 59 r1 :as+ 60 l:te8 l:ta7 6t l:te6+
l ) Black's rook stays at the side and threat 'iii>h7 (D)
ens checks.
2) If permitted, the black king approaches
the pawn.
3) Black must not allow the white king to w
reach the 8th rank, unless White first puts his
rook there.
The correct defence is 55 ...l:tb8! (Rule 3);
e.g., 56 l:td7 (waiting; 56 lte8 ltb7+) 56 ... l:ta8
(also waiting) 57 l:td l l:ta7+ 58 'it>e8 'iti>g6! (D)
(Rule 2).
Now after 59 f7, Black draws by 59 ... l:ta8+ !
(not 59 ... ltxf7?? 60 l:tg l + 'it>f6 6 1 l:tfl + -
never forget tactics !) 60 'it>e7 l:ta7+ 61 'it>e6
l:ta6+. 62 'it>e8??
62 l:tel wins.
Instead, the game provided an infamous com- 62...l:ta8+ 63 �e7 lh7+??
edy of errors. Play continued: Incredible. 63 ...c;i;>g7 (Rule 2) draws.
55 .l:ta6?? 64 'it>f6 1-0
After 64...l:ta8 even on that form, Capablanca
••
Breaking Rule 3.
S6 l:td7?? would presumably have found 65 l:te8.
�I��-� ] The Importance of King-Shelter
As any good Christian will tell you, in life,

sheltering the homeless is more important than
accumulating material possessions. This is
sometimes true in rook endings as well.
A. Zaitsev Hubner
-
Biisum 1 969
The same idea cost a world-class player half

Kotov - Eliskases a point here.
Saltsjobaden IZ 1952 57 l:.bl? 58 �h5 .:tgl
•.•
Now the familiar breakthrough:

Without the black pawn on g5, the position 59 g5 fxg5 60 r5! �rs 61 r6 1-o
is drawn, because Black would simply play There is no defence to 62 �g6.
l . . .l:tfl and then keep checking the white king
from the rear. But here, the king is able to use Larsen subsequently pointed out that Black
Black's g-pawn as a shelter from the checks, could have drawn in the initial position by
and this is enough for White to win. means of the preliminary 57 ...l:.b4 ! 58 f5, and
47 .l1f2 48 �e6 l:.e2+
•• only now 58 ...l:.bl (D), since the breakthrough
Forced, in view of the mate threat, but now idea seen in the game is no longer possible.
the white king is allowed into g6.
49 �5 g4 (D)
For example, 59 �h5 (59 g5 l:.hl +) 59...l:.gl

and the ending after 60 �g6 l:.xg4+ 61 �xf6
50 �g6! l:.f2 51 f7+ � 52 l:tb8+ �e7 53 l:.a4 is a theoretical draw, since Black's king is
l:.e8+ 1-0 on the short side of the pawn.
Know the Basics, Part 2
82 .l:tdS
••
Step 2.
83 �e3 �b2 (D)
Kochiev - Smyslov
Lvov Z 1978
lbis position is a theoretical draw, but White Step 3.

must play accurately. In the game, White played 84 lk4 <J;;b3 0-1
a losing blunder:
78 <J;;e2?? Note that in the original position, the de
Black now wins, using a process known as the fence 78 :tc2?? would also lose, because the
Combined Method. lbis involves three steps: rook then has insufficient checking distance:
1) Black advances his king as far as it can go 78 .. <J;;b5 79 l:r.b2+ �a4 80 l:tc2 'it>b4 8 1 l:tb2+
.
(normally two squares diagonally in front of the 'it>c3 (D) , etc.

pawn, in this case, a3).
2) His rook then steps up to d5, to defend
the pawn, while keeping the white king cut off.
3) Then the black king attacks the enemy
rook, and forces the pawn forward.
The game continued:
78 �bS 79 l:tbl+ <J;;a4 80 l:r.ct <J;;b4 81
•••
l:tbl + <;#;a3
Step 1 completed.
82 .U.ct (D)
So how should White defend in the initial

position? The simplest answer is 78 �e4 !, the
point being that Black cannot carry out Step 2
of the Combined Method, because his rook
does not have access to d5. For this reason, in
such positions, the 3rd and 4th ranks are known
as the Absolutely Safe Ranks, because the de
fender always draws if his king is on those
ranks.
Know the Basics, Part 3
Tai I . Zaitsev 73 .llel! 0-1

In order to secure the requisite checking dis
-
••
USSR Cht (Riga) 1968

tance for his rook, he must evict Black's rook
As we have discussed previously, in such po from the first rank. But then after 74 <Ji>d2 .:te8
sitions, the 3rd and 4th ranks are the Absolutely 75 .:tbl , his king is only on the Relatively Safe
Safe Ranks for the white king, while the 2nd Rank and it is Black's move, and he is able to
rank is the Relatively Safe Rank. In the latter win: 75 ...g5 76 llhl + 'it>g6 77 l:tgl (D).
case, White draws only if it is his move.
In this case, even though his king is cut off
by two files, White is still able to draw, thanks
to the fact that Black's pawn is on the knight's
file. White needs to get his rook in front of the
pawn, with maximum checking distance (i.e. to
g1) and also to keep his king on one of the Ab
solutely Safe Ranks. Since his king is already
on one of them, he should attend to the rook's
position, with 73 l:.bl !, after which he draws in
standard fashion by 73 ... g5 74 �d3 lle8 75
llh l+ ..tg6 76 llgl (D).
Now 76 ...lleS 77 �d4, etc.
77 ...lle5 78 'itd3 �f5 (because the pawn is
Instead, Tal played a highly plausible move: on the knight's file, a slightly amended version
73 'ifi>d3?? of the normal Combined Method is needed) 79
However, he had to resign after Black's reply: �d4 .:te4+ 80 'it>d3 g4, etc.
48 I 0I CHESS ENDGAME TIPS
When Two Extra Pawns Don't Win
The English grandmaster Mark Hebden once

told me that he learned about basic rook end
ings by studying this ending of l:t+rook's
pawn+bishop's pawn vs .:.. a classic case where
most positions are drawn, despite White's ma
terial advantage. Since it is so instructive, we
shall examine it in detail.
With wonderful clarity and lucidity, Botvin
nik summed up the essential points of the end
ing as follows:
1 . White wins if the black king gets cut off
on the back rank, so the defender must avoid
this.
2. When the pawns reach f5 and h5, the 79 l:ta6+ � 80 �g4 l:tb4+ 81 f4 l:tb2 82
black king should be on f7. h5 'it>g7 83 l:.g6+ W 84 l:tg5 l%bl (D)
3. If the threat is to drive it to the back rank,
it goes to g7, so as to have h6 available.
4. Black's rook is generally best around al ,
so as to have the chance of harassing the white
king both from the flank and the rear.
Bearing the above rules in mind, let us now

see how they work out in practice:
85 l:tc5
If 85 h6, Rule 4 again comes to Black's res
cue: 85 ...l:tal ! (but not 85 ...l:tgl +? 86 �f5 l:r.h l
87 l:tg7+! �f8, when the king has been driven
to the fatal 8th rank and 88 'it>g6 wins) 86 h7
l:tgl + (only now, since White no longer has a
rook check on g7) 87 �5 l:thl draws.
85 �6 86 l:tc6+ 'it>g7!
•••
Gligoric Smyslov Rule 3. Instead 86...�f7? loses to 87 �g5

Moscow (Chigorin mem) 1 947
-
llgl + 88 �5 l:thl 89 l:tc7+, because the black

king again ends up driven to the back rank.
70 l:r.g2 l:ta4 71 l:th2 'it>g6 72 'it>g2 l:tal 87 �g5
Following Rule 4. For a long time Black can Threatening 88 %:tc7+, so. . .
simply keep his rook on al , taking time out to 87 ..:r.gl+! 88 <Ms l:tal! (D)
••
check on g 1 whenever the white king occupies Again following Rule 4. Having secured its
a threatening position on the g-file. king from being driven back, the rook returns to
73 �g3 l:tgl+ 74 l:tg2 l:thl 75 'iii>f4+ <M6 76 its best post. Instead, 88 ...l:th 1 is unthematic, as
�g4 'it>g6 77 l:ta2 .l:f.gl+ 78 'it1'4 l:tbl (D) it puts the rook out of play (no flank attack), but
Again, Rule 4. even this draws after, e.g., 89 h6+ �h7 90 �e5
1 01 CHESS ENDGAME TIPS 49
l:.e l + 9 1 'iiif6 'iitxh6. This shows that Black has 92 l:.e5 llal 93 l:.d5 :n
some choice in how he holds this ending. This draws, but 93 ... l:tbl (Rule 4) is the most
thematic course.
94 l:.d4 l:.al! 95 l:.d6 l:.aS+ 96 'itg4 (D)
,-
- -- -- - . . .
w • • . • -
. -
•:• . .
� . ·�· �
• • D •
� . . .
• • • •
' __ . . .
S9 l:.c7+ 'iith6 (D)
96 .l:.al
••
96...l:.b5 also draws.

97 l:.e6 l:tgl+ 98 'ii>t'S l:tal 99 h6+ �h7! 100
l:.d6 l:.a2 101 'iitg5 l:.g2+ 102 'ii>r6 �6 103
�e7+ �h7 104 rs l:.e2+ 105 l:.e6 l:.a2 106 f6
(D)
• • • •
B • • = ••
-
. - � .
•:u
Rule 3. • • • •
90 l:.e7 • • • •
90
.
�f6 l:.a6+ 9 1 �f7 �xh5 is a draw.
� . . .
•• • • •
90 . .l:.bl 91 :es (D)
. • : • --� � - . .
B � • • • 106 JlaS!
. . . �
••
This essential move reaches a standard draw.
� a B<iti>a �
For more on this position, see Tip 37.
B B n .
107 'ii>r7 'iiih6 lOS l:.el l:.a7+ 109 l:.e7 l:taS
� � u 1 10 l:td7 'iiih7 111 l:.dl lla7+ 1 12 �e6 l:ta6+
• • • • 1 13 l:.d6 :as 1 14 l:.d4 'iitgs 1 15 l:.g4+ 'iitf'8
11z_11z
• • • •
• K • ___I_ This may all seem at first sight to be a lot to
learn, but actually, the Botvinnik rules are ef
91. �g7! •• fectively all one needs to know, as I hope this
The king returns to g7 as soon as permitted. excellent practical example demonstrates.
I T•P 42 I Pawns on the Same Side
Petrosian - Keres 51 g4
USSR Ch (Moscow) 1 951 There is no other way for White to make
progress.
Many rook endings with pawns on one side 51 hxg4 52 fxg4 <J;g7 53 'iti>gS :n. 54 l:.e4
.:.n ss hS gxhS 56 gxhs r6+ s1 �g4 .:.n 58
•••
are drawn. This 4 vs 3 structure is typical. The

h6+ �g6
to h5, as he has already succeeded in doing
key defensive idea for Black is to get his pawn
Not 58 ...�xh6'!? 59 e6.
here. lfl.lh.
32 h4
32 h3 and 33 g4 just leads to simplification, Against the f7-g6-h5 set-up, there is one
which is the main point of Black's pawn set-up. plan at White's disposal that needs to be known
32 .:.rs+ 33 'iti>g2 nas 34 'ifi>h3 l:.a4 35 l:.d2
••• in more detail.
'iti>eS 36 l:b2 'it>r6 37 l:.b5 l:.a2 38 'iti>g2 l:!.a4 39
<t>t'3 l:a3 40 'iti>f4 l:.a2 (D)
-- - - �
w B m m•m
• -
- • -
-·�
�•: ••-m•�•,
d d �
- • R R
- � � �
zm m D �
� Bt_ lt . • S. lonov - Karasev
Leningrad Ch 1 983
41 f3
The only plan is to create a passed e-pawn. This ending is very similar to the one from
41 .l:.e2 42 e4 l:!.el 43 l:r.b6+ <3;g7 44 l:.a6 the better-known game Piket-Kasparov, Inter
net 2000. White threatens 56 e6. Black can de
••
l:.bl 45 l:.c6 l:.gl

Tying White to the defence of g3. fend with his rook, either from the e-file, or
46 l:.c2 �6 47 l:!.a2 <3;g7 48 lle2 <3;!6 49 from the a-file. Only one is correct.
l:te3 1;g7 so es � (D) 55 l:.a6!
•••
In the game, Black chose 55 ...l:te3? and lost White's position is cramped by the black h
after 56 'it>g5 .:.e2 57 .:.e7 ! .:.e4 58 e6! (the pawn.
point) 58 ....:.xe6 59 .:.xe6 fxe6 60 h3 ! 'it>f7 6 1 45 g5
�h6 �f6 62 g4 h4 63 g5+ c;t;;>f5 64 q;g7 'ifr>xf4
•••
The plan is to push the pawn to g4, and White

65 c;t;;>xg6 e5 66 �6 e4 67 g6 e3 68 g7 e2 69 can scarcely contemplate stopping this by 46 h3,
g8'iV e l li' 70 'iVg4+ 1 -0 (in view of70...'ifr>e3 7 1 because of the resulting weakness on g3.
'iVe6+) . I t i s important to note that i f the h 46 .:.b2
pawns were missing (i.e. if the position were 3 46 g4+ leads to a lost 2 vs l position after
vs 2), this king and pawn ending would be a 46... hxg3+ 47 hxg3 g4 48 fxg4+ �xg4.
draw, so it would not matter which defence 46...g4! 47 fxg4+
Black adopted when White threatened e6. With 47 .l:.c2 gxf3 48 gxf3 <t>f'4, followed by ...f5,
the h-pawns on the board, however, only the de ...h3 and ... .:.ht wins for Black (Kasparov).
fence from the a-file works. 41...�g4 (D)
56 '/J.e7 .:.b6 57 �g5
Now 57 e6 is met by 57 ...<;t>f6! 58 .:.xn+
<;t>xe6, drawing.
57 l:tb5!
.••
Stopping e6.
58 f5 gxf5!
58 ....:.xe5? leads to a lost pawn ending after
59 .:.xe5 f6+ 60 ..t>f4 fxe5+ 6 1 ..t>xe5 gxf5 62
�xf5.
59 e6
59 �xf5 l:.a5 60 h4 l:.b5 6 1 �g5 � and 59
@xh5 @f8 60 l:.a7 l:.xe5 are also drawn.
59...f4+! 60 �xf4 �f6
with a draw. Now the e-pawn is passed.
48 .:.b6 .:.ai+ 49 'it;gt f5 50 l:.g6+ cM4 51
If the defender cannot get his pawn to h5 (or l:.h6 �e3 52 h3 e4 53 .:.xh4 f4 54 l:.h8 l:.al + 55
h4 in the case of White being the defender), the �h2 <M2 56 .:.rs
defence is much more difficult, if not hopeless, 56 .:.es f3 57 l:.xe4 (D) (51 gxf3 e3!).
as many practical examples have shown.
Now 57 ....:.h l + ! wins.

Harandi - Vaganian 56 ..f3 57 h4 .:.gt 58 gxf3 l:.g2+ 59 <;t>bt e3!
.
Rio de Janeiro IZ 1 979 60 .:.es l:.g3 61 <;t>h2 .:.x0 0-1

Defence from the Side
Although a rook generally belongs behind a

passed pawn, it is sometimes better at the side.
• • • •
• • • •
Here are two cases where the defender uses this
idea. w
. ••• �
. . •:
x• • • �
, �
• • • • <it;
. . . �
� - -• • - - H
Szabo - Tukmakov
Buenos Aires 1970
This Vancura technique is exceptionally valu

J. Vancura able, and is applicable in many other positions.
28 Rijen, 1 924 Here White is two pawns down, but could still
draw by appropriate use of Vancura's idea How
Defending from the rear loses, because the ever, he lost as follows:
white king hides on a8: 1 ...l:.a l ? 2 :e8 � 3 66 'Otg2
l:.e2 l:tbl + 4 'Otc6 .:a 1 5 �b7 l:tb1 + 6 ¢>38 llal Moving his king in the wrong direction, but
7 a7 l:tbl 8 l:tc2 <l;e7 9 l:tc8 �d6 1 0 l:tb8 llal 1 1 the damage is not fatal yet. Mark Dvoretsky
ltb7, etc. showed that White can draw more simply by
However, Black draws by using his root going directly for the Vancura technique: 66
from the side: .:.b5 �d6 67 l%.f5 l:tal 68 ..t>h2!? a4 69 J:[f4!
1...l:tf5+! 2 ltb4 l:tf6! (fastening on to the dangerous a-pawn) 69...a3
Rather than give further, unnecessary, checks, 70 lU3 ! ltc5 7 1 J:[b3 'Otc4 72 l:tf3 ltb4 73
the rook 'fastens on' to the pawn, tying down l:tf4+!, etc.
the white rook. 66-d6 67 �2 l:ta2+ 68 ..t>et?
3 �c5 :e6 4 ..t>b5 l:te5+!
king to possible checks on h2 and makes it im
lbis is the losing move, as it exposes the
The moment White's king defends the pawn,
threatening to free his rook, Black checks it away possible for him to set up the Vaneura defence
and then resumes his surveillance of the pawn. in time. He could draw by 68 ..t>fl or 68 ..t>gl ,
5 'Otb4 l:te6! 6 ..t>c5 l:tf6 7 a7 l:ta6 keeping his king in the comer and leaving the
7 ...l:tf7?? loses to 8 l:tg8+, but after White's rook to do the work of stopping the a-pawn; e.g.
last move, there is no longer a hole on a7 for the 68 ..t>g l l:tal+ 69 �h2 a4 70 l:txh4 a3 71 l:th3 !
white king, so Black can switch to defence �c5 72 l:g3, etc.
from the rear. 68 l:tal+ 69 ..t>e2 a4 70 l:th6+ ..t>e5 71
l:th5+ ..t>f6 72 �f2 a3 73 ..t>g2 l:tcl 74 :SS
..•
8 ..t>b5 l:tal
with a draw. l:tc3 0-1
TIP _4� I The Strength of the Rook at the Side
Keeping the a-pawn as far back as possible,

while he prepares his play.
36 'iit>e3 e5!
Now Black has enough play to draw.
37 'iii>e4
37 fxe5+ is similar: 37 ...�xe5 38 'iii>d3 'iit>d5
39 �c3 �c6 40 'iii>b4 .U.e5 draws.
37 exf4 38 <t>xt'4 �e6 39 �e4 gS! (D)
•••
Yusupov - Timman
Linares Ct (7) 1992
Here, it is the stronger side who benefits

from using his rook at the side of his passed
pawn, rather than behind it.
3S l:.al??
The question marks are Yusupov's own. He
follows the usual rule, but here it fails to win. Black correctly exchanges pawns at every
Instead, there was a simple win by 35 .:.e4!, opportunity, a typical defensive device.
since the rook can also defend White's king 40 hxgS .:.xgS 41 � l:.a5 (D)
side pawns; for example, 35 ...l:.a6 36 'iii>e2 'iii>f5
37 .:.es+ �g4 38 l:.g5+ '>ti>h3 39 a5 (D) , fol
. . . �
� • • •
lowed by the march of the white king to the
queenside. w
. ••• �
,, • • •.
�· . . �
� • •�D
. . . �
�· - · ---- • �
42 llel+ <Ms 43 lle4
With its tail between its legs, the rook returns
to its rightful position, but already too late. The
position is completely drawn.
Note how the white rook both defends the a 43 J::.cS 44 .:.eJ .:.as 4S .:.a3 �eS 46 'iii>e3
••
pawn, and protects the kingside pawns against <iti>e6 47 'iii>e2 'iii>d6 48 <t>f2 c;t>e6 49 lle3+ 'iii>dS
50 .:.aJ <iti>e6 Sl �e3 h4 52 g4 cM6 S3 <t>f4 'iii>g6
54 �3 �gs 55 .:.ai h3 112.112
the black king's raid.
35 ..:.SS !
••
r = --A1: 1
�� The Strength of the Rook at the Side (Part 2)
Afek - Timman 50 hxg4 hxg4 51 'iii>g3 a2 52 .:as l:.xd7 53

Amsterdam 2002 l1bS+
53 'iii>xg4? loses to 53 ...l1d4+ followed by
This is another case where the rook proves 54...l1a4.
more effective alongside a passed pawn than be 53 <it>c2 54 llaS
•••
hind it. Black's passed pawn is the more danger 54 :cs+ 'ifr>dl 55 l1a8 l:d2 and ...�c l -bl is
ous, but subtle play is required in order to win. winning for Black.
44 l:r.b5!
••• 54 'ifr>b3
•••
44 ...�b4 would lose a tempo, and in such a po

Freeing the black king to go to b3. Instead, 54... l:td3+ 55 <it>xg4 <it>b2 56 l1xa2+ 'iii>xa2
(Nunn) looks like an easier win.
sition, every tempo is vital. 55 :tb8+ 'iii>c4 56 l:r.c8+ <it>b5 57 :as (D)
45 l1dS
45 l:e5 l:xe5 46 dxe5 �b5 wins for Black.
45 'iti>b3! 46 d5 a4 47 d6 l1b6! (D)
•••
57 .:d2!
••
Avoiding a stunning trap: 57 ... l1d3+? looks

as though it wins immediately, in view of 58
The key idea. The black rook operates from �xg4? l1d4+, followed by 59....:.a4. However,
the side of the white d-pawn, tying White's White has the fiendish trick 58 f3 ! ! , when
tra pawns, but loses both of them. Do not hurry

rook to the d-file and also allowing the black 58 ...:txf3+ 59 <it>g2 ! draws - Black has two ex
king to shelter from checks.
48 d7 l1b7 49 h3 a3! (D) - look out for every tactical trick !
49 ...gxh3+ may just about win, but there is 5S f4 gxf3 59 'iitxf3 'it>c4 60 �e3 .:.b2 61
no need to allow White a passed f-pawn. .J:Li7 q;,c3 0-1
[TIP-.- -- --
-·
461
�_J
Karpov's Masterclass
46 l:c4!
The key idea. The rook steps up to escort the
a-pawn, while maintaining its influence on the
w kingside.
46 f5 47 a4 c;t>f6 48 'it>a3 �e5 49 l:c5+! (D)
•••
Karpov - Knaak
Baden-Baden 1992
Here, the 1 2th World Champion gives a veri

table masterclass in the value of a rook at the
side of a passed pawn. As in the other cases we Continuing the same process as at move 46.
have seen, the key point is that from the side, 49 <;f;>e4 50 as h4 51 <;f;>a4 <;f;>f4 52 l:c4+!
•••
the rook performs two functions. A subtle manoeuvre, which gains a tempo.
39 l:.aJ!! 52 �es 53 l:tb4 �d5 54 l:b5+ <ifi>e4 55 l:b6
c;t>f4 56 a6 g4 57 <ifi>a5! (D)
•••
Much better than 39 a4, which leaves the

white rook out of play. Karpov wants to use his
rook to the maximum effect, both to support the
passed pawn, and also to influence events on the
kingside, where Black will seek counterplay.
39 g5!? •••
If 39 .. .l:tc2 then 40 l:f3+ 'it>e6 41 a3, fol

lowed by �dl-c l -b l , but 39 ... h5 may be a
better try.
40 <;f;>d2 �g6 41 l:tc3 .:SS 42 a3 h5 43 <;f;>c2
l:ta8 44 <;f;>b3 l:b8+ 45 'iti>a2 :as (D)
57 g3
•••
57 ...gxh3 is also met by 58 l:b4+ ! .

58 l:.b4+ �e5 59 f3 f4 60 l:te4+ Wf5 6 1 l:e2
Wf6 62 <;f;>b6 1-0
If 62 ... l:b8+ then 63 �c7 l:f8 (63 ... l:bl 64
l:a2 :te l + 65 �d6 l:c8 66 a7 l:a8 67 l:a5 is
winning for White) 64 a7 l:tf7+ 65 �b6 l:f8 66
<;f;>b7 l:.f7+ 67 �a6 l:f8 68 l:b2 wins.
A marvellous ending by Karpov.
li·-- ···--·
TIP 47
·-·· ] Shoulder-Charging
We have seen the technique of shoulder-charg assist the rook in stopping Black' s f-pawn.
ing in king and pawn endings. Here, we see However, Black could have prevented this,
how a knowledge of this idea can be vital in had he played. . .
other endings as well. 70 �e4! (D)
•••
.
Like 70... �g4?, this move frees the f-pawn,
but unlike 70. .'0t>g4?, it also blocks the return
path of the white king. Black then draws easily
after a line such as the following:
Alekhine - Bogoljubow
The Hague Wch (19) 1 929
It is clear that Black will have to sacrifice his

rook for the b-pawn. The result will then de 71 b7 f5 72 b81i' l%xb8 73 l:txb8 f4 74 'it>c5
pend on whether he can draw the resulting end f3 75 :t'8 cbe3 (D)
ing of 'iti>+: vs �+£>,. In the game, Bogoljubow
played a losing mistake:
- m -a- H �
� • • •
70 �g4? 11 b7 rs 72 bSii' l%xb8 73 .:.Xb8
•••
f4 74 �d5 f3 75 'iti>e4 (D) w

. . . �
a • • m .. �
� .
- .
� .
�
B � • • • I . . . �
. . . � � . .••
� . . . . . . �
. - �-· � _ - •-
%
• • •••
__
. . . �
76 'it>c4 f2 77 'it>c3 �e2 78 l%e8+ 'ifi>dl!
The shoulder-charging role played by the
. . - -- black king in this last line is clear. Note that
Black's error here cost him a vital half-point in
7S f2 76 l:lf8 'iti>g3 77 ct>e3 1-0
••• a game for the world championship - proof
enough, if any be needed, of the importance of
It is obvious that White won because his thoroughly understanding basic endgame prin
king was able to come back via d5 and e4, to ciples.
Keep the Furthest Pawn
Marjanovic - Mochalov Compare the previous diagram. Here, it is

Minsk 1 982 the g-pawn which will be left after the white
rook gives itself up, and the white king is a cru
It is clear that White will soon be defending cial step further away from it. 65 l:.f8 (65 l:tbl
an ending of rook vs two connected pawns. In 'itf3! 66 l:.b3+ 'itg2 wins for Black) 6S ...'itg2
such passed-pawn races, the defender will have 66 'ite4 g3 67 'ite3 fl 'ii' 68 l:.xfl 'itxfl and
to give up his rook for one pawn (the more ad Black wins.
vanced), and use his king to stop the other. It 61 f4 62 l:.bS l:.hl 63 b7 l:.h8 64 'itc7!
Black wins after 64 b8'ii' l:.xb8 65 l:.xb8 f3.
•••
therefore makes sense to push the pawn nearest

the defender's king, and retain the one furthest 64 f3! 6S l:.hS!
•••
away. Here, that means Black should push the White defends very ingeniously, but Black is
f-pawn and keep the g-pawn. still able to win.
60 'it>g3! 6S ..:ts 66 .:.rs :gs 67 l:.g5 :rs 68 .:.rs
l:.xfS! 69 b8... f2 (D)
••• •
Correct! Not 60 ... g3? 6 1 b7 'it>f3 62 b8'ii'

l:.xb8 63 l:.xb8 g2 (D).
,�
,,,,,�,," �_m m m
w - -
.,,,,.
-�m m m
m m m•m
m m m m
� - - ·
- m m•m
� •
m
� � m �
This is winning for Black, as White will
Now 64 l:.g8 f4 65 �d5 'it>f2 66 'it>e4 draws, eventually run out of checks.
as the white king is able to stop the f-pawn. 70 'itd7+ l:.f4 71 'ii'b3+ l:.f3
61 l:.gS! 7 l ...'ith4 promotes at once.
The toughest defence. By comparison with 72 'ii'b8+ 'it>g2 73 1i'b2 'ithl 74 'ii'b.8+ 'itgl
the last note, Black wins in the line 6 1 b7 f4 62 75 'ii'd4 g3 76 'ite7 g2 77 'ite6 l:.h3 78 'itd5
b81i' l:.xb8 63 l:.xb8 f3 64 'itd5 f2 (D). 'ithl 79 'ii'g4 l:.d3+ 80 'ite4 g11i' 0-1
[· ·--· -· - �
58 JO] CHESS ENDGAME TIPS
.TIP- 49- · - Passed Pawns Mean Counterplay

As any textbook will tell you, the result of a
rook ending with the generic structure of three
pawns each on the kingside, and an extra passed
the rooks. If the stronger side's rook is behind

a-pawn, generally depends on the location of B
his passed pawn, he is usually winning (Alekh

ine-Capablanca, Buenos Aires Wch (34) 1 927
is the classic practical example quoted in most
textbooks). Conversely, with his rook in front
of the passed pawn, he is usually unable to win.
Since these basic principles are so well
known, and so readily available in endgame
textbooks, we shall deal in this book only with
some interesting exceptions. 47 hxg4?!
.••
White now obtains a potential passed pawn,

but even after the superior 47 ...�e5 48 gxh5
w
gxh5, White has a target on h5 to aim at with his
king, and the win is already seriously endan
gered.
48 fxg4+ �e5 49 h4!
Now the passed h-pawn gives enough coun
terplay to draw.
49 �d5 50 h5 gxh5 51 gxh5 (D)
•••
With his rook behind the pawn, Black should

win this position. However, it is easy to go
wrong in such positions, and Botvinnik shows
how to make the most of the defender's re
sources.
45 :t:a4!
Holding back the passed pawn for as long as 51. �e6 52 h6 �7 53 :t:g4!
possible. See Yusupov-Timman (Tip 44) for a
••
This and the next move secure the draw.

similar example. White makes maximum use of his passed pawn
45 �g5?! 46 f3 �f5? to secure counterplay.
53 � 54 lU4 %la6 55 l:[g4 %:.a7 56 l:[f4
.••
This allows a tactical trick. He should go •••
back with his king to h6 and come out via g7. 'iti>g8 57 l:[xf6 a4 58 .:.C2 �h7 59 l:[a2 'it>xh6 60
47 g4+! (D) �2 �g5 61 'it>e3 1/2-1/2
[T�P 5� I The Exception that Proves the Rule

the pawn queening. For the same reason, he
cannot play ... g5, when hxg5 would leave Black
no recapture. He therefore can only watch and
wait.
50 ltal 51 'iii>d4 l:ta5 52 'iii>c4 l:.a3 53 'iii>c5
•••
(D)
kt - - ------ --
B � . . .
�
� iW,1- - i �
iW,1- �
�
� � � . �
- ·�
·.
.
Unzicker - Lundin . . . �
Amsterdam OL 1954
� m m LS B
Here, with the white rook in front of the . . . �
pawn, the position would usually be drawn,
but this is an exception, because of the badly
�- • - ---- =�"'--
"' __J
placed black king. White has a simple, forcing 53 .:at

••
winning plan: 5 3 ...:a6 54 �b5 would achieve nothing. He

49 a7! cannot prevent the white king from penetrating
This ties the black rook to the a-file. Now the all the way to h6.
plan is to march the white king all the way to h6! 54 �d6 l:.a3 55 �e7 :a6 56 � l:ta3 57
49 .:az+
•• <l;g7 :at 58 'iii>b6 (D)
This allows the white king out, but if Black
marks time, the king just marches along the
B
rank to b2 and then out. Black can never take on
f3 because his rook cannot leave the a-file,
thanks to White's last move.
50 '1ti>d3 (D)
The white king has completed his long

march. Now the final step is to sacrifice the a7-
pawn and win Black's kingside pawns.
58 l:a6 59 l:b8 .U.xa7 60 l:tb5+ 'ii;>e6 61
•••
�xg6 .i::tas 62 'ii;>xh5 l:tg8 63 g4 l:th8+ 64 'it>g6

1-0
Black's other problem is that his king cannot White wins after 64. . l:xh4 65 l:tb6+ and 66
.
leave f5 because of a rook check followed by l:xf6.

r : -----, ,
�p S!_J
-
--
More Chances with a Knight's Pawn
45 .:r.c7 l:tb5 46 b7 �e6

46...�5 47 c;f;>c3 'iti>g4 48 l:tc4+ �xg3 49
l::tb4 wins for White.
47 <liic3 f6 48 <liic4 l::tbl 49 <t>cS
Threatening 50 l::tc6+ and 5 1 .:r.b6.
49...�5 50 l:td7! (D)
Hollis - Florian
corr. OL 1972-4
Again, White's rook is in front of the pawn,

and this position should be a draw. However,
with a b-pawn, White has more winning chances
than with an a-pawn. This is because in order to 50 J:tcl+
••
support his passed pawn, the white king has to 50 ... <liig4 5 1 l:ld4+ �xg3 52 l::tb4, winning,
travel one file less far over to the queenside. is the point of White's last move.
This in turn means that he will be one move 51 �d6 .:bl 52 <l;c7 l::tc l+ 53 ci;d8 l::tbl 54
faster in getting back to the kingside, after Black �c8 <t>g4 (D)
has been forced to sacrifice his rook for the b
pawn.
40 ..tf1 .:r.b2 41 �el �6 42 f3 .:r.b3 43 'iti>d2
.:r.xf3 44 c;f;>c2 (D)
55 l::td6 g5 56 lhf'6 gxb4 57 gxb4 'ihb4 58

:g6!
The final touch, preventing the black king
from unblocking his own h-pawn.
45 .:xn+ is already a threat. 58 <liih3 59 <t>c7! 1-0
59 ... h4 60 l:tb6 wins for White, while after
•••
44 J:tf5?
59....:Xb7+ 60 <t>xb7 h4 he loses by a tempo: 61
••
The losing move. Black can draw by means

of 44 . . .:e3 ! 45 :c1 .:es 46 b7 .:r.bs 47 �d3 'ifoic6 <liih2 62 ct>d5 h3 63 <liie4 <t>h I 64 <liif3 h2
65 .l:.a6, etc. If the white king had started on a7,
.
�5 ! 48 .:xn+ �g4 49 l:tf4+ �xg3 50 .:b4

g5 !. this line would draw for Black.
10/ CHESS ENDGAME TIPS 61
�IP 5 2 J Never Forget about Stalemate
Stalemate defences do not only occur in king In his day, Ossip Bernstein had tangled with
and pawn endings. Here are three examples of the likes of Lasker and Capablanca, so he had
grandmasters dropping half-points in rook end learnt a thing or two about both endings and
ings, through missing stalemate resources. tactical trickery. Smyslov played:
59 b2??
•••
His wily veteran opponent now lost no time

in snapping off the pawn:
60 lhb2! 'iti>g4
60 ... l:.h2+ 61 �f3 l:.xb2 stalemate.
61 �n 112-112
Shirov - Morozevich
Astana 2001
White blundered with...

55 'iii>b2?
55 �c2 is correct.
55 �b4?
•••
Missing his chance. Black could have drawn Gufeld - Bronstein

by 55 ... l:.xh5 ! , thanks to the trick 56 l:.a5+ Kislovodsk 1968
'ittb4 ! 57 l:r.xh5 stalemate.
56 l:.b6+ 'ito>c5 57 lhh6 90 �g7??
90 � wins, although it still requires quite a
This mutual oversight is all the more repre
and White won.
bit of work. One line runs 90. .'itg4 9 1 'ifte7 l:.c3
92 'ito>e6 � 93 'ito>d5 'ito>g4 (Black's king can't
.
hensible for the fact that the same trick had pre
viously occurred in a famous game: make it back in time) 94 �d4 l:tc8 95 c4 �5 96
l:.e2 (White is just in time to cut off the black
king) 96...l:.d8+ 97 �c5 (White can win this
position by using the Combined Method - see
Tip 39 for more details) 97 ...l:.c8+ 98 �b5
l:.b8+ 99 'iii>a6 lk8 1 00 l:.c2 'ite6 1 0 1 <Jiib7 l:k5
1 02 'iii>b6 and the c-pawn advances.
90 �g4 91 l:.h2
•••
9 1 l:.d4+ <Jiih5 !? (9 1 ...�f5 92 c4 l:.c7+ 93

'iii>f8 'ite5 94 l:tg4 'itf5 also draws) 92 c4 :Xc4 !
was the point White had missed. Now he loses a
pawn and can only draw.
91 �g3 92 l:.hl l:txc2 93 h5 l:.c7+ 94 �f6
•••
l:.c6+ 95 'ittr7 l:.c7+ 96 'ifte6 l:.c6+ 97 'ito>d5

0. Bernstein - Smyslov l:.h6 98 �e4 'iii>g2 99 l:.h4 'itl>g3 100 l:.bl 'ittg2
Groningen 1946 1f2.1f2
�·-·- --·]
TIP 53
- - -·- Triangulation
rook to stop the f-pawn is hopeless after 55 rz.g8

f4 56 :lg2 t'3 57 IZ.f2 �c5 58 �c3 :lf4, etc.
55 f4 56 'it>e2 �c5 57 :e4 (D)
•••
King triangulation is a standard technique in Temporarily holding everything, but the

king and pawn endings. However, its use is not black king can penetrate further.
confined to such endings. Here, we see it em 57 �b4 58 m c5 59 <M2 �c3 60 'ifi>e2
•.•
ployed to clinch matters in a rook ending. With 60 � 'it>d3 is zugzwang immediately.

a solid extra pawn on the kingside, Black has 60 n+ 61 �2 rus 62 :th4 bs 63 :te4 'it>d3
••.
every hope of winning. The immediate threat is 64 :lh4 (D)

5 1 ...�c5, so White's first move is forced.
51 �d4 �c6+ 52 'iiic3 rl.f7
The rook takes up its ideal position, behind
the passed pawn.
53 �d3 rs 54 gxrs gxrs <D)
- -- -
� - -·- ·
w m•m m m
� - -·
- LS - m m
� -�- m LS Further direct progress does not appear easy,
m m m m
since Black's rook is tied to defending h5 and
f3. However, with White to move, this position
m _ m m_m_ would already be zugzwang and so Black only
needs to transfer the move to White. He achieves
55 :le8 this by a king triangulation:
The white king will have to blockade the f 64. "'d2 65 :le4 <t>cJ! 66 :lh4 'it>d3 0-1
••
pawn, so the white rook comes back to defend White must lose his rook, or allow the f
the c4-pawn. The alternative plan of using the pawn forward.
New York 1924 Revisited
The classic game Capablanca-Tartakower, New By comparison with the previous note, here
the black king comes to f3 , and there are no
checks along the third rank. Thanks to the white
York 1924 showed the power of a passed pawn
on the 6th rank, supported by king and rook.
Capablanca sacrificed two pawns to reach such pawn on f5, nor are there any down the f-file, a
a position, and duly won. Here is a more mod situation similar to that seen in Kotov-Elis
em and even more spectacular example of the kases, Tip 38.
same idea. 46 fS r.t>e4 (D)
Pelletier - Rozentalis Thanks to the support of king and rook,

Erevan OL 1996 Black's one passed pawn proves stronger than
White's four!
41 g4!! 47 f6
Or 47 b4 �t'3 48 .l:.al :t"2+ 49 r.t>gl :g2+ 50
•••
The more obvious 4 l ...f4 does not work, be

cause of 42 gxf4 gxf4 43 .:r.xb4 e3 44 fxe3 fxe3 c.t>hl lhg3 5 1 :n+ r.t>e2 52 .:r.f4 .:r.t'3 winning.
45 .:r.a4 .:r.b2 (D). 47 d5+ � 48 :r4+ r.t>xg3 also does not help.
47-'it>o 48 l:al l:tg2! (D) 0-1
Black wins after both 49 f7 e2+ 50 r.t>el c.t>e3

Now 46 b4! opens the 3rd rank for the white
rook: 46...'iti>e4 47 .:r.a8 ! �d3 48 .:r.a3+ 'it>d2 49
:at and White holds. 5 1 f8'ii' .:r.g1+ and 49 .:r.e1 e2+ 50 .:r.xe2 .:r.xe2 5 1
42 .:r.xb4 f4 43 lta4 .:r.b2 44 gxf4 g3 45 fxg3
_
f7 :f2+.
e3 A beautiful ending by Rozentalis.
-
64 101 CHESS ENIX1AME TIPS
! � - ]
��5 5 _
..
Don't Get in a Huff
One important difference between chess and Now White has no choice but to make the ex
draughts (checkers) is that, in chess, capturing change himself on e6, since the laws of chess un
is not forced. This can have important conse fortunately do not permit him to 'huff' the black
quences. rook from the board, much as he would like to!
47 l:txe6+ �xe6 (D)
Adams - Lautier
TI/burg 1996 Black now has the opposition and draws.
The remainder of the game is a suitable adden
the two kings dance a small pas de deux.

White can win easily by 45 l:tb5, followed dum to the ending Timman-Yusupov (Tip 3), as
by c5, when Black's queenside pawn-structure
collapses. Instead, Adams blundered: 48 �d3 'it>d7!
45 h4?? l:te6+! 46 l:te5 Keeping the distant opposition. 48 ...�d6?
46 �4 l:tf6 is also drawn. 49 �d4 'ifr>c6 50 c5 bxc5+ 5 1 'it>c4 wins for
Adams had seen that the pawn ending after White, as does 48 ...�f5? 49 c5 ! bxc5 50 'it>c4.
the exchange on e5 is winning for White. Un 49 �cl 'it>c7 50 b4 axb4+ 51 �b4 �c6 52
�b3 'it>d6 53 'it>c2 <j;>c6 S4 �d3 'it>d6 55 �cl
Black is not forced to capture. Instead, he played
fortunately for him, this is not draughts, and
'it>cS 56 'it>b3 'it>d6 57 'it>b4 �c6 (D)
the move White had missed:
46 'iti>t6! (D)
•.•
58 c5 bxcS+ 59 'it>c4 �b6 60 'ifi>dS 'it>aS 61

'it>xcS �4 62 �c4 1/2-1/2
I TIP 56] The Active Rook
As noted in the Introduction, in most endings 35 J:C2 36 l:tb5

••
36 �g4! (D)
material counts, and it is rarely worth giving up Stopping 36 ...f4 and 37...f3.
pawns in order to activate pieces. However, in •••
rook endings, the activity of the rook is critical, Renewing the threat. Black's advantage in
and can make the difference between success or piece activity (both rook and king) makes up
failure. This frequently outweighs purely mate for his material deficit.
rial considerations.
37 h3+
Tarrasch - Rubinstein The only chance, but now Black secures a
San Sebastian 1911 passed h-pawn, which is a new source of coun
terplay.
This is one of the classic examples of the 37 �xh3 38 lhf'S l:txb2 39 :C4
•••
theme. Black is a pawn down, with another 39 a4 l:ta2 40 a5 �g4 41 l:.e5 �f3, threaten
pawn en prise, and seems completely lost. Pas ing ...l:te2+, is a draw.
sive defence by 33 ...l:tb8 would be hopeless, 39-..uul 40 lhe4 (D)
but instead Rubinstein drew by playing actively.
33 l:td2!
•••
It is worth another pawn to get the rook ac
34 lhb6+ �g5 35 �el (D)

tive.
Or 35 a4 f4 36 a5.. f3 37 �el l:te2+ and

draws.
40 hS!
•••
Using his trumps. Despite his extra pawn,

White cannot win, because Black's pieces are
all so active.
41 c4 �g2 42 l:.f4 l:.c2 43 l:.h4 � 44 �dl
l:.xf2 45 c5 )f,;>eJ 46 l:.xhS �d4 1/z-1/2
Another Active Rook
to defending g3, and he cannot free his rook

- - - - -
without losing the c-pawn. Once that pawn
goes, most of the resulting 2 vs 1 positions will
B -
- .
R .
R -
� be drawn, thanks to the fact that the pawns are
m:m m - all on one flank.
62 l:.b2 l:.c3+ 63 �g2 'it1'6 64 'ith3 l:.c6! (D)
� - -· Avoiding a sneaky trap. The obvious 64. . f4?
.1 m•m.rihR
m �-
.
loses to 65 l:tb3 ! l:.xc2 66 l:.f3, when White will

� R Rwd reach a theoretically winning ending, with two
-�- - �
connected passed pawns.
�. - - - - - -
- - -
-
Schlechter Em. Lasker w �

• • �
- 1. -
-
Vienna Wch (1) 1910

- �
� - -·-
- - - '{1
This is another example where Black saves
an apparently hopeless position by active play.
It seems that he has no hope, since he is already
· n m m O<;it
a LS m m �
one pawn down and White threatens 54 c4 fol
lowed by 55 �4. Lasker solved the problem by
giving up another pawn to tie White's pieces
down to defence:
� - -- -
54 Jle4! 55 :Z.c5
•• 65 ltb8
There is no other way to make progress, White has no other way to make progress, but
since after 55 c4 a4 Black creates counterplay the resulting position with two pawns against
with his a-pawn. one is a draw�
S5 'it>f6 56 l%xa5 l:.c4! 57 l:.a6+ �eS 58 6S Jlxc2 66 l:.b6+ �gT 67 h5 l:.c4 68 h6+
:as+ 'it>r6 59 l:.a6+ �es 60 :as+ 'it>r6 61 l:.al
••• ••
�h7 69 l:.r6 l:.a4 (D)

�eS (D)
lfi.lfi
Although White has two extra pawns, the White has no way to make progress, since 70
passive placing of his pieces prevents him l:.xf5 <t>xh6 is a dead draw. A perfect comple
from realizing the advantage. His king is tied ment to Tip 56.
[ Tl- - ��
�_J Yet Another Active Rook
38 .l:.d7
Miles understands that passive play will prob
We have just seen two examples where rook ac ••
tivity enabled the defender to save an appar

ently hopeless position. Here, it is the stronger ably lose, and he must therefore seek counter
side which is able to win, simply because he has play. If he just waits, White has options such as
the more active rook. h6, with a potential threat of l:.c8-h8, or �g5
and l:.c6.
39 lhbS l:.d3 40 l:r.b7+ �6 41 lilh7 gS+ 42
W l:.xa3 43 J:.a7 lla2 44 l:.a6+ �eS 45 g4!
(D)
Andersson - Miles
, Tilburg 1 981
White's only real advantage is the more ac

tive rook, but it proves enough to decide the Activating the white king.
game, without a discernible error from Black. 4S...fxg4+ 46 �g4 l:.g2+ 47 �h3 :h2
33 b4 l:.b7 34 h4 47 ...l:.a2 48 h6 wins for White.
Intending h5 (and maybe h6), followed by e4 48 lha4 e4 49 g3 (D)
or g4.
34 a4 3S �4 �e6 36 hS �6 37 l:.c6+ W
Note how White's more active rook has en
•••
.
� � �
�- . �
- -
� . . .
--
� �
abled him to drive the enemy king back, so that
his own king has also become more active than B •
- � �
-
��
its opposite number.
� �- •
• � -
- • LS
w � -- �
�··
38 l:.cS (D)
,: u
B
·�-- ·-•••
. �
•
� . -
� u
- . . �
n ��
. . ··� � -,{_ � �
� i ii B i B LS
.u� � �- �
49...l:.bl
� �
� � d nn
� He also loses after 49...�xe3 50 h6 l:.bl 5 1
�
- u :a3+ �e4 52 �g4, and the rook can interpose
. . . �
on g3 and h3 .
so h6 l:.g1+ Sl �2 l:.ht S2 l:.a6 �dS S3
. . ------- �g3 l:.h4 54 l:.g6 �e4 SS bS �S 56 l:.c6 l:.e4
S7 �2 l:.e7 SS b6 l:.b7 S9 �g3 1-0
5 -
c.---p - -·9]
�·
When Similar is Not the Same
Drawing on one's experience of analogous po �xc3 <Ji>xe4 4 �d2 �f4 5 �e2 �xg5 6 �f2
sitions from the past is highly valuable, but one �h4, etc.
must always take account of the specific details Here, however, the extra h-pawns mean that
of each position. Here, Botvinnik managed to the pawn ending is drawn after 43 ...l%c7 44
salvage an unlikely half-point by so doing. l%xc3 ! :Xc3+ 45 <Ji>xc3 �xe4 46 �c4 <it>f4 47
�d4 �g4 48 �e5 �xh4 49 �6! (D).
49 ...'itiig4 50 �xg6 h4 5 1 �f6 h3 52 g6 h2 53

g7 h I 'if 54 g8'ii'+ and draws.
Despite the material equality, White is strug Euwe is therefore forced to seek another
gling to draw, because Black has the outside path, but is unable to win.
passed pawn, supported by his rook. 44 �e3!
39 h4 h6 40 g5! h5 41 �e3 �e5 42 lk2! c3 Simpler than 44 �xc3 �xe4.
43 �d3 (D) 44 l%d4 45 :Xc3 :Xe4+ 46 ..t>f'3 l:txh4 47
•••
l%c6! l:.r4+ (D)

47 ...'iti>f5 4S l%c5+ �e6 49 l%c6+ also leads to
a draw.
- -- - �
w . - - -
m:m m••
•
� -
�
• m•
n
ui
- - - .
m m mwm
43 J:dS+ - - - .
• - --- _L__
•.
An almost identical position, but without the

h-pawns (and with reversed colours and wings)
:Xg6+! lfl.lfl
had been reached in Em .Lasker-Rubinstein, St 48 'iti>e3 l%e4+ 49 <M3 Wf5 50 J:tf6+ �xg5 51
Petersburg 1 9 1 4 (the only difference was the
'black' rook being on 'c5'). Lasker had played A valuable lesson in taking the specifics of
1 ...l:.c7, after which 2 l%xc3 loses: 2 ...l%xc3+ 3 every position into account.
Seizing Space
It would be better to activate the rook by 35

l::te4, planning 36 l::tb4.
35 f5!
•••
Continuing his space-gaining activity on the

kingside, and also stopping the l::te4-b4 defence.
36 h3
See the note to move 34. White prepares to
exchange off the h-pawn before it becomes
fixed.
36 h5 37 � g4+ 38 hxg4 hxg4+ (D)
•••
Anisimov - S. Ivanov
St Petersburg Ch 2002
Black has a solid extra pawn, but since it is

on the kingside, he does not have the standard
plan of creating an outside passed pawn. In
such positions, it is better to concentrate firstly
on improving one's position as much as possi
ble, before setting about creating a passed pawn.
Ivanov therefore takes the initiative on the
kingside. 39 �f4?
34 g5!
••• This loses immediately. A tougher defence
It is important to seize space on the kingside was 39 �e2 l:.c2+ 40 �el l::tb2 4 1 b4 b5 42 l:.c3
and threaten to fix the pawn on h2. We shall see i:ta2 (D).
ending in Flohr-Pirc, Tip 74. If White remains

a similar strategy employed in a minor-piece
passive, Black can play the pawn to g4, fixing

h2 as a potential weakness.
35 �g2 (D)
However, Black still has excellent winning

chances with the plan of .. .'ifi>f6, ...e5 and .. .f4.
Note how Black's kingside pawn advances have
created targets in White's kingside.
39 .l::tc2 40 f3 i:tf2 0-1
••
70 10I CHESS ENDGAME TIPS
[��&: 11
�� The Weakness of Weak Pawns
As emphasized in the Introduction, weak pawns 27 g5 28 f4 gxf4+ 29 gxf4 lha5 30 lha5

•••
tend to grow in significance in the endgame. :tdl!

Active play i s again the best defence. Against
30. . 'i.ti>b6, White plays 3 1 :es, followed by
.
w
<M3-g3-h4.
31 lha6 :hl 32 b4 (D)
Benko - Darga
Tel Aviv OL 1964
Black's split queenside pawns mean that he

will always have to suffer here, as Capablanca 32 Wd7?!
Benko recommends 32 :txh2, not fearing
•••
demonstrated in a number of similar double ...
rook endings. Benko follows the procedure typ 33 bS cxbS 34 :txe6, when Black can create
ical of such positions. counterplay by 34...l:th3+ 3S 'if.?f2 l:.a3.
Step 1 : tie Black down to his weaknesses. 33 '1t>d4 :tdl+?
17 :tc4 :td6 18 l:ta4 :as 19 :ct a6 20 b3 The last try was 33 ...l:txh2 34 �cS eS !, se
'if.Jb7 21 :tb4+ rj;c7 22 l:g4 curing e6 for the black king, when Black should
Step 2: provoke a second weakness on the be able to draw.
kingside. 34 'if.Je5 :td5+ 35 �6 l:td4 36 e3 lhb4 37
22 g6 23 :th4 h5 24 :14 f5 25 :cS :tb8 26
.•• :ta3 :te4 38 :td3+ 'li;c7 39 h4 (D)
:a5 :tb5 (D)
-- - - -- - �
w - •- -
�
� - � • �
·-·-·-··
'; I. - - · - ·
m m a •
A �- � ft
�
�Q· � u
�- -�D %
�- - - - White's vastly more active king decides mat
Looking for activity. ters.
27 :fa4! 39...cJtb6 40 �g5 �b5 41 llb3+! 'it>c5 42
Better than 27 :txa6 :es+ 28 <Jif3 gS, when 'it>xh5 :tc4 43 llb2 :tel 44 �g5 cJtd5 45 l:f.h2
the rook on a6 is temporarily out of play. 1-0
Evading the Checks
The ending of queen and pawn vs queen is one

of the most complicated in chess, and there is
no space to examine it in detail here. What we
can do, however, is to show an example of the
main techniques in such endings.
70 1id6+? 'iib4+, or 70 'iic l + b2.

70 <ifiia2 71 'i!i'f7 'i!i'el+?
•••
Black throws away the win by allowing the

white king to reach the 6th rank. 71 ...<itiia3 is
correct, when White has no checks thanks to
Donchenko - Miles the crosscheck on b4.
Dubna 1976 72 <ifiih3?
See move 64. Both 72 'iii>g5 and 72 'iii>h5
Black's basic plan is gradually to run White draw.
out of checks and then advance the pawn. Cross 72 'ili'e4 73 1i'a7+ 'iii>bl 74 1i'gt+ '1i>c2 75
•••
checks (answering a check with a check) are a 'ifh2+ <ifiic3 76 'i'c7+ 'iii>d.2 77 'i'aS+ '1i>e2 (D)
key part of the technique in such positions. The
five-piece computer tablebases show that the
w
ending is drawn with correct play, however.
56 'ifn+ <ifiic2 57 'iVf2+ 'it>c3 58 'iff6+ 'ii'd4
Centralizing the queen maximizes its influ
ence, and is another typical device in such end
ings.
59 1i'c6+ <itiib2 60 'ifg2+ <ifiicl 61 1i'c6+ 1i'c3
62 'ifht+ �bl (D)
63 1i'e4
63 'ifh2+ <ifiib l 64 'ii'h l +? 'ii'c l + (the cross
check).
63 b3 64 <ifiih4?
•••
In such positions, the defender's king should 78 1Wal

usually head for the furthest corner from the 78 1Wb5+ 1i'd3+.
pawn, so as to reduce the scope for crosschecks. 78 1i'd3+ 79 'iii>h4 'ii'c4+ 80 'iii>h3
Hence 64 <ifiih6 is correct.
•••
80 '1i>h5 1i'b5+ followed by 8 1 ...b2.

64 'i!i'cS 65 'ii'g2+ 'ita3 66 'iVa8+ 'iitb4 67 80 'itd.2 81 'i'a7
'iVb7+ <ifiia4 68 ika8+ 'ii'aS 69 'ifc6+ 'ita3 70
••• •••
8 1 'ii'b2+ 'ii'c2 82 'ila3 b2 and Black wins.

1i'f3 81...bl 0-1
72 1 01 CHESS ENDGAME TIPS
King Safety is the Key
counterattack against the white king and force

perpetual check, so White has to seek other
ways to strengthen his position.
47 g4 gS 48 'itif2 'ii'd6 49 � 'ilt'a6+ SO �g2
�g7 (D)
Alekhine - Reshevsky
Amsterdam (AVRO) 1938
In this typical structure, with an extra outside

passed pawn, the result usually depends on the
safety of the white king. If his king is secure White's only plan is to bring his king to the
from checks, the win is generally straightfor queenside to support the pawn, but first he cen
ward, since the queen can itself force the passed tralizes his queen so as to maximize its influ
pawn home. However, if the white king is ex ence.
posed, as here, the task is greatly complicated, 51 1i'b2+ �g8 S2 l!ib8+ �g7 S3 'ii'eS+ �g8
and indeed there may be no win at all. Alekhine 54 rJi>f2 'ilt'a7+ SS 'it>e2 'ii'a6+ S6 �d2 (D)
was unable to win this position.
44 1i'a2 �g8 4S a4 'ii'c6 46 aS 'ii'a6 (D)
S6 'i¥c4!
•••
This excellent move secures the draw. Black

This is where we see the importance of prevents the enemy king from coming across to
White's exposed king. If it were adequately the queenside, and White can make no further
protected, White could win simply by 47 'ii'a4 progress.
57 ii'fS ii'd4+ 58 ct>e2 'ilt'b2+ 59 'it>d3 'ii'h3+
60 �e2 'irb2+ •/z-1/2
followed by 'ifb4-b6, queening his pawn. How
ever, in this position that would allow Black to
I T•P 64 King Safety is the Key (Part 2)
��re, ;,e s� another example of the theme of 45 a4 'ii'h l+ 46 'iii>e2 (D)

g s ety m queen and pawn endings.
. . -·�
-0 .0 - --� �, , • • 'm� ,
w �/. . ·· -� - . . �
B
- �
• rui1
· -�% . � . . - -•
.
� ·/. • ,. 8 A . 'p
• • - 'iV �
- 0 .0 ./. - •
� 8 ··· 8 ·%
,£., '• i• ., •<iiiB •
o &, a B�� ,
� . ..
. .
�--- --'l • - -
One� again, the doubled f-pawns do an ex-

Botvinnik - Donner cellent JOb of protecting the white k"mg.
Noordwijk 1965 46 _hS 47 'i!ldS+ 'ith7 48 'it'dl!
•••
A mce idea. White drives the black queen

.
40 1i'd7+ 'itg8 41 'ii'c8+ 'iii>h7 42 1111
wxa6 1 mto passivity.
.
White has a potent.Jal outside passed pawn, 48 'iihl? ! 49 b4 h4? (D)
. . .
•••
. .
�
but it is his better-protect d k"mg which ts the
key factor. Note that dou led pawns are often
as is the case here.
·/. . --- .
w •0 •, • 'm•
-
strong de1ens1vely,
& •
42 'iig5+ 43 <Ml 'ii'xh5 44 \Wrd3R +! (D)

,
-0 . . •
•••
.•0 m, • • �� . . .
/.
•, • • -· 80 • . -
- %%�
-� . .
B
· • oA ,•
.
�-
· . �� -- - .
�0 .·/. • , .... • •�B 0: ,, ,
- . - .
. � miV_m ,m
�0� .,,. 8,.
8 m, a B • -
Missing the threat.
50 'ii'fi !
• -� - � ��:� :°��
ca h o ,
0 · The black queen is
annot ge� out, either to
harass the white king, stop White' s potential
vent Black from setting uP a perpetual. In the
The f-pawns shelter the white king and pre
passed a-pawn' or to get out of the way of its
meantime, White can get on with his plan of own h-pawn.
creatin�a passed pawn on the queenside. 50 g5•••
44 wg8
Not 44...'ii'g6 45 We4!, when Black would
••• 50...h3 5 1 a5 bxa5 52 bxa5 'ii'g2 53 a6 leaves
Black powerless.
soon be forced into a lost pawn ending. 51 aS bxaS 52 bxa5 g4 53 a6 l-O
r:-- - �
LTIP 6�_1 Exchanges to Real ize a Material Advantage
The normal rule when material up is to exchange policy, and that is what White prepares to do
pieces, not pawns. This is one illustration. here.
30 fS 31 :Cal lDe4
••.
White wins after 3 1 ...l:txd5 32 l:.c7, threat

ening l:ta8+.
32 :as :Xa8 33 lhaS+ <iPf7 (D)
Anton - Lecroq
corr. Wch 1977-81
White has an extra exchange but Black is ac

tive. In a previous game Knaak-Postier, E.Ger Without his rook, Black's activity is much
man Ch (Frankfurt Oder) 1977, which had less threatening and it is already clear that only
reached this very position, White tried to hang White has winning chances here.
on to the d5-pawn with 30 ltJb4?, but Black de 34 ltJaJ! lbxc3 35 d6 �e6 36 ltJxc4 lbe4 37
veloped such piece activity that he even went l:te8+! 'iitd t 7
on to win after 30...ltJe4 3 1 'iiir>h 2 lDxc3 32 l:.a7 If 37 . . .'iittd5, then 38 d7 'iiir>xc4 39 d8'ii' .ixd8
f5 ! 33 h4 .if6! 34 ltJc6 l:td6 35 l:tel 'it>f8 36 40 :xd8 wins, because the black king is cut off.
:xh7?! lbxd5, etc. In the present game, White 38 l:th8 ltJr6 39 lbes+ <iie6 (D)
demonstrated the correct plan.
30 l:ta7! (D)
40 d7! .ie3+
40... ltJxd7 4 1 l:te8+ .ie7 42 lbc6 is winning
Exchanges enhance White's material advan for White.
tage, especially exchanging Black's last rook. 41 'it>f1 .tb6 42 ltJxg6 hxg6 43 d8\i' .txd8
When you are an exchange up, exchanging off 44 l:Xd8
the opponent's remaining rook is often the best and White won.
[11P 6-6-J Rooks are Better than Knights - Usually!
Here Black faces the crucial choice. The

text-move leads to a mating-net, but if he runs
with his king, he loses to an exchange sacri-
fice, as Seirawan's analysis shows: 94...�e6 95
.ih2+ �d5 96 c:j;e7 and now either % . .�bl 97
.
cJi;f7 .if5 (97 ....ic2 98 .l:.e6 and .l:txg6) 98

.l:.e5+ �d4 99 .l:.xf5 gxf5 100 �g6 winning, or
96...tfJg7 97 � tfJb5 98 .l:.e5+, followed by
.:.Xf5 and c:j;g6.
95 .ib2 .ie6 96 �c3 (D)
Kasparov - Karpov
New York/Lyons Wch ( 16) 1 990
Endings with an extra exchange, but all the

pawns on one side, are often difficult to win, all
the more so as the number of pawns gets smaller.
Many observers thought this one was drawn, but
during the second adjournment, Kasparov ana
lysed a forced win. Ultimately, Black loses be
cause he faces a fatal dilemma with his king. If it
stays in the comer, it is vulnerable to mating Zugzwang. He must allow the white rook a
threats, while if it runs away, White can set up a decisive penetration.
winning exchange sacrifice on f5. Black is also 96 .trs
••.
handicapped by his knight's bad position on the 96...tDf4? loses on the spot to 97 .l:.f3.
edge of the board. 97 .l:.e7+ � 98 �es .id3 (D)
89 :a1 .ig4 90 �d6 .th3 91 .l:.aJ .ig4 92
.:te3 .trs 93 cJ;;c7 <M.1 94 �d8 (D)
- --
- - - � �
B .---•••
•••
• �.i.n•
� • m U•
- - - .
� • a •
- - - .
�· - - - 99 .:ta7 .ie4 100 l::tc7 .ibl 101 .id6+ �g8
102 �e7 1-0
94 .tg4
••• .l:.c8+ and �e5 leads to mate.
When Two Pieces are Better than One
38 i.d3 W 39 axbS axbS 40 '>ti>g2 i.d6 41

f3 hS 42 i.fl h4 (D)
Szily - Keres
Budapest 1952
Now both the f3- and h3-pawns are fixed as
Rook + pawn vs two pieces is usually about weaknesses.
equal, but much depends on the specifics, and 43 .:.at 'ite6 44 l:.a6 '>ti>d7 45 :as i.c7 46
who has the initiative. Here White should play llal �c8 47 <M2 1'.b6+ 48 �g2 '>ti>b7 49 :tdl
35 f4, getting his pawns moving, when he would i.c7
have reasonable chances of holding the game. The minor pieces continue to render the en
Instead, he plays passively and allows Black to emy rook impotent.
demonstrate the superiority of the two pieces. so '>fi>f2 �b6 51 �e3 '>ti>aS 52 '>ti>d4 �b4
35 a4? gS! Now we see yet another downside to White's
Now Black's knight occupies a secure cen 35th move - it weakened his b3-pawn.
tral outpost, from which it cannot be dislodged. 53 b2 (D)
In addition, White's e-pawn is fixed on e4, ob
structing his bishop.
36 :t.dl 1'.e7 37 :%.cl 3'.c6 (D)
53 tllg6!
•••
Do not hurry! The b3-pawn is not going any

where. The text-move threatens 54...1'.b6+ and
Note how the black minor pieces prevent the 55 ...tt:\f4+.
white rook from penetrating. In such endings, 54 i.n i.b6+ ss <t>d3 tl\es+ 56 'Otc2
the rook needs to develop activity, and here it is 56 'ite2 'it>xb3 is equally hopeless.
unable to do so. 56 ttJxf3 57 l:td6 1'.c7 0-1
•••
[T1; 68. .. .
The Value of Knowing Your Theory
that he should meet 6 1 f7 not with 6 l . . ..ixf7?

The only other thing Black needs to know is
62 �f6 winning, but with 6 I ...�g7 ! preventing

the white king from getting to f6.
61 .td5 62 'iti>f'S <;f;>f7 63 'lites .ib3 64 l:.c7+
<;f;>fS 65 l:.b7 1'.c4 66 l:.b4 .ia2 67 �5 (D)
•••
Szabo - Botvinnik
Budapest 1952
A knowledge of basic endgame theory can

often enable a player to save a seemingly lost
position. Here, Black had been defending for
the entire game, and appears hopelessly placed. Szabo continues to manoeuvre, but if Black
But Botvinnik now produced an exchange sac knows the basic position, it is easy to draw.
rifice: 67....idS 68 �g6 .if7+ 69 �g5 .id5 70
51. .AxaS! 52 lbd7+ .ixd7 53 :xa5 .ixg4
• l:.b4 .ib3 71 l:.b8+ <l;f7 72 l:.b7+ <;f;>fS 73 f7
He knew this position to be a theoretical (D)
draw, even without the h-pawn.
54 'ifiie3 1'.e6 55 <it;>f4 .tc4 56 l:.a7 h5 57 'iti>g5
h4 58 'il;ixb4 1'.b3 59 'iti>g5 1'.c4 (D)
Having done his best impression of the Grand

Old Duke of York, Szabo makes his last try.
73...'iii>e7!
This exact position was analysed as a draw See the note to move 6 1 .
by del Rio as far back as 1 750 ! White cannot 74 'iii>g6 .ic4
make progress, because �g6 is always met But not 74... .tc2+?? 75 �g7 winning.
with a bishop check. 75 l:.g7· 1'.b3 76 f8'ii'++ 'iti>xf8 77 'itf6 'ifiie8
60 l:.c7 .ta2 61 l:.cl 78 l:.e7+ �d8 1'2-l/2
[r1P 691 Drawing with Rook vs Rook and Bishop
�e2 99 .l:.h3 'Wtd2 100 �eS �c2 101 !/4.c5 �d2

102 �e4 .l:.e8+ 103 �d4 l:td8+ 104 �c4 .l:.c8
105 .l:.h2+ ..t>ct (D)
Short - Sandipan
Gibraltar 2006
This ending has been known for many years The key defensive formation, known as the
to be a theoretical draw, but in practice the de Cochrane Defence. The rook pins the bishop
fender frequently loses. Ulf Andersson won it from behind. Black just waits, and when the
against GMs twice in the space of a year, while white king goes to one side, the black king goes
English GM Keith Arkell has apparently won to the other.
the position 1 6 times out of 1 6 ! It is therefore 106 �d4 �bl! 107 1'.b4 .l:.c2 108 1'.d2 �a2
worth being familiar with the drawing tech 109 q;,c13 �b3 110 .l:.hl .l:.c5 1 1 1 .l:.bt+ �a4
nique, which is not really all that complicated. 112 1'.b4 .l:.h5 1 13 .i.c3 l:r.b5 1 14 .l:.at+ �b3
87 �e7 �5 88 �f7 .l:.gl 89 .l:.aS+ ..t>e4 90 1 15 .l:.cl <it>a4 (D)
1'.e7 .l:.bl 91 �e6 .l:.b8 92 .l:.a4+ (D)
�
- %
� •
� -
z
B �•••<ttm •
• �
� . . .
: • • •• �
� . . .
. . . �
. . - . 116 .:Sl+
The defender cannot avoid his king being 1 16 �c4 is met by 1 16 ... l:.b4+ ! , so White
driven to the edge of the board, but this not fa cannot strengthen his position.
tal. 116 q;,b3 117 .l:.gt �a4 1 18 <it>c2 .l:.b7 1 19
92 �e3 93 !/4.c5+ �d3 94 �d5 .l:.d8+ 95
•••
l:.g5 .l:.a7 120 l:.h5 .l:.a6 121 .l:.c5 .l:.a7 122 ..t>d3
1'.d6 �d2 96 'WteS 'Wtd3 97 �d5 �d2 98 .l:.a3 l:.b7 123 :as+ �b3 124 .l:.al .l:.b5 1/2-1'2
•••
[ T1P_70 When Two Rooks Beat a Queen
Two rooks roughly balance a queen, but with 46 ..:.Sa6 47 'ili'c3 l:r.a4
••
equal pawns, the rooks are usually stronger. Black wants to get a rook to f4 or f6.
This is especially true if there are weak pawns 48 'ii'c6 :t8a6 49 'ii'e8+ �g7 so 'ii'b5
to attack, since the rooks can double up and at 50 'ii'e5+ l:r.f6 helps Black.
tack them more times than the queen can de 50 .l:r.4a5 51 'ikb4 l:r.dS (D)
••
fend. This is a typical example.
. . . �
� . • . ·· -
.• . . �
w
� . - .. . - �
mtt •
- . . -��
-
- . � -��.
. . . �
�- - -- - -- -- -
52 'ii'b3
Leko - Kramnik Or 52 f4 :td3+ 53 �2 l:r.a2+ 54 �el :th3,
Brissago Wch (I) 2004 etc.
52 ..:.ad6 53 'ikc4 l:r.d3 54 'Otr>f2 l:r.a3 55 'ikc5
••
44 gS! 45 'ii'f6 h6!

••• :a2+ 56 �g3 l:r.f6!
White is lost. Black just needs to double Step 1 complete. Now he just needs to get the
rooks against the f-pawn, after which he can other rook attacking f3.
liquidate to a winning pawn ending. White fights 57 'ikb4
to prevent this, but is ultimately unable to do so. Stopping .. .:a4-f4.
46 f3 (D) 57...llaa6 58 ci>g2 l:r.f4 59 'ikb2+ l:r.af6 (D)
On 46 'ikxh6??, 46...:t8a6 wins the queen,
while 46 f4 loses to 46...l:.a3+ 47 �2 l:a2+ 48
..w1 (48 �f3 .:.sa3+ 49 �e4 l:r.a4+) 48 . . ..:.cs.
Mating threats are also typical of endings with w
queen vs two rooks.
Mission accomplished. Now f3 falls and

Black forces a winning pawn ending.
60 'ii'e5 lhf3 61 'ital .:.n 62 'ii'c3 l:r.lf2+ 63
�g3 l:r.2f3+ 64 'ifxf3 .:.xn+ 65 � �6 0-1
Transforming an Advantage
It is often the case that the most effective way to typical, but require good judgement and confi
tended benefits fail to materialize!

realize an advantage is to transfonn it into an dence - one can look rather silly if the in
advantage of a different kind. Many weaker
players tend to cling on to a certain advantage 39 'it>d6 40 dxc5+ -.t>xc5 41 a4!
•••
for dear life, whereas giving it up for some Taking b5 under control and preparing to
other benefit is often the best strategy. drive the black king back with b4+.
41. -.t>d6 42 b4 a6 43 �d4 (D)
••
Gligoric - Trifunovit
Budapest 1 948 One can clearly see the effect of White's
39th move. Now a check on c4 will force the
We saw the earlier part of this ending in Tip black king to give way.
3 1 . Thus far, the weak pawn on c5 has been the 43 h5 44 tt:\c4+ 'it>e6 45 'ittc5 i..e8 46 a5
•••
cornerstone of White's advantage, yet he now i..a4 (D)

played:
39 d4! (DJ
47 lbeJ!
The final step. The knight comes to d4, at
Exchanging off Black's weakness. The idea tacking f5 and supporting a breakthrough by b5.
of this surprising move is to open a path for the 47 i..b5 48 l2Jc2 'it>e7 49 tt:\d4 i..d7 50 b5
•••
white king to reach d4, and ultimately, c5 or axb5 51 a6 i..c8 52 a7 i..b7 53 tt:\xb5 �d7 54
e5. Such transfonnations of an advantage are tbd4 i..e4 55 tt:\xf5! 1-0
81
r-------J
101 CHt..'SS ENDGAME TIPS
TIP
--
72
-----
Centralizing the King
w w
Botvinnik - Flohr 21 b4!

Zvenigorod (training) 1 952 Preparing to activate the king. It is obvious at
a glance how far from the action the black king
10 .te3! is, bearing out the note to move 13.
Only in the endgame is White's advantage in 21 f6 22 'lfi>b3 'lfi>f7 23 b5!
•..
pawn-structure significant. With queens on the We have already seen this idea in Tip 7 1 .
board, Black has much more chance to work up Although White exchanges off Black's weak
active counterplay, but once the position sim pawn, it is more important that he opens lines
plifies to an ending, White will have a freer for his own king.
hand to attack Black's weaknesses. The text 23 .tdS 24 bxc6 :txc6 25 .:.Xd8 l:xc5 26
•••
move also stops ....ic5. l:rb8 'lfi>g6

10 l!i'xdl+?! 11 �dl!
••• After 26 ...l:c7 27 lbb5 lbd7 28 l:a8 l:b7 29
The correct recapture. The king is going to 'it>b4 White wins a pawn.
c2. 27 l:b7 tbc8 28 a4! h5 29 f3 l:c6 30 l:d7 a6
11 ...eS 12 'iiic2 .te7 13 .ie2 0-0? (D) 31 tbdS aS (D)
13 ...f6 is correct. The king belongs in the
centre on f7.
32 h4!
Typical prophylaxis. Getting his pawns off
14 l:hdl .ie6 15 .ia6! l:ab8 16 b3 l:fd8 17 the 2nd rank reduces Black's counterplay.
.:.Xd8+ .ixd8 32 :&6 33 'iitc4 l:tc6+ 34 'lfi>b5 l:c2 35 lbe3
••.
1 7 ... l:xd8 1 8 .ib7 is awkward. l:tb2+ 36 'otxaS 'it>h7 37 lbfS 1-0

18 .tc5 .ic7 19 l:dl .tc8 20 .txc8 .:.Xc8 (D) 38 l:k7 will be decisive.
[--TIP- 73
-- ----]
- ------
Transforming an Advantage
A remarkable transformation manoeuvre. It

may have seemed that White gave up his
bishop-pair to weaken Black's pawns, but now
w he promptly exchanges off the freshly-created
weakness on d5. He does so because this is the
only way to get his king in. As noted in Tip 7 1
above, confidence i s important i n such situa
tions. The player must trust his judgement, and
not worry about how he will look if the hoped
for benefits do not materialize.
28 dxe4+ 29 �e4 'Otd7 30 �d5 h5? (D)
...
•••
30 i.d8! 31 ll:le5+ ll:lxe5 32 Jlxe5 would

still leave White very few winning chances.
Botvinnik - Kholmov
Moscow Cht 1969
White has only a very small advantage - the

two bishops and a little more space.
17 :Xd8+!
The plan is to exchange all the rooks and
then to activate his king.
17 ll:lxd8 18 J:tct ll:ld7 19 .Jlb5!
Stopping Black's intended ...Jlf6. A typical
•••
defensive idea when the opponent has the

bishop-pair is to exchange one's own bishop
for its opposite number. White takes steps to
prevent this. 31 .i.g'T! .Jld8 32 Jlf8 .Jlb6?! 33 i.cS lt'Je7+
19 J:txcl+ 20 .Jlxcl ll:lb6 21 'Wti>f1 a6 22
••• 34 �c4 .hc5?! 35 �x� 'ii>c7 (D)
.Jld3 ll:lc6 23 Jle4 g6 24 'ii>e2 �f8 25 �d3 �e8
26 .Jlb2 ll:ld5 (D)
-- -·- �
w · -
�·. .,
_,.
• •�• • m • m
. -�- -
n -
u •i.•- m
� •�DlDm
R A r:,;;
u o ;;;
_
m •
M •
a_m m a The knight ending gives White excellent win
ning chances, thanks to his extra space and more
27 Jlxd5! exd5 28 e4! active king, ;md he duly went on to win.
�74 ] The Perils of Passivity
38 g4!
Now the h7-pawn is a target.
38 tbr6 39 ll'icS tl'ie8 40 g5! lbc7 41 i..d3
••.
ll'ie8 42 tl'ib3 i..d7 43 llld.4 lbc7 44 i..e2 i..c8

45 h4 ttJe8 46 h5 i..d7 47 h6 (D)
Flohr - Pirc
Bad Podebrady 1 93 1
I n endings, passive play i s frequently fatal.

Here, Black stands worse because of his iso
lated d-pawn and inferior bishop, but with ac Since move 37, Black has just sat on his
curate play he should still be able to hold. hands, while White has made huge space gains
Instead, he plays too passively and allows Flohr and fixed another weakness at h7. As a result,
to strengthen his position unopposed. his advantage has now reached decisive propor
31 i..d7? tions.
47 <:J;;e7 48 i..f3 i..e6 49 tl'ib3 '1i>d6 50 ltJcS
••.
Black should try .to place his pawns on dark •••
squares, so 3 l ...a5 ! is correct. i..cS 51 e4 (D)

32 a5!
Now b7 .is weak.
32 tl'ie6 33 b4 tl'ic7 34 i..d3 i..e8 35 f4 i..d7
36 i..e2 tl'ie8 37 tl'ib3 (D)
•••
Another example of the now-familiar device

of exchanging off Black's weak pawn in return
for opening lines. The white bishop can now at
tack Black's weaknesses.
37 i..c6?
••. 51. dxe4 52 i..xe4 tl'ic7
••
Again too passive. In such positions, the de 52...�c7 53 �d4 and the white king marches
fender must fight for space and not allow the in on e5.
'clamping' of his kingside. 37 ...h5 was essen 53 lbxb7+ i..xb7 54 i..xb7 ttJe6 55 i..xa6
tial. lbxf4 56 i..c4 1-0
TIP 75
[----]
-- - --
A Lesson in Not Hurrying
as Black is well placed to meet it by 58 ...cxbS

59 axb5 l:.c7, with counterplay.
5s .:c7 (D)
••
w Black prepares counterplay with . . b6 and

.
...cs.
Flohr - Bondarevsky
Leningrad/Moscow (training) 1 939
White stands better - his minor piece is su

perior, the pawn on h7, despite being a passed
pawn, is just a weakness and White has a poten 59 l:.h6+!
tial minority attack on the queenside. Never Do not hurry; seize every little advantage!
theless, to exploit such an advantage requires White's plan is :.h l -c l , but first flicking in this
outstanding technique, and Flohr gives a true check slightly worsens the position of the black
masterclass. The most important thing to notice pieces.
is how slowly and patiently he conducts the 59 .i.g6 60 .rl.bl .i.fS?!
Passive. Shereshevsky recommends 60. . b6
•••
ending. There is no hurry, since Black has no .
active countcrplay, so White can take his time 61 l:E.c l .i.e8.

and seize every little advantage, no matter how 61 l:t.cl a6
insignificant it may appear. Fon:ed, since 62 a6 was a threat, but now the
54 b4 <it>gS 55 l:thl �6 56 l:th2 :gs 57 a4 b7-pawn is fixed as a weakness and White has
:.g7 (D) the plan of transferring his knight to c5.
.
- - - -- -.
62 llhl l:E.g7 (D)
,v, · - - - · - -- - - �
�·. · •. - -·
w
•
-•- • -- .
••-• - � .
w
� - · �-*-- •-
�u � u � .• . ,v, - · �-*--
u
� un • . •
� m D�m
. . � , • m D�m
8 8 m �
- - �
- -- - - - . . �
S8 a5!
• m . m m:
Aiming to provoke ... a6, leading to a fixed 63 :.b2!
weakness at b7. By contrast, the standard mi See move 59. The plan is l:E.h4, but first he
nority-attack break 58 b5?! is ineffective here, puts Black in a kind of zugzwang and again
forces a slight deterioration in his piece posi There is no hurry for lbct3-c5, so Flohr first
tioning. improves his king's position.
63 .ltc7
•• 69 .14g7 70 lDd3 c;.t>d6 71 lt:JcS J:r.e7 72 :rs
••
Here, the rook is less active than on the open Wc7 73 e5 J:r.e8 74 lhe8 Jlxe8 75 'ifw>r4 b6 (D)
g-file. It may not matter - indeed, as the game
goes, it probably does not - but it is a small ex
tra plus for White at no cost, so why not have it
anyway?
64 l:.h6+! (D)
76 lt:Ja4!
Suppressing counterplay. This keeps the
black king out and the rest is easy.
76...bxaS 77 bxaS .1Lf7 78 lt:Jc5 1'.c4 79 'itg5
.1Le2 80 'it>h6 'itd8 81 'itxh7 <l;e7 (D) ·
The same idea as in the previous note.

64 �g6 65 l::.b4
••
Flohr reasons that the bishop is slightly worse

on g6 than f5.
65 1'.fS
•••
Bondarevsky seems to agree, and can find

nothing better than to return it to that square.
White's 64th move has thus won a tempo.
66 e4 dxe4 67 fxe4 1'.g6 68 l:.f4+ c.ti>e6 (D)
82 �g6
The final stage is to transfer the white king to
the queenside.
82 1'.n 83 'ifw>r5 .1Lh3+ 84 c;.t>e4 1'.g2+ 85
�e3 1'.n 86 'it>d2 �7 87 rlo>c3 'it>g6 88 'it>b4
•••
'i1tf5 89 tt:Jb7 1-0

The king comes in and wins the queenside
pawns. A marvellous endgame tour deforce by
Flohr, and a model example of patient, immac
69 'iite3! ulate technique.
[.=--..Jt:l
�!J Positional Alchemy
39 g4!
Another typical idea - fixing pawn weak
nesses.
39 'i;e7 40 .:tg2
.••
Threatening %tg3-h3, and so tying down the

black king.
40 'iPfs 41 f4 l:.d8 42 ltgJ i.d7 (D)
•••
This move allows the white king in, but if

42 ...'ite7, then 43 f5 i.c8 44 l:.h3 l:.h8 45 �d4
'itd6 46 l:r.h5 and the threat of h4 and g5 wins
material.
Gheorghiu - Mednis
Riga IZ 1979
Here we see another typical example of trans

fonning an advantage. The main thing one no
tices about White's position is his two bishops,
but...
36 i.e7! l:.d7 37 i.xf6 gxf6
White has given up the bishop-pair, but bas
broken up Black's kingside pawns. Even more
. importantly, he has removed the only black
43 'iiid
i> 4!
dark squares c5 and b6. Now Black must con
piece which is capable of defending the weak
Not fearing the discovered check, the white
stantly worry about the white king penetrating king penetrates through the weakened dark
side pawns, these two weaknesses prove too

on those squares. Together with his weak king squares.
43 fS
43 ....i.xg4+ 44 �c5 is similar.
••.
much.
38 'ite3 'iPf8 (D) 44 'i;cS fxg4 45 'i;b6
The king has penetrated decisively on the
- - �, � weak dark squares, a direct result of the ex
w
change on move 37.
� j - :l _ j _ 45 i.e8
•••
·· . ·.
· ...
·� . � Or 45 ...i.c8 46 'itc7 l:te8 47 f5 h5 48 h3 win-
ning.
-� - - -
• �
46 'itxb7
u�·
n . - Now it is all over.
• m.tm � m 46 bS 47 'itxa6 l:td4 48 'it>b6 h4 49 l:txg4!

Jhd3 50 f5 l:tb3 51 c5 .i.d7 52 a6 .i.xf5 53
•••
- g -�� l:txh4 'it>g7 54 a7 l:r.a3 55 'it>b7 i.c8+ 56 �b8

� - -- - 1-0
l 0 l CHESS ENDGAME TIPS 87
I TIP 77 �] The I mportance of Calculation
Calculation of short, concrete variations is inte lbd6. This simple line is not difficult to calcu
gral to good technique. late (Dvoretsky records that it only took him
about three minutes). Given that the game con
tinuation is obviously good for Black (his king
is freed), no further thought is necessary to de
cide on 54...l:tb5.
55 '0ti>b7
•••
Liberating his king has significantly eased

Black's technical problems, and he went on to
win as follows:
56 � l:tb2 57 'Oti>e3 l:tb3+ 58 �2 llb2 59
�e3 l:tb3+ 60 �f2 b4 61 e6 �c7 (D)
Kotkov - Dvoretsky
Moscow Ch 1972
54...l:a3+ is obvious, but does not solve the

problem of the black king being cut off on the
a-file. Clearly. Black would prefer first to liber
ate his king with 54....U.b5. The only question is
whether White can exchange on b5 and then
win the d5-pawn. Answering this question re
quires a little concrete calculation.
54 l:tb5! 55 :at+ 62 l:ta8 h3 63 l:tg8 ftJe7 64 l:tb8 �d6 65
lbgl c3 66 � c2+ 67 ltd2 (D)
•••
It turns out that White cannot exchange

rooks: 55 l:txb5 'iii>xb5 56 lbc3+ 'iii>b4 57 lDxd5+
�b3 58 e6 (58 'iitd2 lbxd4 wins for Black)
58 ...c3 59 lbxc3 �xc3 60 d5 (D).
----- - . -
-- , ;
B � - - -
- -�-·�
• m•m•o
� u
-
�
� n u � �
�-- - - ·
-·= - �
� - -- , y,
67 J:r.g3!
•••
Not 67 ... l:tbl? 68 lbe2 h2 69 lbc l !.

0-1
The white pawns appear dangerous, but A simple but neat example of how a little cal
Black wins by 60...lba5 ! 6 1 e7 lbc4+ 62 'ifiif2 culation can simplify a technical task.
-- --8[l
----
Concrete Play
40 iLcS!
The same policy. Rather than exchanging
the rooks, which of course would win easily
back to f5 and so maintain the bind. Do not al

w enough, Kaev uses tactics to force the knight
low unnecessary counterplay!

40 lbfs 41 iLb6 a4 42 b3 (D)
•••
Kaev - Rovner
Kiev 1938
It is important to understand that endings are

always concrete, i.e. a moment always comes
when general strategy gives way to specific cal
culation. Many technical errors result from fail
ing to adjust to this, and continuing to play by 42...ub3 43 axb3 4Je7 44 iLcS! liJfS 45 b4!
general principles after the time for concrete �7 46 Wc2 <li>h6 (D)
play has come. Here, we have a simple example
of how effective a concrete approach can be.
White has two extra pawns and can win by 36
l:h8 and 37 l:td8. However, since Black's pieces
are awkwardly placed along the 5th rank:, White
plays to exploit this.
36 l:tg5+! Wh6 37 f4
Black is totally tied up.
37 b5 38 iLf2! a5 39 c3 fi:..e7 (D)
•••
Black is completely bound hand and foot,

and has no defence to the following break
through, which sets up a winning passed pawn
on the queenside.
47 c4 ltJd4+ 48 �d3 bxc4+ 49 'iti>xc4 1-0
A fine example of meticulous care in a win
ning position. Yes, White had many ways to
win, but such precision is always best, no mat
ter how overwhelming the position.
Schematic Thinking
Another classic technique in the endgame is the white king." (Vaganian). Only now does
thinking in tenns of schemes. White concern himself with exactly how to
achieve his scheme.
42 lbds 43 'iti>e2 �b7 44 e4 dxe4 45 c;i;>e3
•••
'it>h7 46 i..e7 'iti>h6 47 c;i;>xe4 'iti>h7 48 c;i;>f4 'iti>h6

49 i..f8 tbd8 50 c;i;>g3 �b7 51 c;i;>h4 lbd8 52
i..a3 �b7 53 i..cl + 'iti>h7 54 cbg5 'iti>g8 55 i..a3
c;i;>h7 (D)
Vaganian - Postny
Moscow 2()()4.
Vaganian: "Black thought he had a fortress,

since the white king cannot penetrate to the
queenside. But my plan is simple: the king goes
to g5, the exchange ... hxg4 is forced, then the 56 h6! c;i;>g8 57 i..e7!
white pawn goes to h6 and forces an entry for 57 i..f8? gxh6+ 58 i..xh6 c;i;>h7 59 i..f8 c;i;>g8
the white king via f6." Vaganian does not worry 60 i..b4 c;i;>g7 is a draw. 'Therefore White waits
at this stage about how to achieve the desired until the black knight is on a dark square; only
set-up, but first determines where he wants his then does the white king get to f6."
pieces. 57-.!DaS 58 i..rs! ! (D)
36 f3 �b7 37 g4 hxg4
This exchange is forced sooner or later.
38 fxg4 c;i;>g6 39 h5+ 'iifi>h6 40 i..f8 lLias 41
c;i;>g3 �b7 (D)
58 �c6
58 ...gxh6+ 59 i..xh6 'iti>h7 60 i..f8 'iti>g8 6 1
•••
i..b4 ! (the point, gaining a tempo to allow the

white king to reach f6) 6 l ...�b7 62 c;i;>f6 and
42 � White wins.
"White needs to eliminate the e3-pawn. Then 59 b7 gxh6+ 60 i..xh6 a5 61 'iii>f6 a4 62 i..c l
a check from c l will secure the g5-square for 'iti>f8 63 i..a3+ 'Oti>e8 64 i..d6 1-0
[l
- - -·-
[
[ TIP SQ Bad Bishops are Often Good Defenders
.
- ·
nesses, and the winning plan therefore may oc

Bad bishops are often good defenders of weak l:.f3 l:.h3 53 l:.d3 g4 54 .l:!.e3 :h2 55 l:.el l:tg2 56
l1e3 b6 (D).
casionally involve exchanging off the opponent•s
bad bishop, paradoxical as this may seem.
White is in zugzwang and must lose a pawn,

in view of 57 l:td3 .l:.e2. Note how vulnerable
Eliskases - Larsen his weak pawns are (especially e4), now that
Mar del Plata 1958 the bishop does not protect them.
51-*.el 52 J.xb7 .ig4!
White's bishop is very bad, but his position White's bishop is now trapped, and he lost
is hard to break down. The main reason is that on � 75.
his bishop actually does a good job of defend
ing his weak pawns on e4 and b3. Larsen real Hem is another classic example of the same
izes that the paradoxical winning plan is to theme:
exchange the bishops off, after which White's
pawns are much more vulnerable.
49 .if7! 50 .ig4 .ih5! (D)
.••
- w
- - -- .
w � ... . • •
. ... . . .
"' � m . ...
� � � � .JI.
� - � - � -i.-
� n n
�� u u
. . . /. Petrosian - Bannik
� -- - - . USSR Ch (Riga) 1958
51 J.c8 Petrosian played 18 J.c5! exchanging off

This loses tactically, but exchanging allows Black's bad bishop. He went on to win a long
Black to show his idea. Larsen gives the follow ending, where the weakness of the black pawns
ing illustrative variation: 5 1 .ixh5 lhh5 52 (e.g. f6) proved decisive.
The Principle of Two Weaknesses Revisited
Now Black must also watch for the white

king entering via d3-e4-f5, etc. See also Tip 1 2
for a similar example.
36 ltJd.6
•••
36...tbb6 37 il.c2 tbc8 38 g4 tbb6 39 'ifi>d3

<ifi>d5 40 il.b3+ 'iti>c5 41 'iti>e4 (D), winning, is a
typical example of the importance of the sec
ond weakness.
Dydyshko - Balashov
Minsk 1 982
White has a solid extra pawn on the queen

side, and can immediately obtain a passed pawn
on that side of the board. It seems that nothing
could be more natural than 35 b4+. However,
after the further moves 35 ... axb4+ 36 axb4+
'iti>d5 37 1'.c4+ 'ifi>c6 the knight comes to d6 and 37 'ifi>b3!
White has no clear way to make further prog Now that the black knight has been diverted
ress. The passed pawn cannot get beyond b5, to defend the second weakness, White returns
and Black is able to hold the pawn up and keep to his queenside plans.
the white king out. 37 .fS 38 � e4 39 il.c2 (D)
••
Instead, since there is no hurry to play b4,

White creates a second weakness in Black's po
sition.
35 h4! B
Heading for h5 in order to fix the kingside
light squares.
35 gS (D)
•••
39 tt:Jc4?
•••
A blunder, but even after 39 ...'"1tb6 40 b4

axb4 4 1 'it>xb4 White is winning, as Black can
not cope with the twin threats of the a-pawn's
advance and the white king penetrating into the
kingside.
40 b3 l-O
36 h5 He loses the a-pawn for no compensation .
Pawn-Structure Subtleties
are a disadvantage in endgames, as we have

In general isolated pawns are a weakness and J:.xd4 21 J:.xf7 gxh6 22 g3 l:.d2, Black's rook
would have been able to take the a- and b-pawns
seen many times already in this book. However, on consecutive moves. In the game, however, it
its exceptions, and really strong players are ad

part of the charm of chess is that every rule has will take the rook one extra move to eliminate
both white queenside pawns. In addition, White
ept at finding and exploiting these. Here is a is able to use the b-file.
case where White voluntarily chooses to have 22 l:.bl! J:.g5
his queenside pawns broken up. 22 ... b6 23 l:ba7 l:td3? loses to 24 l:.a8+.
23 h4! l:.b5 24 l:t.xbS cxb5 25 J:.xh7 (D)
w
B
Browne - Bisguier
USA Ch (Chicago) 1974 Wbile's passed pawns are much quicker. He
woo . Wlows:
20 l:.xf7! 2.S lldl+ 2' «i>h2 J:.d2 27 l:.xb6 l:.xa2
20 'ii'xd4 is more natural, keeping White's Sec lbe earlier note!
pawns intact, but not as good, as we shall see. .21 J.s .D.sn 29 l:lb8+ <l;c7 30 h6 'iii>b6 31
20 Ji'xc3 21 bxc3 gxh6 (D)
•• ci>bl llS_(D)
The ending is going to be a race between 32 g4 b4 33 cxb4 axb4 34 J:.e8 .i:tn 35 �g2
White's kingside pawns and Black's queenside rl.f7 36 g5 :rs 37 h7 l:txgS+ 38 'it>f3 l:th5 39
counterplay. If White had played 20 1Wxd4?! b81i' :txh8 40 lhh8 1-0
r:�IP---8�J
---;f
L
More Breakthroughs
In lip 5, we saw some tactical pawn break Black had been suffering in a slightly worse
throughs in king and pawn endings. The same position for the whole game.
idea also occurs in other endings. 59 lli6 'ita7 60 h5 'ita6?!
Now Black's position appears lost. 60...gxh5
61 gxh5 'iita6 is a better try, although Black is still
struggling to save the game. Now White pro
duced the familiar breakthrough idea, thereby
securing a decisive passed pawn on the h-file.
61 g5! hxg5 62 h6 <ha5 63 h7 l:.d8 64 l:r.xf7
b5 65 cxb5 (D)
. - --- �
--.
� • · ![ · �
••• ··�
B
=�· • -
• n
B u B
Averbakh - Korchnoi
Erevan 1965 . �
. -�· .
i • • • �
Here, Korchnoi gave his time-troubled op-
•
ponent a very nasty shock:
36 g4!
-- -
•••
.
This forces a passed h-pawn.
37 hxg4 h3 38 1;c2? 65...'hb5
- -
The only hope was 38 f3 h2 39 li:Jf2, al After 65 ...cxb5, White wins by 66 ltg7 l:.h8
though after Korchnoi's 39 ... a5 ! Black should 67 d5 g4 (67...�b6 68 d6 �c6 69 d7 and 70
still win. l:r.g8) 68 d6 g3 69 (fl g2 70 l:r.xg6.
38 J':.c4+ 0-1 " :b7+! � 67 l:r.g7 l:r.b8 68 1;e4 1;b5 69
After 39 'iii'd2 lie l ! the first rank gets blocked ¢>f3 <k4 70 l:r.d7 �d3 (D)
.•
and the h-pawn queens.
71 �g4 l:r.xh7 72 l:.xh7 �xd4 73 l:.d7+ 1-0

After 73 ...1;e5 (trying to shoulder-charge the
Karpov - Hort white king - see Tip 2) 74 l:r.g7 ! c5 75 l:r.xg6 c4
Tilburg 1 979 76 l:r.xg5+ 1;d4 77 � White wins.
[ TIP 84 The Master at Work
:txh8 �xh8 30 :txf7 lLlc8 31 :tf5 with a large

advantage.
25 hxg6 fxg6 26 l:1hgl lt)bS 27 a4 (D)
• ll - n- -,- ,
�• ··. - - --
B
.t. . - . .t. m'
m m
��- -�- . .
�· •�B�•
� . - ...
�� .
Smyslov Ernst � .
u • . • .
� • _m.: a _
-
London (Uoyds Bank) 1 988
I was present at this game, and after heavy 27 tfJa7

•••
exchanges in the opening, many in the audience

b4 :tc3 29 lk I, with decisive effect.
27 ...lLlc7 loses control of the c-file after 28
(as well as Ernst, I suspect) expected a quick
draw. However, after Smyslov ' s next move, it 28 �d4 a5 29 es lt)c8 30 :ct ll'ib6
became clear that White has a serious advan 30... b6 3 1 f5 is winning for White.
tage. Rooks and bishops tend to work well to 31 :txc5 dxc5+ 32 WxcS lt)xa4+ 33 �d4
gether in such positions. lt)xb2 (D)
20 g4! 33 ... :tc8 34 d6 wins for White.
Seizing space and preparing to open lines on
the kingside.
20 h6 21 h4 tfJa7
--- .
.�·· � -
- -
•.•
He hopes to exchange all the rooks on the c w

file, but White does not oblige. �
-
� -
� ·
�� - ·
� · �,
�
22 :en ! �g7 23 f4 :tc5 24 h5 (D)
� �
-
•�n u • �

-
·� �
� � u
"�·
• •..t• •
- ·
- -
�� - � -
�
.� - �
•- - -
� u
• n _J
34 i..bS!
Now the knight is trapped in enemy territory.
The rest is just a mopping-up exercise.
34 a4 35 d6 exd6 36 exd6 l:1d8 37 d7 �f6
•••
38 :tel a3 39 <J;cJ :as 40 :te8 lt)a4+ 41 i.xa4

1-0
24 .:thS
.• It is hard to believe that the initial position is
Against 24 ... g5, Smyslov gives 25 fxg5 hxg5 lost for Black, but even harder to pinpoint what
26 h6+ 'it>g8 27 h7 <J;g7 28 h8'ilf+ l:1xh8 29 he did wrong thereafter.
Workmanlike Technique
38 �c7 39 a5 l:td6+ 40 i.dS ..ti>c8

•.•
After 40...bxa5 4 1 lhb7+ �c8 42 .nb5 the

a-pawn is weak.
41 axb6 f6?! (D)
Missing his one and only chance, which was
4 l ...tLlxb6; e.g., 42 �e5 �c7 43 l:.c5+ '>ti>d7 44
i.xb7 l:td2 with counterplay (Simagin).
Fischer - Barcza
Stockholm JZ 1 962
White has the better pawn-structure, and

l:.+i. are better than l:[-+2ZJ in such a position.
Fischer's technique is first-rate.
27 lldl � 28 l:td4
Stopping ...tLla4 (do not allow unnecessary 42 �e3 tLlxb6 43 i.g8
counterplay). Winning a pawn in view of 43 ...h6 44 i.f7.
28 l:tc7 29 h3 f5 30 l:tb4
••• 43 �c7 44 l:tc5+ �b8 45 i.xh7 tLld5+ 46
••.
Tying Black down by attacking the b6-pawn. <M3 tLle7 (D)

30 lbd.7 31 <;f;f1
•••
Activate the king.

31 �e7 32 We2 �d8 33 l:.b5 g6 34 �e3
•••
�c8 35 �d4 'ifi>b8 36 �d5?!

The only small blemish on Fischer's conduct
of the ending. As pointed out by Simagin, 36
i.d5, forcing 36...f6, was more accurate.
36 .l:tc6 37 �d4 l:te6 (D)
••
47 h4!
Preparing h5, to free the bishop. Black has
lost a pawn and still has no counterplay.
47 b6 48 l:tb5 �b7 49 h5 'it>a6 50 c4 gxh5
•••
51 i.xf5 l:td4 52 b3
The rest is easy.
52...tL!c6 53 ..ti>e3 l:td8 54 i.e4 lbas 55 i.c2
h4 56 l:th5 l:.e8+ 57 ..ti>d2 l:tg8 58 nxh4 b5 59
38 a4! l:tf4 bxc4 60 bxc4 J:txg2 61 l:.xf6+ <l;a7 62 ..ti>c3
Going after the b7-pawn. l:tg4 63 f4 tLlb7 64 'it'b4 1-0
A Two-Bishops Masterclass
20 i..a6 l:td8 21 b4 cxb4 22 axb4

Now Black must guard the c-file against en
try by the white rooks.
w 22 R.c8 23 i..e2 i..e6 24 'iPf2 .l:.ac8 25 .Ual
••.
lZ'ic6 26 1'.a6 (D)
It may not look as though White has much

here, but "he who has the bishops has the fu
ture", and the asymmetrical pawn-structure 26 :b8
helps to enhance their strength. Kariakin does 26....UC7 27 .Ube I .Udd7 28 1'.b5 leaves Black
•••
not commit any clearly identifiable errors, but badly tied up, since the attempt to free himself
is slowly ground down. tactically fails: 28 ...tt:Jxb4? 29 i..xd7 lZ'id3+ 30
16 e4 <M7 17 1'.e3 l:thc8 18 f3 .te6 19 l:tcl 'iti>g3 lZ'ixc l 3 1 R.xe6+ winning.
b6!? (D) 27 1'.bS lZ'ixb4 28 .Uxa7+ �g8 29 l:te7 1'.f7
Slightly weakening the light squares on the 30 ltal
queenside. The alternative was 19 ..c4, which
. Preparing to double on the 7th, while the
was recommended by Russian GM Sergei Shi b6-pawn is doomed.
pov. This may indeed be an improvement, but 30 .:as 31 l:td7 .Udc8 32 lh:a8 lh:a8 33
..
it is understandable that Black did not wish to i..xb6 (D)

fix a pawn on the same colour square as his
bishop. Such dilemmas are typical of these po
sitions. -
White has won a pawn, and retains the

bishop-pair. The remaining moves can be seen
in Tip 36.
I TIP s7- 1 Appearances ca n be Deceptive
An excellent move, which, amazingly, wins

a pawn. As well as 25 l:.d8+, there is also a
threat of 25 tbc4, when 25 ...b5 can now be an
swered by l:[d8+, winning the a5-pawn. Black
cannot meet both threats.
24 b5 25 l:t.d8+ lhd8 26 lhd8+ �h7 27
•••
tt:Jxf.7 tbc6 28 l:[cS tLlb4 29 l:[b8+ �g6 30

tt:Je5+ 'itf6 31 f4! (D)
Andersson - Gheorghiu
Moscow IZ 1 982
Superficially, this simplified, symmetrical

structure may look drawish, but White's supe
rior piece activity actually gives him a near
decisive advantage. He controls both open files,
his knight is more active, he has penetration
squares along the 6th rank, and his king is The use of tactics to maintain the advantage
nearer the centre. Andersson, the great end is a key element in good technique. Now grab
game grinder, wins in a canter. bing the pawn by 3 1 ...fua2? loses to 32 l:[f8+
20 l:[d6 h6 21 e4 and .:.n+.
Taking the squares d5 and f5 from the black 31-.�7 32 l:[gB � 33 l:[dS+ r:J;e7 34 :as
knight. tLixa2 35 lhaS (D)
21. .:.eS?! 22 � 1138
•.
Black is struggling for a constructive plan.

He settles on ...a5 and ... b5, but he could have
saved a tempo by playing this last move. B
23 �e3 a5 24 l:[cdl! (D)
White is a pawn up with the better position.

The re�t is easy.
35 M 36 Wd3 b4 37 .:.a6 tt:Jb5 38 tt:Jc6+
•••
Wd7 39 tt:Jd4 tLixd4 40 �xd4 �7 41 g4 1-0

TIP
·
·
[-- · 88 .
- - ---- --
Overcoming the Blockade
42 h3 � 43 f3 'ikc7 44 g4 (D)
w
B
P.H. Nielsen - Harikrishna

Hastings 200213 44 hxg4 45 fxg4!
•••
The decisive moment. Instead of the routine

White has an extra pawn, but it is securely capture towards the centre, White takes the other
blockaded and there are no other obvious tar way, setting up a potential outside passed pawn ·
gets in Black's position. Nielsen shows excel on the h-file. This pawn will stretch the defences
lent, workmanlike technique. to breaking point (the two-weaknesses princi
29 l:l.d2 'ili'f6 30 l:te2 ple). We have already seen other examples of
Exchanging pieces usually favours the side non-routine handling of pawns, so as to create a
with the material advantage. passed pawn; for example, Khalifman-Salov, Tip
30 .:xe2 31 'ii'xe2 'ii'd4 32 'ii'dl 'ii'b4 33 20.
45 'ii'c5 46 c;t>r3 'ii'b4 47 �e2 'ii'a3 48 �d3
••
'ii'd2 � (D)
'ii'b4 49 h4! tbe8 (D)
••.
33 ...tbc4? loses to 34 'ii'd4.
m m m -a m•p•-.a- - �
v, • •••• w v, - •••
m • ••a
w
m m ••a
� -�· - � -�� -
ma
- a� ••
� • � mu
- ·
-
� ��
� a
•
�
"
• � . d-
� R
-�� - .�
q ruu
�u n�� , • • • a
� - - -
- u
� • . m •--H
34 b3 tbf5 35 'ii'd3 tt:\d6 36 tbe2 'ii'et 37 50 'ifd4 'ili'a3
tbc3 'ii'eS 38 'ii'e3 'ii'f5 39 'ii'd4 50 ...'i!i'xb3 5 1 d6 wins for White.
White patiently improves the position of his 51 h5 gxh5 52 gxh5 �g8 53 'iie5 'i!i'f8
pieces. It is clear that Black can never exchange 53 ...lbct6 54 h6.
queens, so he is forced to give ground. 54 tbe4 f5 55 tbg5 tbf6 56 'i!i'xf5 'ii'b6 57 d6
39 'iti>g8 40 a4 'i!i'd7 41 'ili'e5 h5 tbxh5 58 d7 'i!i'd6+ 59 �e3 'ii'e7+ 1-0
Not waiting for 60 'ii'e6+ �f8 (or 60...'ii'xe6+
•..
White has patiently probed for a few moves,

and now starts the decisive advance. 6 1 tbxe6) 6 1 d8'ii'+ , winning.
-- ----,I
r:TIP 89
J
- - -
Squeeze Play
�---- _ J
31 exfS 32 tl\es tt:\xe5

.••
Forced, else g6 drops, but now White has a

powerful passed e-pawn, which proves deei
w sive.
33 dxe5 (D)
Aronian - Svidler
Morelia/Linares 2006
This position looks harmless for Black, but

this is not so.
17 g4! 33 b4+
..•
We have seen this idea several times already The pawn will be lost anyway after <itid4-c5.
in this book. White sets up a squeeze on the 34 axb4 axb4+ 35 �xb4 .i.d7 36 .i.b3 .i.c6
kingside. 37 �c5 (D)
17 h6 18 h4 tt:\a5 19 g5 hxg5 20 hxg5 l:r.xcl
•••
21 lhcl l:r.c8 22 lbcS i.xc8 23 tt:\e5

Black's kingside pawns are fixed on light
squares, and f7 is weak.
23 i.b7 24 a3 tt:\c6 25 tt:\g4 tt:\b8 26 �d.2
•••
.i.c6 27 �c3 tt:\d7 28 f4 a5 29 e4 b5 30 .i.c2 (D)
37 .i.eS?
•••
Losing immediately. 37 ....i.f3 is far tougher,

but even then, Marin's analysi s suggests that
White is winning. He plays i.g8-h7, and then
forces the exchange of the e-pawn for that on
g6. In the resulting position, he can drive the
30 fS?!
.•• black king back and get his own king to e5, with
Black loses patience. There are no direct decisive effect. Averbakh analysed a similar 2
threats, so he should sit tight. vs I position many years ago.
31 exf5! 38 e6 l-O
3 1 gxf6+ is more natural, but Aronian sees It is zugzwang, as the bishop has no safe
that the time is ripe to enter a bishop ending. square .
The Power of the King
42 f4 e4 43 i.h3 lt:Jf8 44 lt:Jn i.a4 45 rs

i.e8 (D)
.m_• _.. _m�

w �·- • -·
.w.- � .w�
� u••�u
n u
u � .t. B
�
� �
�
� .�
- �� . � �
- ....
. . . �
Smyslov - Rumens
Hastings 1 976/l
� • •tt:J•
4' Wf4
The power of the king as an active piece is 1he king enters the fray. The black king can
one of the main characteristics of the endgame. not come to e7, because it must keep guard on
Here, the white king marches right into the heart Wbilc's queenside pawn-majority - another ex
of the black position and nets a whole piece. mnt*of tbe principle of two weaknesses in ac
38 gS ltJd7 39 e4 e5 tion.. .
Black tries to confuse matters rather than � 47 lt:Je3 b6 48 cxb6+ 'it>xb6 49 'ifi>eS
wait passively, but on this occasion, this may ..... b5+!
be the wrong approach. Ifhe sits tight, his Iring Ddlccting the enemy king, and allowing a
can protect the central squares and prevent the -.:e1· : penetration.
white king from entering the position. It is not . !IL.�b5 51 'it>d6 (D)
then obvious how White can win.
40 'ito>e3 i.c2 41 exdS cxdS (D)
The white king cannot be prevented from

helping itself to a whole piece after the 52 <J;;e7,
Now the downside of Black's 39th move is trapping the knight or forking two pieces.
clear. The exchange on d5 has given White a 51 ..llbS 52 'it>e7 lt:Jg6+ 53 fxg6 i.xg6 54
•••
potential outside passed pawn on the queen 'it>d6 'it>b4 55 lt:JxdS+ <J;;c4 56 'it>eS 'it>d3 57
side, which proves vital as play proceeds. lt:Jf4+ 1-0
[ TIP 9 l ] Calculate You r Way to Victory
Black must keep the b3-pawn blocked. After

49...i.a3? White frees his bishop by 50 b4!
i.xb4 51 i.b3 'iPb2 52 'it>dl and draws.
50 'iPd3 (D)
After 50 a4 �c 1 the bishop is lost.
General principles are one thing, but most

endgames become concrete at some stage, and
accurate calculation is needed. Here is one of 50 �bl!
•••
the most impressive examples I have ever wit Karpov had to see this move, and the next,
nessed, as I was present at this event, and before trading rook s. Both 50...'it>c l 5 1 'iPe2
watched the whole game from a few feet away. i.e7 52 b4! and 50...'iPxa2 5 1 �c2 only draw.
Karpov had already outplayed his opponent 51 a3 �cl!! 52 <ite2 i.xa3 (D)
from a very drawish-looking position. Here,
he thought for about ten minutes, and then
played the rest of the moves instantly.
45 .:ld2! 46 l:txd2 exd2 47 i.dl 'iti>d4 48
••
�2 'iPc3 49 �e2 (D)

The key to Black's winning idea is that the
white bishop is extremely short of squares.
53 b4
The only way to save the bishop, but by com
parison with the line given in the note to Black's
49th move, here White has had to give up the
a-pawn as well, which proves fatal.
53...axb4 54 i.a4 'iPb2 55 �dl b3 56 i.c6
'iti>al 0-1
Beautiful play by Karpov.
102 JO l CHESS ENDGAME TIPS
c -c �--1
�9� Fighting for Space
With the threat of breaking in down the h-file

space, and not allow one's position to be clamped
In many endgames, it is important to fight for
(two weaknesses).
by an enemy pawn advance. 41 ..:t.d6 42 l:thl (D)
••
Stein - Bronstein Simply comparing with the starting position

Moscow (Alekhine mem) 1 971 shows how much progress White has made in
Although an exchange for a pawn down,

the past five moves.
42 l:t6d7 43 l:th4 :as 44 hxg6 fxg6 45
•••
White stands better, because of his passed d l:tf4+ .U.f7

pawn and Black's weaknesses on a5 and c5. Else the white rook goes to f6 and then c6.
Even so, Black should be able to draw with cor 46 l:txf7+ cltxf7 47 d6+ �e8 48 .td5 l:tc8 49
rect defence. .llc6+ � 50 �c4 (D)
37 �6?
Too passive. It is essential to fight for space
•••
in such positions, so 37 ... h5 gives Black much

better survival chances.
38 g4! (D)
The white pieces dominate and he will sim

ply pick off the c5-pawn and then win with his
two passed pawns. Note how passive Black's
position has been, ever since he allowed White
to clamp the kingside.
Now White gets a clamp on the black king 50 'ito>e6 51 .llxc5 .txc5 52 'itixc5 nrs 53
•••
side and his advantage grows significantly. iLd5+ �d7 54 f3 l:th8 55 .llc6+ �e6 56 b6
38 wg7 39 g5 .:.ad7 40 h4 ..ti>f8 41 h5
••• l:td8 57 b7 l:tb8 58 f4 1-0
r =---� � 1
�_J The I mportance of Playing On
28 .:S? 29 .:t6d2?! (D)

••
White fails to find anything to do;and soon

loses the initiative.
Vallejo Pons - Topalov

Linares 2005
"You don't win by resigning", the saying 29 b4 30 axb4 rLb8!

•••
goes. One can add to this that you don't win by A few vacillating moves by White have given
agreeing draws either. An equal position is not Black a large advantage already.
the same as a drawn position, and many games 31 .:a1 :Xb4 32 lhJ
have been won by a determined player grinding �te is terribly passive after 32 i.a2 i.c4.
on in an objectively equal position. TIIis is a 32 .a4 33 .ia2 .ic4 34 i.bl i.fi 35 i.a2
••
case in point. With a near-symmetrical, heavily ltab7 (D)

simplified position, many players would be
tempted to shake hands, but Topalov just plays
on, and the result is that a world-class player
loses as White inside 20 more moves.
21 l:r.6d4 .:acs 22 lbel a5 23 f3 ltc7 24 e4
l:dc8 25 a3 g5 (D)
36 lZ'ic2?
Collapsing immediately. 36 <iii>f2 keeps White
on the board.
36 .:Xb2 37 l:r.xa4 l::tc7! 38 'it;f2
••
38 lZ'iel l:r.c3+ 39 'itr>d4 1Ud2+ 40 'itr>xc3 .:tdl

is equally hopeless for White.
38 i.b5 0-1
An
Securing e5 for his knight.
26 'itr>e3 ll'ie5 27 l:4d2 h6 28 l:d6?!
•••
object lesson in the advantage of playing

It is better to exchange knights by 28 lZ'id3. on in equal positions.
104 JOI (;HESS ENDGAME TIPS
[;- 94 The Outside Pawn-Majority
Black's rook is the more active, so he avoids

its exchange.
33 l:th3 l:tdl+ 34 'iti>c2 l:tgl (D)
• • • -.
. -
w - - -. .
-
• • ··· · �
�·• • m
.n·•�· �
� u • • • .:
� o�• • �
� •---• /. ,
This may appear a fairly equal position, but 35 1'.d2 g5 36 l:tb6+ <M7 37 c.Pd3 l:tg3+ 38
in fact Black's superior kingside pawn-struc 'itte2 l:tg2+ 39 �d3 g4 40 e5
ture gives him the advantage. White's passed e-pawn is no use at all, so he
25 f5 26 c;t>r2 c;f;>f7 27 �e3
•.. trades it off to get at Black's queenside pawns.
Note that he cannot himself exchange pawns 40 1'.xeS 41 l:txa6 g3 42 1'.e3
..•
because of 27 exf5? 1'.g3+! and Black wins the A sad necessity, as 42 ...l:txd2+ was a threat.
exchange. 42 l:txb2 43 a4
•••
27 fxe4 28 fxe4 (D)

••• White hopes to exchange the queenside
Now the 2 vs I kingside majority promises
so as to reach J:t vs l:t+1'., but Benko is too alert
pawns and sacrifice his bishop for the g-pawn,
Black an outside passed pawn, whereas White's
own passed e-pawn is easily controlled by to allow this.
Black and constitutes more of a weakness than 43-b4 44 cxb4 g2 (D)
a strength.
The rest is easy.

28 'itte6 29 h5 .:CS! 45 �e4 J:.e2 46 � l:txe3+ 47 � J:.b3 48
b5 1'.d6 49 'iii>f2 c.Pe6 50 �e2 �d5 51 :a8 �c4
•••
Preventing White from exchanging bishops

by 30 1'.f4. 52 l:ta6 1'.c5 53 l:tc6 l:te3+ 54 �d2 l:te7 0-1
30 hxg6 hxg6 31 'iti>d3 .i::r.n+ 32 It.e3 nn White loses both of his pawns.
Too Brutal
31 �eS 32 �d2 �d7 33 �c2 �e8 34 �b2

..•
'it>d7
Black is totally helpless, and can only await
developments.
35 �a3 �e8 36 <ifi>a4 �d7 (D)
Tartakower - Asztalos
Budapest 1 913
White has a dream position - a huge space

advantage, a target on c5, and crippled black
37 .i.b8!
Preparing the road for the king. The immedi
bishops. However, he still has to find a way to
break through, and Tartakower accomplishes
this very systematically, with a few elegant ate 37 �a5 Ji..d8+ gets nowhere, so the bishop
touches thrown in. first manoeuvres to b6, to shelter the white king.
28 tt'lh2 �e8 37 ..tr>cS 38 .i.a7 �d7 39 .i.b6 d4 (D)
•••
The attempt to get counterplay with 28 . . .d4 Tartakower shows that further passive play
is nicely refuted by 29 tt'lg4 dxe3 30 Ji..g7+ fol by 39 ... Ji.. a8 loses after 40 �a5 .i.b7 4 1 tbe5+
lowed by tt'lf6. �e8 42 .i.c7, followed by �b6.
29 tt'lg4 Ji..e7 (D)
40 e4 �e8 41 e5 �d7 42 lbf'2 1-0

He has to keep the knight out of f6, else he The knight comes t o e4 and wins the c5-
loses the h7-pawn. pawn. In the first volume of his best games col
30 .i.e5 <iitd7 31 <iite2 lection, Tartakower recalls that he was at first
Bringing the white king into Black's position awarded a special prize for this game, but the
is the final step. The king is ultimately heading judges later changed their mind and took it away,
for b6. on the grounds that the game was too brutal !
[, 96 · J
I. '--;-· --
The Minority Attack
This advance is the climax of Black's minor

ity attack.
34 bxa3
34 l:tbl is better, although still bad after
34... axb2 35 :Xb2 l:tc3+ 36 �e2 lhg3 37 l:txb4
l:tg4.
34 Jlc3+ 35 �e2 l:txa3 36 l:td4 l:txa2 37
••
l:txb4 ltaxc2+ 38 nxc2 l:txc2+ 39 <iite3 l:tc3+

40 <M2 �c6 41 l:tb8 (D)
Martin - Yanofsky
Toronto 1938
Many endings arising from the Sicilian De

fence favour Black, because of the half-open
c-file and a potential minority attack on the
queenside. This position, although it actually
arose from a French Defence, shows the typical
Sicilian endgame. White's problems are exac
erbated by his advanced kingside pawns, which 41-"7
are vulnerable to a black rook getting round the 41 ¢>c5! wins; e.g., 42 l:tb7 ..ti>d4 43 l:.xf7
...
back. That may not seem very likely at present, �e4 44 .:r6 :C2+ 45 �e 1 d4, etc., but Yanof
but Yanofsky soon manages it. sky is in no hurry. He can always return to this
23 l:thc8 24 :C1 l:tc5 25 .i.e2 l:tbc8 26 l:td2
••. plan later.
.i.a4 27 Jid3 'ifi>d7 28 �d4 .i.b5 42 :rs �e7 43 .:as l:tc7 44 g4? (D)
Exchanging the bishop which defends the
c2-pawn.
29 .i.xb5+ l:txb5 30 :n l:tbc5 31 �d3 l:tc4
32 g3 a4 33 .l:tcl (D)
Hastening the end, but White knew that pas

sive play was hopeless anyway.
44 J:c2+ 45 �e3 hxg4 46 l:ia7+ � 47 f5
••
gxf5 48 'iti>r4 lk4+ 49 �g5 g3 50 cM6 �e8 51

IZ.xf7 g2 0-l
101 CHES!Y ENDGAME TIPS 107
�P 9i ] Attacking the King
Naturally, direct attacks on the enemy king are If40...f6, then 4 1 ..id5+ llJf7 42 l:tagl � 43
relatively uncommon in endgames, but the ab l:f.h7 is crushing. In this, as in other variations,
sence of queens does not invariably preclude the power of White's bishop-pair is central.
a kingside attack, as Kramnik demonstrates 41 :b4 (D)
here.
41 ...f6
Kramnik - Ulybin Or 4 1 ...llJh7 42 :ah l liJef8 43 ..ie4 g6 44
Khalkidhiki 1 992 ..id5 :d8 45 f4 and Black has no defence to f5.
42 l:tgl <j;f7 43 ..td5!
Black is somewhat tied up, but his queenside Avoiding the nasty trick 43 f4? liJh3 ! . Tacti-
weaknesses are defended adequately, and it is cal alertness is always important !
not easy for White to make further progress on 43...'it>es 44 :hs+ llJrs 45 f4 liJge6
that side of the board. Kramnik therefore sac 45...llJgh7 46 ..tg8, or 45 ... llJf7 46 ..ic6+.
rifices a pawn to convert his space and mobil 46 'it>f3
ity advantage into a direct attack on the black 46 ..ixe6 :Xe6 47 llxg7 wins, but there is no
king. hurry - Black is so tied up that he can hardly
36 g5! hxg5 37 hxg5 llJxg5 38 ..ic6 l:tc8 39 move a piece without dropping something.
:ht 'itg8 40 l:aal ! (D) 46...fS (D)
Simply threatening to double rooks on the 47 :lg6! llJcs 48 ..txg7 :C7 49 .txrs 1-0
h-file. Black loses a rook after 49..J:txf8 50 Axf8+
40 llJce6
••• 'itxf8 5 1 :g8+.
The Power of Zugzwang
Zugzwang is an important element in many end Now White's passed d-pawn will decide the
ings, and is often the simplest way to clinch a po game.
sition. Here, we see Fischer using the weapon to 27,_<i;gT
perfection, to wrap up a decisive positional ad After 27 ...lt:Jd7, Fischer gives 28 :Z:.c7 lt:Jb6
vantage. 29 .ib3 �h8 30 :Z:.xa6 lLlxd5 3 1 .ixd5 .:.Xd5 32
:Z:.aa7 winning.
28 :Z:.b3 :Z:.f7 (D)
Fischer - Rossetto
Mar de/ Plata 1959
29 d6!
White's trumps are his active bishop, which While punues his initiative energetically,
cooperates well with the rooks, and the queen takiDB advantage of the passive and uncoordi
side pawn-majority. Clearly, the plan is to break nalf.d black pieces.
with c5, but first the black knight has to be di 29-� 30 ::.C7 lt:Jf8 31 l:tbb7 :Xc7 32
verted. dxc7 :Z:.c8 (D)
24 .:aJ!
Threatening to win the a-pawn by 25 hd7,
and so forcing the black knight to unguard c5.
24 lt:Jb8 25 c5 bxc5 (D)
If 25 ... b5, then 26 .ib3 :Z:.f7 2 7 c6 :Z:.c7 28
•..
:Z:.a5 followed by 29 a4, when the a6-pawn will

fall.
33 .ib3!
After this, no black piece can move: rook
and king moves allow l:tb8, and knight moves
allow .ie6. Black is therefore reduced to pawn
moves, and when these run out, he will be in
zugzwang.
33...aS 34 a4 h6 35 h3 g5 36 g4 fxg4 37
hxg4 l-O
JO] CHESS ENDGAME TIPS 109
-·----·-· -]
TIP 99
�-- Passed Pawns are the Key
· positional advantage. 1be pawn on c7 ties Black

hand and foot and will sooner or later result in
significant material losses.
w 24 ..h4 25 l:td6 �6 26 lhb6 <j;xe7 27
••
:tb7 �6
Both 27 ...iLd?? 28 c8'ii' l:.axc8 29 l:tdl and
27 ...'iii>d6 28 iLa6 iLd7? 29 :E.d 1 + 'it>c6 30 l:.xd7
'iti>xd7 3 1 c81V++ lose.
28 l:.cb1 ! i.c6 29 l:.b8 l:.c8 30 l:.cl (D)
Passed pawns are one of the key elements in

endgames. We have already seen various meth
ods of manufacturing passed pawns, and here
is another. The position looks solid and one
would expect a long period of manoeuvring,
but instead White finds a tactical means to
strike inunediately (never forget tactics!) and 30 .liaxb8
.•
set up a winning passed pawn on the c-file. Black must lose material; e.g., 30...i.d7 3 1
16 e5!! dxe5 nxa8 l:.xa8 32 i.a6 .i.c8 3 3 l:tbl ! .i.xa6 34
Now the queenside pawns become rampant, l:tb8.
but allowing exd6 is also unpleasant. 31 cxbS'ii' l:.xb8 32 l:.xc6
17 b5 i.e8 18 c5 lbd7 19 lDdS e6 20 lbe7+ Now White wraps up easily.
<j;g7 21 c6 bxc6 22 bxc6 lbb6 23 c7 iLd7 24 32 J:tb2 33 1'.c4 h5 34 h4 g5 35 hxg5+
••
lUdl (D) � (D)

- - -- - %
B w � - -·-
m : m .t m �
� �. - m m
-·
-ii.- - �
. - -�
�- - -��
�- - · ---=--
This position arises by force after 16 ...dxe5, 36 'iti>h2 J:lb4 37 1'.b3 a5 38 J:la6 J:lb5 39 J:la7
and it is easy to see that White has a decisive 'it>r6 40 1'.a4 1-o
110 10 l CHESS ENDGAME TIPS
Don't Forget Combinations!
now ready to strike with the decisive break

through.
46 .i.f6!!
Here is another example of the importance This is White's idea. The piece sacrifice cre
of tactics and combinations in endings. White ates a passed h-pawn, which proves decisive. 47
has a substantial space and mobility advantage, hg7, followed by 48 h6 is a threat - knights
but it is not apparent how he can actually break bate rook's pawns!
through, since the black knight is playing a key 46...g6
defensive role. Kramnik supplies the answer- a The main line is 46...gxf6 47 exf6 :cs (the
combination! bight cannot get to g6 - 47 ...tt'lh4 48 �g4 !
35 :dS+ :Xd8 36 .:Xd8+ �b7 37 �c3 � - �g5 and h6 wins) 48 :xc8 �xc8 49
Black is totally passive, so White first im ti>p"1d7 50 b6 lbxh6 5 1 'it>xh6 e5 (D) and now
proves his king. � the pawn ending:
37 �a6 38 �d3! :c7 39 �e4 �b7 40 .ld.1
•••
..t>c8 4 1 :dS+ �b7 (D)
52 ..t>h7 ! (triangulation) 52 ...e4 (52 ...�e6 53

�g7 a3 54 f3 zugzwang) 53 <j;g7 'it>e6 54 a3
42 ..t>f4! :cS 43 :td7+ :c7 44 :d3 ..t>c8 45 zugzwang.
:dS+ ..t>b7 (D) 47 hxg6 fxg6 48 ..t>g5 1-0
White has improved the position of his A fine example of the importance of combi
pieces to the maximum extent possible, and is nations and accurate calculation in endgames.
Ill
�------ -1
101 CHESS ENDGAME TIPS
TIP
----- -�
l Q li Positional Draw
The positional draw is a concept familiar from 49 g6?!

•••
the world of endgame studies, where some as At once Black commits an inaccuracy. Bot
tonishing examples are to be found. The idea vinnik gives 49 ...h5 ! 50 l:r.b8+ (50 c6 l:r.c3 5 1
arises much less often in practice, but it should l:.b7 lDh6 52 l:r.d7 lDg4 5 3 l:r.xd6 l:r.c 1 + 54 'iii>g2
not be forgotten, as it can sometimes enable l:r.c2+ is also winning for Black) 50 ...'iii>h7 5 1
precious half-point<> to be salvaged from appar l:.f8 dxc5 ! 52 l:r.xf7 exf4 5 3 gxf4 l:r.xe4, when
ently hopeless positions. Black should win the rook ending, since White's
pawns are too weak.
50 c6 l:.c3 (D)
Botvinnik - Keres
USSR Ch (Moscow) 1951
51 l:r.b7! '1;g7?
White's position is lost, but with some help Now it is a draw. There is still a win by
from his opponent, he manages to escape by 5 I ...'1;f8 ! 52 l:r.b8+ 'iii>g7 ! 53 :.b7 f5 ! 54 fxe5
means of a positional draw. dxe5 55 exf5 gxf5 56 l:.d7 (56 c7 'iii>f6) 56 ...� !
48 c5! 57 c7 �e8, etc.
The only chance, using Black's back-rank 52 c7
weakness to set up a passed c-pawn (passed Now it is a draw, as Black cannot free his
pawns mean counterplay !). After 48 1'.f2 lDd8, king and knight without allowing c8'ii', and his
the knight comes to c5 and White has no coun rook: is tied to the c-file.
terplay at all. 52. .:c2 53 '1;gl hS 54 h4 l::tc4 55 'iii>g2
48 .:xe3 49 lbb2 (D)
•
•• l:.c2+ 56 c;i;>fJ. l:.c4 (D)
Threatening 50 l:r.b8+.

Giddins - 101 Chess Endgame Tips

Uploaded by

Copyright:

Available Formats

Giddins - 101 Chess Endgame Tips

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Giddins - 101 Chess Endgame Tips

Uploaded by

Copyright:

Available Formats

101 Chess

Copyright © Steve Giddins 2007

Gambit Publications Ltd, 99 Wallis Rd, London E9 5LN, England.

Edited by Graham Burgess

Gambit Publications Ltd

Kings and Pawns

2-The Gentlemanly Art of Shoulder-Charging

Tip 7 - Simplifying into a King and Pawn Ending 14

Tip 15-Barrier Reefs (Part l) 22

Tip 29-Knight vs Bad Bishop 36

Tip 3 J -Knights are Better than Bishops at Attacking Weak Pawns

Tip 37 -Know the Basics 44

Tip 54 - New York J 924 Revisited

Tip 62-Evading the Checks 71

Strategy and Technique

Tip 65 - Exchanges to Realize a Material Advantage 74

Tip 72 - Centralizing the King 81

'lip 80 - Bad Bishops are Often Good Defenders

Tip 81 -The Principle of Two Weaknesses Revisited 91

Tip 84 -The Master at Work

Tip 88 -·Overcoming the Blockade

Tip 93 -The Importance of Playing On

Important Endgame Principles

Beliavsky & Mikhalchishin: Winning Endgame Technique

Smyslov & Levenfish: Rook Endings

Shakhmatny Biulleten, 1955-91

R. Reti Tarrasch had exchanged into this position

return via a different diagonal, one which is not

Schlage - Ahues If White tries to bring his king to the queen­

4 <ifi>b7 'it>dS 5 'iha7 'iii'c6 6 �b8 and White

Again, the only move to draw, keeping the

In many king and pawn endings, it is vital to

tion. 5 8 <j;lg3? loses to 58 ...<j;lg5 59 <j;lf2 <j;lh4 60

59 <j;lg3 also draws.

Black tries his last trap.

Passed pawns can be manufactured in the most 42...f4! 43 'ifi>dS

In both of the preceding examples, the de­

Despite the symmetrical pawn-structure,

The Swedish GM casually played an incau-

White's reply left him with a very red face:

'ifi>g2 'ifi>e4 53 h7 1-0

[TW 5 More Pawn Breakthrough Ideas

Were it not for an immediate breakthrough,

42 'iti>g5! 43 d4 �xg4 0-1

�;P 6.] Don't Forget about Stalemate Resources

Stalemate ideas can often rescue an apparently 44 fi'c7+ 45 �??

hopeless position, and should not be forgotten. 45 b6 wins.

Unfortunately, the great Russian champion

The Ukrainian GM had previously been win­

Simplifying into a King and Pawn Ending

Simplifying to a king and pawn ending can fre­

White can win in many ways thanks to his

Things look hopeless, but Smyslov found a

Giving White a protected passed pawn!

50 c3 a5 51 cxb4 axb4 11z_11z

45 ...<;i;>xe5 46 'it>f3 and White wins.

I TIP 8 ] Knights Hate Rook's Pawns

can the black king.

Bonner - Medina Garcia

Black forced a winning passed a-pawn by 65 tlle6 l-O

Schlage - Ahues If White tries to bring his king to the queen

In both of the preceding examples, the de

The Ukrainian GM had previously been win

Simplifying to a king and pawn ending can fre

Sometimes, even getting the king in the cor

36 gxf4 <ii;>xf4 37 f3 .i.e7 and the h-pawn ad

In Basic Chess Endings, Fine quotes this po