Andrew Soltis - Grandmaster Secrets, Endings

GRANDMASTER
BY
GIVI Andrew Soltis

Caricatures hy Roh Long
.:....::..=..:
Thinkers' Press, Inc. Davenport IA u.s.a. 1997
Grandmaster Secrets: Endings
Copyright 1997 by Andrew Soltis

All rights reserved. No part of this
work may be reproduced or trans
mitted in any form or by any means,
electronic or mechanical, including
photocopying and recording, or by
any information storage or retrieval
system, except
as
may be expressly
permitted by the 1976 Copyright

Act or in writing from the publisher.
First Printing: January 1997
ISBN: 0-938650-66-1
Requests for pennissions and republication rights

should be addressed in writing to:
TIrinkers' Press Inc.
Editor, Bob Long
P.O. Box 8
Davenport IA 52805-0008 USA
CONTENTS
JrILC:
1
What Every Grandmaster Knows About the Endgame
2
Getting Started
3
Rules
4 Mismatches
5
Techniques
ti
]Etlans
R()()
8 Queens
"
......
25
..........
41
..................
67
............
77
.........
1 15
141
1!)j7
!J
10
11
12
1E'1LS
16
.....
183
...
205
Knights
Bishops
The Chess Oub
1Url4dl
214
215
Colophon
The Thinkers' ess Catalog
...
ttl
iv
Preface
A lot of readers are under the mistaken impression that authors write
chess books in one fluid motion, moving from an idea to an outline to a com
pleted manusctipt without interruption. No way-as the history of this book shows.
My otiginal idea for it had been lurking around, undisturbed, in the back of
my mind since the Ford Administration: I wondered if you could Wlite a book
teaching the endgame by means of a
Socratic dialogue.
I visualized a conversation, punctuated by diagrams and simple analysis. It
was going to be a master talking at length with a young player who already knew
all the Trends In The Whatsisname Attack and who could improvise through
most middlegames-but who suddenly collapsed in the endgame.
My idea was that by means of Q&A the master and this endgame-challenged
amateur could arrive at certain eternal truths about endings and understand
why
they are true. I knew from enough Class B players that they had heard about
zugzwang or triangulation or the checking distance of a Rook yet didn't appreci
ate how or why these things could work in their own games.
But the idea was difficult to flesh out. After all, a lot of smart people have
been trying to capture the essence of Socratic dialogue since 399 B.C. without
success. I eventually realized this project could tum out to be very wordy, very
philosophic-and not very useful to someone trying to figure out why a Lucena
position is not a Philidor position. Like a lot of my ideas, this one remained as a
bunch of scribbled notes on yellow legal-sized pads at the back of a very deep file
cabinet for more than 15 years.
In the spring of 1993 I was in between books when I exchanged thoughts

with Bob Long about our next project. A really honest, down-to-earth book on
the endgame would be worthwhile, I wrote Bob, because "most people are
bewildered by trying to figure out what they really need to know."
The problem with endgame books, Bob wrote in reply, is that they're usually
"terribly designed and poorly written." The information presented is often ar
cane, the type too small, the wording confusing, the paragraphs too long, the
book just ugly. He wondered if I had any idea for a format that would both
entertain and teach and yet say something about endings that hadn't been said
before.
And so Noah and Pat were born. (The names of the two characters came
from my wife Marcy. Blame her.) I revived the idea of a dialogue but, since Pat
was your typical '90s junior, it was going to be a conversation with an attitude.
Bob and I worked up a plan for separating the analysis from the talk, for using a
lot of quote boxes and visual aids, and the rest fell into place. And the title? Well,
let's say there were several names on this book before we settled on this one.
Andy Soltis
New York
vz
1997
Scene:
A chess cluh, the home hase oj the veteran grandmaster,
Noah Tall.
Enter Pat Sayre, a talented, but young club amateur. Sayre

has just finished a game as White which began:
1. e4 c5 2. f3 d6 3. d4 cxd4 4. xd4 f6
5. 4)c3 a6 6. f4 e6 7. d3 e7 8. twf3 c6
9. 4)xc6 bxc6 10. b3 e5 11. f5 d5 12. b2
g6 13. exd5 cxd5 14. 0-0-0 0-0 15. t!hel
twd6 16. 4)xd5!? twxd5 17. e4 twc5 18.
xa8 xf5 19. twc6 twf2 20. xe5! c8
21. twxc8t xc8 22. xe7 f5 23. e4!
twe3t 24. bl twe2 25. xf6! twxdlt 26.
b2 d7 27. d3 twg4 28. c4 twf5 29.
c3 b5 30. xf7 twxf7 31. xf7t xf7
(oJffiTijr:JI
'!8
mimtl:}f
Ctt t:1 , t[. , ,'f.l'."it;:"

't'l'l'J:-
iIiIfllf
f
ijmit!F:l ...t
What Every Grandmaster Knows
middlegame well.
Pat: Dam!
Noah: You lost again?
Pat: Well? I was crushing him!
Pat: And from a dead-won A * sac even.
game! I should have gotten at
But then something hapleast a draw. Let me show you pened after I got his * back.
I don't know what my blun
on a board.
Noah: Don't be so hard on der was.
yourself.
Noah: It wasn't just a blunder.
Pat: Why not? I must be the
It took many bad moves to
only idiot in the world who lose the win before you lost
can lose such a position.
the draw:
Noah: At least you played the
You put your pawns on the
wrong color.
You dilly-dallied with your
.
32. b4?
With 32. 11,d2 and 33. c4 White
should win swiftly.
32.
33. a3?
34. <Ile1?!
c4
<Ile6
Loses time compared with 11,g7

and Wc3.
42. c3 ?
You left your passive.

To avoid the 42. a4 11,dl skewer,
34
d5
Pat: And the worst part is I
but then 43. xa6 xc2 44. a5
<Ilf5
35. g3
know I'll never get any bet
might still have won.
36. <Ild2
<Ilg4
42
g5
37. <Ile3
<Ilh3
ter. I just hate endgames.
43. d2
g4
38. <Ild4?
Noah: Relax. You're no differ
<Ilg2
44. a4?
With 38. 11,e5 White is still betent from anybody else. All
45. f4
ter.
amateurs loathe when the 'Ills
38
f3
go off.
Chapter One
45
46. b5?
47. axb5
48. c5
Resigns
f3
axb5
h4!
h3
The h- i promotes in two

moves.
Pat: It's not that I'm so great in knowledge are so extremely

the middlegame. But there's rare you don't need to know
just so much you need to it.
know about the ending.
Pat: So what do I really need
Noah: Not that you need to to know?
know. Don't believe every Noah: You can put everything
thing you read in books like you III list absolutely know on
Reuben Fine's Basic Chess a f\irly short list. Basically,
Endings. Seventy percent of you need to know what it
the information in them is takes to win.
impractical.
Pat: I thought I was the only
one who found Fine impos
sible to read. But what do you
mean by "impractical?"
Noah: I mean most of what he
covers will never occur in
your games.
It's nice to master the theory
of corresponding squares or
be able to play <&>+ a +jt v.
+ a flawlessly. But the
times when you can use that
9
The Estimated Odds of Your:
Ever Going Bald if male
............................................
Ever Dying of a Heart Attack

Ever Playing
.................................... 1 in 5
..
t! + + vs. t! +\t>
Being Robbed This Year .

.
Playing Out
......
...
. .. . .
..
..
..........
.....
.... ..
.....
..
1 in 40
.. . . 1 in 500
...........
...
+ +4) vs. This Year ................
Being Diagnosed with Lung Cancer This Year

Playing
1 in 2.5
....
1 in 3,000
. 1 in 7,500
*+4)- ft + vs. *+ This Year ............
1 in 8,000
Being Murdered This Year .................................... 1 in 12,000
10
Chapter One
emerge a clear ahead and win

without much effort.
For example, 2 ... h3t 3. Wd4
h4t 4. Wd5 a4 5. a8= xaH
6. xa8 g5 7. We4 Wg6 8. a()
and Black runs into zugzwang (8 . . .
g4? 9. Wf4 or 8 . . . h4 9. WfJ Wf5
10. a5t and 1 1. Wg4).
Ambrosz-Cioca1tea
Baile Herculane 1982
White to Play
1. a7!
This should end matters quickly
( I... a4 2. h8! a3t 3. Wc4
a4t 4. Wb3 xa7 5. h7t We6
6. xa7).
1.
2.
3.
4.
...
gfBt
a8=*t
*b7t
gxh4
wxf8
W;e7
Drawn
After Black puts his at some

safe square (g4, e5, f5) his fortress is
impregnable.
But with 2. h8! White would
mean like that you

can mate with W+2.Q. v. tJ
bulnot with +24)? Even I
know that sort of stuff.
Noah: Perhaps. But you'd be
SlI rpris('d h ow even veterans
ofinternational play err badly.
Here's a case in point (Diag.
4). White has a choice be
tween having an extra a or
having his w+* v. w+ a
He makes a disastrous choice.
Pat: I don't get it. The differ
ence between * vs. a and
a vs. Nothing is about the
same.
So why did Black draw?
Noah: In a middlegame the
differences are about the
same. But in an ending with
none of your own fts left, *
vs. a can tum out to be just 1
piece vs. 1 piece.
Pat:
You
11
Pat: So what does it take to Noah: As with most rules, this

one is most helpful when
win?
Noah: In most cases, you need making transitions, that is,
when changing the nature of
to be able to queen a t .
Pat: Speak for yourself. I usu the struggle.
ally need an extra or two. Pat: Like trading fls?
Noah: Well, you do usually Noah: Yes, trading anything is
need to be a ahead to win a transition.
if you have no pawns.
In Diagram 5 you see a
But 990/0 of all endgames strong G M throw away his
will have pawns. If you can last chance to advance in the
promote one, you'll probably candidates' matches by forc
WIn.
ing an endgame that wasn't
Pat: How does it help to know
nearly as favorable as the
middlegame he was in.
that?
Pat: Because he didn't know
what it takes to win?
Noah: Because he went with
the instinct that told him that
if you capture the other guy's
fI, there is no endgame. In
stincts are dangerous to have
in the endgame.
.
12
/f5, <I//d7) :
Sax-Korchnoi
Candidates match playoff 199 1

White to Pltiy
1. f8=*t
Much stronger is 1. h5!, which

probably would have caused im
mediate resignation.
1.
2. xf8
3. *xe6

*xf8
",xf8
Ofcourse, 3. *xg6?? g7Ioses.
3.
4)xh4
White can only try to win now

by grabbing jt s and trying to pro
mote his own a- or d- !. Black set
up an impregnable fortress (4)/e7,
After 28
a5.
White played another nine

moves before conceding a draw.
Chapter One
<lie3 13. <lin g3! and wins quickly.

For instance, 14. <ligl <lie2 15.
<lih 1 fJ! 16. gxf3 <Ii2 and ...g2t.
Pat: But don't I have to know trading is when you're

7
a lot of basic positions to get ahead. And in particular you
good?
don't want to liquidate all the
8.
dS
Noah: It helps-but it's far from
is on one side of the board if
9. d3
c5
essential.
you're only up one i on the
10. c3
g4
The more basic positions other.
And Black won after 1 1. <lid3
gxh3! 12. gxh3 <lid5 13. we3
you know, the more opportu Pat: Makes sense to me. But
<lie5 14. <lif3 f4! 15. <lifl <lie4 16.
nities you have to make win why did Kasparov do it?
we2 f3t 17. <lin wf5! (See Dia
ning transitions. Take Dia Noah: Because he was Kas
Alburt-Kasparov
gram on next page.)
parov. And because as a
gram 7, for example.
Daugavpils 1978
Black to play
schoolboy under Mikhail Bot
Pat: Yuch. A '111 ending. They
vinnik "I had to work funda
are impossible.
1. . . .
*dlt
Noah: Actually, '111+ i end mentally on W+ i endings
2. *gl
*xglt
ings are among the easiest to and knew this winning
3. xgl g7
4. fl
f6
method."
play.
Black secures the best <Ii-posi
But the point is: He might
But the point here is that 15tion before he tries to create a
year old Kasparov headed have won this ending with
passed i.
for the diagram-and violated out Botvinnik. Knowing the
S. e3
eS
a general principle-because basic Diagram 8 position just
fS
6. f3
he recognized the resulting saved him time and energy at
7. e3
gS
8. h3
the board.
+ i ending as won.
Or 8. WfJ g4t 9. <lie3 f4t 10.
Pat: What general principle?
<lie2 We4 1 1. 2 Wd3! 12. We I
Noah: That you should avoid
13
"Don't you know this is a

draw?"
-Pal Benko to Walter

Browne during their
game in the 1973 U.S.
Championship.
14
Pat: Well, I don't have a world Noah: You don't need them.
champion as a teacher. And I What you need to do is take
couldn't understand end- stock of what you know and
games if they provided Cliff don't know about the ending.
Notes.
Pat: I don't know much at all.
How much do I really need?
Noah: Well, the endgame
books usually list a lot of "es
sential" stuff, such as when a
*+ beats a + t on the
7th rank, and when it results
in stalemate.
And the notorious business
about the a t and wrong
colored . It can't be denied
that these positions do arise,
and occasionally an alert de
fender can save himself, the
"I know it But I don't
way Black does in Diagram
believe itl"
9.
-Browne, before
eventually drawing on
the 86th move.
..
This position (Diagram 8) has

been appearing in endgame texts
for more than a century. Black wins
because his W decisively reaches
g3 and wins the h- i . For example,
18. Wgl We5 19. 'lftfl 'iffe4 20.
'ifff2 Wf4 (and 2 1 . . . Wg3) or 20.
Wgl We:i 21. Wfl f2.
Chapter One
Pat: So this is stuff I really

need?
Noah: Not at all. You can be
come a master by only know
ing two basic positions, both
of them occuring III
W+ + ft v. W+ .
Pat: Why those?
Noah: Because +ft endings
Shirov-Mascarinas
are so common, and they end
Manila 1990
Black to play
up so often in one of the two
basic positions, the key g
b5!
1.
endings are those with one ft
Otherwise, Black dies the way
and most of them end up
most piece-down players do, from
zugzwang ( I... Wg6 2. We6 f4 3.
either in the winning "Lu
We5 g7 4. Wf5 etc.).
cena" position or the draw
2. xb5 g4
ing "Philidor" position.
3. hxg4
g5!!
With a little training you'll
Drawn
be able to visualize well in
Because after 4. gxf5 h3! 5. gxh3
advance whether you're get
xf5 the black W reaches h8 and
can't be driven out.
ting into Lucena or Philidor.
"Chess has three phases:

Thefirst when one hopes one has an
advantage;
The second when one believes one has an
advantage;
And the third when one knows one is
going to lose. "
-SavieUy Tartakower
15
Black must prevent the from

Lucena-type position
Pat: There must have been a
gaining a flight square (4... f6? 5.

Lucena who discovered this.
10 WfS).

Noah: A 15th century Span
5. e4!
.
iard, as a matter of fact. Actu

And this builds the bridge which
ally this notorious position
enables White's to escape.
was discovered by someone
For example, 5 ... h l 6. f7
fl t 7. wg6 gl t 8. h6 (with a
else and it was named after
threat of e5-g5) h l t 9. g5
him by mistake 400 years af
glt 10. g4 and queens.
ter he was dead. That's chess
for you.
White to play
Pat: The problem for me is all
I. 'lflh7!
h2t
these positions-Lucena, Phil
Black has no other counterplay
idor, whatever-look the
to prevent g7-g8=*.
same.
g2
2. 'lflg8!
3. g7
h2!
v-
@ I
"Bum all the books! Chess

should" 't be a science. "
-David Bronstein
Otherwise 4. h l, followed by
5. h7 and the i queens. For
example, 3 . . . g3 ? 4. h l f6 5.
h8 xg7 6. flt! Kg6 7. glt
and the falls.
This is the infamous "Lucena
Position. "
4. elt
16
'lfld7
"Chess Un't a science. "
-Henri Poincare
Chapter One
with 5. f5?? xf5t 6. Wxf5 '1t>e7

would even have lost.
M. Gurevich-Rechlis
Tel Aviv 1989
White to play
1. 1:!e5??
Of course, 1. g4, threatening
the unstoppable ( I... c6 2. g5
a4 3. Wg7) 2. WgB and 3. f8='it1,
would have forced resignation.
a4
1.
2. 1:!a5
1:!al??
And Black resigned when he
saw 3. a7t WdB 4. aBt Wd7 5.

WgB wins.
But 2 ... f4! would have drawn
(3. ctlg7 g4t 4. Wf6 f4t 5.
Wg6 We6) and bridge building
"Of my 57 years I have

applied at least 30 to
forgetting most of what I
have learned or read. "
-Emanuel Lasker
Noah: Once you learn the ba

sic principles it should sink
in. Then you just have to re
main alert.
In Diagram II you have the
embarrassing <.'xample of the
world's 12 th rank<,'d pla yer,
at the time, becoming con
fused over Lucen4l.
Ifthe a- t were offthe board
he would have seen the win
and won the game instantly.
Pat: But aren't there dozens of
other essential positions in
the textbooks you need to
know?
Noah: Only if you're a mas

ochist.
The vast majority of mate
rial you find in those "Every
Endgame Ever Played" tomes
is useful the way the stuff in
an almanac or library refer.
ence IS.
Pat: How's that?
Noah: You open them up when
you need to know, say, the
average annual rainfall ofBo
livia.
Pat: I thought the G Ms knew it
all by heart-even the rainfall.
Noah: Hardly. Once upon a
time Salo Flohr-one of the
finest endgame players of all
time-was about to adjourn in
a very favorable-looking po
sition against the great Jose
Capablanca.
17
"What are you thinking Noah: Rubbish. There's al

about? The position is ways been too much to learn.
drawn," Capa told his 26The point is you rarely get
year old opponent. "Look in punished for not knowing
the book!"
some obscure "basic" posi
Flohr looked and looked tion.
and didn't see any draw for
Here's something (Diagram
14) played by a modem world
Black. So he got upset, sealed
a 'lfl-move and rushed to his champion-who also asked his
hotel room to analyze.
G M opponent: "Why are you
But he still couldn't find a continuing to play in this
draw. He later admitted he drawn position?"
would have looked in "the
book"-but he didn't have one
with him.
Pat: But that was in the old
days, before databases and
all the junk they have now.
Way back then there prob
ably wasn't that much to learn
about the endgame and none
of the modem technology to
study it.
18
12
Flohr-Capablanca
Moscow 1935
White to play
1. e2
e4!
Now 2. Wf2 h4! 3. gxh4 f4 4.

exf4 Wxf4 will leave White with
only the h- ft s-a draw even with
out the black f- ft .
2. h3
d5!
3. f3
4. h4
e5
White counted on 2 ... We5?,

which allows zugzwangafter 3. Wf3
d5 4. Wf4 We6 5. h4!.
4.
5. \tif4
Drawn
Wd5
e6
White can make no progress,

e.g., 6. e4 fxe4 7. xe4 f5t 8. f4
Wf6.
Chapter One
Pinter-Kasparov
French League 1993

White to play
I.
2.
3.
4.
5.
6.
7.
8.
9.
10.
f3
g4
Wf5
e8
e6
we4
Wd5
e7t
f7
f2
Wd6
Wd7
Wd6
Wd7
wc7
Wd7
wc7
Wb6
h2
g3
Pat: Okay, so he knew it was

drawn and White didn't. Does
that mean I should just bring
lots of reference works to ev
ery tournament?
Noah: No. Having the right
book would only have saved
White from having to ad
journ.
Pat: You mean that aside from
wasting your time-and your
opponent's-it doesn't matter
whether you know the posi
tion on the board is a win or
a draw?
Noah: Most of the time? No.
Pat: But why?
Garry Kasparov
And White struggled another

17 moves before agreeing to a draw.
79
Noah: Because it doesn't tell

minutes to realize the best try
you how to win it. Take Dia to do that is 1. Wb4. That's
gram 15.
true whether it j a win or a
The books will tell you it's a draw.
win with the black W/e 7, but Pat: But there must be times
only a draw if it's on e6 or e5. when you really need to know
That's the type of impractical which it is?
information you find all the
time in texts like Fine's and
Paul Keres.' They give "ba
sic" knowledge a bad name.
Pat: Wouldn't a G M know in
stantly whether it's a win or
not?
Noah: Probably not. But that
would not affect his life at
all-unless he was hired as an
adjudicator.
Pat: But why not?
'n expert is a person who
Noah: A good practical player
has made all the mistakes
that can be made in a very
would figure out the only way
narrow field.
to win is to advance the it . It
-Niels Bohr
won't take more than two
"
20
White to play
1. \flb4
Not 1. t! d4 because 1... t! d8! 2.
t! xd8 (else the black W gets to the
c-file) Wxd8 3. Wb4 Wc8! draws
by using the opposition, discussed
later in this book (4. Wc5 tic7 5.
Wb5 Wb7 6. c5 c7 etc.).
1.
2.
3.
4.
5.
..
waS
\flb5
wa6
d4!
b8t
c8!
b8t
c8
If the Wi had been at e6 in the

previous diagram, Black could now
stop the plan with 5 ... We5, forcing
the t! to give up the file or the
protection of the ft.
But here the ft gets to the fifth
rank (5 ... w;>e6 6. Wb7 We5? 7.
t! d5t or 6... t! c5 7. Wb6! t! c8 8.
c5), reaching a better-known
"book" won position-not far from
Lucena.
Chapter One
Larsen-Torre
Leningrad 1973
White to play
1. e5t!
Now on 1... d7 White plays 2.
f3 and the winning plan of Wg45 and f5-f6 can't be stopped.
1.
2. Wd3!
3. el
Wf6
xf4
And wins because the black W

is cut otT, e.g., 3 ... tla4 4. c6 tla6 5.
tlc l ! tla8 6. Wc4 We6 7. tld l !
followed by bS-b6 and c6-c7,
leading inevitably to "Lucena."
Noah: Sure-when you can se

cure a book draw (if you're
losing) or a win by making
some major decision, like a
trade of pieces or a sacrifice.
Here's a typical example
(Diagram 17):
White was one of the half
dozen best players in the
world, at the time, and needed
a win to make sure of qualify
ing for the Candidates
matches from an interzonal.
Pat: He must have known
whether this was a win.
Noah: Yeah, but that didn't
tell him how to win it. He
actually played an error, 1.
h5?
Yet 1. e5twinseasilysince
the black W has to choose
between two losing sides of
the board.
In fact, White had missed

the same kind of transition
six moves earlier-and did it
again two moves after 1.
h5?
21
Pat: So you're saying I only

41's move away from White's,
need to know a handful of and puts his g on the e-file.
positions?
It's a useful position to know
Noah: Yes, because most of only because it goes against a
the time you can rely on gen crucial general principle in
g -endgames. Keep your
eral principles.
Sure, sometimes that doesn't as flexibly placed as pos
work. For example, Diagram sible.
18, which was analyzed by Pat: This principle is so crucial
Andre Philidormore than 200 that it doesn't work?
Noah: By putting the g on e l,
years ago.
Pat: Is this the famous "Philidor and giving up the chance for
position" I'm supposed to rank checks, Black violates
know?
the principle. But he draws
Noah: Almost. If it is Black's
because he can stop White's
move he draws by keeping only winning plan. Stopping
his g on his third rank until the plan was just more im
the pawn advances. 17zatthe portant here.
By the way, Philidor got the
"Philidor" you need to know.
Pat: So what happens on 1.
analysis all wrong-and con
'tf;f6 ?
cluded that Diagram 18 was a
Noah: After that, Black can forced loss.
draw only ifhe keeps his \t a
22
White to play
1. Wf6!
On 1. e6 Black draws with 1...
a6t! 2. Wf5 b6 and if 3. e6,
then 3 ... b l ! since the W cannot
escape from checks along the files.
1. . . .
el!
Not 1... a6t? because of2. e6,

threatening mate, and then 2 ... Wd8
3. h8t Wc7 4. Wf7 and White
reaches "Lucena."
And not 1... fl t? because of
2. e6 W-moves 3. h8t fol
lowed by We 7 and e6, again headed
to "Lucena."
2. we6!
4. e8
Otherwise the t can't advance

(4. a8 e2! 5. Wd6 Wf7).
4.
al!
And draws (5. b8 a6t 6.

Wf5 Wf7! or 6. Wd7 a7t 7. We8
Wg6).
Chapter One
5. wdl
6. cl!
7. b2
d3
hlt
el
Pat: Yeah, it always seems there

are tons of stuff like that in the
books. Do they ever happen
No better is 7.. . We3 8. c2.
in real life ?
8. d8!
And a draw as agreed after 8 .. Noah: All the time. A good
example of how knowing it
a e4 9. Wc l 'it'e2 10. 'it'c2.
helps was Diabrram 20. Both
players were strong GMs who
well knew the ubook" of Dia
gram 18.
But at crunch time in 20
only one of them remem
bered it. Black erred because
he didn't recognize a book
draw when he saw one.
Pat: I see-3. d5 is the same
as 1. .. e 1 in Diagram 18. All
that's happened is the colors
have been changed and
board's been reversed.
How often does a position
like that come up?
.
Dreev-Belyavsky
Soviet ChampionshiP 1989

White to play
1. h4!
Otherwise 1... Wd3 wins.
1.
xh4?
Not 1... Wd3 2. a xg5. But there

was an elaborate win in 1... a g 1 t
2. We2 d3t 3. wd2 a g2t 4. wd 1
g4.
2. xg5
3. d5!
c3
Drawing (3 ... d3 4. a c5t

Philidor).
3. . . .
4. e2
hlt
h2t
la
23
Noah: You'd be surprised how Noah: Sad but true. The point
often the possibility arises. In is that even walking "book"
that same Larsen game that encyclopedias can misuse the
we just looked at, White little that they need to know.
missed another winning tran
You need to not only know a
sition, in Diagram 21.
few key positions. You also
He could have sacrificed the need to be able to figure out
c- t to bling about something what to do in a slightly differ
almost identical to the Phili ent position.
dor position-except that
White, who clearly knew
Black's t! is so badly placed "Lucena," didn't look hard
enough to see a simple transi
that White wins.
tion to it.
Pat: Badly placed on c5?
Pat: I'll bet all this is a lot
harder than you're making it.
Noah: See for yourself. End
games aren't that scary.
Come back to the club to
morrow and I can show you
what you should be doing.
Interested?
Pat: Hmmm. Lemme think
about it.
24
Larsen-Torre
Leningrad 1973
White to play
1. g5!
Not 1 . a c7? which White
played.
1.
Rxc5
Otherwise 2. Wf6 and 3. a hSt

allows White to advance the f- t to
"Lucena. "
2. g6!
Black loses-because he can't

play 2 ... a f1 !, as in Diagram IS.
2. . . .
3. h8t
4. f6t
cl
e7
e6
6. f7
And wins (6... Rg 1 t 7. Kh5 Rh 1 t

S. Kg4 Rgl t 9. Kh3 etc.).
In which Pat learns about h-pawns and draw offers and

converting advantages-and why thefirst move many grand
masters make in the endgame is with their feet.
fffir;yi.
Getting Started
Pat: Okay, I'm back.

Pat: That's a no-brainer.
Noah: I thought you might be. Noah: Not true. Many players
Pat: Butjust to satisfy my curidon't realize or fully appreci
osity. Suppose I wanted to ate that.
really get into endgames.
In the endgame there are
Where would I start?
big differences in thinking.
Noah: You begin with this not
You have to readjust.
to-be-forgotten warning:
Pat: How?
Noah: You have to start think
An endgame is not a
ing differently about speed
middlegame.
and space, not to mention
pawn structure, and safety.
Your whole sense of how to
evaluate a move or a position
changes instantly when the
fls are removed from the
board.
Pat: How do the G Ms begin
thinking differently when the
endgame begins?
Noah: Some of them take a
walk.
Pat: They take a walk?
26
Karpov-Timman
Candidatesfinals 1989
White to play
1. f4?
A good positional move in the

middlegame-but an error here.
With 1. \1?fJ, threatening to take
control on e4, White has good win
ning chances (1... f5 2. <3;e3 e5 3.
4)b5 Wf6 4. \1?d3 and \1?c4).
1.
f5!
A positional lemon-in the mid

dlegame. The game was drawn
after 14 more moves beginning
with 2. 4)a4 Wf6 3. 4)c5 4)b6 and
eventually ... e5.
Chapter Two
>
''--"""'''''
f"-"'
In the Middlegame
). .
Haste makes waste. Repeating the

position is good technique. Losing a
tempo often helps.
Space counts. A cramped position is

usually a bad position.
Control of more space than your

opponent is relatively unimportant.
J"' ",
:,
t. .
'
ry.
I
In the Endgame
Speed counts. Often the winner is the

player who gets there "fastest with the
mostest."
___________,-__
w .
W safety is a major priority.
The easiest way to win is to accumu

late advantages.
Center its are more valuable than

wing its.
When deciding whether to exchange

something, masters try to trade bad
pieces for good ones, to avoid losing
time, and to exchange its when that
opens lines for their pieces.
-__-
f
i
<efts are generally safe. '3; activity is
more important.
The easiest way to win is to convert

one advantage to another.
Wing its-particularly if they are

passed-are often more valuable be
cause they're easier to queen.
Trading priorities depend on whether

you're winning or losing. Usually the
player with the edge wants to trade
pieces but not its, and vice versa.
27
Getting Started
Noah: It's an old Russian idea,

suggested in the 1930s by a
master named Sergei Bela
venets. He said that as soon
28
as you trade iiYs you should

stand up and walk around the
playing room to calm your
nerves and clear your head.
In this game (Diagram 24),
White had hardly sat down
when the iiYs went off. But he
needed to reorient his think
ing after 9. iiYxb6. A stroll
then wouldn't have hurt.
Pat: Okay, suppose I've traded
iiYs and had my walk. What
then?
Noah: First, figure out who's
better.
One of the things that dis
tinguishes an ending from a
middlegame is that it's usu
ally much easier to see in an
ending whether you're play
ing for a win or just trying to
draw.
Pat: For sure.
v.
Kovacevic-Smirin
Zagreh 1993
1. d4 4)f6 2. g5 4)e4 3. f4
c5 4. f3 *a5t 5. c3 4)f6 6.
4)d2 cxd4 7. 4)b3 *b6 8.
*xd4 4)c6
9. *xb6
10. a3?
axb6
A low priority move. Better was

10. e4.
10.
d5!
1 1. c7?
Playing for advantage in what is
now an inferior position. Better
was 1 1. ctld4.
1 1. . . .
e5!
12. xb6 d4
And Black soon had an over
whelming initiative ( 13. cxd4 .Q.e6!
14. ctlc5 ctld5 15. ctlxe6 fxe6 16.
.Q.c5 ctlxd4! 17. .Q.xd4 exd4 and
... ctle3/ ... .Q.b4t.
Chapter Two
\t1b7 cct>e6 14. .Q.b6.
Kmic-Flear
Wijk aan Zee 1988

White to play
1. Draw accepted??
White is winning:
1. f4 cct>f7 2. e5 invading, or
1... 4)c8 2. g5 cct>f7 3. Wh6 and
wins by zugzwang.
For example, 3 ... 4)b6 4. Wh7
4)c8 5. a4! a6 6 . .Q.c5! or 5 ... 4)b6
6. as 4)c8 7. h4!.
Black does better with 3 ... a6 4.
.Q.c5 cct>f6 but is still losing: 5. h4
cct>f7 6. h7 as 7. a4! f6 8. Wg8
Wf5 9. Wf7 g4 10. e8 xh4
11. d7 Wg4 12. \t1xc8 Wf5 13.
"It is a well known

phenomenon that the
same amateur who can
conduct the middlegame
quite creditably, is
usually perfectly helpless
in the endgame. "
-Aron Nimzovich
Noah: Yet it constantly shocks Noah: G Ms accept draw of

amateurs to see how often fers when they see how they
masters accept draws in win can lose, not when they're
ning positions, like Diagram unable to see how to win.
25, or resign in drawable ones. Pat: Okay, so you evaluate the
Pat: What should White be position. Then what?
Noah: Then you try to under
thinking in that position?
Noah: That he has a very good stand what kind of advantage
1t and Black has a limping it is.
Pat: Usually that's obvious.
fl.
That if he wins either of the Right?
black pawns he'll win the
game.
That he has excellent
chances for imposing zug
zwang.
And, most important, that
with no visible Black count
erplay, it costs White abso
lutely nothing to play on.
Pat: But don't you have to
know what you're doing next
to refuse a draw offer?
29
Getting Started
retreat the like that and let

Noah: Sometimes yes, some
the Black tJ in.
times no.
For instance, with the Ex Noah: Yes, but very logical. A
is superior to a 1t only
change for two fts, White
can't be entirely sure he has when it can act like a .
That is, only when it has
any real edge in Diagram 26.
Black's King is closer to open files. To make such a
White's W-side fts and file-and create a winning ft
Black's can attack the 'i!1- White had to force a trade.
side fts.
Pat: I sense a "but" coming up.
Noah: But White has the Ex
change. A basic rule of thumb
"Fight the enemy
IS:
When you're up the

Exchange you want to
trade a pair of : s.
That's because a trade mag
nifies the mismatch between
the remaining v. piece.
Pat: It's weird that White can
30
with the weapons

he lacks. "
-Field Marshall
Aleksandr Suvorov of
the Russian Imperial
Army
Hmelnicky-Romanishin
Herson 1989
White to play
1. hh 1!
lflg5
Or 1... b6 2. axb6 xb6 3.

hb l c6 4. b8 and the a- i
falls.
2. hbl Wf4
No better is 2 . . . xb I 3. xb l
Wf4 4. xb7 xfJ 5. a7 Wxe4
6. xa6 Wd5 7. a8 g5 8. a6 and
wins.
3. xb5
4. dl!
axb5
Better than 4 . b 1 b4 5. cxb4

WxfJ with drawing chances for
Black.
Now the creates a winning
passed a- i -4 Wxf3 5. d5 b4!
6. b5! (not 6. xd6?? b3t) b3t
7. \t>b2 Wxe4 8. xb7 Wd5 9. a6
Wc6 10. b5! and 11. a7.
Also 7... g5 8. xb7 c5 9.
xc7 Ae3 10. xc4 Wxg4 11. a6
Wf4 12. c7 g4 13. a7 xa7 14.
xa7 Wxe4 15. g7 f5 16. Wxb3
Wfa 17. c4 and wins.
Chapter Two
Arulaid-Gurgenidze
U.S.S.R. 1955
White to play
This position was adjourned and

White resigned-a blunder since he
can draw by threatening mate with
1. Wd6! and checking.
For example, 1... WcB 2. a c I t
Wb7 3. a b l t Wa6 4. Wc6! (threat
of 5. a a l mate).
Then 4... Wa5 5. Wc5 Wa4 6.
Wc4 Wa3 7. Wc3 Wa2 8. a fl ! and
since B ... g2?? loses the f- i with
check, Black has nothing better
than 8... h5 9. Wd3! and White
draws with We2-fJ.
Pat: What happens ifl'm worse takes to win.

when the endgame begins? Pat: You mean like winning a
i or something else?
Noah: Same sort of thing. You
need to know what your Noah: No, you generally don't
opponent's advantage con need to win anything in an
ending.
sists of.
But you also have to recog Pat: You don't?
nize where your own assets
lie-and not to underestimate
them.
Pat: You got an example?
Noah: Sure. In Diagram 28
White was so afraid of the
black is that he completely
overlooked the power of the
strongest piece on the board,
his own .
Pat: Okay, suppose I know I
have the edge and understand
you generally don't
what kind of edge it is. What
need to win anything in
next?
an ending.
Noah: A good next step is to
GMTaU
figure out what specifically it
31
Getting Started
lng.
Noah: No. In the endgame the
side pressing for a win usu Noah: Actually he eliminates
ally has some kind of advan
the only good black piece left
tage already.
to stop the c-pawn.
What a player needs to do, After t! xe6, Black must try to
blockade with his t! -and a
like White in Diagram 29, is
to convert one advantage into
t! , as you'll leam, is a terrible
a more significant one.
blockader.
So once White knew what it
Pat: Seems like Whitejust gives
away the Exchange for nothtook to win, the rest was easy.
A winning conversion.
Pat: But what if you're playing
for a draw?
Noah: Then you need to know
what it will take for your oppo
nentto win.
Don't try to figure out what
it takes for you to draw.
Usually there is no such
thing.
Pat: I'll believe that. Okay,
what's happening in Diagram
30?
32
1 1. c3 and 12. Wb6.

Too late is 1 1... xc7 12. xc7
xg3 13. xg7t Wxf4 14. h7.
Gelfand-Bareev
Linares 1992
White to pla,
1. e4!
Threatening 2. xe6!, e.g., 1...
h5 2. xe6! fxe6 3. c5 b4t 4.
Wa3 b8 5. c6 g5 6. c7 c8 7.
Wa4 and xa5-b6-b 7.
1.
b4t
2. 'It'a3
g6
Or 2 . . ,!td7 3. e7.
3. xe6t! fxe6
4. c5
And wins, e.g., 4 ... f6 5. c6

b8 6. c7 c8 7. a4 e5 8.
xa5 d4 9. c6 e3 10. f4 f2
Chapter Two
30
Marin-Minasian
Debrecen 1992
White to play
1.
2.
3.
4.
a4!
as
a6
a a3 ??
c6
c5
c4
White wins with 4. d3! !, e.g.,

4 ... c3t 5. xe3 d I=* 6. xdi
xd I t 7. d2 and 8. a7.
Or, 4... f4t 5. xe3 xd3 6.
<it'xd2.
4.
c3t
5. a xc3t xc3
6. a7
c2
Drawn
Because the *+! ending is

dead after 7. a8=* d I =*t 8.
xe3 *gI t 9. e2 *xh2.
Noah: A strange but not at all

rare misunderstanding.
White may have asked him
self at the diagram "Am I
better and, if so, why?"
The answer would be: "Of
course, I am-but because of
the a- i, not so nluch be
cause of the Exchange." That
would have given him the
winning plan.
Pat: I guess that makes sense.
Noah: But when it was crunch
time, at move four, he tried to
keep a !! he didn't need
instead of eliminating is
the only counterplay Black
had.
Pat: It sure must help if you've
won this kind of position be
fore.
33
Getting Started
Noah: Listen, Pat, it helps more

if you've lost this kind before.
Now, take Diagram 3 1 .
What would you do here?
Pat: I'm not sure. But 1. !! xc3
looks right.
Noah: Why?
Noah: Because it eliminates a
it -and a pretty dangerous
it, it looks like. And it avoids
a !! endgame after ...1t xe4.
Noah: When you've had more
experience you'll know 1.
!! xc3 is a blunder. Trading
!! s only enlarges the advan
tage of the remaining 1t over
the 4).
Pat: I guess I see that now. The

c- it isn't running away and
White's is much more ac
tive. But what about the !!
ending?
Noah: With time you'll also
appreciate how notoriously
drawish !! + it endings are.
Pat: What if you haven't a clue
Shirov-Karpov
about what's going on in a
Moscow 1992
White
to play
position?
Noah: Then a good way to
1. <&>e2!
start-after you've had your
Not 1. xc3? xc3 2. 4)xc3
walk-is to figure out what 1txc2 which leaves Black with ex
pieces you want to trade and cellent winning chances.
For example, 3. Wd2 1tf5 4.
which you don't.
we3 1th3 5. Wd4 1tg2 6. f4 WfS 7.
Pat: For example?
e5 e7 and with 8 . . . f6t Black

is making progress.
1.
2. d3
"It makes no difference whether you

win or lose-until you lose. "
-Anonymous Loser
34
g7
g5
Or 2 ... a4 3. 'itxc3 1txe4 4.

fxe4 cct>f6 5. f3t with enough
counterplay to draw.
3. <&>d4!
4. fxg4
5. xc3 !
g4
hxg4
xc2
Chapter Two
Noah: For example, Diagram

33. White concluded that he
could draw if he managed to
trade the fls.
Pat: He was wrong?
Noah: No, the minor-piece
endings is drawn because
White can liquidate the fI..
side is and keep Black from
Vaganian-Lautier
creating a passed f- i .
Manila 1990
White to Pill,
What he overlooked was
that Black can force a \fl+ i
1. '/tc3 ??
ending-which was quite lost.
Passing, such as with 1. *a6
,tf3 2. *a5!, makes it hard for Pat: So some trades are very
Black to make progress.
good . . .
But White counts on 1... *xc3t Noah: . . . and some trades are
2. 'lilxc3 ,ta4 3. d2 and 4. d4
just awful. Remember what I
with a draw.
said about the importance of
1. . . .
a4!
And Black wins, e.g., 2. d2
*xc3t 3. Wxc3 ,txb3 4. Wxb3
Wg6 5. a4 h5 6. b5 'iftg4 7.
Wc6 xf4 8. 'iftxd6 'ifte3.
"Exchanging is the
soul of chess. "
-Georg Kieninger
transitions.
An example of good trades

on both sides is Diagram 34.
35
Getting Started
Pat: What's happening here?

Noah: Your basic plus-over-
36
equals () edge for White. He

has the better A and his op
ponent has the weaker tar
gets, at a5 and c6.
If he can trade his al- for
the b8- , his remaining
pieces will be much better
than Black's, and he can take
his time with a plan of, say,
creating a passed a- ft , with
b2-b4.
Pat: But Black makes his own
good trade, a bad piece for a
good one-even if it louses up
his ft structure.
Noah: Not so loused up. In the
end the f5- ft is not much
worse than the ones at g2 and
b2. A draw makes sense.
Pat: You coulda fooled me. I
lose positions like the ones
grandmasters agree to draws
In.
White's best chance was 4. f1 !.
4.
gxf5
White can make no progress,

e.g., 5. c3 (intending 6. f3)
b8t 6. Wa3 g8. Mter 7. g3?
xg3 Black is even better.
A draw was agreed soon after 5.
c2 f4 6. b3 h5 7. d3 b8.
Hodgson-Oll
Dos Hermanos 1992

White to play
1. a3 !
A good idea (2. b3).
1. . . .
g6 !
Black also sees a good trade,

... f5!'
2. b3
xb3t
Black can avoid the trade with
2 . . . e8, but after 3. Wd2 (to stop
3 . . . el ) White will have a big
edge with 4. b7 0r 4. b6(or 3 . . .
c7 4. bb5!).
3. xb3
4 .Q.xf5
.Q.f5!
Chapter Two
Burnett-Kaidanov
U.S. 1992
Black to play
1.
e4!
Not 1... !! xd I ? 2. !! xc6 with

very slim winning chances.
2 .Q.e2
3. a2
4 .Q.f1
d4
a el
gS!
Black prepares to create a passed

e- t { .. f7-f5-f4}.
.
S.
6.
7.
8.
9.
d2
a e2
xel
wgl
wg2
f3
fS!
xel t
f3t
f6
Noah: With experience you'll with few winning chances.

learn what a drawable posi Pat: And a trade of just t! s?
tion looks like. Often the ex What happens then?
act same position-same is, Noah: Let's work it out.
same <it> positions-is a win
Having is on only one side
of the board makes the 4)'s
with 4)s on the board or *s
on.
inferiority disappear. In fact,
But it may be a dead draw the 4) turns out to be a super
with the addition of a pair of piece on d2.
t! s or a pair of ..Q.s. Or vice
White had nothing better
Now 1 1 ... xfl, creating a
than a t! exchange-but then
versa.
w+ t ending, would win easily.
Pat: How do you know the he was simply lost.
1 1 .Q.bS
f4
12. gxf4
gxf4
vice from the versa?
And Black won after 13. h4 f3 Noah: By looking ahead and
14. h3 h5 15. Ac6 <3te5, e.g., 16.
seeing which piece situation
Ae8 e3 17. fxe3 fxe3 18. Ab5 e2!
would be the easiest to win.
19. Axe2 gl t.
In Diagram 35 Black can
see that a trade of all the
pieces leads to a <it>+ i end
ing that should be won.
But without a *-side pawn,
a trade of minor pieces only
gets him into a t! + i ending
37
Getting Started
Pat: You're talking a lot about and . . . g5 to create a passed

pieces, but what always con
i.
fuses me in endings is where Noah: With a likely draw. Trust
to put my is.
me.
I mean, what's the deal with
As the game goes, Black
the h- is?
forces a favorable trade of is
In just about every G M
because of the threat of 9 ...
game I've ever seen, it seems h3.
White plays h2-h4! ? and
In other games with a simi
Vadasz-Rogers
Black replies ...h 7-h5!? .
lar i structure you may see
Kragujevac 1985
Black to play
Noah: There are reasons for Black winning by attacking
that. For one thing, it makes it the enemy \t>-side with ... \t>g4
1.
gS!
harder for your opponent to
and ...hxg3.
If Black had allowed 2. h4 his
create a passed i , as Black Pat: Is there an ideal endgame advantage would be microscopic.
2. f3
a4
does in Diagram 37.
i formation?
3. e2
Pat: I see. If White had a i at Noah: That usually depends
White has nothing better. On 3.
h4 then Black would have
on whether you're playing to
h3 Black creates a passed i with
had to trade two sets of \t>
win or draw.
3 ... f5 and ... h4.
side is, starting with . . . f6
For example, in Diagram
g6
3. . . .
39 Black ends up with the
fS
4. d3
S. c2
g4!
exact same \t>-side structure
a3t
6. c3
as his opponent-and loses be
7. d4
gS
cause of it.
38
8. e2
8. . . .
9. gxh4t
h4!
Otherwise 9... h3 followed by

.. a6-b6-b I -g l -g2!
9. . . .
xh4
Black won after 10. e5 cctlg5

I I. cctld4 a4t 12. d3 cctlf6 and
...cctle5.
Chapter
Vaganian-HeUers
New York 1990
White to p14y
1. h4
2. 3
3. e3
ab2
we6
<tIi'f6
Passing is Black's only policy.
4. 3 1
5. f4
ag2
ab2
There followed some more ma

neuvering as White prepared to
push his passed ! : 6. b6t 1;g7
7. b4 b3 8. b8 <1ftf6 9. b6t
<1ftg7 10. 1;e4 bl 11. bS.
Two
Pat: Let's do this one slowly. It

seems like I've seen positions
like this a jillion times.
Noah: Okay. White starts by
restraining Black's g- ft with
1. h4. That also creates a
chain, so he can protect all his
W-side pawns with a W at e3.
Pat: Why does he take his time
f6??
1 1.
in pushing the b- ft ?
A blunder. More passing, with
1 1... b3 12. b8 Wf6 and ... We6- Noah: Because that would al
low Black's to cut off his W
f6 would draw.
12. ab7t h8
by going to the sixth rank like
13. ab8t <tIi'g7
on 2. b4? b3!
14. b6
ab3
Pat: Okay, so White can move
15. d5!
a xm
his <Ii> to e3, and then push his
16. c4
Noah: Because his only win

ningplan would have been to
advance the ft to b6 and
then run his <Ii> to defend it.
Black's would have eaten
most of the <Ii>-side in the
meantime, and he would then
draw by sacrificing his for
the b - ft j ust before it
queened.
Pat: So he lost because ..
Noah: He gave his opponent a
safer and easier winning plan.
Pat: But what if he... ?
Noah: That's enough for you
to absorb in one day. Ifyou're
not sick ofendgames yet, we'll
f- t .
And Black loses because his W
start again tomorrow. Okay?
plays no role: 16 ... fl 17. b7t Noah: That allows his W to go
Wh6 18. d7 b I 19. Wc5 c 1 t
to f4 and protect his entire W Pat: I guess so.
20. Wd6 d lt 21. rJic7 c 1 t 22.
side while the b- ft advances.
Wd8 a b l 23. a d6 1-0.
But it shouldn't have been
enough to win.
Pat: Why?
39
Getting Started
40
In which Pat learns that all rules have exceptions, that some
rules are exceptional-and that no rule is a substitute for
thinking.
Read and Forget: Actual rules laid

down in "Basic Chess Endings"In
+2 i vs. + i : "when White has

is on the 1t - and 4)-files, with only one i
more, if Black's i is on the 1t -file, he can
usually draw but if it is on the 4)-file he will
as a rule lose."
"With 3 connected passed is (against a 1t)
a win is possible only if all the is can
succeed in crossing the 4th rank (except for
certain special cases)."
In + t! +2 connected
i vs. + t! : "if
neither i has reached the sixth rank this is
always a draw, but if one is on the sixth and
the other is on the fifth, a win is possible
provided there are no t! is."
-
(ifff.lr;}l
Rules
Pat: I haven't forgotten what

we talked about yesterday
yet.
But today I'd like to move
on to another reason I hate
endgames-the rules. There
seems to be a million differ
ent rules you have to memo
nze.
Noah : You mean like
"*+<it>+ it beats *+<it> ifit's
a c- it that reaches the sixth
rank during a month with an
"r" in its name?"
Pat: Yeah, that type.
Noah: Here's a rule of thumb
on rules-The more specific
the rule, the more exceptions
there will be-and the more
useless it will be to you.
Pat: I'll buy that.
Noah: The best rules are the
simplest, like "Passed Pawns
42
Must be Pushed," a maxim

so ancient no one remem
bers where it came from.
But we don't forget that rule
because it works, as Diagram
41 shows.
Pat: Lemme check this out.
White's a it ahead but it's
doubled. And Black has a
good .Q. against the 4).
Noah: But Black has the
weaker its. If White had
pushed the c- it before move
six the win would have been
routine.
Pat: I thought speed doesn't
count in endings.
Noah: Don't think of these
kinds ofpositions as it -races.
Pushing the c- it is more like
converting one kind of ad
vantage into another.
5. . . .
a4!
And Black managed a draw af

ter 6. bxa4 axa4 7. Wd2 aa7 8.
\t>d3 a c7 9. f5 \t>xg5! 10.
xg7t \t>f4 because his W is
active and White only has two is
left.
Kamsky-M. Gurevich
Belgrade 199 1
White to play
1. g5?
With 1. c4 bxc3 2. bxc3 White
creates a winning passed i . In
stead he plays to win the a- i .
1.
2. hxg5
3. 4)c4
hxg5
g6
!! d7!
Now on 4. a xa5 a d5! 5. a xd5

1txd5 and it will be Black who gets
a passed i rolling first.
4. e3
5. cl?
!! a7
White is still winning after 5 . c4!
Chapter Three
Drawn
Pat: What other kind?

It's only a draw because the Noah: Closeness of a i to its
remaining white l s are eliminated
queening square. The reason
after 4. WXaS hxg4 5. fxg4 We4 6.
is explained by another rule:
c 1 g5! and 7. .. Wf3.
is Increase in Value
as they Advance.
D. Gurevich-Franzoni
Lucerne 1989
White to play
1. dle3?
Going after a i that has no

significance. No better was 1. as?
1txaS! 2. 1txaS f4 which wins
both remaining white is.
But winning was: l. 1taS! 1tf2
2. 1tc3t Wf4 3. as, e.g., 3 ... Wxf3
4. a6 Wxg4 5. 1tb4 and 6. c5.
1.
g6!!
Now 2. as xaS! 3. xa5 h5

draws, as in the game.
2. dlxg5
3. as
h5
dlxaS!
fts Increase in
Value
as they Advance.
In fact, a i on the sixth

rank is often twice as valu
able as one on the fourth,
particularly in minor-piece
endings.
Pat: Why particularly those?
Noah: Because unlike heavier
wood, a -'l. or cannot con
trol more than one or two
squares in the path of a i at
a given time.
Pat: So?
Noah: So in Diagram 42 the
only square in the a- is' path
that the -'l. controls is a5, and
once it is kicked away, the i
reaches a6 quickly.
Then all White needed to
win was to drive the -'l. off the
g l-a7 diagonal and promote.
The increased value that a
i picks up as it advances is
illustrated best by your old
friend, "the square."
Even with several minor
pieces, you have to keep in
mind whether your <it> is in
the square to catch a passed
i.
Pat: Yeah, the square. That's
one ofthe first things I learned
about endgames-and one of
the few I remember.
43
Rules
"The lord ofthe endgame

is the passed pawn. "
-Cecil Purdy
44
Noah: And you'd be surprised

how much it figures in grand
master games.
It comes up in many forms,
like Diagram 43. Because
Black's rtJ is one step out of
the square of the a- ft , White
can win with an amazing two
piece sacrifice.
Pat: Yeah, pretty amazing. Es
pecially since White can win
a ft , at least temporarily with
1. c4.
Noah: But that allows the en
emy rtJ into the square and
complicates the win.
As it goes, Black can't ac
cept the second piece sacri
fice and the rest is easy.
Pat: Okay, so, fts increase in
value as they advance" is a
good rule to know. But how
do I use it?
Karpov-Hansen
Biel 1992
White to play
1. 4)xb5!!
Too slow is 1. c4 Wd7 2. cxb5
cxb5 3. xb5 1a,f8.
1.
2. xb5
cxb5
And 2 . . . axb5 loses to 3. a6

because Black's W is out of the
square and 3 . . . e6 is met by 4.
b6.
So Black had to try something
like 2... g5 but resigned soon
after 3. xa6 ge6 4. c4 Wd8
5. a6.
Chapter Three
Wf6 6. Wg4 and the g- ft will fall

after 7. Wg5 and 8. a6t.
Noah: Once you realize the

is and, finally, win with the
power of a passed i you'll e- i .
!
be alert to the opportunities Pat: I like general rules like
:
of trading into a won game, is increase in value ... " Got
c6
6. wg5
like Diagram 44.
any more?
7. wxg6 Resigns
Because of 7... Wb6 8. a al Pat: This looks confusing. Noah: You're probably famil
a xa7 9. a xa7 Wxa7 1O. Wf6 wb7
Everybody's queening.
iar with this one:
I I. Wxe6 and wins.
Noah: It's actually simple once
you see that White doesn't
Trade Pieces When
have to promote immediately l'Ou're Ahead Material,
but can get a paralyzing
Trade Pawns When
position with 1. xb2 and 2.
You're Behind.
Trade Pieces
a2.
When YOu're
Since White will win any Pat: I've heard something like
Ahead Material,
W + i ending arising from it. What's the reasoning?
Trade Pawns
... Wb7 and ... xa7, Black
can't do anything but pass
When YOu're
with his tt.
Behind.
Pat: But how does White queen
the a- i ?
Noah: Doesn't need to. He'll
run the black W and is out
of moves, then pick up both
: :::
Psakhis-Vaganian
Rostov on Don 1993
White to play
1. Rxb2!
If you're willing to calculate,
you'll find: 1. a xh8 b l=* 2.
a8=*. White wins because Black's
checks stop after 2 . . *f1 t 3. Wg3
*gl t 4. *g2! or 3 ... *e l t 4.
Wh2.
But not: 1. a8=* b l=* 2.
xb l ? xa8 which is drawn.
.
1.
2. a a2!
3. a a4
.
a a8
wg5
Wf6
Or 3 ... Wf5 4. aSt e5 5. e4t
45
Rules
Noah: Remember what I said Pat: That generally you need

about what it takes to win?
to promote a i .
Noah: Correct. Well, the more
pieces you trade, the less en
emy resistance to keep a i
from promoting.
Pat: And the more is you
trade?
Noah: The fewer is that can
be promoted-and the fewer
targets to go after.
In Diagram 45 Black, even
though two is down after 1 ...
g6/3... exd5, has drawing
chances. But he has none af
ter it becomes just ..Q. v. 4).
Pat: The 4) sure looks sick
after 5. <tfj>xf3.
Noah: And White can pro
mote either on the 'itt-side or
W-side.
That's another unique thing
about endgames.
46
4. t3 !
5. xt3
xt3
White won after S... weS 6.

We3 gS 7. 1t c2 h6 8. 1t e4! a5
(else the a- t advances) 9. Wd3 hS
10. 1tg2! (pass) Wd6 11. ",d4 etc.
Kindermann-MueUer
Bundesliga 199 1
Black to play
1.
4Je7?
Black can afford to give up a t

in order to obtain counterplay with
1... g6 2. a xh7 a f5!, threatening
3 ... f4t.
Mter 3. 1txd5 exd5 Black, with
his own passed t , has excellent
drawing chances.
2. b6
c6?!
He should keeps a s on the

board with 2 . . c6 and possibly
. . . d4.
.
3. xc6
4Jxc6
Chapter Three
Pat: What's that?

Noah: Since the basic winning
technique is to promote a i ,
the defender can usually draw
ifhe eliminates all the enemy
is.
That's what saved White in
Diagram 46-and should have
cost White a win in Diagram
Karpov-Anand
47.
Roquehrune 1992
Black to play
Since is increase in value
as they advance, Black's f- i
1. . .
xg4?
is decisive on 12. But it turns
With 1... f2 Black wins quickly:
out to be only a target on f3.
2. 1te2 .Q.xg4! 3. 1txg4 fl=* or 3.
1tfl f4, etc.
2. xf3!
And White forces a draw after:

2 ... .Q.xf3 3. 'lWc4 and 4. xb4 or,
as the game went, 2
.Q. e6 3. wc2
<1fld4 4. b2.
.
"The older I get,

the more I value
pawns. "
-Paul Keres
47
Rules
Pat: The first one looks pretty

easy but I don't know about
47.
Noah: It's a classic example of
48
trying to win an endgame with

middlegame moves.
Instead of taking his time
and carrying out a slow-but
unbeatable-plan, White sim
plified with a flashy tempo
rary sacrifice of a i .
Remember, is are cheap
when you have eight ofthem
but prohibitively expensive
when you have one or two.
Pat: Is this trading rule always
true?
Noah: No, like most rules, it is
only generally true. Sometimes
there are higher laws.
Pat: Like what?
Noah: Like Fine's Law:
If you are one i ahead, in
99 cases out of 100, the game
is drawn if there are is on
only one side of the board.
Pat: And is that true?
bxc4 7. 4)a4 Wf6 and ... g6-f6

(but Black blundered later and lost).
Lautier-Grunfeld
Palma de Mallorca 1989

White to Pilly
1. g4??
White should win with 1. h4!,
sealing the W-side, followed by
bringing his to c5.
1. . . .
2. hxg4
hxg4
<alh6!
Much better than 2 .. fxg4 3.

4)e5. Now Black's W gets to g5.
.
3. gxf5
4. g3
gxf5
g5
And White, with only one it

left, will allow Black to draw with
...4)c4. E.g., 5. 4)c5 4)c4 6. 1a,xc4
Chapter Three
Speelman-Short
1.
!J,.c7?
With 1... ab8! Black should

draw.
2. f6
3. f7t
4. xh7
Forexample, 7. . xd4 8. exd4t

and 9. a a7 wins another i or
trades as.
Or, if 7... a f6, White can also
win with 8. 4)fJt We6 9. ag7
1a,d8 and now 10. Wi'e2 followed
by 1 1. g4 and Black's g- i eventu
ally falls.
.
Candidates Match 199 1

Black to play
e5
Wd6
And White won soon after 4

,a,b6 5. 4)c2 ae6 6. an We5 7.
4)d4.
Noah: Maybe not 99 out of Noah: But as played, Black

100, but in an awful lot of leaves himself with two weak
situations. That's why Black
its and an inferior .
loses in Diagram 48.
That leads to another rule of
The minor-piece endgame thumb on rules:
with only one passed it , at e3
after 1... a b8!, is drawable
Every rule
because Black's remaining its
has exceptions.
are hard to attack thanks to
Fine Law.
Pat: I was afraid you were
White would have a hard going to say that.
time trying to promote his e Noah: It's true. And some
pawn after 2. a xb8 and 3. times in chess it seems that
4)xd5.
the exceptions outnumber the
Pat: But?
rules. Just like in life.
Pat: Uh-huh.
Every rule has

exceptions.
49
Rules
Noah: Some have more ex
ceptions than others, like
Capablanca's RulePut Your Pawns on the

Opposite Color of Your .Q..
In Diagram 50 Black threat
ens to put aU his .fts on light
squares and draw by keeping
the white W at bay and by
blockading the 'if1-side .fts.
But White beats him to it,
and eventually wins by pick
ing up the Black .fts on dark
squares.
Pat: But you were talking about

"Put Your Pawns on
the Opposite Color of
Your Bishop.
"
-Capablanca
exceptions. Why isn't
Capa
blanca '5 Rule always true?

Noah: Because there are al
ways going to be special
cases-like when creating a
passed .ft or locking out the
enemy W-when "bad" .ft
moves work.
50
BJatny-Adams
Adelaide 1988
White to play
1 . h3!
Not I. Wf3? e4t 2. we2 g4! and

... Wd5-c4 blockades.
1.
Wd 5
f4
2. g4!
.Q,.f8
3 . .Q,.f6
Or 3 . . . Wc4 4 . .Q.g7 Wxc3 5.
.Q.xh6 .Q.e7 6. b5 and wins.
4. f3
The game ended soon after 4
e4 5. fxe4t wxe4 6. b5 Wd5 7. b6
wc6 8. .Q.d4 .Q.e7 9. c4 .Q.d8 10.
eS and Wf3-e4-f5.
Chapter Three
3. c5!
4. b6
5. g5!
bxc5
c8
Pat: Can you think of an ex Noah: True enough. But the

51
ample?
proper winning plan involves
Noah: For example, in Dia creating two passed is at b6
The point.
gram 51, the chromatically and h6, and the only way to
5.
hxg5
correct moves, c4-c5 and g4- do that it to make White's -'t
f6
6. h6
g5, would allow Black to temporarily into a very bad
7. c4
Wg6
8. xe6! b7
blockade the position and -'t by first putting the is at
Of course, not 8 ... .Q.xe6 9. b7.
b5 and h5.
draw.
9. d5
c8
Pat: Even though he would be
This shows how endgame
Sulskis-Slekys
xb7
10. b7
making his own -'t bad and theory changes: Howard
Lithuanian Championship 1994
1 1. xb7 c4
White to play
giving White protected Staunton, in the best-selling
12. c8
'1f(xh6
chess book of the 19th cen
passed is?
13. g4
Resigns
1. b5!
In view ofthe unstoppable '1t'e3tury, said if you have the infe
Not 1. cS because L. bS! locks
d2-c3
and
wins.
rior game you should put your
the '/!'1-side.
Now 1.. . .Q.d7 allows 2. c5! dxcS
is on the same color as your
3. b6 .Q.c8 4 . .Q.c4 followed by S.
-'t -so you could protect them
Other Bits of
g5t, which trades a \tl-side t for
much better.
Vintage Advice
the black e6- t. Mter .Q.xe6 White
wins with .Q.dS! and b6-b7.
1. . . .
2. h5!
We7
Here 2. gS? hS! seals the '1t'

side.
2. . . .
d7
"Tofree your game,

take offsome ofyour
adversary's men, if
possiblefor nothing. "
-Joseph Bertin, 1735
"One Knight, at the end ofthe game,

is generally superior to one Bishop. "
-Howard Staunton, 1847

51
Rules
to Siegbert Tarrasch. It's not

a bad rule-as long as you
realize it doesn't work a lot
All E! + t Endings are
of the time.
Drawn.
Atleast they always tum out 52, even at the cost of a t ,
that way when I'm one or two White would have won the
ts up.
+ t ending after 1... -'ixe5.
Noah: We owe that bit of irony Pat: But he won in the long
run ...
Noah: . because he still man
aged to trade pieces and cut
the black <ct> off along the c
file.
White's tl and b- t were
then a mismatch for the en
emy , and Black resigned
when he saw he'd have to
surrender it for the t .
If you're looking for useful
rules, I'm sure you can cre
ate a few of your own.
Pat: Like what?
Pat: I just remembered a rule.

I heard a long time ago:
52
Bareev-Ivanchuk
Linares 1994
White to play
1. 4)e5!
a7
Black must begin to retreat be

cause the l1 + ft ending is lost after
1... Axe5 2. dxe5t Wxe5 3. b5
l1 a8 4. b6 Wd6 5. Wb5 and 6. b7
WinS.
2. b5
1td8
3. c6
4. Wb3!
a4t
a8
Else the ft advances decisively.

Best play by Black but...
5. 4)xd8!
6. c2
xd8
Black resigned after 6 l1 a8 7.

Wb4 l1aI 8. b6 g5 9. Wb5 l1blt
10. wa6 l1alt II. Wb7.
Chapter Three
3. b5
4. d7t
wf6!
e 7
And White had nothing better

than 5. d6! \tldS 6. f6 \tlxd7!
7. xf7t \tle6 S. xh7 b3 9.
b7 b2!. Black drew thanks to
his superior and W.
Yudasin-Kramnik
Candidates Match 1994

White to play
1.
xe3?
Making it easy for White. With

I ... b3! Black can still fight. E.g.,
2. b6 WfS 3. d6 WeB and 4 ...
Wd7 or 2. d6 Wg7 3. cB d3 4.
dB Wf6.
2. d6??
Clearly winning is 2. cBt (con

trolling the queening squares) Wg7
:l b5!, e.g., 3 ... b3 4. d6 and 5. d7,
or 3 ... d3 4. b6.
2.
wg7
Now 3. d7 d3 halts White.
Of course, some rules

apply only to the player
with the advantage. And
some rules apply only to
the one trying to draw.
Noah: Oh, how about "It's
easier topromote a distantpassed

i than a near one. "
Pat: Isn't that obvious? Seems
to me a i on, say b4 or c4, is
always going to be easier to
queen than one on g4.
I mean, if both sides castled
-side.
Noah: True, but even if they're
not on the same rank, the
distant i often has better
chances.
54 Black simplifies the task
by permitting White to ad
vance two is to the fifth rank,
ensuring that one could
queen. He should have
stopped the distant one ( 1...
b3!).
Pat: Then where did White go
wrong?
Noah: By trying to promote
the nearer i . If he had
pushed the b- i instead of
the d- i , Black would have
had to commit his -and
one of the is would have
sailed through.
Pat: Seems like you really en
joy collecting endgame rules.
Noah: You might too if you
gave them a chance.
Of course, some rules apply
only to the player with the
advantage. And some rules
apply only to the one trying
to draw.
-GM TaU
53
Rules
Pat: How does that work?

Noah: Well, something like
the last rule goes:
When Defending, Go
After the Most Dangerous
Enemy i .
Pat: That is too obvious.
Noah: Not really. In Diagram
55, with only a few is left on
the board, Black took the i
that didn't threaten him in
stead of the one that did.
And this was played by the
No. 3-ranked player in the
world at the time in the clos
est thing to a world champi
onship match.
Pat: It seems like everyone has
some personal rules they need
to remember. Do you have
any?
Noah: Sure, my Three Com
mandments.
54
55
Karpov-Sokolov
Candidates superfinals 1987
Black to play
1.
4Jxe4??
Mter 1... 4)xa4! White's only

winning chance is to trap the 4) by
2. ,Q,d4.
But 2... \ftd6 followed by 3 ...
c6 and 4 ... 4)c5 will draw.
2. \fib5
3. 1tf8!
4Jc5
Now 3. ,Q,d4? 4)xa4! would

have drawn (4. \ftxa4 f5 5. ,Q,e5
g5 or 5. ,Q,e3 \ftg4).
3.
4Jd7
And Black resigned after 4.

,Q,a3 d5 (4... f5 5. ,Q,d6) 5.

,Q,e7 \ftd4 6. ,Q, d8.
Chapter Three
Mnatsakanian-Vogt
Stary Smokovec 1979
White to play
1. Resigns?
But 1. We2 We5 2. Wd3 Wd5 3.
c 4t Wc5 4. Wc3 as 5. Wd3! is a
draw after all.
For example, 5 ... f5 6. Wc3 f4 7.
Wd3 fJ 8. We3 Wxc4 9. WxfJ
Wb4 10. We3 \t;xa4 1 1. \t;d2 Wb3
12. Wc I Wa2 and Wc2- I-2 etc.
Pat: Okay, let me start with

Commandment No. 1.
How come I see G Ms resigning all the time in positions that don't look lost to
me?
Noah: In some cases it's a
pride thing. They don't want
to play out ugly positions in
front of spectators. But amateurs can't afford pride.
It's far more embarrassing
to resign prematurely. As a
practical matter, whether
your position is a forced loss
should not affect you.
Here, in Diagram 56 for
example, White should play
the obvious 1. We2 and see
what happens.
Pat: But what if you know it's a
book loss?
Noah: Never resign until you
know your opponent knows

how to win. I'd never stop the
clocks in '/tt+W v. +W, for
example.
Sure, it's a book loss. But
almost no one knows the
book.
"Let no man surrender so

long as he is unwounded
and can fight. "
-Field Marshal Montgomery to
his troops on the eve of the
Battle of EI Alamein.
55
Rules
Thou Shalt Not Trade

Down to w + it Unless
You Can Safely Bet
Your First-Born Child
On The Result.
Pat: Every G M must.

Noah: Then why did Ljubomir
Ljubojevic play it out with
the against Gata Kamsky
at Linares 1991 ? Or Nigel
Short against Valery Salov at
Barcelona 1989, as in Dia
gram 57.
In fact, Short played out
Salov-Short
v.
w
to
the
very
end.
If
+W
if1
Barcelona 1989
White to play
there's any doubt in your
mind that your opponent can
1. -ra8+1 Wh7
win, let him prove it.
2. -re8!
Pat: I guess that makes sense.
A zugzwangposition that's been
Noah: That leads to Com in the books for 300 years.
Yet, Black played on: 2 a7
mandment No. 2:
3. *h5t wgS 4. *g4t Wh7 5.
*h3t wgS 6. *g3t WhS 7. *h2t
Thou Shalt Not Trade
wgS S. *bSt Wh7 9. *xa7t
Down to + .ft. Unless
You Can Safl,-- Bet Your WhS 10. *g7#.
First-Born Child On The
Result.
56
Chapter Three
59
Waulin-Grozspeter
Kecskemet 1993
Black to play
1.
2. Wh3!
d5
xe5t xe5
.txe5 xe5
h4
d5
xh5 wc4
Resigns
It's obvious after 7... Wxb4 8.

wg6 a5 9. h5 a4 10. h6 { 1 1. h7.
.txe5??
Black may have assumed 2.

xe5t xe5 3. xe5 Wxe5 4.
Wg3 Wf5 5. Wh4 Wg6 draws.
Or he mayjust have miscounted
what happens here now.
2.
3.
4
5.
6.
7. h4
When In Doubt,
Don't
promote the h- i . Total: 8

Pat: That's perfect for me. I
always find \t+ .ft endings
moves."
hard.
By the same count, you can
Noah: Yes, good players mess see it also takes Black 8 moves
to queen-but White moves
them up, too-even when the
position is largely a matter of first.
counting, like Diagram 58.
That means 12. h8=*
Pat: Counting?
comes in time to answer 12 ...
Noah: Sure. If Black had seen
al=* with 13. *xa l ! .
2. <It;>h3, a neat but simple Pat: I couldn't see all that.
gain of a tempo, he could Noah: That's a corollary to
my second commandmenthave said to himself: "Okay,
all the pieces are going off in
two moves.
When In Doubt, Don't.
"The result is + is. But I
can't stop <It;>xh5. Therefore,
It's funny how often people
my only chance is ... \txb4."
rush headlong into lost or
Pat: So far, I'm with you.
drawn + i endings long
Noah: Well, then he contin before they've exhausted
ues: "Mter 4 ... <It;>xe5 it will
their options, such as 1. .. Wd5
take White two moves to cap back in Diagram 58. It's the
ture the h- i , one to get the worst trading decision you
can make.
W out of the way, and five to
57
Rules
Pat: But don't people make Pat: VVhat about 2. e4?

other bad trades, into end Noah: Then Black locks up
the enemy with 2 . . . f5t!
ings or endings?
and VVhite's odds of survival
Noah: Yes, but in practical
are about one in a million.
terms there is nothing as final
as W+ ft s. Except maybe Pat: Speaking of odds, I'll bet
that game made an impres
death.
sion on Black he'll never for
You almost always have
some ways of complicating get.
your opponent's life when
there's wood on the board.
Yet in Diagram 60 Black
had an unstoppable winning
plan of 1 . . a3 followed by
bringing his W to g7 and .. .f6.
That wins the g- ft .
.
"Nobody ever
died/rom
playing a King
andpawn
ending-but
why take a
chance?"
-GM TaU
58
Yusupov-Ljubojevic
Linares 1992
Black to play
1. .
fS?
White assumes that 2. xf5 is

an easy win for Black (it is) and that
2. We4 xg5 is almost as good (it
isn't).
2.
3.
4.
S.
6.
We4!
hxgS
gxf6
Wf4
Wf3!
xgS
f6
Wxf6
gst
And the game is quickly drawn,
by 6 Wf5 7. g4t! hxg4t 8. wg3

or 6... Wf7 7. Wf2, with distant
opposition.
.
Chapter Three
mescas-Ljubojevic
Linares 1988
White to play
1.
2.
3.
4.
!! d4?
\t>d3
!! d5t
as
!! c7t??
\t>g5
f5
And White drew after 4 cl

5. Wd2 c6 6. b4 b5 7. axb6
xb6 8. as etc.
Only in the post mortem did a
kibitzer point out that 1... xd4 2.
Wxd4 Wg5 wins because White
soon runs out of "passes" (3. We4
a5! 4. We5 b6! 5. We4 f6! 6. Wd5
Wxh5 7. Wc6 f5).
Noah: As a matter offact, Black

had made a very similar er
ror in the same event four
years before.
In Diagram 61 he was of
fered a cold win with 1. ..
xd4-and gave it away.
Pat: That's pretty embarrass
ing for a top G M.
Noah: It was worse than you
think.
The diagram was an ad
journed position so Black lit
erally had hours to count out
moves like 1. d4?
Pat: Ouch. So why did White
play 1. d4?
Noah: Simple. He also blun
dered. People get tired in
endgames simply because
they are the ends of games.
Pat: What if there's no choice
but to go into 'lt1+ t ?
"Fatigue makes
cowards of us alL "
-General George s.
Patton Jr.
59
Rules
Noah: There's almost always

a choice.
In Diagram 62 White prob
ably thought he was at least
drawing-until it was too late.
Pat: Sure looks good after 1.
h 1. There's no way Black
can save the h- ! . Then a
trade of s would leave
White with the outside passed
f- ft .
Noah: But it remains the out
side ! for about a femto
second-until S . . . dS.
I'm always careful about
trading off my last . You
can take it off the board but
you can't put it on.
Pat: Whatever.
f6
3. g4
4. xh3 xh3
5. <I1xh3 d5!
Resigns
White is out of the square (6.
cxd5 c4).
Obukhov-Ibragimov
U.S.S.R. 199 1
White to play
1. hl??
White sees he can win the h- ft

and thinks that must be a draw.
But if he calculates it out, he
would force himself to find the
active defense: 1. b l ! h3 2. b 7t
and now 2 . We8?? 3. b8t and
2 . \tIg8?? 3. \tIg6 actually lose.
So the game should end in a
draw after 2 <&>f8! 3. \tIg5! and
checks at b7 and b8.
..
..
...
.
2. h2
I.
60
h3
a6
Chapter Three
Grivas-Minasian
Dehrecen 1992
White to pilly
1. g4??
It's not entirely clear that White
is winning after 1. '1fle4 and '1fle5e6. But there's an excellent chance
of it, particularly if the W reaches
Noah: That's another danger

Pat: Yeah, but usually you
in W+ i endings: the margin
see G Ms just shifting pieces
of error is much smaller than
back and forth for hours.
other endings.
That's not hurrying. But what
Pat: Meaning what?
is it?
Noah: Meaning that one error Noah: Subterfuge. Sometimes
is often fatal-and if you're
it's advantageous just to renot winning, you're probably peat the position when an
losing.
opportunity arises.
That brings up my third Pat: Why? Because you want a
commandment:
draw?
Never Shalt Thou Hurry.
Pat: What means "hurry?"

Noah: Anything that rushes
1.
into a significant change in
1. .
hxg4t
the number of is or pieces.
2. cifixg4 .tel
For example, in Diagram 63
The game was drawn 22 moves
White should have tried to
later: 3. Wt3 g5 4. d4 r:Jg7 5.
exploit the target at g6 before
We4 c1 6. Wd5 Wf6 7. Wd6
e3 8. e4 cl etc.
he traded off one of his two
remaining is.
"Rashness often
succeeds. Still more
often itfails.
-Napoleon Bonaparte
61
Rules
Noah: No, it gains time on the Noah: Usually it's when you
clock and .,cores mind-points.
can't significantly improve
your position further. In this
Psychological warfare.
Often the side fighting for a
case, White waited until he
draw will make further weak
had exhausted all of the
nesses-to avoid repetition
tempo-moves Black could use
because he starts doubting in the inevitable w+ it end
himself.
Ing.
In Diagram 64 White un
Of course, this applies best
Gleizerov-Salai
derstood that the best win
to positions like Diagram 64
Parduhice 1992
White to play
ning plan was to sacrifice the
when the enemy has little or
Exchange at some point. The
no counterplay.
1. a4!
problem is figuring out the Pat: As if you're playing with
Not 1. !3 d6t r:t;e7 2. !3 xg6??
right point.
out an opponent?
xg6 3. Wxg6 We6 and there's no
Pat: How do you know when Noah: Exactly. White has that win (e.g. 4. e4 as as in the note to 3.
it is right?
luxury here, so he can wait as).
i,te4
1.
until the a- it has gone as far
2. d6t e7
as it can go.
3. a5!
"The most valuable commodity in the endgame

next to a protected passedpawnis patience. "
-Anonymous
62
Still too early (3. !3 xg6?? xg6

4. r:t;xg6 We6 S. e4! as! and Black
draws-6. r:t;gS r:t;d7 7. '\t>fS wc6 8.
WxeS '\t>xcS 9. '\t>f6 bS!).
3. . . .
a6
The can't move (3 ... c6 4.

!3 xg6; 3 .. fS 4. a6! bxa6 S. !3 xa6
and the a- it falls) .
4. t! xg6! i,txg6
.
5. xg6
And White wins through the

opposition: S ... we6 6. e4 r:t;-moves
7. fS or 5 ... r:t;d7 6. WfS c6 7.
'lflxeS xcS 8. e4 bS 9. axb6 xb6
10. d6 and he queens first and
wins with checks at b8 and a8.
Chapter Three
g4-g5-g6 etc.
Charushin-Rosengol
Correspondence 1986
White to play
1. Wf4!
Not 1. .Q.xg4 .Q.xa6 2. Wf4 Wc7
:l .Q.fJ Wd6 4. g4 We7 5. g5 Wf7
and 6 ... Wg7, Black can then draw
just by moving his .Q. forever. The
g7 blockade can't be broken.
1. . . .
wa7
Worse is 1... Wc7 2. a7 .Q.fJ 3.

.Q.xg4. A .Q.-move like 1.. . .Q.b5
allows 2. Wxg4 followed by Wf41
g4-g5-etc.
And a .Q. -move like 1.. . .Q.d 1
allows 2 . .Q.xg4 .Q.c2 3 . .Q.c8 and
Noah: Remember, that unless

2. 1txg4 1txa6
the position has deteriorated
Wb6
3. 1tf3
into some mad t -queening
It's a race but this time Black's
race, you can take your time
r:t; loses.
with a maneuver or two-or
4. g4
'it'c5
four or five.
5. g5
Wd6
6. g6
we6
Also, because ofthe curious
7. wg5
1tc4
nature of endgames, you can
gain tempi by taking your
time.
Pat: Huh?
Noah: You heard me right.
And Diagram 66 is a perfect
example of it.
Pat: Do this one slowly. I have
real problems with 1t end
ings, even simple ones.
Black resigned after 8. g7 be
cause after 8 ... f7 9. h6 Wg8 Noah: Okay, what's happen
the W can be driven off g8 by 10.
ing is that White can take on
.Q.e4 Wf7 1 1. Wh7 Wf6 12. Wh8!
g4 whenever he wants.
followed by .Q.h 7-g8.
And the only way Black can
then draw is to meet 1txg4
with ... 1txa6 followed by
rushing his W; back to the W;

side.
Pat: That's a long way to go.
Noah: Sometimes W;s are sur
prisingly fast. Anyway, if
White takes on g4 immedi
ately, Black gets his W; to g7
in plenty of time.
Pat: But...
Noah: But by taking his time
with 1. Wf4! White actually
gains two moves. Black then
has to try for a blockade on g8
instead of g7. But g8 is a light
colored square and the W
can be driven away by
White's 1t of the same color.
Pat: I didn't think there was
that much calculation in the
endgame.
63
Rules
Noah: Sometimes you must

count-but not generally. If
you take your time you may
not have to. It's like Jose Ca
pablanca and his endgame
eyes.
Capa once came upon a
couple of amateurs trying to
figure out some complex end
ing, like Diagram 68.
Without even asking whose
tum it was to move, Capa
began shifting pieces around
until he had found a winning
formation for White.
Then he went back to the
original position to see how
to get from point A to point B.
Pat: What's the point of the
story?
Noah: This-in an endgame
Visualizing is often more
important than calculating.
64
13 moves.
Pat: How's that?
Noah: In the opening, you can Noah: Because he didn't know
what the lost position looked
get by on memory. But then
like. White did.
you start calculating. In the
middlegame it's often a mat Pat: That's what you mean by
endgame eyes.
ter of how well you calculate
your W-side attack as it races
to beat his * ,side attack, for
example.
Pat: And in the ending?
Noah: Time is cheap. But the
ability to visualize is often
crucial.
Here White is shooting for a
zugzwang position-his .!i on
d3 and Black's on d7.
Pat: He can't force that, can
Noah: Sure. As I said about
he?
how a master sees a end
Noah: No, but he can try. And
ing with one t and in most
by not hurrying he sets little
cases he can visualize-with
traps.
Pat: Which Black eventually out really calculating-wheth
er it will reach the Philidor
falls into-even though it takes
drawing position or the Lu.

cena WInnIng one.
Pat: One thing I'm sure of-I
wasn't born with endgame
eyes.
Noah: Nobody is a born end
game player. Bobby Fischer
was already a two-time world
championship candidate
when he decided to master
endgames-by locking him
self in a Manhattan hotel
room for three months with
just a set and a mess of books.
Pat: I think I'll pass on that
idea, or at least I won't hurry.
Noah: There are better ways
than Fischer's. Come back
tomorrow and I'll show you
something invaluable you
won't find in any book.
.
Chapter Three
t7
5. g2
So far Black has avoided all the

little traps. For instance, 5 ... .Q.g6??
would actually lose to 6 . .Q.h3! and
Black is in zugzwang.
For example, 6 ... Wc6 7. We5,
or 6 ... We6 7. Wc5 or 6 ... .Q.h 7 7.
.Q.fl !
6. t3
Pinter-Alterman
3. e2
4. fl
d5
b7
c6
e8
7. ...
8. d5!
9. b3
10. dl !!
d7
e8
d7
e6
70
'N n
.////.
Jl

t
//h

Rr)
Because Black's is are on tar

get squares White has good win
ning chances.
Black has other good passes,
such as 2 ... .Q.c4 and if3 . .Q.c2 .Q.e6
.l. .Q.d3 then 4 ... .Q.d7.
What White is looking for is a
zugzwang position: White .Q./d3,
Black's .Q./d7 and Black to move.
And Black resigned in view of

1 1. .Q.e2 .Q.d7 12 . .Q.d3 and:
Beer-Sheva 199 1
White to play
1. h5
2. dl
White continues 9. .Q.b3 and

now 9 ... .Q.c6 10 . .Q.c2 .Q.d7 1 1.
.Q.d3 or 9 . .Q.d7! 10 . .Q.d l ! ! .Q.e6
1 1. .Q.e2 .Q.d7 12 . .Q.d3!
The actual game finish was very
similar.
Now the right pass was 6... .Q. c4.
e8??
6. . .
7. b7!
Here 7... rj;c7 8 .Q.d5 Wd6 al

lows the key zugzwang position
the one White has been playing
for.
.
. .
. .

White finally reaches the win

ningposition: 12 ... We6 13. rj;c5 or
12 ... Wc6 13. We5.
65
Rules
66
In which Noah explains why a chess King is like a defensive

hack in foothall-and that in either game, two usually heats
one.
(OJffil]1{;}I
I
OOY:l ct
Mismatches
Pat: Okay, what's this mysteri

ous secret I won't find in any
book?
Noah: It's the single most im
portant principle ofendgame
play-the mismatch.
Pat: The mismatch?
Noah: It's another reminder
that an endgame is not a mid
dlegame. Mismatches are
unique to the ending.
Pat: How so?
Noah: In middlegames, the
board is filled with pieces and
the battle is like a tug-of-war.
White may have, say, five
pieces attacking a d5- t but if
Black has five defenders, the
situation doesn't change until
68
White adds a sixth. And if

Black has one more defender,
it's in balance again.
Pat: And in the ending?
Noah: It's more like hand-to
hand combat, one piece
against another.
And if you can make it two
vs. one, you have a favorable,
if not winning, mismatch.
Pat: But it must depend on
which two against which one.
Noah: A bit. But, as you can
see from Diagram 71, even a
'l!J cannot make progress
alone against a mere W+ t .
Against other single pieces,
the W+ t can force progress,
even against theoretically
stronger pieces.
Pat: But creating a 2 vs. 1
situation only happens if
you're ahead material, right?
The d8-piece is a Again, promotion can't be pre

vented. The .Q. can sacrifice itself
for the i after 1.. . .Q. c 7.
The d8-piece is a 4)The 4) can still sacrifice itself,
e.g., 1... 4)b7 2. WfS 4)d6t 3. We6
4)c4!
A 2 vs. 1 Mismatch
Black to move
What happens if:

The d8-piece is a 'lt1Black cannot win the i. The
best an unaided 'lt1 can do is stop it
from advancing, e.g., 1... 'lt1d6 2.
Wd4 'lt1b4 t 3. WeS 'lt1cS 4. We6
'lt1e3t etc.
The d8-piece is a The i can't be captured or
blockaded and its promotion can't
be stopped, e.g., 1... e8t 2. WfS
d8 3. We6 e8t 4. Wd7 and S.
d6, etc.
The d8-piece is a WThe i can be halted only by

opposition-and then only because
of stalemate (1... Wd7 2. WeS We7
3. d6t Wd7 4. WdS Wd8! S. We6
We8 6. d7t Wd8 7. Wd6).
Chapter Four
Kasparov-Timman
Linares 1992
White to plo,y
1.
2.
3.
4.
e8t
xf6!
g5t
h6
Wfl
Wxf6
Wfl
Resigns
For example, 4 ... a4 (4 ... We6?

.1. h7) 5. We5 d l 6. Wd6 b3 7.
Wic5 a4 8. Wb6.
White must create a winning
ft : 9 ... b5 9. a4! xa4 to. Wxa6
.Q.d7 1 1. b5 c8t 12. Wa7 and b6b7-b8=*.
Noah: Actually, sometimes

you can create a mismatch by
giving up material.
In Diagram 72 White wins
by taking the black W out of
the game with moves 2-4.
Pat: I see, because he must
prevent h 7-h8=*.
Noah: Precisely. Black re
signed when he saw there
was nothing to stop the W
from reaching the *-side and
creating a passed b- 1t .
Black must eventually give
up the .Q. for the 1t , on b5 or
b7, after which White's ti
returns triumphantly to the
W-side, where there's another
mismatch.
Pat: That must be a pretty rare
situation.
"Chess is the movement of

pieces eating each other. "
-Marcel Duchamp
69
Mismatches
\flg6
5. c5
Noah: Not really. It occurs
6. d6
often when one player gets
Black fell on time in a lost posi
his W involved in the game
tion after 6 d2 7. ctlc7 d5 8.
long before its opposite num
b4 as 9. d7.
ber-as in Diagram 73.
It looks like White's W
should just win easily but
there are tactical problems
( 1. We4 e7t).
Pinter-Popovic
Pat: But how can he get away
Thessaloniki 1988
White to play
with sacking the Exchange?
What's that about?
1 . c5!
Noah: It's basicallyjust a trade
White makes no progress by 1.
of one piece for another-but We4 e7t (2. Wf5? e5t).
f8
1.
it allows a 2 vs. 1 mismatch to
2. Wd4!
decide the game.
Now 2 . . . Ag7t is best.
Pat: Okay. I get your point
xc5t?
2.
that one guy has a big edge
3. \flxc5
. c7t
because his W plays and the
Otherwise the is just over
other W doesn't. But in most whelmed and the W is too slow (3 ...
endings both players have f7 4. Ad4 Wg8 5. d6 ctlfS 6. Wc6
We8 7. h5).
Ws.
. c2
4. Wb6
"Thefundamental
object of all mili
tary combinations
is to gain local
suPeriority by con
centration. "
-Rear Admiral Alfred
Thayer Mahan
70
Chapter Four
Wg7.
Salov-M. Gurevich
Biel 1993
White to play
1. e4
Not 1. Wc4? ctlg6 2. h5? ctle5t.
1. .
2. h5
3. Ae2!

g6
e7
Waiting: 3 ... Wc6 allows 4. We5!

and 5. 'iflf6. Also 3 ... We6 4. a4
(threat of as-6-7) ctlc6 5 . .Q.c4t
Wd6 6. Wf5 and 5 ... Wf6 6. Wd5.
3.
4. f5!
.
d5
Now 4... ctlc3 5. c4 and 6.

Wg6, or 4 ... ctle7t 5. Wf6 and 6.
Noah: Yes, but in most cases

Black can just barely hold
4. . . .
ra; e7
there is a latent threat of one the fort on the <&,-side-where
5. e5!
f6
player's <&' to run to one wing he has his extra piece-but
Also 5 ... ctlc3 6. -'tc4 ctlxa2!? 7.
or the other. That's the case
he's three tempi from salva
-'txc2 WfS 8. Wf6 and the Black W
in Diagram 74.
tion on the -side.
never reaches h8 while White plays
Pat: I don't get it. White has a Pat: So if mismatches are the
Wg6xh6-g6 and h5-6-7-8.
passed a- ft but he starts by greatest thing since the in
going to the -side.
vention of the 4J move, how
Noah: He needs a second win
do I make them?
ning plan, such as winning Noah: You can walk into a
the h- ft . If Black then shifts
losing mismatch of your own
his W to the defense of h6, he making if you play with only
allows a W+,g, vs. 4J imbal
one pIece.
ance on the -side.
Pat: For sure.
Pat: It's even worse after 5.
6. a4
d7t
<&,e5. It looks like W+,g, vs.
7. ra;f5!
c5
nothing.
8. as
e6
Also 8 ... ctlb3 9. a6! ctld4 t 10. Noah: That's when Black
wg6 ctlxe2 1 1. a7 and queens.
rushes the 4J back into action
Resigns
9. a6!
but it's too late. The game's
finish, and the similar 8. 4Jb3
9. a6! line are typical of mis
matches.
.
..
71
Mismatches
4. cSt!
Noah: But not always obvi
a h6? <&>c5).
No draw!
ous. In Diagram 76 Black Pat: I get it. Mter 2. a c8 Black
4.
dxcS
plays a losing move because is playing without a rlJ. And
5. d6
he thinks it will actually give Black loses if he allows
And Black resigned after 5
him the edge ( 1... e3 2.
rlJe6xd6 by playing . . . a xb3.
a d3 6. We7 a e3t 7. <;ftd8 in view
So the rest of the game...
of8. d7 followed by queening after
9. a c7 and 10. Wc8.
Noah: . . .is another mismatch
between his a and the White
W+d- t .
Browne-Biyiasas
Pat: Tell me, do mismatches
U.S. Championship 1980
Black to play
occur in all endings?
Noah: No, you usually need
1.
e3?
t s on both wings for a true
With 1... Wc5 Black's W-activ
mismatch. If all the t s are on ity ensures a draw. For instance, 2.
the same side it's much easier a h6 a e3 or 2. a h7 b6 3. ab7
for a mismatched defender to a e3 (a better try is 3. ac7t Kb4 4.
rush from one key square to a c6 but after 4 . . . Wxb3 5. a xb6t
Wxc4 6. a xd6 wb4 both sides
another one nearby.
will have to give up their a s even
But with t s on both wings, tually for the last enemy ft ).
particularly in + t and
2. c8!
Now 2 ... a xb3 3. We6 wins.
4J+ t endings, it ends up
2.
e5t
more like football than chess.
3. <Ilf6
e3
Pat: How's that?

72
Chapter Four
A: 6. a d4!, cutting off the W again,

e.g., S ... WcS? 6. a c4t WbS 7.
wd6 a xb3 B. a c 1 and eS-e6 leads
to Lucena.
13 Wd6 because the Black W will

be cut off ( 13 ... a b2 14. e6 a xb6t
IS. We7 or 13 ... Wd8 14. Wc6t
WcB IS. b7t).
Speelman-Gulko
Hastings 1988-89
White to play
1. d6t!
We7
It's easier after 1... Wc7 2. a d3

when Black's W is cut ofT, leaving
Black's a futilely trying to stop the
WI from advancing the e- .ft .
2. d5!
b2
Otherwise White reverses di

rection and advances the b- .ft after
3. Wd4!.
3. b5
4. b7t
5. b4!
Wd7!
wc6
Threatening to go back to Plan
Noah: Well, in Diagram 77,

you can see that if White pro
motes a i , he wins. If he
fails, it's a draw.
Pat: I've had positions like
this-and never figured out
how to get the two is mov
Ing.
Noah: You don't need two'*s,
just one. So White tries to
Otherwise Plan B (9. Wc6 and
stick his 'it' next to one i .
10. bS) soon brings about Lucena.
Pat: Where's the football come
9. e5!
el
To stop 10. e6t.
in?
10. b5
Noah: Think of Black's W as a
free safety. He's the last line
a e2 11. weSt we7 12. b6t we8
of defense.
And like a good free safety

he has to move to where the
ball-that is, the enemy <If-is.
Pat: So if White moves his W
to the side of the b- i ...
Noah: ... Black must follow suit.
Otherwise: touchdown.
In this case, White con
stantly threatens to cut off the
enemy W until he finally iso
lates it. It turns into <If+ + i
against <If either with the b
i or the e- i . In the main
line Black has a choice be
tween allowing "Lucena
with the e- i -or getting
mated with the b- i .
Pat: So the secret to mis
matches is to get your pieces
fighting weaker ones in a par
ticular section of the board.
73
Mismatches
Noah: Yes, but it's not always

Here the only clear winning
clear which are the weaker try is a piece sacrifice to set up
ones.
a W+ i vs. 4) battle on the
For example, Diagram 79 *-side.
starts out with Black trying to Pat: Cute.
eliminate the *-side as a fac Noah: Often a 4) can stop a
tor.
W+ i , if only by commiting
Pat: Is it so clear that 2. axb4
suicide. But Black's blunder
would have been a draw?
at move 3 was decisive.
Noah: Usually a 4)-ending
He allowed a final sacrifice
with only one i is a draw.
of a i , that left both his 4)
There are a few exceptions and W on the wrong side of
you'll learn. But usually: draw. the board.
Pat: You said mismatches oc
cur often in . + is and
4)+ is. Why?
Noah: Both 4)s and . s can be
overpowered by Ws. Re
member a 4)-unlike a .!t -is
a short-range piece that
moves slowly from one wing
to the other.
Pat: Almost as slow as a W.
74
4)xb6! or 5. c5! r:j;xe6 draws.
4. b5
5. W/c7
4)c8t
4)e7
Maksimenko-Baikov
U.S.S.R. 199 1
Black to Plily
1.
b4!
Standard drawing strategy. If

White has only one jt , Black may
draw (2. axb4 axb4 3. 4)xb4 4)e 7
4. d7 4)g6).
And if 5. 4)d5t r:j;g7 6. d8,
Black has 6... 4)e5! 7. e7 4)c6t.
2. 4)xb4!
3. axb4
axb4
4)e7??
To avoid a mismatch, the 4) has

to attack the b- jt from behind.
That is, 3 ... 4)e5 4. b5 4)c4t 5.
r:j;d7 4)b6t 6. '\t>c6 4)c4 7. b6
6. W/d7!
Also winning is 6. b6 4)d5t 7.
WId8'. .
..
6.
7. e7!
8. b6
4)d5
4)xe7
Resigns
Chapter Four
4. h5!
5. Wxf5
gxh5
And White wins (5 ... 4)d6t 6.

We6 b5 7. 4)c5 or 5... Wd6 6.
4)xb6 Wc5 7. 4)d7t Wxd5 8. 4)f6t
Wd4 9. 4)xh5 We3 10. We5).
Shirov-Hiibner
Manila 1992
White to play
1. a4!
Wc7
On 1... We7 2. Wd4 Wd6 White

wins with 3. 4)b2 and 4. 4)c4t.
The main alternative is 2 ... Wf6
after which White creates a mis
match with 3. 4)xb6! 4)xb6 4. Wc5
and wins (4 ... 4)d7t 5. Wd6 4)b6
6. Wc6 4)c4 7. d6 or 4... 4)a4t 5.
c6 e7 6. d6t Wd8 7. b6).
2. Wd4
d6
Again, 2 ... Wd6 3. 4)b2 and 4.

c4t wins.
3. we5!
xb5
Noah: That's why the key to

from being carried out.
winning most 4)+ .ft endings
Sometimes the key to win
is to create a mismatch.
ning-or drawing-is to pre
In Diagram 81 the threat of vent your opponent from
4)a4-b2-c4t forces Black to creating a mismatch.
choose between putting his Pat: How's that?
rtJ on the W-side, around f6
or e7, or the -side, around
c7.
Pat: And whichever way he
chooses, White runs the other
way. Is that it?
Noah: Yes-although it may
involve sacrifices (2 ... Wf6 3.
4)xb6! or, in the game, 3.
We5) for the W to penetrate.
Pat: If mismatches are such a
basic winning technique, how
come I've never heard of
them before?
Noah: Beats me. It also plays a
major defensive role by pre
venting the only enemy threat
75
Mismatches
Noah: Here's a case of win .Q. battle on either wing, with

Black's W on the wrong side.
ning by avoiding a mismatch.
In Diagram 82 Black uses zug
Then White would be able
zwang to make sure the en to sacrifice his h- i to get rid
emy W goes in the same of the c5-pawn.
direction as Black's.
Pat: So how can Black win?
Pat: Why is that good?
Eventually he must use his W
Noah: Because White can
to win the c- i .
draw if he can get in a W vs. Noah: True, but by that time
Black's i is on the sixth rank
and he can afford to give up
the .Q..
Pat: Not bad. But tell me, if I
ace this mismatch business
will I finally be able to under
stand what goes on in the
ending.
Noah: Not so fast, my young
grossmeister. I said the mis
match is the most important
technique in the endgame.
But it ain't the only one-as
we'll see next time.
76
82
We6 and White eliminates the last

ft by 6. d5 and 7. h8='/W!.}
2. <l?d7
d4
<l?xc4
3. e6
<l?d5
4. f6
And White resigned after 5. g5
c4 6. g6 c3 because of 7. g7 1td4t
Christiansen-Kengis
Manila 1992
Black to play
It appears 1... 1tg5t wins-(2.

\t>d7 \t>d4! or 2. f7 f4 with a
mismatch coming up}.
But 2. Wf7 f4 3. e6! xg4
4. d5 1te3 5. h6 1txh6 6. xc5
draws.
1.
Ae3!
Zugzwang. Now on 2. f7 f4
3. h6 Black had 3 ... g5!, which
leads to a winning + ft ending,
4. h 7 1td4 5. wg8 '\ffxg4 or 4. '\ffe6
xh6.
(But not 3 ... xg4 4. h7 1td4 5.
and 8 ... 1txg7 is a won + ft , or

rather *+ vs. + ft ending (9.
xg7 c2 1O. h6 c l-* I l. h7 *c7t
12. g8 e6!).
In which Pat learns ahout elhows, triangles, fortresses and

other stratagems unique to the endgame.
(iJffiW(:lI
Techniques
Little choice for Black now:

Pat: Okay, Noah, I'm feeling ners know how to win <ttt +
5.
g5
sharp today. Like I'm ready vs. W.
6. a5t
to nail all the other key tech
You probably learned it be
And White repeats the process
niques you were talking fore you knew what a
until, say, 18. a8*.
fianchettoed or the Caro
about.
Kann Defense was.
Noah: Good. Let's start by
recalling something one more Pat: But I never saw that as
zugzwang.
time...
Pat: Lemme guess. "An end Noah: Zugzwang is actually a
White to move
misnomer the way we use it.
game is not a middlegame."
White makes no progress with
Noah: Correct. There are a True ZZ is relatively rare and
few techniques which are occurs only when neither side "threats," such as 1. wd3 (threaten
ing 2. h5t) because of 1... Wc5!
unique to endgames that you has a beneficial move.
(or 1... Wle5!).
have to learn-and zugzwang
But zugzwang has come to
And if 2. c3, then 2 ... d5.
is at the top of the list.
mean when one player is run
1. a4!
But you already know a out of moves and it's such a
This leads quickly to fatal zuggood deal about zugzwang.
useful term it hardly seems zwang (1... Wc5 2. a5t).
1. . . .
e5
worth correcting the error.
Pat: I do?
2. d3
f5
ZZ is one of the relatively
Noah: You must. Even begin'1t>g5
3. e3
modem contributions to end
Black continues to be forced
game theory. In Philidor's day (3 ... e5 4. a5t).
hardly anyone understood it.
h5
4. '1t>fJ
5. g3
Pat: Even the great Andre?
78
Chapter Five
Noah: Even him. He gave Dia

gram 84 as a mate in 8. If
Philidor had appreciated 7Z
he could have saved White 5
or so moves.
Pat: So sometimes you need
zugzwang to win and some
times it just makes the win
faster?
Noah: You got it. Zugzwang
occurs all the time in piece
up endings. In fact, you often
can't win even when a piece
ahead without ZZ.
2. h7!
And mates next, thanks to Pat: Why piece-up endings? I
thought they won themselves.
zugzwang (or 1 . . Wb8 2. Wc6
.
Was 3. Wb6).
Other endgames that

can't be won without
zugzwang include:
+ 'If1 vs. W + 2s
1; + 'If1 vs. + 24J s
1; + 'If1 vs. + 4J
+ 'If1 vs. 1; + t!
+ 2s vs. W
+ + vs.
79
Techniques
Noah: No, they are hard to

win in other ways because of
the nature of 2-vs.-l battles
that I mentioned yesterday
even when it's two is vs. one
-'1,.
The player with the extra
piece can't always count on
picking off the enemy is
unless he can force them to
advance, as in 2. d5 in Dia-
80
gram 85. Nor can he count on

using his extra piece to crowd
the enemy W out.
After 2. d5 there is no way
for Black to force the win of
any i -except by running
White out of moves.
As Emanuel Lasker put it, a
right, in this case, the light to
move, is also an ohligation.
Pat: Sounds like a high school
graduation speech.
Noah: Well, Lasker thought
ZZ was one of the few things
that didn't have a parallel in
real life.
But he was wrong. There is
something in life like this zug
zwang: (see Bisguier pg. 81)
Pat: Cute. But, tell me, how
does zugzwang-the chess
kind-happen?
Seems like luck to me.
ltf5t 6. We 7 g4 7. d6 WfJ and

... Wg2.
2.
3. c5
4. h4
Ag2!
we7
Af3!
Not 4 ... Wd7 because of 5. f5!

gxf5 6. h5 e4 7. 'tt>d4!, drawing.
S. wc6
Djurhuus-Gipslis
Gausdal 1993
Black to Pilly
The direct 1... WfS allows 2.

'Ile7 and the d- i and W offer
White serious counterplay. Black
must try to win is while stopping
both dS-6-7 and Wf6.
1.
2. dS
Ae4!
Or 2. '1t>cS '1t>e6 3. h4 ltfJ and

White is out of moves.
If the white W temporizes (2.
'tt>d 7 '1t>fS 3. 'tt>d6) Black can begin
a queening race, which he wins
after 3 ... 'tt>xf4 4. d5 gS S. We6
Wf6!
Now 6. 'Ild6 Wf5 runs White

out of useful moves: Zugzwang.
6. 'tflc5
'tflfS
And here 7. d6 We6 is another

zugzwang since B. f5t gxf5 9. h5
loses to 9 .. ltxh5 10. Wc6 lteBt.
.
7. Wd6
Ag2!
And not 7. .. Wxf4 8. We6! g4t

9. Wf6 f5 10. d6.
'tflxf4
8. Wc5
White resigns.
In view of 9. d6 h3 10. Wc6

<ltg3.
Chapter Five
4.
5.
6.
7.
8.
9.
...
a a5!
g5
hxg5
a c5
a4
a b7
a e7
hxg5
wg7
e4
The threat of a4-aS and a c8g8# quickly runs Black out of

moves.
The game ended with 9 a b7
10. a b5 a6 11. ab4 as 12. ab5
a b8 13. e3! and Black forfeited.
Romanishin-Benjamin
Groningen 1993
White to play
1. a a3
a c7?
With 1... a c4! Black keeps ac

tive and can lure White into passiv
ity (2. 4)c3? as!).
2. 4)f6t!
Wh8
Not 2 ... rt/g7?? 3. 4)e8t or 2 ...

Wf8 3. 4)xh7t.
3. g4!
4. h4
h6
Now g4-g5, rendering Black's

W useless, can't be avoided.
Mter 4 ... 4)d7 White wins a ft
with 5. 4)e8 a b7 6. 4)d6.
Noah: Good players make

their own luck. It takes only a
minor slipup by Black in Dia
gram 86 to allow White to
paralyze him with simple
moves.
Pat: Why should Black lose
such a position?
Noah: There's nothing wrong
with it from a matelial or
structural view.
But by allowing his pieces
to become passive, his op
tions are slowly but steadily
reduced.
Pat: I see. First he ties his a to
the second rank. Then he
can't move his W from the
-side.
Noah: Notice that White didn't
even need his W to win this
ending.
Pat: It's hard to imagine zug-
zwang happening with so

much material on the board.
Noah: Zugzwang actually oc
curs a lot more in pure 4)
endings or 1i endings, and
less so with heavy pieces.
Pat: Why is that?
"ZugzUJang is like getting

trapped on a safety island in
middle of a highway when a
thunderstorm starts. You don't
want to move. But you have to.
"
-Arthur Bisguier
81
Techniques
"What is difficult about ma

neuver is to make the devious
route the most direct "
-Sun 1Zu
The Art of WaT
Noah: Simply because with

big wood there's less chance
to run out of moves.
Yet it can happen in any
ending-and with quite a bit
at stake, as Diagram 87 shows.
Boris Spassky advanced to
his first world championship
match with the help of ZZ in
this position.
Pat: Looks like White could
1.
Wd2
give it up immediately. He
White has only one useful move.
doesn't have a move.
Wc3
2. a e4
Noah: So all Black needed to
Now a -move along the e-file
create ZZ was to make it loses the c- 1 .
3. a h4
d3!
White's tum in the same po
And with no move, White
sition.
must allow the f- l to advance to
Pat: Neat.
victory, 4. e l f2t 5. fl e3.
Noah: And even the foremost
endgame authorities can miss
a ZZ-as in Diagram 88.
''1 destroyed the enemy
merely by marches. "
-Napoleon on the 1805 Austria

Campaign
82
Chapter Five

Noah: Yuri Averbakh, one of
the finest endgame analysts
ever, once gave 88 as a draw,
saying 1. d7 was the bestbut insufficient-winning try.
Yet it's an easy win.
Pat: I get it. Whichever direc
tion the black goes, the
White to play
white goes the opposite.
1. Wd7?
Noah: A very common theme
Correct is 1. d l ! after which
in the endgame.
Black's W must give way, e.g., 1...
Pat: How do you know when
We4 2. It;>d6 d3 3. Wc5 It;>e3 4.
you're close to zugzwang?
Wc4 d2 5. It;>c3.
A better try is 1... <tt?d5! so that Noah: There are usually clues.
2. lt;>f6 We4! blocks the W out and Pat: Like what?
Several English-language re
placementsfor zugzwang have
been suggested. Among them:
Move-bound
Plank walk
Movicide
Squeeze (a bridge term)
draws. But 2. Wd7! wins (2 ... Wc4

3. <tt?e6 d3 4. \t>e5 etc.).
1.
2. c6
3. Wb5
d3!
Wd4
c3
And draws (4. h3 Wc2; 4.

c 1 t <tt?b 2).
83
Techniques
Noah: Like when you see your

opponent has just about run
out of good moves, as Tal had
in the previous example.
In Diagram 89 you can
readily see that Black's only
good "pass" with the * is
... *c7.
With a bit more examina
tion, you can see that if the
white * reaches a7, where it
attacks 3 is, one will fall.
Pat: So hejust maneuvers until
Black runs out of useful
moves-and that happens af
ter 2. g I !. It looks a lot
easier than I thought.
Noah: It usually takes more to
win thanjust random maneu
vers-although it often looks
exactly like that. Consider
Diagram 90.
What do you think is hap84
Countering the threat of

...*b 1 txa2 with a bigger threat of
*h841=.
pening here?
Pat: Let's see. White only has

two is left and they're both
- is. This must be hard to
WIn.
Noah: In fact, there are some
endings in which - is are
better than center is. Re
member, endgames ...
Pat: . are not middlegames.
Enough already! I got it.
But what's the winning plan
here?
Noah: The best plan, some
times, is no plan-or in this
case, several plans.
White wins by shifting back
and forth, attacking the a- and
c- is while threatening to
bring his into Black's camp,
at e4 or c4.
. .
4. . . .
5. h2
6. *f8t!
*blt
*b2
Resigns
A.N. Panchenko-Grabarczyk
Katowice 199 1
White to pilly
1. *e5!!
h7
Zugzwang is relatively rare in

*+ 1 endings but here is one: 1...
*c6 2. *b8t g7 3. *a7, win
ning a 1 , e.g., 3... *fJ 4. *xcS.
2. wgl !
Establishingzugzwang since 2 ...

*c6 3. *e7 g7 4. *a7 is again
lost and so is 2 ... g8 3. *e7 and
4. *a7.
2.
3. *e7
4. *d8!
h6
*f5
Ftacnik-Browne
San Francisco 199 1

White to pilly
1. g a4
2. Wd3
Nc7
h5
Other moves lose the a- 1 or

permit the white W to go to e4/c4.
3. gaS
g4
Or 3 ... h4 4. a4 and Black
Chapter Five
give way (4 ... e5 5. c4

',"/1"4 6. Wxc5t and 7. c6).
Pat: But he can't seem to break

RealizingthatBlack's is be
91
through. He just keeps re come progressively weaker
\1leS
4. !! a4
peating the position.
and were in no danger of
White's and W are too fast
Noah: It only looks like repeti promoting-helped White
a fter 4 ... e6 5. xa6 h4 6. e3.
tions. No positions occurs WIn.
S. !! e4t!
more than once.
Pat: But ultimately his "plan"
Forcing a choice: On 5 ... Wd5
f i. !:! e7 e6 7. a7 White picks off
Pat: And yet after 1 1 moves of was just to go from one win
t he a- i .
shifting back and forth with ning idea to another, until
S.
\1lfS
Black was out of moves.
his W and ...
e6
6. !! c4
Black is lost because he must
Noah: White wins through Noah: True. And sometimes
7. \t>e3
move ( 1 1 ... c7 12. a5).
weS
creating zugzwang requires
zugzwang, pure and Simple.
Also: 7. .. h4 8. a4 g3 10ses to
1 1. . . .
h4
another special technique
Pat: Even though he started
!,. hxg3 hxg3 10. Wf3 e5 I I.
12. !!xa6 g3
g xa6 d4t 12. xg3 c4 13. f2.
13. hxg3
hxg3
tempo gaining.
with only two is. Amazing.
8. !! a4
c7
14. t3
Noah: But he had enough win Pat: And there's also tempo
9. !! e4t! fS
And Black resigned soon after
ning plans. In fact, he had losing, right?
And here 9... Wd5 10. e7 14... d4t 15. wxg3 we4 16.
more plans than is.
Wd6 1 1. h7 or 9 ... Wd6 1O. Wd3 a3! and Wt2.
In this case White had three
e6 I I. a4 c7 12. Wc4 and
ways to win:
Black is out of good passes ( 12 ... h4
13. Wd3! wins a W-side i ).
(a) by capturing the a- i ;
10. !! c4!
e6
(b) by eliminating all the
1 1. !! a4
side is; and
(c) by penetratingatc4 0re4
with his W.
I I l llst
85
Techniques
Noah: Basically, they're the

same thing.
Here in Diagram 92 is a
double error which delayed
the career ofone ofthe world's
best players.
White has virtually run a
once and future world cham
pion out of moves and has
two chances to force a win.
Pat: Where's the tempo busi
ness?
Noah: After 1. g c8 he could
have won by gaining a tempo
with 3. g gBt! and 4. g a8!.
Pat: Which he missed.
Noah: Yes, but when your op
ponent is this tied up there's
86
often a second chance.

White could have used it by
gaining a tempo with 2. g g8t
before c8.
Pat: I never realized how eas
ily you can win material or
promote is by doing noth
Ing.
Noah: It's more than "doing
nothing. Besides, often when
you have a solid material edge
you still need a special tech
nique to make it matter.
Pat: Why is that?
Noah: Because when you don't
have a passed i , you need a
point ofpenetration, an entry
square for your rlf into the
enemy stronghold.
Pat: Shouldn't there be entry
squares all over the board
when there are so few pieces
and is left?
won the match.

But 2. g8t! would have won
a tempo and gotten back into the
previous note-2 ... Wh7 (not 2 ...
Wh5 3. g5t Wh6 4. xf5t) 3.
c8 Wg6 (else 4. 1ig5) 4. 1if4 etc.
Anand-Karpov
Candidates 'Match 199 1

White to play
1. a8?
White can win with 1. c8!
(threatening lig5-e7-d6) Wg6 2.
lif4-zugzwang. Then 2 ... f6 3. exf6
and 2 ... Wh7 3. 1ig5! lose quickly,
so 2... c4 is best.
But 3. g8t! Wh7 4. a8!
c6 5. 1ig5 Wg6 6. 1ie7 wins
since 7. 1ic5, 8. c8 and 9. 1id6
must gain material.
1.
2. c8?
g6
6!
And Black eventually drew and
Chapter Five
3. fxe3
And White has established his
entry point-f4. The game ended
with 3 <ct;>d5 4. <ct;>f4 <ct;>e6 5. e4!
fxe4 6. <ct;>xe4 1-0, since White
wins the c- i .
Yermolinsky-Komarov
U.S.S.R. 1986
White to play
1. f4t!
The slow method doesn't work:
1 . wg2 Wd5! 2. h3 <Ile6! 3. e4
rxe4 4. fxe4 We5 5. f3 f4 6. <ct;>g2
g4! or 3. f4 gxf4 4. exf4 <ct;>f6! 5.
Wg3 wg6 6. WfJ Wh5.
1.
2. f3!
gxf4t
Not 2. exf4t?? We4 followed

by 3 ... Wd3 and White has created
a winning entry square-but for his
opponent.
2.
fxe3
Noah: Not necessarily. There Noah: Equal in material, but

are only c;t>s and is in Dia won for White because he
gram 93 and White is one has both an entry and zug
button ahead, as Bobby Fisch zwang. Mter he gets the c;t> to
f4 White can force the win of
er used to put it.
Now c;t>+ i endings are a the c- i and eventually get a
bit unusual because an extra position just like diagram 8.
i really means a lot in them.
The flip side of this is that
But here the win is hard be often the only way to draw is
cause White can't create a to seal off all entry points.
passed i and can't penetrate Pat: That must be hard with
is on both wings.
with his c;t>.
Pat: I see. The c;t>-side is sort of
closed. And d3 is off limits to
the white W.
Noah: Also, playing the c;t>
around to b4 doesn't work
because when it gets to the
-side Black plays ...g5-g4and makes an entry for his
own c;t> at 3.
Pat: Yet he wins by creating an
equal-material ending.
87
Techniques
Noah: Not that difficult. Dia

gram 94 is typical: White has
the better piece and a much
better-placed t;.
Pat: Seems all he needs to win
is to penetrate at e5.
Noah: That's deceptive. White
can penetrate-but accurate
play by Black can stop him
Alburt-Seirawan
from going any further.
u.s. Championship 1990
Black to play
White ends up with plenty
of squares for his but no
1. . .
d7!
targets for it. And he has
Or else 2. gS and WeS-f6 wins.
plenty of targets for his W but
2. g5
f5t
4)d5
3. e5
no squares to go to.
4. Acl!
This leads to another tech
Playing for zugzwang, e.g., 4...
nique that is exclusive to the
rt/ e7 S. Aa3t Wd7 6. AfS bS 7.
endgame: thefortress.
AcS! and wins.
Pat: Never heard of it.
b5!
4.
And S ... b4 will seal the posi

tion. A draw was agreed after 5.
Aa3 b4 6. Acl rt/e7 7. Ad2 Nc3
8. Ael.
88
Chapter Five
Sokolov-Vaganian
Candidates'Match 1986
1. a al!
2. an!
3. af3
e6
d5
Completing an impregnable for

tress. White shifted his between
g 1, h I and h2, and his along the
third rank for the remaining 26
moves of this drawn game.
Noah: No surprise. Fortress

arises much less often than
zugzwang or triangulation.
But it is crucial in many
positions in which there are
few is left and all of them are
on the same side of the board.
In Diagram 95 this enabled
White to avoid defeat despite
Black's big material edge
and thereby win a candidates'
finals match.
Pat: Neat. Black's '1t has no
way of breaking in.
Noah: It can get as far as g4 or
e4 but no further.
Pat: And an exchange of is
on g5 doesn't help.
"What is the object of defense?

Preservation. It is easier to hold
ground than take it.
"
-Prussian General Carl von Clausewitz
89
Techniques
Noah: Okay, you try this one.

by asking yourself: what are
Find an improvement on the good things about Black's
position?
Black's play in Diagram 96.
Pat: I'm clueless.
Pat: Hmmm. I guess you could
Noah: Then you should start say he controls a lot of light
squares thanks to the A.
Noah: Then look for a light
square fortress. A A at c6,
protects the b- t and can be
protected, in tum by a at
c4. Looks like a fortress to
me.
Pat: Me too, I guess.
"Dig or die. "
Black can lose the a- t , but
-Thaditional army
as long as he keeps the b- t
advice for
defender
and protects his pieces, White
constructing
doesn't have a winning idea.
fortification.
Noah: But his failure to secure
the A allowed the tactical
trick 2. g4! that cost him the
game.
Pat: Fortresses must happen a
lot in blocked positions.
90
moves, beginning with 4... d2t

5. c&>e3 d3t 6. '1f;e2 g3 7. '1f;rl
g4 8. Wt3 gl 9. xa6 1txf5
10. xb5.
Dorfman-Bonsch
Lvov 1984
Black to play
1.
f5?
One improvement is 1... a1 2.

xa6 b4, in order to eliminate the
a- t o
But 1... c 1 ! followed by 2 ...
1tc6 and perhaps ... c4 may cre
ate a real fortress.
2. g4!
h7
Not 2 ... fxg4? 3. e6t and 4.

xe4.
bl
3. 'lWe6
4. gxf5
And White won after 28 more
Chapter Five
hxg4 3. <ale3 Wf6 4. <alf4 wins for

White.
Noah: But many fortresses can

be stormed. Sometimes you
2. f4
4)f6
have so many positional ad
On the crucial 2 ... <alf6 White
vantages that you can afford
plays 3. e5t! xe5 4. xe5 dxe5t
a sacrifice or two to penetrate,
5. <ale4 and wins.
as in Diagram 97.
Or 3 ... dxe5t 4. Will g5 5. hxg5t
<alxg5 6. xc5 and 7. d6.
Pat: Why is this hard?
3. eS!
Doesn't 1. 4)xeS reach a
And White won soon after 3
+ i ending a clear i
dxe5t 4. xe5 Wd6 5. xg6.
ahead?
Noah: Very true-except that
after 1. .. dxeS White has no
entry and the game's
deader than the dodo. If
White then plays g4 Black
ignores it.
Pat: I don't really understand
how he can afford to sacrifice
two i s here.
Noah: It's clever. In the 2 ...
f6/3... 4)xeS line, Black
ends up a i up-but that ex
tra i is the useless one at g6.
.
Razuvaev-Ostojic
Berlin 1988
White to pilly
1. g4!!
A good plan would seem to be

I . f2 followed by 2. <alf4 and 3.
g4 .
But White can't dally too far
rrom the *-side because of
... xc4! (and if bxc4, then ... b3b2).
1.
4)xg4t
Now 1... hxg4 2. xe5 dxe5 is

a win for White because of the
entry squares atg4/g5 (after 3. <alf2).
Similarly 1... xd3 2. <if;>xd3
Pat: And only the positions of

the s matter.
Noah: Correct. What happens
in that line is that after S. e4
White has triangulated-his
going from e3 to f4 to e4.
Pat: That's another mystery to
me-triangulating.
Noah: It shouldn't be. Trian
gulating is just a way of creat
ing an entry point by lOSing a
move.
Pat: Sounds Simple-in words.
But in moves...
91
Techniques
Noah: Well, Diagram 98 is a Noah: He also knows some

relatively simple example.
other things: that White can't
If it's White's move in the play \t>c5 for tactical reasons.
diagram, he must allow one Pat: And White can't play Wc3
of two winning moves, for legal reasons.
... Wd4 or ... b3.
Noah: So by putting it all to
Pat: So what you're saying is gether Black can find a way
Black knows he has to lose a to drive the white W back
tempo.
and manufacture an entry
point. QE.D.
Pat: What I have trouble with
in endgames is that matelial
matters so much. Or so it
seems.
But then you get one ofthese
blocked positions-and the
number of 1s doesn't count.
Noah: Exactly. And sometimes
you can give up material in a
fairly open position to deny
access to your opponent. As
in Diagram 99.
92
Not 6 ... c4 because of 7. e5!.
98
7. Wd3
8. e5
Wb4
Wa3!
And White resigned in view of

9. exf6 b2 10. c2 a2.
Seirawan-Kasparov
Niksic 1983
Black to pia,
1. . . .
2. 'It'c4
'It'c6
'It'c7!
Now maintaining opposition

with 3. c5 1oses to 3 ... b3. And 3.
c3 is impossible.
3. Wd3
Wd7!
Now 4. d4 d6 5. c4 e5
6. d3 b3 wins.
4. we3
5. Wd3
wc6!
Wc5
Recreating the position in the

diagram but with White to move.
6. we3
b3
"In the endgame,

there is no 'later. ' "
-Pal Benko
Chapter Five
Or 4. 1tf2 g3 5. 1te 1 and now

Black loses a tempo with 5 ... 1td6
6. Wd2 1tc7! 7. We2 1te5! to cre
ate a zugzwang position in which
White must lose (8. Wd2 1tc3t; 8.
Wd l Wd3; 8. 1td2 1tf6 and 9. .
1txh4).
.
4.
5. -'tf2
.
I. Ivanov-Christiansen
U.S. Open 1983
White to play
-'td6
-'t c5
A white W move now allows

the winning entry at f3 or d3.
6. -'tg1
f4!
1. -'td2??
With 1. Wd3 White keeps the
enemy W out.
The key variation is 1... 1tc3 2.
e4t! fxe4t 3. We2 and White may
draw because e4 is unavailable to
Black's W.
1.
2. -'tel
we4!
g5
Now 3 ... h4 and 4 ... 1tg3 be

come a danger.
3. h4
4. g3
g4
This creates two connected

passed is. Black won after 7. gxf4
g3 8. 5 1t e7 9. Wfl Wt3! (not 9 ...
1txh4 10. wg2 Wxf5 I I. WfJ).
White tried 10. e4!? g2t 11.

Wei 1txh4t 12. Wd2 but after
12 Wxe4 13. We2 1tf6 14. Wt'l
1td4t he resigned.
Pat: I see. In the 1 . . . -'1.c3 line

Black ends up an extra ! but
it's on e4 where it blocks his
W's entry.
Noah: And to prevent a W
side entry, White ends up
putting more ! s on dark
squares-as much as he hates
to do it. Eventually it costs
him the game.
Pat: Why does it seem there
are so many more blunders
in endgames than in mid
dlegames?
I mean, you see a load of
question-mark moves early
on-but in the ending the mis-
takes rate double question

marks.
Noah: There really aren't
more blunders in the end
ing. The reason it may seem
so is that in post-mortems we
can tell more clearly that
endgame errors lead to im
mediate losses.
Middlegame errors arejust
as common-but they only
lead to bad positions that
may be saved later on.
93
Techniques
Pat: Okay, we've coveredzug

zwang, tempo gaining, entry
points, fortresses and trian
gulation. That's an awful lot
for one afternoon.
Noah: But there's one more
technique you need to know
elbowing out.
Pat: News to me.

Noah: Elbowing out is just
using your pieces-usually
your <&>-to crowd the enemy
out of action.
It happens all the time in
<&>+ a vs. <&>+ i endings.
Pat: Why those in particular?
Noah: Because the way the
94
side with the usually wins

is to elbow out the enemy
'iff-as in Diagram 101.
If White had played cor
rectly (1. a b3t) the black <&>
would have been elbowed
into zugzwang by move 10
and forced to lose both is by
move 17. Trust me.
Pat: Yeah, but how often do
you play such an ending?
Aren't you being the im
practical one now?
Noah: It's learning the basic
technique that matters. White
didn't need to capture the d
i in order to win. But he did
need to reduce Black's moves
to zero. That's why he had to
play 4. 'iffd4! in the winning
line.
Maybe Diagram 102 is a
more practical example.
lflf2.
1. . .
2. e5
g4
g3
And Black draws (3. b3t lflf2

4. lflf4 g2) .
Kalinichev-Sinyavsky
U.S.S.R. 1982
White to play
1. xd5??
Correct was 1. b3t e4 2.
g3 f4 3. g l ! and now the
white lfl elbows in until the black
lfl has no moves.
For instance, 3 ... g4 4. lfld4!
lflf3 5. l\>d3! g3 6. f1 t l\>g4 7.
lfle2 g2 8. d l l\>g3 9. lfle3! and
Black is out of moves (9 . lflh2 10.
lflf2).
He can delay matters only with
9 ... d4t! 1O. l\>e2 d3t I I. lfle3! d2
12. l\>e2 l\>h3 13. lflf3! l\>h2 14.
..
Chapter Five
Or 6... Wd6 when 7. Wf5! el

bows the W out once more.
Pat: Walk me through this one.

that diagram you should con
Noah: White has an extra i
sider lending an elbow.
7. xb2!
but Black's is excellently Pat: More like a body block.
And White converts to a won
posted to stop the enemy is Noah: True. Averbakh calls
"basic" position (7... xh6 S. d2!
and also promote the b- i .
this "the hockey technique"
g6 9. Wf4 f6t 10. Wg5 f8 1 1.
This means White has two because it's just like a de
g4: "Lucena" can't be stopped).
winning scenruios. In one he fenseman blocking an ad
gives up his for the black vancing forward.
i just before it promotes
Gavrikov-Hiibner
creating a 'itJ+2 i vs. mis
Swiss League 1992
Black to play
match on the -side.
Pat: llnd the second scenario?
I.
b4?
Noah: It's what happens in the
Mter 1... c3!, the issue is very
game-he trades one of his
much in doubt (2. cst Wd2!/3 ...
b4) .
is for Black's, leading to
2. c8t b5
"Lucena."
Worse is 2 ... b3, blocking the
Pat: Was it all forced?
i , e.g., 3. h5 Wa2 4. g4 b3 5. g5 b2
Noah: No. The key was when
"No man lives without
6. c2! Wb3 7. xb2t Wxb2 S.
jostling and beingjostled;
Black allowed his to be
h6 and the is score.
in all ways ke has to
elbowed onto the b-file.
3. h5
b3
elbow
himself
through
tke
b2
4. cl
Pat: What's the moral here?
world.
"
5. bl
wc6!
Noah:Just this-whenever you
-Thomas Carlyle
6. h6
h7
see the two s lined up as in

95
Techniques
is with 1. t! xb2, but he just

Noah: In Diagrams 103 and
104 you see elbowing at the ends up in that Philidor posi
tion again.
candidates' match level.
Pat: Let's see. White can trade Noah: But with 1. t! b 7t! and
4. 'it>g6 he creates a model
version ofelbowing out. Very
pleasing.
Pat: I don't know about that. It
just seems that in the mid
dlegame you have crunching
sacrifices, but in the ending
you have to use all sorts of
methods that aren't very
pretty.
Noah: "Winning ugly" in an
endgame is redundant.
Pat: Okay, okay, so every time
I can get my 'it> lined up like
that in a t! ending I should ...
"They don't give medals
for endgame technif[lul"
Noah: Not just t! s. Elbowing
-Bobby Fischer
is also one of the key tech
niques in 1t endings, as in
Diagram 105.
96
To stop 5. a e2tL
Dolmatov-Yusupov
Candidates' Match 199 1

White to pla,
On 1. a xb2 Black can't defend
his i but draws by attacking the
enemy's is with 1... a a3L
For example 2. ag2 a a5! or 2.
ltlxf5 a xg3 3. a b7t ltlg8 4. ltlf6
a a3 and the lateral checks draw.
1. b7t!
\tle6
Or 1... e8 2. a xb2 a a3 3.
a g2! a a5 4. ltlf6! and Black can
not avoid a h2-h5.
2. xb2
3. g2!
4:. '1f;>g6
a3
b3
e3
5. g4:!
6. f5t
fxg4
'1f;>e5
Or 6... ltle7 7. ltlg7, continuing

to elbow, with f6-f7-f8='/t.
7. xg4:
8. gl
f3
And Black sealed 8... a t'l (8 ...

a xf5 9. a e lt ltlf4 to. a nt) but
resigned in view of 9. a e 1 t and
to. f6.
Chapter Five
6.
7
8.
9.
f8
.Q.e7
c5
we4
9.
h5
g3
,Q,el
Short-van der Wiel

Rotterdam 1989
Black to play
1.
Wf5!
White threatened we4-f3-g2

and xb4!. Now 2. h6 gS
Black breaks the blockade.
2. Wd4
3. Wd3
4. h6!
wg4!
Wf3!
d6
With ... f4 to follow, e.g., S.

Wc4 f4 6. f8 hS 7. e7 (else 7. .
I d and queens) d2!, leading to a
position very similar to the next
(Iiagram.
.
5. Wd4
f4
Wg2 !
Not 9 .. wg3 10. Wd3 h4 1 1.

We2! c3 12. Wfl and 13. Wg 1
will draw.
Now, however, 10. Wd3 h4 is
hopeless.
.
10. ,Q,d8
Wh3!
So he can play ... h4. The game
ended with 11. Wd3 h4 12. a5
e7 13. we2 Wg2! White re

signs.
Pat: Where's the logic in this get his in front of Black's.

one? Black only has 2 fts and Pat: So White knew he could
one of them is the . - ft which always make progress if he
queens on a light-colored kept the white 'it from fl.
Noah: Elbowing is a natural
square.
Noah: Correct. If the white component of mismatches. A
gets to control h 1, all he needs very powerful weapon.
Just don't get canied away
to draw is the elimination of
the b- ft -which he can do with it.
with xb4!
Pat: But. . . ?
Noah: But Black's dances
along the light squares and
elbows the other out with
3 . f3! he's half way to
victory then.
Pat: Black takes an awful lot of
time to push his ft . He even
lets it get blockaded.
Noah: There's no race here.
As you'll ieam, in such posi
tions White's only chance is
to get his own passed ! or to
. .
97
Techniques
Pat: What's that supposed to

mean?
Noah: I mean all these tech
niques, like zugzwang, trian
gulation and elbowing, are
only means to an end-not an
end in themselves.
Pat: So?
Noah: So you have to be on
98
guard against cases like Dia

gram 107, where White de
cides that elbowing out the
black 'iff is better than creat
ing a passed t -a winning
t.
White won squares with 1.
xf4, but only drew the
game. He could have won by
1. 'iffxg6 and 3. xf4 even
though that loses material.
Pat: Endgames didn't seem so
confusing when we started
this afternoon.
Noah: The more you learn,
the more you realize how little
you know-like everything
else in life.
Maybe it will all be a bit
clearer tomorrow. I've got
something in mind that puts
all these techniques in per
spective.
in a few moves. He cannot keep

both his is and make progress.
Ivanchuk-Belyavsky
Linares 1992
White to pla,
1. xf4??
After 1 . <cflxg6 White queens by
force: 1... e3 2. WgS <cflf2 3. Wxf4!
<cflxe 1 4. WgS and f4-fS-f6; 2 ... Ac4
3. hS Ag8 4. h6 Ah7 S. 4)g2t
<cflxf3 6. 4)xf4 and 7. 4)g6 followed
by Wf6-g7.
1.
2. g5
Ab5
No better is 2. 4)g2 Ac6 3. 4)e3

<cflcS! and ... d6-e6-f6.
2.
.Q.e8
And White conceded the draw
In which Noah explains how much a good endgame plan is

worth-and why you often need two or three o/them to win.
Plans
Pat: Okay, coach, what's on

What Lasker meant was in
the agenda today?
that deceptively simple end
Noah: How to plan.
ing you win only by follow
Pat: Oh, yeah? I never thought ing a very logical plan, of
there was any planning in the limiting the enemy 'it rank
ending.
by rank until there are no
Noah: Quite the contrary.
more ranks.
That's where strategic think
It was, he said, the first true
ing began. As a wise man chess plan.
once put it:
Pat: Yeah, but that must be a
rare case. Real planning hap
pens in the middlegame, with
junk like minority attacks and
jt chains and so on.
Noah: Not true. The endings
are filled with little plans
although we don't always rec
ognize them as plans. And
they're extremely valuable,
"Planning in chess started on
almost a tangible asset like an
its career with the theory of the
endgame:
extra jt .
King and Rook vs. King. "
Pat: Oh, come on.
-Emanuel Lasker
Noah: No, it's true. As a prac100
4.
g5!
The W must now be able to

invade, say at f4.
5.
6.
7.
8.
b7t
Wh2
wg2
'lflh2
9. wg2
Resigns
Speelman-Korchnoi
Hastings 1988-89
Black to play
1. . .
bl
The first step in an unstoppable

plan-advance the b- i to the sev
enth rank, tying up the enemy W
and S .
Now 2 . Wg4 b3 3. S b7 b2 4.
S b8 loses to 4 ... f5t! followed by a
S check and queening.
2. g4
3. 'lflf3
4. wg2!
b3
b2
The W is now tied to h2 and g2

(4. Wf2? S h l ! 5. S xb2 S h2t).
wg8
WfB
We8
Wd8
Wc8
It's clear after 10. S h7 S c I and

queens, or 10. S b3 Wc7 and
. . . Wd6-e5.
Chapter Six
109
Karpov-Kasparov
World Championship 1990

White to play
I.
xd6!
e7
Normally 1... a5 would be much

better than this passive move-but
here it loses to 2. 1:! d5!
2. a6!
3. g l
tical matter a grandmaster ending I'm too busy playing

may pass up the opportunity moves to look for a plan.
for a "better position-even Noah: You don't have to come
a materially superior posi
tion-in favor of one with a
clearer plan.
1 09 White can get two con
nected passed is with 1.
f7t. But he has no clearcut
plan in the variation that ends
with 3 . dS.
Pat: So he went with the slower
idea, 1. xd6.
Noah: Not really slower, as it
turned out. Once his oppo
nent appreciated how easily
the White winning plan could
be carried out he resigned.
And Garry Kasparov is not
known for resigning prema
turely.
Pat: My problem is that in the
. .
Much better than 1. 1:! f7t Wg6

2. 1:! xa7 1:! e2 after which 3. Wgl
d5 gives Black excellent drawing
prospects due to his good 1:! and
passed t .
1.
the obvious winning plan of advancing the <Ii> to g4, e.g., 3 ... Wf7
4. f2 e8 5. f3, or 3 ... 1:! f7 4.
g3! WfS 5. g2 and 6. 1:! a4.
Wlg7
And Black resigned in view of
707
Plans
up with some Grand Design,

20 moves long.
But you need to have an
idea of what your ultimate
goals are.
If Black had taken a mo
ment in Diagram 1 10 and
asked himself, "How am I
going to win this?" he would
have answered: "By promot
ing the h- i ."
And that would have made
calculating 1... xh2 a lot
easier. Instead, he ended up
trading too many is and al
lowed White to draw by
blocking his only passer.
Pat: It helps that his was a
lot more active than White's.
Noah: You can execute good
plans even if your opponent
has the active pieces-pro
vided he doesn't have a good
102
plan of his own.

Pat: Show me.
Noah: Sure. In Diagram 1 1 1

White is a i down but his
pieces are placed on excel
lent squares.
Yet he can't go much fur
ther: He lacks a good plan.
He can only make minor ha
rassing threats.
That means Black has time
to execute an elaborate reor
ganization that involves free
ing the .fl from the defense of
the a- i so it can go to g6 and
then h4.
Pat: That should take years.
Noah: Actually only nine
moves. Time flies when your
opponent has no counterplay.
Pat: Doesn't counterplay come
HeUers-Adorjan
Esjherg 1988
Black to play
I.
...
e5?
Mter 1... xh2 2. ,Q,xe6 h5

there is no stopping the h- t .
2. dxe5
3. g e l
fxe5
gxh2?
If Black takes on f2 first Black

retains winning chances.
4. g xe5
5. h5!
g xf2
And in view of5 ... b7 6. e6,

Black played 5 ,Q,xh5 6. xh5
f6 but agreed to a draw after 7.
Wb2 Wc7 8. Wb3 (and 9. a4).
Romanishin-Timman
Taxco 1985
Black to play
1.
2. \t>c4
3. g al
4. b4

g b7!
4) b8!
\t>c6
a6
Intending ... a7 and ... 4)d7.
5.
6.
7.
8.
9.
g d3
gal
el
f2
g a5
4)d7
g a7
g a8
4)f8!
4) g6
Completing the first plan. Now

Black can attack the f3-pawn and
convert his advantage.
Chapter Six
Here 10. Wd3 ctlh4 1 1. We2

was White's last real hope.
10. a a3?!
I I. e3
12. d2
13. c3
as
4)h4!
a d8
Or 13 . .Q.xaS a d4tL
13.
14.
15.
16.
17.
axaS
xd2
a a6t
a a4
4)xf3
4)d2t!
a xd2
Wb7
a dl
1 13
Black took his time and won

with 18. a3 a d2 19. Wc3 a g2 20.
we4 a d2! (not allowing 21. Wd5
with counterplay).
Eventually the e4 or g4- i s had
to fall: 21. Wc3 a e2 22. a e4 we6
23. a4 ag2 24. as ag3t 25. Wb4
axg4 26. a6 a gl Resigns.
from well-placed pieces?

around in your head.
Noah: No, it usually comes
For example, if you spent
about when you have tar the last three hours looking at
the -side or trying to win
gets to attack-and here White
doesn't.
the enemy d- t , you'll need a
Or it occurs when you have clear head if the best plan in
a plan, like making a passed the endgame is to create a
t -which White also lacks passed b- t .
here.
Pat: How quickly should a plan
emerge?
Pat: So Black can take his time.
Noah: "Never shalt thou hur Noah: Often one suggests it
ry." Remember.
self as soon as fts go off.
Pat: How can I forget? But tell
In Diagram 1 14 Black had
me this: at what point in a just traded fts on d4 when he
game should I start planning? began his plan.
Noah: Very soon after the end Pat: Which was?
game begins.
Pat: After my little stroll around
the tournament room.
Noah: Don't laugh. One of the
reasons that's a good habit to
get into is that it helps get rid
of any middlegame plans left
103
Plans
l favorably.
Noah: The most basic of all
7. hxg5t wxg5
endgame scripts-to create a
8. dl
f8
passed t and promote it.
g4!
9. we3
Pat: Black didn't actually
queen the g- t .
.'m // _
1 15
Noah: No, but a good plan can
'0.W# W#
always be modified. In this
%I
t
case it forced a won + t

ending.
Ehlvest-Karpov
White was so concerned that
Linares 1991
Black
to play
the enemy passed t would
advance that he traded down
f4!
1.
10. xg4
hxg4
to a dead loss.
Other "good-looking" moves,
1 1. hl
f6!
Pat: I can never figure out my like { I.. . . e8 2. Wf2 . e3 3. . xd4
Another finesse, and better than
opponent's plans until it's too . c3, achieve little (4. . d2).
1 1... g3 12 . . h3! Wg4 13 . . h6
2. f2
fxg3t
late. Maybe I'm better off not
followed by . g6t or . xd6. E.g.,
3. xg3 e8
trying to guess what he's up
13 . . . . g8 14 . . xd6 g2 15. f2 or
4. xd4 f5
to.
Threatening . . . . e2, which 13 . . . . e8t 14 . . e6.
i
" r Jl
i
;W
would have been answered here

by 5 . . e4!.
5. 'itf2
6. f3
Wf6
g5!
Now 7. xh5 . h8! regains the

104
12. We2
g3
And Black won-13 . n ! .f4!

14. a3 Wg4 15. b4 axb4 16. axb4
g2 17 .xf4t Wxf4 18. f2 we5
19. wxg2 d4 etc.
Chapter Six
Sumjakina-Berezjuk
Bratislava 1992
White to play
1. a8t!
2. g5!
W';g7
Now 2... a2, threatening 3 ...

g flt and 4... al='t, can be an
swered by 3. Wg2 or 3. We4.
gl
2. . .
3 . xa3
xg5
4. a7
Black tried for another 35 moves
before agreeing to a draw (4 gf5t
5. wg4 g f1 6. wg3 g5 7. wg4
Wg6 8. g a6t f6 9. gb6! ghl lO.
g a6 etc.).
Noah: No. It is very important has more than one plan?

Noah : Then it becomes
to know enemy intentions.
See what happens in Dia tougher to thwart him-but
gram 1 16 when White takes not impossible.
away Black's main plan.
Usually one plan is more
Pat: Which was... ?
dangerous than the other, as
in Diagram 1 17.
Noah: Simply to activate
Black's W so it could win
White's is or shepherd his
own i to al. If Black had
gotten his W to f6 or e 7 it
would have been fairly simple
to carry out one ofthose ideas
since White's would be
burdened with three tasks.
And that's two too many.
Pat: Three tasks?
Noah: Sure. No. 1 : Stop the a
pawn; No. 2: Cut off the <ttl ;
and No. 3: Failing all else,
create counterplay by gob
bling up the W-side.
Pat: What if your opponent
' plan, like a

tree, must
have
branches-if it
is to bear
fruit "
-Basil Liddel
Hart
105
Plans
Pat: I don't even see one plan,

just a blockaded mess.
Noah: That pawn at c6 creates
it's own plan.
Remember what I said
106
about is increasing in value

the more they advance? That
inspires White to devise two
plans, the 1ixe5/d6 one that
worked and the b5/as-a6/
1iaS one he didn't need.
Pat: Black's 1... 1if2 would
have stopped the second plan.
Noah: But unfortunately for
him, not the first.
Pat: I'm a little surprised at
there being two plans avail
able.
Noah: Don't be. Many posi
tions can't be won with only
one.
Pat: Why?
Noah: Because as pieces de
part the board, the defender
has less to defend. And it's
easier to thwart one plan.
Pat: What's Diagram 1 18 got
to do with this?
passive 1... f6, e.g., 2. bS Wg5 3.

as ctJg6 4. a6 WgS S . .Q.aS ! (threat
ening .Q.xb6!) d8!'
Lagky-Task
Chanae 199 1
Black to play
f2?
1. .
2. c3 !
Threatening 3 .Q.xe5 dxe5 4.

d6 and a i queens.
.
2.
3.
4.
S.
...
xeS!
d6
'1t>xeS
h4
dxeS
d8
And the .Q. can't stop all the is:

5 ctJfl 6. b5 f6t 7. c&>d5 d8
8. g5! Resigns.
Back at the diagram Black

should be able to draw with the
__
__
__
__
__
__
__
__
__
__
__
__
__
__
____
__
__
__
__
__
__
__
__
__
__
__
____
Chapter Six
Noah: White wants to pro

mote the h- i but that idea is
easily stopped. So he finds a
second plan, winning the b
i.
Pat: Which is also easily

stopped.
Noah: Yes, but then with
Black's t! badly placed,
1. b7!
White wins by returning to
The direct 1. Wg6 fails to 1...
the first plan. On b5, with his
g3t 2. ",f6 b3! and White
own i on b6, the black t!
makes no progress.
1.
xb5
can't give "Philidor" checks
If Black allows 2. xb6, White
on the third rank or checks
has a variety of winning plans.
on the g-file.
2. Wg6!
WfB
You have to be aware of
Else 3. b8 mates.
multiple plans-even in so
3. h6
e5
called simple positions.
4. b8t e7
5. h7
Pat: I suppose that's a subtle
And the t queens.
hint about Diagram 1 19.
107
Plans
Noah: Quite right. If White
had only one goal-Wc5xc6b5xa4-he would have lost to

... We4-f3-g2.
Pat: That sorta makes sense.
White has to capture two fts
in order to promote his a- ft ,
but Black only needs to rip
off the h- ft .
Noah: You got it.
And if Black had only one
goal-to promote the h- ft White would have drawn by
going after the a4- ft . But this
time both players had a choice
of weapons.
Pat: Why should one plan
work and the other one fail?
Noah: When your opponent
108
stops one plan, he usually

expends so much energy-that
is, piece activity-that he al
lows the second one to work.
Pat: How many plans can there
possibly be in one position?
Noah: Sometimes, several.
For example, you may have:
(1) a mating idea,
(2) a ft that can promote,
(3) enemy fts to attack and
win, and
(4) chances to trade down to
a clearer position, like a won
W+ ft .
And that doesn't include
some ideas that are available
if you're trying to draw-such
as perpetual check, building
a fortress or trying for stale
mate.
Pat: Okay, what's the story in
Diagram 121 ?
Velea-Vidoniak
Rumania 1992
White to play
1. c5
\t>e4!
Not 1... g4 because both sides

promote after 2. b4! h3 3.
xa4 Wxh2 4. b3 g3 S. a4 h4
6. a5.
2. h4!
Hopeless is 2. xc6 fJ 3.
WbS h4 4. xa4 g2.
2.
3. <lib4
4. <lixa4
5. <liaS
<lid3!
d4
<lic4!
c5
A fine example of elbowing out.
6. b6
7. a4
\t?d4
c4
Both players promote but Black

trades 'Its into a won + 1 with
1 1... *cst and 12 ... *dSt or 12 . . .
*a3t.
Chapter
121
c3 3. e 1. [to avoid mate] d2t).
2.
3. c7t
c3
b4
And now 4. d7 allows 4 ... g3

and it is Plan No. 2 that triumphs.
4. ag7
Wiesniak-Kholmov
Warsaw 1991
Black to play
1.
g4!!
Plan No. 1, advancing the d- i

with the aid of the W, fails to 1...
Wc3 2. c7t b4 3. b7t! be
cause now 3 ... Wxa4 4. d7 Ieads,
at best to a drawn + i ending
(4 ... d2 5. xd2 0r 4 ... h3 5. Wd2
g4 6. d4t!). So he tries to pro
mote the g- i .
2. a d7t
The point of Black's last move

is that foiling Plan No. 2 (2. xg4)
makes Plan No. 1 decisive (2 ...
g3!
Anyway. Here White resigned

in view of5. xg3 c3-Plan No.
1 again-6. We I (else 6 ... h I t)
Wc2 and the i queens.
Six
Noah: Another case of how
two plans, each a failure on

their own, create a forced win
when combined. Here Black
sees that going for mate with
I". 'itfc3 and 2". h l can be
foiled by White checks. So he
tries to promote the d- ft -by
first threatening to queen the
g- ft .
Pat: And that diverts White's
temporarily.
Noah: But when Black pushes
the ft to g3 White either has
to capture it or let it promote.
Capturing allows Black to
win control of the key square,
d2.
"The most important
squares in the endgame

are the squares in front of
passed pawns. "
..
-Cecil Purdy
709
Plans
Pat: Seems pretty complicated.

Noah: Not compared to a mid
dlegame. There you're deal
ing with more than a dozen
objectives we might call plans,
such as weakening enemy fts
or harassing his * or creat
ing a pawn majority or Lord
knows what else.
In an ending usually only
one or two arise. In Diagram
122, for example, White's
chief plan involves his prime
asset.
Pat: The b- ft .
Noah: Of course. And if there
were no -side fts-and he
was just a ft up-the game
would easily be drawn. It can
only be won via a second
plan: attacking the -side
with his and a , with ma1 10
chine-like accuracy.
Pat: Computers are machines.
How come they play so lousy
in unknown positions, like
some blocked middlegame,
but they're so flawless in
"book" positions?
Noah: That's easy to explain.
A computer is a book.
Pat: That doesn't explain why
their weakest play comes in
unclear positions.
Noah: That's also easy. Com
puters can't plan. White
eventually wins by doing
something that doesn't seem
remotely possible at first: he
promotes the g- ft .
Of course it takes awhile.
Pat: Twenty moves from Dia
gram 122 is a lot! How long
should it take to carry out a
plan?
Going for zugzwang. If Black

retreats to d6, the 4. Wh51f6 idea
works.
3. . . .
4. fxg6
5. gb6!
g6
fxg6
d4
It was zugzwang again: 5 ... We4

6. e6t (elbow) and xg6/h6
reaches Lucena with the g- l .
6. g xg6
7. g xh6
Larsen-Browne
Las Palmas 1982
g xb7
White to play
1. g4!
White gets nowhere after 1. \t'e4
\t'e7.
1.
e6
As good a pass as any. Mter 1...

g6 White creates an entry for his W
with 2. b6t and 3. f5, e.g., 2 .
Wg7 3. f5 c&>h7 4. Wf4 and c&>e5.
.
2. f5t
e5
Ifthe rtJ goes to d 7, White breaks

into the W-side with 3. Wh5 and 4.
f6!
3. gb4!
______
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____________
____
____
____
____
__________
The result is a relatively clear

won game since the enemy rtJ is
crowded out and the 1 can reach
the fifth rank.
The rest went: 7 g7t 8. wf4
Chapter Six
t! t7t 9. wg5! we5 10. g4 t!f8 11.
Wh5 t!t7 12. g5 Wf5 13. ah8
we6 14. t! e8t Wf5 15. g6!.
Black played it out to Lucena,
another eight moves: 15 t! a7 16.
t! f8t We6 17. t! f1 t!a2 18. Wh6
we7 19. g7 and so on.
.
Noah: That depends on the
position. It can take 30 moves

even in an apparently simple
position like Diagram 124.
"Anticipation is 60 per cent of command. "

-u.s. Army cliche
"But like 10 years in the penitentiary, it is

very easy to say 'Anticipate!' and very
hard to do it "
-S.L.A. Marshall
111
Plans

Noah: An exception to an old
Capablanca rule.
Capa said that a clear Ex
change up is a win in the vast
majority of cases since you
can give it back at the right
point to win a ft -and that
should make it easy.
Pat: Is that the plan here? To
sac the t:! for the 1t?
1 12
Noah: No, the 1t is too frisky
rtie7
12. a c4
13. r:t;e8
I I.
e5
b2
5
to be traded off. This requires

a plan of several steps. You
Else the t! reaches the 7th rank
start by finding a target. Here,
and wins the f- t , e.g., 13 ... .Q.e5 14.
there is only one, 17. So the
t! c5 .Q.d6 15. t! d6 .Q. -moves 16.
frrst step is to get the <it> the
t! d7.
al
14. r:t;e7
closest it can to 17.
15. r:t;e6
b2
Pat: Won't he just be checked
16. a c7t r:t;g8
away by the 1t?
Lputian-Sideif-Zade
c3
17. a d7
Noah: Not if you head for e8.
USSR 1979
18. f3
b2
White
to
play
Pat: That takes a lot of time.
c3
19. g3
Noah: Time flies when the
r:t;g7
1. f1
opponent has no counterplay,
2. r:t;e2
r:t;g8
remember? Black's fts are
3. e3
g7
4. e4
g8
on their best squares and his
g7
5. d5
<it> cannot reach the White
6. a b3
al
fts. So all he can do is: pass.
c3
7. a b l
Pat: But White can't win the f
8. a cl
b2
ft by force, can he?
al
9. a c2
Noah: No, but after Black
Now that the .Q. is pushed to a
moves his f- ft , he gives e6 bad square White can easily reach
e8.
up.
10. d6
d4
Pat: I don't see progress yet.
Chapter Six
125
*
//
///,
R ?//h
;

\ff t
tt
66

20.
21.
22.
23.
24.
a d3
a dl!
agl!
ag2
g4!
1a.al
1a. c3
1a.d4
wg7
fxg4
On 24 ... hxg4 25. hxg4 fxg4 26.

t! xg4 threatens to win with f4-f5.
If Black defends with ...Wh7 or
... Wh6, he allows Wf7! .
25. hxg4
26. gxh5!
h6
Not 26. Wf7 h4! 27. t! d2 -'te3

28. t! d6 wg5 29. t! xg6t Wf4 and
Black draws.
26.
gxh5
Or 26 ... Wxh5 27. Wf7, because
27. .. g5 allows 28. t! h2=11= .
27. f5
h4
Noah: Wait for the second
stage. The new target is g6, so

White lines his t:! up against
it, on g2-another light
square ...
... and then blows the de
fenses apart with his one g
t.
Then the W reaches f7 or
f5-and it suddenly becomes
easy.
28. a d2
1a. c3
Pat: Seems it should have a lot
29. a c2
1a.el
30. a c6t g7
easIer.
Or 30 ... Wh5 3 1. t! c8 and 32. Noah: No, actually Black had
t! h8t.
an excellent t structure, con
3 1. g5
1a.g3
sidering the color of his .t. If
32. f4
h3
Black had begun with a light33. ag6t f7
colored .il White would have
h2
34. a h6
35. g4
Resigns
found another plan.
Pat: Such as?
Noah: Such as penetrating at
f6 with his 11 or advancing
his f- t to the sixth rank.
It was Mikhail Botvinnik

who said that masters gener
ally find the plans that ama
teurs can't.
Pat: A grandmaster would say
something like that.
Noah: But Botvinnik didn't
believe masters were infal
lible. As he put it, sometimes
"the plan which is so thor
oughly explained in the com
mentaries is only formulated
after the game is over."
Pat: You mean, after the mas
ters win, they lie about it.
Noah: You know, Pat, for a
young player you're showing
remarkable signs ofprogress.
1 13
Plans
1 14
In which Pat discovers why the most common ofall endgames

deserves its reputation for heing the trickiest.
'
i
,
'
;
Y
\
;
;
'lli!f.i;;
.,
:,
;
/;
.. ... .... -,::'":':'''
....
"
. ':.t,';t.' " .t'...

il
? '.'
::;:,f';t{.:;'::"
'
,.
: . . ,,',
'
. : 'e
' .
';i : 't' ):.
.
'
",
"
.:
." .
:;.:.
' ..
.
,
'
;
.;
'
:
t
.
,
.'
. ,
. .
..
.;
"
';;.' ;
; . .
'
.
":'
, ' ..
'
.'
: :'1: " :; :::'r.
' .
.
.f'
' . . .. ' '' ' " .
'1'
1 t
. " ;t:; ,' > I . : . ,

..
:
';:.:1) :r j!f; :" '

"
,.
,.
.
..
; :.
';.. ..
'.
, . . .... f}k .: .- ::' .'

. >
".
...
I.
. ..
'
'
.
f"
)
.. ,
..
'; :':;:ig :' ::" , ':

-I'
..
. of
.::0\:}1, :.' .'
- ..
"
. ,,'
..
'
,.
,' .
Rooks
are relatively rare.

Pat: Today I've got a question
minor pieces.
The figures don't lie. But
that's been bugging me. With Pat: But there are all those
a limited amount of time- books on, oh, + t (s), and sometimes endgame books
do.
and attention span-what
should I be studying?
Noah: Look at the numbers. Pat: Okay, you talked me into
Here's a chart of major tour starting with + t (s). But
Noah: Study the endings,
either pure + t (s) v.
naments showing which end I've got to tell you this is one
+ t (s), or endings with ings occur often and which endgame I'd like to take pass
fail.
-PERCENTAGE OF AU. ENDGAMESTournament
+ ts +m* t(s) v.Q.
Karlsbad 191 1
Moscow 1936
Santa Monica 1966
Leningrad 1973
Wijk aan Zee 1975
First four KasparovKarpov matches
US Chmpshp 1993
1 16
m*
4J
-'t -'t t
same opp
19
26
20
26
14
21
30
26
3
29
30
54
5
14
9
6
8
6
14
9
6
12
10
9
3
3
6
3
14
2
4
3
3
0
3
2
.7 1.5
3 3
3 3
3 3
0 0
0 0
19
26
22
minor pieces and
Sax-'ISeshkovsky
RovinjlZagreb 1975
Blilck to Play
1.
Resigns??
But 1... <&>h 7 2. f7 Rc8 and

White's W can neither advance
safely nor attack the a . For ex
ample, 3. We7 a c7t 4. We8 a c8t
5. Wd7 a a8 6. Wc6 a f8 and so
on.
Chapter Seven
Kasparov-Short
London 1993
PCA World Championship

White to Play
1. e4??
Hurrying. With 1. a2 (pass)
Wf5 2. e4t If/e5 3. It'e3 (elbowing)
White wins eventually. For ex
ample, 3 . . . It'e6 4. It'd4 If/d6 5.
It'c4 lf/e5 6. Wb4 as 7. as Wxe4
S. It'c5 and the a- t wins. Or 5 . . .
Wc6 6. It'b4 Wb6 7. h2 and
h6t makes progress.
1.
2. c.tfe3
3. c.tfd4
c.tfe6??
c.tfd6
And Black resigned after 3...

If/d7 4. If/c4 If/c7 5. lf/b4 e5 6.
cl t! It'b6 7. c4 and It'c3 -d4
(mismatch) leads to Lucena with
the e- t .
But 1... c5! would draw: 2.
a3 c4 3. as xe4 4. a6 f4t
and 5 ... f8.
Or, 2. as c3t 3. It'e2 It'xe4 4.
a6 cS 5. a7 as.
I. Keep the . Active I
Noah:Join the club. Everyone another in one move, not

finds them hard-because three like a 4) or five like a .
they are the hardest endings.
A can cross the board in
Even with advantages that one move-if it's on the right
would be sufficient to win diagonal. Even then a may
any other endgame-such as be useless if, say, you have a
White's two solid passed is dark-squared trying to at
in Diagram 128, the very best tack a i on a light square.
players in the world can blun- Pat: So ifyou don't make maxi
der big time.
mum use of a ...
Pat: If they do it, is there any
hope for me?
Noah: Well, there are some
tips I can offer, like that prime
directive of endings:
Keep the Active
Pat: Why is that a bigger deal

with s than with 4)s or s?
Noah: Because of a 's ex
traordinary range.
It can go from one wing to
117
Rooks
Noah: ...you're wasting the
strongest piece you've got in

most endings. That's part of
the reason a makes such a
lousy blockader-like 1. ..
a7?? here. The stronger the
piece, the weaker the block
ader.
Here the passive cost

Black only a tempo-but it
turned out to be fatal.
Pat: So how come so many
endings end up in hand
shakes? You know, "All + t
endings are drawn!"
Noah: Because an extra t -or
even two-doesn't count as
much in these endings. Active pieces do.
In fact, you could codify
another lute:
A More Active Piece In

+ Endings Is
(usually)
Worth A Pawn
"Nothing is more disas
trous in a Rook ending
than a passive attitude. "
-Rudolph Spielmann
1 18
Pat: How usual is "usually?"
2.
3.
4.
5.
6.
Alterman-Miles
Debrecen 1992
Black to Play
1.
a7??
Black obtains enough counter

play to draw after 1... a d6! 2.
a xa5 a xf6.
For example, 3. a a6 a fS 4. b6
a bS (or even 4 ... g7 S. b7 a bS!
6. a a7 gS! 7. f3 g6 8. e4 fSt
9. d4 f4 10. <a;>c4 a b I l L <a;>cS f3
and both passers may queen).
2. a4!
Now Black sees that 2 ... <a;>f8 3.

f4 e8 4. eS d7 S. dS
c 7 6. cS c8 7. b6 is hopeless.
_______
___
___
___
___________________________
___
___
___
___
___
___
___
___
___
___
___
_________
...
xa5
f4
e5
d5
b7!?
b6
xf6t
g7
White wins even if Black gets

the h-pawn, e.g., 6... a f3 7. a a4
a xh3 8. b6 gS 9. b7 a b3 10. c6
g6 1 1. a a5! threatening 12. abS.
Similarly, if Black gets his
going-6 ... h6 7. a a4 hS-he
still loses: 8. cS a fSt 9. c6
a f6t 10. c7 a f3 1 1. a b4! a xh3
12. b6 gS 13. b7.
Chapter Seven
Or 3. a3 (stopping checks)
xg2 and Black emerges a win
ning t up.
3. . . .
4. f2
Flear-Lyogky
Le Touquet 1991
Black to Play
1.
f6!
Mter 1... xg2 2. xa5 xh2

White draws with 3. a7t Wf6 4.
d7! (4 ... We5 5. e7t!).
2. xa5?
He still draws with 2. Wf4!

f2t! 3. We3! xg2 4. xa5
xh2 5. a7 h6! 6. d7 Wg7
because of 7. e5! dxe5 S. We4. Two
active pieces versus two passive
ones.
2. . . .
3. a7
e5!
b3t
xe4
b2t!

'<!.f
:
::
,,,
,
.
l . ir
If White tries to promote the d

t instead, Black stops him with S.
c6 c3 9. xd6 b5.
8. . .
Now 5. h4 b2t and Black's c

t is faster than White's h- t be
cause it gets W support. For ex
ample, 6. g3 c4 7. h5 c3 S. c7
Wd3! 9. h6 c2 10. h7 bS.
5. xJ7
8. h4
131
c3
And Black won after 9. h5 c2

10. h6 bl 11. f4 cl=ftt 12.
xcl xcl 13. g4 f1 t (elbow)
14. wg5 We4. E.g., 15. Wg6 gl
16. h7 xg4t 17. h5 gl and
IS . . . h l t. Or 15. h7 h l 16.
Wg6 xd5 17. g5 We6 IS. g7 d5
19. g6 f5.
A form of elbowing out. Now 6.

We i xg2 7. h4 h2 and ... xd5
wins.
6. g3
7. c7
c4
d3
Noah: An awful lot. Here's a
good example, Diagram 130.

After 1. .. xg2 Black would
have been a i ahead, but
White's active t! on d7 is
enough to draw. Even after
the superior 1... <&>f6 White
can draw with 2. <&>f4 even
though he may have to sac a

second t ( 7. e5t).
Pat: And he loses even though
he ends up a t ahead.
Noah: is are cheap in heavy
piece endgames because they
can be blockaded and cap
tured so easily.
Notice that in an ending a
changes character slightly.
In the middlegame he's a
long-range controller of files,
a distant general. In the end
ing he's a vicious attacker of
is, an invader of seventh
ranks.
Incidentally, another reason
computers are such poor end
game players is that they think
too much about material.
Pat: And not enough about
piece activity.
1 19
Rooks
Noah: Quite right. In a highly
publicized exhibition game

Diagram 132 arose.
Black lost because he
thought like a machine, pre
serving his h- is.
Pat: I thought only humans
would set a cheap trap like
1. .. h5 2. t!xh5?? t! c5 t .
Noah: Sometimes a player'S
"Pawns cannot escape

Rooks by running away
A Rook moves in the same
direction as a pawn. "
-CecU Purdy
. .
120
true character comes out in

the ending. Even a comput
er's_
Pat: So, what's the point? I
should just keep shifting my
pieces around searching for
activity?
Noah: No, the object is to find
the ideal squares for them. In
Diagram 133 Black finds a
superb place for his W to
blockade the passed is. And
he can cut the W off on the c
file. The makes up a two- i
deficit.
Pat: And White's t! is horri
bly passive at fl.
Noah: True. The only try to
win is to put White's W on a
forward square, like b 7, and
then try to activate the t!
behind the e- i .
But that shouldn't work.
Karpov-Deep Thought
Exhibition Game 1990
Black to play
h5??
1.
Black draws easily after 1... h3!,
threatening 2 ... h2 3. xh6 c2,
after which the a- i wins. For ex
ample, 2. xh6 a3! 3. xh3 a4
(behind the t !) 4. h l a2 5. a l
We 7, a W+ vs. W mismatch.
2. we5!
3. f5
h3
The threat of 4. f6 Wg8 5. f7t

forced a quick finish (3 wg8 4.
xh5 a3 5. xh3 a2 6. a3
c5t 7. Wf6 Resigns).
Chapter Seven
Anikaev-Karacev
Severodoentsk 1982
Black to play
1.
2. Wb4
3. h4
4. b5
a c8!
h5
f4
e3?
By simply holding the c-file

barrier, Black has a solid draw,
e.g., 4 . . . c7 5. Wb6 c8 6. Wb7
c5 7. e 1 e5! (blocking the !
and threatening 8 ... WxfJ) 8. f1
c5 draw.
5. e5
6. agl!
7. e6
e2
xf3
a e8
Pat: So why did he lose?

t! a6 to the Philidor drawing
Noah: By deciding to play
trick. Or he can draw with the
"pseudo-actively, as the win active defense, attacking the
ner put it, when just about only undefended target, the
anything else would draw.
f4- i .
Remember, active isn't al Pat: You mean he had two
ways better than passive.
ways to draw?
Pat: You mean Black tried to Noah: Why is that surprising?
win back material with When all the is are located
With his W now freed, White
. . . e3-e2 when he could close together and there's no
proceeded to win with 8. g6 Wf4
have maintained a rock-solid passed i , the defender has
9. wc6 Wf5 10. Wd7!.
blockade with . . . t! e5! at the all sorts of resources in t!
right moment.
endings that he doesn't have
''Experience in
Noah: Quite so.
in other endgames.
Pat: Aren't there a lot of times
Sometimes the hardest part
Rook-endings is
when passive play with the of defending is trying to
what you getfor
or t! is as good-or even choose between three con
failing to find
better-than active play?
vincing lines that all seem to
those mate-in
Noah: Naturally. Chess isn't
draw.
fours in the
that Simple.
middlegame. "
In Diagram 135 Black has a
-Anonymous
classic choice between a pas
sive defense, leading after 3 ...
12 1
Rooks
essentially passive plan that

Pat: But Black loses here and I
gave White a golden penetra
still don't understand why.
tion square for his tl at g6.
Noah: He lost because he
chose an active-looking-but Pat: That's funny. Black loses
because his own .t. shields
the white from checks
along the files.
Noah: It's sort of like defeat
ing the ghost of Andre Phil
idor.
Pat: Whatever.
I know a 4J stands badly on
the edge of the board and
well in the center. Are there
any best places for t!s?
Noah: Yes. As I mentioned,
t!s are long-range pieces, the
field-cannons of endgames.
They often belong a good
distance away from the center of the action so they can't
be attacked or shut out. Like
in Diagram 136.
122
Black.
4.
S. f6
<tt>f8
Resigns
There's no defense to 6. Wg6.
Zaitsev-Hubner
Busum 1969
Black to pia,
Black draws with 1... t!a5 (pass)

and if 2. g5 fxg5t 3. fxg5 t! a6
(Philidor!) 4. h5 t! b6 5. g6 t! b l
and an unstoppable series ofchecks.
He also draws with 1... t! b4! 2.
f5 t! b l ! 3. g5 t! h l t! or 3. Wh5
t!gl ! (4. wg6 t! xg4t 5. Wxf6
t! a4 and lateral checks).
...
2. <tt>h5
3. g5
4. fS!!
I.
a bl?
agl
fxgS
Undoubtedly overlooked by
Chapter Seven
And Black offered a draw after
another 30 moves (7. <&>c3 c5 8.
.Q,b6 c8 9 .Q,d8 c6 10. <&>c2
e6 11. Mhc3
1\ b6 Mhc6
e4 12. ...
19'
19'
13 .Q,d4).
Dokhoian-H. Olafsson
Socki 1988
Black to play
1.
a a8??
Black wins easily after 1... a l

2. e7 e l t 3. d4 b5 4. c3
e4 with the first of many zug
zwangs to come.
2. e7
3. d4
4. c3
d7
a c8
e8
Black has no clear way to im

prove his pieces, particularly the
passive .
5. wc2
6. d2
a c6
Wd7
Pat: Yeah, I see. The a works

much better operating from
e l than from as.
Noah: On the e-file it cooperates with the black to advance the c- ft . On as, or as it
turns out on the c-file, the a
loses much of its power. And
the <ttl loses virtually all of its
range because it is tied to the
enemy ft .
Pat: So White can draw with
little more than <ttl-moves.
Noah: That why a -manage
ment is so important.
Since so many a + t end
ings end up as man-on-man
battles, the misplacement of
the a can lead quickly to a
fatal mismatch.
For example, do you think
the white a stands well in
Diagram 137?
"Most terrible, or
rather, most
effective, of all
maneuvers IS
sudden attack
against the
enemy's rear. "
-Onasonder,
1st century Greek
military writer
123
Rooks
"Rook endings are the most

democratic endgames of all:
Every players gets a chance to
badly misplace his Rook.
-Anonymous
124
Pat: Don't as always do well

on the seventh rank? This
one can be used to stop the b
ft or harass the .
Noah: True, but after two ac
curate moves the game would
have been over. The a just
wasn't active enough. On eS
the a would have been able
to get behind the ft at bS.
Pat: Seems you're saying now
the more active a always
WIns.
Noah: Not quite. But what you
want to avoid is a serious
imbalance-your very pas
sive, almost immobile a
against his flexible one.
Pat: What's happening in Dia
gram 13S?
Morozevich-Van Wely
Tilburg 1993
Black to play
1.
2 .B: e l
3 .B:bl
b2??
b4
And by the time White has to

sacrifice the a for the b- i , his It;>
side is will be too far ahead. Draw.
But in the diagram Black wins
immediately with 1... a d3t 2. g4
a e3! ! since the white a has no
way of getting to the b-file or first
rank (3. a f7 b2) .
Chapter Seven
3. a a8
e4t
Resigns
The rest would have been 4.
We3 Wg4! S. gst Wh3 and
now 6. gS is White's only de
fense. But then 6 . .. c3t 7. Wf2
f3t S. We2 wg2! creates zug
zwang (9. gS xg3 or 9. hS
Wxg3 or 9. We I f2).
Miles-Ermenkov
Aegina 1993
Black to play
1.
e5
Played only after 20 moves of

jockeying for better W position
without success.
Noah: White's !! is more "long Pat: Only sometimes?

distance" and should play for Noah: Certainly not always.
the maximum checking pos But usually the !! belongs
sibilities, on files as well as on behind a passed i -whether
ranks.
it's your i or his.
But he violates the principle Pat: White looks very bad in
of flexibility by checking at Diagram 139 because his
a6-and allowing the to can't get off the first rank.
reach the terrific g5 square.
Pat: Now you're going to tell
me the thing about !!s be
longing behind passed is.
Noah: Well, they do. Some
times.
2. a a6t??
Mter 2. as! White can meet

2 ... WgS with checks along the files
and 2 ... e4t 3. We3 c3t 4. Wf2
f3t S. Wg2 WeS with checks
along the ranks (6. aSt).
2. . . .
g5
The W must reach g4 now with

decisive effect.
"The best SfJUllTesfOT the
defender's Rook is in the

comer, awayfrom the
scene of the battle. "
-Rudolph Spielmann
125
Rooks
Noah: Yet he can draw be
cause in the key 1. .. !! e4 line
he gets his !! to b 7 where it
immobilizes both fts-and the

enemy !! as well.
All Black has left to move
then is his 'it>, but he has to
remain near the f- ft or it may
promote.
That illustrates another key
principle:
Keep The Enemy Passed

ft From Advancing
Savchenko-Naumkin
Pula 1988
White to play
1. b4!
Pat: Another version of

Not 1. t! c l c5 or 1. t! d l t! xb2
"Passed ftsMustBe Pushed?" 2. d8t f7 3. t! d7t e6 4.
Noah: Sort of the reverse im t!xc7 t! xa2 followed by ... d6age. Remember hows fts in c6 and ...b5.
c2
1.
crease in value as they get
Mter 1... t! e4 2 . t! c l t! xb4 3.
closer to the 8th rank? In t! xc7 Black gets nothing from 3 ...
Diagram 140 for instance, t! xf4 4. t! xa7 and little more from
Black has an extra ft and has 3 ... t! a4! 4. t! b7! t! xa2 5. f5!.
2. dl!
f7
managed to blockade White's
3. d3
a- ft -the only source of en
And White drew because 3 ...
emy counterplay so far.

126
______________________________________________________
___
Keep The
Enemy Passed
:it From
Advancing
Chapter Seven
6. as!
7. a6!
8. Wc3!
Kupreichik-Sveshnikov
USSR 1985
Black to play
1. . . .
f5?
With 1... Wf8! Black can hold

the a-pawn on the fourth rank: 2.
We3 We7 3. Wd3 Wd6 4. Wc4
Wc6 5. Wb4 Wb6, and with his
free Black has excellent winning
chances.
2. we3
3. d3
4. wc4
r7
e6
f4
The Black is too slow (4 ...

Wd6 5. b4) so he uses his is.
5. Wb4
a a8
e4
we5
Pat: Why do I think there's Pat: Seems like there are al

another "but" coming up?
ways mismatches in . + !
Noah: You're right. "But" he
endgames.
spends a tempo on pushing
his own f- ! and that costs a
tempo.
White uses the tempo to
break the a5-blockade and
tie up the black t=! -not the
-on the a-file.
Pat: And that means what?
White threatens to use his Noah: That means a mismatch
greater mobility and eliminate
on the -side where Black's
is via 9. a5t and 10. Wd4.
lone W is trying to fight two
8. . . .
f3!
mobile pieces without the
9. gxf3
exfJ
benefit of his t=! .
10. d2
f4
Pat: Black is even in danger of
1 1. a7!
lOSing in Diagram 141.
White draws. E.g., I I... g5 12.
We 1 Wg3 13. Wfl ( 13 ... Wxh3 14. Noah: Deservedly. He com
Wf2 g4?? 15. h l =It) or I I... wg3
mited the sin of taking too
12. gl t and 13. xg7.
long to use his W-and in an
endgame that's often a mor
tal sin.
127
Rooks
Noah: Naturally. One of the
most common themes is
shown in Diagram 142.
128
White's b- t , with the sup

port of and 11, must win a
on the * side. Black's
is no match for three enemy
units.
Pat: But if he'd played accu
rately?
Noah: Then there would be a
balancing mismatch on the
W-side. Black's <if? would cre
ate at least one passed t that
would eventually cost White
his .
But he lost a vital tempo-in
fact, he lost three tempi-with
his faulty -move.
Knowing when to reposi
tion your -from in front of
a t to the side of it or behind
it-is one of the most subtle
skills in rookdom.
Diagram 143 is a fine illus
tration.
..
______
____
____________
____
____
____
____
____
____
____
____
____
____
____________
____
____
____
____
____
____
____
____
________
Short-Speelman
Hastings 1987-88
White to play
1.
e6??
With 1... a e8 2. b7 a b8 Black

can draw because by the time White
plays rtIa7 (and forces ... a xb7t)
he will have created an equalizing
passed i on the other wing.
2. b7
Resigns
Because now 1... a a6t 2. rtIb2!

ab6t 3. a b3 a xb7 is forced,
after which Black is three tempi
behind the previous line (4. a xb7
rtIf6 5. c3 h5 6. d4 g5 7. e4).
Chapter Seven
7 .B: d2
S .B:a2
a3
Pat: Looks drawish to me since Pat: So sometimes the is the

<;t>d6
xg7 can't be stopped.
star ofthe endgame and some
Noah: True, but Black can re
times the t; is.
position his just in time.
Noah: But ideally they should
Pat: Behind the a- i ?
coordinate-as they do after
Noah: Ideally, yes. But it can
13 . . . d3! in this case.
go to c5 if needed, as in the 3.
as line in order to free the
t; so it can chase the white
Pr. Nikolic-Vaganian
from its best square, as.
Lucerne 1989
Black to play
Pat: So both s get shifted,
9. <;t>d2!?
Black's during moves 1-4 and
To avoid the immediate zug
1.
.B: a2t
White's during 6-S. And
Here 2. '1t>d3 xg2 is hopeless zwang which would occur after 9.
Wd3
Wd5.
White gets back just in time.
for White.
9.
d5
2. <;t>f3
.B: c2!
Noah: But the difference in
10.
<;t>d3
as
Now on 3. a8 Black wins with
s in Diagram 144 is termi
Black wins the opposition ( 1 1 .
3". c5 and ".Wc6-b7.
nal. White's is almost out
'1t>e2 Wc4).
3 .B:xg7 .B: c7!
of moves and space, and he
1 1. <;t>c3
.B: a7
4. .B:gS
12. <;t>d2
can only pass with his t;.
a4
5. <;t>e2
"Always put the Rook
Resigns
It's entirely appropriate that
6 .B: dSt
behind the pawn Except
Because of 13. We2 d3!
The only way to get the back
Black finishes offwith another when it is incorrect to do so. "
in time.
repositioning of the -from
-Siegbert Tarrasch
<;t>e7
6.
the a-file to the third rank.
729
Rooks
Pat: So when doesn't the

tions by following the old
cliche.
belong behind the t ?
Noah: When it can be more Pat: I don't get it. What can
useful on a rank, as in Dia possibly be wrong with 1.
al?
gram 145. White spoiled his
chance to advance in the Noah: The a becomes a pas
world championship elimina- sive bystander on al once
Black blockades on as.
But on e4 it would have
kept the e- t under attack,
creating the possibility of a
<&>+ t vs. a mismatch on the
'l!J-side.
Pat: I guess a isn't that much
stronger than a <&>.
Noah: As I mentioned earlier,
when we were talking about
mismatches, usually a can
not stop a tJ from advancing
a nearby t . But there's a big
exception.
Pat: Let me guess. It happened
in Diagram 146.
130
Yusupov-Timman
Candidates Match 1992
White to play
Mter 1. a e4! White's W reaches
the 'It-side decisively, e.g., 1... <lifS
2. a eSt and 3. a5 or 1... <lie7 2.
<lie3 \ftd6 3. d3 and 4. <lic3.
I.
. al??
2. we3
. a51
e51
And Black draws, e.g., 3. fxeSt

<lixeS 4. \ftd3 \ftdS S. c3 c6 6.
<lib4 a eS and ... <lib6 followed by
a checks.
The game actually went3. <lie4
exf4 4. \ftxf4 <lie6 5. <lie4 g51 and
was eventually drawn.
Chapter Seven
Noah: Yes, this decided a crusupporting the t 's advance.

Bear in mind the checking
cial match in a European
2. a f5
a h8t
Team Championship. Hours distance only applies to a
Now 3. Wg3 g8! keeps the 1
ofadjoummentanalysiswere in front, not in back as in 1 . . .
back, and 4. f4 f8! leads into
dl.
rewarded when the diagram
what happened.
Pat: The problem for me is
was reached.
3. g5
ag8t
4. f4
a fB!
that all these positions look
White wins if he gets the t
Draw
5. a xf8
to the fifth supported by his alike.
Black keeps the crucial opposi while the black is not in Noah: Same for most of us.
tion after S ... xf8 6. gS wg7 7.
the ts path. But Black stops Even the great authorities on
fS Wf7.
that because he has the the endgame get mixed up,
as in Diagram 147.
"checking distance."
Pat: What's that?
Pat: Who played this?
Noah: It's a device that en
ables a to prevent a
" the checking
supported t from advancing
by checking the from in
distance only
front of the t .
applies to a Rook
If the enemy t is no further
In
firont "
than the fourth rank, as it is
GM Tall
here, there's plenty of room
in front of it for a to check
and keep the enemy from
h8t S. g6 g8t 6. Wh7 and 7.

g6.
Adams-Dreev
Dehreczen 1992
White to play
1. a h5!
The only try. Otherwise 1... f7
guarantees Black the drawn Phil
idor position.
1.
a d8!
The seemingly "more active"

1... d 1 2. fS h l t?? allows the
1 to advance decisively (3. gS
g1 4. WhS h l t S. Wg6 and 6.
gS).
If the pawn reaches the fifth,
White can force Lucena. E.g., 2 ...
d4?? 3. WhS d8 (too late) 4. gS
13 1
Rooks
Noah: No one. It's a study by
B 147
a famous composer named

Nikolai Grigoriev.
He showed that White can

"'1<' 'II %
%
win-even though Black has
- ;.
"the distance"-by cleverly

seizing the a-file for his W's

approach.
Pat: So?
White to pllly
Noah: What he missed is the
1. c6?! a b8
main distinguishing feature
2. a6?! d5
of the position: That Black's
3. a4
wc4
W is cut off by rank, not just
4. c6t Wd5
by file. So 1. h5 wins much
5. b5
a8
faster.
6. b4
Pat: So much for the great
And White wins.
But 1. !! h5! makes 2. b5! un
Grigoriev.
Noah: Worse than that. The stoppable and therefore saving a
same study was used in the bit of time and energy.
Even 2. !! c5 and 3. !! h5 wins
most respected book on the faster than the book line.
subject, RookEndings, by Lev
enfish and Smyslov-with no
mention of 1. h5.
"The simplest Rook

endgames contain a
lot oJsurprising little
secrets.
"
-Mikhail
132
Botvinnik
Chapter Seven
1 48
the last note.
4. xa4
5. d8!
6. b4
g4
g2
g3
Yudasin-Zlotnik
Ponfirrada 1992
White to play
1. b8!
Much better than 1. b4 xg3
2. xa4 which prevents the white
from holding up the g- i's ad
vance (2 ... g4!).
1.
Kg6
Mter 1... xg3 2. Wxa4 g4 3. c4
fJ 4. g8! the slows the g- i
while White promotes on the other
wmg.
2. c8!
3. c4
f7
xg3t
Otherwise 4. Wxa4 wins as in
7. d3!
a2t
Or 7... gl 8. Wb5 g2 9. d2
and c5-c6 wins.
8. b5
9. g3!
g2
Not 9. d l f2 (threat of
... fl) 10. gl We7 and Black has
two pieces to fight the connected
is.
9.
e6
10. c5
And wins after 10 .. b2 I I. c6
.
Wd6 12. g6t.
Pat: Aren't there any super

moves that work in every t!
ending?
Noah: I'm afraid not. Chess
isn't a multiple choice test.
White's task in Diagram 148,
for instance, is quite hard
even though he not only has
an extra t but has the better
placed t! .
Pat: But doesn't Black's t! get
in the way on the g-file?
Noah: Yes, and thanks to his
accurate first two moves, the
long-distance white t! is able
to make a crucial transfer.
Pat: Where?
Noah: At move 7. The t! can
not get to g8 where it would
slow the march of the enemy

t . So it goes to d3 where it
cuts off the black from the
-side and still slows the g
t.
Pat: And that's it?
Noah: No, there's one more
subtlety at move 9. But there
it's just another case of keep
ing the t! more active than
Black's.
Pat: I'm getting confused by
where or when the t! be
longs.
133
Rooks
Noah: No surprise. The role of
the t:! may constantly change
in the course of an ending.
In Diagram 150 Black cor
rectly places his t:! on d5,
where it cuts off the white
from the -(t-side and tries to
immobilize the enemy t:! by
tying it to the defense of the
c5- i .
Pat: So White trades is.
Noah: And makes a fine trans
fer with 4. t:! a3. The t:! can
then (a) cut off the black W
from the -(t-side, (b) shield
White's own W on its way
left, and (c).
Pat: What's (c)?
Noah: That's (c), defend the
W-side singlehanded against
Black's only source of coun
terplay-an attack by his <it
and is.
134
Pat: What is Black's t:! doing

in the meanwhile?
Noah: It was correctly used to
prevent the W and i (S ...
t:! as!) from advancing.
Pat: But it looks like he missed
something later.
Noah: Something big. In Dia
gram 151 he could have pen
etrated along the d-file-an
active plan. Instead, he kept
his t:! flexible-yet passiveon as. This allowed White to
use his and t:! to advance
the i to the fifth rank.
Pat: And Black's only chance
was to find an entry on the side for his W to gobble is.
Noah: A good plan but easily
stopped by White. It may
seem like magic that White
makes progress in moves 1621 and Black doesn't.
7. a3
8. w,>c2
h5
a8!
This gains checking distance to

stop the threatened 9. b3 b5t
10. c4 a5 1 1. b4 and 12. a4.
9. b3
b8t
10. a2
Karpov-Knaak
Baden Baden 1992
Black to play
1. . . .
d5!
Not 1... d4 or any other
move because of2. a6! (2 ... d7
3. e2 and White makes progress).
2. a6
xc5
3. xa7t f6
4. a3!
Now 4 ... c l t 5. d2 a l
gets the behind the i but it will
be ousted after 6. c2 and 7. b2.
4.
5. d2
6. c3!
g5
g6
a5!
10.
a8?
Much better was 10... d8!,
threatening 1 1... d2t. Then 1 1.
c2 allows 1 1... d3! and the W
cannot advance with his a- i .
1 1. c4!
5
12. a4
w,>f6
13. Wa3
e5
14. c5t! we4
Chapter Seven
h4
15. as
1 6. Wa4
Wf4
Not 16 ... g4 because of II hxg4
fxg4 18. h5.
1 7. c4t
We5
18. b4!
Wd5
1 9. b5t
We4
20. b6
Clears the way for 21. a6. Now
20. . g4 again allows 21. hxg4 fxg4
22. h6.
20. . . .
4
2 1 . a6
g4
22. waS
But 23. b4t! kills the idea

22.
g3
23. b4t! We5
24. fa
Black resigned after 24'M f4 25.
E! e4t WfS 26. e2 and 27. <;t'b6.
Pat: Not magic, just tactics.

Noah: Well, to quote a wise
Once White had tactically

killed off all counterplay his
opponent threw in the towel.
Pat: Does the defender have to

be just as flexible?
Noah: Sure. He usually has

even more useful things he
can do with his t! than the
player with the advantage.
Pat: I don't see that at all.

Noah: Try Diagram 153 on for
size. You tell me what should
"Good
good tad'!
is
happen.
-Vasser Seirawan
135
Rooks
Pat: Okay. First Black tries-to
get his behind his i .
Noah: Correct. That allows
his W to fight the -side is,
a fight he can hold (3. g4)
even when it's a case of
and two connected is against
a mere .
Pat: But White crosses him up
with 3. b6t, which cripples
both the and i .
Noah: And what else does that
do?
Pat: It forces Black to use his
to break the b6-blockade.
Noah: Right again. Once the
blockade is broken the W can
return to the W-side.
Pat: Then White can win if he
cuts off the black with his
.
Noah: But notice how differ
ently Black uses his . At
136
move 10 it swings back to

provide a bridge for the to
fl.
As usual the defender's is

best stationed in front of the
enemy is. And that frees the
black for one last duty.
Pat: I see. The checks along
the b-file. White can't take
his own away from the
first rank or Black queens.
Noah: And if and when he
removes his far from
checks, Black attacks the
side pawns with his , creat
ing a final mismatch.
Pat: Aren't there any kind of
general rules that you can
learn instead of trying to re
member all these key posi
tions?
Noah: Hmmm. There is one
mildly useful rule for + i
endgames. But almost no one

west of Kiev and south of a
2400 rating knows it: the
Rule of Five.
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
_
Dvoiris-Novikov
Polanica Zdroj 1989

Black to play
1. . . .
g8!
Mter 1... bS? 2. b2! Black's
is immobilized and he loses (2 ...
<IIe6 3. g4 <IId 6 4. <IIg3 <IIcS 5. h4
gS 6. gS ,&>c4 7. g4 b4 S. hS-a
<11 +2 1 vs. mismatch).
2. b2
3. b6t!
b8
On 3. g4 bS 4. g3 b4 S. h4? b3
6. <II f4 b4 t or S. b3 '&>g6 6. h4
<IIf6 White can be stopped.
'It'e5
3.
Not 3 ... '&>g7? 4. g4 and Black is
Chapter Seven
fatally passive, or 3 ... <ligS 4. g4
<lif4 5. '1t>g2.
4.
5.
6.
7.
8.
9.
g4
g5
abl
h4
\flg3
h5?
\fld5
\flc5
b5
\fld6
b4
The last chance for a win was

the cutoff, 9. e l !. If9 ... b3 10. hS
b2, then I I. h6! b l='/t 12. xb l
xb l 13. h 7 h l 14. g6 and
queens.
9.
10. an!

b l='/t 13. xb l xb I 14. h8='/t

or, 12 ... h8 13. g6.
1 1. elt
Of course, I I. xf8?? loses the

<Ii+ t race.
1 1.
12.
13.
14.
. .
g6t
\flg4
\flg5
\flO
\flg8
b3
b8!
And the game was drawn after

15. h6 b5t 16. <liffi b6t 17.
<lif5 b5t 18. <lie4- b2 19. d4-!
(mate threat) b8! 20. b 1 b6!
21. '1t>c3.
\fle7
a f8!
Not 10 ... b3 I I. h6 b2 12. h7

137
Rooks
Pat: Never heard of it.
Pat: I'm absolutely lost.
Noah: It applies only to end Noah: You shouldn't be. Look
ings with one t . The rule at Diagram 155. Take the rank
states:
of the t . That's four. Add it
to the number of files the W's
cut offby. That's two. It makes
RULE OF FIVE
D ,\:-sv;,
1:"
_
&<d.
Add the number

of the rank of the
pawn to the num
ber offiles the
defender's King is
cut off. /fthe sum
is more than five
it's a win.
138
Pat: Six. So it's a win. But what

if the black W isn't cut off?
White to play
Noah: The rule doesn't apply
1. c4
c8t
to any position in which it
Black has the checking distance.
isn't-because the defender
can force the Philidor posi Otherwise the Wi promotes the t
with 2. d5, 3. \t>c5, 4. d6 etc.
tion.
2. b5
d8
Pat: A dead draw.
3. c5
c8t
4. b6
d8
This is classic defense by Black,
checking the <eft away, then attack
ing the t .
__________________
____________________________________
Chapter Seven
Schmittdiel-Brennjnkmeijer
Groningen 1991
White to play
I.
:xe4t??
With 1. c5t and 2. h5 White

kills the mate and should draw.
..
xe4
2. :dl
e5
3. : elt f4
Resigns
I.
Black could also win by 3 ...

Wd4 4. d l t c3 5. e l g5
and 6... d2. But White resigned
when he saw the faster 3 ... f4 4.
fl t We3 5. e l t r2 and 6 ...
h6*.
Noah: Exactly. Like most rules, Pat: Can I count on this Rule of
the main benefit of the Rule of Five always working?
Five comes in transitions: it
"Rules arefor when
tells you when to trade t!s or
brains run out "
is to reach a simpler posi
-Anonymous
tion.
Pat: Don't you still have to Noah: Of course not. It
calculate positions out?
wouldn't be a rule if it didn't
Noah: Of course-if you have have exceptions. It should
time to do it. Otherwise let only be used as a guideline.
the rule be your guide. In
For example, it never works
Diagram 157 White only saw with t! - is, and there are a
that mate was threatened on few times it doesn't work with
the h-file and assumed that 4)- is. In fact, Andre Cher
he would draw thanks to the on, a famous French analyst,
checking distance.
started with a Rule of Five,
He couldn't-so he traded then modified it to a Rule of
straight down to a dead loss. Six-and then gave up when
Pat: Because his was so
he found there were too many
useless at hI.
exceptions.
Noah: Once more the player
Still, the Rule of Five is a lot
with the more active tri easier than memorizing Basic
umphs.
Chess Endings.
"Set even a strong club

player down to play a
1IUlSter in a Rook and
pawn ending and give one
player a pawn advantage.
Every time the 1IUlSter has
the extra pawn he will
win; every time the club
player has it he will draw
(unless he manages to
lose). "
-C.H.O'D. Alexander
lJ.tJ
Rooks
Pat: Lemme try Diagram 158.

Third rank plus three files,
makes six. It's a win, right?
Noah: Very good. The advan
tage of knowing it's a win
140
gives you a huge psychologi

cal edge. You still have to find
the win, but you know if you
look hard enough, it'll be
there.
Pat: I've had enough of look
ing hard today.
Noah: Just try to remember
the important things-the ac
tive a , the possibilities for
mismatch, the best place for
the a .
If you can also remenlber
stuff like the checking dis
tance and what Lucena looks
like, you'll be better than most
2100+ types.
Besides there is a silver lin
ing to studying a endings.
Pat: What's that?
Noah: Almost every other type
of ending is easier-as we'll
see tomorrow.
to the fourth rank.

But now 2. b4 f5 forces the !!
to give way: 3. !! d4 e5 or 3. !! c4
e6 4. b3 d7 5. ,&>a4 !! b8!
and now 6. b5 !! cB! draws.
f5
2. c3
3. e3
4. e 1
f4
Resigns
Vaisman-Adamski
Bucharest 1981
White to play
1. e4!
Black's !! would be better
placed at bB but even here he has
the checking distance. White can
not make progress with 1. c3
because of 1... !! c7t 2. d4 !! b7
or 2. b4 !! b7t 3. c4 !! c7t 4.
d5 !! b7 5. !! e3 !! bB (checking
distance) 6. c5 !! cBt 7. 'tflb6
!! bBt B. c7 !! b4.
1.
f6
White was threatening to use

his !! , not his , to advance the t
__________________
____________________________________
Black threw in the towel be

cause of 4 ... !! c7t 5. ,&>d4! !! b7 6.
,&>c4 !! c7t 7. d5 !! b7 B. !! b l !
and the t advances undisturbed.
In which Pat learns the two distinct types of'll1 + .ft endings
and why longer isn 't the same as harder.
(if{:}t
[!{fljJt
I
Why
Queen endings seem hard

In a book called Practical Chess Endings, Paul Keres
devoted 10 of the 35 pages on -endings to +
vs. + ft (!) and another 7 pages to + vs.
\t7+ (no pawns) or vs. \t7+ other pieces.
In 600 Endings, Lajos Portisch devotes 76 to
endgames-but most of them to the same endings as
Keres.
And in QJleen Endgames, Yuri Averbakh spends 25
of the 133 pages on \t7+ vs. w+ ft (or fts).
..
Queens
The is now close enough to
Pat: I hope what you have Noah: They're simple because
the * to be shielded from checks:
today, Noah, is easier than the player with the advan
6. *c6t d2 7. *dSt c 1 8.
s.
tage usually has only one or
*gst <ctb l 9. *fSt *c2! and
two ways to win: He can pro
Noah: Much. Today we do 'l!1
Black wins.
mote a i or he can trade 'l!1s
endings.
Pat: That's easier?
down to a won <tt> + i endNoah: People only think lng.
The other plans and tech
they're hard because the end
game textbooks are filled with niques you often see winning
Korchnoi-Lobron
all sorts of arcane and virtu an endgame-including zug
Biel 1993
White to play
zwang and triangulation-are
ally useless information.
fairly rare with 'l!1s on the
Pat: Useless?
1. 'l!\'d7t f6!
Noah: I'd say \t>+'l!1 vs.
board.
2. 'l!\'d8t e6
<tt> +24)s is about as useless as Pat: What about the other guy?
3. 'l!\'b6t
Other checks make little differ
you're likely to find in a $28 Noah: The defender's job is
also fairly simple-he either ence since Black's reaches safety
book.
Yet people who master the tries to deliver perpetual around his 'IfI. For example, 3.
'IfIg8t <ctd6 4. 'IfIf8t <ctdS s. *fSt
real 'l!1 endings say they're check or create a passed i . <ctc4 6. 'IfIf4t? *d4t! or 6. *e6t
among the easiest to leam Other defenses, such as sacri d3 7. *fSt c3 as in the game.
and perhaps the simplest to ficing your remaining piece
3. .
d5!
4. 'l!\'xa5t we4
to stop your opponent from
play.
5. 'l!\'a6t c3
queening, clearly don't work
Pat: I don't see simple, even in
Resigns
positions like Diagram 160.
in 'l!1+ i endings.
.
142
Chapter Eight
Kaidanov-P. Nikolic
Groningen 1993
White to play
1. xc6?
White wins soon after 1. e6.
1.
d4!
2. e6
xb2
Now 3. e7? allows 3 ... *al t 4.

Wh2 *e5t and 5 ... *xe7.
3. g3
e5
Both sides have passed is and

after 4. *d7 b3 both are on the
sixth rank. A draw was agreed soon
after both sides made a second *'
Pat: How about rules? Are Noah: And sometimes they

there any that work only in last 40 or 50 moves until one
* endings?
player runs out of checks or
Noah: No iron laws, I'm afraid.
the other one runs ot of pa
But it helps if you understand tience.
this general guideline:
Pat: That's me-the one with
out patience.
Material In + i s Isn't As
But tell me: What do I really
Important As How Advanced
have to know about *-end
Your i s Are
ings?
Pat: Sort of like "Pawns In
crease in Value as They Ad
vance."
Noah: True. With a passed t
"Patience: A minorfo'l71l of
on e6 White would have won
despair, disguised as a virtue. "
Swiftly in Diagram 161. But
-llnlbrose Bierce
picking up the irrelevant c- t
let Black escape with a draw.
Pat: I guess the reason I find
*+ t endings so confusing
is that there are so many
checks and so many t situa
tions.
143
Queens
does in Diagram 162 after 1.

Noah: What you really need
to know is that there are only \fj>f2 and so on up to 6. \fj>e4.
two basic types of 'lI1 end Noah: You would ifyou played
a few type No. 2 'lI1 endings.
ings-and how to play each.
The No. 1 approach, such as
They are:
No. 1 : The guy trying to win
1. 'lI1e6- to break the block
hides his \fj> from checks while ade on d6-fails. So White
only has one other way to
his 'lI1 does all the work.
Pat: That sounds easy enough.
make progress: bringing the
What's No. 2?
\fj> to the support of the d- t .
Noah: In No. 2 the guy with Pat: But why should he win
the edge parks his 'lI1 on a after Black manages to get
good centralized square while back to equal material?
his \fj> advances to help out. Noah: He wins because of one
Pat: Hmmm. I'd never think of the basic rules of 'lI1+ t
ofbringing my \fj> out into the endgames:
center ofthings the way White
Rank Has Its Privilege
Or to put it another way, it's
usually better to have, say, a
passed t on the sixth rank
than to be a t or two ahead.
144
5. <t;f3
6. <t;e4!
*f6t
*h4t
White finally had a threat, 7.

f1d4. No better was 6 ... f1f5t 7.
<ltd4 f1f4t 8. <ltc5.
7. g4
8. <t;d4
9. <t;c5
Rashkovsky-Krasenkov
*elt
*alt
*xa4
163
Cappelle la Grande 1990

White to play
1. \t>f2!
Usually a f1 trade will win when
you're a i up. But here 1. f1e6?
f1xe6 and 2 ... <ltf6 or 1. f1d4 t f1f6
2. f1xf6t? <ltxf6 and 3 . . <lte5 kills
the t .
1.
*f6t
White's <It is headed for e4 or
d4 and Black has nothing but
checks.
.
2. <t;e2
3. <t;e3
4. *d3
End of stage one: White has

centralized his <It at the cost of a t
less important than the d- t . Now
10. f1d4 t?? f1xd4 t I I. Wxd4 <ltf6
boomerangs for White.
10. *c3t
<t;fl
1 1. d6
More accurate, now or a move
Chapter Eight
before, was c&>b6 and d5-d6.
'lM4t
1 1.
12. d7
b5
No more checks, so Black has

only one other defensive idea.
13. d6
14. i;Yc7!
b4
Now 14 ... b3 10ses the 1 to 15.

c&>c6t 'l1?g8 16. *b8t.
Black played 14 *e4 but re
signed after 15. c&>c8t c&>f6 16.
*c5 ! in light of d7-d8=*. After
16 ... *a8t or 16... *e8t White
ends the checks with 17. C&>C 7.
Pat: But White's d- t is still on

the fifth rank in Diagram 163
and he's lost a t on the 'I!!
side.
Noah: And that means Black
has a passed b- t . All true.
But White has made huge
progress between the two
diagrams. The advance of the
c&> creates a 3-to- l mismatch.
T h e c o m b i n ati o n o f
c&>+'1!!+ t beats Black's 'I!! .
745
Queens
Pat: Aren't you afraid of get
casual with your in an end
*c5t 9. f6 *d4t 10. wgS Wc7
ting mated when you bring ing-but you gotta be so care
11. *t7t c6 12. e7 *glt 13.
your into the center?
ful in the middlegame?
h6 *h2t 14. Wg7 in view of
Noah: No. Mates are very, Noah: Because a can't mate
14 ... *g3t 15. *g6t or 14 . . *b2t
very rare in endings. As by herself. In fact, the ac
15. *f6t.
tually declines a bit in power
another wise man once said:
in the endgame when com
"Nobody ever died of a
pared to other pieces-and in
check. "
particular to the <tf}.
Piket-Ljubojevic
-Anonymous
Pat: How so?
Monaco 1994
Black to play
Noah: The strength of a is
White's dances all over enhanced by its ability to at
White's '1ft is not badly placed
the board in Diagram 164 tack two targets at once. For to promote a -side t . So. . .
example, a at d4 can
and lives to tell the tale.
I. . . .
*dlt
2. c4!
*xa4t
Pat: How come you can be so threaten a t at a7 and an
3. d5!
*d7t
other at g7-as well as enemy
*c6t
4. '1fle4
pieces at a4 and h4 and gl.
5. '\fIf5
*xc3
But in the endgame there
Or 5 . *d7t 6. e6 and 5 ... *c5
are fewer targets, so the of 6. g6.
fensive power of the de
'1flb6
6. e6
Or 6 ... *d3t 7. f6t and 8. e 7.
clines a bit-while that of a ,
a5
7. f4
no longer afraid of mate, in8. *g6!
creases.
146
___
______
__
__
__
__
__
__
__
__
__
__
__
__
__
____
__
__
__
__
__
__
__
__
__
__
Chapter Eight
fld5t Wa4? 9. Wa2 ! threatening
10. flb3=H=.
Mainka-Vaganian
Bundesliga 1992
Black to play
1. .
2. "rxh6
"re4!
f4
Rank has its privilege.
3. "rg7t
The best defense now is a per

petual check. Clearly too slow is 3.
h4.
dbc
e 6.1
Seeing that 3 ... Wa6 4. flf7! fJ

5. h4 and White may have per
petual checks once the black fI
clears a path for the f- ft , e.g., 5 ...
fle2 6. fld5! t2 7. fla8t Wb5 8.
Pat: I think I was better off type No. 1 position-at move

before I learned about the 3-he wins in just six moves.
4. "rg4
Wd6!
two types of'Q/ endings. How His perfectly centralized 'Q/
Threatening 5 ... fie 1 t 6. Wa2
am I supposed to tell them does the trick by shielding
fle6t!
apart?
him from all sorts of checks.
5. "rh4
Wc5
Even on a square like d4 the
Noah: Well, ifyou had a choice,
6. Wa2
which would you try to win is safe and when it reaches
with?
e2 there is no way of stopping
Pat: Me? I'd take the one where
the promotion of the f- ft .
the W is safely covered with Pat: Okay, so the perpetual
fts and I only have to move check defense is crucial. What
the 'Q/.
else is unique about 'Q/ end
ings?
Noah: That should tell you
something: If you can't safe
guard your castled W posi
6. . .
Wd4!
tion, then it's type No. 2-as in
7. "rf6t
'ti7d3
Diagram 165.
8. "rd6t
Or 8. flc3t We2 and White's Pat: The trouble for me with
out of checks.
type No. 2 is that I hate to
8. . . .
e2
w
give my opponent hours of
9. "rc7
f3
checks.
Resigns. White's only defense
fI checks-run out after 10. flxb6 Noah: You won't. After Black
figures out he doesn't have a
fld5t and 1 1 ... t2.
147
Queens
Noah: Well, one thing to know

is the significance of diagonals.
,....-----...
"All diagonals are created

equal-and that lasts until
White'sfirst move.
"
-Anonymous
Queen-endgames
everything is different "
"/11,
-Mikhail
148
Botvinnik
That's shown in Diagram

167. Mter three moves White
causes Black to resign-because he seizes the right di
agonal.
Pat: Is it so obviously hopeless
when he gives up?
Noah: See for yourself. Once
White nails down h2-b8,
Kasparov-Salov
Black has no checks.
Dortmund 1992
White to play
And without checks, he has
no counterplay. The only
1. *xa7! *cl t
other defense would be to
*xc4
2. 'i7h2
blockade the b- t . That's a
Or 2". '/Wf4t 3. Wh l '/Wc l t 4.
second best defense and it '/WgI ! and then 4." '/Wxc4 5. '/Wb l !
fails badly here because followed by 6. b6, etc.
3. *b8!
Resigns
White also controls the three
There is no defense to 4. b6 CIlld
squares in front of the t .
5. b7 now.
Pat: I see that-but what about
other cases?
Mter all, a fI travels on
diagonals and files. Why are
diagonals more important?
______
___
_________
___
___
___
___
___
___
___
___
___
___
___
___
___
___
___
___
___
___
____
Chapter Eight
9. h5 *f6t 10. e5 *flt (e.g. 1 1.
<&>c4 *f1t and queens).
Hiibner-Karpov
Baden-Baden 1992
Black to play
1. . .
-lftd4!
Now 2. *b8t Wg7 accomplishes nothing {3. *a8 f5t!}.
a5
2. <1tlf3
3. -lftc7
4. -lfta5
a4
Otherwise Black wins immedi

ately with 4 . .. a3 followed by 5 ...
*b2 {diagonal!} and 6 . . a2.
.
4.
5. <1tlf4
-lftd3t
a3
And White conceded after 6.

h4 *h3 7. w e5 *e6t 8. Wd4 a2
Noah: Well, in Diagram 168 it

should be easy to see that
there is one line of squares
that's much more important
than any other.
Pat: You must be talking about
al-h8 because it controls the
queening square for Black's
a- t .
Noah: Not bad. You nlay have
a future in '* endings after
all.
And yes, that diagonal also
controls a key checking line if
and when the black W is
driven to g7.
Pat: White's moves look pretty
feeble. What was he doing?
Noah: The best he could.
When you don't have checks
or a passed t there's not
much a defender can do.
Black's a- t goes from the
second rank to the eighth in

1 1 moves, pretty fast for a '*
ending.
149
Queens
Pat: Okay, so diagonals are a
big deal. What about mis
takes? What are the major
blunders I have to avoid in 'JJ1
endings?
Noah: For the player with the
edge the principal errors are
allowing perpetual check amd
failing to create and push a
passed ft .
For the defender, it's the
opposite-missing a chance
for perpet, for instance.
And one of the ways both
sides can err is to decentral
ize the 'JJ1 .
In Diagram 169 White wins
because Black puts his 'JJ1 on
the side of the board, where it
can neither deliver perpet nor
advance the a- ft .
Moral: The center is a very
nice place for a 'JJ1 to be.
150
after 4 'I/1g5t 5. Wh3 'I/1e3t 6. g3

'I/1el 7. 'I/1f4.
169
Bareev-Cvitan
Tilhurg 1993
Black to play
1.
h5t??
With 1... a4! Black has excellent
drawing chances despite the ad
vanced h- it .
2. wg3
a4
3. c6t
Not 3. 'I/1g7 '1/1f7! and Black is
alive.
3. . . .
we7
4. xa4
And since 4 ... 'I/1xh6 allows the

fatal 5. 'I/1h4 t !, Black had to play
on two its down. He resigned soon
Chapter Eight
p1 170
Rivas-Littlewood
Hastings 1981-2
Black to play
-lte4?
2. dSt! xdS
3. -ltxdSt Wh7
4. -ltdl
1.
Thwarting checks and preparing al and a4-5-6-etc.

4.
fS

S. a4
6. -ltc2
-ltb4
-ltaS
The blockade is easily broken

in most endings. Here White's
idea is b5!, breaking the block
ade.
7. -ltc6!
S. g2
eS
-ltel?
Pat: Why is that? In end

swapping?
ings you said the defender Pat: I'd say he's better offtrad
Relinquishing the blockade and
and sometimes his opponent ing *s, like with 1. .. *c 1.
losing quickly.
as well-is supposed to keep Noah: Very good. Either that
9. -ltdS!
-lte2
his on the fringes, like a8.
or keeping all the heavy
-ltc2
10. -ltf3
Noah: Right. But remember pieces on with 1. .. Wh 7.
I I . as
And Black soon resigned.
that the real power of a * is
The pseudo-active move he
to land on a square and sud chose only allowed White to
denly attack two things at trade down into a * ending
once. You can't do that much and a very favorable one at
if you keep the * at, say, a8 that.
or al.
Once White got control of
Pat: I also thought *+ it end
the key g2-a8 diagonal and
The blockade is
broke the a5-blockade, it was
ings were a lot like + it
easily broken in
endings-because they both all over.
most Yf1 endings.
get drawn alot.
Pat: But what happens when
Noah: Not nearly as much as you don't have any it pro
endings. Or for that mat tection for your W? How are
ter * endings with other you supposed to advance a
pieces.
passed it -and avoid per
For example, in Diagram petual-then?
170 which pieces, if any,
should Black be interested in
151
Queens
Noah: You have to use the one
crucial winning technique of
* endings.
And Diagram 171 illustrates
how it works.
Pat: Where's the black 1/?
Noah: Right now there is none.
Pat: How can that be?
Noah: To make a point.
This diagram is sometimes
used in books to show the
many different ways White
has to win-all depending on
where the black 1/ is located.
Pat: Interesting, but is this re
ally useful?
Noah: It can be-but the most
instructive thing about a po-
152
sition like Diagram 171 is

rarely if ever mentioned.
Pat: And that is?
Noah: That it can only be won
if a black W is on the board.
Pat: Excuse me?
Noah: You heard me. Usually
Black's only defenses in such
positions are checks and pins,
for example, 1. Wg7 *a7.
And White can only get out
of those checks with cross
checks. Black loses because
he has a <at.
Pat: But what if there are no
cross-checks?
Noah: There almost always is
one. The trick is finding
where.
A simple example is Dia
gram 172. How long do you
think it should take White to
win?
171
White to play
If the black is on h2, d2 or e2White trades s with 1. Qg2t.
If the black W is on a l, hl, el, dl
or e 1-White trades s with 1 .
gl t and 2 . g2t.
If the black 1; is on a3 or a
White trades with 1 . a6t.
If the black W is on fl,f2, f3 or
fWhite promotes with 1 . Wg7
al t 2. f6t! or 1... a7 2. f6t
and 3. WhS.
If the black W is on h 1, h2, h3 or
hWhite promotes with 1. g7
and 2. WhS because checks on the
h-file allow 3. h7t.

If the black W is on h4 or e3White establishes zugzwang with
1 . c6 or 1. e4.
If the black W is on eWhite
promotes with 1. Wg7 since there
is no check on al or b2 and 1 ... a7
allows 2. Wh6!.
Similarly, ifthe black W is on d4
or e3-White wins with 1. g7 or 1.
Wh7 respectively.
Chapter Eight
Again on 3 . . We7 White has 4.
e5t '\flfB 5. c4.
.
4. *f5t
Resigns
Because of 5. a5 or 5. b5.
Mirkovic-Nesic
Yugoslavia 1987
White to play
1. *d3t! we7
Not 1... We6 2. g6t or 1...
'\fle8 2. b5t.
Also 1... Wc8 2. f5t! '\flc7 3.
a5 and 3 .. Wb7 4. b5, trading
s.
Finally, 1... W;c7 2. c4t W;b8
3. b5 or 2 ... '\fld8 3. d5t We7
(3. .. We8 4. b5t) 4. e5t and 5.
c4.
2. *e4t
Wd7
Not 2 ... '\flfB 3. b4 t.
Pat: Let's see. With the it only

on the second rank and
Black's 1; close to the queen
ing square I'd say maybe 30
moves-if at all.
Noah: You're a bit off. White
has all sorts of ways of trading
'its, with cross-checks at b3,
b4 and b5.
And he doesn't want to es
cape to the \t>-side because
then the c- it advances at will.
Since any <it>+ it ending is
hopeless, Black throws in the
towel at the right time.
3. *d5t
Wc8
153
Queens
Pat: Amazing. It only takes
four moves.
Noah: 't!J endings usually are
longer-and amateurs often
mistake that for being harder.
Here at Diagram 173 is a
longer example. White be
gins with the premise that he
wins if he can trade 't!Js.
Pat: In all cases?
Noah: Maybe not all. But all
W+ i endgames in which he
keeps his a- i are won, even
if he loses both other is (2".
't!Jxg3 3. 't!Jg5t!).
What this means is that
White can play with two
plans-promoting and trading
't!Js. And to achieve the latter
he has to find a magic square
or two.
Pat: Magic how?
Noah: Magic because it will
154
allow White to meet a check

with a cross-check ofhis own,
thereby forcing a trade.
Pat: Which square is it here?
Noah: Figure it out for your
self. Black's <ttl is pretty well
hidden at h6 and can't be
checked on files.
But h6 also has two diago
Dlugy-Benjamin
nals and since we mentioned
U.S. Championship 1988
White to play
g5, then . . .
Pat: That only leaves f8.
1. '/hb5!
'/hc3t
Noah: Yes, so as hard as it may
2. '/hc5
'/hf6t
seem, White just has to figure
On 2 ... xg3 White has the
a way to cross-check on f8- winning check at g5.
3. '/h d6
'/hf5
and the game's over.
Now the cross-check to aim for
Pat: So the best place for the
is at f8: 3 ... c3t 4. Wd7 g7t 5.
defender's <ttl in these end WeS! gst 6. f8t.
ings is as far from the action
4. a7
'/he4t
as possible.
5. Wic7
'/hc4t
6. Wid8
'/hg8t
Noah: Usually. But having a
7. Wie7!
W on the board isn't always
a liability for the defender.
Now if Black wins the it (7...

g7t and 8 ... xa7) White forces
a won W+ ft endgame (S. We8
and 9. f8t Wh7 10. f7t or 9 ...
g7 10. xg7t Wxg7 1 1. We7
Wh6 12. f6 Wh7 13. Wf7 Wh6
14. \figS!).
7.
8. '/hd4
9. '/hd7
.
'/has
'/hb7t
'/hb4t
The blockade plan (9 ... aS) is

so passive even 10. Wf7, with zugzwang, wins.
10. Win
'/hc4t
1 1. WfB
'/hc5t
12. wg8!
Resigns
Chapter Eight
Because mate is threatened on
h 7, and the only defense to it is 12 .
c4t, after which 13. f7 c8t
14. f8t! reaches the magic square.
..
If he's close enough he can

help in stopping the enemy
from queening, as in a t!
ending-orjust about any end
Ing:
If the defender can get his
1; to the queening square,
it'll probably be a draw. Just
like your niend Philidor's po
sition.
Pat: But it must be a lot harder
in a ending because the W
can be checked away so eas
ily.
Noah: He doesn't always have
to reach the queening square
itself. In Diagram 175 it looks
like Black's is badly placed.
155
Queens
Pat: You mean because of pos Noah: At least they're not as
sible cross-checks, like on the bad as the books make them
second rank?
look. On the other hand, there
Noah: Correct. But the black
are endgames with . . . but
can perform other func
let's leave them for tomortions, such as elbOwing the row.
enemy aside-orjust block
ading the t itself.
Pat: In the end, Black nearly
175
elbows him off the board.
Noah: So much so that his 'iff
can operate at long distance
at h 7 or c2. White takes the
draw when his opponent's W
took part on the attack on his
t.
But, again, all you really
need to know is a few things
like cross-checks, the two
kinds of endgames, and the
importance of diagonals.
Pat: Okay, I admit 'iff endings
aren't as horrible as I thought.
156
White's 't out of play-and 5.

'txh6?? 'ta6t takes it off the
board}.
2. 'IIIc5
3. c7!
'IIIa6t
e2
l!:J
There was no other useful move,

and 4. 'tc6t was threatened. Now
on 4. 'td6t Black can't approach
the queening square (4 ... Wf7 5.
'tg6t We7 6. 'tg7t WeS 7. 'txh6
'tc4 t 8. 'tc6t!).
4. d5!?
'IIIxh5
5. 'IIIf8t
e5!
Now Black's 't is out of play

and he would lose quickly after 5 ...
'It;g5 6. d6 and 7. d7.
6. d6
'IIIh2!?
176
Machulsky-Smirin
Pula 1989
White to play
1. 'IIIc7t
<ltf6!
Now on 2. 'te7t Wf5 3. 'te6t
Wg5 the black W provides coun
terplay (4. 'tg6t Wh4 takes
Now 7. 'teSt wd5 S. 'tc6t

We6 9. 'te4t drives the W away
and paves the way for the d- l .
7. d7??
8. wc8
9. 'lfld8
We6t
'IIIc2t
'IIIh7!
And White cannot make pro

gress. The game ended with 10.
'teSt Wd6! and adrawwas agreed.
In which Pat learns that not all .ft.s are created equal-and
why chess im't like golf
(jf{;JI
j
Wfi 'i?'
Pawns
are more like the par-seven

holes of chess-they're rare
and very hard to prepare for.
The reason is that W+ i
endings are 90% calculation.
No matter how much 'book
theory you know, even a
world class player can err
badly, as White did in Dia
gram 177.
After he miraculously saved
the game the spectators were
sure White had blundered (he
had). But halfofthem claimed
his error was 1. 'l:17e2 instead
"Pawn endings are
of 1. Wc3 (it isn't).
to chess wha "ing
And the other half correctly
is to golJ "
-Cecil Purdy
blamed 2. gxffit.
As Anand said after they
Pat: Yeah, like that.
shook hands "In this game
God helped me."
Noah: Actually, i endings
Pat: I could use some divine
help in my endgames.
Pat: I don't know if I'm up to

anything heavy today, Noah.
You know, midterms and
stuff.
Is there any endgame you
can sum up in a few minutes?
Noah: Sure, + i endings.
You already know my advice
here: Avoid them.
Pat: Why? Everyone says
they're, oh, you know, the
most basic endgames of all.
Noah: You mean...
158
bowed out.
1.
f3t!
The best try in a lost position.
2. gxf3 t??
Smirin-Anand
Moscow 1994
White to play
Some $5,000 was at stake in this

game and White should win with
out much difficulty.
1. e2
Now 1... We5 2. WfJ Wf5 re
gains the opposition for Black but
White wins with 3. h4 h5 4. g3 fxg3
5. Wxg3!
Things are obvious then after
5 ... We4 6. f4!
The key line is 5 ... We5! 6. WfJ
Wf5 7. We3 Wg4 8. f4 Wxh4 9.
WfJ! and Black's W is fatally el-
Here 2. Wfl ! would have won:

2 ... fxg2t 3. Wxg2 Wf4 4. fJ h5 5.
h4 We5 and now 6. Wfl ! leads to
the same won position as in the
last note.
For example, 6 ... Wd5 7. We2
Wd4 8. Wd2 We5 9. We3! Wf5 1O.
f4 Wg4 1 1. We4 Wxh4 12. WfJ.
f4
2. .
3. el
\t>xf'3
White can no longer win. A
draw was agreed soon after 4. Wfl
h5 5. wgl Wf4! 6. wg2 h4 7. Wfl
W.
Chapter Nine
g6t 7. Wf6 b6! with zugzwang.
Or 5. WfJ Wg5 6. wg3 g6! 7.
Wi'f3 Wh4 8. Wf4 g5t 9. Wf5 b6,
ditto.
Vlahovic-Pikula
Yugoslavia 1993
BlIlck to Pilly
1. .
wg5??
2. e5!
Wg6
Now 3. Wi'e4?? would give Black
a second chance.
3. We6
wg5
4. Wfl
Wh6
5. h4
Wh7
Drawn
But in the diagram Black is win

ning with 1... We6!. For example,
2. g4 hxg4 3. hxg4 Wf6 4. Wi'f4 Wi'g6
and White has no good move.
For example, 5. g5 Wh5 6. Wi'f5
Noah: Actually, '\t>+ i end

games should be very easy.
There's only one way to
win-to queen a i -and very
few techniques.
There are no mismatches,
for example, when the only
pieces on the board are the
s.
But + i endings tum out
to be notoriously difficult to
evaluate.
Pat: Can't you just count the
is?
Noah: Sure, and an extra i or
two is almost always deci
sive.
But the endings in which a
player is most likely to mis
take a win for a draw-or a
loss-are equal-material i
endings.
Pat: You think?
Noah: Sure. Look at Diagram

178.
Black jumped at the chance
to take a draw by repetition
never suspecting that he not
only had the edge but, in fact,
had a dead-won game.
He grossly overestimated
the power of White's central
ized W and underestimated
the benefits of his "pass"
moves.
Pat: But Black still has to find
a lot of good moves to win.
Noah: Yes, but the point is he
didn't even think of looking
for them.
Pat: Okay, so he was tired. He
was hungry. The dog ate his
copy of Basic Chess Endings...
159
Pawns
Noah: I know, there are a lot Pat: Doesn't that depend a lot
of excuses in chess. But what on other factors?
explains Diagram 179? Any Noah: Sure. If you're half a
one who can see up to 3. h4- dozen pawns up, the opposi
and count up to seven-can tion is not going to matter.
But in any W+ i ending
realize the position should be
a draw.
whose outcome is in doubt,
Yet White resigned even the opposition is a very high
though he could have ana priority. It's basically the sim
Klovan-Elistratov
lyzed the adjourned position plest and most powerful dem
Spartakiad 1963
Black to pilly
for hours. The other reason onstration of elbOwing out.
W+ i endings should be
Before Black could seal his
easy is there's only one golden

move, White resigned (!) because,
rule to remember:
Pat: I take it 180 is all about he said, it was "clear" that Black
wins.
opposition.
Actually 1... WcS 2. fS Wxc4
Noah: Yes. It illustrates a basic
Never give up the
(or 2 ... <alb4 3. cS!) and now 3. h4!
point:
is a draw.
opposition
White can't win by making
Both sides queen after 3 ... b4
(without a good
a passed i because it is eas 4. g6 Wxa4 S. xhS b3 6.
reason).
ily blockaded, such as after 4. g6 a4 7. hS and t!le result is a
draw.
g5t?
160
Chiburdanidze-Watson
Brussels 1987
White to pilly
1. Wl??
Either 1. g4 <aleS 2. e3 or, the
other order, 1. We3 eS 2. g4
keeps the opposition and wins.
Mter 2 ... Wf6 3. <alf4 g6 4.
WeS g7 S. fS <alf7 White can
finally play 6. hS! because his
gets to g6 (6 ... g7 7. e6 h7 8.
f6) .
There's no saving grace in 1. g4
e6 because of 2. e2! and if 2 ...
d6 then 3. f3 (3 ... <ale6 4. We4;
3 ... <aleS 4. e3).
Chapter Nine
1. . . .
e7!
Black can draw now by taking
the opposition after 2. g4 Wf7! (or
block further progress after 2. Wg4
Wf6).
Wle6!
2. Wlf4
Not 2 ... \fj>f6??, because White
gets the opposition back with 3. g4!
3. g4
4. <Ilf3
f6
Of course, not 4. gst hxgSt S.

hxgSt wg6, a book draw.
4.
we7??
The draw was there with 4...

Wf7 (S. We3 We7).
5. e3!
Wlfl
6. d4!
Keeps the opposition (6. We4?
We6!).
6. .
7. Wd5
f6
e7
Or 7... Wf7 8. WeS We7 9. Wf5

Wf7 10. hS!.
8. e5
9. f5
1 o. e6
1 1. h5t!
<Ilfl
Wlg7
<Ilg6
g5
Or 1 1... wg7 12. We7 Wh7 13.

Wf6 as in the last note.
12. Wlfl!
And Black resigned after 12 ...
Wxg4 13. Wg6 Wf4 14. Wxh6.
One rule of thumb worth

remembering is you have to
have the opposition in front
of the is before you can
afford that kind of liquida
tion.
Pat: And pushing the h- is
doesn't help anybody.
Noah: Right. If Black plays
... hS he makes the i a target
for the enemy and gives
White access to the gS square.
Ditto, White. Ifhe plays h4hS too early he gives away gS.
Pat: So what's left?Just a lot of
jockeying for position.
Noah: Correction-jockeying
for opposition. Both players
managed to give it away with
blunders.
White loses the opposition
with his first move, but Black
returns the favor at move 4.
Pat: And White wins because...

Noah: ...as yet another wise
man once put it:
"The winner is the

player who makes
the next to the last
blunder. "
-Savielly Tartakower
Pat: It seems like any time a

master makes a blunder in a
+ i ending it gets pub
lished.
Noah: That's because people
believe the lie that this is a
"simple" ending.
One reason you see so many
blunders is that even masters
may not appreciate the dif
ferent values of is.
Pat: What different values? A
i is a i .
16/
Pawns
Noah: No, a protected passed

i is no mere i . In fact, it
alone is usually enough to
win a + is ending, like
Diagram 182. Black's c:ti is
immobilized by the con
nected passed is after 1. h6
Pat: What about other kinds of
is?
Noah: Next in value comes
outside passed i s-like
162
Black's c- i in Diagram 183.

Reuben Fine set down the
rule that an outside passer
wins if the enemy also has
other attackable i s-like
White's W-side is in the 5.
We4 line.
Then c o m e s norm al,
healthy is. And finally, the
weakest is are doubled and
backward ones.
Pat: Are they equally weak?
Noah: No, a pair of doubled
is on, say, a g-file can stop a
pair of healthy f- and g- is
from creating a passer.
It's tricky to evaluate some
is. Nevertheless, most 1700players know better than to
do what White did in Dia
gram 183,justongeneral prin
ciples.
Pat: What's so terrible?
Kha1ifman-Belikov
Podolsk 1992
White to play
1. h6!
2. f3
gxh6
Even though Black has two

passed 1s, one of them an outside
passer, he loses (2 ... xf5 3. g7).
Play continued 2 h5 3. g3
c5 4. 'lflh4 e4 5. g3 ! Resigns
(5 ... 'lflg7 6. f4 and 7. xe4).
.
Sokolov-Korchnoi
Interpolis 1988
White to play
1. Rxc5t??
With 1. Rb8 White has good
drawing chances even a 1 down.
1.
wxc5
2. e4
3. h4
c6
Wd7!
Intending 4 ... 'lfle6 ... c5-c4-c3,

etc.
4. d5
Or 4. f5 c5 5. h5 We7 and
zugzwang arrives after 6. g3 g6t!7. hxg6 hxg6t 8. xg6 c4 or 7.
e4 'lfle6 (not 7. .. gxh5 8. d5) 8.
Chapter Nine
h6 g5!
4.
h5
9. <&>e5
10. <&>f4
xh4
Or to. Wf5 g5, zugzwang.

10
g6
1 1. <&>13
g5
White forfeited.
5. e6t
Also 5. We4 We6 6. Wd4 c6! is
zugzwang.
White can then try 7. Wc5 Wxe5
8. \txc6 but he's dead after 8 ...
\tf4 9. \td5 wg3 10. \te5 Wxg2.
5. . . .
6. wc6
we7
And here 6. \te5 c6! 7. Wf5

Wd6 8. g3 We7 9. We5 g6 is an
other zugzwang.
6. . . .
7. Wxc7
8. Wd6
wxe6
Wf5
wg4
Noah: He commited a hanging offense for a grandmas

ter. He traded down from a
drawable g + i ending to a
W+ i ending in which Black
held the outside passer.
Pat: Inexperience? Or time
pressure?
Noah: Neither. White spent
35 minutes over 1. g xc5t??
And he was about to be
named by some professor the
19th greatest player in his
tory.
Pat: It still took an awful lot of
calculation to win. And it
wasn't clear for 10 moves.
Noah: But it should have been
obvious after 1. xc5t that
Black had excellent winning
chances.
As I said, W+ i endgames
are more about calculation
than any other endgame-and

White just miscalculated very
badly.
''Never underesti
mate the power of
human stupidity. "
-Robert Heinlein
Pat: Aren't there some ancient

rules that work all the time
here?
Noah: What did you have in
mind?
Pat: Oh, I don't know. Maybe
"Passed pawns must be
pushed."
Noah: But only if they're
passed. A lot of times you see
masters blundering away by
advancing a i -like in Dia
gram 185-that only becomes
a target for the enemy to at
tack.
163
Pawns
Pat: What should he have Noah: Unfortunately it was

his opponent who had the
done?
Noah: He could pass with his
tempo moves, 4. a4 and 6. g3,
that made all the difference.
<it'.
That's usually a safe policy Pat: Okay, so far you've told
when your opponent doesn't me that in '\t'+ i endings the
have a penetration point for most important things are W
his <it'.
position and i quality.
Pat: Yeah, but the way he
Anything else?
played, Black had the oppo- Noah: 1 thought you could
sition after 3. . . <it'g6.
only afford a few minutes?
There is one thing you need
to appreciate and 1 just men
tioned them-tempo moves.
Pat: Which are what, exactly?
Noah: Moves that do nothing
but allow you to say "I pass"
and start your opponent's
clock.
In i endings you often
reach a standoff like Diagram
186 in which neither side
wants to move.
164
then 2 ... e6, Black can draw.

For example, 3. g4 Wf6 4. g5t
hxg5t 5. hxg5t Wf7 6. e5 We7!
(7. d4 e6 8. c3 We5! 9. Wb4
Wd4!).
Polugaevsky-Ermenkov
Black to play
1.
g5t??
Also losing is 1... h5 (although it
takes several more moves) because
of 2. a4! and Black reaches zug
zwang.
2. hxg5
3. wg4
4. a4!
hxg5t
Wg6
Wf6
No better is 4... h6 5. f5.
5. Wh5
6. g3!
Wf5
Resigns
In view of 6 ... <ifj>f6 7. g4! But

passing, with 1... a4 and if 2. a3,
Chapter Nine
W-side. So both sides must use
their tempi.
Pat: But somebody has to.

Noah: Right, so what usually
3.
occurs is the player with the
4. f3
most tempi ends up forcing
h5
5. h4
his opponent into a very disBoth sides are out of passes.
agreeable situation.
6. We5
a4
7. bxa4t Wxa4
In Diagram 186 White sets
up the standoff with 3. a3. He
could force matters at any
point with We5. But what he
should have done is exhaust
Black's tempo moves before
moving his 11.
Pat: That's another thing that
bugs me. White has to see as
far as a 'Q1 ending to win that
What White probably saw at
position.
move two is that 8. lfj>f6? c4 loses.

Timman-Sveshnikov
Tilburg 1992
White to play
1. f41
Begins the zugzwang process
(1... Wd6? 2. Wb5).
1.
Wb6
Here White played 2. a3??, ad
mitting he had no idea what to do.
The game was agreed drawn after
2 . Wc6 3. h3 Wb6 4. h4.
The only winning try was:
..
2. Wd5
3. a3
Wb5
f:
What he missed is 8. Wd5!,

after which the best Black can get is
a lost 'It+ it ending after 8. .. Wb5!
9. a4t! Wb4 lO. a5.
Now 3 ... a4 4. bxa4t c.txa4 5.

Wxc5 allows White to win on the
1 65
Pawns
Noah: I'm afraid that comes tion.

with the territory of + i . Pat: Hmmm. I'd be looking
You have to be able to count for something like a 'f)-check
in queening races. In this case at c6 or d3.
White would have to see that Noah: But when you have a
2. Wd5 leads to 13. a8=
queening race you have to
after which he can start pick look for chances to exploit
ing up the remaining black the very first move after you
is with his and well-cen promote-as Black would
Salov-Short
Mohr-Conquest
tralized W.
have had here with the 4 . .
Linares 1992
Gausdal 1989
White to play
Black to play
And sometimes you have to b3t! hick.
cctfcl?
1.
recognize a window of op Pat: Too much calculation for
1. Wli
f6?
bl=*
2. g7
portunity after you promote.
me. I guess you're right about
Best was 1... g6 2. Wf4 f6, after
3. g8=* Drawn
Black failed to notice in 188 avoiding W+ is.
The extra piece means little which only 3. h5! offers White
how he could have reached a
But one last thing. These here. But Black could have won chances (3 ... gxh5? 4. Wf5 and
won +W vs. + posi- endings are bad enough with with 1 ... 4)d5! threatening 2 ... 4)e7. Black is soon in zugzwang) but 3 ...
Then 2. Wxd5 Wc 1 and after both w;g7 may hold).
a few is.
2. h5!
Wlfl
Yet sometimes my oppo sides queen Black has 4 ... *b3t!,
3. Wlf4
skewering the W and *'
nents offers a trade of s or
Now White is threatening to
s or whatever, and so I end
occupy f5 and later go decisively to
e6 or g6.
up with something impossible
cctfe6
3.
like +6 is vs. +6 is.
4. g4
a6
What do I do then?
.
1 66
Chapter Nine
Black begins to run out ofpasses.
Mer 4... We7 S. WfS Wf7 6. a4 we
get something like the garne.
1 1. e6
12. f5t!
xg6
'tilg5
Noah: Those endings aren't

much more difficult than end
Otherwise 13. Wxd5 wins.
ings with a few is. The tech
5. a4
13. fl
as
xg4
niques we've talked about
b6
6. b3
14. xf6
h5
apply also to multi- i end
Resigns
7. f3!
we7!
15. <ct>e5
Ings.
Black can see that after both is
fl
8. 'tilf5
queen White wins with 19. ftg8t.
In 189, White hasjust traded
*s, on d5, so he gets to move
his t; first-a big advantage.
He can reach f4 by force
and that forces Black to find a
way of keeping him out of e5.
Pat: Which he can do with his
W or f- i .
Noah: But the real reason Black
loses this game is that White's
W gets to f5 and he has
enough passes to run Black
out of moves.
9.
g6t!
Pat: It looks like Black is sav
10. hxg6t wg7
ing himself when he sacks a
Black sets a small trap: 1 1. gS??
i with 9 ... g6t.
fxgS 12. fxg5 hS!.
Noah: But remember, end-
games are all about convert

ing one advantage to another.
So White gives back the i in
order to penetrate and win
either the f6- or d5- is.
Pat: In the end Black loses
even with an outside pass
er. How did that happen?
Noah: I warned you + is
were almost impossible. But
look on the bright side.
Pat: Bright side?
Noah: Sure, we've done *,
, and t endings.
There ain't that much left.
1 67
Pawns
168
In which Noah investigates the trickiest piece of all, and

explains how it changes your thinking ahout good fts and
had fts.
(ifffifflr:F
j
1fi!)'
Knights
Pat: Okay, Noah. So which

endings are you going to tell
me to avoid today?
Noah: None. Today we'll do
the ending with the trickiest
piece-and find they're really Tigran Petrosian quickly be
came frustrated and just
fairly easy.
Pat: By tricky you must mean wanted to know the solution.
Vassily Smyslov found the so
a , right?
lution in 10 minutes.
Noah: Naturally. There's acute
story G M Leonid Shamko Bobby Fischer didn't want to
vich loves to tell about show be shown anything-he solved
ing Diagram 191, a position it in a few minutes.
from a study by our old friend Garry Kasparov solved it al
Nikolai Grigoriev, to world most immediately in his head.
champions Shamkovich met He explained that all he
over the years. Their reac needed to do was figure out
how to reach the key squares
tions were quite different(fl or g4) with his .
Pat: Is this typical of all end

games with s?
Noah: No, but it illustrates a
few important features of
them.
It shows the technique of a
W elbowing out a . It shows
how valuable a wing it , rath
er than a center it , is.
Pat: Anything else?
Noah: It also shows that both
a 11 and f) can sometimes be
pretty fast pieces.
But most of all it shows that
speed isn't as important as
reaching the right squares.
191
Whitt' to play and draw
1. 4)b4!
h5
Now 2. c2t loses to 2 . Wf2!,

setting up a kind of W vs. oppo
sition.
Another version of that is 2.
d5t? WfJ! and the is blocked
out.
..
2. 4)c6
So far, so good. Now on 2 ... h4

White has 3. e5 h3 4. g4t WfJ
5. h2t wg2 6. g4 and draws.
The key point is that 6 ... Wg3
can be met by 7. e3! and 7. h2 8.
fl t followed by 9. xh2.
.
2. . . .
1 70
we4!
Chapter Ten
Another form of W v. 4) oppo
sition. Now 3. 4)e7 h4 4. 4)g6 h3 is
hopeless, as is 4. 4)b4 h4.
3. aS!!
Paradoxical but effective. Now

3 ... Wd5 allows 4. 4)b3 h4 5. 4)d2
and 6. 4)fl, and White reaches the
same fl -h2-g4 cycle.
3. . . .
4. c4
h4
Wf3
Or 4... h3 5. 4)d2t w;>e3 6.

4)fl t and 7. 4)h2, drawing.
5. e5t
But losing was 5. 4)d2t? Wie2

6. 4)e4 h3 7. 4)g5 h2 8. 4)e4 It/f3!
- not 8 ... h l='/t 9. 4)g3t.
5. . . .
6. c4
wg3
And the 4)d2-fl t drawing idea

is on once more (6 ... Wf4 7. 4)d2).
Pat: Yeah, but it's a trick posi

tion, a composition. Stufflike
that doesn't happen in the
real world.
Noah: Don't be so certain.
In Diagram 192 Black won
a very subtle ending because
he could see that a 4) on h6
can't get back to as in three
Gheorghiu-Yusupov
moves, but a 4) on h5 couldLucerne 1985
Black to play
much like Grigoriev.
Pat: I always thought of 4)s as
l. . . .
wg6!
clumsy and slow and not
Not 1... 4)xh6?? because of 2.
much use in an ending.
xb4! (2... axb4? 3. a5).
2. ,a,d2
xg3
Noah: Not so. Remember
3. ,a,xb4
when I said, in the 19th cen
Nothing better.
tury it was widely believed
3. . . .
axb4!
even by great players-that a
h5!
4. as
4) was better than a ..Q. in the
And Black won (5. a6 4)f4t 6.
ending, because a ..Q. is lim
Wie3 4)e6 and 7. . 4)c7, or 5. w;>e3!
4)f6 6. Wf4 It/xh6 7. a6 4)d7 8. a7
ited to half the squares on the
4)b6 9. Wxg4 Wig6).
board.
Pat: Yeah, right, the good old
days, like when they used

sandglasses instead of clocks.
So, anyway, tell me: what do
I need to know about 4) end
ings?
1 71
Knights
It's zugzwang: 3 ... 4)e6 4. h7

etc., or 3... 4)g6 4. e6t e8 5. h7
4)h8 6. 4)f6t.
Noah: Two things. The first is

a basic principle:
Black lost because he made

a very inflexible first move.
Pat: You said there were two
things I need to know about
Keep You,
4) endings.
As Flexible As Possible.
Noah: Yes, the other is Bot
vinnik's Law:
Black loses in Diagram 193
purely because he violates this
rule-and because he forgets
A Ending Is Really
a basic technique of 4) end
A :ft Ending
Ings.
Pat: Which is?
Pat: What the heck does that
mean?
Noah: Triangulation. White
zugzwangs Black by taking Noah: It means that most of
two moves to get his W from the general rules and tech
d5 to d6.
niques of i endings apply
Pat: And that's a big deal in 4)
here.
endings?
For example, a one- i ad
Noah: Yes, because a W usu
vantage is usually decisive if
ally can gain or lose a tempo there are lots of is on the
when fighting a 4).
board.
It usually can't do that to a
A good example of bad
and rarely to a heavy piece.
thinking is Diagram 194.
1 72
Chernin-Panno
Buenos Aires 1992

White to play
1. d5
4)f4t??
Black draws with the flexible

1... f7, leaving his 4) free to go to
f4 or d4, or to retreat.
For example, 2. c6 4)d4t! 3.
Wld6 4)b5t 4. Wd7 4)d4 or 3.
Wld7 4)e6 4. h7 4)f8t 5. Wld6
Wlg7! and draws.
2 We611 cih'P7
e .l l
Otherwise a 1 advances deci
sively, e.g., 2 ... 4)e6 3. d6 Wf7 4.
h7! and wins.
o
','
00
3. Wd6
Resigns
B OTVINN I K' S
L A W
A Ending
Is Really
A it Ending
Chapter Ten
4)e3 9. wxg6 4)d5 10. f5! 4)xb4
1 1. f6 and wins.
Zsu. Polgar:J. Polgar

Monaco 1994
Black to play
1. .
4)f1t??
Black could have virtually

forced a draw with 1... a g3 since
White has nothing better than 2.
4)e2 a h3 3. 4)gI ag3.
2.
3.
4.
5.
6.
4)xh3
\tla4
<iflb5
<It'c6
<It'd5!
4)d2t
4)xe4
e7
d8
And the black \t?-side collapses6... 4)d6 7. e6! 4)f5 8. Wxf6

4)xh4 9. \t?g5 or 7. .. 4)xc4 8. Wxf6
Pat: With s on the board.

Noah: Yes, but Black walks
into a disastrous transition
from a drawable + 4) end
ing to a dead lost 4)+ i one.
Pat: In what other ways is a 4)
ending like a i ending?
Noah: There's a lot of zug
zwang, a great emphasis on
gaining tempi and a high pre
mium on outside passed is.
Pat: Aren't outside passers al
ways valuable?
Noah: They're important, but
they alone don't give you a
win in, say, a ending. In
fact, a - i can be a real
liability in some -endings.
A Lucena-like position with a
- i is only a draw, for ex
ample because it's easier for
the defender to cut off the
W's escape.
But since 4)s can't defend a

distant pOint the way a or
..Q. can, a fast - i or 4)- i
can end a game quickly. Both
sides missed this point in Dia
gram 195.
1 73
Knights
Pat: I thought you're never

supposed to hurry.
Noah: White wouldn't be hur
ryingwith 2. a4!, he would be
creating a positive asset.
But he was hurrying with 8.
h5 ?-and that's what cost him
a half point.
Pat: So is everything that's true
Nunn-Timman
about t endings also true
Reykjavik 1988
BlIlck to pilly
about 4) endings?
Noah: Not quite. Being a t or
4)dS?
1. . . .
two behind in a t ending is
Mter 1... c4! Black has real
usually hopeless-but that's drawing chances.
2. 4)e2?
not so true with 4)s.
But 2. a4 ! offers winning
This explains why the supe
rior side in a 4) ending is chances, e.g., 2 ... 4)xf4 3. b4 and 4.
as.
always looking to trade 4)s
4)e3
2. . . .
into a more winnable t end3. 4)d4
4)c4!
game.
4. b3
4)xa3
4)bl
S. c4
4) c3
6. <tt;>g2
f7
7. h4
1 74
With 8. Wf2 White can prepare

a decisive h4-h5.
8. hS?
9. 4)xfS
gxhS
4)e2!
Now 10. 4)d6t We6 1 1. 4)c8

gives Black serious counterplay
( 1 1 .. 4)xf4t 12. Wg3 e3!).
The game went 10. Wf2 4)xf4
11. we3 4)g6 12. wxe4 h4 and it
was Black who had the outside
passed i . White offered a draw
with 13. 4) xh4.
.
Chapter Ten
Wd4! (7. f6 4)d5t).
Kindermann-Lautier

White to play
1. 4)g2?
White can play 1. xb4! be
cause 1... 4)d3t 2. Wc4 4)xf4 al
lows 3. 4)d5t!, trading into an
obviously won t ending.
1.
2. 4)el??
4)a2
White still wins with 2. 4)e3!,

e.g., 2 . We6 3. 4)d5 Wf5 4. xb4.
"Some Knights
don't leap; they
limp. "
-Proverb
Noah: That's what cost White

in Diagram 197.
Staying in the ending only
drew-even though White
might have been two is up
in one line. Yet he missed a
great chance to trade horses
(s)-and that would have
won.
Pat: But generally don't the
kind of advantages that are
decisive in W+ ft endings also
win here?
Noah: True. Usually a two- i
advantage is an easy win with
s-but sometimes it requires
some precautions as in Dia
gram 198.
..
2.
c&'fS
And draws: 3. 4)d3 We4 4.

xb4 cl 5. 4)d5 xb3! or 4.
Wc5!? xd3 5. f5 c3 6. xb4
1 75
Knights
goes in the other.

Pat: But Black has no is.
Shouldn't this be easy?
Noah: Exactly. But in this case
Noah: No, as you recall our it's a lot harder because of
rule about trading when you that ancient saying:
are behind, the fewer is left
on the board the easier it is to
"You can't dance at two
weddings. "
draw.
-Anonymous
Add a bunch of is to either
chessmaster/wedding
wing and White would score
guest
easily by setting up a mis
match somewhere.
Pat: Another ofyour wise men?
Pat: By forcing Black's to go Noah: Ofcourse. The black
in one direction while his own can stop one of the is, but
not both.
Pat: Okay, so that's a 2- i
edge. What about one i ?
Win or draw?
Noah: There's no simple rule.
The result varies depending
on how many total is there
are on the board.
We can draw up a chart:
1 76
3. g4
4. g5!
Wd3
Now 4... xc2 transposes into

the previous note, so ...
4.
we4
Lautier-Yusupov
Baden-Baden 1992
White to play
1. 4)el!
Not 1. f1 c4 winning a t ,
or 1. g4 We4! 2. 4)f2t Wd4, with
drawing chances.
1.
2. 4)c2!
c4
Now 2". d3 allows 3. g4! and

the is too far away (3 ... xc2 4.
g5 4)d5 5. g6 4)e7 6. g7 and 7. b5).
A better try is 3 ... e4! but then
the white enters with 4. g2
f4 5. Wh3.
2. . .
4)d5
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
_____
5. b5!
Since 5 ... f5 allows 6. 4)e3t!
4)xe3 7. b6 and queens, Black
played 5 e5 .
But he resigned after 6. f2
4)c3 7. b6 4)e4t 8. g2 d6 9.
g6 4)f6 10. g7 d7 [Ed.: 10 . . .
c6? and both 1 1. b7 and 1 1.
4)d4t will win for White.] 11. 4)b4.
Chapter Ten

PAWN POWER
4 fts vs. 3 fts III. Almost always

.
a WIn.
3 fts vs . 2 fts
III.
A win if there's
a clear passed ft .
2 fts vs. 1 ft
I II.
A draw most of
the time.
1 ft vs. 0 fts I II. A draw unless
the ft reaches the 7th rank

(when it usually wins) or the
6th (when it often does) .
_______
1 77
Knights
Pat: Doesn't the quality of the ings. What works in them and
fts matter a lot, like it did in only in them?
Noah: 4) endings have a few
ft -endings?
Noah: Often it does. But even unusual qualities. The first is:
in a case like Diagram 200,
where White's extra ft is The Winning Side Sacrifices
backward, a win can be His Piece Much More Often
achieved by accurate use of Than In Any Other Ending.
the W and 4).
Serper-Suba
Pat: Why?
Pat: Okay, so far you've told
Hastings 1990- 199 1
Noah: A couple of reasons.
White to play
me 4)+ ft endings are usu
The 4) is closest in value to
ally like w+ ft and some
White's extra l , at h3, looks
a ft , so a 4) sac isn't as big a
times like + fts or whatever.
useless. But:
material
loss
as
with
other
But what about the differ
1. el!
sacrifices.
Mter 2. 4)3 Black's \fl will be
ences? There must be unique
Also,
after
a
sac
a
defender
tied to g5, leaving a \fl vs. 4) mis
things about W and 4) endoften must use both W or 4) match in the center.
1.
c3
for blockade duty, leaving a
<ilfg6
2. f3
mismatch elsewhere on the
3. <ilfd4
d5
board.
f4
4. c5
Pat: I guess a 4) isn't such a
xh3
5. <ilfd6
bad blockader, as blockaders
Passive defense, with 5 ... \flh6,
go.
allows White to elbow the \fl off
1 78
______
_
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
____
the board: 6. d7! '1t'g6 7. '1t'e7

h6 8. f6!
6. xe6
7. h2
f2
It's 2 vs. 1 now, but White has

created a winning passer, e.g., 7.
4)e4 S. \fle7 and 9. e6.
.
7.
8. d6
9. e6t
.
d3
Wfl
And after 9 e8 10. 4)3 4)12

White could afford 11. 4)xg5!
4)xg4 because 12. 4)h7! wins, e.g.,
12 .. '1t'dS 13. e7t '1t'e8 14. '1t'e6!,
zugzwang.
Chapter len
Noah: Right. And another rare

feature of l) endings, as I
mentioned earlier is:
Wing is Are Better

Than Center is.
- is Are The Best,
Followed By 4)- is, etc.
Suba-Zapata
Pat: That sounds like the exact

opposite of what happens
with heavy pieces.
1. h4??
White wins with 1. ctle8t We 7 Noah: Just about. With only
an a- jt or an h- jt it is really
2. ctlxd6! Wxd6 3. 'llg5.
The main line runs 3 . e7 4.
difficult, if not impossible to
wxg6 ctld7 5. h4 ctlf6 6. f5 and
win a ending or a * end
Black's ctl is no match for White's
ing because it's so hard to
W and passed is.
find a good place for your w.
1. . . .
Win
But the ground rules are dif
2. WfJ
Wie7
ferent with the l)s, rather than
3. wg3
Wif8
4. wg4
Win
heavy pieces, for a few rea
And White agreed to a draw 3 1
sons:
moves later.
1) No perpetual check is
Tunis 1985
White to play
possible, so the best method

of defense is to reach a point
in front ofthe queening square
with your 11 or 4). And it's
harder for them to reach a
point in front of a wing jt
than a central one.
2) If the defender's 11 can't
get in front of the jt , he'll
have to use the 4) . But the
closer the jt is to the edge of
the board, the greater the
chance that it will be kicked
away, as in Diagram 202.
..
1 79
Knights
Noah: The resource that might

have saved Black was a check
after 6 ... 4)c2 7. a6 4)b4 8.
a7-but the check would have
to be on a fifth-rank square
that is to the left of the a-file.
Pat: You mean off the board.
Noah: You're lucky. Some day
they may add another rank
or file to the board and you'll
have to relearn a lot about
endings.
Pat: I think I'll stick to 64
squares-for now.
Noah: One other thing you
ought to keep in mind. In one
other way a 4) ending is quite
unlike a i ending.
Pat: How's that?
Also lost is 6... Wd6 7. a6 4)c6

because of 8. 4)b7t d7 9. 4)a5.
7. a6
So that 8. a7?? allows 8 ... 4)c8t
and 9 ... 4)xa7, drawing.
8. lflc6!
4'l
Cifuentes-Brenninkmeijer
Dutch Championship 1992
White to pilly
1. 'it'c6!
2. 'it'b5
'it'g4
4)b3
White can answer 2 ... f3 3.

'If;xa5 'If;e4 with 4. 'If;b4! since 4 ...
'If;xd3 5. a5 4)c4 6. f5 wins.
3. 4)c5
4)d4t
The 'If;+ t ending is clearly lost

and 3 ... 4)d2 4. 'If;xa5 'If;xf4 loses
because the 4) can't catch the a- t
after 5. b5 and 6. a5.
4. \flxaS
5. \flb6
6. as
180
4)d6
'it'xf4
\fle5
4)f5
4)c8
///m 203
'Ii
/
.
..
nhn

9. 4)b3!
White sees that when he attacks

the 4) it can only safely go to d6 (9.
b7 4)d6t 10. c7 4)b5t).
Therefore his own 4) heads for
c4 where it stops ... 4)d6.
9.
10. \flb7
1 1. \flb6
12. \flc5
.
4)a7t
4)b 5
4)d6
4)c8
______
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
____
_______
Now 13. 4)a5 and 14. 4)c4

would win. But White prefers an
other way to cover d6.
13. 4)d4
14. 4)b5
\fle4
1-0
The end would be 14 ... Wd3 15.

4)d6 4)a7 16. b6.
Chapter Ten
Noah: In endings an out

side passer is better than a
protected passer, as that an
cient example, Diagram 204,
shows.
Pat: Why is it that every end
ing has to have something
unique-and something extra
to remember about?
Pillsbury-Gunsberg
Noah: You wanted to play an
Hastings 1895
Black to play
easy game?
But don't give up on it now
1. . . .
a5!
we're in the home stretch.
Much better than 1... 4)b8?,
There's only one piece left to
which lost quickly {2. f5! gxf5 3.
gxf5 exf5 4. 4)f4 regaining a ft
confuse you with.
favorably, or 2 ... g5 3. b4 a5 4.
c6!, threatening 5. c7, 4 . . . \Vd6 5.
fxe6! axb4 6. e7 \Vxe7 7. c7!}.
2.
3.
4.
S.
fS?
c6
4)c5
gxfS
gS!
4)b6
exfS
Wd6
Black is better! (6. b7t \Vxc6

7. 4)xa5t \Vc7 and ... 4)c4-d6).
181
Knights
182
In which Noah explains how some s dance and others

blockade, and why connected passed is are often weaker
than distant isolated is.
(Otffil{:}1
Bishops
Pat: I can't believe we've cov

ered everything but s.
Noah: See, Pat, I told you there
wasn't that much you abso
lutely had to know about
endgames.
Pat: Okay, so let's start with
my favorite question: What
makes endings different
from others?
Noah: Well, in general it's more
common in endings with the
s-and with 4Js for that
matter-for the defender to
draw because his ft/ is active.
Pat: Why is that?
Noah: Because you can often
give up your final piece for an
184
opponent's passed ft and

then draw by raiding his fts
on the other side ofthe board.
Pat: And that just won't work
in * endings, and in most g
endings, because those pieces
are so valuable.
Noah: Exactly.
In Diagram 20S Black's 4J
is tied to the hS-pawn and the
invasion square fS. His only
winning try is to create a
passed *-side ft .
Pat: Yeah, but he can do that
by force whenever he wants
to.
Noah: True, but if White then
gives up his for the ft on a
distant square, like a4, he
should draw.
He loses because his first
move means any raid
(dS) will be met by ... 4JxfS.
a5 5. Ad7 \tilc5 6. Aa4 Wb4 7.

Ad7 a4 8. Axa4! Wxa4 9. d5.
Pigusov-Epishin
Biel 1993
White to play and draw
1.
2.
3.
4.
fS??
bxc6
,Q. dl
,Q.a4t
c6
xc6
as
c5
And Black shepherds the a- ft

(5. Ad7 Wb4 6. \tild3 a4 7. \tild2 a3
8. \tilc2) before turning his \til back
to scoop up the \til-side (8 ... \tilc4 9.
\tilb 1 Wd4 10. \tila2 \tile4 and 1 1...
xf5, winning) .
But in the diagram 1. An or 1.
Ad3 should draw, e.g., 1... c6 2.
bxc6 \tilxc6 3. Aa6 b6 4. Ac8!
With the White ft on f5 this

drawing raid doesn't work because
of ... xf5xh4.
But here it does: 9... b5 10.
Wd6 Wc4 1 1. \tile7 f5 12. \tilf6 or
10... f5t 1 1. \tile6 xh4 12. xf6
f3! 13. f5 Wc6 14. \tile4
moves 15. f5 and \tilf4-g5xh5.
Chapter Eleven
5 .Q.d5
Kveinis-Vetemaa
Lithuania 1986
Black to play
1.
b4!
Not 1... b6? because Black

loses even if he reaches g8.
For example, 2. fJ c2 3. g4
Wc7 4. e4 b3 5. We5! (elbow)
Wd8 6. Wf6! We8 7. g5 Wf8 8. g6
1fg8.
Now 9. g7 wins by 1fg6-h6/
h7t as we saw way back in Dia
gram 67.
2 .Q.f3
3. g4
4. e5
.Q.a4
c3 !
d3
Now 5 ... We3 loses to 6. f7

(which stops ... e8!) c2 7. g5
d3 8. e6 g6 9. <1flf6.
5.
6. g5
7. Wf6

.Q.e8!
e3
Wf4
White cannot make progress

because when he plays f7 Black
trades and takes the t . And, no
better is 8. e6 h5! (pass) 9.
d7 f7!
curely in front of the ft then

Pat: How often does the active
he absolutely must attack it
W save the defense?
Noah: A lot. Even in the most from the rear.
basic positions of w+- ft Pat: And that's good enough?
vs. W+ the defender draws I don't remember seeing this
if he can get his W in front of attack-from-behind much in
other endgames.
the ft .
Pat: Just like with + ft . I
haven't forgotten about the
Philidor position-yet.
Noah: I'm sorry to relay the
news that there's no iron-clad
"Philidor" in endings.
Still, an alert defender can
draw even if he can't block
ade the enemy fts.
The reason is that he can
give up his for his oppo
nent's last ft .
Pat: Like the way Black tries to
do in Diagram 208, I guess.
Noah: Yes. Generally, if the
defender's W can't get se185
Bishops
Noah: It doesn't usually work Noah: Because Black's W is

in others. In fact, it only works then elbowed out, leaving him
in certain .Q. endings. For with only one defense, the
example, Diagram 209 bears attack from behind.
a strong resemblance to the Pat: Which he can do with 7...
previous example if White 'ifff5. So why does he lose
forces an exchange of his d- anyway?
ft for the f- ft
Noah: Because this time we're
Pat: Why is that good?
dealing with a - ft , closer
Prasad-Anand
to the edge of the board.
India 1985
White
to play
This means the defender's
.Q. has fewer squares and can
d6
1. f5
be nudged off the board-like
2. xf6
wxd5
after 8 . .Q. c 1 !.
f8
3. g5
4. b2!
Pat: How does it help me to
Now 4 ... l\'d6?? 5. a3t or 4 ...
know that?
Noah: It helps you know when l\'c4 5. l\'e6 and 6. l\'f7 make it
easy for White.
to make good transitions-like
e7
4. .
White did with 1. Wf5 and 2.
5. g6
f8
Axf6.
6. Wf6
It also helps you set good
traps, as Black did in Dia e4 7. l\'f7 f5 8. cl! and
Black has no moves.
gram 210.
.
186
______
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
_____
Chapter Eleven
8. wg7
Resigns
In view of 9. h5, 10. h6 and

11,h5t-g6.
Belyavsky-Kotronias
Belgrade 1993
BlIlck to pllly
1. . .
2. xg3!
.
g3!
Mter 2. hxg3 xa6 it's much

harder for White to end the block
ade (3. 11,e4 Wb5 4. 11,fJ .Q.b3 5. g4
Wc4 6. g5 Wd4 as in Diagram
208).
2. .
3 .Q.g4??
Wxa6
A basic mistake: White should

advance the i first, then drive the
enemy .Q. away (3. h4 Wb6 4.
11,g4).
Now 3 ... .Q.a4! draws because

the .Q. has an extra tempo to set up
a blockade of h5. For example, 4.
h4 .Q.e8! 5. f4 b6 6. g5 r:j;c7
7. .Q.f5 Wd6 8. 11,g6 We 7! or 8. Wf6
11,h5!
Pat: I'm not used to capturing

away from the center. Why is
2. xg3 correct? This isn't a
c2??
3.
4') ending.
Another basic blunder: Black
tries to stop the 1 from reaching Noah: Because while a 4')- ft
is more queenable than a h 7 when he could have stopped it
ft , a g - ft is even better.
from getting to h5 with 3 ... .Q.a4!
and 4 ... 11,e8.
The reason is that once
4. Wf4
<llb6
White blocks the d I-h5 di
5. <llg5
<llc7
agonal, the only places for
6. h4
<lld8
Black's to stop the g - ft
<lle8
7. <llf6
are along c2-h7 and e8-h5.

Pat: But Black missed a draw

later on.
Noah: That's because White
delayed getting his ft to h4,
and the delay should have
allowed Black to set up shop
with ... e8.
Once Black plays ... e8,
White's only winning chance
is to play g6. But by the
time he gets that in...
Pat: ... Black's arrives to the
rescue at e 7.
So what else do the G Ms
and other Informant types
know about making a draw
here that I don't?
Noah: They make draws be
cause good defenders know
what a lost position looks like
several moves in advance
and avoid them.
Take 212, for example.
187
Bishops
7. g4!
fxg4
is. That's pretty much what
Pat: Hmmm. White eliminates
8. xg4 f7
happens here, right?
the is that make Black's A
9. d3
Noah: Exactly. A endings with
"bad." Explain that one.
one i are pretty easy once
Noah: He does it because an
other rule takes precedence. you figure out how to screen
the defender's A by inter
Do you rememberposing your own.
Pat: So that screening thing is
Trade t s When
a big deal in A -endings.
You're Behind
Belyavsky-Pr. Nikolic
Noah: It is the technique to
Barcelona 1989
White to play
learn in these endings, as
Pat: Sure.
important as learning about
Noah: Well, it works here be
Now 9 . . b3 10. \t'c3 \t'd6 is
1. f4!
cause White sees that by the cross-checks in *-endings, or
Mter 1. fl c5 the b3- 1 falls drawn-not because of -'td lxb3
{which leads to a lost \t'+ i end
time Black picks off the b- i , the opposition in i -endings. and with it the game.
ing)-but because of 1 1. h5! For
exf3
1.
Make sense?
White's <ft1 will be very close
example, 1 1... gxh5 12. -'txh5! or
4)xf3t
2. 4)xf3
to the i's queening square Pat: I guess so. But A endings
White has an easier time after 1 1... g5 12. h6 -'tg6 13. xb3.
and can count on setting up a are over my head when there
d5
9.
2 ... 4)d3 3. -'tfl.
successful blockade on a are several is and all sorts of
10. lt
e5
3. xf3
<Ilc5
Black's best chance is a \tl-side
confusing things like ZZ.
square such as b2.
e6
4. <Ilt2
Then all White needs to do Noah: Zugzwang does come
Now 5. -'td 1 b4 and 6... -'txb3 mismatch.
1 1. e4!
to make a draw is trade offthe up a lot in A + i endings in is lost.
Threatening 12. -'txg6! -'txg6t
xb3
5. <Ile3!
Black g- i .
which one side's A is "bad"
13. \t'c4, killing the last i .
6. h4
b4
hemmed in by its own is.
Pat: Or give up his A for both
.
188
Chapter Eleven
1 1.
12. c6
13. \tIc3
14. e4!

b3
\tIf4
\tIg4
Pat: How bad does a bad

have to be?
And since 14 ... xh4 15. xg6

is dead, a draw was soon agreed.
789
Bishops
Noah: Not very. In Diagram

214 Black only has two fts on
dark squares-but he loses.
Pat: Seems like White bored
him to death.
Noah: No, it was more artistic
than that, even though it in
volves a very elaborate set of
maneuvers.
What you see here is a re
markable technique that is
unique to these endings-the
Dance of the As.
Pat: Why is it a "dance"?
Noah: Because it's a sort of
pirouette between the two
enemy As.
White is looking for anyone
of a series of zugzwang posi
tions.
Pat: I actually see one of them.
It's zugzwang when White's
ji is on f6 and Black's is on
190
ft .
c7, and it's Black's tum.
A ji move loses a ft and a Pat: I know there are always
move allows b5 or d5. games in the Informant like
Noah: It's also 'lZ with a jifc3
that-but I can promise you
versus a jifc7. And there's I'll never win anything like it.
one with the black jifg7 and Noah: Sometimes the position
wins it for you. Certain fac
a jif h2fg3.
tors always give you good
Pat: I'll take your word for it.
But what's the point? They winning chances in ji end
Shabalov-Varavin
dance and they dance...
Ings.
U.S.S.R. 1986
White to play
Noah: ...untilatmove 10White
gets one of the positions he
Winning Factors
1. .1ld2!
wanted-the one in the origi
(I) Better ji
Headed for the h4-d8 diagonal,
nal diagram but with Black to
to take away squares for Black's .
(2) Outside passed ft
d8
1.
move.
(3) Wi on the fourth
2. el
b6
Pat: White lost a tempo.
rank
Not 2 ... .Q.e7 3. xa5, or a 2 . . .
Noah: You got it. Then all
W,move which would allow Wd5
White needs to win is repeat Pat: I'd add another-( 4) An or WbS.
opponent who's dumber than
the process-the same A
3 .1lh4
e3
dance-and lose another you in the endgame.
Black must avoid a zugzwang
position of 3 . .. .Q.c7 4 . .Q.f6. For
tempo.
example, 3... .Q.b8 4 . .Q.d8 and
Black runs out of passes at
WIns.
move 21 and fatally loses a
Chapter Eleven
4 .Q.g3!
.Q.d4
Or 4 ... f4 5. e I ! and wins

(5 ... Wb6 6. 'lfld5) .
S .Q.h2
6 .Q.gl!
.Q.b2
The threat is f2-e 1. Now 6 ...

c 1 7. f2 d2 allows 8. g3!
(8... f4 9. e 1 or 8 ... Wd6 9.
WbS and f2-b6).
6.
7 .Q.f2
So
d8.
.Q.a3
.Q.e7
8. e 1 can be met by 8 ...
8.
9 .Q.el!
10 .Q.c3

.Q.d6
.Q.c7
hS
Looks like the last diagram but

here 19 ... d6 loses quickly to 20.
e l c7 21. c3!, zugzwang.
Only the h- l could move.
1 1.
12.
13
14
IS
.Q.d2
hxg4
.Q.el
.Q.h4
.Q.g3
hxg4
.Q.d8
.Q.b6
.Q.e3
.Q.d4
16
17
18
19
.Q.h2
.Q.gl
.Q.f2
.Q.g3
19.
20. .Q.h2!
21. gS!
.Q.f6
.Q.g7
.Q.f8
Or 2 1... h8 22. g3 g7 23.

e 1 Wb6 24. \tldS.
The game went 22. xe5 e7
We've seen this before except

with h- ls on the board. Again IS ...
f4 loses to 16. e 1.
8 .Q.g3
.Q.b2
.Q.a3
.Q.e7
23. f6 b4 24. c3 e7 25.

xa5 xg5 26. b4 f4 27. b5t
Wd6 28. c3 g5 29. e5t c7 30.
a5t and 31. d5.
Noah: No doubt that would

also help-as well as crucial
disadvantages for the de
fender, like having only two
square diagonals.
Pat: What does that mean?
Noah: When a 1t is severely
limited in scope, like Black's
in Diagram 217, there is tre
mendous potential for zug
zwang.
216
Now 8 ... f6 9. h2 g7 10.

gS! and White wins with g3-e 1.
"Chess is a fairy tale of 1,001 blunders. "
-SavieUy Tartakower
19 1
Bishops
Pat: I'm with you so far. I

think.
Noah: Black loses this game
192
because when the A sits on

c8 he only has two squares he
can go to. And on. e8 he only
has three.
Pat: So what?
Noah: So White can gain or
lose a tempo, like with Ad 1ffi-g4-d 1, and break through.
Pat: Still looks like Black was
bored to death.
But I'll admit what really
surprised me is the way White
gave up his dark-squared A
at the very beginning.
Noah: It's very logical. The
advantage a good A has over
a bad one grows as the rest of
the pieces are traded off.
The same goes for other
transitions. In Diagram 219
the real difference between
White and Black may not be
instantly evident.
1a,b3-c4.
5. <CIteS
6. 1tdl
Wic7
Kharlov-mybin
Soviet Championship 1991

Black to play
I.
2. 1txe7!
4)e7
There is no W entry after 2.

<&>d4 4)c6t.
2.
3. e5
4. Wid4
Wixe7
<CItdS
WicS!
The position Black must avoid

arises after 4 ... <&>c7 5. 'if(c5 and
now 5 ... 1a,c6 6. 1a,xe6 or 5 ... <&>moves 6. 'if(d6, or, finally 5 ... 1a,cB
6. a4! bxa4 7. 1a,xa4 followed by
1a, eB or creating zugzwang with
Now Black lost with 6 1a,e8 7.

WI 1a,d7 8. 4 1a,c8 9. 1a,d1!
since a 'if( move allows 'if(d6 or
<&>b6, and a 1a, move leads to the
zugzwang mentioned above (9 ...
1a,d7 10. 1a,b3 1a,cB 1 1. a4).
The game could have gone:
6.
7. 1tc2!
1tc6
1t eS
Not 7. . 1a,d7 because B. 1a,b3 is

that zugzwang again.
.
S. 1te4
1td7
Going to f7 allows 9. 1a,c6 with
Chapter Eleven
9.
10.
11
12
13
.Q.g2!
.Q.f3
.Q.g4
.Q.dl!
.Q.b3
counterplay for White.

Pat: So that's all I really need
to know about .1l endings?
Noah: That's enough to get
by. Except.
Pat: Except what?
Noah: Except that most ev
erything I've said about .1l

Pat: You mean the fact that all endings flies out the window
of White's four is are on when we're talking about As
dark squares and all ofBlack's of-opposite-colors.
are on light ones.
Pat: It doesn't apply?
Noah: You're catching on. This
means that Black has.a mod
est edge with *s on the board
because the *s mask the bad
ness of the white .1l. But it's a
stone-cold win after the *s
are traded.
Pat: Then Black picks up ei
ther the h- i or the e- i -or
both.
Noah: With a minimum of
must lose. For example, 6. f6 f4

and 7. . fJ or 6. d3 e7 7. f4
f6 and ... We7xe6.
a different zugzwang.
.Q. e8
.Q. d7
.Q.c8
.Q.d7
And wins, as noted above (13 ...

c8 14. a4 bxa4 15. xa4).
Browne-Ivanovic
New York 1988
Black to play
1. . . .
2. *xf5
*f5t!
Mter 2. Wc3 "ii1 xf4 3. xf4

Wf7 Black wins the h- i and the
game. For example, 4. d2 wg6
5. Wd4 f8 6. b4 xh6! 7. wc5
c 1 8. xb5 xb2 9. wxa4 h5
and Black is faster.
2.
3.
4.
5
.
e6
c3
.Q.g5
exf5
f8!
.Q.c5
e8!
And White, out ofgood moves,
193
Bishops
Opps, " as we call them, Black

Noah: Almost none of it. For
example, with same-colored has drawing chances because
lis Diagram 220 is child's he can sacrifice the li for
both is if White isn't careful.
play.
Pat: Well, maybe even I could Pat: How'd White screw up?
Noah: He violated a basic "Bs
win it.
Noah: But with "Bees of of opps" policy: when you
have connected is you need
your in front to advance
them.
Otherwise the defender can
set up a successful blockade
on a square iike b6 in that
example.
Even huge advantages in
is-and even the presence of
other material on the boardcan prove inconsequential
when you're stuck with " Bs of
Opps."
Pat: Hard to believe Black can't
win in Diagram 221. Where'd
he go wrong?
194
7. b5
Bf2!
White can make no progress

and agreed to a draw after 8. b6t
Bxb6 9. axb6t rtIxb6.
Somlai-Sherzer
Zalaegerszeg 1990
White to play
Wb3". .
Necessary was 1. b5 We5 2. b6
Wd6 3. <&>b4 followed by 4. a4 and
\fIb5, and the a- t goes through.
I
..
1.
2. a4
3. we4
4. Wb5
Bel
we5
Wd6
Bd2!
Not 4 ... rtIc7?? because of 5.

Wa6! followed by 6. b5, 7. as, 8.
b6t, 9. Wb5 and a6-a7.
5. a5
6. wa4
we7
Bel
Chapter Eleven
Noah: Earlier. In the diagram

White draws, even though
four is behind and even with
a pair of s still alive.
Pat: Looks like an optical illu
sion. Can't Black do anything,
like sac a piece, to keep his
winning chances alive?
Noah: Well, he might have
kept s on the board, but
after 1. t! c5! he's losing the
b- i and the f- i then same
Black has just captured on b4, a
story: draw.
mistake that allows:
1. g c5!
g4t
2. e2
g b2t
3 . Wfi
' special weakness, and at
times a saving grace, ofBishops

is that two opposing ones may
.find themselves unable to attack
each other. They operate each in
a different diocese, and ignore
each other. "
-Gerald Abrahams
Note that White's W keeps to

light squares.
The game was drawn after 3
,td4 4. S xb5t S xb5 5. ,txb5
w;c7 6. ,td3 since 6 .. f4 7. ,tf5!
sets up a classic blockade, 7. g3 8.
,txh7 and wg2-f3.
..
795
Bishops
Pat: So, are all Bees-of-opps claiming Black can't win.

endings a draw?
Pat: He can?
Noah: The drawing ability of Noah: He can ifhe can get two
Bs of Opps is legendary-but of his is to dark squares.
sometimes exaggerated.
Black can break the dark
Emanuel Lasker used Dia square blockade by triangu
gram 222 in his Manual of lating and then sacrificing the
Chess to illustrate the drawish d- 1 . His W then reaches d3
nature of this kind of end and the rest is more or less
Black to play
game.
easy.
<it?d4
1.
But these endings are so Pat: Are you saying extra is
2. b2
tricky even Lasker erred- don't usually mean anything?
White cannot allow 2 ... c3.
Noah: No, they do. But you
2.
e3
could state this rule about
Now 3 ... d2 must be stopped,
Hs of Opps with confidence: and 3. c3 fails to 3 ... We2 4. d2
Material
Advantages
Don't Count
(much)
Material Advantages
Don't Count
(much)
b4t! So ...
Spassky-Yusupov
3. cl
Lasker only gave 3 .. We4 4.

a5 <3;d5 5. Wb2 Wc5 6. Wa3 and
draws.
Linares 1990
White to play
3. . . .
4. 1txd2
d2t!
Wd3
A case in point is Diagram

223. White loses after he es
And wins since 5. Wb2 hangs
the .
tablishes material equality.
196
White must allow the is to

advance: 5. b4 c3 6. Wb 1 Wc4 7.
a3 b4 8. c l b3 9. a3 g6t
10. Wa l <3;d3! (not 10 ... b2t 1 1.
xb2) 1 1. cl c2!.
___
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
1 .!te3?
White likely draws after 1. f1 !

<3;e6 2. We 1. If 2 ... f5 then 3.
g7 g5 4. h5! and 5. xh6, pre
venting the creation of a passed h-
Chapter Eleven
Threatening 9 . . c5t!.
that doesn't create a passer

cSt!
9. ,Q. d2
for him.
Now 10. bxc5 dxc5t 1 1. Wxc5
Noah: No, not immediately.
loses to 1 1... Wc2. For example, 12.
White lost because he al
.te l ?? Wd l or 12 . .te3 <&>xc3 and
lowed Black to create two
...b4 or ... as.
connected passers. He missed
10. e3
wc2
1 1. f4
dl!
a fundamental B-of-Opp tech
And White forfeited in this
nique.
.
hopeless position.
Whitefailed to force the

enemy fts onto the color
of the enemy jt so they
could be blockaded.
Pat: Because he doesn't use his Pat: Like with .(tfS-e 7-d8!
W to protect a vulnerable * Noah: Correct. That would
side.
have drawn in Diagram 224.
Noah: Right. He goes after the
By the way Boris Spassky,
The position may still be saved
Black h- i -but that's little who played White here, once
by 7. .tf8! and B . .te7. For exhelp because he's miles from explained why his first two
ample, 7... Wa4 B . .te7 d5 9 . .tdB,
creating a passed <ai-side i .
marriages failed.
or even B ... as 9 . .tdB!!.
Pat: I don't get it. Black goes
The reason, he said, is he
7. e3?? it d5
after the b2- i even though and his wives were "Bishops
8. Wd4
itc4
of opposite color.
Pat: Very funny. So if material
doesn't matter...
Noah: Doesn't matter much.
Pat: Have it your way: "doesn't
matter much.
But if it doesn't, what does?
Noah: Proper piece place
ment-trying saying that three
times fast.
Pat: Uhh, I'd rather not.
Noah: All it means is getting
the .(t and W to the right
squares.
For instance, in Diagram
225 Black takes his time in
bringing his <aI forward be
cause he knows there is some
thing much more important
to do in the diagram.
Pat: That's to stop the White
.(t from reaching g4, right?
197
Bishops
"Bishops of opposite
color are not much good
at supporting the
advance of their own
pawns For this reason,
the Bishop usually holds
up enemy pawns, leaving
to the King the task of
supporting the advance
of his own pawns. "
-Paul Keres
198
Noah.: Correct. When your

opponent has two passed is
you try to find the diagonal
that covers squares in front of
both is.
In this case, it means dl-h5.
Pat: Can't White draw by set
ting up some sort of blockade
on the light squares?
Noah.: You means with a A on
h3 and on f3? It doesn't
work because the Black is
are so far apart.
That's another unique as
pect of "Bs of Opps"-the
power of widely separated
is.
In some other positions,
such as with s, it's only a
draw if the only is left on the
board are your a- i and h- i .
Pat: But not with As.
as
many is as possible.
6.
7.
8.
9.
...
axb6
-'to
b5
We5
axb6
Wd4
h4
Slekys-Panchenko
Bratislava 1992
Black to play
1.
h5!
White can draw ifhe gets his 1a,

to the key h5-d 1 diagonal. For ex
ample, 1... Wg7 2. WfJ 1a,f4! 3.
1a,g4! 1a,g5 4. 1a,h5 Wf6 5. we4
We6 6. a4 Wd6 7. Wd4 etc.
2. Wt3
-'tf41
Of course, 3. Wxf4 e2.
3.
4.
5.
6.
-'ta6
b4
a4
as
-'tg5
Wf7
Wf6
This is White's best bet to trade
Now 10. 1a,h3 Wd3 allows the

W to invade ( 1 1. 1a,n t Wd2 12.
1a,c4 h3).
10. We2
I I. -'th3
We4!
Wf4
And Black's W gets to g3 to

advance the is: 12. 1a,d7 wg3 13.
Wfl h3 (with the threat of 14 ...
e2t! 15. wxe2 g2 and ...h2 wins)
14. 1a,c6 Wh2 15. 1a,t3 1a,h6! 16.
1a,e2 wg3 17. 1a,c4 e2t and White
Chapter Eleven
resigned in view of lB. xe2 h2
4.
S. g4!
6. S!
7. h5
or lB. Wxe2 g2.
EhIvest-Kupreichik
Moscow 1987
White to play
1.
2.
3.
4.
bSt!
as
Wb4
eS!
axbS
Wb7
e2
A key move: White is going to

sacrifice two is to create a passed
h- i , but must make sure Black's
new passed is don't advance far.
c4
hxg4
gxfS
The threat is B. h6. For example,

7. f7 B. h6 g6 9. xb5 and 10.
a6t 1;a7 1 1. c6 wins.
.
7.
8. xf4
9. h6
4
d3
Wc6
Black was in zugzwang. He lost

after: 10. a6 e4 11. a7 Wb7 12.
Wxb5 d3t 13. waS xa7 14.
c6 'It>a8 15. 'It>b6 and wins.
Noah: No, the farther apart

the better.
See for yourself in Diagram
227. At first glance it looks
easy because White is three
is ahead.
On second glance it looks
hard because White has no
passers.
Pat: I have a feeling there's
gonna be a third glance.
Noah: On third glance White
wins even after he creates
material equality-and gives
Black three passers ofhis own.
But Black's passers are
stopped-thanks to 4 .te5and Black's W and .t help
lessly try to stop promotions
at a8, c8 and h8.
Pat: So unlike a middlegame,
isolated is are good in these
endings?
Noah: They're usually better

than connected is. And the
more isolated the better.
Pat: Is there a rule here?
Noah: Not hard and fast. But
you can say that if only two
is are left and they are sepa
rated by one file, it's usually a
draw-because the defender's
<it> can easily shift from block
ading one to blockading the
other.
Pat: But with two or more files
in between ...
199
Bishops
Noah: . there's more chance endings it's usually best to

for a mismatch. The defend attack enemy is from be
er's 11 has to choose between hind.
But Black could still draw at
which i to protect.
In Diagram 229, for ex various points by playing his
ample, Black correctly goes g- i to the fourth rank.
Pat: What does that do?
into a B-of-Opp ending.
He can draw at various Noah: It means that when the
points even though he misses inevitable i trades comes,
the fastest way, forcing the White will be left with a d- i
W-side is to light squares.
and most probably an f- i .
Pat: With .. b4-e 1.
Pat: Separated by only one
file-a draw, right?
Noah: On the button. In these
Noah: Right. But as the game
goes Black allows him to cre
ate an h- i -three files from
the d- i .
Pat: A win. Makes sense so far.
And I guess that for the guy
with the advantage, breaking
the blockade is all-important.
Noah: Yes, and that should be
clear in Diagram 231.
5. C{f;xd4
6. h4!
..
200
Wle7
1tb8?!
Black can draw if he plays ...g5,

preventing the creation of a passed
g- or h- i , e.g., 6... \fi>f6 7. CjJ(e4 g5!.
7. C{f;e4
C{f;6
Now 8. f4 g5! 9. fxg5t hxg5 10.

h5 xg3 is drawn.
8 .Q.e8
Ftacnik-Xu Jun
Thessaloniki 1988
White to play
1. 4)5?
With 1. 4)g6, White has good
winning chances.
1.
2. 1txf5
3. C{f;f3
xf5!
C{f;xb7
C{f;e7?
The easiest way to draw is 3 ...

Bb4! and 4 ... e 1. For example, 4.
We4 e 1 6. f3 f2.
4. C{f;e4
<tt?d8!
Black cannot allow Wd5-e6.

The loss of the d- i is serious, but
far from fatal.
___
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
Here again 8 ... g5! draws (as

does 10 ... g5!).
8.
9.
10.
1 1.
12.
d4
4
g4
g5t
1te7?!
.Q.d6
1tb4?
1td2
hxg5?
Chapter Eleven
Clearer is 12 ... Wfi 13. Wf5
e l ! and if 14. h5, then 14 ... d2!
and at best Whites gets a passed f1.
13. fxg5t!
we7?
The final error. With 13 ... wg6

Black still draws. For example, 14.
d5 wh5 and 15 . . . wxh4 0r 14.g4
e 1.
14.
15.
16.
17.
18.
Wf5
d5
1te6
Wg6
Wh7
1tc3
1tb2
1tc3
1td4
1tc3
Or 18 ... WfS 19. h5 d2 20. h6

gxh6 21. g6!. Now the h- 1 queens.
19. h5
20. Wg8
1td4
Resigns.
201
Bishops
Pat: Okay, I think I under

stand why White doesn't need
the f- i here.
Butwhy doesn't Black draw?
He has two passed is of his
own after 1... xf2.
Noah: White's fine fourth
move prevented them from
advancing without W sup
port.
Pat: But Black has material
equality.
If he can draw a i or two
down in other positions, why
not here?
Noah: Because White's is are
not blockaded and threaten
to advance.
Bee-of-opps endgames are
all about i mobility. Immo
bile is are useless is.
Pat: But in order to advance
the c- i White has to allow
202
Black's W to get in front.

Noah: True. But White gets
his i to c6 and there's al
ways a threat to promote the
c- i in two moves.
That means Black's is
tied to the h2-c7 diagonal. It
also means the white W no
longer has to defend b6 and is
free to reach d7-decisively.
Pat: Maybe I'll just talk my
opponents into playing
endings with same-colored
s only.
Noah: Or just mate them in
the middlegame. Saves time.
Pat: Yeah, but somehow I'm
not so scared of endings any
more. Or so it seems now. I'll
let you know Monday, after
the club Swiss.
Noah: Good luck.
Pat: I'll need it.
6. Wb7
ittl
Correctly restraining the c- t .
7. we6
8. b6
9. c5
Wd8
We8
\t;'b8
10. WbS
1 1. e6
12. it dS
it e3
itf4
Kaminski-Prandstetter
Berlin 199 1
White to play
1. Wf4!
The W must reach d6 now. A
bad error is 1. f4 ??, preserving a i ,
but allowing 1... Wg7 2. Wf:i Wf6
3. We4 We6, drawing.
1.
2. weS
3. Wd6
4. it e4!
itxf2
wg7!
WfS
Now on 4 ... Wg7 5. c5 h5 6. b6

White's is are much faster.
4.
S. we7!
we8
itg3t
White gets the 1t to the h3-c8

line. For example, 13. 1tg8 h5 14.
1tf7 h4 15. 1te6! as in the game.
12.
13. itt7
14. it e6!
IS. itg4
hS
h4
gS
itg3
Not 15 ... 1te3?? 16. c7t, which
Chapter Eleven
explains White's maneuver.
16. 'iflc4
17. 'ifld5
18. 'ifle6
.Q.h2
.Q.g3
Threat of '1t>d 7.
18
19 . c7t
Resigns.
203
Bishops
204
Scene:
The chess cluh, three days later. Pat (possesor of the white
pieces) is showing offa tournament game. It hegan:
1. e4 c5 2. 4)3 d6 3. d4 cxd4 4. 4)xd4 4)f6 5. 4)c3

4)c6 6 .a.e3 e5 7. 4)3 .a.e7 8 .a.c4 0-0 9. 0-0 4)a5 10.
lte2 .a.e6 11. h3 4)c4 12 .a.cl c8 13. d3 h6 14.
b3 b6 15. dl a6 16. ltb2 c7 17. a4! d5? 18. exd5
4)bxd5 19. 4)xd5 .a.xd5 20 .11xe5! xc2 21. xc2
xc2.
(iJ{:lI
'1
Tiro:UIJ,,*
CIJ(ffi-
The Chess Club
Noah: Let me get this straight.

the room I still liked my posi
You forced an equal-material tion-because he couldn't stop
ending?
my 4) from getting to f5.
Pat: I guess so. Itjust looked so Noah: True. You'd have had
good I couldn't pass it up.
great chances even if Black
Even after I walked around could keep a a on the sev
enth.
Pat: But I did pass up the a
ending I coulda got after 24.
4)xe7t. You know how draw
ish a endings are.
Noah: Vh huh.
Pat: The hard part was decid
ing to trade away the two-
advantage at move 27.
Noah: But you were able to
visualize how strong your
chances were after 28. a f3
and how little counterplay he
had.
Pat: And it took me a while to
see that what I really wanted
was a W+ + is. I mean, I
206
knew enough not to trade a

lot of is.
Noah: Right. Pieces, not is.
You also made a good deci
sion in bringing your a s to
the most flexible squares, f3
and dl.
Pat: Yeah, even in the post
mortem he couldn't under
stand why his a on the 7th
didn't mean anything.
Noah: So, he ended up trading
it off. I like the way you got
the other pair of a s off-and
then ran your W to the *
side.
Pat: Yeah, the way I saw it, I
could always create a passed
i on the W-side when I was
ready.
Noah: In other words, you
didn't hurry.
Pat: Yeah, I guess I didn't.
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
__
___
22. d4!
a cc8
Or 22 ... a b2 23. 4)f5! a xe2

24. 4)xe7t Wh7 25. xf6 xb3!
26. a db l c2 27. a xb7 gxf6 28.
4)d5!.
23. f5
24. xg7!
a fe8
White is only slightly better af

ter 24. 4)xe7t a xe7 25. xf6
gxf6 26. a xd5 a xe2.
24. . . .
25. axd5
26. a d3
xg7
g8
c5
Or 26 ... a c2 27. fJ and the b

i is a target (27. .. b5 28. axb5 axb5
29. a a6 and a b6).
Chapter Twelve
27. .Q.xf6!
28. a3
29. a dl!
a xe2
a e6
Now 29... xf2t allows 30.

Wfl ! a ec2 31. a d8t. For example,
3 1... Wh7 32. ah8t Wg6 33. ag8t
Wh7 34. ag7t and a winning dis
covered check, a xf7t.
29.
30.
3 1.
32.
33.
34.
...
.Q.h4
.Q.g3
a xd4
a d3
a d6!
.Q.b6
a e4
a d4
.Q.xd4
.Q.f6
Now 34 ... a c I t 35. Wh2 wg7

36. a b6 wins another ft .
34.
35 .Q.xd6
a xd6
.Q.d4!
Noah: I also like the way you

he went I could create a, you
forced the second pair of s know, mismatch.
off at move 34 by threatening Noah: He did use his .Q. nicely
to win a 'It-side i .
to keep your from c5.
Pat: Well, I didn't really need Pat: Yeah, but that's no for
to win another i . I just had tress. So I just shuffled around
to make the power of my is for a few moves until I fig
grow by trading material.
ured out what to do.
Noah: And you figured out
After all, there was no rush
where your '\t-side is were in an ending like this. Did
you know an ending is not
best placed.
The only way to activate Black's
Pat: Sort of.
like a middle game?
W and restrict White's.
36. Wfl
Wg7
Noah: What was your think Noah: I've heard.
37. we2
Wg6
ing after 43. '\tc4 ?
Or 37. .. Wf6 38. g4 We6? 39.
Pat: I thought I had two ways
f8.
to win. I could create a passed
38. .Q.f4
h5
g- i . Or I could penetrate at
39. g4
hxg4
c5 and b6 and win the b- i .
40. hxg4
Wf6
.Q.b2
41 .Q.e3
Noah: You couldn't win with
<lie5
42. <lid3
just one plan.
.Q. a3 !
43. <lie4
Pat: Yeah. And it turned out it
Otherwise 44. Wc5 and 45.
all depended on what he did
with his W . Whichever way
Continued on page 208

207
The Chess Club
Pat: When I found 48. h6 I

knew I was winning. I saw I
could elbow his t; from the
center with the help of g7t
and AfS.
Noah: And that last A maneu
ver to the a5-d2 diagonal was
also very nice. It stopped
... a5 and prepared the ad
vance of the -side is.
Pat: There was really nothing
he could do after that. His
was on the wrong side to stop
the g- i .
Noah: Sometimes a good plan
wins by itself.
Pat: I'd like to say I saw it all
back when I traded *s ...
Noah: but you're not strong
enough to lie that well.
Pat: I just want to be strong
enough not to embarass my
self.
...
208
Noah: About that, I wouldn't Wb6 wins the b- it .

44. b4
1rl.b2
worry any more.
EPILOGUE
Scene: The chess club, three
weeks later. Enter Terry Bel
fisch, a promising, but inex
perienced amateur who has
just lost a tournament game.
Terry: Darn!
Pat: You lost again?
Terry: And from a won game.
I must be the only idiot in the
world who could lose a 4)
endgame two is up.
Pat: Don't be so hard on your
self. Show me how it hap
pened and maybe I can spot
what you did wrong.
You know, as a famous man
once said...
Not 44". We4? 45. b3!, trapping the .
45. Wc5
46. we4
47. t3 !
1rl.a3
1rl.b2
Or 49." e6 50. Wc5.
50.
51.
52.
53.
54.
55.
b5t
itb4
it d2
f4
g5
f5
Wb6
itb2
g7
itf6
itg7
b2
237
t t
t.
;ft
;ft '
r:
v-
,

/// J
fM

Stopping ... e4 and setting up
a mini-zugzwang. Now 47". We6
48. Wc5 a3 49. d2 <l;>d7 50.
Wb6 Wc8 51. g5 and White creates
a winning g- it .
47. . . .
48. ith6!
1rl.al
Wd6
The threat of 49. g7t forces a

W retreat.
49. itf8t!
we6
56. itc3
57. g6
58. fxg6
it el
fxg6
ith6
59. g7
60. itxg7
61. Wc5!
itxg7
Wa5
Resigns
t trap: 59. a5t? Wa7 60. b6t

WaS 61. g7 xg7 and White can't
WIn.
It's clear after 60 ... xa4 61. b6.
Index
600 Endings
141
99 out of 100 49
99% of all endgames
68, 74, 76, 87, 91, 94, 104, 108- 109,

1 16, 130, 141 - 142, 154, 158- 164, 166,
175, 178, 180
4) ending (s) 172, 173, 175, 180- lSI
4)+ 1 endings 75, 178
1 -queening 63
1s increase in value 47
1s on the same color 51
* can't mate by herself 146
* ending (s) 13, 141, 143- 144, 147, 149152, 154-155, 165, 184, 188
* sac 8
*+ 1 end (game lsi, ing lsI) 33, 141- 142,
144, 151, 165
*s go off 8
behind 136
end (game lsI, ing Is]) 34, 1 16, 1 18, 132,
155, 173, 184, 206
mobility 127
on the fringes 151
+ t endgame (s) 15, 34, 37, 52, 1 16,
1 18, 127, 136, 151, 163, 17'2- 173, 188
structure 36
12
A
Symbols
ending (s) 63, 186, 188, 193, 202

s of opposite color 198
<;g safety 27
<;g+ 1 13, 3.1, 37, 45, 54, 56-59, 61 -62,
Abrahams 195
activity 27
Adams 50, 131
Adamski 140
adjournment analysis 131
adjudicator 20
AdOIjan 102
advancing forward 95
agree to draws 36
Alburt 13, 88
Alexander 139
Alterman 1 18
Altermann 65
Ambrosz 1 1
analyst 139
Anand 47, 86, 158, 186
ancient rules 163
Anikaev 121
Anonymous 121, 124, 13!}, 148, li6
anticipation I I I
Arl o/War
82
Arulaid 31
Aunt Sophie 54
authorities 131
Averbakh 83, 95, 141
avoid repetition 62
B-of-Opp 194, 197- 198, 200

bad piece 36
Baikov 74
bald 10
Bareev 32, 52, 150
Basic Chess Endings
9, 41, 139, 159
basic positions 13
basic principles 17
basic technique 47, 75, 17'2
Battle of EI A1amein 55
Bee(s)-of-opps 194, 196- 198, 200, 202
behind the 1 130
Belavenets 28
Belfisch 208
BeIikov 162
Belyavsky 23, 98, 187, 188
Benjamin 81, 154
Benko 14, 92
Berezjuk 105
Bertin 51
best squares 125
better 36
Bierce 143
Bisguier 80, 81
Biyiasas 72
B1atny 50
blockade (ed, ing, s) 63, 68, !}5, 97, 126127, I:O, 136, 144, 148, lSI, 156, 183,
187, 196, 202
blocked positions 90, 92
blunder (ed, s) 150, 158, Hil
body block 95
Bohr 20
Bonsch 90
book draw 2 1
book encyclopedias 24
book loss 55
born endgame player 64
Botvinnik 13, 1 13, 132, 148
BOlvinnik's Law
172
brains 139
Brenninkmeijer 139, ISO
bridge 16, 17, 136
Bronstein 16
Browne 14, 7'2, 84, 1 10, 193
Bs of Opps 194, WS
Burnett 37
C
calculat (e, ing, tion) 45, 60, 63-64, 102,

139, 158, 163, 166
Capablanca 17- 18, 64, 1 12
Capablanca 's Rule 50
Carlyle 95
centralized square 144
Charushin 63
checking distance 131, 1:9- 140
checks and pins 152
Chemin 17'2
Cheron 1:9
Chiburdanidze lfiO
Christiansen 76, 93
Cifuentes 180

Ciocaltea I I
classic defense 138
CliffNotes 14
club player 139
Commandment No. 1 55
complex ending 64
computers 1 10
connected ft (s) 41, 133, 194
connected pass (ed, ers) ft (s) 41, 93, WI,
162, 183, 1}7
Conquest 166
convert 27
corresponding squares 9
count 64
counterplay 62, 80, 102, 103, 1 12, 1 1 8,
134, 135, 148, 156, 206
counting 57
create a passed ft 13
cross-check (s) 152-154, 156, 188
crossing the 4th rank 41
crunch time 23, 33
Cvitan 150
D
dead draw 37, 138

deader than the dodo 91
Deep Thought 120
defender's 125
defenseman 95
defensive back 67
2 10
deny access 92
devious route 82
diagonal (s) 148- 15 1, 156, 198, 208
dilly-dallied 8
diocese 195
distant opposition 58
distant passed ft 53
Djurhuus 80
Dlugy 154
doing nothing 86
Dokhoian 123
Dolmatov 96
Dorfman 90
doubled fts 162
drawable position 37
Dreev 23, 131
Duchamp 69
Dvoiris 136
E
Ehlvest 104
elbow (ing) 94, 96-98, 108, 1 10, 1 17, 1 19,
156, 158, 160, 170, 178, 185, 208
Elistratov 160
end (game lsi, ing lsI) 27, 31-32, 38, 4243, 46, 5), 63, 72, 98, 109, 1 15- 1 16, 1 19,
124, 142, 146, 156
endgame authorities 82-83
endgame books 14
endgame eyes 64
endgame plan 99
endgame scripts 104
endgame texts 14
entry (point lsi, square) 86-8i, 91, 94
Epishin 184
equal-material ending 87
Ermenkov 125, Hi4
estimated odds 10
exceptions 139
expert 20
F
fastest with the mostest 27

femto-second 60
field-cannons 122
Fine 9, 20, 162
Fine's Law 48, 49
finesse 104
First-Born Child 54, 56
Fischer 64, 87, 96, 170
FIear 29, 1 19
flight square Hi
Flohr Ii, 18
football 67, 72, 73
forgetting 17
fortress 1 1, 12, 88-91, 94, 108, 20i
Franzoni 43
free safety 73
Ftacnik 84, 200
G
gain tempi 63
Gavrikov 95
Gelfand 32
general principle (s) 22

Gheorghiu Iii
Gipslis 80
Gleizerov 62
God 158
golf 157- 158
Grabarczyk 84
Grand Design 102
grandmaster 44, i6, 1 13, 163
Grigoriev 132, 170, 171
Grivas 61
grossmeister 76
Grozspeter 57
Grunfeld 48
Gulko 73
Gunsberg 181
Gurevich, D. 43
Gurevich, M. Ii, 42, 71
Gurgenidze 31
H
handful of positions 22
Hansen 44
Hart 105
haste makes waste 2i
heart attack 10
Heinlein 163
HeHers 39, 102
Hmelnicky 30
hockey technique 95
Hodgson 36
hours of checks 14i
Index
how to plan
100
Hubner 75, 95, 122, 149
human stupidity 163
I
Ibragimov
IlIescas
60
59
imbalance
71, 124
202
immobile is
important squares
109
impractical (information)
Informant
9, 20
190
instinct 12
iron laws 143
isolated is 183
Ivanchuk 52, 98
Ivanov, I. 93
Ivanovic 193
jillion times
39
Kengis
76
Keres 20, 47, 14 1, 198
Khalifman 162
Kharlov 192
Kholmov 109
kibitzer 59
Kieninger 35
Kindermann 46, 175
Klovan 160
Knaak 134
Komarov 87
Korchnoi 12, 100, 142, 162
Kotronias 187
Kovacevic 28
Kramnik 53
Krasenkov 144
Krnic 29
Kupreichik 127
Kveinis 185
L
37, 143
Kalinichev 94
Kaminski 202
Kamsky 42, 56
Karacev 121
Karpov 26, 34, 44, 47, 54, 86, 101, 104,
120, 134, 149
Kasparov 13, 19, 69, 92, 101, 1 17, 148,
170
106
Larsen 21, 24, 1 10
Lasker 17, 80, 100, 196
last line of defense 73
Lautier 35, 48, 175, 176
Levenfish 132
lie 1 13, 208
light-colored square 63
limping 4) 29
Littlewood 151
Ljubojevic 56, 58, 59, 146
maximum checking possibilities
142
local superiority
medals
long distance
memory
70
125, 156
long-range controller 1 19
lord of the endgame 44
loser 34
Lputian 1 12
Lucena 15-17, 20-22, 24, 64, 73, 95, 1 10I I I, 1 17, 131, 140, 173
lung Ccancer 10
Lyogky 1 19
125
96
64
middlegame
Miles
27
1 18, 125
military combinations
70
42
million different rules

Minasian
33, 61
62
mini-zugzwang 208
mind-points
minor-piece endings
35, 43
minority attacks
M
machine-like accuracy
Machulsky
1 10
156
magic (square)
100
153
mismatch 30, 68, 6), 71 -76, 95, 120,
123, 127- 128, 130, 136, 140, 145, 159,
176, 178, 188, 200, 207
mistakes 20
Mnatsakanian 55
Mohr 166
Montgomery 55
Morozevich 124
mortal sin 127
move-bound 83
movicide 83
Mueller 46
multiple choice test 133
Murdered 10
mysterious secret 68
Mirkovic
mad i -queening race
63
134, 135, 154-155
Mahon
70
Mainka 147
Maksimenko
Lagky
Kaidanov
Lob ron
74
maneuver (s) 82, 84, 123, 203, 208
Manhattan hotel 64
Manual of Chess 196
margin of error 61
Marin 33
Marshall I I I
Mascarinas 15
masochist 17
master (s) 27, 29, 64, 1 13, 139, 161, 163
master endgames 64
mate-in-fours 12 1
material advantages 196
N
Napoleon
Naumkin
82
126
21 1

Nesic 153
Nikolic, P. 129, 143, 188
Nimzovich 29
no-brainer 26
Noah 67, 99, 169, 183
Norwood 19$
Novikov 136
Nunn 174
o
object or derense 89
Obukhov 60
Olafsson, H. 123
011 36
Onaso{lder 123
one error is orten fatal 61
opponent's advantage 31
opposite color $0
opposition 20, 62, 129, 131, 158, 160- 161
Ostojic 91
outside pass (ed, er) jt 60, 162, 174, 181
p
Panchenko 84, 198

Panno 172
paralyzing !:! position 45
pass (ed, es, ing) 39, 46, 59, 65, 84-85,
1 10, 1 12, 1 17, 122, 159, 167, 185, 190,
196
pass (ed jt lsi, erlsll 39, 42-45, 86, 102104, 109, 1 17- 1 18, 120, 125-126, 128,
143-144, 150, 162- 163, 173, 178-179, 184,
2 12
198- 199, 202, 206

Pat 25, 41, 1 15, 1 4 1, 1$7, 20$
patience 143
Patton 59
penetration point 164
penitentiary I I I
perpet 150
perpetual 151
perpetual check 108, 142, 150, 179
perpetual check derense 147
Petrosian 170
phenomenon 29
Philidor (position) 15, 22-24, 64, 96, 107,
121-122, 131, 138, 15$, 18$
Pigusov 184
Piket 146
Pikula 159
Pillsbury 181
Pinter 19, 65, 70
plan (s) 100- 107, 109- 1 10, 1 12, 1$4, 207208
plank walk 83
playing to win or draw 38
Poincare 16
Polugaevsky 164
Popovic 70
Portisch 141
positional lemon 26
postmortem 59, 206
PractiClli Chess Endings 141
practical player 20
Prandstetter 202
Prasad 186
Preservation 89
promote 12, 86, 104, 107- 109, 126, 146,
166
promote a jt 46-47, 68, 102, 133, 138,
147, 1$2
protected passer 181
Psakhis 45
pseudo-actively 121
psychological edge 140
psychological warfare 62
Purdy 44, 109, 120, 158
QE.D. 92
Q.
queen (ing, s, square Isll 12, 16, 27, 43,

45, 53, 57, 62, 71 -72, 80, 97-98, 100,
106- 107, 109, 1 18, 136- 137, 142, 149,
153, 155- 156, 159, 160, 166, 176, 179,
188
R
rainrall 17
Rashkovsky 144
rashness 61
Razuvaev 91
real life 23
Rechlis 17
resign prematurely 55
Rivas 151
robbed 10
Rodgaard 195
Rogers 38
Romanishin 30, 81, 102
Rosengol 63
Rule or Five 136, 138-139
Rule of Six 139
rule or thumb 30, 42, 49
rules 12, 41 -42, 53, 139, 143, 144
Russian Imperial Army 30
s
safety island 81
Salai 62
Salov 56, 71, 148, 166
Savchenko 126
Sax 12, 1 16
Sayre 7
Schmittdiel 139
science 16
seal off 87
Seirawan 88, 92
Serper 178
Shabalov 190
Shamkovich 170
Sherzer 194
shifting pieces back and rorth 61
Shirov 15, 34, 75
Short 49, 56, 97, 1 17, 166
Sideif-Zade 1 12
silver lining 140
Sinyavsky 94
sixth rank 39, 41
Slekys 51, 198
Index
Smirin 28, 156
Smyslov 132, 170
Sokolov 54, 89, 162
Somlai 194
soul or chess 35
Spassky 82, 196, 197
speed (and space) 26-27
Speelman 49, 73, 100
Spielmann 125
squeeze 83
stalemate 68, 108
star or the endgame 129
Staunton 51
strategic thinking 100
stroll 103
Suba 178, 179
suicide 74
Sulskis 51
Sumjakina 105
Suvorov 30
Sveshnikov 127, 165
T
tactical trick 90
Tal 82, 84
Tall 7, 31, 53, 54, 58, 131
target (s) 103, 1 12, 146, 206
Tarrasch 52, 129
Tartakower 15, 161
Tasic 106
temp (i, moves, 0) 62-63, 71, 86, 92-93,
127- 128, 164- 165, 173, 190, 192
tempo gammg 85, 94

tempo losing 85
terrible blockader 32
theory of the endgame 100
Three Commandments 54
threw (throws) in the towel 140, 153
time is cheap 64
Timman 26, 69, 102, 130, 165, 174
Torre 21, 24
touchdown 73
tournaments 1 16
trade pawns (pawns) 45
trading 12
trading material 207
trading rule 48
transition (s) 12- 13, 2 1 , 24
triangulat (e, ing, ion) 89, 91, 94, 98, 142,
172, 196
Tseshkovsky 1 16
tug-of-war 68
two vs. one 68
two-.Q advantage 206
two-square diagonals 191
Tzu 82
u
ugly positions 55
Ulybin 192
unclear positions 1 10
useless information 142
Vadasz 38
Vaganian 35, 39, 45, 89, 129, 147
Vaisman 140
van der Wiel 97
Van Wely 124
Varavin 190
Velea 108
Vetemaa 185
Vidoniak 108
virtue 143
visualizing 64
Vlahovic 159
Vogt 55
von Clausewitz 89
Yudasin 53, 133

Yusupov 58, 96, 130, 171, 176, 196
z
Zaitsev 122
Zapata 179
Zlotnik 133
zugzwang
(s) 1 1, 15, 18, 29, 56, 64, 65, 76,
80-81, 83-85, 87, 88,-89, 93, 94, 98,
1 10, 123, 125, 129, 142, 152, 154, 159,
162- 166, 172-173, 178, 188, 190-193
ZZ 80, 82, 188, 190
Waulin 57
weapon (s) 30, 97, 108
Wiesniak 109
winning check 154
winning plan (s) 71, 85
winning ugly 96
wise man (men) 135, 176
wood on the board 58
wrong-colored .Q 14
x
XuJun 200
y
Yermolinsky 87
2 13
Grandmaster Secrets:
Endings
Colophon
Grandmaster Secrets: Endgame was typeset in Berthold
Baskerville and Boton for the copy. The title page is
in Microgramma, Crimmon, and Bremen. The dia

grams were done in our
C.R. Horowitz.
Layout and design: Bob Long
Cover and Photography: Rob Long

Caricatures: Rob Long
Keyboarding and editing: Bob Long
Proofing: Bob Long and Andy Soltis

A 3.5 inch.
ChessBase disk
is available with all the
analyzed positions contained within this work-the
comments of the G M and his protege have been left

out. This disk is available from Thinkers' Press, Inc.
at the nominal price of
2 14
$ 10.00.
Accelerated Dragon B34B35 ... $ 1 8.95

s 1Editrict editors.
368 pages, t 1995.
OP97597
3.1 main tines and 234 variations! .177 root

notes. Black seems to do as well as White with a
Tiviakov claim in 199.1 of equality. Black plays
... g6 early, on the second move!
Accelerated Dragon Maroay Variation
(136139) .................................... $ 1 3.95
s 1Editrice editors.
147 pages, 1 995.
OP77882
17 main lines beginning with l. e4 c5 2. NfJ

Nc6 3. d4 cd4 4. Nd4 g6 i c4. J09 variants with
376 footnotes. Many Black wins.
little-known gambit for White... which I know

from experience to be dangerous. Black can
equalise if he's careful but there's no known refu
tation and Black can easily be hammered if he
treats the gambit without respect. .. essential read
ing for players who want an aggressive way to
meet the French and for defenders who need a
guide through the gambit's maze or variations."
Answer Guide to How to Become a

Candidate Master ........................ $4.95
Dunne.
Alekhine's Defense as White, Four

Pawns Attack ............................. $ 1 2.95
Christiansen, Raingruber,joseph.
Benoni Defence Taimanov Variation

(A67) ........................................... $ 1 2.50
Tirabassi.
OP57882
OP77529
63 pages, t 1 989.
The book which changed a lot of the think

ing on A1ekhine's Defense. Included are: Intro-
Against the Maroczy Bind ......... $ 1 9.95

CrouchiKulogin.
OP75958
lOB pages, 1995.
This is the middlegame rererence book about
the Maroczy system or the Sicilian Derense. There

are 60 basic games and 70 supplementary ones.
Games by Larsen have some prominence. The
annotations vary in length. Black will find this
book useful.
1 16 pages, 1993.
A dynamic system for causing great grief for

the Black side of the Benoni is presented by the
sequence: l. d4 Nf6 2. c4 c5 3. d5 e6 4. Nc3 ed5
5. cd5 d6 6. e4 g6 7. f4 Bg7 8. Bb5. In symbolic
notation except ror the Introduction (English).
Hundreds of games and much analysis.
293 pages, 1995.
.............................................
OP27572
50 complete and deeply annotated games!

220 rragments. White is scoring 77%! Ir Black
tries to play the French to avoid facing the BOG
he can end up with his worst nightmare! Tim
Harding says: "... a neat volume ... covering a
TR57885
The follow-up book to Dunne's successful

How to Become a Candidate Master. The ratings of
the players, answers to questions posed in the
book, and corrections to the first edition.
Big Book of Busts, The .............. $22.95

Watson&Schiller.
Alapin French, Tactics for White ........

$ 1 3.95
Sawyer.
87 pages, t 1995.
23 pages, t 1986.
duction, Knight Retreat Variation, Pawn Ex

change Deferred Variations, Ljubojevic Variation,
King's Bishop Variation, and four sections on
the Queen's Knight Variation. An essential "prob
lems" section is also included.
OP92878
Offbeat variations do gain points, but many

of them are rife with flaws! Some or the systems
included here are: Grob, Dunst/van Geet, Gloria,
Durkin, From's Gambit, Lisitsyn, Englund,
Woozle, Franco-Sicilian, Mexican Derense, Colle,
Budapest, Vulture, Staunton Gambit, St. George,
Sicilian 2 b3, Smith-Morra, Portuguese, Fried
Liver Attack, Veresov, the Torre, Albin, Djin,
King's Indian Four Pawns, and a lot more. In
cluded is a symbology index (such as the skull
and crossbones!) ranking the "status" of the open
ing from horrible to quite playable. Some are
dull. There are refutations to stupid junk.
Romanian correspondence GM Rotariu.

Blackmar-Diemer gambit keybook ....
..................................................... $2 1 .95
Sauryer.
137 pages, {: 1992.
OP57887
A book of real gambits! 700 games (includ

ing many by the author) are dissected and of
fered in the avoided and declined lines, the
Ziegler, Euwe, Bogoljubow, Gunderam (not even
mentioned in the Chess Digest book!), and the
popular Teichmann. This system barely merits a
line in ECO and yet this book is rife with White
wins, many at the master level!
Big Book of Combinations, The .........
..................................................... $ 1 7.95
Schiller.
.
280 pages, t 1994.
TR92879
There are 1000 combinations inviting solu

tion by you or your computer (sometimes the
human finds a better and more efficient win).
The author discovered in his "rating testings"
that many lower-rated players have excellent tac
tical ability but poor opening knowledge-your
cue to improvement and knowledge of yourself.
Bird Variation in the Ruy Lopez, The .
..................................................... $ 1 4.00
Rotariu & Cimmino.
1 28 pages, t 1 992.
OP72987
Making a strong comeback ever since the

release of McCormick's and Soltis' book on the
Bird Variation: I. e4 e5 2. NO Nc6 3. Bg5 Nd4.
Many games and considerable analysis by the
Bobby Fischer I Knew, The ....... $ 1 9.95

Denker&Parr.
359 pages, t 1995.
EN92972
Great stories and writing about Fischer and

the rest of the best in chess. Other players of
note are: Fine, Reshevsky, Botvinnik, Don
MacMurray (the "brain"), and more in an exten
sive index. Captioned photos and fabulous sto
ries about Capablanca, Alekhine, Simonson,
Kupchik, Lasker (the former world champion
was Denker's bridge partner), Whitaker, and
many more. Also included is a good and over
due section on Irving Chernev, one of chess'
best authors. Descriptive notation.
Cambridge Springs Variation in the
Queen's Gambit (D52-QO 1 5.2) ..........
..................................................... $ 1 2.50
74 pages, tl 1994.
OP92788
Ten major chapters + the Anti-Cambridge

Springs. This rough and tumble chess variation
begins: l. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Bg5 Nbd7
.i e3 c6 6. Nf3 Qa.i 282 games and analysis.
Caro-Kann Defence Advance Variation
(1 1 2) ............................................ $ 1 4.00
TiraiJassi.
140 pages, t 1994.
OP77558
Much more than Seirawan's ECO book. I.

e4 c6 2. d4 d5 3. e5, and unlike the French De
fense, the Queen's Bishop gets out. The lines of
coverage are: 3 ... Br.1, 3 ... c5, and 3... Na6. Also
includes the hot 3 ... Bf5 4. Nc3 e6 5. g4. 251
games and lots of analysis. One of the best varia
tions of the Caro-Kann.
the points of this variation. There are 277 varia

tions and 579 footnotes. 546 games, some anno
tated. White's chances seem to be better.
Caro-Kann Defence Knight Variation
4 f6 (B I 5-B I 6) ..................... $ 1 2.50
Curtacci.
96 pages, !: 1992.
OP72982
Popular with such GMs as Karpov and

Seirawan, this "secret" weapon contains lines not
as numerous nor as complex as others in the
Caro-Kann. lOOs of decisive games. Indices, figu
rine algebraic notation, English intro.
Caro-Kann Defence Classical (B 1 8-B 1 9)

$ 1 9.95
...............................................
176 pages, t' 1995.
OP97727
A considerable sum was spent on the re

search for this book and the information is worth
it. The 14 main lines run: I. e4 c6 2. d4 d5 3. Nc3
de4 4. Nxe4 Bf5. There are 85 variations and
316 footnotes. A very important system in the
Caro-Kann.
Caro-Kann Exchange Variation (B 1 3)
..................................................... $ 1 6.25
Luccioni.
199 pages, t 1995.
OP77857
52 lines in 18 sections are documented be

ginning with: I. e4 c6 2. d4 d5 3. ed5 cd5. An
extensive two-page English introduction explains
Chess Master
Wel<,eU.
at any age ........ $23.95
301 pages, t 1994.
TR79778
One buyer, an investment counselor, told

us this was one of the two best books he had
read on chess because there is a definite plan-a
plan which worked for Rolf Wetzell. After years
of frustration and trying and trying, Wetzel! dis

covered it was time to change approaches from
all the traditional methods. At age 50 he made it.
Showing exceptional insight into human nature,
Wetzell works on his system throughout the book
and appends 16 of his own games. He is con
cerned with the study of the buyer'S games-noth
ing else-what an innovative idea! If you want to
be a master, you have to master yourself. GM
Lev A1burt told me has recommended this book
to his students.
Chess of Bobby Fischer, The .... $24.95

Burger.
345 pages, 1975, 1994.

B192972
Few have been qualified to dissect Fischer's
play the way master/psychologist Burger is. The
Complete Guide to Correspondence

Chess ........................................... $ 1 4.95
Dunne.
RE58285
147 pages, t 199 1 .
For far too long postal players have been
neected in both their history, games, and analy
ses. In this handbook Dunne sets the records
straight, explains the rules, includes famous
games, and offers an extensive glossary of the ins
and outs of the various types of correspondence
play. Some of the best play comes from corre
spondence chess.
Dutch Defence Leningrad Variation 7...

4)c6 (AI') ..................................... 1 4.00
Luccioni.
157 pages, t 1994.

OP87982
One of the most important lines of the Dutch
Defense, a fighting defense! The 7 . Nc6 line is
extremely critical, and is considered the main
line. 515 complete games!
.
English Opening (A2 1 ) .............. $ 1 2.50

Tirabassi.
OP77557
96 pages, 1993.
A really hot variation in the English, cov
ered in much less detail in New in Chess #28.
Features 200 unannotated games plus plenty of
analysis. The introduction is in English and there
are three major lines and four minor variations.
contents: The Endgame, The Name of the Game,
Combinations, Ideas,Just for Fun!, and The Logic
of Chess. The material is very revealing and help
ful to all levels of players. Full of ideas.
English Opening Lukin's Variation, The

(A21 ) ........................................... $ 1 6.50
60 pages, t 1994.
OP95525
White's having major problems with: 1. c4

e5 2. Nc3 d6 3. Nf3 f5 4. d4 e4. This novel idea is
being propagated by Andrei Lukin from St. Pe
tersburg. There are 132 games (including par
tials) and lots of evaluations ending in the "un
clear" appellation of this "killer system" for Black.
the Winawer variation. A very comprehensive

study showing how complex this system is to
tackle. Necessary for all French players. Exten
sive bibliographic listings are included.
FlohrMikenas (A 1 8A 1 9) .......... $ 1 4.00

Konikowski.
TR72872
346 pages, 1993.
Unlike many books on improvemen the
author takes you through a subject (an ending, a
combination, etc.), shows you diagrams from real
play, and then ofTers an explanation of what re
ally happened or could have happened. It's done
from the viewpoint of the endgame/ middlegame
emphasizing the winning capabilities of pawns
as well as all the usual tactics and strategies of
chess. 8 chapters, exercises, and solutions. 700
diagrammed examples.
145 pages, 1994.

OP92787
The system goes: I. c4 Nf6 2. Nc3 e6 3. e4
and there are 180 selective games by good play
ers and tons of analysis. Black's two moves of
merit are 3 ... d5 and 3 ... c5. It is very important
for Black to realize the implications of these varia
tions-White can easily set Black on his ear be
cause the system looks "weird." It isn't.
French Defence Tarrasch 3 a6 Varia

tion COl ...................................... $ 1 7.00
Luctioni.
Genesis of Power Chess, The ... $25.95

Ault.
...
2 15 pages, 1995.
OP97527
Black stops the Bishop from going to b5,
although allowing the possibility of an isolated
pawn which often gives Black good play. 24 lines
to consider. Often Black will transpose by play
ing 3 ... c5 first. 24 lines with 405 games/partial
games. An excellent English introduction.
French Defence Winawer Variation

C 1 5 1 9 ......................................... $ 1 7.50
Myers.
120 pages, 1994.
OP87987
27 chapters containing one or two exten
sively annotated games on critical responses to
Grunfeld Indian Exchange Variation

DIS .............................................. $ 1 6.50
s 1Editrict editors.
OP95299
74 pages, 1994.
The wholesale liquidation of forces begins
with: 7. Nf3 c5 8. Be3. White/Black is trying to
eke out a slight edge. With the introduction of
7 . . . c5 it seems play is now in White's court.
Important contributions from Karpov, Kaspar
ov, et al. 247 games/game fragments.
Guide to Good Chess (I I th ed.) .........

$ 1 6.95
Purdy.
.............................................
IN78295
138 pages, 1996.
Perhaps the best guide to chess improve-
ment ever written by one or its greatest chess

writers. Purdy had a knack for knowing where
players needed improvement. He uses excellent
examples and uses exactly the right words you
need to read. He believed chess has certain "real"
principles which will help you find the right
moves. They are listed with their exceptions. One
of the major Purdy books. Now in an algebraic
edition.
Henrique Mecking, Latin Chess Genius
..................................................... $ 1 7.95
Gordon.
172 pages, t 1993.
BI72989
Included are 344 games, many seldom seen,

many not seen, and most missing from the major
databases. 24 games are annotated and dia
grammed. Also included are: a biography,
crosstables, his opening repertoire, and recent
news or his comeback attempts. The most com
plete biography or a modem day chess player in
quite some time-the best to come rrom South
America. His games epitomized tension and pre
paredness-never one arraid or current theory.
and Dale Carnegie-like "positive" messages have

made this a perennial favorite among aspiring
Expert chess players.
How to Create Combinations .. $ 1 7.95
Pafoutieff.
200 pages, C 1996.
TR90255
This instructive work attempts to break down

the combinational process into its components.
Then example afier example drills the ideas into
your chess brain. Fourteen chapters illustrate the
various kinds or combinations. Complete solu
tions are also given. There are also 70 games
rrom the author's own play.
Italian Game, The (C53C54) ... $ 1 6.50
103 pages, t 1994.
2 1 2 pages, t 1995.
OP58295
More games, problems, and solutions.

KGAW is built like a textbook. It's purpose has
been to successfully teach how to play winning
systems in the King's Gambit and to deflect all of
Black's counterproposals, 16 chapters of them! It
is loaded with restorative analysis, many new
games, and sections offering totally new theory
OP95528
Journal of a Chess Original ....... $ 1 9.95

Ger<Ptiowia..
TR58288
Postal correspondence master and wielder

of the literary pen, Gerz takes us through his
stint in the 8th U.S. Correspondence Chess
Championship Finals with the strongest players
in the country. During his Midlife Crisis(?) he
plays l. e4!!. The USCF 88 Absolute pitches the
strongest chess gladiators against each other. Fi-
Fifty rully annotated games offering a wide

variety of openings, with the view or taking a
Class A player and turning him into an Expert
(2000-2199 Elo). Dunne shows how the expert
"thinks" and how much you have to exert your
self to become one. The very interesting notes
King's Gambit As White, The 3rd ed.

..................................................... $22.95
RIlingruher&Maser.
Includes the Giuoco Piano and the Evans

Gambit (see Conquest-Kaidanov, an important
game). 288 games/partials rererenced. Besides
new games, many older ones are included which
have withstood time's tests. Ofien the "main"
lines are the older games!
How to Become a Candidate Master

(3rd printing) .............................. $ 1 8.95
Dunne.
252 pages, 1992.
nally, the Gerzadowicz Cup has games against

strong Northeastern players ror one or the weird
est trophies ever seen! Packed with analysis and
humor. His first book Journal ofa Chess Master is
now out or print.
173 pages, t 1995.
King's Indian Four Pawns Attack (A66.

A69) ............................................... 1 4.95
79 pages, t 1995.
OP97785
A transpositional Goliath, the "main" way

goes: l. d4 Nf6 2. c4 c5 3. dj e6 4. Nc3 edj i
cdj d6 6. e4 g6 7. r4. The variation A67 has been
dealt with in The Benoni Defence Taimanov
Variation A67. 8 "main" lines have jj variations
and 137 rootnotes.
Labourdonnais System in the Sicilian
Defence (B32) ............................ $ 1 8.95
Neumann.
269 pages, t' 1995.
OP78289
Harping on the move ... e5 for Black, this

time it comes early: l. e4 d 2. Nf3 Nc6 3. d4 cd4
4. Nxd4 e5. English/German introduction. 416
games. There is much new theory to give Black's
game a strong impetus. 45 lines of play with 675
rootnotes. There are many wins by Black and
improvements for Black's play.
LDL Sicilian .................................. $5.00
Dunne.
34 pages, t 1987.
B189778
as Black desperately looks for a way to combat

the Kieseritzky system. Even 2 ... Nc6 is included.
Nearly 20010 new material.
OP58298
Master (over the board and in correspon

dence play), educator, and writer, Dunne has
uncovered a little known line against the Sicil
ian, based on a suggestion by rormer world cham
pion Emanuel Lasker. Dunne has gone on to win
many games with this variation. In this book he
gives the variations and 18 of his own games. A
players' index with relevant lines are included
on its # 10 business envelope-sized pages.
Modem Chess Brilliancies

Evans.
281 pages, 1910, 1994.
.......
$ 1 9.95
GC92882
Fantastic games with penetrating notes. It

be frenetic to see how fast one's game can
collapse due to an inaccuracy. The best and the
unknowns are seen here in 101 games. Tal's play
is represented 21 times-the master of !!
can
Modem Postal Masterpieces .... $ 1 6.95

Dunne.
1 4 1 pages, 1994.
CR81158
Fifty games by some of the best in the corre

spondence chess business. Of those, twenty seven
are deeply annotated. Among these gems you
will find Aronoff, Eisen, Bovay III, Dunne, Zil
berberg, Spitzer, Palciauskas, Osbun, and oth
ers. Some of these people have real lives and
you'D get to find out about them (photos of many)
and a wide variety of openings. There are also
23 amazing unannotated games.
New York 1 99 1 ........................... $ 1 1 .00

62 pages, 1994.
. GC95289
100 games with notes to the last ("best") two.

A selection of cream from the 600+ participants.
Won by G M Goldin, ranked 20th at the begin
ning of the event.
Nimzo-Indian Capablanca Variation
El2-El' ....................................... $ 1 9.95
321 pages, 1995.
OP91128
From 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2

with improvements for White. 56 "main" lines
and 357 variations. 795 footnotes. Many games
and evaluations. A monster-sized book.
On Top of The Chess World, The 1 995
World Chess Championship
$ 1 4.95
Christwnsen, Fedorowia.&I. Gurevich.
....
129 pages, 1995.
MA98121
An informative and entertaining book on

the big event of 1995, the match between Kas
At one time Viktor Kortchnoi was the sec

ond strongest (some say "strongest") chess player
in the world ( 1974-1978). After being the first
chess defector from the Soviet Union, he played
Karpov, in effec two more times for the world's
chess title. In fac according to the Divinskyl
Keene scale, Kortchnoi is the 7th strongest chess
player of all time and was the only non-world
champion in the top 10! This book was pub
lished to detail the trials and tribulations of an
ex-patrio a man who became very much the
anti-Soviet establishment. His story, and the
games from the controversial 1978 match in Ma
nila, present a picture of a complex man who
had an unswerving desire to win and yet is just
as able to believe in the arcane "arts" of mysti
cism to accomplish his goals. Contains previ
ously unpublished photos, and many documents
in the appendix not previously published. Origi
nally published in a smaller edition, in German,
as "Anti-Schach."
parov and Anand. Experience everyone's exas

peration at the contested "draws" in the first 8
games, and then the unleashed fury which en
sued after that. Relieve Kasparov's enthusiasm
for using the Sicilian Dragon in crucial games.
Wonderful commentary, analyses by three of the
best in the U.S., and photos make this a keep
sake to look at many times.
I
hhlr.(
Persona Non Grata ..................... $8.95

Kortchnoi & Cavallaro.
1 41 pages, 198 1 .
BI58522
Petrosian lays it on the line, in his experi

ences with Fischer, San Antonio 1972, and what
he believes is necessary to teach chess players
how to play better and learn more. Very frank.
Pirc Defence Czech Variation 3 c6
107 .............................................. $ 1 6.25
..
Pieri.
203 pages, 1995.
OP18292
49 lines in this "chameleon-like" defensive

system for Black. 265 variations with .:;91 foot
notes. Its characteristics are the indirect attack
on the center, and defense and counter-attack on
the King-side. An interesting introduction about
the vicissitudes of this system.
Play the Schliemann Defence! .. $ 1 9.95
V. Ivanov&Kulagin.
112 pages, t 1994.
OP28828
A detailed study of this wily attack (oo.f.:;)

against the venerable Ruy Lopez. 330 games are
annotated There are 35 core or basic games, 95
supplemental games, and 205 study games.
Queen's Gambit Accepted (D20-D29)
..................................................... $ 1 8.95
s 1E editors.
119 pages, t; 1995
Petrosian's Legacy
Petrosian.
123 pages, 1990.
..........
1 2.95
BI59881
OP29958
Black looks for a way to relieve his cramped

game in the regular Queen's Gambit, and the
Accepted version allows this. White and Black
have so many different plans at their disposal
that this book covers them in six main lines and
566 games. A very limited supply.
Russian for Chessplayers ........... $ 1 2.95

Russell.
53 pages, 199 1 .
RE58529
such new stars as Kamsky and Shirov many op

portunities to wrest the initiative from White's
hands. Analysis is very extensive for both sides.
\eteran translator of Russian books on chess,

Russell has updated a book he had published
twenty-five years earlier. 2,000 copies sold out in
a short amount of time. Now a considerable
amount of new info has been added to the book
including a very much expanded dictionary of
chess-related tenns. You will see how to translate
chess newspaper columns and "read" verbs,
nouns, and adjectives. To help with the pronun
ciation of many grandmaster names appearing
in the Cyrillic alphabe a six page addition gives
the correct spelling of 17j names and how to
pronounce them (with the proper stress).
60 pages, t 1996.
Sicilian Defence Closed Vol. I (B23)

..................................................... $ 1 6.95
Neumann.
176 pages, t 1996.
OP87985
A Grandmaster Invitational Chess Tourna

ment in the U.S. is a rare event. To commemo
rate the 50th year of the signing of the UN's
Charter in San Francisco, champions from all
over the world were brought together. The event
was won by Kortchnoi. Full of fighting chess,
new ideas, and weird public relations! A treasure
trove of excitement.
Ruy Lopez Arkhangelsk System (C78)

..................................................... $2 1 .95
Konikowski.
283 pages, 1995.
OP52552
206 annotated games arranged from 12 chap

ters of possible play by White. The fianchettoing
of Black's Queen's Bishop is a system begun by
players in Russia in the 1960s and has offered
OP33466
From the opening l. e4 6 2. NfJ d6 3. d4

cd4 4. Nd4 Nf6 i fJ. Black responds i .. ej 6.
Nb3, 6. Nbj, or 6. Bbi There are 6 lines, 93
columns and 77 footnotes.
San Francisco 1 995 .................... $ 1 4.95
132 well-annotated games plus reams of

analysis. Variation 3. Bbj a6 4. Bc6. Nine chap
ters including the 5. 0-0 and i Nc3 variations.
This opening has not faded at all and must be
known by all Lopez players.
137 pages, t 1996.
OP99279
The final volume. l. e4 c5 2. Nc3 Nc6 3. g3

g6 4. Bg2 Bg7 i d3 (there also is 5. Nge2) and
replies i . d6, 5 ... e6, and 5. . . RbS. There are
16 lines/variations with 2jl columns spreading
out the infonnation ECO style. 274 footnotes.
.
Sicilian Defence 5.fl (B54) .......... 1 4.95

s 1Editriee editors.
Ruy Lopez vol. I Exchange Variation

(C68) ........................................... $ 1 5.50
Falchttta.
1 75 pages, 1994.
rigor in the tactical fight and by the complexity

of its strategy." Fifty-two lines offer the player
with the black pieces plenty of chances to win.
Em1e.
1 47 pages, 1995.
MA98872
SemiSlav Defence Botvinnik Variation

(D44) ........................................... $ 1 5.50
Konikowski & Thesing.
244 pages, t 1993.
OP77572
A huge book, twice its announced size. The

Botvinnik Variation is "characterized by extreme
OP98879
23 lines. White does not aim for d4 but in

stead usually plans d3 in combination of f4 or a
King-Side fianchetto (see vol. 2). The primary
purpose is to force the "tactical" player with the
Black pieces to play positionally. 124 variations
coupled with 412 footnotes.
Sicilian Defence Closed Vol. 2 (B24.
125) ............................................. $ 1 8.95
Neumann.
234 pages, t 1996.
0P98892.
White will often try to expand from his

fianchettoed position by means of f4-g4-f5. Black
plays on the Queen-side. 30 lines with 182 varia
tions and 473 footnotes.
Sicilian Defence Closed Vol. 3 (B26) ...
....................................................... 1 6.50
Neumann.
Sicilian Defence Najdorf Variation 7 ...

if1c7 (7 ... 4)bd7) [B96] .............. $ 1 2.50
Curtaea'.
93 pages, t 1993.
OP72985
The Najdorf, as a counterattacking defense,

is complicated. It made its appearance for the
first time in a while in the Short-Kasparov world
championship match. It was played at least six
times. English introduction and two important
chapters covering eight significant variants.
Sicilian Defence Najdorf Poisoned Pawn
(B97) ............................................ $20.50
s 1Editria editors.
1 47 pages, t 1995.
0P97787
12 lines from: 1. e4 cj 2. Nf3 d6 3. d4 cd4 4.

Nd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Qb6. Black
wins a pawn and tries to hang on in defense of a
less developed position. Fischer won a lot of
games with this line, milking Black's advantages
to the last drop. 76 variations and 299 footnotes.
Sicilian Defence NajdorfVariation (B98.
199) ............................................. $ 1 8.50
Curiaea'.
292 pages, t 1993.
OP77559
Over twice the size which was originally

planned, this book starts with most variations on
King" Griffiths, GM Browne, 1M Donaldson, 1M

Watson, and world-champion-to-be Bobby
Fischer-and, oh yes, the "infamous" James
McCormick! What a book! Theoretical open
ings are included.
the 10th move. The index alone is 10 pages and

21 important lines. jjO complete games which
make this book the definitive publication for the
variation which begins 6. Bgj e6 7. f4 Be7.
Soltis Variation of the Yugoslav Attack,

The .............................................. $ 1 9.95
Mayer.
333 pages, t 1995.

0P92877
An extremely well-researched book on the
blocking of the h-file against White's attack in
the Dragon. This is one of the most popular re
plies used by Dragon players. Over 600 games
are examined. All important transpositions have
been noted. This has become a main line. 29
chapters. Many masters recommend this book.
Strategical Themes ................... $ 1 2.95

Unger.
TR59828
102 pages, t 1989.
Master Unger believed that there were at
least subjects four which were poorly understood
by most players: the "bad" Bishop; the double
fianchetto; the central pawn roller; and central
ization. Tom even included a few of his own
games. Another Master, a former state cham
pion, thought this was one of the best discussions
of these subjects. The "data" is in the form of j8
annotated games and 7 supplemental non-anno
tated games. Deserves to be widely read. First
rate game selections.
Thinkers' Chess .......................... $ 1 7.95

GerqulowiC(.,
160 pages, t 1 995
TR87277
Tbe Correspondence
Cbess Yearbooks
These books from s1E in Italy contain the
best of correspondence chess for each period they
cover. The games are annotated. Many issues
contain special theoretical articles, ICCF infor
mation, rating lists... these are the Chess Infor
mants of correspondence play.
eeya #02 (OP28552) ............... $25.95

eeya #03 (OP72855) ............... $25.95
Twenty six games, deeply annotated to show
the inner workings of players rated between 1400
and 2400. The games were all chosen from the
readership of The Chess Ga<ette magazine for
reader interest and instructional value and supple
mented with ideas and brain waves from postal
master Stephan Gerzadowicz, a very popular
writer on the chess scene. The author probes
each game with humor, analysis, parallel think
ing, and compliments. A totally different con
cept in book authorship.
Viktors Pupols, American Master .......

....................................................... $6.50
Parr.
89 pages, t 1983
8159852
Unlike any chess book you have read! Meet
the Latvian-American "Unkel Vik," strong mas
ter, winner of nearly 100 tournaments, poe open
ings dialectician, brawler, and terrific gamesman.
Supporting cast includes: GM Seirawan, "Speed
eeya #04 (OP72857) ............... $25.95

eeya #05 (OP78227) ............... $25.95
Win at Chess! ............................. $ 1 8.95

Curry.
1N79779
272 pages, t 1995
This expert has been teaching improvement
to amateurs and average players for years with
his own special brand of philosophy, techniques,
and openings. The success and comments of his
students offer proof of his methods. Besides the
usual necessities concerning chess (notation, sym
bols, etc.), the other parts discuss the opening,
the middlegame (with special emphasis on tac
tics, combinations, and sacrifices), the endgame,
and practice and progress.
eeya #06 (OP72858) ............... $25.95

eeya #07 (OP72859) + Chess Theory
#1 ................................................. $22.95
#2 ................................................. $23.95

#3 ................................................. $23.95
eeya # 1 0 (OP87897) ............... $22.00
eeya #1 1 (OP89779) ............... $22.00
eeya # 1 2 (OP92958) ............... $22.00
eeya # 1 4 (OP77852) ............... $22.00

eeya # 1 5 (OP98895) ............... $2 1 .00
Any books or equipment which you cannot
obtain through this business may be ordered
rrom: Chesseo, 125 Kirkwood Blvd., Daven
port, Iowa 52803. A he copy or our catalog
will be included,
Titles in this catalog shipped postage

paid.
..Jf?r'"IJ!A.
'
'.
,.
.
. .
....
'
' ,
" ,:.
! .. :
, .'
( ..
.;.:
"
. .
' ,;
' .
.
fi.:
"
Thinkers' Press, Inc.

p.o. box 8
davenport
IA 52805-0008
united states of america
1967
52000

Andrew Soltis - Grandmaster Secrets, Endings

Uploaded by

Copyright:

Available Formats

Andrew Soltis - Grandmaster Secrets, Endings

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

Andrew Soltis - Grandmaster Secrets, Endings

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics does the book cover?

What topics does the book cover?

GRANDMASTER

GIVI Andrew Soltis

Thinkers' Press, Inc. Davenport IA u.s.a. 1997

Grandmaster Secrets: Endings

Copyright 1997 by Andrew Soltis

permitted by the 1976 Copyright

Requests for pennissions and republication rights

Grandmaster Secrets: Endings

Grandmaster Secrets: Endings

Grandmaster Secrets: Endings

teaching the endgame by means of a

I visualized a conversation, punctuated by diagrams and simple analysis. It

Grandmaster Secrets: Endings

In the spring of 1993 I was in between books when I exchanged thoughts

bewildered by trying to figure out what they really need to know."

"terribly designed and poorly written." The information presented is often ar

Enter Pat Sayre, a talented, but young club amateur. Sayre

Ctt t:1 , t[. , ,'f.l'."it;:"

What Every Grandmaster Knows

Loses time compared with 11,g7

You left your passive.

The h- i promotes in two

Pat: It's not that I'm so great in knowledge are so extremely

What Every Grandmaster Knows

The Estimated Odds of Your:

Ever Going Bald if male

Ever Dying of a Heart Attack

Being Robbed This Year .

+ +4) vs. This Year ................

Being Diagnosed with Lung Cancer This Year

*+4)- ft + vs. *+ This Year ............

Being Murdered This Year .................................... 1 in 12,000

emerge a clear ahead and win

After Black puts his at some

mean like that you

What Every Grandmaster Knows

Pat: So what does it take to Noah: As with most rules, this

Candidates match playoff 199 1

Much stronger is 1. h5!, which

Ofcourse, 3. *xg6?? g7Ioses.

White can only try to win now

White played another nine

<lie3 13. <lin g3! and wins quickly.

Pat: But don't I have to know trading is when you're

What Every Grandmaster Knows

"Don't you know this is a

-Pal Benko to Walter

This position (Diagram 8) has

Pat: So this is stuff I really

"Chess has three phases:

What Every Grandmaster Knows

Black must prevent the from

iard, as a matter of fact. Actu

"Bum all the books! Chess

"Chess Un't a science. "

with 5. f5?? xf5t 6. Wxf5 '1t>e7

+4)- ft + vs. + This Year ............