The Keres-Botvinnik Case: 
A Survey of the Evidence 
 
by Taylor Kingston

Part I

The Chess Cafe is pleased to welcome Taylor Kingston as a regular
contributor. Taylor is a senior systems engineer by vocation, 
an avid chess historian by avocation. A national correspondence 
chess master, he lives with his family in the northeastern United 
States in Shelburne, Vermont.

The case of Paul Keres and Mikhail Botvinnik is a chess equivalent
of the John F. Kennedy assassination: an unsolved historical
mystery full of dark implications, about which conflicting opinions
and theories abound. Debate about the case has variously simmered
or boiled for decades, and interest in it has recently increased. 

Background of the Case 

For those not familiar with the case, some background. In the
1930s Paul Keres (1916-1975) of Estonia emerged as one of the
top chessplayers in the world. Based on his victory at AVRO, 1938
(then the strongest tournament ever held), he challenged world
champion Alexander Alekhine. Title match negotiations were
under way in 1939, but were disrupted by World War II. During
the war Estonia was seized first by the USSR, then by Germany,
then the USSR again. 

As a Soviet citizen, Keres' results dimmed somewhat, as he was
continually eclipsed by Botvinnik (1917-1993), strongest of the
Russian players. After the war Keres faced the threat of arrest, for
the crime (by Soviet standards) of having played in Nazi-organized
tournaments during the German occupation of Estonia. He was
under virtual quarantine during the investigation. Eventually he
was cleared, but he got no match with Alekhine, and was absent
from some major post-war tournaments.

Alekhine died in 1946, and the Hague-Moscow tournament of
1948 was held to fill the vacant title. Botvinnik (14-6) dominated,
ahead of Smyslov (11-9), Keres and Reshevsky (10«-9«), and
Euwe (4-16), scoring 4-1 against Keres. In the following years
Keres never got a title match, though he came very close, finishing
2nd four times in Candidates tournaments (an unequaled record). 

Keres' failure in 1948 gave rise to speculation that he was forced
by Soviet authorities to throw games to Botvinnik, who as a native
Russian and loyal Communist, was a more desirable world
champion by Party standards than the non-Russian, politically
tainted Keres. Keres remained virtually mum on the subject, and
Botvinnik denied any chicanery in what little he said, but the
controversy has never died. Some insist the KGB (Soviet secret
police) held a gun to Keres' head, others dismiss this as anti-
Communist paranoia, saying Botvinnik was simply the better
player. 

Seeking Smoking Guns and Sifting through Mountains 

Recent articles in major periodicals, discussing newly uncovered
evidence, have again raised interest in the Keres-Botvinnik case to
a high level. Debate centers on two major questions: one, was
Keres indeed coerced to lose to Botvinnik, in 1948 or any other
time? Two, if he was, what was Botvinnik's degree of complicity?
A suitably dramatic solution would be discovery of a "smoking
gun": an incriminating document or a confession from someone
privy to the conspiracy. However, none has appeared. Were I a
well-funded Russian-speaking investigative reporter, I would be
searching for one. 

In lieu of that, I have, for over a year, been analyzing every relevant
source I could find: books, websites, magazines as far back as
1948, correspondence with international players and journalists.
This has meant sifting through mountains of opinion and
conjecture to glean a few nuggets of fact, and sometimes finding
maddening paradox and contradiction. Yet it has yielded some very
interesting results. 

Principles and Prejudices 

In this, I've tried to adhere to the principle of "innocent until
proven guilty." As an American raised during the Cold War era, it
would be very easy to let anti-Communist prejudice steer my
thinking along stereotyped lines of nice guy Keres versus cold fish
Botvinnik, Stalinist avatar of the Evil Empire. Evidence must lead
one to conclusions, not vice versa, and Botvinnik must be given the
presumption of innocence. We are, after all, discussing allegations
which could indelibly taint him with shame and scandal. One
should not jump to such conclusions lightly. 

Another caveat. Certainly the totalitarian Stalin regime under
which Keres and Botvinnik lived was oppressive and brutal, with
overt and covert coercion being standard procedure. Chess was
highly politicized and regimented. However, the prevalence of
coercion does not prove its role in any particular case. To assert it
only on grounds of probability, without evidence, is sloppy and
irresponsible, and such arguments would be thrown out of any
decent court. The burden of proof is on the prosecution. 

The Extremist Viewpoint: Schroeder

Such principles apparently are alien to James Schroeder, who in
the 4/96 issue of the British monthly "CHESS" espoused the most
extreme view I have found on the Keres-Botvinnik case.
Schroeder, an American from Ohio, accuses Botvinnik of doing
"everything in his power to destroy Keres". 

His allegations include: that Keres was barred by Botvinnik
personally from postwar tournaments such as Groningen 1946; that
Botvinnik was "the only reason why the KGB arrested Keres"; and
that the KGB forced Keres to lose not only in 1948, but also in the
1941 Soviet Absolute Championship and in all his later FIDE
Candidates competitions (the last of which was in 1965). 

This is at best speculation and at worst rubbish. Schroeder's forays
into chess history often exhibit strong prejudice, and are rife with
fallacies. To detail just a few from his "CHESS" article: 1) It is
well established that Keres agreed to, or had imposed on him, a
condition not to interfere with Botvinnik's attempt at the world
title. However, as stated clearly by Hooper and Whyld in "The
Oxford Companion to Chess," this applied only to a match with
Alekhine. In 1946 Keres was, by virtue of AVRO 1938, still an
official FIDE challenger. He ceded this to Botvinnik, a right which
became moot when Alekhine died. 

Schroeder, however, goes far beyond this, asserting that "The KGB
... told [Keres] he could NEVER become World Champion [nor]
win the Candidates Tournament", for which he cites no support
whatever. Nor, as we will see, has the most diligent firsthand
research found any. 2) The idea that the KGB "arrested Keres"
only at Botvinnik's instigation is almost certainly nonsense. 

Firstly, as will be seen, there is doubt that Keres was ever actually
arrested. Secondly, his offenses: playing chess for the Germans,
and naively making mildly anti-Soviet remarks which appeared in
the Nazi press, seem small or excusable in Western eyes, but were
serious by the harsh standards of Stalinist Russia, where a hint of
disloyalty could mean death. Keres was also suspected of aiding
Estonian patriots. During 1945-49 the Soviets "deported" at least
60,000 Estonians, and they were not all rivals of Botvinnik. The
KGB needed no prompting from him to investigate Keres, and
again Schroeder cites no support. 3) His assertion that Botvinnik
personally kept Keres out of several postwar international
tournaments is equally baseless. These were during Keres' political
quarantine, and authorities would have routinely denied him
permission to travel abroad. In 1945 Keres was not even allowed to
go to Moscow for the USSR Championship. 4) The only
authoritative source Schroeder cites is a 1995 "New In Chess"
article by Valter Heuer (more about which later). 

Schroeder writes: "Heuer states that Leonard Barden and other
masters euphemistically state '... the price for Keres' reprieve from
execution was an agreement with Soviet officials not to hinder
Botvinnik's ambitions to be World Champion and remain World
Champion.'" In fact no such passage exists in the article. Its only
similar quote from Barden says "Yet apparently the price for
Keres' return was just a vague promise with Soviet officials not to
hinder Botvinnik's campaign", which has a very different sense.
Note especially his replacing "return" by "reprieve from execution"
and his exaggeration of the meaning of "Botvinnik's campaign"
(see point #1).

Fabrication? At best, flagrant inaccuracy. Schroeder's article is
based more on opinion, perhaps even imagination, than fact.
Unless authoritative support surfaces, his views must be regarded
as more akin to tabloid headlines like "Elvis was a UFO Alien".
Even "CHESS" now disowns him. In response to detailed critique
I sent to London, a senior "CHESS" editor admitted that in places
"Schroeder is just speculating ... [or] is just wrong," and that
publishing the article "was a clear error of judgement". 

Somewhat Less Extreme: Evans 

The interest of many "Chess Life" readers was rekindled by GM
Larry Evans' article "The Tragedy of Paul Keres" in the October
1996 issue. While Evans (mostly) avoids Schroeder's extremes, he
finds "little doubt that [Keres] was forced to take a dive." At first I
found Evans' case persuasive (explaining two highly laudatory
letters of mine written in late 1996 but not published in CL until
8/97 and 10/97). Since then my research has raised doubts. Among
his supports Evans cites: (1) "Newly opened KGB files"; (2)
Schroeder's article; (3) an article by Soviet master Feodor
Bohatirchuk; and (4) analysis of Keres' 1948 games against
Botvinnik. Let's examine these. 

A careless reading might lead one to think that Evans possesses
once-secret Soviet documents on Keres. If so, he is not
forthcoming about it. The only one quoted, a letter from Keres to
Foreign Minister Molotov, was unearthed and published by the
aforementioned Valter Heuer. In it Keres begs not to be
stigmatized for playing in Nazi-organized tournaments, and pleads
for reinstatement as a Soviet grandmaster. Relevant, but not, as
Evans implies, directly illustrative of a plot to make Keres throw
games. 

I asked Evans if he had a "smoking gun," stating clearly the KGB
told Keres "lose or we kill you." He admitted (CL, 4/97, p. 28) "I
doubt such a document will ever surface." He then cited the
aforementioned "Oxford Companion" entry: "In return [for official
forgiveness] Keres promised not to interfere with Botvinnik's
challenge to Alekhine." That, printed in 1984, is nothing new. Its
meaning is also clearly narrower than Evans' blanket statement
"the price of [Keres'] reprieve was to abandon his quest for the
crown." Evans is claiming more than it appears he can document. 

Amazingly, Evans mentions Schroeder's "CHESS" claim that
Botvinnik barred Keres from Groningen et al. As we have seen,
this is like citing a comic book in the Warren Commission Report.
I am frankly surprised that Evans would report this unsupported
claim as fact without question or comment.

The Ukrainian Bohatirchuk, writing in "Chess Review" in 1951, is
an informed character reference but not an eye-witness. Six times a
USSR championship participant from 1923 to 1934, and co-winner
in 1927, he had first-hand acquaintance with Soviet chess politics,
but he left the USSR during WW II and thus had no involvement in
Keres' postwar situation. We will return to Bohatirchuk later, but
for now suffice it to say that the Bohatirchuk excerpts Evans gives
are not definite evidence; they are rather only an earlier statement
of Evans' main thesis: that the low quality of Keres' games against
Botvinnik in 1948 indicates a fix. The bulk of Evans' article is
devoted to analysis of points in those games which strike him as
suspicious. It appears valid, insofar as he finds inferior moves by
Keres.

However, inference based on quality of play is highly problematic.
Chess is a hard game, and even the greatest players have made
horrendous mistakes. Take this position from the 1977 Korchnoi-
Spassky Candidates match. (See Diagram)

Korchnoi-Spassky, 1977
White: Kc1, Qc2, Rd1, Rh1, Bd3, Bd4; pawns - b4, c5, e3, g3
Black: Kg8, Qe7, Rc8, Rf7, Nf3,Bh7; pawns - c6, d5, f5, g4 

Most D-players would see that 32. Bxf5?? Rxf5 33. Qxf5??? Bxf5
is dreadful, yet Korchnoi, a player arguably superior to Keres, did
just that. Those seeing conspiracy in this game involving a Soviet
defector should note that Korchnoi still won the match. Alekhine
and Petrosian each once hung his queen. Nimzovitch and
Reshevsky each allowed mate in two, which Rubinstein and Szabo,
respectively, overlooked, and such greats as Smyslov, Bronstein,
Gligoric, Short and Yusupov have missed mate in one. Capablanca
once hung a piece on move 9. 

If coercion is not required to explain these blunders, why must it be
invoked to explain Keres' more subtle errors? Coercion can cause
bad play, but bad play does not prove coercion. Yet Evans insists
that Keres "left a trail in his first four losses [in 1948] for those
who are knowledgeable enough to follow it to an inescapable
conclusion." Apparently British GM John Nunn, generally
considered a stronger player than Evans, is not "knowledgeable
enough". 

Seeming to reply to Evans in his foreword to "Paul Keres: The
Quest for Perfection" (ICE, 1997), Nunn says "Attempts to
analyze the games themselves for evidence of Keres' suicidal
efforts ... appear to be misguided" and notes that Vishy Anand, in
his 1995 title match with Kasparov, committed two "elementary
errors far worse than any committed by Keres in ... 1948 ... and
nobody seriously suggests that Anand deliberately lost". Evans said
in the 10/97 CL "Obviously a player of Keres' genius isn't going to
make stupid mistakes that are easy to detect." 

Yet ease of detection was his basis in 10/96 for suspecting Keres'
21. Re1 in this position (See Diagram): 

Keres-Botvinnik, 1948
White: Kg1, Qd2, Ra1,Rf2, Ne2, Ng5, Bg2; pawns - a2, b5, c4, d3, e4, g3, h4
Black: Ke8, Qd6, Ra8, Rh8, Nf6, Be7, Bg4; pawns - a7, b7, c5, d4, e5, g6, h5 

"Incomprehensible ... 9 out of 10 grandmasters would double rooks
by 21. Raf1 without thinking twice." Evans wants it both ways:
obvious errors show Keres taking a dive, subtle mistakes show him
using his genius to cover the tracks. In this kind of logic, any and
all evidence is interpreted to support an a priori conclusion. I do
not agree with recent anti-Evans comments by "Chess Life"
readers, calling his article "wild charges", "crackpot theory" and "a
fire sale on paranoia". Evans is raising important issues. 

However, by citing Schroeder he uncritically accepts dubious
support. In the 4/97 CL he disturbingly misrepresents Hooper &
Whyld, and later (10/97) Heuer, implying the Estonian supports his
conclusion, when as we will see, the opposite is true. His analyses
are perhaps necessary to establish his case, but are not sufficient.
He admits he has no direct evidence. His conclusion is far from
"inescapable." 

Alternate Explanations and Occam's Razor 

Still, Evans is correct that "To put it mildly, Keres didn't play his
best against Botvinnik in 1948," and one wonders, if not because
of coercion, then why? Research suggests some alternatives. One
explanation may lie in military history. 

Chess masters, like most of us, do best in peace and prosperity.
Estonia for decades had little of either. Its recent history runs:
1920-39, independent; mid-1940 to mid-1941, Russian occupation;
mid-1941 to September 1944, German occupation; then Soviet
occupation until recent years. During these turbulent times, Keres'
chief rivals were Alekhine, then Botvinnik. 

A correlation of their games with historical periods is revealing. 

Keres vs. Alekhine         
1935-39     +1 -2 =5
1942-43     +0 -3 =3

Keres vs. Botvinnik
1938        +0 -0 =2
1940-41     +0 -1 =4
1947-48     +1 -5 =0
1951-56     +2 -1 =2

During Estonian independence Keres' results against Botvinnik are
equal, and nearly so against Alekhine. However, his score dips
during foreign occupations, against Alekhine under the Germans,
then twice vs. Botvinnik under Russia, finally recovering years
later. There are precedents for this. 

A reason for Lasker's poor showing against Capablanca in 1921
was his suffering in the aftermath of Germany's defeat in WW I.
Flohr's decline after 1938 was due to the travails of his homeland
Czechoslovakia. That during occupation, a distracted Keres would
do poorly against the favorite of the occupying power, is not hard
to grasp. If only Communist coercion explains his decline against
Botvinnik, logic requires that Nazi coercion be invoked to explain
his dip against Alekhine, but no one has suggested that. 

During WW II, Keres and Botvinnik faced different levels of
competition. From late 1941 to late 1945 the only top player Keres
met was Alekhine; the rest were decidedly lesser masters and a
washed-up Bogolyubov. Botvinnik meanwhile stayed keen against
fellow Soviet GMs such as Smyslov, Kotov, Boleslavsky,
Bronstein, and Flohr. Thus, like Muhammad Ali in 1971, Keres
returned to competition out of shape, and his post-war quarantine
limited his chances to play back into form. 

Botvinnik may simply have had Keres "psyched". Botvinnik states
in "Achieving the Aim": "Keres had failings which were well
known to me ... when his mood was spoiled he played below his
capabilities." When their rivalry resumed in 1947, Botvinnik
"resolved with the help of this game [against Keres in the Chigorin
Memorial] to rob my main rival ... of his confidence." (p. 110). 

Botvinnik did indeed win, and may have succeeded in "spoiling
Keres' mood" for some time. This again is not uncommon: Lasker
had Capablanca psyched at St. Petersburg 1914; Capablanca had
Alekhine's number before 1927; Tal had Fischer's in 1959. Heuer
says "There is no doubt that ... Keres suffered a certain 'Botvinnik
complex'". Botvinnik also observed that Keres had "a tendency to
fade somewhat at decisive moments in the struggle," or in common
parlance he wasn't a "clutch player". 

Sports history is rife with talented teams and players who can't or
don't "win the big one": the Brooklyn Dodgers of the 1950s, the
Los Angeles Lakers and the Dallas Cowboys in the 1960s for
example. I don't recall that cries of "Fix!" were raised when they
consistently lost to the Yankees, Celtics and Packers, respectively. 

Keres was highly variable. Immediately after sharing 1st with
Alekhine at Bad Nauheim 1936, he placed 8-9th at Dresden. After
beating the world's best at AVRO 1938, he next placed only 12-
13th against a much lesser field. Hague-Moscow may have been a
similar downswing. Finally, as Keres himself said, "Anyone can
lose to Botvinnik, he is a very strong player." 

Consider his triumph at Groningen, 1946, ahead of most of the
world's best including many un-coercible non-Russians. To insist
that Botvinnik could not succeed without a fix is a very shaky
stance. I do not claim these facts disprove coercion, but they do
show that many simpler, mundane, non- conspiracy explanations
are plausible. Occam's Razor requires that more complex
hypotheses have supporting evidence.

Copyright 1998 Taylor Kingston. All Rights Reserved.
[END, PART I]