Are twelve games enough?

Are twelve games enough?

By Michael von Keitz

One of the more controversial aspects of the world championship today lies in the actual length of the title match. Since the 1886 match between Steinitz and Zukertort, when Steinitz pronounced himself the World Champion, officially baptizing the title, the vast majority of the matches to decide the title have lasted more than twelve games. This brings to question the fairness of the current system. How efficient is it at determining the rightful champion? Would more games really change the final outcome? As chess players, we would love to see a longer match to decide the greatest title in the game, but the question here is one of truth. Are twelve games really enough, or is that too few?

We could ask the game’s many statisticians to produce graph after graph on the topic, discussing probability and variance, but we propose a different approach: we have taken the 43 title matches since 1886, not including the mini-matches from the FIDE Knockout events of the 90s, and compared the final result with the score after the first twelve games. How often would the result have changed, and a different champion been crowned?

In the results below, it is assumed the match is twelve games long, and any player reaching at least 6.5 points would be declared the winner. In many cases, this actually means the matches would not have gone the full twelve games. Until recently the champion benefited from drawn matches and retained the title if the challenger failed to actually win.

World Championship Matches after twelve games:

1886 - Steinitz v Zukertort (+6 =2 -4)

1889 - Steinitz v Chigorin (+7 =0 -5)

1891 - Steinitz v Gunsberg (+4 =5 -3)
Hypothetical best-of-twelve would have ended after Game eleven (6.5-4.5).

1892 - Steinitz v Chigorin (+3 =4 -5)

Note: The first match where the eventual World Champion would have been match-loser in a best-of-twelve format.

Mikhail Chigorin: A should-have-been World Champion?

1894 - Steinitz v Lasker (+2 =3 -7)
Hypothetical best-of-twelve would have ended after Game ten (3-7)

1897 - Lasker v Steinitz (+7 =4 -1)
Hypothetical best-of-twelve would have ended after Game eight (6.5-1.5)

1907 - Lasker v Marshall (+5 =7 -0)
Hypothetical best-of-twelve would have ended after Game nine (6.5-2.5)

1908 - Lasker v Tarrasch (+6 =3 -3)
Hypothetical best-of-twelve would have ended after Game nine (6.5-2.5)

1910 (early) - Lasker v Schlechter (+1 =8 -1)
Only lasted ten games

1910 (late) - Lasker v Janowski (+8 =3 -0)
Only lasted eleven games

1921 - Capablanca v Lasker (+3 =9 -0)
Hypothetical best-of-twelve would have ended after Game eleven (7-4)

1927 - Alekhine v Capablanca (+3 =7 -2)

1929 - Alekhine v Bogoljubov (+6 =4 -2)
Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

1934 - Alekhine v Bogoljuboc (+3 =8 -1)
Hypothetical best-of-twelve would have ended after Game eleven (6.5-4.5)

1935 - Euwe v Alekhine (+4 =3 -5)
Hypothetical best-of-twelve would have ended after Game eleven (4.5-6.5)

Note: The second match where the eventual World Champion would have been match-loser in a best-of-twelve format.

Had the match in 1935 lasted only twelve games, Max Euwe might never have
been world champion

1937 - Alekhine v Euwe (+5 =5 -2)
Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

1951 - Botvinnik v Bronstein (+3 =7 -2)

1954 - Botvinnik v Smyslov (+4 =4 -4)

1957 - Smyslov v Botvinnik (+4 =6 -2)

1958 - Botvinnik v Smyslov (+5 =5 -2)
Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

1960 - Tal v Botvinnik (+4 =6 -2)
Hypothetical best-of-twelve would have ended after Game eleven (6.5-4.5)

1961 - Botvinnik v Tal (+6 =3 -3)
Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

1963 - Petrosian v Botvinnik (+2 =9 -1)

1966 - Petrosian v Spassky (+2 =10 -0)
Hypothetical best-of-twelve would have ended after Game eleven (6.5-4.5)

1969 - Spassky v Petrosian (+3 =6 -3)

Note: Another match where a hypothetical best-of-twelve would have potentially seen the eventual World Champion end up match-loser (assuming the World Champion retains title in the event of a 6-6 draw)

Boris Spassky might have been stricken from the list in a twelve-game match

1972 - Fischer v Spassky (+5 =4 -3)

Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

If Petrosian had defended his title, then today we might be reading about the
1972 Petrosian-Fischer match instead

1978 - Karpov v Korchnoi (+1 =10 -1)

1981 - Karpov v Korchnoi (+4 =7 -1)

Hypothetical best-of-twelve would have ended after Game ten (6.5-3.5)

1984/85 (abandoned) - Karpov v Kasparov (+4 =8 -0)

Hypothetical best-of-twelve would have ended after Game nine (6.5-2.5)

In twelve games, Anatoly Karpov would have kept his title in 1984...

1985 - Kasparov v Karpov (+2 =8 -2)

Note: Third match where a hypothetical best-of-twelve may have been insufficient

1986 - Kasparov v Karpov (+2 =9 -1)

1987 - Kasparov v Karpov (+3 =7 -2)

... and Garry Kasparov would have had to wait until 1987
to get a second shot at it.

1990 - Kasparov v Karpov (+1 =10 -1)

1993 (FIDE) - Karpov v Timman (+3 =8 -1)

Hypothetical best-of-twelve would have ended after Game eleven (6.5-4.5)

1993 (PCA) - Kasparov v Short (+5 =7 -0)

Hypothetical best-of-twelve would have ended after Game nine (7-2)

1995 (Classical) - Kasparov v Anand (+2 =9 -1)
Going into Game twelve: +2 =8 -1

1996 (FIDE) - Karpov v Kamsky (+5 =5 -2)

Hypothetical best-of-twelve would have ended after Game nine (6.5-2.5)

2000 (Classical) - Kramnik v Kasparov (+2 =10 -0)

Hypothetical best-of-twelve would have ended after Game 11 (6.5-4.5)

2004 (Classical) - Kramnik v Leko (+1 =9 -2)

Note: Another example where a best-of-twelve format would have produced a different World Champion

Peter Leko: A would-have-been World Champion

2006 - Kramnik v Topalov (+3 =6 -3)

Was a best-of-twelve (plus tiebreaks)

2008 - Anand v Kramnik (+3 =7 -1)

Was a best-of-twelve

2010 - Anand v Topalov (+3 =7 -2)

Was a best-of-twelve

2012 - Anand v Gelfand (+1 =10 -1)

Was a best-of-twelve

Of the 43 World Championship matches held over the course of chess history, a total of four may have seen different outcomes under a best-of-twelve format. In 1892, Chigorin would have been World Champion after a +5 =4 -3 result, but under the strain of a prolonged schedule, his lead eventually evaporated. We will avoid alternate history speculations such as: Had Chigorin been World Champion, would Steinitz have been his subsequent challenger and, if so, would Steinitz have regained the title? In 1935, Euwe would have been denied his world title, since Alekhine edged him by a point at the twelve-game mark. For his part, Petrosian would have kept Spassky's name off the hallowed register in 1969, as the result of a deadlocked match, while Kasparov, in turn, would have suffered to wait until the next world championship cycle to materialize his claim. Leko, in 2004, serves as the most recent would-have-been, and the extra games gave Kramnik his chance to rally and retain his crown.

Two World Champions that never were and two that might have been. In stark contrast to reality, to be sure, but in the face of the larger pool of matches, perhaps a worthy footnote to history. By and large, despite strong feelings expressed against it, the best-of-twelve format seems a fair and reliable means of determining a worthy champion. Discarding the matches that were shorter than best-of-twelve, we are still left with an overwhelming number that would have produced the same decision that ultimately came to pass. In a world of instant gratification, this may be good news for the prospects of the truncated schedule.