Monday, January 4, 2010


Once again there is another article on the ChessBase website to advance the hypothesis that Garry Kasparov was the greatest chess player of all time. Maybe he is the "greatest chess player of all time who had the help of seconds and banks of computer engines and databases from ChessBase." Despite all that assistance his Elo, at its highest, was only 2851! Garry did this in 1999.

Fischer on the other hand had an Elo of 2785 after his match with Spassky in 1972, a difference of 66 points and 27 years. In the interim many have suggested there has been considerable rating inflation. Today there are 30 people (at their maximum rating) who are rated at or over 2731, a difference of 120 points between Kasparov's highest and 2731. Fischer was rated at 2785 which was 120 points (or so) higher than his nearest competitor, Spassky! And Fischer lost a few rating points for not beating Spassky even worse! (Kasparov's nearest max. rating competitor, Topalov, was 2813, a difference of 38 points.) Elo's system is really about comparative "relative" performances.

What's more, Fischer's disdain for seconds, and there being no computers available then, seriously puts into question that Kasparov was the best player of all time--I think objectively, Kasparov knows this and why he writes so deferentially about him.

The next thing we know someone will write that Kasparov was stronger than Paul Morphy, whom many had previously regarded as close to "unbelievable." Why this unprovable emphasis on who is/was the best? I think there is a better question: If you were a grandmaster and your family's life was on the line, who would you fear more as an opponent, for one game, an alive Fischer (1972) without a computer or seconds, or Kasparov (1999) without a computer or seconds?


  1. I fail to see why people are putting so much effort into trying to answer the "Best of the Best" question. Please note that as a professional statistician, I am totally addicted to "best of list" etc.

    They are all amazing chess players! That is good enough for me.

    Bob, Thank you very much for interesting blog posts!

  2. The worst part of the ChessBase story and of the current rating list is very obvious when one looks at the kind of players currently rated 2700 and up, i.e. above all of the past World Champions, as rated by Arpad Elo, except for Lasker Botvinnik and Fischer!

  3. Just being better than your contemporaries doesn't make you the greatest ever. Maybe all of your contemporaries stunk!

    Nowadays, Chess is played more more people, in more countries, who have access to the best computers and theory. It is entirely possibly and even likely that todays crop of 2700+ GMs would have crushed Fisher, Lasker, Botvinnik etc.

  4. Actually a far better experiment would be to *objectively* compare the games of Fisher against the games of modern GMs like Anand, Kramnik, Carlsen etc. See who actually played the best moves (sound familiar to Fisher fans ? :)).

    I think you would find the quality of todays games are much much better than in Fisher's day (i.e. today's players play better moves than Fisher).

    Here's an experiment, even if you are not a GM, you can pull out your favorite chess engine and replay a game of Fisher and then replay a game of Anand, and see who made the most mistakes. You would find that easily Fisher played far more bad moves than Anand.

    One thing Fisher was right about is that chess is about playing good moves, now we can objectively judge who played the best moves, and see that Fisher was not as good as todays GMs. That's what is important, not some subjective *feeling* that Fisher was the greatest.

    1. Today's GM's will surely be "better" than fischer if you use that method. first of all, today's GM's are trained by computers in the first place. so they know the best movies according to the computers and will therefore their moves will be "rated" higher by computers. This method is incredibly partial to today's GMs. On the other hand, Fischer's moves were simply gut feel, and pure creativity, and not jsut a chess program telling you what to move.