Author: Otello Gnaramori
Date: 02:34:06 07/21/01
Go up one level in this thread
On July 20, 2001 at 17:56:01, Miguel A. Ballicora wrote: >As an example, a young Tal said that he was going crazy to understand one >position by analysis. Botvinnik said to him something like, "you have to >exchange queens and you win". This is the kind of annotations that a strong GM >is capable to give and a computer is still weak (non-existent) in that >department. I think that in that case the line itself from the computer analisys (queen exchange) combined with the value (very high) gives the same annotation of Botvinnik. There is an interesting anecdote by Sam Sloan on Botvinnik and computer chess : <snip> Botvinnik "retired" to devote himself to computer chess. Botvinnik wrote extensively on this subject. He also developed computer chess playing programs. The first world computer chess championship was won by a program developed by a team headed by Botvinnik. Although I have personally met and spoken to almost all of the world's top chess grandmasters (except for the newest, youngest ones) I never met Botvinnik. I missed one chance when he attended a computer chess conference in Montreal, which I decided not to attend. Unfortunately, after that, Botvinnik no longer entered programs into the world computer chess championship. The reason was obvious. Botvinnik was working with old style computers using vacuum tubes, which had limited speed, computing power and memory. Integrated circuits were not available in the Soviet Union. It is a wonder that the computers which the Soviets had were even able to play chess at all, much less compete for the world championship. Botvinnik was not going to enter a computer program into a competition unless he felt that his program had a good chance to win. Botvinnik nevertheless wrote about computer chess. Every serious chess computer programmer in the world has studied his works. Hans Berliner recently characterized Botvinnik as a "fraud" in the fiend of computer chess. I disagree. I do agree, however, that what Botvinnik wrote has little value to chess computer programmers today. Botvinnik wrote about the "Minimax" theory, which is the cornerstone of every strong computer chess program in existence. He used as an example to explain this theory his famous win over Capablanca. However, I feel certain that Botvinnik did not invent the minimax theory. He merely explained it to the layman. Botvinnik was working with "selective search" computer programs. We now know that such programs do not work. All strong chess playing programs today use "full breath search", which means that the computer searches every possible legal move and every possible legal position for several ply deep, before even starting to evaluate the resulting positions. However, with the equipment which Botvinnik had, it was not possible to run a full breath search program. Today, every computer chess program analyzes millions of positions per second. The machines Botvinnik had could only analyze a few positions per minute. Therefore, Botvinnik had to try to develop general principles which his computer could understand and follow. Even though his programs were weak by modern standards, they still produced a number of remarkable and interesting games. Having never met Botvinnik, I do not know what kind of man he really was. However, my perception based open what I have read and have been told about him is that he was extremely serious, that he never told a joke, that he had few personal friends and that he must have spent his days sitting around his apartment thinking pure thought. He was indeed a great thinker. His writings show that. At chess, he was a profound strategist. One story which Botvinnik wrote has always stuck in my mind. It was how, as a young man, he had defeated Grandmaster Spielmann in just 12 moves. The way this happened was that Botvinnik published an article about the Gruenfeld Defense. He included analysis about a new move that he had discovered. However, his published analysis was not complete. Instead, he deliberately left out one move from his analysis. This was a promising looking move for Black which he did not discuss. Botvinnik had seen that move and had found a refutation. However, in his article, Botvinnik had discussed neither the move or its refutation. Later, Botvinnik played Spielmann. Spielmann, of course, had seen the article. Spielmann had also noticed the move for Black which Botvinnik had failed to mention. Needless to say, Spielmann thought that this was an oversight on the part of Botvinnik. Spielmann played the move, and Botvinnik immediately played the refutation. Grandmaster Spielmann was forced to resign on the very next move, which was move number 12. Ever since then, generations of chess grandmasters have followed this fine example established by Botvinnik. This is the reason that one can never completely trust published grandmaster opening analysis. Grandmasters are paid to write chess articles, but they are also paid to win chess games. Sometimes, these two objectives conflict. I personally see nothing ethically wrong with this practice. Grandmasters devote their entire lives to looking for and finding these little traps and tricks. When some chess editor hires them to write an article, they are under no moral or ethical obligation to spill their guts and reveal everything they know in detail about chess. Mr. Crowther, at the end of his article, left a suggestion hanging that Botvinnik might have "saved" Keres. This is something that I would like to learn more about. This concerned a difficult period in history. During the Second World War, Petrov, a grandmaster strength player from Latvia, was executed by the Soviets. Keres, as Estonian player only slightly stronger than Petrov, could have been executed as well. It must be explained that at this time, the official grandmaster title did not exist. The title of "Grandmaster" had been conferred by the Czar of Russia upon the five finalists at the great tournament of St. Petersberg, 1914. By this criteria, Frank Marshall of the U.S.A. was a grandmaster, but Rubinstein of Poland, who did not make the finals but was probably stronger, was not. These titles were not conferred by any official body, however. Then, in about 1950, FIDE conferred the title of "grandmaster" on about 80 players. This number has increased every year since. Therefore, strictly speaking, both Petrov and Spielmann were never grandmasters, because they died too soon. By modern day standards, they were clearly grandmasters, however. <snip> Regards.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.