Author: Uri Blass
Date: 16:54:08 09/07/01
Go up one level in this thread
On September 07, 2001 at 17:40:41, Christopher R. Dorr wrote: >On September 07, 2001 at 16:26:01, Uri Blass wrote: >> >>Deep Fritz cannot do it because of the unfair conditions. > >Again, if the CB team thought the conditions were unfair, they were under no >obligation to accept them. I guess that chessbase prefers to earn money and they believe that they can get a lot of money by losing against kramnik. The match is unfair for people who are interested in chess. > >> >>Suppose Deep fritz can do it. >>Kermnik is going to discover it after having the program and learn to avoid the >>weaknesses. >> >>If kramnik can take advantage of the program's weaknesses the program has no way >>to learn it from games against kramnik before the match in order to fix the >>problem. > >The if it has weaknesses that are that easily exploitable, then it deserves to >lose doesn't it? You think that if Fritz demonstrated some astonishing defect in >Kramnik's play, he would be able to change a fundamental part of how he plays >chess in the course of a match? Not a chance. he would likely change his >openings to avoid being exposed to it, just as the CB team can change theirs. Chessbase cannot change their book based on the private games of kramnik against Deep Fritz because chessbase does not know about the preperation of kramnik against them. It >seems quite fair to me in every way. We do not agree here. One of the qualities of good players is ability to learn from their mistakes and not to repeat the mistakes again. If kramnik is losing in part of his private games against Fritz then he is going to learn to be a better player and not only about the opening book. Deep Fritz is not going to learn from this private match even if it has perfect learning function that is not only about learning to avoid losing lines but also about learning to change the evaluation function after analyzing games The reason is simple Deep Fritz does not get the games of kramnik against it. Deep Fritz probably cannot learn to change it's evaluation. I can understand if you say that this is a weakness of the program and the program should suffer from it but suffering from the fact that it even cannot learn to change it's book based on private games that kramnik played against it is simply unfair. Uri
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.