Author: Bob Durrett
Date: 08:48:30 11/21/03
Go up one level in this thread
On November 21, 2003 at 03:26:41, Mig Greengard wrote: >Had a long, fascinating session with Garry on Wednesday. He went over all the >games and talked deep and wide about computer chess and his matches. It was all >video-recorded for future display as a Playchess.com multimedia segment. Some of >it is recapped in the article I just put up here: > >http://www.chessbase.com/newsdetail.asp?newsid=1313 > >More will go up in text later. From a chess perspective my suspicions during the >game were confirmed when he said that he had analyzed just about all of game >four a few years ago when looking at that QGA line. He mostly spent his time >double-checking things to see if the Fritz team had found something nasty >somewhere. That one of his former analysts, Kharlov, had played that game up to >move 18 in 2001 was a tip-off. The position was already equal by that point. > >Probably of more interest to you all is Kasparov's interest in computer-computer >games beyond simply preparing for a specific opponent. That part I recap in the >article linked to above. I'm seeing him again before he leaves Sunday so if you >have an interesting, (non-flame) question for him on comp chess I'll be happy to >put it to him (and take credit for it myself of course). My biggest concern is that the beginning of the game was NOT a match between Kasparov and a chess engine. While Fritz is "in book," Kasparov is, in effect, playing against human opponents who have had infinite time to create a formidable anti-Kasparov opening book. The part that should interest chess computer people is whether or not it is possible to put together an anti-Fritz book which would put Fritz into early middlegame positions which would be disadvantageous to Fritz. Maybe the sponsors should fund a team of anti-Fritz book makers to work with Kasparov many months in advance of the games to prepare this anti-Fritz book. Then the first part of each game would be a competition between the anti-Kasparov book-making team and the anti-Kasparov book-making team. Unfortunately, the public might not understand. It seems unfair that Kasparov should have to rely on his memory of a vastly huge opening book. He should have access to his notes [on a computer] during that phase of the game when Fritz is still "in book." Maybe then he would play 1.e4. The objective of the competition in the "in-book" phase should be to see if it is possible to force the game into early middlegame positions unfavorable to the opponent. After Fritz is out of it's opening book, Kasparov should be required to close his book too and then the real competition could begin. I'm not sure what to do about endgame tablebases, but the situation seems similar to me. Bob D.
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.