Author: Enrique Irazoqui
Date: 05:18:32 10/29/00
Go up one level in this thread
On October 28, 2000 at 17:43:44, Jorge Pichard wrote: >Mr Enrique you have analyzed most of Kasparov vs Kramnik games using the >latest betas of Gambit Tiger 1.0 Deep Fritz and others... What is your >perception of how close these Betas, or should I asked which beta so far has >predicted more accurately the next moves of the best two players in the World. I didn't keep track of which program predicted most moves, but globally I have the impression that Deep Fritz, Century 3 and Gambit did an excellent job, not only at predicting but at suggesting moves pertinent to the positions. The PVs of Gandalf posted by Mogens were quite impressive too. I guess I must revise my idea about the general incompetence of programs in positional play. I also have the impression that these programs would have done better than Kasparov and Kramnik in yesterday's game and in those two games in which Kramnik failed to win. I am not implying that they play better than Kasparov and Kramnik, of course not, but I don't have it clear that in the cases of divergence the moves they picked were necessarily inferior. It was also interesting to see how programs evaluated Kramnik's attack yesterday. The King, Deep Junior and Gambit "believed" in it, while Deep Fritz and Gandalf, in agreement with Shipov, did not. >Now my next question even if you don't have the beta version of Shredder 5 I had it, but not any longer. >what chances do you really give to Shredder 5 in scoring a least 33% of the >games against Kasparov. This is anybody's guess. I find important in this regard the recent match Deep Fritz - van der Wiel, in which a great anti-computer expert was unable to make the program look like an idiot. DF won 3-1. Enrique > If Kasparov accept the Challenge after the match with >Kramnik is over, even if he is detroned from the Title, it seems to me that >current P.C. chess programs are not too far from selecting the same moves that >super GM can produce given normal standard time control. > >Pichard.
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.