Author: Mark Young
Date: 10:10:16 06/20/99
Go up one level in this thread
On June 20, 1999 at 12:47:29, Robert Hyatt wrote: >On June 20, 1999 at 09:37:19, Michael de la Maza wrote: > >>The machines are up 1.5 - 0.5 and in the remaining three games, the computers >>all have, at minimum, decent positions. There is a legitimate possibility that >>the GMs will not win a single game. >> >>I think that the Ferret/Vaganian game is the most interesting from the computer >>chess point of view. Against a human, Vaganian would have had a "crushing >>kingside attack." However, Ferret is not a human and so it defended with elan >>and precision and is now two pawns up. >> >>The day will soon come when GMs are forced to concede that they can no longer >>launch successful mating attacks against computers when there are a substantial >>number of pieces on the board. > > >Boy do I disagree. You just haven't seen GMs do this very much yet. But of >all the weaknesses computers currently have, king safety is at the _top_ of >the list, not the bottom. Computers will always have decent chances, but >attacks can and do work all the time. IE drop over to ICC some time with your >favorite program and ask a GM to play you a few games to see what I mean. Or >even a few IMs that I can think of over there that are _deadly_ with slow- >developing kingside attacks that have to be parried almost before they start.. Still again, a 2.5 (GM humans 2600+) vs 2.5 (computers) point result is not suggestive of chess programs being only at a 2400 rating. P.S. The games were not played at G/30 or G/60, nor were they "out of date" weak GM's.
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.