Author: Keith Kitson
Date: 12:45:25 09/08/99
There was some mention recently of SSDF results and their relevance to the real world. I reckon the work they do is very valuable but it cannot possibly be the de-facto standard for ratings of chess computer programs as they only use one time control. So programs get better results at faster time controls, and others at slower time controls. IMHO the only true way to get a true reflection of a chess game's rating is to play against humans at a variety of time controls. However failing that a substantial number of games against recognised chess computer competition at varying time controls would give a more accurate comparison. I do find that of the top programs some do manage sudden death time controls better than others, surely this must be taken into account when assessing the overall strength of the program? After all, every move calculated by a chess computer in a match is a compromise in time or we would never finish a game....:-)) Keith
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.