Author: Uri Blass
Date: 10:21:31 07/03/03
Go up one level in this thread
On July 03, 2003 at 12:47:16, Christophe Theron wrote: <snipped> >As for the progress between version 14 and version 15, all depends on the amount >of work you invest in testing. While it is easy to get the impression that an >already strong chess program has not made progress, Chess Tiger 15 is definitely >clearly superior to version 14. I have measured this improvement and it is above >what the SSDF has found out. The question is if you measured the same thing that the ssdf did. I know about cases when a program that did better in blitz did worse at longer time control because of a bug that caused it to play stupid mistakes only at long time control(this happened with a version of Terra). I know from experience with movei that a program can play and play well inspite of bugs of calling something that is undefined like A[-1]. It is possible that the bug does not cause stupid mistakes but cause slightly inferior moves so it is hard to find it's influence or even the fact that it exists. I simply see no way to measure the same thing that the ssdf did unless you do the same as the ssdf and I believe that the games that you got was at faster time control. You may have a method to get an estimate for the improvement but the only way to test improvement is by games. 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.