Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Testing Rybka 1.01 Beta 11 at 40/20+40/20+10min finish.

Author: James T. Walker

Date: 07:17:29 01/22/06

Go up one level in this thread


This looks like a clear case of personal Bias to me. :-)  You didn't like the
result with Fritz so you did a repeat?  You liked the other results so it was ok
to let it stand.  To avoid biasing your results maybe you should pick an exact
number of games to play vs each and stick to it.  The final overall result is
what counts.  In short matches anything can happen.  I've seen Rybka get beat by
Fritz 8 +8 -0 =2 for 9-1 score.  But after 100 games Rybka won 67% of the games.
Regards,
Jim



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.