Author: Uri Blass
Date: 00:59:37 04/25/01
Go up one level in this thread
On April 24, 2001 at 15:56:22, Thorsten Czub wrote: >first you should ask which evidence johan has that cm8 is stronger than e.g. >cm6. > >i am sure if you take the cm8 ssdf games, and replay the same openings against >the same opponent, but instead of cm8 use cm6, the results get better. >try out yourself. >same games, same opponent, but different cm-version and you will see >which program is better. Even if cm6 does better result then it does not prove that cm6 is better. Deep Fritz repeated some wins against Chessmaster8000. The choice of the opening of Deep Fritz was not a random choice and Deep Fritz chose openings that Chessmaster8000 does not know to play because of learning. It is possible that Chessmaster6000 can play the same openings that were played against Deep fritz better but plays other openings worse and the fact that they were not played is Deep Fritz's learning function. In other words the test of using the same opening is biased against Chessmaster8000. 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.