Author: Charles Roberson
Date: 10:16:09 07/09/03
Go up one level in this thread
Years ago I noticed that using tests of a new version on Noonian vs and older version to indicate strength improvment was prone to inaccuracy. Much better has been to test against multiple different programs. Guy Haworth published a paper that suggests accurate measures from self-play requires more than 1000 games.
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.