Author: Thorsten Czub
Date: 02:32:56 12/13/05
Go up one level in this thread
On December 13, 2005 at 04:31:27, Joachim Rang wrote: >So far I was never able to tune Fruit against a specific opponent which resulted >in a signficant difference not to say in doubling the score (40% to 80%). So >either I simply lack the understanding how to do this or you played not enough >games to get a statistical significant result and observed just noise. Given the >little time between your first post about Rybka and this one I give both >possibilities the same probability. ;-) tuning styles is not easy joachim. and it works not with ANY program. IMO it works less good with search based programs. whenever you bring new evaluation into a program you need to tune it in relation to the other evaluations. and tuning style needs lots of time because you have to proof the idea in games versus all kind of opponent on all kind of hardware/levels. this takes. but it works. I have found that there are programs i can do it, and programs i can't do it. maybe because i cannot FEEL for these programs.
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.