Author: Shaun Brewer
Date: 15:08:52 03/05/06
Go up one level in this thread
Marc, I have been using GF and like it a lot - it seems to play quite differently from the Toga & Fruit versions I have tested. From vaious rating lists it seems slightly weeker than TogaII 1.1a - however in test suites I have been investigating it scores better. If someone was to make any tweaks I would be happy to test for them - using both test positions and games I could run a gauntlet at blitz and slower time controls. I already have tournaments in play at blitz including GF (My tournaments are based on 16 early start positions 32 games per engine pair) - the idea being it should be easier to attribute score improvemnts to the engine rather than book improvements/luck. However I have noticed that even running tornaments like this a re-run does not repeat identically as sometimes an engine gets fractionally more/less CPU and a different path is chosen. I have stopped the obvious background task and the machine is dedicated for Chess for the duration however this does not seem to be enough to be able to repeat consistently. Shaun
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.