Author: Komputer Korner
Date: 16:04:26 12/31/97
Go up one level in this thread
I have the solution for one of your questions at least. At the start position with opening books off, look at the score evaluation after a good night's think. If the evaluation isn't in the range of +0.12 to +0.20, your evaluation is either skewed asymmetrically or else the evaluation function is out of whack. This is based on the fact that White scores 56 % out of the total mass of games at the higher levels. The 56 % is equivalent to a score of about 1/2 a tempo or 0.16 pawns. If your program doesn't attain this score ( adjusting for variance is why I gave the range from 0.12 to 0.20) then there is indeed something wrong. On December 30, 1997 at 07:13:39, Chris Whittington wrote: snipped >This doesn't solve the problem of how to test for relatively neutral >positions. Like the start position, e4, d4, nf3 and various other moves >are perfectly acceptable, and the choice depends on a bunch of factors, >importantly what type of chess game you like to play. > >How are you going to test for this or any other position where there's a >'steering' choice ? > >Chris Whittington
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.