Author: Dan Ellwein
Date: 05:02:33 04/16/02
Been trying an experiment of late... Took the default settings of CM8000 and lowered the point value of Q,R,B,N by approx 15%... This kept the ratio the same between Q,R,B,N and at the same time brought the piece values closer to that of the P... Wanted to see if by lowering the value of the pieces closer in line to the P if the pieces as a whole would co-ordinate better with each other... I also increased the default setting of Selective Search from 6 to 12... The opening book used for CMLite was the Depth4 book... I used this setting because I wanted to see how CMLite would play on its own being out of book after the first 2 moves... The name of this new setting I call - CMLite (sort of like Bud Lite except without all those calories)... This was played on a 400Mhz Celeron, 64M Memory, Windows 98 2nd Ed. computer... At a time control of 5/3 there was no appreciable difference... The score was 6 wins for CM8(default) and 5 wins for CMLite... 9 draws... At a time control of 10/3 I started to notice some difference... The score was 4 wins for CM8(default) and 8 wins for CMLite... 8 draws... The next time control will be played at 40/40... This will take awhile to play 20 games... I'm curious to see if the score will be even more lopsided in favor of CMLite or if CM8(default) will make a come back... CMLite Opening Book: Depth4 Selective Search: 12 Own Q 7.7 Opp Q 7.7 Own R 4.3 Opp R 4.3 Own B 2.6 Opp B 2.6 Own N 2.6 Opp N 2.6 Own P 1.0 Opp P 1.0 Dan
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.