Author: Terry Presgrove
Date: 18:52:38 06/01/99
Go up one level in this thread
On June 01, 1999 at 17:05:25, Hristo wrote: >Has anyone compared the two versions of crafty(15_20 and 16_8)? >So far I had them play 6 blitz games and 1 one_hour game(the second one >should be done when I get home). >The score: >Blitz- 15_20 +3 =2 -1 > 16_8 +1 =2 -3 >1 hour-15_20 +1 =0 -0 > -16_8 +0 =0 -1 > > >While the results of a few games are not very important the style >of play was in favor of 15_20. It seemed that 16_8 had some problems >entering endgame with Bishop and pawns against Bishop\Knight with pawns. >FYI I'm interested only in the strength of the engines, so there is no >opening-book or endgame database used. To be fair, I set a position >from which the programs continue(similar to the Nun test). Otherwise >the opening lines are rather funny .. :))) >I'm asking this to know if I should keep track of the score. > >Computer: > PII450, 128M Ram, default engine settings. > Windows NT. > > >Hristo I have run a number of tests with v15.20 vs. 16.6 the results being mixed. It seems 15.20 played better at 3 0 through 15 0 and 16.6 breaking even around 30 0 and playing better at 3 hour matches. I have been running a number of matches between 16.6 amd 16.8 and it also appears 16.8 plays better at faster blitz but I'm getting a different result from 30 0 and up 16.6 winning 14-7 so far in 30 0 matches. And 3.5 out of 4 at 3hour+ matches. This on k6-350 . I recognize this small sample is not sufficient to draw a reliable conclusion but these early results seem to indicate 16.6 may still be the strongest version at longer time controls on a single cpu machine. TP
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.