Author: Uri Blass
Date: 12:18:09 07/22/04
Go up one level in this thread
On July 22, 2004 at 14:10:33, Uri Blass wrote: >On July 22, 2004 at 13:56:43, Rémi Coulom wrote: > >>On July 22, 2004 at 13:46:09, Uri Blass wrote: >> >>>I may consider crazybishop later and another experiment is to use different time >>>control of x minutes/40 moves. >> >>I expect The Crazy Bishop to be stronger at short time control than at slow time >>control. Its search is fast, but the evaluation is really very poor. I am >>looking forward to the result of your experiment. > >note that poor evaluation is only one possible reason to do relatively worse at >long time control and even if the evaluation is the same it is possible that >program A will do better at long time control when B will do better at blitz >(for example if A has better order of moves but B has some tricks to be faster >in nodes per second). > >Uri Some questions: 1)How much hash should I give crazybishop for these matches. I plan to give it 256 mbytes like I give Jonny. I plan to play later matches of 1,2,3,4,10,20 minutes per game when movei get the same conditions. 2)Am I correct to assume that the positive score of crazybishop against Jonny at blitz was inspite of inferior hardware(I know that Jonny has superior hardware when crazybishop used the hardware of the tournament but if I remember correctly Jonny started with the hardware of the tournament so I am not sure if Jonny used superior hardware relative to crazybishop). Result so far against Jonny in 20 minutes per game is 4.5-3.5 for Jonny and it will probably be 4.5-4.5 because movei is 2 pawns up in the endgame Uri
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.