Author: Thomas Mayer
Date: 18:30:47 09/23/01
Go up one level in this thread
Hi Vincent, >With all respect this testset is way too simple for nowadays >chessprograms! I think the goal behind that testset is to have a test that you can use to test if you have made tactical progress with your engine... and you can test it fast and it didn't take several days to get any results. >>I tried the suite with Bringer up to 1 minute/move. >> >>1 second/move 76/184 >>5 seconds/move 109/184 >>10 seconds/move 124/184 >>30 seconds/move 143/184 >>1 minute/move 156/184 look at this, that is very nice, with more time there is a real difference between the amount of solved positions... At WAC we have something like: 1 second - 265/300 5 second - 289/300 10 second - 296/300 1 minute 298/300 1000 hours 299/300 that does not help any programmer - he can not see if his changes have helped at anything, there are to many positions way to easy... By the way: indicate your results of Diep ?! (With hardware please... :) Greets, Thomas
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.