Author: Les Fernandez
Date: 13:13:59 01/03/01
Go up one level in this thread
I am in the process of running 2 identical crafty engines (crafty1 and crafty2 both with no egtb's) using winboard playing against one another. The tests are based on a Pentium II 350 machine on a Windows 98 platform for those that are interested. The experiment is set up to play 256 games each at 1,2,4 and 8 minutes where crafty1 is white for game 1 (odd number games)and crafty2 is black (even number games) till 256 games have been played. For each new time setting the book.bin file and books.bin file get reinitialized to promote better continuity in the results and in addition the (I think) learn.bin file gets deleted to eliminate any learning variations. After the entire 4 sets are completed I now set egtb=3 for just crafty1 and run the exact same test as above to look for any effect a 3 piece egtb has on the results when compared to the first set of tests which were based on egtb=0. I have completed egtb=0, egtb=3, egtb=4 and am currently working on egtb=5. I can not test egtb=6 since not only dont I have the space for the egtb=6 but the "entire 6 piece egtb is far from being completed. Anyone interested in these results when they are finished can email me at vb4@prodigy.net. I dont think I should post them here since the pgn files are all about 256k in size. I have also generated some general stats on the pgn's and if anyone has any ideas as to what other types of info I should take a look at please let me know. May this year be better then last year! Les Fernandez
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.