Author: blass uri
Date: 21:57:14 04/20/00
Go up one level in this thread
On April 20, 2000 at 09:23:51, Robert Hyatt wrote: >On April 20, 2000 at 03:52:10, blass uri wrote: > >>On April 19, 2000 at 22:52:01, Robert Hyatt wrote: >> >>>On April 19, 2000 at 18:40:03, Mogens Larsen wrote: >>> >>>>On April 19, 2000 at 18:12:34, Robert Hyatt wrote: >>>> >>>>>Looked hard to me... I couldn't find anything useful... :) >>>> >>>>I couldn't find Crafty either, but I think I found a clone :o). >>>> >>>>Carfty 17.07 - Junior 6.0: 2 - 10 >>>> >>>>Best wishes... >>>>Mogens >>> >>> >>>I finally found it... goes to show how intransitive chess programs are. >>> >>>crafty got pounded by junior, crafty pounded rebel 8 on a mmx, then rebel8 >>>and junior were pretty close... >>> >>>which is why it never surprised me when a beats b, b beats c and c beats a. >>> >>>:) >> >>I have the games Crafty(k6-450)-Rebel8(p200) >> >>One of the reason for the good result of crafty is that crafty with white >>repeated the same winning opening against Rebel8 12 times and repeated the same >>winning opening with black 6 times. >> >>Uri > > >that isn't book learning... it is more likely a narrow book in rebel 8. It >requires a special crafty.rc file to make book learning behave like that, re- >playing won games immediately, and never replaying a lost line..... > >I didn't send anything special to the SSDF testers to change this, and since >I don't know anything about the fritz gui and opening book, this seems to be >a problem that rebel 8 was falling into, rather than a problem that crafty was >causing... You did not do it but chessbase did it and crafty tries to repeat the same opening if it wins like other chessbase programs. Crafty did not repeat exactly the same games because it probably used learning by position but the opening was the same. 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.