Author: blass uri
Date: 23:07:09 05/01/00
Go up one level in this thread
On May 01, 2000 at 18:46:20, Robert Hyatt wrote: >On May 01, 2000 at 14:38:22, Alexander Kure wrote: > >>On April 30, 2000 at 20:02:03, Robert Hyatt wrote: >> >>>On April 30, 2000 at 19:32:41, Chris Taylor wrote: >>> >>>>Here, at a longer time control, and on two comps! Crafty does, OK.....? >>>> >>>>Or, should I say, "Is doing, not too badly"(!!!!!!) >>>> >>>>Looking forward to the next version! >>>> >>>>Chris Taylor........ >>> >>> >>>14.5-18.5 is not horrible. Take the two book losses, turn them around and >>>the match is dead even. Or make them draws and it is 15.5-17.5, which is >>>nearly equal. >>> >>>I'd rather see 20-0 of course, but that isn't realistic, unless a big SMP box >>>is used for crafty. >> >>Hi Bob, >> >>Actually, there has been only _one_ book loss so far in this match (game 17), >>where unfortunately Crafty was on the loosing side. >>As both programs play with the same book (general.ctg from Fritz 6) as i found >>out now after a little research, chances of so called book losses are equally >>shared between them ;-) >> >>Greetings >>Alex > > >I was going by Tony's comments. He said -1.4 in one book line, -2.5 in another >one. I didn't look very carefully at either. And I don't worry about them as >I see them with my book just as often, although I don't know whether it learns >to not play them again, when running in the Fritz GUI. IE book learning most >probably is not working... I think that book learning is working for all the chessbase engines in the same way(Fritz's engines try to repeat opening that they won and avoid opening that they lost). 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.