Author: Uri Blass
Date: 03:28:27 10/16/00
Go up one level in this thread
On October 16, 2000 at 04:02:45, Thorsten Czub wrote: >On October 15, 2000 at 10:49:41, Enrique Irazoqui wrote: >>Gambit "risks something" against others too, and wins. Not with Tiger. > >?!? > >tiger risks less. >gambit-tiger risks more. >this works against all except tiger itself, since they have to many >similarities. > >>Careful: you are going down the path of bean-counting... :) > >and ? i thought this is the future :-)) > > >>I don't have a problem. Gambit does. With Tiger. > >pah - no problem. gambit-tiger will not have to play anywhere against >tiger itself. thats not happening. > >>No, no. Read again. Gambit is very succesful against programs with weak king >>safety code, but much less against the others > >no - it is as good as tiger. > > >> and loses to its brother Tiger. > >no problem, as i told you gambit-tiger will not have to play >anywhere against itself. I guess that it is going to play also against itself when the ssdf test both tigers. I do not think that losing against the default tiger is not a peoblem. It is a problem because it means that other programs also may win gambittiger if they are similiar to the default tiger. Christophe knows that gambittiger is not perfect and he explained that he is going to improve it to in order to continue to play the good sacrifices and avoid the bad sacrifices. 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.