Author: martin fierz
Date: 11:00:51 11/08/02
Go up one level in this thread
On November 08, 2002 at 02:13:25, Dann Corbit wrote: >On November 08, 2002 at 01:38:30, martin fierz wrote: > >>On November 07, 2002 at 20:43:32, Dann Corbit wrote: >> >>>On November 07, 2002 at 19:48:46, Russell Reagan wrote: >>> >>>>So you're kind of building a giant opening book? >>> >>>Yes and no. It's just a pile of data. It could be used as an opening book, but >>>it would be unlikely that such an approach would be successful. >> >>i would be surprised if such an approach were not successful! my checkers >>opening book generator analyses opening positions and somehow decides which >>position to analyse next, and this approach works extremly well for checkers. of >>course, chess is much more complicated, but if you get 1000s of computers >>working for you, and a powerful engine to search with, then this approach will >>work for chess too. > >See what computer programs think about the Evans Gambit. The evaluation is >often off by about 2 pawns. of course. the same is true in checkers. >On the other hand, if you minimax the same data, it will turn out the bugs. and of course again. that is what my book generator does. if you expand your tree deep enough, you will eventually "see the light". aloha martin
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.