Author: Ulrich Tuerke
Date: 14:26:06 05/11/01
Go up one level in this thread
> >PS : Do _you_ know why Comet does so well in this test ( and want to tell ? ) ? I think it's just by chance. There are many suites, where Comet does'nt do very well (e.g. lct2). So there should also be a few where it does well because the program isn't that bad. In order to reliably determine the playing strength of a program with the help of positions, I believe that a very large set of positions (several hundreds) is needed. It should be a mixture of tactical, positional and endgame positions like lct2 but far larger. On the other hand, the current well known suites are either too small or aren't appropriate mixtures. Most of them are purely tactical. I'm afraid, that it's very difficult to compose positions with a unique best move for only positional reasons. These suites measure only some special fraction of the program's playing abilities. Kaufmann belongs to these. By chance, kaufmann hits a fraction which is perhaps well done in Comet. But don't ask me what this fraction is. Uli
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.