Author: Peter Berger
Date: 13:58:05 09/19/03
Go up one level in this thread
On September 19, 2003 at 14:58:53, Vincent Diepeveen wrote: >On September 19, 2003 at 11:35:32, Peter Berger wrote: > >>On September 19, 2003 at 11:14:29, Vincent Diepeveen wrote: >> >>>I'm looking for positions to search very deep with a cpu or 500 which are >>>interesting to search deep. >> >>Why not try the Nolot positions? http://www.seanet.com/~brucemo/nolot/nolot.htm >>has a survey with results, including the ones of Deep Thought, that still look >>very impressive. > >most of them too easy and they are just tactical. All you need is a beancounter >for most of the pos. Of course, but even with an impressive eval you should be able to solve some of them on 500 CPUs I suppose. There would be the great additional advantage that there are other results to compair with. And positions that take Ferret 93 hours are certainly not trivial. Another idea would be to use the CSS "Weltmeister-Test" hosted at http://computerschach.de/test/index.htm. The positions seem to be mostly correct but they are too easy for hours of search. But a test run at 20min/move should be possible - and it would allow to compair results with other programs. The problem with the terribly difficult non-tactical ones is that proving they are even correct is near impossible. [D]6k1/7p/4r3/R7/3P1bb1/2P3N1/1P3K1P/8 b - - 0 1 I believe that in this position ..Kf7 and only ..Kf7 wins. Analyzing it I can prove +6 with Yace after many hours. But to prove that it is the solution and the only one is near impossible - and a program might very well choose it for the wrong reasons anyway. But I _do_ believe it is correct :). Peter Peter
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.