Author: Dann Corbit
Date: 10:22:02 01/17/01
Go up one level in this thread
On January 17, 2001 at 11:09:27, David Dahlem wrote: > >On January 17, 2001 at 06:25:04, Uri Blass wrote: > >>On January 17, 2001 at 05:28:37, allan johnson wrote: >> >>>On January 16, 2001 at 20:09:53, Dann Corbit wrote: >>> >>>>Here are the remarks that describe the actual positions: >>>>http://home.interact.se/~w100107/pos.htm >>>> >>>>It also gives the names of the annotators. >>> >>>Dann Any suggestions as to why Century3 did so badly [relatively] in >>>the test suites? >>>Allan >> >>Part of the people who analyzed the games are not close to be GM's. >> >>For example one of them Magnus holm has only 2198 fide rating(I searched his >>name in the online database of chessbase). >> >>I think that it is a bad idea to trust this test. >> >>I do not trust even GM's. >> >>I prefer to see a test suite that was generated by chess programs because I >>believe more in their moves. >> >>Uri >> >>Uri > >I agree. Why not use these positions in your suggestion for top engines to >analyze for 24 hours? Better yet, analayze the entire trace of: 1. The games actually played 2. The suggested alternatives 3. The moves the engines suggest It would take about a month of effort with several computers to complete. But when finished, the results would be a worthy test suite.
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.