Author: Robert Hyatt
Date: 09:36:22 11/21/97
Go up one level in this thread
On November 21, 1997 at 10:53:47, mike cooter wrote: > >On November 20, 1997 at 16:59:03, mike cooter wrote: > >>On the Chessbase website, there is a test suite developed by John Nunn. >> >>According to Chessbase these are equal positions and are good starting >>points for a program vs. program game because it eliminates opening >>books and tests the true playing strength of the program. >> >>Is this true? If it is, then Under 2/40 time conditions Fritz 5 is >>stronger than Rebel 9, MChess 6, and Genius 5. > >Thanks for the feedback guys. But I still have questions. IS a test like >this truly ojective? Are they equal positions which eliminate poor or >excellent opening books? the main point of this sort of test is that you play *two* games for each position, with each program playing both the black and white sides. This means the position "ought" to be book-independent, although I don't know that there is any guarantee that the test position is not in either program's book...
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.