Author: David Dahlem
Date: 14:27:43 06/20/04
Go up one level in this thread
The subject in this thread is "test suites", which have a proposed solution move, which may or may not be correct. There are only two possible results from each position in a test suite, either the proposed solution move is found, or is not found. All i'm saying is that counting the number of so-called solved positions is a totally unreliable way to test engine strength. An engine may not find the proposed "best" move, but might find a "better" move than the comparison engine(s). And this isn't reflected in test suite results. That's the problem with test suites, as i see it. "Analyzing" a single position is a totally different concept than "test suites". Regards Dave
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.