Author: Mogens Larsen
Date: 07:59:38 12/12/00
Go up one level in this thread
On December 12, 2000 at 10:20:35, Dave Gomboc wrote: >No need, it's pretty easy. Just take an opening book by a strong player and >enter the contents of it into a computer program while having a playing engine >analyzing. For example, Crafty does well with some parts of the Complete >Richter-Rauzer (Wells and Osnos), but it's overly pessimistic about Black's >chances when Black's king stays in the bunker in the centre and has piece >activity on the flanks, occasionally assessing white to be up a full pawn >positionally where the GM authors assess the position as equal. I guess that's one way of doing it. Problem is that the knowledge may not be especially useful unless you're trying to improve the evaluation of your own program. If you're analysing a game on the basis of a book then there's a point even though the importance of computer aid would be limited. This depends on your ambitions of course. I'm not sure exact evaluation would be especially important in this case. If you're analysing without a book then the problem is twofold. When you know that the program is wrong then you didn't really need it. When you don't, you're just being mislead by the evaluation if it's incorrect. The importing thing for me when analysing games is knowing when I'm dead lost or winning. A 100% exact evaluation isn't needed for that. Mogens.
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.