Author: Danniel Corbit
Date: 09:50:26 10/02/98
Go up one level in this thread
Weighting systems work for computers but I don't think that they will add anything for a human. For me, anyway, I have to really understand how an opening *works* to use it. Just having some number like +.63 for an opening would not be beneficial. On the other hand, to have a ranking by computer evaluation and by wins and losses in actual games -- using both of these measures humans could figure out which openings to *study* and that could be of great benefit. Trying to apply this sort of knowlege to humans may be difficult and certainly will be imprecise. There are probably some GM's that can just glance at an unfamiliar position and know how it *works* whereas there may be persons who will never master more than one or two openings. Also, the computer analysis can be faulty. In particular Zugzwag positions freak out a large number of algorithms [those that rely on null move pruning]. Also, the win-loss statistics can be faulty. Suppose a great GM like Petrosian likes a certain opening and plays it a lot. That opening will win a lot, just because it is being played by a great player. Some other very good players may see him winning a lot and try it too. So an opening could become in vogue by good players. Later a hole gets discovered. But *in the past* it was played successfully so many times that the statistic is skewed. If that opening is now avoided, the statistic could remain stuck at the old value. So to directly answer your question: Statistics and studies have been done on openings. In fact, I am doing one right now. But to get a really good set of answers is problematic at best.
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.