Author: Uri Blass
Date: 00:22:42 11/30/02
Go up one level in this thread
On November 30, 2002 at 03:11:05, Dan Andersson wrote: >It is not trivial to test by random generation of pieces. The proportion of >legal positions is influenced by the mix of pieces on the board. To get an >accurate measure you need to multiply the fractional contribution of each class >of position with the frequency of legal positions. And to find what classes of >positions that you should include is another problem. All this could be >automated. A Bayesian learning process maybe. But you would need quite some time >to get a viable result. > >Mvh Dan Andersson I agree. I also said that it is not trivial but is not impossible. Uri
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.