Author: J. Wesley Cleveland
Date: 11:54:42 09/17/99
Go up one level in this thread
On September 17, 1999 at 11:28:04, Robert Hyatt wrote: >On September 17, 1999 at 09:55:43, Claudio A. Amorim wrote: > >>On September 16, 1999 at 23:59:25, Robert Hyatt wrote: >> [snip] >> >>Plain brute force is the answer. Given a sufficiently powerful hardware and a >>search algorithm good enough, the problem shall be over (as any other >>positional-judgement problem, for that matter)... >> >>Or not? > > >One day, yes. Seeing a passed pawn queen takes search. Seeing how an outside >passed pawn distracts the opposing king takes even more search. Seeing how an >offside majority turns into an outside passed pawn takes even more search. More >than I can do at present, so I am taking the evaluation approach. > >odd for a 'bean counter' as some would say. :) How about a tablebase-like approach. I would guess that 10^8 >the number of interesting pawn structures >10^4, so with a clever encoding scheme (Eugene ?), the whole table would fit in less than a gig, perhaps much less. It could contain enough hints to allow a program to have a good estimate of the value. Creating this table would not be easy (perhaps a PhD project).
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.