Author: Dann Corbit
Date: 11:44:55 04/20/04
Go up one level in this thread
On April 20, 2004 at 14:35:01, James Murphy wrote: >Thanks for the replies and the help you have given. However, I was more >interested in a Mathematical proof of the system. Some way I could abstractly >reason about the program without having to use very specific testing. > >Any thoughts? In theory, chess is a perfect information game. In practice, it is not. You have to estimate. Some things to test for are 3-position repeat draws and 50-non-reversible move draws. Make sure that your program understands underpromotions. If you release a binary (and your program is Winboard or UCI compatible) the problems will get shaken out in a hurry.
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.