Author: William Bryant
Date: 14:56:34 07/28/99
Go up one level in this thread
IMHO, I agree. In reading posts and learning much of the history of this hobby/, it is quite clear that the more competitive programmers have Master/Grandmaster help in tuning their programs for computer-human play. Dr. Hyatt had Grandmaster help for Cray blitz, and he constantly reports on feedback from Grandmasters playing crafty on ICC. Ed Schroeder (sp) has monthly tournaments for Rebel. The list goes on and on. Even the Sprackens were a programmer/player team. The point is the programmer may or may not be good at OTB chess -- (I hope to improve my chess as I improve my program)-- but that at some point a chess players knowledge will be essential to improve the quality of any program's play. I don't (yet) have an interface for my program to ICC (WinBoard/XBoard uses i/o redirection available in unix/linux/dos that is not available in MacOS)but it is on the list of things to do. I often wonder if the changes I make based upon test suite performance select for tactical sufficiency but not necessarily for better play. I am also not sure I have the knowledge (yet) to fully interprete every win or loss I would achieve on server play. A Master or GM would be able to look at the games however, and say, "Wow, you king safty stinks" or "You should push you passed pawns more agressively and you might win games."
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.