Author: Tim Mirabile
Date: 13:13:22 10/08/97
Go up one level in this thread
On October 08, 1997 at 14:18:45, Ed Schröder wrote: >During this year I have done some experiments finding combinations >faster. The idea: if the current game is a position and no moves >played yet, extend *some* moves from the move list with 3 plies. > >I got wonderful solution times but if the algorithm failed a slow >down of 50-100% was noticed. I finally rejected implementation of >the idea because people would consider it as cheating. Absolutely! :) Unless you make it an option and turn it off by default. Otherwise it makes the whole point of running a test suite to try to estimate playing strength in real games useless. On the other hand, isn't this what humans do when confronted with a "white to move and win" position in a book of combinations? You could say you've made your program more "human-like" - it would find combinations easier from a book (EPD file) than when playing a game. :) (This is what spoils a book like Encyclopedia of Chess Middlegames for me. When there are chapter headings like Bishop sacrifices on h7/h2, it makes it a bit too easy to guess the key move.)
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.