Author: Stuart Cracraft
Date: 12:45:29 08/19/04
Go up one level in this thread
On August 19, 2004 at 15:42:05, Robert Hyatt wrote: >On August 19, 2004 at 15:27:56, Stuart Cracraft wrote: > >>Hi, >> >>I am looking around for those last few "big-ticket" items >>to implement in my program and am wondering about these. >> >>The program currently knows nothing about pawns other >>than preference for center pawns and pawns near the friendly >>castled king. >> >>So the question is how much improvement for games vs. >>normal middle-game position tests do you see coming from >>the three items square-of-pawn, pawn races, and general >>pawn structure including passers, isolani, doubled, color >>complex weakness, etc. as well as implementing a pawn >>hash. Be detailed. I'm interested in what kind of results >>before and after you added these that you saw. >> >>For example, I once heard one programmer comment that >>square of pawn was worth a full rating class in competitive play. >> >>And how do you implement square-of-pawn and and pawn races >>and what gotchas did you encounter? >> >>Thanks, >> >>Stuart > >Once you get your program tactically sound, you will lose every game you can't >win tactically, if you don't understand pawn structure. It is a _huge_ issue. > >If you don't understand weak pawns, isolated pawns, doubled pawns, pawn >majorities, distant majorities, outside/distant passers, etc, IM/GM players will >have you for lunch... Well, fortunately the program will never play against anyone stronger than me and it can whip me now so an IM and GM isn't needed; however, I'd like to add them anyway (without consulting existing code) in the hopes that it might raise my own level of play.
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.