Author: Don Dailey
Date: 15:13:19 12/16/97
Go up one level in this thread
On December 16, 1997 at 00:19:29, Dan Homan wrote: >Hi, > > My program, EXchess, is a little more than a year old. (It started as >a simple one-ply evaluator until I learned more, thanks to rgcc >and ccc.) I've had it playing on FICS for a few months now where >it has slowly achieved a rating of ~1900 blitz. However, until recently >- other computer programs, such as crafty clones have defeated it >easily. > > I'd like to report my program's first win against a crafty clone on >roughly >equal hardware. JPCrafty is crafty v14.2 running on a pentium 120 with >64 MB of ram. I have no idea what opening book it is using, if it has >any endgame tablebases, or if the operator has made any modifications to >the code. My program is EXchess v2.08 running on a Cyrix MediaGX >processor with 48 MB of ram. This processor runs my program about >10-20% slower than the pentium 133's we have at school, so my guess is >the hardware is about even. Congratulations! It's a lot of fun isn't it? I remember the first time my program beat a Richard Lang program. His program seemed so strong at the time that I wondered if I would ever win a single game! But all it takes is hard work and some solid engineering. You will problably find that most of your progress is in small incremental improvements but sometimes you will find a major improvement or learn of some important thing missing that will be a big leap. I've spent a lot of time looking for that secret "one line improvement" that adds 100 rating points. I never found it (except when it was a bug fix) but it's not wasted time. You learn a lot during the process. Good luck. -- Don
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.