Author: Graham Laight
Date: 03:16:01 05/03/01
Go up one level in this thread
On May 02, 2001 at 12:59:47, Robert Hyatt wrote: >On May 02, 2001 at 12:34:51, Graham Laight wrote: > >>On May 02, 2001 at 12:28:09, Robert Hyatt wrote: >> >> >>Hi Bob, >> >>You're probably right - but when I studied the eval.c in the version of Crafty >>you sent me for the WMCCC 2000, I remember that some "anti-human" code had been >>removed. From memory, the code to prevent the Stonewall attack in particular. >> >>Apologies if I'm wrong. >> >>-g > > >If you still have the source, you might check. It should _not_ have been Just to confirm - Bob is right about this. When I briefly looked at the program before WMCCC 2000, I wrongly thought that the Stonewall attack code was commented out - but it isn't. -g >removed. When I am playing a computer, the program behaves differently in >two ways: > >1. it uses a more reasonable draw score than it does against humans. IE >for humans I might use -.3, but for a computer opponent, I always use 0.0.. > >2. In time management, sometimes humans try to play "shuffle-chess" and move >instantly to try to run crafty out of time. It recognizes this and speeds >itself up dramatically to avoid the problem. Against computers I disable this. > >I don't have any evaluation differences between computer and non-computer >opponents at all. > >The "stonewall" code wasn't removed to my knowledge. I did rewrite it and move >it somewhere else in the evaluation. It is certainly possible that the comments >were misplaced or missing for a while, but I have never knowingly played without >that as it is dangerous to let a computer play white and start a Stonewall >attack, which is quite easy to play even for a computer.
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.