Computer Chess Club Archives


Search

Terms

Messages

Subject: Solution to my Null Move bug !!

Author: Geoff

Date: 16:11:19 06/07/03


Hello

Thanks to everyone that had a read of my Null move code and tried to help

The prize for spotting my deliberate mistake goes to Frank Schneider ;-)

In hindsight my mistake was embarassingly obvious !

The top level think routine was calling

  x = search(alpha, beta, depth, TRUE);    <--- Bug

meaning that on the first search it will have doNull = TRUE and try to make a
Null move at the root, which was basicially just wrong. I just changed that last
param to FALSE, still getting a huge speed up from the Null move, but my playing
strengh is back up to par !

Test Results               Points

Null Move with buggy line    8.5
Original TSCP	            12.5


Null Move with Bugfix line  16.0
Original TSCP	             5.0

          Regards Geoff



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.