working on the eval, but after following a discussion between H.G.Muller
and Shawn Chidester in the technical forum, I decided to implement futility pruning first.
And this gave a nice elo boost.
To get an idea of the new strength, here are the results of 1000 games:
Openingbook: Adam-Hair-10moves-199041.pgn
Ponder On
TimeControl: 40/60
gui: winboard
Galjoen - Jan Willem (1956 +/-11): 162 - 89 - 49 (W-L-D)
Galjoen - Floyd 0.6 (2072 +/- 25): 117 - 129 - 54
Galjoen - Parrot (2075 +/- 15): 109 - 163 - 28
Galjoen - Monarch (2055 +/- 12): 33 - 45 - 22
Using ordo plus the ratings and errors from ccrl 40/4:
Code: Select all
# PLAYER : RATING POINTS PLAYED (%)
1 Parrot : 2083.9 177.0 300 59.0%
2 Monarch v1.7 : 2058.0 56.0 100 56.0%
3 Floyd 0.6 : 2049.8 156.0 300 52.0%
4 Galjoen_v0.31 : 2028.5 495.0 1000 49.5%
5 JanWillem : 1953.0 116.0 300 38.7%