Revert "Double Impact of Gain tables"
authorMarco Costalba <mcostalba@gmail.com>
Fri, 5 Apr 2013 06:56:22 +0000 (08:56 +0200)
committerMarco Costalba <mcostalba@gmail.com>
Fri, 5 Apr 2013 06:59:38 +0000 (08:59 +0200)
This reverts commit 36c82b751ce227c05bfb

Seems a regression against 2.3.1 tested with 20K games at 60"+0.05

With patch applied
ELO: 15.44 +-2.8 (95%) LOS: 100.0%
Total: 20000 W: 3928 L: 3040 D: 13032

Without patch applied
ELO: 18.76 +-2.8 (95%) LOS: 100.0%
Total: 20000 W: 3903 L: 2824 D: 13273

bench: 4781239

src/search.cpp

index e246512e985bbccff13c91c307454f680dd24356..60667c8764e32a01f904477e66ad614b57b73b57 100644 (file)
@@ -876,7 +876,7 @@ split_point_start: // At split points actual search starts from here
           // but fixing this made program slightly weaker.
           Depth predictedDepth = newDepth - reduction<PvNode>(depth, moveCount);
           futilityValue =  ss->staticEval + ss->evalMargin + futility_margin(predictedDepth, moveCount)
-                         + 2 * Gain[pos.piece_moved(move)][to_sq(move)];
+                         + Gain[pos.piece_moved(move)][to_sq(move)];
 
           if (futilityValue < beta)
           {