Adjust best value after a pruned quiet move

Vizvezdenec · vondele · commit 0ef5d05102ce · 2024-03-29T10:15:17.000+01:00
Logic somewhat similar to how we adjust best value after pruned captures in qsearch, but in search this patch does it after pruned quiet moves and also to not full scale of futility value but to smth in between best value and futility value. Passed STC: https://tests.stockfishchess.org/tests/view/6601cf900ec64f0526c55c30 LLR: 2.93 (-2.94,2.94) <0.00,2.00> Total: 59936 W: 15722 L: 15369 D: 28845 Ptnml(0-2): 182, 7097, 15112, 7340, 237 Passed LTC: https://tests.stockfishchess.org/tests/view/66029b2d0ec64f0526c566f1 LLR: 2.96 (-2.94,2.94) <0.50,2.50> Total: 118362 W: 29953 L: 29460 D: 58949 Ptnml(0-2): 68, 13159, 32249, 13622, 83 closes #5141 bench 1772608
diff --git a/src/search.cpp b/src/search.cpp
@@ -995,12 +995,17 @@ Value Search::Worker::search(
 
                 lmrDepth += history / 5637;
 
+                Value futilityValue =
+                  ss->staticEval + (bestValue < ss->staticEval - 59 ? 141 : 58) + 125 * lmrDepth;
+
                 // Futility pruning: parent node (~13 Elo)
-                if (!ss->inCheck && lmrDepth < 15
-                    && ss->staticEval + (bestValue < ss->staticEval - 59 ? 141 : 58)
-                           + 125 * lmrDepth
-                         <= alpha)
+                if (!ss->inCheck && lmrDepth < 15 && futilityValue <= alpha)
+                {
+                    if (bestValue <= futilityValue && abs(bestValue) < VALUE_TB_WIN_IN_MAX_PLY
+                        && futilityValue < VALUE_TB_WIN_IN_MAX_PLY)
+                        bestValue = (bestValue + futilityValue * 3) / 4;
                     continue;
+                }
 
                 lmrDepth = std::max(lmrDepth, 0);
 
