Stockfish Testing Queue

Finished - 38881 tests

17-01-12 Elb step13 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 1561 W: 213 L: 331 D: 1017
sprt @ 10+0.1 th 1 Take 2
17-01-12 sni singular4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 44520 W: 7998 L: 8003 D: 28519
sprt @ 10+0.1 th 1 Use constant margin for singular move extension
17-01-12 Voy cft-cm diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13181 W: 2351 L: 2420 D: 8410
sprt @ 10+0.1 th 1 Try color-from-to for cm
17-01-12 Mys bk diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23216 W: 4192 L: 4218 D: 14806
sprt @ 10+0.1 th 1 backward/backward opposed pawns suck even more on ranks 2 & 3.
17-01-12 lbr counterMoves diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21975 W: 3177 L: 3360 D: 15438
sprt @ 30+0.3 th 1 do we need counterMoves ? Rerun STC at 30+0.3 (Throughout x1/3), because I suspect counterMoves do not scale, and CMH take over at longer tc.
17-01-10 SC lazyEval_tune diff
41729/50000 iterations
84712/100000 games played
100000 @ 30+0.3 th 1 It looked like diminishing returns. Try to tune at ATC.
17-01-12 Mys kn diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 31760 W: 5700 L: 5691 D: 20369
sprt @ 10+0.1 th 1 bonus for knights in semi symmetrical med/high pawn positions
17-01-12 pb0 clearance_stat diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25942 W: 4717 L: 4731 D: 16494
sprt @ 10+0.1 th 1 Take 2 with higher bonus
17-01-12 sni backward_value diff
LLR: -1.45 (-2.94,2.94) [0.00,4.00]
Total: 10884 W: 1921 L: 1961 D: 7002
sprt @ 10+0.1 th 1 Raise backward penalty
17-01-12 pb0 clearance_stat diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 17364 W: 3109 L: 3160 D: 11095
sprt @ 10+0.1 th 1 When a sliding-piece bestmove crosses (or ends on) the square which had been cleared by previous own move, this is a sign for a clearance motif. If the other quiet moves for ply-2 still have to be scored, give a bonus to the other moves that clear this square too.
17-01-11 sni weak_pawn_threats diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7957 W: 1379 L: 1470 D: 5108
sprt @ 10+0.1 th 1 Take 2 (submitted with priority -1)
17-01-09 sg sharing_cmh diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 31979 W: 3846 L: 3865 D: 24268
sprt @ 20+0.2 th 7 LTC: Update also counterMoveStats of neighbour thread. The neighbourhood is organized as a circle, so that the last thread triggers also updates for the main thread. Park with prio -1 to not interfere with the running SMP LTC test.
17-01-11 moj KnightBlockTune diff
19768/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 Tune knightblock score
17-01-12 Roc SimpleLazy diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 54317 W: 10008 L: 9651 D: 34658
sprt @ 10+0.1 th 1 Quick (0, 5) experiment to understand if we need the shrinking factor when above the lazy threshold. Tested against functionally equivalent version of # 966
17-01-12 Voy research diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11996 W: 2122 L: 2196 D: 7678
sprt @ 10+0.1 th 1 stc
17-01-12 sni king_tropism3 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 37094 W: 6775 L: 6742 D: 23577
sprt @ 10+0.1 th 1 Bonus for each defensive piece in our king flank
17-01-11 lbr counterMoves diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27570 W: 4844 L: 5047 D: 17679
sprt @ 10+0.1 th 1 do we need counterMoves ?
17-01-12 Voy research diff
LLR: -2.48 (-2.94,2.94) [0.00,5.00]
Total: 6842 W: 1218 L: 1294 D: 4330
sprt @ 10+0.1 th 1 fix patch
17-01-11 Mys rrp2 diff
LLR: -1.37 (-2.94,2.94) [-3.00,1.00]
Total: 31748 W: 5699 L: 5826 D: 20223
sprt @ 10+0.1 th 1 take 2 & last, compensation in bonus for popcount removal
17-01-11 sni weak_pawn_threats diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8679 W: 1550 L: 1638 D: 5491
sprt @ 10+0.1 th 1 Weak pawns multi-threats (after mojinumber1)
17-01-11 Voy research diff
LLR: -1.22 (-2.94,2.94) [0.00,5.00]
Total: 4140 W: 755 L: 789 D: 2596
sprt @ 10+0.1 th 1 stc
17-01-11 sni unstoppable4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 30123 W: 5353 L: 5407 D: 19363
sprt @ 10+0.1 th 1 Raise the passed pawns evaluation when the opponent has no pieces left (3*score/2)
17-01-11 Voy lmrt diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 24585 W: 4401 L: 4422 D: 15762
sprt @ 10+0.1 th 1 Fixed Faulty logic...
17-01-11 SC behindPawns diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 14360 W: 2553 L: 2617 D: 9190
sprt @ 10+0.1 th 1 Fixed version. Thanks to Rocky.
17-01-11 IIv c_lazy_eval diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 3844 W: 591 L: 753 D: 2500
sprt @ 10+0.1 th 1 Apply lazy eval globally. Take 2.
17-01-11 Mys rrp diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 107829 W: 19282 L: 19622 D: 68925
sprt @ 10+0.1 th 1 get rid of popcount on the rookonpawn, think it's weaker at ltc too
17-01-10 lbr counterMoves diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 85688 W: 11144 L: 10985 D: 63559
sprt @ 60+0.6 th 1 LTC: Use (from,to) instead of (pc,to) for MoveStats
17-01-10 Roc WeakPawnThreat diff
LLR: -2.60 (-2.94,2.94) [0.00,4.00]
Total: 47442 W: 8504 L: 8481 D: 30457
sprt @ 10+0.1 th 1 See if we can extend mojinumber1 backward threat S(4, 7) to weak pawns in general. The recent ThreatByRank bonus did not apply to pawn, so it might make sense to recalibrate those threats on weak pawn.
17-01-11 Voy lmrt diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8599 W: 1505 L: 1594 D: 5500
sprt @ 10+0.1 th 1 One more shot at this...
17-01-11 Elb step13 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 24071 W: 4328 L: 4351 D: 15392
sprt @ 10+0.1 th 1 Try joergoster's solution in https://github.com/official-stockfish/Stockfish/issues/760
17-01-11 IIv c_lazy_eval diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 190 W: 2 L: 174 D: 14
sprt @ 10+0.1 th 1 Continuous and simple lazy eval. Take 1.
17-01-11 Mys bis diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13219 W: 2322 L: 2391 D: 8506
sprt @ 10+0.1 th 1 penalise bishop in symmetrical pos
17-01-11 pb0 pinners diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10153 W: 1750 L: 1832 D: 6571
sprt @ 10+0.1 th 1 Take 2 (bugfix)
17-01-11 Roc SplitThreat2 diff
LLR: -4.20 (-2.94,2.94) [0.00,5.00]
Total: 29881 W: 5274 L: 5326 D: 19281
sprt @ 10+0.1 th 1 Score combo threats by N + B, but this time skip the rank threats for this case.
17-01-10 moj BackwardThreat diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 87262 W: 15786 L: 15541 D: 55935
sprt @ 10+0.1 th 1 A bonus for Backward pawn threat
17-01-10 moj knightblock2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8522 W: 1459 L: 1548 D: 5515
sprt @ 10+0.1 th 1 Decrease score if passed pawn is blocked by enemy knight, only in middle game.
17-01-10 sni unstoppable4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 64292 W: 11689 L: 11626 D: 40977
sprt @ 10+0.1 th 1 Double the passed pawns evaluation if the opponent has no pieces left
17-01-10 Roc StormDanger4Files diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18843 W: 3483 L: 3527 D: 11833
sprt @ 10+0.1 th 1 Bug fix on take 2.
17-01-10 Voy lmrt diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 55015 W: 9956 L: 9848 D: 35211
sprt @ 10+0.1 th 1 take 3
17-01-10 sni singular3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10843 W: 1904 L: 1983 D: 6956
sprt @ 10+0.1 th 1 More margin for singular captures or promotions: take 2
17-01-10 fau QueenThMo diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9757 W: 1705 L: 1789 D: 6263
sprt @ 10+0.1 th 1 Queen Threat + Queen Mobility
17-01-10 SC behindPawns diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7633 W: 1320 L: 1413 D: 4900
sprt @ 10+0.1 th 1 Avoiding computing bitboard behind in evaluate_space could give a small speed-up (and it is more idiomatic, too).
17-01-10 mbo stats_brq diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 6975 W: 1177 L: 1273 D: 4525
sprt @ 10+0.1 th 1 Update queen stats for bishop and rook moves.
17-01-10 Voy lmrt diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17154 W: 3034 L: 3087 D: 11033
sprt @ 10+0.1 th 1 take 2
17-01-10 Roc StormDanger4Files diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 1891 W: 334 L: 455 D: 1102
sprt @ 10+0.1 th 1 Take 2: In the 4 files model, only use half weight for the most distant file to the king.
17-01-10 Roc KingOnTheRim diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 23993 W: 4262 L: 4337 D: 15394
sprt @ 10+0.1 th 1 For a king on the h file, don't look at "closeenemies" attacking on file E Run as parameter tweak
17-01-10 Roc StormDanger4Files diff
LLR: -2.58 (-2.94,2.94) [0.00,5.00]
Total: 4431 W: 786 L: 877 D: 2768
sprt @ 10+0.1 th 1 Always compute 4 files for stormdanger
17-01-10 pb0 no_pinaware diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 43825 W: 7724 L: 7954 D: 28147
sprt @ 10+0.1 th 1 Since Atumanian's see_ge-implementation I'm not sure if my pinaware patch brings a benefit anymore (at least local test results were unclear) so trying to simplify away my own contribution.
17-01-10 lbr counterMoves diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 39202 W: 7113 L: 6823 D: 25266
sprt @ 10+0.1 th 1 Use (from,to) instead of (pc,to) for MoveStats
17-01-10 sni unstoppable4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 13358 W: 2341 L: 2452 D: 8565
sprt @ 10+0.1 th 1 Complete the circle: rank-dependent bonus for unstoppable pawns