Stockfish Testing Queue

Finished - 46979 tests

14-03-26 Mys iid_adjust diff
LLR: -2.21 (-2.94,2.94) [-1.50,4.50]
Total: 6672 W: 1208 L: 1269 D: 4195
sprt @ 15+0.05 th 1 Reduce more of the depth for internal iterative deepening as it saves going over already solid search.
14-03-26 mco asp_restore diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 67918 W: 10590 L: 10541 D: 46787
sprt @ 60+0.05 th 1 Restore old window size: test at LTC as a no-regression
14-03-26 sg queen_hanging diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 24338 W: 4541 L: 4582 D: 15215
sprt @ 15+0.05 th 1 Variation of Ralf's tests. check hanging queen only at ALL nodes
14-03-26 sg queen_hanging diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 12893 W: 2330 L: 2402 D: 8161
sprt @ 15+0.05 th 1 Variation of Ralf's tests. check hanging queen only if TT hit is upper bound
14-03-26 jos seepruning diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 13088 W: 2386 L: 2457 D: 8245
sprt @ 15+0.05 th 1 No SEE pruning near our king.
14-03-27 sni pawns_on_semi_open_file diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 3905 W: 705 L: 801 D: 2399
sprt @ 15+0.05 th 1 Evaluate pawns on semi-open files. Take 4.
14-03-27 Mys candidate_passed diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 32737 W: 6059 L: 5870 D: 20808
sprt @ 15+0.05 th 1 Give less bonus to what I think is lower ranks.
14-03-27 jos double_rooks^^ diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 40992 W: 7524 L: 7523 D: 25945
sprt @ 15+0.05 th 1 Bonus for doubled rooks on an open file. Take 1.
14-03-27 jos double_rooks^ diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 4983 W: 873 L: 965 D: 3145
sprt @ 15+0.05 th 1 Bonus for doubled rooks on an open file. Take 2.
14-03-27 jos double_rooks diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 4578 W: 805 L: 898 D: 2875
sprt @ 15+0.05 th 1 Bonus for doubled rooks on an open file. Take 3.
14-03-27 Mys candidate_passed diff
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 13395 W: 2039 L: 2080 D: 9276
sprt @ 60+0.05 th 1 LTC: give less bonus to what I think are lower ranks. Low priority while framework running on empty.
14-03-28 sg king_safety_blocked_che diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 32243 W: 5950 L: 5971 D: 20322
sprt @ 15+0.05 th 1 penalty for blocked checks (including discovered checks)
14-03-28 Mys delta diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5396 W: 956 L: 1047 D: 3393
sprt @ 15+0.05 th 1 Revisit delta. When we have large search time consider more moves, otherwise slight cut. (Delete if unacceptable).
14-03-28 sg king_safety_blocked_che diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 26180 W: 4745 L: 4782 D: 16653
sprt @ 15+0.05 th 1 penalty for blocked checks (including discovered checks). weights = 4
14-03-28 sg king_safety_pinned diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 10047 W: 1867 L: 1737 D: 6443
sprt @ 15+0.05 th 1 penalty for pinned pieces
14-03-28 Mys misplaced_minor diff
LLR: -2.94 (-2.94,2.94) [-1.50,4.50]
Total: 13686 W: 2477 L: 2546 D: 8663
sprt @ 15+0.05 th 1 Small penalty to misplaced minor pieces on edges of board.
14-03-28 Meh ext_zero diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7217 W: 1269 L: 1355 D: 4593
sprt @ 15+0.05 th 1 trying to see if extending a bit more when value is exactly zero worth investigating. (worked locally LTC, 500)
14-03-28 sg king_safety_pinned diff
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 10419 W: 1692 L: 1543 D: 7184
sprt @ 60+0.05 th 1 LTC: penalty for pinned pieces
14-03-28 Mys misplaced_knight diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 6739 W: 1211 L: 1299 D: 4229
sprt @ 15+0.05 th 1 Give penalty to knights only where they are out of play on A or H file.
14-03-28 sg king_safety_pinned2 diff
ELO: 1.49 +-2.1 (95%) LOS: 92.1%
Total: 40000 W: 7449 L: 7278 D: 25273
40000 @ 15+0.05 th 1 pit against succesful patch. weight = 1
14-03-28 sg king_safety_pinned2 diff
ELO: -1.71 +-2.3 (95%) LOS: 7.2%
Total: 32555 W: 5918 L: 6078 D: 20559
40000 @ 15+0.05 th 1 pit against succesful patch with weight=3 (Sorry, wrong weight commited on first attempt)
14-03-28 gli dangerous_edge_pawn diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 27368 W: 5013 L: 5047 D: 17308
sprt @ 15+0.05 th 1 50% increase file shelter penalties if dangerous edge pawn.
14-03-28 gli dangerous_edge_pawn diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 6910 W: 1246 L: 1333 D: 4331
sprt @ 15+0.05 th 1 Double file shelter penalties if dangerous edge pawn.
14-03-28 gli dangerous_edge_pawn diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 11195 W: 2030 L: 2106 D: 7059
sprt @ 15+0.05 th 1 Static penalty -50
14-03-28 gli dangerous_edge_pawn diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 12570 W: 2401 L: 2263 D: 7906
sprt @ 15+0.05 th 1 Static penalty -100
14-03-28 gli dangerous_edge_pawn diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 16970 W: 3156 L: 3217 D: 10597
sprt @ 15+0.05 th 1 Static penalty -150
14-03-28 hxi piece_values diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 33133 W: 6083 L: 6102 D: 20948
sprt @ 15+0.05 th 1 PawnValueMg -20 with compensation in pawn evaluation
14-03-28 sto 3linear_gamephase diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 3721 W: 615 L: 710 D: 2396
sprt @ 15+0.05 th 1 Yet another try to improve game phase calculation. piece-wise linear function this time.
14-03-28 sg king_safety_pinned2 diff
ELO: 0.58 +-1.9 (95%) LOS: 72.8%
Total: 40000 W: 6118 L: 6051 D: 27831
40000 @ 60+0.05 th 1 LTC: pit against succesful patch. weight = 1
14-03-28 Mys misplaced_knight diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 14385 W: 2632 L: 2700 D: 9053
sprt @ 15+0.05 th 1 Give penalty to knights only where they are out of play on A or H file. Change penalty Take 2.
14-03-29 Mys attack_pin diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 5249 W: 932 L: 1024 D: 3293
sprt @ 15+0.05 th 1 Try Stefan's idea in attacking evaluation.
14-03-29 sto gamephase_tweak3 diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10229 W: 1846 L: 1924 D: 6459
sprt @ 15+0.05 th 1 Yet another tweak at the game phase calculation. Local testing gives +100 -86 =233. 2 piece-wise linear interpolation.
14-03-29 sni supported_pawn_2 diff
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 43529 W: 8159 L: 7941 D: 27429
sprt @ 15+0.05 th 1 Slightly different implementation of hxim's idea for supported/unsupported pawns.
14-03-29 mco tt_rule diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7230 W: 1286 L: 1372 D: 4572
sprt @ 15+0.05 th 1 Replace TT entry even if new entry has same depth
14-03-29 Fis asp_interpolate diff
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 13621 W: 2071 L: 2111 D: 9439
sprt @ 60+0.05 th 1 Interpolate aspiration over a certain depth range. Take 2 likely final. Needs LTC.
14-03-29 sg king_safety_pinned diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 15991 W: 2895 L: 2959 D: 10137
sprt @ 15+0.05 th 1 add penalty also to supported pinned pieces
14-03-29 sg king_safety_pinned diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 96554 W: 17563 L: 17572 D: 61419
sprt @ 15+0.05 th 1 Simplification: drop not defended by pawn condition
14-03-29 sni supported_pawn_2 diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 17687 W: 3281 L: 3340 D: 11066
sprt @ 15+0.05 th 1 Slightly different implementation of hxim's idea. Take 2 : be more cautious with weak pawns in midgame.
14-03-29 sni supported_pawn_2 diff
LLR: -1.74 (-2.94,2.94) [-1.50,4.50]
Total: 11026 W: 2029 L: 2062 D: 6935
sprt @ 15+0.05 th 1 Slightly different implementation of hxim's idea. Take 3 (be even more cautious with weak pawns in midgame).
14-03-29 Fis tt_store_nosearch diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 27236 W: 4918 L: 4953 D: 17365
sprt @ 15+0.05 th 1 When storing TT entries don't search again for those we already have a pointer to from probe.
14-03-29 rst bishop_vs_knight diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 17775 W: 3215 L: 3274 D: 11286
sprt @ 15+0.05 th 1 Endgame bonus for bishop vs knight and pawns on both wings. Take 1
14-03-29 gli dangerous_edge_pawn diff
LLR: -2.97 (-2.94,2.94) [0.00,6.00]
Total: 8547 W: 1291 L: 1354 D: 5902
sprt @ 60+0.05 th 1 LTC: Static penalty -100
14-03-29 jos bishop_knight diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 16372 W: 2974 L: 3037 D: 10361
sprt @ 15+0.05 th 1 Endgame bonus for bishop (vs. knight) with pawns on both wings. Suggested at Talkchess.
14-03-29 rst bishop_vs_knight diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 6591 W: 1217 L: 1305 D: 4069
sprt @ 15+0.05 th 1 Endgame bonus for bishop vs knight and pawns on both wings. Take 2
14-03-29 tzi trapped_rook diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 6743 W: 1183 L: 1270 D: 4290
sprt @ 15+0.05 th 1 generalize trapped rook
14-03-30 sni keep_material diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5461 W: 977 L: 1068 D: 3416
sprt @ 15+0.05 th 1 Keep material if ahead.
14-03-30 Fis tt_rule2 diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 20196 W: 3660 L: 3713 D: 12823
sprt @ 15+0.05 th 1 Allow NOT overwriting any entry on TT::store(). Idea from Marcos tt_rule patch. Test w/ 16MB to simulate more normal hash pressure.
14-03-30 sg king_safety_pinned diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 18418 W: 3259 L: 3317 D: 11842
sprt @ 15+0.05 th 1 lower penalty for pinned pawn
14-03-30 sg king_safety_pinned diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 57761 W: 10401 L: 10358 D: 37002
sprt @ 15+0.05 th 1 lower penalty for pinned minor piece
14-03-30 Mys negative_contempt diff
LLR: -2.13 (-2.94,2.94) [-1.50,4.50]
Total: 9495 W: 1591 L: 1642 D: 6262
sprt @ 15+0.05 th 1 Contempt Factor = -15