Stockfish Testing Queue

Finished - 35508 tests

14-01-07 dor psqt^^^ diff
ELO: -3.75 +-3.0 (95%) LOS: 0.6%
Total: 20000 W: 3689 L: 3905 D: 12406
20000 @ 15+0.05 th 1 Measure value of rook PSQT
14-01-07 dor psqt^^ diff
ELO: -3.01 +-3.0 (95%) LOS: 2.4%
Total: 20000 W: 3724 L: 3897 D: 12379
20000 @ 15+0.05 th 1 Measure value of queen PSQT
14-01-07 dor psqt^ diff
ELO: -39.64 +-3.2 (95%) LOS: 0.0%
Total: 20000 W: 3257 L: 5529 D: 11214
20000 @ 15+0.05 th 1 Measure value of king PSQT
14-01-07 jki kpsqt2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 31779 W: 4911 L: 4967 D: 21901
sprt @ 60+0.05 th 1 kpsqt, final
14-01-08 gli kingpp diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 4078 W: 718 L: 813 D: 2547
sprt @ 15+0.05 th 1 6 * enemy, 2 * friendly
14-01-08 gli kingpp diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5759 W: 1088 L: 1179 D: 3492
sprt @ 15+0.05 th 1 5 * enemy, 3 * friendly
14-01-08 gli kingpp diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 10382 W: 1901 L: 1979 D: 6502
sprt @ 15+0.05 th 1 4 * enemy, 2 * friendly
14-01-08 jos game_phase^^^ diff
ELO: -2.48 +-2.6 (95%) LOS: 3.0%
Total: 30000 W: 6419 L: 6633 D: 16948
30000 @ 5+0.05 th 1 Limits 4. Begin shifting MidgameLimit. 4 tests planned. If there is no improvement to be seen after these tests, idea is cancelled.
14-01-08 jos game_phase^^ diff
ELO: -4.40 +-2.6 (95%) LOS: 0.0%
Total: 30000 W: 6265 L: 6645 D: 17090
30000 @ 5+0.05 th 1 Limits 5.
14-01-08 bin pawn_psqt diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 6177 W: 1164 L: 1254 D: 3759
sprt @ 15+0.05 th 1 pawn psqt EG
14-01-08 jos game_phase^ diff
ELO: -6.64 +-3.1 (95%) LOS: 0.0%
Total: 20866 W: 4461 L: 4860 D: 11545
30000 @ 5+0.05 th 1 Limits 8.
14-01-08 sg update_stats diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 12187 W: 2333 L: 2196 D: 7658
sprt @ 15+0.05 th 1 update stats for pv moves too
14-01-08 uri play_even_faster diff
LLR: 0.04 (-2.94,2.94) [-1.50,4.50]
Total: 9803 W: 1820 L: 1793 D: 6190
sprt @ 15+0.05 th 1 test changing a parameter in my time management that means to play faster when you are stuck for some time in the first move without a fail low.
14-01-08 rst se_remove diff
ELO: -30.97 +-4.2 (95%) LOS: 0.0%
Total: 8740 W: 1080 L: 1857 D: 5803
10000 @ 60+0.05 th 1 LTC rough estimate of value of SE after removal of easy move exclusion search. Low prio.
14-01-08 joa mbonus diff
ELO: -10.22 +-3.1 (95%) LOS: 0.0%
Total: 20000 W: 3839 L: 4427 D: 11734
20000 @ 15+0.05 th 1 PP: Measure base mbonus
14-01-08 joa ebonus diff
ELO: -29.34 +-3.2 (95%) LOS: 0.0%
Total: 20000 W: 3497 L: 5182 D: 11321
20000 @ 15+0.05 th 1 PP: Measure base ebonus
14-01-08 jos game_phase diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 8127 W: 1492 L: 1576 D: 5059
sprt @ 15+0.05 th 1 Limits 10. Final try.
14-01-08 sg update_stats diff
LLR: -2.95 (-2.94,2.94) [0.00,6.00]
Total: 30076 W: 4733 L: 4697 D: 20646
sprt @ 60+0.05 th 1 LTC: update stats for pv moves too
14-01-08 pec f14cd1bb89d080f36a11df3 diff
ELO: 1.76 +-4.8 (95%) LOS: 76.5%
Total: 8670 W: 1875 L: 1831 D: 4964
20000 @ 5+0.05 th 1 Test Marco simplification of TM
14-01-08 lbr psqt^^^^^^ diff
ELO: -3.35 +-2.8 (95%) LOS: 1.0%
Total: 20000 W: 3397 L: 3590 D: 13013
30000 @ 60+0.05 th 1 LTC: Measure value of pawn PSQT
14-01-08 joa unsafe_squares_1 diff
ELO: -21.88 +-3.0 (95%) LOS: 0.0%
Total: 20000 W: 3284 L: 4542 D: 12174
20000 @ 15+0.05 th 1 PP: Measure unsafe squares 1
14-01-08 joa unsafe_squares_2 diff
ELO: -1.10 +-2.6 (95%) LOS: 20.3%
Total: 26014 W: 4826 L: 4908 D: 16280
30000 @ 15+0.05 th 1 PP: Measure unsafe squares 2
14-01-08 joa unsafe_squares_3 diff
ELO: 0.65 +-2.4 (95%) LOS: 70.2%
Total: 29845 W: 5585 L: 5529 D: 18731
30000 @ 15+0.05 th 1 PP: Measure unsafe squares 3
14-01-09 lbr update diff
LLR: -2.94 (-2.94,2.94) [-1.50,4.50]
Total: 620 W: 76 L: 180 D: 364
sprt @ 15+0.05 th 1 update_stats() regardless of whether bestValue >= beta or not.
14-01-09 uri play_even_faster diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 5289 W: 962 L: 1054 D: 3273
sprt @ 15+0.05 th 1 test changing a parameter in my time management that means to play faster when you are stuck for some time in the first move without a fail low.
14-01-09 gli master diff
ELO: 29.85 +-2.0 (95%) LOS: 100.0%
Total: 40000 W: 8593 L: 5165 D: 26242
40000 @ 60+0.05 th 1 Regression test after time management fix
14-01-09 sg update_stats diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 10142 W: 1884 L: 1963 D: 6295
sprt @ 15+0.05 th 1 update stats for pv moves too. Additionally use seperate Pv and Non-Pv countermove stats.
14-01-09 jki update_stats diff
ELO: -0.15 +-2.3 (95%) LOS: 45.0%
Total: 40000 W: 9318 L: 9335 D: 21347
40000 @ 2+0.05 th 1 Measure value at super-blitz: update stats for pv moves too
14-01-09 fwi scale_pow_by_mobility diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 3749 W: 673 L: 769 D: 2307
sprt @ 15+0.05 th 1 Scale Stefan Geschwentner scale_pawns_on_wings according to relative mobility differences
14-01-09 pec tm_simple diff
ELO: -6.40 +-4.0 (95%) LOS: 0.1%
Total: 12483 W: 2587 L: 2817 D: 7079
20000 @ 5+0.05 th 1 simplification: remove one constant from TM
14-01-09 jos rook_pst diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 6100 W: 1100 L: 1189 D: 3811
sprt @ 15+0.05 th 1 Try to improve Rook PSQT.
14-01-09 uri tune_time diff
ELO: 1.15 +-2.2 (95%) LOS: 84.5%
Total: 40000 W: 8489 L: 8357 D: 23154
40000 @ 5+0.05 th 1 try to tune the patch and before deciding about sprt I try to save time by using faster time control to see what direction to try next
14-01-09 uri tune_time diff
ELO: 0.20 +-2.2 (95%) LOS: 57.0%
Total: 40000 W: 8502 L: 8479 D: 23019
40000 @ 5+0.05 th 1 opposite direction
14-01-09 bin pawn_psqt diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 41528 W: 7912 L: 7697 D: 25919
sprt @ 15+0.05 th 1 pawn psqt EG take 2
14-01-09 mco f14cd1bb89d080f36a11df3 diff
ELO: -1.08 +-2.9 (95%) LOS: 23.7%
Total: 20000 W: 3716 L: 3778 D: 12506
20000 @ 15+0.05 th 1 Test Marco simplification of TM (I am going to apply the patch, but I want to be sure there is no regression)
14-01-09 joa defended_sq_1 diff
ELO: -3.95 +-2.6 (95%) LOS: 0.1%
Total: 26358 W: 4786 L: 5086 D: 16486
35000 @ 15+0.05 th 1 PP: Measure defended squares 1
14-01-09 joa defended_sq_2 diff
ELO: -1.46 +-2.9 (95%) LOS: 16.4%
Total: 20000 W: 3650 L: 3734 D: 12616
31000 @ 15+0.05 th 1 PP: Measure defended squares 2
14-01-09 joa defended_sq_3 diff
ELO: 0.98 +-3.0 (95%) LOS: 73.9%
Total: 19497 W: 3705 L: 3650 D: 12142
30000 @ 15+0.05 th 1 PP: Measure defended squares 3
14-01-09 sg update_stats_hist diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 51827 W: 9691 L: 9660 D: 32476
sprt @ 15+0.05 th 1 update stats for pv moves too. Additional double their bonus for history update.
14-01-09 rst pv_instability diff
ELO: -22.53 +-3.0 (95%) LOS: 0.0%
Total: 20000 W: 3241 L: 4536 D: 12223
20000 @ 15+0.05 th 1 TM: measure value of pv_instability.
14-01-09 rst pv_inst_tune diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 3825 W: 638 L: 733 D: 2454
sprt @ 15+0.05 th 1 Change pv_instability formula.
14-01-10 pec tm_simple diff
ELO: -1.39 +-3.1 (95%) LOS: 19.1%
Total: 20000 W: 4140 L: 4220 D: 11640
20000 @ 5+0.05 th 1 simplification: remove couple of constants from TM. This patch is also having effect of redistributing ~27% of stable PV time in favor of unstable PV
14-01-10 lbr kpp diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 2540 W: 459 L: 559 D: 1522
sprt @ 15+0.05 th 1 KPP distance: take 1
14-01-10 pec tm_simple diff
ELO: -3.65 +-2.9 (95%) LOS: 0.7%
Total: 20000 W: 3566 L: 3776 D: 12658
20000 @ 15+0.05 th 1 fixed number of games at 15+0.05 .simplification: remove couple of constants from TM. This patch is also having effect of redistributing ~27% of stable PV time in favor of unstable PV
14-01-10 uri sometimes_faster diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10289 W: 1877 L: 1955 D: 6457
sprt @ 15+0.05 th 1 testing an idea to play twice faster in some moves when you spend a significant time in the first move with no fail low(I also added the simplification of marco because I do not believe that it cause a regression and the test of 20000 games did not show a significant difference)
14-01-10 gli pawn_psqt diff
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 5400 W: 782 L: 859 D: 3759
sprt @ 60+0.05 th 1 Long TC for BI: pawn psqt EG take 2
14-01-10 lbr 9bafb268c633c59aca892c7 diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 17578 W: 3254 L: 3313 D: 11011
sprt @ 15+0.05 th 1 KPP distance: take 2
14-01-10 lbr 3eefb67da334da7203ec7bd diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 3736 W: 651 L: 747 D: 2338
sprt @ 15+0.05 th 1 KPP distance: take 3
14-01-10 uri sometimes_faster diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 1019 W: 149 L: 252 D: 618
sprt @ 15+0.05 th 1 this time I test playing twice faster if more than 95% of the time is used for the first move(meaning that stockfish refuted relatively fast the rest of the moves)
14-01-10 pec tm_simple diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 11396 W: 2075 L: 2150 D: 7171
sprt @ 15+0.05 th 1 use 20% more time for stable pv at the expense of unstable pv