Stockfish Testing Queue

Pending - 0 tests 0.0 hrs

None

Active - 0 tests

Finished - 41 tests

18-03-14 Gar pawn_endings_scale diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5857 W: 1156 L: 1258 D: 3443
sprt @ 10+0.1 th 1 STC: rerun before LTC now that a variant on the idea is in master
18-03-09 Gar pawn_endings_both diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 70399 W: 10915 L: 10772 D: 48712
sprt @ 60+0.6 th 1 LTC: pawn endings are decisive, scale both sides
18-03-09 Gar pawn_endings_scale diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 55706 W: 11798 L: 11404 D: 32504
sprt @ 10+0.1 th 1 STC: Pawn endings are decisive, smaller factor (fixed branch)
18-03-09 Gar pawn_endings_both diff
LLR: 3.01 (-2.94,2.94) [0.00,5.00]
Total: 22174 W: 4731 L: 4494 D: 12949
sprt @ 10+0.1 th 1 STC: pawn endings are decisive, scale both sides
18-03-09 Gar pawn_endings_win_3 diff
LLR: -1.12 (-2.94,2.94) [0.00,5.00]
Total: 17185 W: 3574 L: 3542 D: 10069
sprt @ 10+0.1 th 1 STC: pawn endings are decisive, smaller scale factor
18-03-09 Gar pawn_endings_win_4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15598 W: 3173 L: 3229 D: 9196
sprt @ 10+0.1 th 1 STC: pawn endings are decisive, large factor
18-03-08 Gar pawn_endings_win diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 102313 W: 15777 L: 15517 D: 71019
sprt @ 60+0.6 th 1 LTC: Pawns endings are usually decisive.
18-03-08 Gar pawn_endings_win2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8552 W: 1683 L: 1772 D: 5097
sprt @ 10+0.1 th 1 STC: Extend when entering pawn ending
18-03-08 Gar dynamic_time_checks diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32908 W: 5071 L: 4969 D: 22868
sprt @ 60+0.6 th 1 LTC: Avoid time losses when machine speed is unreliable or just slow.
18-03-07 Gar dynamic_time_checks diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 57857 W: 11880 L: 11827 D: 34150
sprt @ 10+0.1 th 1 Avoid time losses when machine speed is unreliable or just slow.
18-03-07 Gar pawn_endings_win diff
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 39238 W: 8264 L: 7949 D: 23025
sprt @ 10+0.1 th 1 Pawns endings are more decisive.
17-10-19 Gar missing_endgames diff
LLR: 0.55 (-2.94,2.94) [-3.00,1.00]
Total: 263973 W: 32941 L: 33247 D: 197785
sprt @ 60+0.6 th 1 LTC: replace bitbases by small heuristic
17-10-22 Gar missing_endgames_2 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 88142 W: 15422 L: 15725 D: 56995
sprt @ 10+0.1 th 1 Based on Fishtest posts, remove KPK KPKP and don't even add a heuristic
17-10-19 Gar running_passers diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3501 W: 607 L: 718 D: 2176
sprt @ 10+0.1 th 1 Evaluate running passers
17-10-18 Gar tune_running_passers diff
19770/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 Evaluate running passers: initial value tuning
17-10-18 Gar missing_endgames diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 37600 W: 6742 L: 6649 D: 24209
sprt @ 10+0.1 th 1 Replace bitbases by small heuristic
17-10-18 Gar outside_passers diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 14101 W: 2492 L: 2557 D: 9052
sprt @ 10+0.1 th 1 Evaluate outside passers drawishness
17-10-17 Gar tune_outside_passers diff
28826/30000 iterations
59963/60000 games played
60000 @ 20+0.2 th 1 Outside passers: fix base branches, don't tune material threshold, increase games
17-10-17 Gar tune_outside_passers diff
2434/10000 iterations
4964/20000 games played
20000 @ 20+0.2 th 1 Evaluate outside passers: tune starting values
17-10-14 Gar futility_reduction_3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8781 W: 1574 L: 1662 D: 5545
sprt @ 10+0.1 th 1 Variant 3 on more LMR in futile positions
17-10-14 Gar futility_reduction_2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5907 W: 1001 L: 1101 D: 3805
sprt @ 10+0.1 th 1 Variant 2 on more LMR in futile positions
17-10-13 Gar futility_reduction diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15609 W: 2770 L: 2829 D: 10010
sprt @ 10+0.1 th 1 More LMR in futile positions
17-10-13 Gar lmr_restructure diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3963 W: 648 L: 756 D: 2559
sprt @ 10+0.1 th 1 More LMR on captures (variant 2, lower throughput)
17-10-13 Gar lmr_uniform diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 2809 W: 442 L: 555 D: 1812
sprt @ 10+0.1 th 1 More LMR on captures (variant 3, lower throughput)
17-10-13 Gar lmr_all_caps diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 3640 W: 605 L: 715 D: 2320
sprt @ 10+0.1 th 1 Variation on more lMR for captures
17-10-09 Gar clop_lmr_1 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 99623 W: 17896 L: 17601 D: 64126
sprt @ 10+0.1 th 1 Wider SPRT bounds
17-10-09 Gar clop_lmr_1 diff
LLR: -0.07 (-2.94,2.94) [0.00,4.00]
Total: 13638 W: 2445 L: 2403 D: 8790
sprt @ 10+0.1 th 1 Tune LMR values
17-10-03 Gar futility_condition diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 17042 W: 2931 L: 3116 D: 10995
sprt @ 10+0.1 th 1 Remove an old, weird condition that's been there for 5 years (fix bounds)
17-10-03 Gar diff_lazy_eval_3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 3697 W: 626 L: 736 D: 2335
sprt @ 10+0.1 th 1 Only do LE in the 2nd level of QS
17-10-03 Gar futility_condition diff
LLR: -0.11 (-2.94,2.94) [-1.00,3.00]
Total: 50 W: 8 L: 14 D: 28
sprt @ 10+0.1 th 1 Remove a weird condition that's been untouched for >5 years
17-10-03 Gar diff_lazy_eval_2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9887 W: 1769 L: 1852 D: 6266
sprt @ 10+0.1 th 1 Differential lazy eval, with margin for qs futility
17-10-03 Gar diff_lazy_eval_1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4142 W: 703 L: 811 D: 2628
sprt @ 10+0.1 th 1 Differential lazy eval
17-10-02 Gar aspiration_update diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 13664 W: 2390 L: 2457 D: 8817
sprt @ 10+0.1 th 1 Small tweak to aspiration window opening
17-10-02 Gar no_pv_pruning diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6507 W: 1152 L: 1250 D: 4105
sprt @ 10+0.1 th 1 Don't prune in PV nodes.
17-10-02 Gar aspiration_update_2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5586 W: 969 L: 1071 D: 3546
sprt @ 10+0.1 th 1 Do a probe search to find the direction before opening the window
17-10-02 Gar lmr_capture_see diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23168 W: 4205 L: 4231 D: 14732
sprt @ 10+0.1 th 1 Using wining/losing capture knowledge in LMR decision. Variant 1
17-10-02 Gar lmr_capture_see_2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15294 W: 2744 L: 2804 D: 9746
sprt @ 10+0.1 th 1 Using wining/losing capture knowledge in LMR decision. Variant 2
17-09-30 Gar really_skip_quiets diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 32705 W: 5807 L: 5795 D: 21103
sprt @ 10+0.1 th 1 skip Quiet movegen/sort if skipQuiets is set early
17-09-26 Gar king_safety_3 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 11479 W: 2040 L: 2217 D: 7222
sprt @ 10+0.1 th 1 Tuning indicates it may be optimal to just remove this code (at least with a book). Verify on fishtest at low priority.
17-09-23 Gar king_safety_1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 21004 W: 3713 L: 3750 D: 13541
sprt @ 10+0.1 th 1
17-09-23 Gar king_safety_3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13256 W: 2352 L: 2421 D: 8483
sprt @ 10+0.1 th 1