Stockfish Testing Queue

Finished - 45441 tests

16-08-26 sni pawn_weaknesses2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15757 W: 2932 L: 2989 D: 9836
sprt @ 10+0.1 th 1 Small midgame bonus for pressure on weak pawns, even if the pawns are protected
16-08-26 tvi id diff
LLR: -2.94 (-2.94,2.94) [-3.00,1.00]
Total: 40548 W: 7445 L: 7673 D: 25430
sprt @ 10+0.1 th 1 Take 4, (Take 3 at prio -1)
16-08-26 sni pawn_chains diff
LLR: -0.19 (-2.94,2.94) [0.00,5.00]
Total: 77 W: 11 L: 19 D: 47
sprt @ 10+0.1 th 1 Prefer pawn chains towards the center
16-08-26 sni no_null_move_trick diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20584 W: 3718 L: 3911 D: 12955
sprt @ 10+0.1 th 1 See whta happens if we remove the null move trick
16-08-26 pec narrow diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5547 W: 861 L: 962 D: 3724
sprt @ 5+0.05 th 7 Narrow window fast search instead of iteration skipping
16-08-26 sni pawn_weaknesses2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 6351 W: 1124 L: 1222 D: 4005
sprt @ 10+0.1 th 1 Small endgame bonus for pressure on weak pawns, even if the pawns are protected
16-08-24 tvi id diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 113325 W: 21126 L: 21166 D: 71033
sprt @ 10+0.1 th 1 Take 3
16-08-24 SC simpleOppbis diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 231529 W: 42918 L: 43164 D: 145447
sprt @ 10+0.1 th 1 Fixed base bench
16-08-25 IIv deep_search_tune diff
14700/15000 iterations
30000/30000 games played
30000 @ 10+0.1 th 1 Two parameters require further tuning with lower ck values.
16-08-25 Elb simple_eg2 diff
LLR: -0.84 (-2.94,2.94) [-3.00,1.00]
Total: 71800 W: 10566 L: 10713 D: 50521
sprt @ 10+0.1 th 1 Endgame simplification: retire KBPPKB endgame, let opposite bishop endgames be handled in KBPsK endgame. And in the process, improve evaluations for them.
16-08-26 cru evaluate_space_wt18 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 52151 W: 9824 L: 9799 D: 32528
sprt @ 10+0.1 th 1 Continue tuning evaluate space weight factor. Dividing by lower factor of 18 means even larger score.
16-08-25 luc weaknesses diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16858 W: 3137 L: 3189 D: 10532
sprt @ 10+0.1 th 1 Take 2.
16-08-26 Voy nme diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 28730 W: 5358 L: 5359 D: 18013
sprt @ 10+0.1 th 1 Null Move Experiment: If nullmove < beta and R is high... try null move again but with reduced R.
16-08-26 cib ft diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 6204 W: 1102 L: 1201 D: 3901
sprt @ 10+0.1 th 1 Futility tweak
16-08-26 cib opp_stat1 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7085 W: 1214 L: 1309 D: 4562
sprt @ 10+0.1 th 1 update opponent stats after null move or prubCut
16-08-25 luc weaknesses diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 12315 W: 2228 L: 2301 D: 7786
sprt @ 10+0.1 th 1 Merge and tweak St├ęphane's weak pawn patch and LooseEnemies bonus (linear bonus for a start).
16-08-25 cru evaluate_space_weight diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 47693 W: 8803 L: 8795 D: 30095
sprt @ 10+0.1 th 1 Change parameter to see how this behaves, sensitivity and direction. This is 20, i.e. increasing score.
16-08-25 cru evaluate_space_wt_2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 31658 W: 5845 L: 5892 D: 19921
sprt @ 10+0.1 th 1 Sensitivity and direction test, this tests 24, i.e. decreasing the evaluate_space score.
16-08-25 SC updateStatsTweak2 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 19417 W: 3599 L: 3640 D: 12178
sprt @ 10+0.1 th 1 And the same for opponents stats
16-08-25 Voy snm diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8870 W: 1602 L: 1776 D: 5492
sprt @ 10+0.1 th 1 Simplify Null Move Dynamic Formula...
16-08-25 SC updateStatsTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8274 W: 1453 L: 1543 D: 5278
sprt @ 10+0.1 th 1 Tweak quiets tt stats update (and thanks to Macro to make it understanble).
16-08-24 IIv deep_search_tune diff
28697/30000 iterations
58105/60000 games played
60000 @ 10+0.1 th 1 This patch solves some difficult and mating positions several times faster than master. So, trying to tune and test in the framework.
16-08-25 sni space_formula diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 5131 W: 902 L: 1042 D: 3187
sprt @ 10+0.1 th 1 More space evaluation and bigger space area
16-08-25 sni pawn_weaknesses diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 31580 W: 5912 L: 5900 D: 19768
sprt @ 10+0.1 th 1 Principle of multiple weaknesses: endgame
16-08-25 sg hanging diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11947 W: 2184 L: 2258 D: 7505
sprt @ 10+0.1 th 1 tighter futility margin if opponent has hanging pieces
16-08-25 SC smoothSpace diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18933 W: 3531 L: 3574 D: 11828
sprt @ 10+0.1 th 1 Reduce range and combine with increase in threshold.
16-08-25 lbr test diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23339 W: 4258 L: 4283 D: 14798
sprt @ 10+0.1 th 1 penalize king on blockSq
16-08-25 SC FauziTuning diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 5535 W: 1014 L: 1153 D: 3368
sprt @ 10+0.1 th 1 Tuning for Fauzi Akram see https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/FawXB3zUooY
16-08-25 lbr test diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25365 W: 4699 L: 4902 D: 15764
sprt @ 10+0.1 th 1 unify king distance rules
16-08-25 sni pawn_weaknesses diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 59381 W: 11059 L: 10927 D: 37395
sprt @ 10+0.1 th 1 Take 2: pawn weaknesses in endgame
16-08-25 sg hanging diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5631 W: 1004 L: 1106 D: 3521
sprt @ 10+0.1 th 1 Sorry, the hanging condition was added at the wrong if statement. I have stopped the last test and retry it now. No nullmove pruning if opponent has hanging pieces
16-08-25 sg hanging diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4147 W: 737 L: 845 D: 2565
sprt @ 10+0.1 th 1 no nullmove pruning if opponent has hanging pieces
16-08-25 mco nofmh2 diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7908 W: 1399 L: 1571 D: 4938
sprt @ 10+0.1 th 1 Drop fmh2
16-08-24 luc beyond_oppB diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 28357 W: 5280 L: 5283 D: 17794
sprt @ 10+0.1 th 1 One take at opposite bishops with other minors
16-08-25 sni pawn_weaknesses diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4063 W: 719 L: 828 D: 2516
sprt @ 10+0.1 th 1 Pawn weaknesses in midgame: quadratic bonus because of the principle of multiple weaknesses.
16-08-25 sni storm2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22642 W: 4205 L: 4232 D: 14205
sprt @ 10+0.1 th 1 Another definition of pawn advances
16-08-25 sni storm diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 21532 W: 4001 L: 4033 D: 13498
sprt @ 10+0.1 th 1 Try to favor kingside pawn majorities by also counting pawn advances in king tropism bonus
16-08-25 Voy cft-rp diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9392 W: 1694 L: 1779 D: 5919
sprt @ 10+0.1 th 1 Update refutation penalty for color-from-to stats.
16-08-24 tvi id diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 45646 W: 8560 L: 8558 D: 28528
sprt @ 10+0.1 th 1 Take 2
16-08-25 sni pawn_weaknesses diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5357 W: 988 L: 1091 D: 3278
sprt @ 10+0.1 th 1 Take 1: pawn weaknesses in midgame
16-08-24 cru evaluate_space_close diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 12497 W: 2286 L: 2358 D: 7853
sprt @ 10+0.1 th 1 Evaluate space when sides are close in material. Curious. Defect: might be called even later in the game, and might be a bad idea if temporary imbalance.
16-08-24 sni blockers_instead_of_pin diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11900 W: 2195 L: 2269 D: 7436
sprt @ 10+0.1 th 1 Use both pinned pieces and discovered check candidates for estimating attackUnits. Tested against the code in pull request #765, which makes this change trivial and fast.
16-08-24 SC smoothSpace diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22879 W: 4172 L: 4199 D: 14508
sprt @ 10+0.1 th 1 Smooth space contribution.
16-08-24 Elb simple_eg diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 11945 W: 1713 L: 1882 D: 8350
sprt @ 10+0.1 th 1 Simplification of bishop endgames that gives better evaluations.
16-08-24 SC simpleOppbis diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 28222 W: 5234 L: 5293 D: 17695
sprt @ 10+0.1 th 1 Try out suggestion by Luca Brivio. Test as a parameter tweak.
16-08-24 cru evaluate_space_4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11448 W: 2088 L: 2164 D: 7196
sprt @ 10+0.1 th 1 Evaluate space only until at most two minor pieces have been exchanged. In effect, this replaces 12222 with 13390 (at current MG values).
16-08-24 mco remove_stats2 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8830 W: 1561 L: 1734 D: 5535
sprt @ 10+0.1 th 1 Drop stat stuff: take 2
16-08-24 Voy futileCaptures diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8374 W: 1503 L: 1593 D: 5278
sprt @ 10+0.1 th 1 STC
16-08-24 mco remove_stats diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 1765 W: 253 L: 413 D: 1099
sprt @ 10+0.1 th 1 Drop some stat stuff
16-08-24 Elb tune_opp diff
21653/30000 iterations
43999/60000 games played
60000 @ 10+0.1 th 1 Tuning session for opp bishops endgame