Stockfish Testing Queue

Pending - 0 tests 0.0 hrs

None

Active - 0 tests

Finished - 891 tests

24-03-17 jo simplify_threat diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 42557 W: 5634 L: 5542 D: 31381
sprt @ 60+0.6 th 1 LTC: Since the tuning was pretty neutral, let's see if we can simplify the scoring of ThreatBySafePawn.
24-03-17 jo imbalance1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 28483 W: 5183 L: 5187 D: 18113
sprt @ 10+0.1 th 1 Now with bishop values.
23-03-17 jo simplify_threat diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 18895 W: 3361 L: 3237 D: 12297
sprt @ 10+0.1 th 1 Since the tuning was pretty neutral, let's see if we can simplify the scoring of ThreatBySafePawn.
23-03-17 jo threats_tuning1 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 42171 W: 7568 L: 7581 D: 27022
sprt @ 10+0.1 th 1 Check some values for ThreatBySafePawn.
22-03-17 jo imbalance1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 24955 W: 4524 L: 4543 D: 15888
sprt @ 10+0.1 th 1 Now with knight values added.
18-03-17 jo imbalance1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 17648 W: 3129 L: 3180 D: 11339
sprt @ 10+0.1 th 1 Re-add Linear imbalance table.
17-03-17 jo no_pruning_when_mate diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 29844 W: 5335 L: 5334 D: 19175
sprt @ 10+0.1 th 1 Final take. (beta < VALUE_MATE_IN_MAX_PLY)
17-03-17 jo no_pruning_when_mate diff
LLR: -0.27 (-2.94,2.94) [0.00,5.00]
Total: 2263 W: 405 L: 407 D: 1451
sprt @ 10+0.1 th 1 Besides bestValue, make sure alpha is greater than mated score, too. Take 2. (Although bench is the same it's a functional change.)
17-03-17 jo no_pruning_when_mate diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4295 W: 719 L: 826 D: 2750
sprt @ 10+0.1 th 1 Helps resolving mates. Does it eventually have a positive effect, in general?
17-03-17 jo BishopSet diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 26873 W: 4871 L: 4882 D: 17120
sprt @ 10+0.1 th 1 Fixed version.
16-03-17 jo BishopSet diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 1679 W: 286 L: 407 D: 986
sprt @ 10+0.1 th 1 Try the bishop set. Last try on this. Tuning hints to re-adding the formerly LinearCoefficient might be useful.
14-03-17 jo nmp_exp diff
ELO: 0.16 +-2.1 (95%) LOS: 55.8%
Total: 20000 W: 1912 L: 1903 D: 16185
20000 @ 60+0.6 th 1 Add some preconditions and be more cautious when to allow null-move pruning. Check for elo-gain/-loss at LTC. (Test with 8-moves book and lower throughput.)
15-03-17 jo KnightSet2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 5396 W: 945 L: 1048 D: 3403
sprt @ 10+0.1 th 1 Try the same for knights. (Test against passed PawnsSet.)
13-03-17 jo outpost_new diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 52199 W: 9311 L: 9291 D: 33597
sprt @ 10+0.1 th 1 Check some locally tuned values.
10-03-17 jo PawnsSet diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 100478 W: 13055 L: 12615 D: 74808
sprt @ 60+0.6 th 1 LTC: Final values and last try.
10-03-17 jo PawnsSet diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16155 W: 2980 L: 2787 D: 10388
sprt @ 10+0.1 th 1 Final values and last try.
08-03-17 jo PawnsSet diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 62328 W: 11045 L: 10909 D: 40374
sprt @ 10+0.1 th 1 Another try with some handmade values.
07-03-17 jo PawnsSet diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10153 W: 1763 L: 1845 D: 6545
sprt @ 10+0.1 th 1 Now with some quickly tuned values. Take 2.
05-03-17 jo PawnsSet diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 49136 W: 8656 L: 8576 D: 31904
sprt @ 10+0.1 th 1 Try GuardianRM's idea in imbalance calculation. Start with some guessed values.
05-03-17 jo aspiration_eg diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8549 W: 1499 L: 1588 D: 5462
sprt @ 10+0.1 th 1 Open aspiration window when material gets low.
01-03-17 jo tune_outpost2 diff
58539/60000 iterations
119999/120000 games played
120000 @ 10+0.1 th 1 Retry the fresh tuning on the framework, STC, lower throughput. Let's see, if the increased ck values work in practice.
04-03-17 jo outpost_new diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 11150 W: 1950 L: 2069 D: 7131
sprt @ 10+0.1 th 1 Check the values from fishtest.
03-03-17 jo rook_bonus_new diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 26656 W: 4705 L: 4771 D: 17180
sprt @ 10+0.1 th 1 Also check new values for RookOnFile bonus.
28-02-17 jo outpost_new diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 20534 W: 3616 L: 3703 D: 13215
sprt @ 10+0.1 th 1 Values from a fresh SPSA tuning starting from zero.
28-02-17 jo pawn_push_lmr diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11339 W: 1934 L: 2011 D: 7394
sprt @ 10+0.1 th 1 Less LMR for pawn pushes with low material.
24-02-17 jo initiative diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 2972 W: 492 L: 605 D: 1875
sprt @ 10+0.1 th 1 Initiative tweak.
17-02-17 jo matimb_fix diff
ELO: -3.66 +-2.4 (95%) LOS: 0.2%
Total: 30000 W: 5642 L: 5958 D: 18400
30000 @ 7+0.07 th 1 Ready for going sprt?
08-02-17 jo matimb_fix diff
ELO: -1.53 +-2.4 (95%) LOS: 11.0%
Total: 30000 W: 5752 L: 5884 D: 18364
30000 @ 7+0.07 th 1 Now with updated queen values. Interestingly, some of them show no real sign of convergence and may contribute nothing at all.)
04-02-17 jo space_eg diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 2221 W: 373 L: 525 D: 1323
sprt @ 10+0.1 th 1 Try an endgame bonus in space eval.
04-02-17 jo entry_points diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 5067 W: 841 L: 945 D: 3281
sprt @ 10+0.1 th 1 Try a variation of Stéphane's patch.
02-02-17 jo udm1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9071 W: 1539 L: 1626 D: 5906
sprt @ 10+0.1 th 1 My try on piece development. Take 1.
31-01-17 jo matimb_fix diff
ELO: -3.82 +-2.5 (95%) LOS: 0.1%
Total: 30000 W: 5735 L: 6065 D: 18200
30000 @ 7+0.07 th 1 Values have noticeably changed. To the better or worse?
29-01-17 jo smp_verification diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 1178 W: 134 L: 251 D: 793
sprt @ 5+0.05 th 7 Avoid doing null-move/LMR always at the same ply. This reliably solves 2k5/p7/Pp1p1b2/1P1P1p2/2P2P1p/3K3P/5B2/8 w - - 0 1, but seems to be a good idea in general at longer thinking times.
28-01-17 jo matimb_fix diff
ELO: -4.38 +-3.0 (95%) LOS: 0.2%
Total: 20000 W: 3774 L: 4026 D: 12200
20000 @ 7+0.07 th 1 Any progress?
28-01-17 jo matimb_fix diff
ELO: 0.90 +-7.5 (95%) LOS: 59.3%
Total: 3095 W: 586 L: 578 D: 1931
20000 @ 7+0.07 th 1 Any progress?
26-01-17 jo smp_verification diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 27394 W: 4296 L: 4313 D: 18785
sprt @ 5+0.05 th 7 SMP tweak; less reduced verification search for every second thread.
25-01-17 jo matimb_fix diff
ELO: -8.95 +-3.0 (95%) LOS: 0.0%
Total: 20000 W: 3726 L: 4241 D: 12033
20000 @ 7+0.07 th 1 Just a quick check to see if it makes sense to continue. (Tuning in groups and Chess960 games.)
22-01-17 jo matimb_fix diff
LLR: -3.11 (-2.94,2.94) [0.00,5.00]
Total: 43232 W: 7808 L: 7756 D: 27668
sprt @ 10+0.1 th 1 Changed implementation and decreased bonus. Take 2.
21-01-17 jo matimb_fix diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 30262 W: 3908 L: 3928 D: 22426
sprt @ 60+0.6 th 1 Speculative LTC: Check for 'real' pair of bishops in material imbalance.
20-01-17 jo matimb_fix diff
LLR: -3.09 (-2.94,2.94) [0.00,5.00]
Total: 90600 W: 16629 L: 16372 D: 57599
sprt @ 10+0.1 th 1 Check for 'real' pair of bishops in material imbalance.
19-01-17 jo matimb4 diff
Pending...
sprt @ 10+0.1 th 1 Standalone tuning of material imbalance values using L-BFGS. First try, took less than 5 minutes. (See also Álvaro Begué's thread at talkchess.)
17-01-17 jo no_drawsore_to_tt diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12743 W: 2236 L: 2307 D: 8200
sprt @ 10+0.1 th 1 Don't save draw scores in endgame. Take 4.
17-01-17 jo no_drawsore_to_tt^ diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11305 W: 2014 L: 2091 D: 7200
sprt @ 10+0.1 th 1 How is this one doing?
16-01-17 jo pawn_tune_spsa2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 21873 W: 3997 L: 4079 D: 13797
sprt @ 10+0.1 th 1 Test some preliminary values of another tuning experiment after 50k games, starting from S(20, 20). Take 1.
14-01-17 jo skip_pruning2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 30561 W: 5472 L: 5468 D: 19621
sprt @ 10+0.1 th 1 Don't prune king moves in pawn endgames only. Take 2.
14-01-17 jo skip_pruning2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 10168 W: 1818 L: 1900 D: 6450
sprt @ 10+0.1 th 1 Use shallowPruning flag to additionally skip early pruning. Final try.
06-01-17 jo MCP1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 8186 W: 1468 L: 1559 D: 5159
sprt @ 10+0.1 th 1 Try the same with movecount based pruning.
06-01-17 jo FUT1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 11234 W: 1989 L: 2067 D: 7178
sprt @ 10+0.1 th 1 Slowly increase amount of futility pruning. Is this too oldschool?
05-01-17 jo less_lmr diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 2176 W: 346 L: 463 D: 1367
sprt @ 10+0.1 th 1 Try a less restrictive version.
04-01-17 jo skip_pruning2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 1410 W: 189 L: 307 D: 914
sprt @ 10+0.1 th 1 Don't prune king moves.