Stockfish Testing Queue

Finished - 4560 tests

14-10-25 sg fix_see diff
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 149153 W: 30448 L: 29925 D: 88780
sprt @ 15+0.05 th 1 use in SEE the same piece value for knights and bishops to avoid that an exchange of a biship with a knight is considered a bad capture. This affects the move order because bad captures are sorted at the end of the move list.
14-10-27 sg see_fix2 diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 25452 W: 5213 L: 5249 D: 14990
sprt @ 15+0.05 th 1 use in SEE the same piece value for knights and bishops to avoid that an exchange of a biship with a knight is considered a bad capture. This affects the move order because bad captures are sorted at the end of the move list. Use now standard piece value array (Take 2)
14-10-27 sg fix_see diff
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 13660 W: 2345 L: 2381 D: 8934
sprt @ 60+0.05 th 1 LTC: use in SEE the same piece value for knights and bishops to avoid that an exchange of a biship with a knight is considered a bad capture. This affects the move order because bad captures are sorted at the end of the move list.
14-11-12 sg move_order diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 16525 W: 3300 L: 3361 D: 9864
sprt @ 15+0.05 th 1 give bonus to underpromotions
14-11-14 sg move_order diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 9432 W: 1910 L: 1991 D: 5531
sprt @ 15+0.05 th 1 add bonus to double checks
14-11-17 sg knight_fork_threats diff
ELO: -0.21 +-3.1 (95%) LOS: 44.7%
Total: 20000 W: 4029 L: 4041 D: 11930
20000 @ 15+0.05 th 1 Add bonus for safe knight fork threats. Parameters tuned by population base incremental learning (see also my post at the forum).
14-11-18 sg spsa_knight_fork_threat diff
19493/20000 iterations
40000/40000 games played
40000 @ 15+0.05 th 1 tune now with spsa starting at my values get from PBIL. This is my first SPSA run, so check if everything is ok please.
14-11-19 sg knight_fork_threats diff
ELO: -2.21 +-3.0 (95%) LOS: 7.8%
Total: 20000 W: 3947 L: 4074 D: 11979
20000 @ 15+0.05 th 1 Use the SPSA tuned values. some parameters show a straight trend and so seems not be converged. So first a elo measure is done.
14-11-19 sg spsa_knight_fork_threat diff
19381/20000 iterations
39888/40000 games played
40000 @ 15+0.05 th 1 The first SPSA tuning attempt seems to give worser result than the initial values. Perhaps with higher resolution (c = 5) we get better results. Prio -1
14-11-20 sg knight_fork_threats diff
ELO: -1.02 +-3.0 (95%) LOS: 25.5%
Total: 20000 W: 3987 L: 4046 D: 11967
20000 @ 15+0.05 th 1 Measure values from second SPSA. Prio -1
14-11-21 sg spsa_knight_fork_threat diff
18999/20000 iterations
40000/40000 games played
40000 @ 15+0.05 th 1 use now a two dimensional table (indexed by piece types) for knight fork bonus
14-11-22 sg knight_fork_threats2 diff
ELO: 0.09 +-3.0 (95%) LOS: 52.2%
Total: 20000 W: 4006 L: 4001 D: 11993
20000 @ 15+0.05 th 1 measure elo of new approach using SPSA tuned values
14-11-22 sg ks_opposite_bishops diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 35906 W: 7096 L: 7104 D: 21706
sprt @ 15+0.05 th 1 opposite bishops favorites the king attacker
14-11-22 sg spsa_move_order_checks diff
19339/20000 iterations
39560/40000 games played
40000 @ 15+0.05 th 1 I have measured that quiet check moves gives a cutoff in 5-8%. On all quiet moves only 2% delivers a cutoff. So checks should be moved up on the move list. My double check patch failed, so i let tune the bonus for the different types of checks (starting at a moderate bonus of 50).
14-11-23 sg move_count_pruning diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 21547 W: 4269 L: 4317 D: 12961
sprt @ 15+0.05 th 1 exclude killers from move count based pruning
14-11-23 sg move_count_pruning diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 36269 W: 7280 L: 7287 D: 21702
sprt @ 15+0.05 th 1 exclude countermoves from move count based pruning
14-11-23 sg move_count_pruning diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 58279 W: 11745 L: 11691 D: 34843
sprt @ 15+0.05 th 1 exclude followupmoves from move count based pruning
14-11-23 sg move_order_checks diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 17514 W: 3508 L: 3567 D: 10439
sprt @ 15+0.05 th 1 use SPSA tuned values for checks move order
14-11-23 sg ks_opposite_bishops diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 5342 W: 1013 L: 1105 D: 3224
sprt @ 15+0.05 th 1 opposite bishops favorites the king attacker (Take 2)
14-11-24 sg ks_opposite_bishops diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 19250 W: 3779 L: 3833 D: 11638
sprt @ 15+0.05 th 1 opposite bishops favorites the king attacker (Take 3). Idea from Rocky640.
14-11-29 sg lmr diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 19180 W: 3834 L: 3888 D: 11458
sprt @ 15+0.05 th 1 less reduction if move has same target square as ttMove
14-12-01 sg king_safety diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 8478 W: 1693 L: 1777 D: 5008
sprt @ 15+0.05 th 1 use quadratic kingAttackersWeight
14-12-03 sg aspiration_window diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 17535 W: 3480 L: 3539 D: 10516
sprt @ 15+0.05 th 1 counter search instability: if fail highs and lows alternate widen the window by don't changing opposite bound
14-12-03 sg aspiration_window diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 11685 W: 2301 L: 2376 D: 7008
sprt @ 15+0.05 th 1 counter search instability: if fail highs and lows alternate widen the window (change opposite bound less) (Take 2)
14-12-03 sg aspiration_window diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 33826 W: 6851 L: 6864 D: 20111
sprt @ 15+0.05 th 1 try opposite. if NOT fail highs and lows alternate use tighter window (Take 3)
14-12-03 sg aspiration_window diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 47917 W: 9749 L: 9509 D: 28659
sprt @ 15+0.05 th 1 try opposite. if NOT fail highs and lows alternate use tighter window (Take 4)
14-12-04 sg aspiration_window diff
LLR: -2.95 (-2.94,2.94) [0.00,6.00]
Total: 21378 W: 3560 L: 3560 D: 14258
sprt @ 60+0.05 th 1 LTC: try opposite. if NOT fail highs and lows alternate use tighter window (Take 4)
14-12-11 sg stormdanger diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 40589 W: 8370 L: 8363 D: 23856
sprt @ 15+0.05 th 1 25% more Stormdanger bonus for blocked f6/c6 pawn (After Garys compile fix)
14-12-11 sg stormdanger diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 5899 W: 1168 L: 1259 D: 3472
sprt @ 15+0.05 th 1 25% less Stormdanger bonus for blocked pawn on b or g file (After Garys complie fix)
14-12-11 sg stormdanger diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 45064 W: 9115 L: 9097 D: 26852
sprt @ 15+0.05 th 1 Double Stormdanger bonus for blocked pawn on f5/c5
14-12-11 sg history_bonus diff
ELO: -3.23 +-3.1 (95%) LOS: 1.9%
Total: 20000 W: 3937 L: 4123 D: 11940
20000 @ 15+0.05 th 1 use half history bonus, so updates can occur up to depth 31 (instead depth 22). I am expect no significant difference on STC, but do a quick measurement as baseline for later attempts (Take 1)
14-12-12 sg stormdanger diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10472 W: 2075 L: 2153 D: 6244
sprt @ 15+0.05 th 1 25% more Stormdanger bonus for blocked pawn on g6/b6. Try some opposite because this test fails badly: http://tests.stockfishchess.org/tests/view/5489fa5b0ebc591511eb6ef2
14-12-13 sg history diff
LLR: -4.39 (-2.94,2.94) [-1.50,4.50]
Total: 19474 W: 3886 L: 3991 D: 11597
sprt @ 15+0.05 th 1 add half bonus if Max exceeded (so higher depth influence history)
14-12-17 sg big_king_safety diff
ELO: 3.46 +-2.2 (95%) LOS: 99.9%
Total: 40000 W: 8275 L: 7877 D: 23848
40000 @ 15+0.05 th 1 Measure big king safety tuning with corrected values (see forum). Prio -2
14-12-17 sg big_king_safety diff
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 10311 W: 1876 L: 1721 D: 6714
sprt @ 60+0.05 th 1 LTC: Measure big king safety tuning with corrected values
14-12-19 sg king_block_pawn diff
ELO: -2.18 +-2.4 (95%) LOS: 3.7%
Total: 32659 W: 6502 L: 6707 D: 19450
40000 @ 15+0.05 th 1 big_king_safety: its unlikely but perhaps the change from 200 to 300 for king blocks pawn is responsible for the elo gain, so go for safety and test this version against the current master (see pull request comment)
14-12-14 sg spsa_big_king_safety diff
47490/50000 iterations
90852/100000 games played
100000 @ 60+0.05 th 1 Big king safety tuning. Stormdanger and Shelterweakness indexed by file pairs (a/h,b/g,c/f,d/e). Special case where king blocks pawn is incorporated in Stormdanger. LTC because TC-dependant. There are 93 parameters so i use following SPSA configuration: Games=100000 Gamma=0.159 Alpha=0.558 C=5 (except in maxSafety C=10 is used) Prio -1
14-12-26 sg passed_pawns diff
LLR: -2.95 (-2.94,2.94) [-0.50,3.50]
Total: 9498 W: 1829 L: 1963 D: 5706
sprt @ 15+0.05 th 1 Because stockfish underestimates passed pawns in middle game raise base bonus factor slightly
14-12-28 sg spsa_big_king_safety diff
53749/50000 iterations
85254/100000 games played
100000 @ 60+0.05 th 1 retune king safety further because some parameters not converged in first run and the -300 handset values are tuned not at all. Additionally attack units, king danger array and king safety weight are included too, so the complete king safety related stuff is tuned. Low prio. Remark: the other tuning attempt http://tests.stockfishchess.org/tests/view/54a0620d0ebc597884f6935e from n_persson is wrong because of the reordering of stormdanger indices by Marco (see my comments on his repo).
14-12-29 sg pruning diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 40438 W: 8071 L: 8066 D: 24301
sprt @ 15+0.05 th 1 allow move count pruning only if distance to next Pv node is greater than one
14-12-29 sg pruning diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 4693 W: 959 L: 1054 D: 2680
sprt @ 15+0.05 th 1 allow move count pruning only if distance to next Pv node is greater than two (Take 2)
14-12-29 sg pruning diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7185 W: 1395 L: 1482 D: 4308
sprt @ 15+0.05 th 1 allow move count pruning only if distance to next Pv node is greater than one or move is no killer-, counter- and followupmove (Take 3)
14-12-31 sg pruning diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 2671 W: 494 L: 594 D: 1583
sprt @ 15+0.05 th 1 allow razoring only if distance to next Pv node is greater than one
15-01-01 sg big_king_safety diff
ELO: 9.22 +-3.1 (95%) LOS: 100.0%
Total: 19514 W: 4340 L: 3822 D: 11352
20000 @ 15+0.05 th 1 Quick measure of tuned values. For attack unit stuff i use finer granularity ( factor 4) because of very low values.
15-01-02 sg big_king_safety diff
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6399 W: 1192 L: 1056 D: 4151
sprt @ 60+0.05 th 1 LTC:Because STC seems so far very good (and i want to go to bed),i set up the LTC now. For attack unit stuff i use finer granularity ( factor 4) because of very low values.
15-01-07 sg spsa_pawns diff
51922/50000 iterations
99911/100000 games played
100000 @ 15+0.05 th 1 Tune pawn structure (except passed pawns and king shelter)
15-01-07 sg pawns diff
LLR: 2.95 (-2.94,2.94) [-0.50,3.50]
Total: 164666 W: 33319 L: 32703 D: 98644
sprt @ 15+0.05 th 1 Test pawn structure tuned values
15-01-09 sg pawns diff
LLR: -3.21 (-2.94,2.94) [0.00,4.00]
Total: 29850 W: 4948 L: 5020 D: 19882
sprt @ 60+0.05 th 1 LTC: Test pawn structure tuned values
15-01-10 sg pruning diff
ELO: 1.17 +-2.5 (95%) LOS: 82.2%
Total: 30000 W: 6029 L: 5928 D: 18043
30000 @ 15+0.05 th 1 Measure the effect of allowing move pruning at PV nodes. Inspired by following talkchess discussion: http://www.talkchess.com/forum/viewtopic.php?t=54761&start=80
15-01-10 sg pruning diff
ELO: 1.71 +-2.6 (95%) LOS: 90.2%
Total: 23404 W: 4008 L: 3893 D: 15503
30000 @ 60+0.05 th 1 LTC: Measure the effect of allowing move pruning at PV nodes. Little gain for STC. Because i'am interrested how this scales i prefer a fixed games test instead of a no-regression-sprt.