Variant Stockfish Testing Queue

Filter: variant = seirawan

Finished - 190 tests

02-02-18 ce seirawan sf9_merge diff 
ELO: 2.08 +-11.8 (95%) LOS: 63.6%
Total: 2000 W: 603 L: 591 D: 806
 2000 @ 30+0.3 th 1 Merge to SF9 minus 5 commits
01-02-18 ia seirawan seirawan_iid diff 
LLR: -2.97 (-2.94,2.94) [0.00,10.00]
Total: 9689 W: 2983 L: 2939 D: 3767
 sprt @ 10+0.1 th 1 more IID
01-02-18 ia seirawan seirawan_see_gates diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 2498 W: 732 L: 775 D: 991
 sprt @ 10+0.1 th 1 tweak SEE for captures on gates
28-01-18 ce seirawan sc_see diff 
LLR: 2.98 (-2.94,2.94) [0.00,10.00]
Total: 1074 W: 342 L: 258 D: 474
 sprt @ 30+0.3 th 1 LTC: Another try at SEE in Seirawan
28-01-18 ce seirawan sc_see diff 
LLR: 2.95 (-2.94,2.94) [0.00,10.00]
Total: 1267 W: 451 L: 360 D: 456
 sprt @ 10+0.1 th 1 Another try at SEE in Seirawan
26-01-18 ce seirawan sc_undo_see diff 
LLR: 2.97 (-2.94,2.94) [-10.00,0.00]
Total: 3753 W: 1036 L: 1008 D: 1709
 sprt @ 30+0.3 th 1 Seirawan see_ge looks buggy so I propose SPRT(-10,0) to revert to plain chess version. See commit message for more details.
26-01-18 ce seirawan sc_undo_see diff 
LLR: 2.95 (-2.94,2.94) [-10.00,0.00]
Total: 4648 W: 1420 L: 1403 D: 1825
 sprt @ 10+0.1 th 1 Seirawan see_ge looks buggy so I propose SPRT(-10,0) to revert to plain chess version. See commit message for more details.
23-01-18 ia seirawan tune_sc_gate_penalty diff 
19611/20000 iterations
40000/40000 games played
 40000 @ 10+0.1 th 1 tune weights for gate count based penalty
24-01-18 ce seirawan sc_eval_hand diff 
9723/10000 iterations
20000/20000 games played
 20000 @ 10+0.1 th 1 Tune alternative evaluation for hand/gates
25-01-18 ce seirawan sc_eval_hand2 diff 
LLR: 2.96 (-2.94,2.94) [0.00,10.00]
Total: 230 W: 122 L: 42 D: 66
 sprt @ 30+0.3 th 1 LTC: New hand evaluation
25-01-18 ce seirawan sc_eval_hand2 diff 
LLR: 2.95 (-2.94,2.94) [0.00,10.00]
Total: 261 W: 129 L: 50 D: 82
 sprt @ 10+0.1 th 1 STC: New hand evaluation
23-01-18 ce seirawan no_lazy diff 
LLR: 2.97 (-2.94,2.94) [-10.00,5.00]
Total: 3212 W: 906 L: 877 D: 1429
 sprt @ 30+0.3 th 1 LTC: I worry about LazyEval obscuring the large penalties for running out of gates so try to remove.
23-01-18 ce seirawan no_lazy diff 
LLR: 2.96 (-2.94,2.94) [-10.00,5.00]
Total: 1848 W: 578 L: 540 D: 730
 sprt @ 10+0.1 th 1 I worry about LazyEval obscuring the large penalties for running out of gates so try to remove.
21-01-18 ia seirawan seirawan_mobility_h diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 5814 W: 1583 L: 1587 D: 2644
 sprt @ 30+0.3 th 1 tweak mobility bonus for hawk. LTC
21-01-18 ia seirawan seirawan_mobility_h diff 
LLR: -2.98 (-2.94,2.94) [0.00,10.00]
Total: 9986 W: 3046 L: 2999 D: 3941
 sprt @ 10+0.1 th 1 tweak mobility bonus for hawk
21-01-18 ia seirawan seirawan_mobility_e diff 
LLR: -2.94 (-2.94,2.94) [0.00,10.00]
Total: 2290 W: 689 L: 735 D: 866
 sprt @ 10+0.1 th 1 tweak mobility bonus for elephant
18-01-18 ia seirawan seirawan_danger diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 5276 W: 1449 L: 1459 D: 2368
 sprt @ 30+0.3 th 1 tweak king danger parameters. LTC
19-01-18 ia seirawan tune_sc_mobility_e diff 
19060/20000 iterations
40000/40000 games played
 40000 @ 10+0.1 th 1 tune elephant mobility bonus
18-01-18 ia seirawan tune_seirawan_mobility_ diff 
19203/20000 iterations
40000/40000 games played
 40000 @ 10+0.1 th 1 tune hawk mobility bonus
20-01-18 ia seirawan seirawan_queen2 diff 
LLR: 2.96 (-2.94,2.94) [-10.00,5.00]
Total: 8389 W: 2341 L: 2342 D: 3706
 sprt @ 30+0.3 th 1 simplify away special handling of queen. LTC
20-01-18 ia seirawan seirawan_queen2 diff 
LLR: 2.97 (-2.94,2.94) [-10.00,5.00]
Total: 930 W: 308 L: 264 D: 358
 sprt @ 10+0.1 th 1 simplify away special handling of queen
19-01-18 ia seirawan seirawan_queen diff 
LLR: -2.97 (-2.94,2.94) [0.00,10.00]
Total: 15027 W: 4570 L: 4462 D: 5995
 sprt @ 10+0.1 th 1 remove special handling of queen
20-01-18 ia seirawan seirawan_inhand diff 
LLR: 2.96 (-2.94,2.94) [0.00,10.00]
Total: 3126 W: 918 L: 811 D: 1397
 sprt @ 30+0.3 th 1 bonus for pieces in hand. LTC
19-01-18 ia seirawan seirawan_inhand diff 
LLR: 2.95 (-2.94,2.94) [0.00,10.00]
Total: 6397 W: 2036 L: 1884 D: 2477
 sprt @ 10+0.1 th 1 bonus for pieces in hand
19-01-18 ia seirawan seirawan_promotions diff 
LLR: -2.96 (-2.94,2.94) [0.00,10.00]
Total: 4758 W: 1415 L: 1431 D: 1912
 sprt @ 10+0.1 th 1 generate elephant and hawk promotions as captures
19-01-18 ia seirawan seirawan_material diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 3909 W: 1180 L: 1206 D: 1523
 sprt @ 10+0.1 th 1 tweak material imbalance parameters
18-01-18 ia seirawan tune_sc_imbalance_he diff 
9583/10000 iterations
20000/20000 games played
 20000 @ 10+0.1 th 1 tune imbalance parameters for hawk and elephant
18-01-18 ia seirawan seirawan_danger diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 7032 W: 2139 L: 2127 D: 2766
 sprt @ 10+0.1 th 1 tweak king danger parameters
16-01-18 ia seirawan tune_seirawan_danger diff 
19287/20000 iterations
40000/40000 games played
 40000 @ 10+0.1 th 1 tune king danger
17-01-18 ia seirawan seirawan_mg_limit diff 
LLR: -2.98 (-2.94,2.94) [0.00,10.00]
Total: 386 W: 83 L: 152 D: 151
 sprt @ 30+0.3 th 1 tweak midgame limit. LTC
17-01-18 ia seirawan seirawan_mg_limit diff 
LLR: -2.95 (-2.94,2.94) [0.00,10.00]
Total: 534 W: 140 L: 209 D: 185
 sprt @ 10+0.1 th 1 tweak midgame limit
17-01-18 ia seirawan seirawan_pv diff 
LLR: 2.97 (-2.94,2.94) [0.00,10.00]
Total: 138 W: 99 L: 16 D: 23
 sprt @ 30+0.3 th 1 tweak piece values. LTC
17-01-18 ia seirawan seirawan_pv diff 
LLR: 2.96 (-2.94,2.94) [0.00,10.00]
Total: 178 W: 124 L: 33 D: 21
 sprt @ 10+0.1 th 1 tweak piece values
17-01-18 ia seirawan seirawan_danger_qeh diff 
LLR: 2.95 (-2.94,2.94) [0.00,10.00]
Total: 485 W: 179 L: 102 D: 204
 sprt @ 30+0.3 th 1 more Q<->E<->H symmetry in king danger. LTC
17-01-18 ia seirawan seirawan_danger_qeh diff 
LLR: 2.96 (-2.94,2.94) [0.00,10.00]
Total: 678 W: 267 L: 182 D: 229
 sprt @ 10+0.1 th 1 more Q<->E<->H symmetry in king danger
15-01-18 ia seirawan tune_seirawan_pv diff 
9297/10000 iterations
20000/20000 games played
 20000 @ 10+0.1 th 1 initial tuning of piece values
15-01-18 ia seirawan master diff 
ELO: -0.42 +-5.4 (95%) LOS: 44.0%
Total: 10000 W: 3148 L: 3160 D: 3692
 10000 @ 10+0.1 th 1 test new variant Seirawan chess (aka S-Chess)
15-01-18 ia seirawan master diff 
ELO: 17.91 +-12.2 (95%) LOS: 99.8%
Total: 2000 W: 695 L: 592 D: 713
 2000 @ 10+0.1 th 1 regression test for evaluation changes
15-01-18 ia seirawan d45e7e6bcb0c9c2f1adb5f6 diff 
ELO: -18.43 +-12.5 (95%) LOS: 0.2%
Total: 2000 W: 616 L: 722 D: 662
 2000 @ 10+0.1 th 1 regression test against first stable version
15-01-18 ia seirawan e0a287adb870de4ca511192 diff 
ELO: 17.91 +-12.1 (95%) LOS: 99.8%
Total: 2000 W: 685 L: 582 D: 733
 2000 @ 10+0.1 th 1 regression test of bugfixes