Stockfish Testing Queue

Finished - 38987 tests

14-02-09 inf razor_margin diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 1342 W: 209 L: 311 D: 822
sprt @ 15+0.05 th 1 Extreme Try: Try to make razor_margin more exponential with respect to <depth> and see if it scales better. Also, use vd's tweaked pre-condition.
14-02-09 joa c_checks_stm diff
ELO: 0.43 +-2.4 (95%) LOS: 63.8%
Total: 40000 W: 9985 L: 9935 D: 20080
40000 @ 5+0.05 th 1 Measure stm bonus for contact checks
14-02-09 Fis checkextless_pvinstabil diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 22359 W: 4104 L: 4151 D: 14104
sprt @ 15+0.05 th 1 Combo of check_ext_less and pv_instability both of which passed STC and were positive on LTC.
14-02-09 rst null_tweak diff
LLR: -2.95 (-2.94,2.94) [0.00,6.00]
Total: 12722 W: 2380 L: 2418 D: 7924
sprt @ 60+0.05 th 1 Final take 2. Direct LTC because TC dependent. 2moves_v1book. Low prio.
14-02-09 inf pv_instability diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 1500 W: 228 L: 329 D: 943
sprt @ 15+0.05 th 1 Take 4: Decay PV faster when depth is greater
14-02-10 inf pv_instability diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 2132 W: 337 L: 436 D: 1359
sprt @ 15+0.05 th 1 Take 5: Decay PV faster when depth is greater. Little bit twisted, but works at short TC.
14-02-10 uri null_modify diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 12528 W: 2343 L: 2206 D: 7979
sprt @ 15+0.05 th 1 trying to do always qsearch in null move pruning with the idea that starting with qsearch before search help to get better order of moves.
14-02-10 hon master diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 46790 W: 8541 L: 8525 D: 29724
sprt @ 15+0.05 th 1 Avoid calls to pos.legal() in most cases in search()
14-02-10 rst null_intermediate diff
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 19321 W: 3643 L: 3488 D: 12190
sprt @ 15+0.05 th 1 Search at intermediate depth if remaining depth is high. Take 1
14-02-10 rst null_intermediate diff
LLR: -1.44 (-2.94,2.94) [-1.50,4.50]
Total: 6522 W: 1191 L: 1225 D: 4106
sprt @ 15+0.05 th 1 Search at intermediate depth if remaining depth is high. Take 2
14-02-10 uri null_modify diff
LLR: -2.97 (-2.94,2.94) [0.00,6.00]
Total: 9439 W: 1437 L: 1496 D: 6506
sprt @ 60+0.05 th 1 trying to do always qsearch in null move pruning with the idea that starting with qsearch before search help to get better order of moves.
14-02-10 rst null_intermediate diff
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 4119 W: 760 L: 844 D: 2515
sprt @ 60+0.05 th 1 Search at intermediate depth if remaining depth is high. Take 1
14-02-10 uri null_modify diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 21558 W: 3953 L: 4002 D: 13603
sprt @ 15+0.05 th 1 second try of null move pruning change we do not use null move pruning when eval<beta and I want also not to use null move pruning when result of qsearch is smaller than beta by the same logic
14-02-10 rst probcut diff
LLR: -1.03 (-2.94,2.94) [-1.50,4.50]
Total: 7507 W: 1382 L: 1399 D: 4726
sprt @ 15+0.05 th 1 Attempt to improve move ordering for probcut search
14-02-10 jos razor_margin diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 6975 W: 1554 L: 1644 D: 3777
sprt @ 15+0.05 th 1 Try SPSA values for razor_margin. Take 3.
14-02-10 rst iid_tweak1 diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 28894 W: 5356 L: 5386 D: 18152
sprt @ 15+0.05 th 1 Do IID search also if ttMove is from qsearch
14-02-10 uri null_modify1 diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 3188 W: 555 L: 652 D: 1981
sprt @ 15+0.05 th 1 use null move pruning in more cases and not only when the evaluation is at least beta.
14-02-11 hwi see_while diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 21865 W: 4038 L: 4086 D: 13741
sprt @ 15+0.05 th 1 Another attempted optimization of Position::see.
14-02-11 alw KingAttackW diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 26781 W: 4954 L: 4989 D: 16838
sprt @ 15+0.05 th 1 king attack values from CLOP set 1
14-02-11 pec tm_simple diff
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 34102 W: 6184 L: 6144 D: 21774
sprt @ 15+0.05 th 1 TM simplification
14-02-11 pec tm_simple diff
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 16518 W: 2647 L: 2545 D: 11326
sprt @ 60+0.05 th 1 LTC. TM simplification
14-02-11 gli master diff
ELO: 38.63 +-2.0 (95%) LOS: 100.0%
Total: 40000 W: 9385 L: 4956 D: 25659
40000 @ 60+0.05 th 1 Regression test after king capture see speedup
14-02-11 joa c_checks_stm diff
ELO: -0.28 +-2.4 (95%) LOS: 41.1%
Total: 40000 W: 10018 L: 10050 D: 19932
40000 @ 5+0.05 th 1 double only queen contact checks
14-02-11 joa c_checks_stm diff
ELO: -0.33 +-2.4 (95%) LOS: 39.4%
Total: 40000 W: 9966 L: 10004 D: 20030
40000 @ 5+0.05 th 1 double only rook contact checks
14-02-11 joa tm_simple diff
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 22406 W: 4390 L: 4312 D: 13704
sprt @ 40/10 th 1 TM simplification (x in y)
14-02-12 luc rounded_time diff
ELO: 0.28 +-2.2 (95%) LOS: 59.8%
Total: 40000 W: 8321 L: 8289 D: 23390
40000 @ 5+0.05 th 1 minor tweaks affecting time management: assess at very short TC (only intended as measurement)
14-02-12 uri reduction_modify diff
ELO: -5.32 +-2.9 (95%) LOS: 0.0%
Total: 20000 W: 3584 L: 3890 D: 12526
20000 @ 15+0.05 th 1 less late move reduction for small branching factor(surprisingly it has also a smaller bench). before using SPRT I prefer to measure elo change and I decide to measure elo change with 20,000 games before using 40,000 games because maybe the change is big and it is a waste of time to play 40,000 games for measuring.
14-02-12 jki tm_simple diff
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 21625 W: 3212 L: 3124 D: 15289
sprt @ 60+0.5 th 1 LTC. TM simplification, higher increment, regression test.
14-02-12 pec tm diff
LLR: -3.35 (-2.94,2.94) [-1.50,4.50]
Total: 111508 W: 20484 L: 20315 D: 70709
sprt @ 15+0.05 th 1 In case of pv instability don't think longer than 5 times.
14-02-12 jhe tempo_bonus diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 8941 W: 1590 L: 1672 D: 5679
sprt @ 15+0.05 th 1 Reduce tempo bonus.
14-02-13 uri time_formula diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 16522 W: 3005 L: 3067 D: 10450
sprt @ 15+0.05 th 1 I also care to have maximal number for BestMoveChanges but by a different way(4=1.2+0.7*4 so no increase from 4).
14-02-13 pec tm diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 35576 W: 6375 L: 6389 D: 22812
sprt @ 15+0.05 th 1 In case of pv instability don't think more than 4 times longer.
14-02-13 pec tm1 diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 8491 W: 1531 L: 1614 D: 5346
sprt @ 15+0.05 th 1 Hard stop iteration at 5 times more than usual time
14-02-13 pec tm diff
LLR: -4.38 (-2.94,2.94) [-1.50,4.50]
Total: 15244 W: 2744 L: 2860 D: 9640
sprt @ 15+0.05 th 1 In case of pv instability don't think more than 6 times longer. This I suppose very rare.
14-02-13 joa c_checks_stm diff
ELO: -8.88 +-2.5 (95%) LOS: 0.0%
Total: 38842 W: 9306 L: 10298 D: 19238
40000 @ 5+0.05 th 1 1.5 bonus for stm contact checks (now hopefully with correct bench)
14-02-13 pec tm1 diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 38867 W: 7117 L: 7121 D: 24629
sprt @ 15+0.05 th 1 Hard stop iteration at 7 times more than usual time
14-02-13 hxi checks_stm diff
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 17829 W: 3303 L: 3153 D: 11373
sprt @ 15+0.05 th 1 stm bonus also for safe checks
14-02-13 uri limit_ply diff
ELO: -130.86 +-5.7 (95%) LOS: 0.0%
Total: 10000 W: 1510 L: 5108 D: 3382
10000 @ 5 th 1 test the value of the demage with maximal ply of 18 plies
14-02-13 pec tm diff
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5759 W: 1150 L: 1029 D: 3580
sprt @ 15+0.05 th 1 Try in case of pv instability not thinking longer than 4.5 times
14-02-13 jos repfix diff
ELO: 1.45 +-2.1 (95%) LOS: 91.7%
Total: 40000 W: 7360 L: 7193 D: 25447
40000 @ 15+0.05 th 1 Final try to reduce early draws. First measure the impact on performance and draw rate.
14-02-13 hxi center_control diff
LLR: -2.97 (-2.94,2.94) [-1.50,4.50]
Total: 11705 W: 2118 L: 2193 D: 7394
sprt @ 15+0.05 th 1 bonus for attacked squares in center
14-02-13 jhe scale_outposts diff
LLR: -2.96 (-2.94,2.94) [-1.50,4.50]
Total: 8818 W: 1580 L: 1662 D: 5576
sprt @ 15+0.05 th 1 Scales outpost bonus based on number of enemy knights.
14-02-13 inf checks_stm diff
LLR: -2.95 (-2.94,2.94) [0.00,6.00]
Total: 12313 W: 1903 L: 1948 D: 8462
sprt @ 60+0.05 th 1 LTC for hx: stm bonus also for safe checks
14-02-14 Fis score_instability diff
LLR: -2.95 (-2.94,2.94) [-1.50,4.50]
Total: 5292 W: 930 L: 1021 D: 3341
sprt @ 15+0.05 th 1 Give extra time to moves for which the score changes by a relatively significant amount from the previous depth iteration.
14-02-14 pec remaining diff
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 33685 W: 6247 L: 6206 D: 21232
sprt @ 15+0.05 th 1 Remove unclear stuff from remaining. Test as simplification
14-02-14 jos maxply diff
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 17434 W: 3345 L: 3194 D: 10895
sprt @ 15+0.05 th 1 Return static eval when reaching max depth. Test against modified master MAX_PLY = 30.
14-02-14 uri limit_ply diff
ELO: -246.80 +-6.7 (95%) LOS: 0.0%
Total: 10000 W: 719 L: 6828 D: 2453
10000 @ 5+0.05 th 1 test the value of the demage with maximal ply of 18 plies
14-02-14 uri limit_ply diff
ELO: -22.96 +-4.9 (95%) LOS: 0.0%
Total: 10000 W: 2213 L: 2873 D: 4914
10000 @ 5 th 1 test the demage with 22 plies at very fast time control
14-02-14 pec tm diff
LLR: -2.94 (-2.94,2.94) [0.00,6.00]
Total: 33818 W: 5263 L: 5211 D: 23344
sprt @ 60+0.05 th 1 LTC. Try in case of pv instability not thinking longer than 4.5 times
14-02-14 vdb probcut diff
LLR: -2.95 (-2.94,2.94) [-4.00,0.00]
Total: 4362 W: 704 L: 875 D: 2783
sprt @ 15+0.05 th 1 Move probcut inside the main move loop. If it works then it is a simplification.