Stockfish Testing Queue

Pending - 22 tests 0.0 hrs

19-08-04 Lol lpt diff
LLR: -1.79 (-2.94,2.94) [0.00,4.00]
Total: 29750 W: 6583 L: 6565 D: 16602
sprt @ 10+0.1 th 1 Lower lazy threashold after recent pawns changes
19-08-24 sg ks_attacks diff
LLR: -1.50 (-2.94,2.94) [0.50,4.50]
Total: 15000 W: 3332 L: 3341 D: 8327
sprt @ 10+0.1 th 1 Safe queen king attacks. Bonus 70.
19-08-24 sg lmr_delayed diff
LLR: -1.88 (-2.94,2.94) [0.50,4.50]
Total: 18000 W: 3927 L: 3943 D: 10130
sprt @ 10+0.1 th 1 ttPv reduction: -1 ply directly, -1 ply delayed.
19-08-24 sg ks_attacks diff
LLR: -1.62 (-2.94,2.94) [0.50,4.50]
Total: 36750 W: 8205 L: 8113 D: 20432
sprt @ 10+0.1 th 1 Bonus 100
19-08-24 sg lmr_delayed diff
LLR: -2.25 (-2.94,2.94) [0.50,4.50]
Total: 30500 W: 6782 L: 6756 D: 16962
sprt @ 10+0.1 th 1 Retest cut node split with clearance for full depth search.
19-08-23 sg lmr_delayed diff
LLR: -2.30 (-2.94,2.94) [0.50,4.50]
Total: 43750 W: 9729 L: 9640 D: 24381
sprt @ 10+0.1 th 1 Cut node reduction: one ply additional delayed reduction.
19-08-22 xot vary4pr diff
LLR: 0.43 (-2.94,2.94) [0.00,4.00]
Total: 26250 W: 4163 L: 4061 D: 18026
sprt @ 20+0.2 th 8 Remove added complexity, just pure param changes.
19-08-23 sg lmr_bestmove diff
LLR: -2.34 (-2.94,2.94) [0.50,4.50]
Total: 30250 W: 6678 L: 6659 D: 16913
sprt @ 10+0.1 th 1 Merge remote-tracking branch 'origin' into lmr_bestmove.
19-08-22 MJZ psq_npm diff
LLR: -0.71 (-2.94,2.94) [0.00,3.50]
Total: 75000 W: 12444 L: 12278 D: 50278
sprt @ 60+0.6 th 1 Increase score in late EG. +1/2, NPM==0 and abs(v) > 120 - LTC
19-08-22 sg tuned_king_danger diff
LLR: -1.96 (-2.94,2.94) [0.00,4.00]
Total: 15000 W: 2527 L: 2581 D: 9892
sprt @ 60+0.6 th 1 As expected the STC failed yellow so test my at intermediate TC tuned values at LTC (Low TP).
19-08-22 sg lmr_bestmove2 diff
LLR: -1.68 (-2.94,2.94) [0.50,4.50]
Total: 8750 W: 1945 L: 1995 D: 4810
sprt @ 10+0.1 th 1 Do less reduction at root according to the numbers the move was already a best move in the last iterations.
19-08-21 MJZ psq_npm2 diff
LLR: -2.24 (-2.94,2.94) [0.50,4.50]
Total: 61500 W: 13755 L: 13574 D: 34171
sprt @ 10+0.1 th 1 Increase score in late EG. +1/4, NPM==0
19-08-22 xot vary4pr diff
LLR: 0.24 (-2.94,2.94) [-3.00,1.00]
Total: 10250 W: 1689 L: 1693 D: 6868
sprt @ 60+0.6 th 1 LTC test for non-regression on 1 thread. Hope to get elo gain on multiple threads.
19-08-21 sg lmr_cutnode_cap diff
LLR: -2.24 (-2.94,2.94) [0.50,4.50]
Total: 40750 W: 9040 L: 8963 D: 22747
sprt @ 10+0.1 th 1 At LMR for captures replace cut node condition with capture history <= 0. (Park at prio -1)
19-08-21 MJZ psq_npm diff
LLR: -1.58 (-2.94,2.94) [0.50,4.50]
Total: 33250 W: 7340 L: 7264 D: 18646
sprt @ 10+0.1 th 1 Increase psq score in late EG. +1/4, NPM==0
19-08-21 sg lmr_cutnode_cap diff
LLR: -2.02 (-2.94,2.94) [0.50,4.50]
Total: 33000 W: 7321 L: 7270 D: 18409
sprt @ 10+0.1 th 1 Variation of Vizvezdenec's passed patch and do LMR for captures at cut nodes only if last move was no capture.
19-08-18 Roc PinnedQueen2 diff
LLR: -2.02 (-2.94,2.94) [0.00,4.00]
Total: 25651 W: 5686 L: 5697 D: 14268
sprt @ 10+0.1 th 1 See git notes.
19-08-18 Roc 64738f80c5b7e53eebf8320 diff
LLR: -0.85 (-2.94,2.94) [0.50,4.50]
Total: 59891 W: 13348 L: 13099 D: 33444
sprt @ 10+0.1 th 1 Take 2b; leave RookOnFile as master, and reduce ThreatOnPawn.
19-08-05 Lol lpt_2 diff
LLR: -1.70 (-2.94,2.94) [0.00,4.00]
Total: 5500 W: 1161 L: 1234 D: 3105
sprt @ 10+0.1 th 1 Reduce and trust pawns even more
19-07-31 xot searchstats1 diff
LLR: 0.57 (-2.94,2.94) [0.00,4.00]
Total: 21250 W: 4359 L: 4249 D: 12642
sprt @ 20+0.2 th 1 Information only. Does the big search param tune help at 20+0.2? See discussion at https://github.com/official-stockfish/Stockfish/pull/2260 -1 prio to run when fishtest is quiet
19-07-31 Roc ROFControl diff
LLR: -0.88 (-2.94,2.94) [0.50,4.50]
Total: 9825 W: 2195 L: 2195 D: 5435
sprt @ 10+0.1 th 1 R_1 Bonus for restricting rook access to fully open file. Since the rook may be already on open file, increase RookOnFile[1] bonus accordingly to compensate.
19-07-31 Voy SE-CMH diff
LLR: -1.04 (-2.94,2.94) [0.00,4.00]
Total: 1936 W: 295 L: 343 D: 1298
sprt @ 60+0.6 th 1 One more SE LTC test... (Place the other at -1) Low TP

Active - 135 machines 931 cores 1.28M nps (1195.20M total nps) 1590 games/minute

Machine Cores MNps System Version Running on Last updated
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 queen_distance seconds ago
TueRens
7
1.30 Linux 4.15.0-58-generic 69:2 queen_distance seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 queen_distance seconds ago
ChessDBCN
8
1.18 Linux 3.10.0-957.21.3.el7.x86_64 69:2 queen_distance seconds ago
0xB00B1ES 2 1.85 Windows 10 69:2 queen_distance seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 tweak_singular seconds ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 tweak_singular seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 tweak_singular seconds ago
ChessDBCN
8
1.53 Linux 3.10.0-957.21.3.el7.x86_64 69:2 tweak_singular seconds ago
ChessDBCN
8
0.00 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
leszek 7 0.92 Windows 7 69:2 only_pawns seconds ago
mhoram 1 2.16 Linux 4.19.0-5-amd64 69:2 only_pawns seconds ago
mhoram
1
1.01 Linux 4.9.0-9-amd64 69:2 only_pawns seconds ago
rpngn 1 1.86 Windows 7 69:2 only_pawns seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.13 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
0.00 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns 30 minutes ago
ChessDBCN
8
0.00 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
0.00 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.29 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.38 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.23 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
cuistot 2 1.43 Windows 7 69:2 only_pawns seconds ago
velislav 3 1.96 Linux 5.0.0-25-generic 69:2 only_pawns seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.29 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 only_pawns seconds ago
nordlandia 1 0.81 Windows 10 69:2 only_pawns seconds ago
dew
3
0.00 Linux 3.12.13-gentoo 69:3 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
gvreuls
4
0.87 Linux 5.2.9-200.fc30.x86_64 69:3 quanta seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
robal
2
2.12 Windows 10 69:2 quanta seconds ago
TueRens 3 2.15 Linux 5.0.0-23-generic 69:2 quanta seconds ago
racerschmacer
1
1.63 Windows 10 69:2 quanta seconds ago
fkraiem 1 0.93 Linux 5.0.0-25-generic 69:2 quanta seconds ago
ChessDBCN
8
1.20 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
TueRens
3
1.92 Windows 7 69:2 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 quanta seconds ago
dew 2 0.90 Windows 2003Server 69:2 quanta seconds ago
racerschmacer
1
1.39 Windows 10 69:2 quanta seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst3 seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.27.2.el7.x86_64 69:2 searchconst3 seconds ago
nordlandia 7 2.37 Windows 10 69:2 searchconst3 seconds ago
ChessDBCN
8
1.29 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst3 seconds ago
ChessDBCN
8
1.38 Linux 3.10.0-957.21.3.el7.x86_64 69:2 contdyn2 seconds ago
ChessDBCN
8
1.08 Linux 3.10.0-957.21.3.el7.x86_64 69:2 contdyn2 seconds ago
robal
5
1.80 Darwin 18.6.0 69:2 contdyn2 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 contdyn2 seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 contdyn2 seconds ago
ChessDBCN
8
1.21 Linux 3.10.0-957.21.3.el7.x86_64 69:2 contdyn2 seconds ago
dew
3
1.02 Linux 3.17.7-hardened-r1 69:3 contdyn2 seconds ago
torbjo 2 1.00 Linux 5.1.11-x86_64-linode127 69:3 contdyn2 seconds ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm4 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm4 seconds ago
ChessDBCN
8
1.29 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm4 seconds ago
dew
6
1.15 Linux 3.17.7-hardened-r1 69:3 psq_npm4 seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm4 seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm4 seconds ago
thijsk 4 1.42 Linux 4.15.18-14-pve 69:2 psq_npm4 seconds ago
sunu 2 1.13 Linux 4.20.0-0.slh.1-aptosid-amd64 69:2 psq_npm4 seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.21.3.el7.x86_64 69:2 evalT seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 evalT seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 evalT seconds ago
AdrianSchmidt1990 4 2.43 Windows 10 69:2 evalT seconds ago
dew
1
0.74 Windows 2003Server 69:2 evalT seconds ago
dew
6
1.03 Linux 4.0.5-gentoo 69:3 lmr_bestmove seconds ago
gvreuls 16 1.54 Linux 5.2.9-200.fc30.x86_64 69:3 lmr_bestmove seconds ago
ChessDBCN
8
1.27 Linux 3.10.0-957.21.3.el7.x86_64 69:2 lmr_bestmove seconds ago
mgrabiak 3 1.30 Windows 10 69:2 lmr_bestmove seconds ago
leszek 3 1.12 Windows 2003Server 69:2 lmr_bestmove seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.32 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.38 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.26 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 searchconst2 seconds ago
manap 7 0.00 Windows 10 69:2 psq_npm3 2 minutes ago
ChessDBCN
8
1.52 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.51 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
tvijlbrief
7
1.88 Linux 4.15.0-59-generic 69:3 psq_npm3 seconds ago
dew
14
0.75 Linux 3.15.10-hardened-r1 69:3 psq_npm3 seconds ago
dew
20
0.82 Linux 4.19.57-gentoo 69:3 psq_npm3 seconds ago
nordlandia 3 2.46 Windows 10 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
TueRens
15
1.33 Linux 4.15.0-58-generic 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.18 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.35 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
bcross 15 1.20 Windows 10 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.32 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
grandphish2
5
1.63 Windows 10 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.38 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.40 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.19 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.38 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.39 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
ChessDBCN
8
1.02 Linux 3.10.0-957.21.3.el7.x86_64 69:2 psq_npm3 seconds ago
davar
3
1.68 Windows 8.1 69:2 psq_npm3 seconds ago
Vizvezdenec 1 1.90 Windows 8.1 69:2 psq_npm3 seconds ago
nordlandia 1 1.40 Windows 10 69:2 psq_npm3 2 minutes ago
ChessDBCN
8
1.28 Linux 3.10.0-957.21.3.el7.x86_64 69:2 ps_connected32_tune seconds ago
dew
3
0.00 Linux 3.18.7-gentoo 69:3 ps_connected32_tune 18 minutes ago
ChessDBCN
8
1.31 Linux 3.10.0-957.21.3.el7.x86_64 69:2 ps_connected32_tune seconds ago
dew
5
1.02 Linux 3.18.9-hardened 69:3 ps_mobility22_tune seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.21.3.el7.x86_64 69:2 ps_mobility22_tune seconds ago
nordlandia 3 1.08 Windows 10 69:2 ps_mobility22_tune seconds ago
dew
14
0.82 Linux 4.8.17-hardened-r2 69:2 ps_mobility22_tune seconds ago
fastgm 4 1.24 Linux 4.4.0-18362-Microsoft 69:2 ps_mobility22_tune seconds ago
19-08-24 xot searchconst2 diff
LLR: 0.88 (-2.94,2.94) [0.50,4.50]
Total: 7631 W: 1191 L: 1118 D: 5322
sprt @ 20+0.2 th 8 The pure constant changes from the 2nd big search tune appear to only give a small gain, test how much gain the added fractions give.
19-08-24 MJZ psq_npm4 diff
LLR: 0.74 (-2.94,2.94) [0.50,4.50]
Total: 35138 W: 8070 L: 7851 D: 19217
sprt @ 10+0.1 th 1 Increase score in late EG. +1/2, NPM==0 and abs(v) > 100, extend PV until abs(alpha-staticEval) > 400
19-08-23 sg lmr_bestmove diff
LLR: 0.61 (-2.94,2.94) [0.50,4.50]
Total: 110561 W: 24975 L: 24389 D: 61197
sprt @ 10+0.1 th 1 Allow at root LMR if best move count <= 1 but use rootNode weight 2 in move count condition.
19-08-24 sni queen_distance diff
LLR: 0.36 (-2.94,2.94) [0.50,4.50]
Total: 20445 W: 4612 L: 4490 D: 11343
sprt @ 10+0.1 th 1 Linear formula (weight 2)
19-08-24 sni quanta diff
LLR: 0.24 (-2.94,2.94) [0.50,4.50]
Total: 22752 W: 5077 L: 4951 D: 12724
sprt @ 10+0.1 th 1 Round evals to multiples of 8
19-08-24 xor evalT diff
LLR: 0.13 (-2.94,2.94) [0.00,4.00]
Total: 40022 W: 8957 L: 8791 D: 22274
sprt @ 10+0.1 th 1 try stc first.
19-08-24 xot searchconst3 diff
LLR: -0.03 (-2.94,2.94) [0.50,4.50]
Total: 31744 W: 7058 L: 6903 D: 17783
sprt @ 10+0.1 th 1 Tweak reduction adjustment for statScore.
19-08-24 sni only_pawns diff
LLR: -0.18 (-2.94,2.94) [0.50,4.50]
Total: 22134 W: 4971 L: 4871 D: 12292
sprt @ 10+0.1 th 1 Tweak initiative() for pure pawn endgames
19-08-24 sni only_pawns diff
LLR: -0.59 (-2.94,2.94) [0.50,4.50]
Total: 21440 W: 4812 L: 4738 D: 11890
sprt @ 10+0.1 th 1 Tweak scale factor for pure pawn endgames
19-08-24 sni quanta diff
LLR: -0.73 (-2.94,2.94) [0.50,4.50]
Total: 22792 W: 5124 L: 5051 D: 12617
sprt @ 10+0.1 th 1 Round evals to multiples of 4
19-08-24 sni only_pawns diff
LLR: -0.80 (-2.94,2.94) [0.50,4.50]
Total: 21953 W: 4865 L: 4801 D: 12287
sprt @ 10+0.1 th 1 Tweak initiative() for pure pawn endgame (take 2)
19-08-24 sg tweak_singular diff
LLR: -0.84 (-2.94,2.94) [0.50,4.50]
Total: 21265 W: 4755 L: 4696 D: 11814
sprt @ 10+0.1 th 1 Decrease reduction in singular search if at cut node.
19-08-23 MJZ psq_npm3 diff
LLR: -0.91 (-2.94,2.94) [0.00,3.50]
Total: 112845 W: 19050 L: 18787 D: 75008
sprt @ 60+0.6 th 1 Increase score in late EG. +1/2, NPM==0 and abs(v) > 100 - LTC
19-08-24 sni only_pawns diff
LLR: -0.96 (-2.94,2.94) [0.50,4.50]
Total: 21063 W: 4692 L: 4641 D: 11730
sprt @ 10+0.1 th 1 Tweak scale factor for pure pawn endgames (take 2)
19-08-24 xot contdyn2 diff
LLR: -2.14 (-2.94,2.94) [0.50,4.50]
Total: 31722 W: 7132 L: 7093 D: 17497
sprt @ 10+0.1 th 1 Different dct gradient depending if winning or losing.
19-08-16 pro ps_mobility22_tune diff
48368/100000 iterations
100776/200000 games played
200000 @ 60+0.6 th 1 big tune linear mobility equations.
19-08-24 pro ps_connected32_tune diff
12064/50000 iterations
25153/100000 games played
100000 @ 20+0.2 th 1 tune scores for connected array.

Finished - 54342 tests

19-08-24 sg lmr_delayed diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 30616 W: 6849 L: 6861 D: 16906
sprt @ 10+0.1 th 1 ttPv reduction: -2 ply directly, -1 ply delayed.
19-08-24 sg ks_attacks diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 17432 W: 3838 L: 3916 D: 9678
sprt @ 10+0.1 th 1 safe queen king attacks. Bonus 130.
19-08-25 Viz futPrCaptures3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 2789 W: 565 L: 717 D: 1507
sprt @ 10+0.1 th 1 Take 2
19-08-24 MJZ TTsave-bestMove diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 15393 W: 3389 L: 3477 D: 8527
sprt @ 10+0.1 th 1 Skip TT save if no bestMove and TTdepth > depth
19-08-24 sni queen_distance^ diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 14864 W: 3272 L: 3363 D: 8229
sprt @ 10+0.1 th 1 Revert to linear formula (weight 3)
19-08-24 Viz futPruningPNtweak1 diff
LLR: -2.32 (-2.94,2.94) [0.50,4.50]
Total: 25901 W: 5729 L: 5730 D: 14442
sprt @ 10+0.1 th 1 Take 1
19-08-25 Viz futPrCaptures2 diff
LLR: -2.64 (-2.94,2.94) [0.50,4.50]
Total: 5118 W: 1076 L: 1198 D: 2844
sprt @ 10+0.1 th 1 Take 1
19-08-24 sg tweak_singular diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12812 W: 2784 L: 2885 D: 7143
sprt @ 10+0.1 th 1 Increase singular search depth at cut nodes.
19-08-24 sg lmr_delayed diff
LLR: -2.94 (-2.94,2.94) [0.00,3.50]
Total: 32493 W: 5443 L: 5527 D: 21523
sprt @ 60+0.6 th 1 LTC: cut node reduction: 2 ply drectly, -1 ply delayed.
19-08-24 pro ps_captures1 diff
LLR: -3.88 (-2.94,2.94) [0.50,4.50]
Total: 31583 W: 6959 L: 7019 D: 17605
sprt @ 10+0.1 th 1 scale down scoring of captures. 3, /2
19-08-24 MJZ psq_npm5 diff
LLR: 0.18 (-2.94,2.94) [0.50,4.50]
Total: 14167 W: 3171 L: 3091 D: 7905
sprt @ 10+0.1 th 1 Increase score in late EG. +1/2, NPM==0 and abs(v) > 100, extend PV until abs(alpha-staticEval) > 200
19-08-24 sni only_pawns diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 2227 W: 454 L: 611 D: 1162
sprt @ 10+0.1 th 1 Add 50% to passed pawn endgame value
19-08-24 Viz badCaptureLMR1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 17359 W: 3812 L: 3891 D: 9656
sprt @ 10+0.1 th 1 Take 1
19-08-24 Voy nmpEG2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 23995 W: 5341 L: 5386 D: 13268
sprt @ 10+0.1 th 1 stc
19-08-24 sg lmr_delayed2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 10078 W: 2206 L: 2321 D: 5551
sprt @ 10+0.1 th 1 My LTC struggles so try a rewrite by not using search stack but instead a function parameter for delayed reductions to avoid some interactions (bench changes). Cut node reduction: +2 ply directly, -1 ply delayed.
19-08-24 MJZ psq_npm diff
LLR: -0.96 (-2.94,2.94) [0.50,4.50]
Total: 12086 W: 2709 L: 2702 D: 6675
sprt @ 10+0.1 th 1 Increase score in late EG. +v.abs(v)/256, NPM==0 and abs(v) > 80
19-08-24 sg lmr_delayed diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 14286 W: 3184 L: 3278 D: 7824
sprt @ 10+0.1 th 1 ttPv reduction: -2 ply directly, +1 ply delayed.
19-08-24 Sop mob diff
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 11388 W: 2553 L: 2661 D: 6174
sprt @ 10+0.1 th 1 fix ub
19-08-23 xor lastPiece diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 78090 W: 12969 L: 12927 D: 52194
sprt @ 60+0.6 th 1 LTC: Do not prune conversion to endgame.
19-08-23 xot vary4pr diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 11676 W: 1841 L: 2013 D: 7822
sprt @ 60+0.6 th 1 Remove added complexity, just pure param changes. Test for non-regression on 1 thread. Not sure what correct throughput is since 8 thread test hasn't passed yet. Running this now since a fail here makes the cpu intensive 8 thread test pointless. (Feel free to set the 8 thread to prio -1 for now if you think that is appropriate until this test finishes.)
19-08-24 Viz seePruningEndgame3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 19207 W: 4264 L: 4333 D: 10610
sprt @ 10+0.1 th 1 Take 2
19-08-24 sg lmr_delayed diff
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 18709 W: 4315 L: 4056 D: 10338
sprt @ 10+0.1 th 1 cut node reduction: 2 ply drectly, -1 ply delayed.
19-08-24 Sop mob diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 3638 W: 755 L: 903 D: 1980
sprt @ 10+0.1 th 1 exclude squares with our pieces on when counting mobility
19-08-24 Viz seePruningEndgame2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 24606 W: 5540 L: 5582 D: 13484
sprt @ 10+0.1 th 1 fixed version
19-08-23 Viz negCapHistLMR3 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 49179 W: 10938 L: 10859 D: 27382
sprt @ 10+0.1 th 1 Take 3
19-08-24 sni queen_distance diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 24773 W: 5520 L: 5561 D: 13692
sprt @ 10+0.1 th 1 Scale by number of pawns (divisor 4)
19-08-24 sg lmr_delayed diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 7939 W: 1705 L: 1830 D: 4404
sprt @ 10+0.1 th 1 Experiment: use the negative of current reduction as delayed reduction. So the current reduction will be somehow cancled out at child node LMR:
19-08-24 sni queen_distance diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 19118 W: 4276 L: 4346 D: 10496
sprt @ 10+0.1 th 1 Scale by number of pawns (divisor 3)
19-08-24 Viz captureNumberLMR1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 8551 W: 1893 L: 2016 D: 4642
sprt @ 10+0.1 th 1 Take 1
19-08-24 Viz lmrStatBonusTweak2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 7660 W: 1638 L: 1765 D: 4257
sprt @ 10+0.1 th 1 Take 2
19-08-24 sg lmr_delayed diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 6131 W: 1329 L: 1464 D: 3338
sprt @ 10+0.1 th 1 Also clear delayed reductions for full depth search.
19-08-24 Viz zeroMcLMR1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 4851 W: 1009 L: 1150 D: 2692
sprt @ 10+0.1 th 1 Take 1
19-08-24 Viz seePruningEndgame1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 236 W: 8 L: 177 D: 51
sprt @ 10+0.1 th 1 Variation of xoroshiro idea
19-08-24 Viz lmrStatBonusTweak1 diff
LLR: -3.07 (-2.94,2.94) [0.50,4.50]
Total: 6621 W: 1425 L: 1564 D: 3632
sprt @ 10+0.1 th 1 Take 1
19-08-24 Viz cutNodeExtBound2 diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 12864 W: 2036 L: 2176 D: 8652
sprt @ 60+0.6 th 1 des
19-08-24 sni queen_distance^ diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 14618 W: 3216 L: 3308 D: 8094
sprt @ 10+0.1 th 1 Scale by number of pawns (divisor 2)
19-08-23 sni shelter_signature diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 65864 W: 14630 L: 14596 D: 36638
sprt @ 10+0.1 th 1 test a version of PR #2277 for no-regression
19-08-23 MJZ QueenBlockers diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 30051 W: 6732 L: 6747 D: 16572
sprt @ 10+0.1 th 1 Try to improve the evaluation queen blockers : separate pin and discovered attacks. MG 55 / 44 / 52
19-08-23 sg lmr_bestmove diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 60829 W: 13518 L: 13381 D: 33930
sprt @ 10+0.1 th 1 Allow at root LMR if best move count <= 1 but use old move count condition.
19-08-23 Viz cutNodeExtBound2 diff
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 20171 W: 4559 L: 4295 D: 11317
sprt @ 10+0.1 th 1 Take 2
19-08-23 Voy nmpEG diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 48321 W: 10756 L: 10681 D: 26884
sprt @ 10+0.1 th 1 stc
19-08-24 Fis WeakLeverSTM diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 11966 W: 2629 L: 2734 D: 6603
sprt @ 10+0.1 th 1 Try single attacks
19-08-18 xor ltcTune diff
29938/31250 iterations
62500/62500 games played
62500 @ 60+0.6 th 1 LTC: for comparison: same number of params, low number of games but LTC (low TP).
19-08-19 MJZ Alpha-ReductionB diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 158086 W: 26528 L: 26261 D: 105297
sprt @ 60+0.6 th 1 LTC of http://tests.stockfishchess.org/tests/view/5d5a96be0ebc5925cf110841
19-08-23 sg lmr_delayed diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 23593 W: 5180 L: 5228 D: 13185
sprt @ 10+0.1 th 1 Introduce delayed reductions which allow to delay a reduction to the child node. First split cut node reduction: one ply directly and one ply delayed.
19-08-23 sni queen_distance diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 42203 W: 9459 L: 9413 D: 23331
sprt @ 10+0.1 th 1 Try now a linear formula
19-08-23 Viz cmhPdepthCutNode4 diff
LLR: -2.31 (-2.94,2.94) [0.50,4.50]
Total: 18893 W: 3107 L: 3152 D: 12634
sprt @ 60+0.6 th 1 STC with depth threshold reduction passed twice and both LTCs seems to be failing fast - try LTC for this version which increases threshold. 1/3 TP
19-08-23 Viz futMcCutNode3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 3745 W: 783 L: 930 D: 2032
sprt @ 10+0.1 th 1 Take 3
19-08-23 Viz futMcCutNode2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 2492 W: 474 L: 626 D: 1392
sprt @ 10+0.1 th 1 Take 2
19-08-23 sg ks_attacks diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 18079 W: 4002 L: 4077 D: 10000
sprt @ 10+0.1 th 1 Consider safe queen attacks on the king adjacent squares for king danger. Bonus 200.