Stockfish Testing Queue

Pending - 5 tests 0.0 hrs

19-01-19 SFi tune_assorted_groupA diff
11307/20000 iterations
23500/40000 games played
40000 @ 60+0.6 th 1 Tune some bonus/penalty in assorted. Low TP.
19-01-19 SFi tune_eval_param diff
27979/40000 iterations
58258/80000 games played
80000 @ 60+0.6 th 1 Tune Eval Parameters. Low TP.
19-01-22 pro ps_threads40 diff
LLR: -0.45 (-2.94,2.94) [-3.00,1.00]
Total: 17500 W: 3381 L: 3437 D: 10682
sprt @ 5+0.05 th 8 Linearly distribute thread skipping scheme according to # of threads (max group size of 4).
19-01-20 vdv threadPick diff
LLR: -1.42 (-2.94,2.94) [0.50,4.50]
Total: 20750 W: 3963 L: 3947 D: 12840
sprt @ 5+0.05 th 8 Voting scheme tweak.
19-01-23 Ala MultiImbalanceTune diff
1350/75000 iterations
2792/150000 games played
150000 @ 20+0.2 th 1 Starting from smoothed values of previous tune, continue the tuning with better adapted range for variables, smaller variance (the previous one was great to get a quick approximation starting from all 0 but was too sensitive to converge). Also, as the subcases in 3Mv2R and 3MvQ proved important, add subcases for 2RvQ and MRvQ

Active - 141 machines 971 cores 1.41M nps (1371.29M total nps) 1835 games/minute

Machine Cores MNps System Version Running on Last updated
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
dew
18
1.47 Linux 3.12.13-gentoo 67:3 TrappedRookFile2 seconds ago
Bobo1239 7 1.72 Linux 4.20.3-arch1-1-ARCH 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.43 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.56 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.57 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
gvreuls
3
1.78 Windows 10 67:3 TrappedRookFile2 seconds ago
ChessDBCN
8
1.54 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.33 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
mhoram 3 0.71 Linux 4.9.0-8-amd64 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.1.3.el7.x86_64 67:2 TrappedRookFile2 seconds ago
mgrabiak 3 1.54 Windows 10 67:2 TrappedRookFile2 seconds ago
robal
2
1.92 Windows 10 67:2 TrappedRookFile2 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.56 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
jbwiebe 3 1.97 Linux 4.9.0-8-amd64 67:2 prune3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
Vizvezdenec
6
2.16 Windows 10 67:2 prune3 seconds ago
ChessDBCN
8
1.58 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.56 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.31 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 prune3 seconds ago
mhacleth 7 1.54 Linux 4.4.0-141-generic 67:2 prune3 seconds ago
dew
5
1.09 Linux 3.18.9-hardened 67:3 prune3 seconds ago
leszek 3 1.18 Windows 2003Server 67:2 prune3 seconds ago
dew
6
1.11 Linux 3.8.13-gentoo 67:2 prune3 seconds ago
ChessDBCN
8
1.42 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
Fisherman 32 1.46 Windows 10 67:2 bish_block seconds ago
MaZePallas 4 2.38 Linux 4.19.13-300.fc29.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.31 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.55 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
ChessDBCN
8
1.57 Linux 3.10.0-957.1.3.el7.x86_64 67:2 bish_block seconds ago
mgrabiak 3 1.65 Windows 7 67:2 bish_block seconds ago
ChessDBCN
8
0.00 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.32 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
nordlandia 7 2.51 Windows 10 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.36 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.56 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
horst.prack 1 1.14 Linux 4.19.0-1-amd64 67:2 trapped_rook3 seconds ago
dew
3
1.22 Linux 3.18.7-gentoo 67:3 trapped_rook3 58 minutes ago
leszek 1 1.17 Windows 2003Server 67:2 trapped_rook3 seconds ago
gvreuls
15
2.93 Linux 4.20.3-200.fc29.x86_64 67:3 trapped_rook3 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
jmdana
3
2.16 Linux 4.9.0-8-amd64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
dew 5 1.09 Linux 4.9.0-8-amd64 67:2 trapped_rook3 seconds ago
davar
3
1.78 Windows 8.1 67:2 trapped_rook3 seconds ago
dew
1
0.76 Windows 2003Server 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.47 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.57 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
TueRens 3 2.26 Windows 10 67:2 trapped_rook3 seconds ago
ChessDBCN
8
1.32 Linux 3.10.0-957.1.3.el7.x86_64 67:2 trapped_rook3 seconds ago
TueRens
3
2.05 Windows 7 67:2 trapped_rook3 seconds ago
ctoks 3 2.09 Windows 10 67:3 trapped_rook3 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest1 seconds ago
ChessDBCN
8
1.43 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest1 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest1 seconds ago
MaZePallas 4 2.54 Linux 4.19.13-300.fc29.x86_64 67:2 percentageRest1 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest1 seconds ago
ChessDBCN
8
1.32 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest2 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest2 seconds ago
robal
5
1.87 Darwin 18.2.0 67:2 percentageRest2 seconds ago
ChessDBCN
8
1.33 Linux 3.10.0-957.1.3.el7.x86_64 67:2 percentageRest2 seconds ago
leszek 7 0.87 Windows 7 67:2 percentageRest2 seconds ago
dew
6
1.10 Linux 4.0.5-gentoo 67:3 percentageRest2 seconds ago
mhoram
1
1.10 Linux 4.9.0-8-amd64 67:2 ps_moveimportance1_tune 2 minutes ago
dew
3
1.09 Linux 4.9.76-gentoo-r1 67:2 Blocked_posG4 9 minutes ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 Blocked_posG4 3 minutes ago
MaZePallas 6 2.59 Linux 4.19.13-300.fc29.x86_64 67:2 Blocked_posG4 seconds ago
mhacleth 3 1.95 Darwin 18.0.0 67:2 Blocked_posG4 seconds ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 Blocked_posG4 seconds ago
jromang 3 2.13 Linux 4.19.13-1-MANJARO 67:3 Blocked_posG4 3 minutes ago
ChessDBCN
8
1.34 Linux 3.10.0-957.1.3.el7.x86_64 67:2 Blocked_posG4 seconds ago
dew
14
0.83 Linux 4.8.17-hardened-r2 67:2 Blocked_posG4 seconds ago
gvreuls
4
1.69 Linux 4.20.3-200.fc29.x86_64 67:3 Blocked_posG4 9 minutes ago
dew
3
1.14 Linux 3.12.13-gentoo 67:3 Blocked_posG4 seconds ago
dew
6
0.86 Linux 3.17.7-hardened-r1 67:2 Blocked_posG4 2 minutes ago
horst.prack 1 1.13 Linux 4.19.0-1-amd64 67:2 Blocked_posG4 seconds ago
dew
6
0.94 Linux 4.0.5-gentoo 67:2 Blocked_posG4 7 minutes ago
pb00067
2
2.16 Windows 10 67:2 Blocked_posG4 5 minutes ago
nordlandia 1 0.89 Windows 10 67:2 ps_moveimportance1 2 minutes ago
CSU_Dynasty
31
0.00 Windows 10 67:2 singular11 seconds ago
cuistot 2 0.00 Windows 7 67:2 singular11 seconds ago
vdv
31
1.35 Linux 4.15.0-43-generic 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.35 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
nordlandia 1 1.57 Windows 10 67:2 singular11 2 minutes ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
mhacleth 7 0.94 Linux 4.15.0-1026-gcp 67:2 singular11 2 minutes ago
tvijlbrief
7
2.08 Linux 4.15.0-44-generic 67:3 singular11 seconds ago
dew
3
1.22 Linux 3.18.7-gentoo 67:3 singular11 11 minutes ago
ChessDBCN
8
1.35 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
xoto
12
1.32 Linux 4.13.0-38-generic 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.43 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.58 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.46 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.56 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.42 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 singular11 seconds ago
mhoram 3 2.06 Linux 4.9.0-8-amd64 67:2 singular11 seconds ago
Vizvezdenec 3 1.77 Windows 10 67:2 singular11 seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.30 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.45 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
ChessDBCN
8
1.44 Linux 3.10.0-957.1.3.el7.x86_64 67:2 noskip seconds ago
dew
3
1.06 Linux 3.17.7-hardened-r1 67:2 Blocked_posG5 3 minutes ago
dew
1
1.10 Linux 4.9.76-gentoo-r1 67:2 Blocked_posG5 9 minutes ago
dew 3 0.94 Windows 2003Server 67:2 Blocked_posG5 13 minutes ago
MaZePallas 2 1.62 Linux 4.19.13-300.fc29.x86_64 67:2 Blocked_posG5 seconds ago
0xB00B1ES 3 0.88 Windows 8.1 67:2 Blocked_posG5 seconds ago
nordlandia 1 0.88 Windows 10 67:2 Blocked_posG5 23 minutes ago
racerschmacer
1
1.70 Windows 10 67:2 Blocked_posG5 6 minutes ago
dew
3
1.24 Linux 2.6.36-gentoo-r5 67:2 Blocked_posG5 2 minutes ago
ali-al-zhrani 2 1.73 Windows 10 67:2 Blocked_posG5 seconds ago
19-01-23 sni trapped_rook3 diff
LLR: 1.16 (-2.94,2.94) [0.00,4.00]
Total: 10960 W: 2482 L: 2374 D: 6104
sprt @ 10+0.1 th 1 Stronger penalty for non-castling case
19-01-20 Cof noskip diff
LLR: 0.27 (-2.94,2.94) [-3.00,1.00]
Total: 20744 W: 3274 L: 3292 D: 14178
sprt @ 20+0.2 th 8 LTC 20+0.2 8 thread for removing thread skipping scheme. No functional change for 1 thread. Well, 5+0.05 test failed quickly, skipping scheme seems to be ~4 Elo at 5+0.05 8 threads. Reason for this test is, that the skipping scheme does NOT scale well with time, so I expect between 0 and -2 Elo. In my local tests (Amd FX CPU) it was 0 Elo, I would like to see it run on fishtest noob hardware. I will limit max games to 20000, to limit the resource usage. If I am wrong that scaling of the skipping scheme is bad, this test will fail quickly (like 5+0.05 run before).
19-01-23 Viz TrappedRookFile2 diff
LLR: 0.08 (-2.94,2.94) [0.50,4.50]
Total: 8570 W: 1932 L: 1885 D: 4753
sprt @ 10+0.1 th 1 Take 1
19-01-23 Elb bish_block diff
LLR: -0.37 (-2.94,2.94) [0.50,4.50]
Total: 10688 W: 2362 L: 2330 D: 5996
sprt @ 10+0.1 th 1 Bonus for pawn move that reduces bishop mobility.
19-01-22 MJZ Blocked_posG4 diff
LLR: -0.53 (-2.94,2.94) [0.50,4.50]
Total: 7807 W: 1085 L: 1085 D: 5637
sprt @ 240+2.4 th 1 Blocked position. One last try with reduction for reversible moves in potentially blocked position. Limited test at VLTC with low throughput (test needs high depth search).
19-01-23 Viz percentageRest1 diff
LLR: -0.54 (-2.94,2.94) [0.50,4.50]
Total: 14200 W: 3143 L: 3103 D: 7954
sprt @ 10+0.1 th 1 Take 1
19-01-20 MJZ Blocked_posG5 diff
LLR: -0.57 (-2.94,2.94) [0.50,4.50]
Total: 17414 W: 2415 L: 2384 D: 12615
sprt @ 240+2.4 th 1 Blocked position, it seems that the problem come from poskey in Qsearch. Third try at higher depths (d > 28). Limited test at VLTC with low throughput (test needs high depth search).
19-01-22 sni singular11 diff
LLR: -0.84 (-2.94,2.94) [0.00,3.50]
Total: 33713 W: 5604 L: 5560 D: 22549
sprt @ 60+0.6 th 1 Speculative LTC: Make singular margin a function of (d/2)
19-01-22 pro ps_moveimportance1 diff
LLR: -1.28 (-2.94,2.94) [-3.00,1.00]
Total: 22148 W: 3559 L: 3660 D: 14929
sprt @ 60+0.6 th 1 try a simple sigmoid for move importance.
19-01-23 Elb bish_block diff
LLR: -1.35 (-2.94,2.94) [0.50,4.50]
Total: 9742 W: 2124 L: 2151 D: 5467
sprt @ 10+0.1 th 1 Take 2
19-01-23 sni trapped_rook3 diff
LLR: -1.39 (-2.94,2.94) [0.00,4.00]
Total: 10210 W: 2214 L: 2251 D: 5745
sprt @ 10+0.1 th 1 TrappedRook = S(47,20)
19-01-23 Viz percentageRest2 diff
LLR: -1.65 (-2.94,2.94) [0.50,4.50]
Total: 14167 W: 3114 L: 3136 D: 7917
sprt @ 10+0.1 th 1 Take 2
19-01-23 sni trapped_rook3 diff
LLR: -1.66 (-2.94,2.94) [0.00,4.00]
Total: 11189 W: 2438 L: 2486 D: 6265
sprt @ 10+0.1 th 1 stronger default for TrappedRook
19-01-23 xor prune3 diff
LLR: -2.12 (-2.94,2.94) [0.50,4.50]
Total: 8730 W: 1878 L: 1953 D: 4899
sprt @ 10+0.1 th 1 take 1.
19-01-22 pro ps_moveimportance1_tune diff
11223/40000 iterations
23993/80000 games played
80000 @ 60+0.6 th 1 Tune a sigmoid equation for move_importance.

Finished - 42781 tests

19-01-23 Fis KnightVSBishop diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 7288 W: 1546 L: 1675 D: 4067
sprt @ 10+0.1 th 1 Knight vs Bishop
19-01-22 MJZ Blocked_posG7 diff
LLR: -0.98 (-2.94,2.94) [0.50,4.50]
Total: 9232 W: 1383 L: 1399 D: 6450
sprt @ 120+1.2 th 1 Blocked position. Try n°8 with progressive decrease from ply 24 (+remove depth limit). Limited test at VLTC with low throughput (test needs high depth search).
19-01-22 Viz BlockedPosEval1 diff
LLR: -1.36 (-2.94,2.94) [0.50,4.50]
Total: 18499 W: 3097 L: 3094 D: 12308
sprt @ 60+0.6 th 1 I was unable to come up with anything that changes bench really for this patch... At least for now. Let me check how it scales. 1/6 TP and STC SPRT bounds for it being faster in terms of converging.
19-01-20 sni singular_tuning4 diff
16033/20000 iterations
33521/40000 games played
40000 @ 60+0.6 th 1 Tune singular margin, starting from a flat 40. Tuning with A=500, to be compared to the previous run "singular_tuning2" at http://tests.stockfishchess.org/tests/view/5c409a820ebc5902bb5d029a
19-01-20 sni singular_tuning3 diff
16187/20000 iterations
33713/40000 games played
40000 @ 60+0.6 th 1 Tune singular margin array, but with a smaller A=500. To be compared to the previous tuning run "singular_run" at http://tests.stockfishchess.org/tests/view/5c40996b0ebc5902bb5d0282
19-01-23 Ala MultiImbalanceTune diff
30534/75000 iterations
63891/150000 games played
150000 @ 20+0.2 th 1 The quadratic imbalance handles how two piece type interact but not specific imbalances. There are some imbalance scoring patch attempt. This tune is for a much more generalized version. Starting values at 0 with high variance to hopefully not get stuck.
19-01-23 pro ps_pawnattacks103 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 70873 W: 15418 L: 15722 D: 39733
sprt @ 10+0.1 th 1 remove pawnAttacks array and use pawn_attacks_bb.
19-01-23 31m backward1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 65685 W: 14467 L: 14308 D: 36910
sprt @ 10+0.1 th 1 Half Backward penalty if there is only one stopper (about 10% of bench positions).
19-01-23 Viz FullyBlockedShelter1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 29190 W: 6559 L: 6578 D: 16053
sprt @ 10+0.1 th 1 Take 1
19-01-23 Viz UselessEGBishop2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 25380 W: 5586 L: 5625 D: 14169
sprt @ 10+0.1 th 1 Take 2
19-01-23 Viz UselessEGBishop3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 22479 W: 4924 L: 4977 D: 12578
sprt @ 10+0.1 th 1 Take 3
19-01-23 Viz FullyBlockedShelter2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 17693 W: 3852 L: 3929 D: 9912
sprt @ 10+0.1 th 1 Take 2
19-01-23 31m TrappedRook_castling3 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 40996 W: 9080 L: 9042 D: 22874
sprt @ 10+0.1 th 1 I generally try to avoid respins of old tests, but there's a fairly obvious, large (multiplicative) interaction between this patch and the recent TrappedRook simplification. Therefore, respin this 60K yellow.
19-01-23 Jek rook_semiopen diff
LLR: -0.13 (-2.94,2.94) [0.50,4.50]
Total: 1092 W: 240 L: 242 D: 610
sprt @ 10+0.1 th 1 rook_semiopen
19-01-23 31m backward1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 7265 W: 1538 L: 1667 D: 4060
sprt @ 10+0.1 th 1 First attempt looks initially promising. Double effect: no Backward penalty at all if there is only one stopper (but still possibly increment weakUnopposed).
19-01-23 Fis RookVSMinor diff
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 23570 W: 5240 L: 5287 D: 13043
sprt @ 10+0.1 th 1 Roook vs Minor
19-01-19 Viz enPassant2 diff
LLR: -1.89 (-2.94,2.94) [0.00,3.50]
Total: 162972 W: 27207 L: 26865 D: 108900
sprt @ 60+0.6 th 1 I really rarely go for TCs but I would like to see how this goes... 1/10 TP and this is all really logical stuff.
19-01-23 Viz UselessEGBishop1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 9300 W: 1980 L: 2098 D: 5222
sprt @ 10+0.1 th 1 Take 1
19-01-23 Fis RooksVSMinors diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 41590 W: 9207 L: 9166 D: 23217
sprt @ 10+0.1 th 1 2 rooks vs 3 minors
19-01-22 pro ps_comboarrays1 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 44745 W: 9780 L: 9704 D: 25261
sprt @ 10+0.1 th 1 remove 4 arrays (FileBB, RankBB, PawnAttacksSpan, Mask).
19-01-22 pro ps_pawnarrays1 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 42267 W: 9321 L: 9240 D: 23706
sprt @ 10+0.1 th 1 remove some pawn arrays, check performance.
19-01-22 Viz BlockedPosEval4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 31704 W: 6963 L: 6971 D: 17770
sprt @ 10+0.1 th 1 try to return some material value + ignore also psqt score.
19-01-22 pro ps_forwardfile1 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 34324 W: 7498 L: 7401 D: 19425
sprt @ 10+0.1 th 1 Remove ForwardFileBB arrays.
19-01-22 sni master diff
ELO: 13.98 +-1.8 (95%) LOS: 100.0%
Total: 40000 W: 6323 L: 4714 D: 28963
40000 @ 60+0.6 th 1 Regression/progression test against SF10 after "Simplify TrappedRook" of January, 22th
19-01-22 Ala KRBKRB3 diff
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 14373 W: 3104 L: 3197 D: 8072
sprt @ 10+0.1 th 1 Take 3 : try a quadratic bonus
19-01-22 pro ps_comboarrays1 diff
LLR: -0.07 (-2.94,2.94) [-3.00,1.00]
Total: 3118 W: 694 L: 704 D: 1720
sprt @ 10+0.1 th 1 remove 4 arrays (FileBB, RankBB, PawnAttacksSpan, Mask).
19-01-22 Ala KRBKRB2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 16345 W: 3541 L: 3625 D: 9179
sprt @ 10+0.1 th 1 Take 2 : double bonus
19-01-22 Ala KRBKRB diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 19407 W: 4185 L: 4254 D: 10968
sprt @ 10+0.1 th 1 Take 1 : bonus for king closer to pawns in rook+bishop endgames, scaled with distance and pawn relative rank
19-01-21 vdv threadDist5 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14558 W: 2859 L: 2726 D: 8973
sprt @ 5+0.05 th 8 small tweak. (8 threads)
19-01-22 31m pawns_passable~10 diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 12887 W: 2058 L: 2198 D: 8631
sprt @ 60+0.6 th 1 Speculative LTC for a nearly 84K yellow, because it is an endgame-only bonus and therefore more effective at LTC than STC. Low throughput (166), at least to start.
19-01-22 sni singular11 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 80172 W: 17638 L: 17408 D: 45126
sprt @ 10+0.1 th 1 Tighter array values
19-01-22 Viz BlockedPosEval3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 39197 W: 8631 L: 8602 D: 21964
sprt @ 10+0.1 th 1 require material to be equal
19-01-22 Viz BlockedPosEval1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 62112 W: 13596 L: 13455 D: 35061
sprt @ 10+0.1 th 1 First and the least sophisticated attempt to correctly evaluate blocked positions.
19-01-22 xor tt_probcut diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 36923 W: 8097 L: 8116 D: 20710
sprt @ 10+0.1 th 1 promotion in probcut.
19-01-21 xot pawns2b diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 22647 W: 4939 L: 5014 D: 12694
sprt @ 10+0.1 th 1 tuned pawn arrays.
19-01-22 MJZ Blocked_posG6 diff
LLR: -0.67 (-2.94,2.94) [0.50,4.50]
Total: 3868 W: 557 L: 577 D: 2734
sprt @ 120+1.2 th 1 Blocked position. Try n°7 with progressive decrease from ply 20 (+remove extension). Limited test at VLTC with low throughput (test needs high depth search).
19-01-22 Fis sortTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 21441 W: 4617 L: 4697 D: 12127
sprt @ 10+0.1 th 1 Tweak sorting threshold
19-01-22 Viz RBNimb1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 16646 W: 3657 L: 3739 D: 9250
sprt @ 10+0.1 th 1 take 1
19-01-22 Viz BlockedPosEval2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 9540 W: 2053 L: 2170 D: 5317
sprt @ 10+0.1 th 1 To my big surprise this doesn't fail fast, so is probably promising, try a more broad applying of this.
19-01-16 vdv threadDist2 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 48654 W: 8169 L: 8093 D: 32392
sprt @ 5+0.05 th 31 Take 2, param 2.26. On 31 threads
19-01-22 sni singular11 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 23178 W: 4974 L: 5025 D: 13179
sprt @ 10+0.1 th 1 Make singular margin a function of (d/4)
19-01-22 xot initiativemg1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 13441 W: 2907 L: 3005 D: 7529
sprt @ 10+0.1 th 1 only initiative if mg > -50 or npm low.
19-01-21 MJZ Blocked_posG6 diff
LLR: -2.61 (-2.94,2.94) [0.50,4.50]
Total: 7802 W: 1177 L: 1282 D: 5343
sprt @ 120+1.2 th 1 Blocked position. Try n°6 with extension for potentially blocked positions + limit to {d < 6 and ply > 32+d}. Limited test at VLTC with low throughput (test needs high depth search).
19-01-22 pro ps_ranksfiles1 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 62532 W: 13720 L: 13679 D: 35133
sprt @ 10+0.1 th 1 try to remove FileBB and RankBB arrays.
19-01-22 Viz snMargin2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 24227 W: 5285 L: 5330 D: 13612
sprt @ 10+0.1 th 1 Take 2
19-01-22 Ala UnpinnedAttackBB diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 21815 W: 4802 L: 4859 D: 12154
sprt @ 10+0.1 th 1 Take 2 with different parameters balance. Testing against version with the new bb overhead to first find something improving eval and try and get to have several places with such improvement to overcome overhead.
19-01-22 pro ps_moveimportance1 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21656 W: 4806 L: 4684 D: 12166
sprt @ 10+0.1 th 1 #2: try a simple sigmoid for move importance.
19-01-22 Viz snMargin1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 14697 W: 3203 L: 3295 D: 8199
sprt @ 10+0.1 th 1 Take 1
19-01-22 pro ps_moveimportance1 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 13083 W: 2951 L: 2812 D: 7320
sprt @ 10+0.1 th 1 try a simple sigmoid for move importance.
19-01-21 xot initiativemg1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 19060 W: 4177 L: 4247 D: 10636
sprt @ 10+0.1 th 1 only initiative if mg score also reasonable, >-20.