FuryBench Testing Framework

home toIntegralPlentyChessObsidianCaissaQuanticadeSerendipitySpaghetSchoenemannClarityIlluminaPawnocchioVineAnura
No logs for event exist
Active : 31 Machines / 500 Threads / 237.7 MNPS
Priority 0
PlentyChess Yoshie2000
npv-fds-ext
diff40.0+0.40
ELO    1.08 +- 1.58 (95%)
LLR    0.88 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 41970 W: 10084 L: 9954 D: 21932
PENTA  [3, 4663, 11526, 4787, 6]
Quanticade DarkNeutrino
fp-correction
search.c: Correction margin to futility pruning
diff60.0+0.60
ELO    -0.23 +- 2.04 (95%)
LLR    -1.21 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 24598 W: 5618 L: 5634 D: 13346
PENTA  [6, 2702, 6897, 2690, 4]
Finished
Caissa Witek
fifty-move-fit
Search: Fitted 50-move eval damping Replace the linear (234 - hmc) / 234 eval damping with a curve fitted on game results: 1 - 0.221 x - 0.713 x^2 with x = hmc / 100, and 1 - x for pawnless positions. The multipliers are a table built at startup (indexed by pawnless and half-move counter), so the search does one lookup, a multiply and a shift. Fit: for each half-move-counter bucket, the slope k of the logistic E(k * eval) that best predicts the game result, relative to HMC 0-7, using the raw static eval of the master net. 69M positions from OpenBench STC games, 2.1M from LTC games and 13.6M from selfplay data give nearly the same curve (multiplier at HMC 48: 0.70 / 0.61 / 0.71, at HMC 80: 0.36 / 0.32 / 0.40; previously 0.79 and 0.66). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
diff10.0+0.10
ELO    -5.90 +- 2.77 (95%)
LLR    -2.95 (-2.94, 2.94) [0.00, 2.00]
GAMES  N: 13024 W: 3025 L: 3246 D: 6753
PENTA  [1, 1510, 3714, 1283, 4]
PlentyChess Yoshie2000
npv-fds-ext
diff8.0+0.08
ELO    0.91 +- 0.69 (95%)
LLR    2.89 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 243308 W: 59420 L: 58786 D: 125102
PENTA  [334, 28319, 63719, 28943, 339]
Quanticade DarkNeutrino
fix-history-enpas
search.c: Fix capture history indexing on enpassant
diff10.0+0.10
ELO    1.11 +- 2.02 (95%)
LLR    2.97 (-2.94, 2.94) [-3.00, 1.00]
GAMES  N: 27626 W: 6597 L: 6509 D: 14520
PENTA  [62, 3065, 7470, 3155, 61]
PlentyChess Yoshie2000
badcap-fds-ext
diff8.0+0.08
ELO    -0.71 +- 1.68 (95%)
LLR    -2.26 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 40756 W: 9879 L: 9962 D: 20915
PENTA  [56, 4850, 10644, 4777, 51]
PlentyChess Yoshie2000
recap-fds-ext
diff8.0+0.08
ELO    -1.39 +- 2.06 (95%)
LLR    -2.26 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 26998 W: 6571 L: 6679 D: 13748
PENTA  [61, 3188, 7081, 3136, 33]
PlentyChess Yoshie2000
corr-fds-ext
diff8.0+0.08
ELO    -9.34 +- 4.59 (95%)
LLR    -2.26 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 5468 W: 1255 L: 1402 D: 2811
PENTA  [13, 707, 1437, 568, 9]
Quanticade DarkNeutrino
fp-correction
search.c: Correction margin to futility pruning
diff10.0+0.10
ELO    0.10 +- 1.50 (95%)
LLR    -1.63 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 50666 W: 11967 L: 11952 D: 26747
PENTA  [90, 5886, 13355, 5923, 79]
Quanticade DarkNeutrino
see-ray-fix
see.c: Fix SEE pinned ray
diff10.0+0.10
ELO    -0.13 +- 0.95 (95%)
LLR    2.94 (-2.94, 2.94) [-3.00, 1.00]
GAMES  N: 121730 W: 28838 L: 28884 D: 64008
PENTA  [223, 13521, 33428, 13465, 228]
Quanticade DarkNeutrino
[Stockfish] master
Adv datagen vs SF 6000 fixed nodes
time odds diffN=5000
ELO    -5.38 +- 0.19 (95%)
GAMES  N: 10000006/10000000 W: 3782460 L: 3937423 D: 2280123
Quanticade DarkNeutrino
16k-history-limit
Switch to 16K history limits
diff10.0+0.10
ELO    -1.95 +- 1.94 (95%)
LLR    -2.97 (-2.94, 2.94) [-3.00, 1.00]
GAMES  N: 31132 W: 7316 L: 7491 D: 16325
PENTA  [61, 3801, 8003, 3654, 47]
PlentyChess Yoshie2000
refuted-by-probcut
diff40.0+0.40
ELO    1.09 +- 0.89 (95%)
LLR    2.89 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 133196 W: 32346 L: 31928 D: 68922
PENTA  [24, 14692, 36740, 15126, 16]
Integral Fury
corrhist_fmr_buckets
diff10.0+0.10
ELO    -1.58 +- 2.41 (95%)
LLR    -2.27 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 19356 W: 4719 L: 4807 D: 9830
PENTA  [36, 2248, 5191, 2174, 29]
Integral Fury
singular_ext_defer_tt_move
diff10.0+0.10
ELO    -0.67 +- 1.89 (95%)
LLR    -2.30 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 32628 W: 8017 L: 8080 D: 16531
PENTA  [59, 3872, 8543, 3753, 87]
Integral Fury
moar_hist_bonus
diff10.0+0.10
ELO    -0.51 +- 1.75 (95%)
LLR    -2.40 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 38510 W: 9502 L: 9558 D: 19450
PENTA  [68, 4613, 9968, 4519, 87]
PlentyChess Yoshie2000
0194
diff40.0+0.40
ELO    -3.67 +- 2.96 (95%)
LLR    -2.26 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 12316 W: 2976 L: 3106 D: 6234
PENTA  [7, 1466, 3343, 1334, 8]