FuryBench Testing Framework

home toIntegralPlentyChessObsidianCaissaQuanticadeSerendipitySpaghetSchoenemannClarityIlluminaPawnocchioVineAnura
No logs for event exist
Active : 15 Machines / 174 Threads / 105.71 MNPS
Priority 0
Illumina Mergener
fix-increments
diff40.0+0.40
TUNE   5 parameters
ITER   259 / 500
GAMES  4,148 / 8,000
Quanticade DarkNeutrino
master
25K node datagen.
diffN=25000
ELO    -0.11 +- 0.12 (95%)
GAMES  N: 24517926/30000000 W: 8603410 L: 8610889 D: 7303627
PlentyChess Yoshie2000
no-pv-razor
diff8.0+0.08
ELO    -0.05 +- 2.89 (95%)
LLR    -0.38 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 13638 W: 3298 L: 3300 D: 7040
PENTA  [18, 1605, 3570, 1613, 13]
PlentyChess Yoshie2000
no-ttpv-razor
diff8.0+0.08
ELO    0.25 +- 1.73 (95%)
LLR    -0.59 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 37942 W: 9141 L: 9114 D: 19687
PENTA  [62, 4405, 9999, 4454, 51]
Integral Fury
lmr_eval_alpha_diff
diff10.0+0.10
ELO    -0.04 +- 2.15 (95%)
LLR    -0.97 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 24722 W: 6025 L: 6028 D: 12669
PENTA  [46, 2869, 6524, 2886, 36]
Caissa Witek
pairwise-rounding
Neural net: Use rounding to nearest in pairwise multiplication
diff10.0+0.10
ELO    0.32 +- 0.66 (95%)
LLR    -1.83 (-2.94, 2.94) [0.00, 2.00]
GAMES  N: 287914 W: 68593 L: 68328 D: 150993
PENTA  [1070, 34404, 72714, 34729, 1040]
Priority -1
Caissa Witek
qs-conthist-fix
QSearch: Fix uninitialized cont history pointers
diff10.0+0.10
ELO    -0.30 +- 1.82 (95%)
LLR    -0.95 (-2.94, 2.94) [0.00, 2.00]
GAMES  N: 36178 W: 8486 L: 8517 D: 19175
PENTA  [96, 4322, 9283, 4293, 95]
Finished
Integral Fury
lets_think_about_killers_differently
diff10.0+0.10
ELO    -2.72 +- 2.95 (95%)
LLR    -2.25 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 12920 W: 3110 L: 3211 D: 6599
PENTA  [18, 1541, 3448, 1430, 23]
Integral Fury
seval_update_shenanigans
diff10.0+0.10
ELO    -5.93 +- 4.18 (95%)
LLR    -2.19 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 6564 W: 1551 L: 1663 D: 3350
PENTA  [14, 819, 1721, 721, 7]
Integral Fury
seval_update_shenanigans
diff10.0+0.10
ELO    0.26 +- 1.82 (95%)
LLR    -0.85 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 34744 W: 8517 L: 8491 D: 17736
PENTA  [59, 4067, 9099, 4083, 64]
Integral Fury
bad_noisy_capthist_bonus
diff10.0+0.10
ELO    -0.58 +- 1.84 (95%)
LLR    -2.25 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 33922 W: 8268 L: 8325 D: 17329
PENTA  [60, 3989, 8919, 3934, 59]
PlentyChess Yoshie2000
hp-cap-only-bad
diff8.0+0.08
ELO    0.13 +- 1.47 (95%)
LLR    -1.09 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 53176 W: 12853 L: 12833 D: 27490
PENTA  [73, 6269, 13906, 6245, 95]
Integral Fury
mp_gives_check_see
diff10.0+0.10
ELO    -3.12 +- 3.14 (95%)
LLR    -2.26 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 11702 W: 2836 L: 2941 D: 5925
PENTA  [20, 1431, 3047, 1340, 13]
Integral Fury
[PlentyChess] main
diff40.0+0.40
ELO    -2.09 +- 4.48 (95%)
GAMES  N: 5324/10000 W: 1295 L: 1327 D: 2702
PENTA  [5, 613, 1455, 587, 2]
PlentyChess Yoshie2000
rfp-cn-margin-2
diff8.0+0.08
ELO    -0.64 +- 1.61 (95%)
LLR    -2.29 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 43710 W: 10585 L: 10665 D: 22460
PENTA  [48, 5187, 11477, 5083, 60]
Integral Fury
main
Bad Noisy Futility Pruning (#259)
diff40.0+0.40
ELO    32.48 +- 4.65 (95%)
GAMES  N: 5000/5000 W: 1443 L: 977 D: 2580
PENTA  [2, 367, 1302, 821, 8]
PlentyChess Yoshie2000
fp-search-first-quiet-check
diff8.0+0.08
ELO    -0.00 +- 1.15 (95%)
LLR    -2.27 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 87518 W: 21268 L: 21268 D: 44982
PENTA  [134, 10448, 22584, 10470, 123]
Integral Fury
bad_noisy_fp_attempt_1bajillion
diff10.0+0.10
ELO    2.12 +- 1.65 (95%)
LLR    2.92 (-2.25, 2.89) [0.00, 3.00]
GAMES  N: 43212 W: 10786 L: 10522 D: 21904
PENTA  [101, 4934, 11258, 5226, 87]
PlentyChess Yoshie2000
fix-one-singular
diff8.0+0.08
ELO    1.07 +- 1.92 (95%)
LLR    2.94 (-2.25, 2.89) [-3.00, 1.00]
GAMES  N: 26674 W: 6611 L: 6529 D: 13534
PENTA  [35, 2634, 7926, 2698, 44]
Integral Fury
tm_fixes
diff40.0+0.40
ELO    11.06 +- 10.87 (95%)
LLR    0.61 (-2.25, 2.89) [-3.00, 1.00]
GAMES  N: 660 W: 171 L: 150 D: 339
PENTA  [0, 45, 219, 66, 0]
Caissa Witek
eval-ml-8-dtm-b0-v2
Endgames: Remove KBNvK endgame evaluator KBNvK was preferred more than other, more winning endgame, leading to draws in winning positions. STC, KBNvK_dtz60.epd: Results of New vs Old (6+0.06, 1t, 8MB, KBNvK_dtz60.epd): Elo: 95.95 +/- 12.31, nElo: 234.36 +/- 28.57 LOS: 100.00 %, DrawRatio: 45.42 %, PairsRatio: 37.75 Games: 568, Wins: 164, Losses: 11, Draws: 393, Points: 360.5 (63.47 %) Ptnml(0-2): [0, 4, 129, 145, 6], WL/DD Ratio: 0.06 LLR: 2.95 (100.0%) (-2.94, 2.94) [0.00, 5.00] STC, endgames_cdb95105.epd: ELO 11.02 +- 3.49 (95%) SPRT 10.0+0.10s Threads=1 Hash=16MB LLR 3.00 (-2.94, 2.94) [0.00, 2.00] GAMES N: 7758 W: 1948 L: 1702 D: 4108 PENTA [2, 667, 2293, 917, 0] https://furybench.com/test/7502/
diff10.0+0.10
ELO    0.04 +- 0.71 (95%)
LLR    -2.98 (-2.94, 2.94) [0.00, 2.00]
GAMES  N: 195736 W: 46754 L: 46733 D: 102249
PENTA  [38, 20427, 56918, 20446, 39]
PlentyChess Yoshie2000
rfp-cn-margin
diff8.0+0.08
ELO    -1.23 +- 1.96 (95%)
LLR    -2.30 (-2.25, 2.89) [0.00, 2.50]
GAMES  N: 30014 W: 7256 L: 7362 D: 15396
PENTA  [53, 3593, 7815, 3499, 47]