Stackruns

Run T-0071 rejected

· Evidence: SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.

Hypothesis

If negamax clamps its alpha-beta window to the mate values reachable at ply [-Mate + ply, Mate - ply - 1], the fixed-depth result and mate distance are preserved, while impossible subtrees are eliminated.

translated from German. Original: Wenn negamax sein Alpha-Beta-Fenster auf die bei ply erreichbaren Mattwerte [-Mate + ply, Mate - ply - 1] klemmt, bleiben Fixed-depth-Ergebnis und Mattdistanz erhalten, während unmögliche Teilbäume entfallen.

Stages

Elo against the predecessorElo intervals: short stage +2.71 ± 2.11; long stage +1.05 ± 1.65.0+2+4short stage+2.71 ± 2.11short stage · passed+2.71 ± 2.11 Elo · 48,340 gameslong stage+1.05 ± 1.65long stage · rejected+1.05 ± 1.65 Elo · 76,396 games
Point estimate with the reported 95-percent half-width, per stage. The dashed line is zero. Estimates from a stopped sequential test are biased by the stopping rule; the interval is nominal.
Field short stagelong stage
Verdictpassedrejected
Elo Strength difference to the opponent, estimated from the games, with a 95-percent half-width where the oracle reported one. Never an absolute rating.+2.71 ± 2.11+1.05 ± 1.65
nElo Normalised Elo: the Elo difference divided by the spread of the game results, so that the SPRT bounds mean the same at different time controls.+3.99 ± 3.10+1.56 ± 2.46
Games48,34076,396
Wins / draws / losses16,366 / 15,985 / 15,98924,382 / 27,862 / 24,152
pentanomial Games are played in pairs with swapped colours; the five counts are the pairs scoring 0, ½, 1, 1½ and 2 points.1566 / 4601 / 11582 / 4732 / 16892282 / 7861 / 17752 / 7951 / 2352
LLR Log-likelihood ratio, the running evidence of an SPRT. It starts at 0; a stage passes at the upper bound and fails at the lower one.+2.97−2.97
Bounds−2.94 … +2.94−2.94 … +2.94
SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.[0.0, 5.0][0.0, 5.0]
Error rates α / β0.05 / 0.050.05 / 0.05
Modelnormalizednormalized
time control Base time plus increment per move in seconds, e.g. 8+0.08: eight seconds per game plus 0.08 seconds per move.8+0.0840+0.4
BookAB
Opponentvs predecessorvs predecessor
Duration5 h 47 min46 h 30 min

Provenance

Date
2026-09-05
Started
2026-09-05 09:35:55 UTC
Duration
52 h 18 min
Author
Agent B
Epoch
4
Gatekeeper
passed
Bench nodes
3,537