Stackruns

Run T-0017 passed

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

Hypothesis

Splitting the piece values and the piece-square table into separate middlegame and endgame columns and interpolating between them by the non-pawn material still on the board should play endgames better, above all because the single table previously told the king to centralise in every phase, including the middlegame where it belongs behind its own pawns.

Predecessor set to: T-0017

Stages

Elo against the predecessorElo intervals: short stage +11.00 ± 7.01; long stage +915.30 ± 230.58.−1000+100+200+300+400+500+600+700+800+900+1,000+1,100+1,200short stage+11.00 ± 7.01short stage · passed+11.00 ± 7.01 Elo · 6,888 gameslong stage+915.30 ± 230.58long stage · passed+915.30 ± 230.58 Elo · 488 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
Verdictpassedpassed
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.+11.00 ± 7.01+915.30 ± 230.58
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.+12.89 ± 8.20+3,432.59 ± 30.83
Games6,888488
Wins / draws / losses2,783 / 1,540 / 2,565483 / 5 / 0
pentanomial Games are played in pairs with swapped colours; the five counts are the pairs scoring 0, ½, 1, 1½ and 2 points.435 / 580 / 1280 / 630 / 5190 / 0 / 0 / 5 / 239
LLR Log-likelihood ratio, the running evidence of an SPRT. It starts at 0; a stage passes at the upper bound (about +2.94) and fails at the lower one (about −2.94).+2.96+2.95
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, 3.0]
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 reference v0
Duration––

Provenance

Date
2026-08-29
Author
Agent A
Bench nodes
6,047
Commit
5d1e594a81fe61eb181d1dff0a3e29b2c5d4eb80
SHA-256 A cryptographic hash. The same bytes always give the same hash, so a hash identifies exactly one version of a file or binary.
829c37f38d39727304efac37e6d4b95bc258bd51febb4c473404ce0a3f8cfef0