Setup · League 3

GPT-6.1 Solhigh

FinishedPublished

Results f24d1f41, updated 16 November 2026, 21:24 UTC; scored by GNU Backgammon 1.08, 2-ply.

How we called it

Model ID
Missing
Provider or endpoint
Missing
Date run
2026-10-19 to 2026-10-24
Price
Missing
Model type
Missing
Weights
Missing
Runs on
Missing
Reasoning setting
Missing
Confidence source
Missing
Rules text
Missing
Option groups
not recorded; in this data 242 decisions were split into groups of 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26
Sampling settings
Missing
Notes
Missing

Missing B7 fields: every B7 field

Results in League 3

Error rate (PR)
9.98.6 to 11.3
Decisions
1,389 (1,389 scored, 0 invalid)
Engine's best play
64.4%
Blunder rate
7.8%
P95 equity loss
0.103
Mean loss, worst 5%
0.187
Median / P90 s per decision
13.1 / 56.5
$ per 1,000 decisions
15.573API-equivalent (subscription)
Risk judgement, margin 0.001
72.2 (95% interval 69.5 to 74.9); risks found 50.6% (212 / 419), right safety 93.7% (568 / 606)
Risk judgement, margin 0.02
72.4 (95% interval 69.7 to 75.1); risks found 52.4% (188 / 359), right safety 92.5% (616 / 666)
No trade-off
364 (left out of every risk metric)

JSON

Risk call card

72.2risk judgement, margin 0.001

right risks212
wrong risks21
needless risks38
missed risks186
right safety568

Hit risk only: being hit. Gammon risk and priming are not in it.

Close choices

A decision is a close choice when the second most likely play has at least half the probability of the first (internally: split).

Probabilities over plays, not sampled reasoning. Method

Decisions
1,389
Eligible (full probability distribution)
0
Left out, by reason
one letter only, or a program 1,389
Close choices
N/A no full probability distribution

By game phase

PhaseError rate (PR)Decisions
OpeningToo few decisions183
Contact13.6824
Holding gameToo few decisions85
RaceToo few decisions150
Bear-offToo few decisions147