program · League 2

GNUbg weak

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
GNU Backgammon 1.08, Beginner level (0-ply, noise 0.060), x86-64 build (see About the anchor)
Date run
2026-10-16 to 2026-10-25
Price
Missing
Model type
program
Weights
Missing
Runs on
CPU
Reasoning setting
Missing
Confidence source
none
Rules text
Missing
Option groups
not recorded; no decision in this data was split
Sampling settings
Missing
Notes
GNU Backgammon 1.08, Beginner level (0-ply, deterministic noise 0.060); fixed anchor in every league

Results in League 2

Error rate (PR)
27.024.1 to 29.8
Decisions
2,136 (2,136 scored, 0 invalid)
Engine's best play
46.2%
Blunder rate
22.1%
P95 equity loss
0.263
Mean loss, worst 5%
0.384
Median / P90 s per decision
0.01 / 0.01
$ per 1,000 decisions
Not measured
Risk judgement, margin 0.001
62.3 (95% interval 59.5 to 64.8); risks found 62.7% (483 / 770), right safety 61.9% (558 / 902)
Risk judgement, margin 0.02
61.7 (95% interval 58.7 to 64.6); risks found 64.8% (420 / 648), right safety 58.5% (599 / 1,024)
No trade-off
464 (left out of every risk metric)

JSON

Risk call card

62.3risk judgement, margin 0.001

right risks483
wrong risks176
needless risks344
missed risks111
right safety558

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

League 2: complete letter-readout records are mostly few-option positions, a selected subset.

Decisions
2,136
Eligible (full probability distribution)
0
Left out, by reason
one letter only, or a program 2,136
Close choices
N/A no full probability distribution

By game phase

PhaseError rate (PR)Decisions
Opening36.3271
Contact30.31,502
Holding gameToo few decisions25
Race5.0247
Bear-offToo few decisions91