registrycompare

Role then format against Strict schema anchor

Both scored on the same cases against claude-fable-5, same judge. The intervals tell you whether the gap is real or noise.

suite reasoning-chainmodel claude-fable-5n=168 public cases each

role-then-format

v2 by @ilya
91.7
86.5 to 95.0 · holdout 90.3

strict-schema-anchor

v4 by @marek
94.0
89.4 to 96.7 · holdout 93.1
Intervals overlap. The 2.3 point gap is not significant at this sample size.
Metricrole-then-formatstrict-schema-anchorBetter
Public score91.794.0right
Holdout score90.393.1right
Interval width8.57.3right
Cost per run$1.850$1.850same
Median latency2.4s1.6sright
Sample size168168same

Reading this verdict

Overlap
Two entries are only meaningfully apart when their confidence intervals do not overlap. A higher number with an overlapping interval is a coin flip, not a win.
Holdout
The public score can be gamed by tuning to visible cases. A gap between public and holdout is the tell.
Reproduce
evalness compare \ role-then-format strict-schema-anchor \ --model claude-fable-5