The skill profile honest edition, sample sizes included how?

Percentile among all 105, per task type. Dimmed spokes rest on fewer than 5 tasks: the model barely knows.

The map strength vs volatility, all 105

Hover or tap any dot. Tap the name to open that dossier.

The record standout tasks, both directions

The season episode by episode, against the Series 7 cast

Statistical kin who they play like, era regardless

Nearest neighbours in strength and volatility. Cross-series comparisons lean on the model's prior; treat as resemblance, not head-to-head evidence.

The tape them, at their best

The evidence what the number rests on

62 scored tasks, 2 disqualifications, one series alongside James Acaster, Jessica Knappett, Phil Wang, Rhod Gilbert. Strength is fitted from who beat whom on every task, and carries a standard error of ±0.0963. More tasks would narrow it; a Champion of Champions appearance would anchor it across eras. Until then, the uncertainty is part of the answer.