D40-F05-A10 · GUIDED LEARNING LAB · SYNTHETIC DATA

Rare-Event Backtest and Confidence Bounds

Change events, trials, and confidence to see why zero or few observed events still imply material uncertainty.

Teaching stage

Rare-Event Backtest and Confidence Bounds teaching stageTopic-specific metric geometry calculated from synthetic teaching data.

Choose → step → observe → explain → compare

    Try
    Observe
    Do not concludeThis synthetic lab validates metric mechanics, not production fitness, fairness, compliance, causation, or policy approval.

    Compact evidence trace

    Only the rows needed to explain this state are shown; the canonical implementation remains the source of truth. Accessibility QA can force deterministic reduced motion with ?reduced-motion=1.

    EvidenceCurrent valueWhy it matters