Keep only the evidence.

Objective: turn a prior population into an observed population and build the correct conditional fraction. Compare bets, base rates and checks.

Reading your device evidence model…