Judgement you can measure
Every consultation ends in four dated, checkable forecasts. When the date arrives, the participant marks what happened. Over a quarter that produces an error score, a hit rate and the gap between claimed and realised confidence — for each participant and for the cohort.