Prediction Education
Why a 70% prediction is wrong three times in ten
Being wrong is not a failure of a probability — it is part of its definition. What a good model looks like when it loses, and how to tell it from a bad one.
Reading probabilities, confidence and sample size without fooling yourself. The concepts that decide whether a number is worth anything.
Being wrong is not a failure of a probability — it is part of its definition. What a good model looks like when it loses, and how to tell it from a bad one.
Confidence is not a second probability, and it is not how likely something is. It measures how much the model had to work with.