Event Prevalence
Event prevalence is the proportion of events that resolve YES in a dataset. It is essentially the base rate of outcome = 1 and drives uncertainty and baseline scores.
Definition
Event prevalence is the fraction of events in a dataset that resolve as YES (outcome = 1). It is equivalent to the base rate for outcome = 1.
Why it matters
Prevalence shapes what “good” looks like. If almost everything resolves YES, simple forecasts can achieve low error. If outcomes are near 50/50, scoring is inherently harder and baseline error is higher.
How it is used
• As a baseline probability for Brier skill score.
• To understand the uncertainty term in Brier decomposition.
Related
See base rate, class imbalance, and benchmark for how prevalence affects evaluation.