Base Rate
Base rate is the underlying frequency that an event occurs in a reference set. It is a natural baseline forecast and is often used to compute Brier skill score.
Definition
The base rate is the observed frequency of an outcome in a relevant reference set. If a type of event happens 30% of the time historically, the base rate is 0.30.
Why it matters
Base rates are the simplest useful benchmark for probabilistic forecasting. They also help you interpret results: if your dataset is dominated by yes outcomes, naive forecasts can look strong unless you evaluate skill relative to base rates.
Base rate as a baseline
When computing Brier skill score, a common baseline is “climatology”, which predicts p0 = base rate for every event in the set. This is often a tougher baseline than 50/50.
Common pitfalls
Using the wrong reference set: Base rate should match the question type and time period. Mixing unrelated events can give misleading baselines.
Ignoring regime changes: If conditions change, historical base rates may not apply going forward. Test out of sample when possible.
Related
Base rate connects directly to calibration and to skill metrics like Brier skill score.