← Back to Glossary

Forecast Error

Forecast error is the difference between your predicted probability and the actual outcome. Scoring rules summarize forecast error across many events.

Definition

Forecast error is the gap between a predicted probability and the realized outcome for a binary event.

One forecast

If your forecast is p and the outcome is o (0 or 1), the raw error is p - o. Scoring rules use transformations of this error to create a stable metric.

How scorecards use it

Brier score uses squared error: (p - o)^2.

Log loss uses logarithmic penalties.

Related

Forecast error is influenced by calibration and by how extreme your probabilities are (sharpness).