# Proper Scoring Rule

A proper scoring rule is a loss function whose expected value is minimized when the predicted probability distribution coincides with the true probability distribution. These scoring rules are widely used in probability forecasting. The use of proper scoring rules encourages the forecaster to be honest, as his expected loss is minimised when he reports his personal probability as the prediction. Among examples of proper scoring rules are the square-loss (Brier loss function) function and the logarithmic loss function.

A proper scoring rule is one of the ways to measure the quality of forecasting. It is related to calibration and resolution.