In decision theory, a scoring rule is a measure of someone's performance at making predictions under uncertainty. A proper scoring rule encourages the forecaster to be honest, as expected payoff is maximized by accurately reporting personal belief about the predicted event, rather than by gaming the system.