It describes methods for predicting time usage in situations with high uncertainty, explains why two plus two is usually more than four in time prediction contexts, reports on research on time prediction biases, and summarizes the evidence in support of different time prediction methods and principles.