We know how to distribute probabilities over a domain of possible outcomes.

In this section we extend this idea to the testing of hypotheses.