This note is a short survey of likelihood methods used in partial wave analysis followed by a longer discussion about using the resulting log-likelihoods as goodness-of-fit measures. While there are several approaches that can be used to give a goodness-of-fit measure, most of them suffer in the lack of an absolute calibration on the goodness. Rather, they provide a relative difference, and some guidance on whether the absolute value might be close to the best value.
