The authors do this by providing “lessons” drawn from their own research program on inverse prediction and related statistical method for estimating postmortem interval (PMI). The first lesson is to use an unbiased sampling technique for generating training data. This is an elementary aspect of good design for many kinds of experiments; the implications were examined for a carrion insect age prediction model. The second lesson is to exceed the minimum sample size for a categorical response, because below the smallest sample size, the training data do not provide sufficient statistical power for a prediction. The third lesson is that the practical significance of a covariate or the practical value of a response should be evaluated by predictive model performance. Overall, the authors argue that predicting PMI should be mathematically explicit and yield a range of values rather than a single value. Defining this range as a confidence set would conform to mainstream scientific practice.