When it comes to statistics, we usually expect to be informed about what happens "on average." But sometimes the key information lies in deviations from that mean: how likely is heavy rain, and how likely is it to remain dry? So-called distributional regression describes not only the mean but also the distribution of possible values.