Paola Janssen formulates judicial conviction formation as a subjective statistical decision problem, thus referring to a methodological interface between statistics and jurisprudence. The application of Bayesian statistics ensures the logical consistency of information processing and decision-making. However, it requires the establishment of a priori probabilities, over which intersubjective consensus may not be achievable. Sensitivity analyses can mitigate this problem. The author illustrates these issues using the criminal trial against former weather presenter Jörg Kachelmann, by representing and evaluating the evidence as a Bayesian network.