Over the past few years, a movement has been initiated among forensic science researchers and practitioners, in Switzerland and across Western criminal justice systems more generally, to shift away from categorical assertions of common source (e.g., «this crime scene mark and this control print come from the same person») towards new reporting formats that consider expert conclusions as decisions (e.g., an «identification» or «exclusion» is a decision made by the forensic examiner). As this movement gains momentum, there remains disagreement on how exactly the notion of decision ought to be understood. The call for improvement of the understanding of the logical tenets of forensic identification decisions faces the obstacle that many forensic practitioners shy away from formal and quantitative approaches (e.g., decision theory). The purpose of this contribution is to show that the logical essentials of forensic identification decisions can be captured and conveyed without going into the details of the mathematics of decision theory. We will then present and defend the view that forensic identification requires assessments and value judgments that go beyond the forensic practitioner's area of competence and that this fact requires a reassessment of the distribution of responsibilities between experts and other participants in the legal process.