This article, published in the Web of Science History first-decile journal Historical Methods, is a significant outcome of the GAČR-EXPRO and ERC Consolidator Grant-funded Dissident Networks Project (DISSINET), based on comprehensive formal data analysis. It disentangles the intricate relations between crime and punishment in earliest medieval inquisitions. It builds on the fact that despite significant research on the techniques of repression employed by medieval inquisitors against religious dissidents, the case-level influences on the penances they meted out are understood only vaguely: the extent to which sentencing “systems” existed is unknown. To overcome this, we apply formal methods – an exploratory analysis supported by crisp-set Qualitative Comparative Analysis, and statistical modeling founded on multiple linear regression – to the large and historically significant register of Peter Seila (1241–2), captured as structured data via a statement-based approach entitled “Computer-Assisted Semantic Text Modelling” (CASTEMO). The results show that Peter systematically weighted different types of crimes and dissident interactions when sentencing; they do not suggest, however, that he was influenced by accomplicity or kinship among the sentenced.