The article introduces a new area of the transdisciplinary scientific study of religion which combines methods from the humanities and sciences with a special focus on the study of complex adaptive systems. It discusses the area's theoretical aims and conditions while offering a short review of selected case studies which demonstrate the merit of the approach. The paper joins present epistemological discussions about combining the expertise of the sciences and humanities and takes a pragmatic stance between the normative exhortation for the scientific study of religions and the discussion of innovative methodological horizons emerging from digital humanities. The central position is dedicated to the practice of "formalizing modeling" and its possible research utility in the historically grounded study of religions. On the one hand, the article notices the limited contribution of the cognitive science of religion project to a historiographical study of religions. On the other hand, the article discusses the limits of conventional historiography arising from an orientation to the histories of events and persons rather than to histories of so-called "long duration". Generally, the paper argues for the complementing of the "close reading" of historical sources with "distant reading" and hypothesis-driven research utilizing a variety of formal modeling approaches (GIS, agent-based modeling, complex networks) and computer-based methods. The discussed methods offer new ways of representing data and are understood not only as innovative means of solving historical problems but also as a platform for asking new questions based on the fusion of scientific and humanistic imaginations.