This paper combines agent-based modelling with coin-level data from The Coin Hoards of the Roman Empire Project, presenting early findings of doctoral work on coin circulation during the third and fourth centuries AD.
Computational methodological approaches are still in their relative infancy within Roman numismatics yet are being increasingly employed within the field. Leveraging the accessibility of computational power and the growing digitisation and expansion of available datasets in numismatic studies, researchers are now able to test sophisticated theoretical frameworks against reproducible computational simulations.
Specifically, this paper presents several case studies based on coin hoard data from the third and fourth centuries AD. Using a custom-built agent-based model employing a wide range of GIS spatial data of the Roman transport network, it will be explored how numismatists can investigate complex phenomena in both spatial and temporal settings using a bottom-up approach, thereby expanding our understanding of Roman economic systems.