In recent years, the use of Bayesian statistical analysis for modelling radiocarbon dates has become much more commonplace within archaeology, however there remains very little literature to aid new practitioners with understanding the archaeological assumptions that underpin many of the modelling choices. Using a suite of radiocarbon dates from a sample of seven ovens, from the 18 excavated across the site at Ysgol yr Hendre, Caernarfon, North Wales, this paper explores the various methods of modelling this group of features. It stresses the underlying archaeological assumptions of each model, while also highlighting some of the tools that should especially be considered when developing very precise models. Finally it compares the results of the models.