Recent case studies have revealed that purely formal analogies have been successfully used as a heuristic in physics. This is at odds with most general philosophical accounts of analogies, which require analogies to be physical in order to be justifiably used. The main goal of this paper is to supply a philosophical account that justifies the use of purely formal analogies in physics. Using Bartha’s (2010) articulation model as a starting point, I offer precise definitions of formal and physical analogies (...) and propose a new submodel of analogical reasoning that accounts for the successful use of purely formal analogies in the development of renormalization group methods for use in particle physics in the early 1970’s (Fraser 2020). Two distinctive features of this new applied mathematics submodel for analogical reasoning are that the conclusion of the argument from analogy includes both an entire model (and not only a hypothesis or a prediction) and the construction procedure for this model. A third important difference from arguments from physical analogy is that only the prima facie plausibility of the conclusion is established, and not stronger types of plausibility associated with confirmation. The use of purely formal analogies is justified because they are suited to supporting conclusions of this sort. Formulating a general philosophical account of analogies that covers purely formal analogies also serves two additional purposes: (1) to highlight the features of this case that are novel in the context of examples of analogies traditionally considered by philosophers and (2) to establish that physicists should not automatically dismiss purely formal analogies when evaluating heuristics for the development of new models. (shrink)