It is often claimed that the medievals had a 'statistical' model of modality; or, at least, that they saw Aristotle as having a 'statistical' model of modality. This view is due to Knuuttila. Now, Knuuttila is certainly more an expert in the history of modality than I am, and by a long shot, but I am skeptical. On the statistical model:
(1) p is necessary iff for all times t, p occurs at t.
(2) p is possible iff there is some time t such that p occurs at t.
And the medievals often say things that could be interpreted in this way; a common point is that what happens always or for the most part is necessary. And it may well be that some medievals really do buy into precisely this model of modality. But I think we should be very cautious about saying a particular scholastic does in a particular context. For one thing, the sort of necessity and possibility that they are discussing in these contexts is natural necessity; so the modality here is tangled up tightly in the scholastic notion of a 'nature'. Really, in many cases (1) should read:
(1) p is necessary if and only if there is an operative nature such that (ceteris paribus) for all times t, p occurs at t.
The 'ceteris paribus', I think, is needed given that natural action can be impeded (this is the 'for the most part' of the 'always or for the most part' criterion); and the occurrence of p is dependent on the operation of a nature. And, while the reminder that 'natures' are involved here may not make much difference, it isn't really statistical in any sense at all. And the necessity here is rather tricky (as is the possibility); it is a necessity that is bounded by the nature of the thing in question. But it's always possible in these things that I am missing something.