Hamilton (the founding father) considered political science a science, and explicitly argues in the federalist papers how it has the equivalent of axioms that one can build and reason from

How well have the predictions made by reasoning from those axioms worked out?

It's true that the most egregiously wrong prediction in the Federalist papers, that the US would not have political parties because it was too large, was the work of Madison, not Hamilton. But AFAIK none of Hamilton's predictions turned out very well either.