Ah ha! You just identified the great debate in Political Science. It's very hard to empirically test game theoretical models since a lot of the model components

Ah ha! You just identified the great debate in Political Science. It's very hard to empirically test game theoretical models since a lot of the model components are about the utility of the actors which is very hard to get at empirically. Plus, things like repeated game have multiple equilibria and we don't really know why the equilibrium that is our world gets to happen.