I think the problem is that economics needs to take human behavior into account, which it tries to do via game theory, which in turn relies on rational actors. I don't think individual human behavior is predictable like that. You could argue that we don't need to model individuals, only large groups of people. But the problem is that often the behavior of some individuals has a huge effect on large groups.
Contrast that to physics, where we have been very successful with predicting the behavior of masses both on small and on large scale, with relatively few exceptions.
