As I said, the models are tested on historical data. Of course that is far from perfect, but it is also better than nothing.
Not sure what other things they do, I am not in that area of research.
But it is not as if nothing at all is known about climate. There is a certain knowledge of physics, and climate responds to the rules of physics. So some things can be derived (heat refelecting from CO2 layer and stuff like that, I suppose). It is not just a bunch of religious types making up stuff.
My question is, how would one know if they were wrong?