> I think a better thing to look at is if a theory can correctly calculate things that weren't built into the theory. I remember reading [1] that string theory

> I think a better thing to look at is if a theory can correctly calculate things that weren't built into the theory. I remember reading [1] that string theory correctly calculated some things concerning black holes that none of the people who had developed string theory to that point knew about. Well if that's true, then string theory definitely isn't useless, so I'd be interested in the specifics. This prediction from string theory wasn't explained by any of the accepted physical models? > Second, I don't like it from a philosophical point of view. It suggests that if say some pre-Newton gravity theorist had just included enough epicycles, science should have rejected Newtonian gravity because it would not have explained anything that the epicycle theory didn't. Well yeah, but why not reject Newtonian gravity if everything observable can be predicted by the epicyclical view? The goal is not to generate pretty models, the goal is to make accurate predictions of the physical universe, ideally through the simplest model we can come up with just to guard against overfitting. Anyway, it's not that I (or anyone, I don't have any physics authority certainly :P), am rejecting string theory, it's just that as far as I understand string theory hasn't had any real use to our physical models. I certainly don't have an alternative to string theory.