Epicycles were certainly a modelling technique for making predictions. They weren't much of a theory since there was nothing in epicycles about causation -- no

Epicycles were certainly a modelling technique for making predictions. They weren't much of a theory since there was nothing in epicycles about causation -- no explanation of why, just a model for what-where-when. I'm sure it must have occurred to more than one back then that the planets go around the Sun in elliptical orbits. The Antikythera mechanism even models the elliptical orbit of the moon around the Earth, so it's not farfetched that they could have figured out the rest. However, convincing all the right "authorities" of the copernican view back then would have been culturally very difficult.