I show this video to students in my Introduction to the Philosophy of Science courses when discussing topics like the underdetermination of theory by empirical data (cosmological models of the solar system, in this case). I think it really gets the point across when I tell them that one can design a geometrical model based on Ptolemy’s system that can describe any motion whatsoever; hence, the data (the observed motion to be described) can be accommodated by a myriad of incompatible models (differing numbers of epicycles, deferents, equants, etc.) and the complexity of the data won’t solve the problem (what’s more complex than a Homer Simpson motion?). I also use it as a nice example showing how sophisticated Greek geometry was back then.
So enjoy (like me!), or don’t (like many of my students, I’m sure).