Quote Originally Posted by Montmorency View Post
There's a (metaphysical) difference between assembling data and variables to project outcomes in an abstracted system (this candidate would hire that advisor, that candidate would prioritize holding rallies...) and retroactively adjusting or eliminating observed features to reconstruct an alternative present.
There is certainly a difference, but it is not so obvious that it is one that matters to what we were looking at.

A better defence of the methodology than that I've provided thus far may require some serious effort. Maybe later...