> > 1964: nominate Southerner. Win.
> > 1968: nominate Northerner. Lose.
> [Etc. etc. until 2000.]
>
> Sampling fallacy: JFK was from the North, and so was FDR (a total of five
> terms). Truman was from Missouri -- in some ways, a border state, but
definitely
> not deep South.
Strictly speaking, there is no "sampling" involved, since the observation is a "census" (i.e. a total survey) of region of origin of successful US presidential candidates since 1964.
The cut off point is valid on the grounds of relevance, especially in light of social changs since that time.