One of my pet theories is that math and (applied) statistics require very different brains. People whose brains are wired for math make poor (applied) statisticians and people who are really good at stats tend to be poor at math.

This is partly an empirical observation and partly, I think, is a consequence of the fact that math deals with "hard" objects (e.g. numbers) that might not be known at the time, but they are not going to mutate and change on you, while statistics deals with uncertainty and "soft"/fuzzy/nebulous objects (e.g. estimates). Moreover, for applied statistics the underlying processes are rarely stable and do mutate...