Arguments can prove too much as well as too little.

Selection bias doesn't explain everything, but it explains more than it seems like it should!

I notice an annoying empirical regularity. (The films I want to see at the cinema are the ones most likely to be sold out. Buses take too long to arrive, and when they do they come in twos or threes. Things go wrong most often when I'm trying to get something done in a hurry.) Chances are it's neither coincidence, nor reality magically conspiring against me — instead there's usually some mundane, obvious-in-hindsight explanation. The game theoretic analogue of this rule is Scott Aaronson's observation that a situation often sucks because its not sucking wouldn't be a Nash equilibrium.