0 / 60 seg.

We want to pick one that creates the most constraints on its own, and thus gives us the most information when it turns out to be right or wrong.