Remember Donald Rumsfeld’s speech from 2007 on the things we know and the things we don’t? His enumeration of known knowns, known unknowns and unknown unknowns garnered ratings from trivial to deeply meaningful. Fishing for truth in such murky waters is not just a governmental problem, but help (to a certain degree) is at hand. So Donald, how about a little sensitivity analysis to at least see what happens if the variables we know about deviate from expectations? And better still, how about some uncertainty or importance analysis to go with that?