Chad caught a funny error in this BIll James piece on Slate. (He meant "more than 1/10 the number of seconds," not "more than ten times the number of seconds.")
Bill James is arguably the most misunderstood figure in baseball analysis. He doesn't pretend to have all the answers, but then doesn't pretend that anyone else does either. (Think about how he went against the grain on the question of Pete Rose's guilt. It just dawned on me today how perfectly consistent that is with the rest of his analytic approach.) That level of humility is the perfect starting point for scientific inquiry.
On the other hand, his actual forays into number crunching are frighteningly ad hoc and sometimes more complicated than they needed to be, a conspicuous result of a human being (albeit an insightful one) doing a lot of trial and error until he trips over something that works (but then of course tests it further to see how robust it is).
Never trust any Bill James formula until you've sanity-checked the explanation behind it (unlike some of his peers, he uses plain English very well and his work is reasonably transparent). On the other hand, that's exactly the advice he'd probably give himself. He'd probably chide a few other stat geeks for pretending to be a lot more certain of the future than is plausible.
His sense of humor is also marvelous, but perhaps the most misunderstood element of his public identity. Given his very subtle humility, I'm sure his equally subtle dry wit is often grievously mistaken for crankiness, and his self-deprecation mistaken for everyone-deprecation.
Posted by Matt Bruce at March 17, 2008 11:40 AM