Myth: Most of the Numbers We Care About Matter

A long time ago, I wrote on this blog that so many statistics are meaningless, but I was only really scratching the surface of the real problem. I was talking about the "win" statistic. Well, it ends up that most most baseball statistics we care about really mean nothing, because they're based on too small a sample size to point to anything but pure luck.

I had always mildly suspected this, reading Bill James baseball books before I was 10 years old, but never really pursued it. For some reason, I had never read Moneyball by Michael Lewis until this week, and so I never really pursued my belief about the bullshitidness of statistics until now. Call me a Billy Beane convert, call me whatever you want, but let me tell you, the arguments in Lewis's books aren't just arguments: They're the truth.

I'll do this slowly, but I'm going to try and start taking a different kind of look at baseball, a look that puts a low Earned Run Average on par with a winning lottery ticket. My last blog entry, which talks about aces going against other aces, is still typical of something I'll write about. You know why? Because most of the crap we've learned about baseball, we've learned from uninformed, babbling sportscasters. Think about that for a second.

Think about that for a while, actually, because I'm done here.