King James

No, not the basketball player in Ohio, the baseball writer/researcher who apparently is the sport's John the Baptist. At least that is what he seems to be based on this review by Matt Welch of the book Moneyball.

James approached statistics not as a method to generate trivia answers -- of the who holds the rookie record for home runs among players whose last name begins with "ph" variety -- but rather as a tool to ask fundamental questions about how the game is played and won, and as a bludgeon to mock baseball observers who used hoary clichés that couldn't possibly be true.


The age loves an iconoclast so it is not surprising that James's renown is growing. And i don't doubt that his research has deepened our knowledge of the game. But in his desire to praise James, Welch caricatures the old guard and gives too much credit to the obsessive researcher.

James was the first to really hammer home the notion that different ballparks have vastly different effects on statistics (Fenway Park and Wrigley Field were great for hitters, the Astrodome and Dodger Stadium protected pitchers).

Look, the Baseball Abstract first came out in 1977. Yet, even children's books on baseball in the 1960s noted that Dodger Stadium was a hard place to score runs and the Astrodome was cruel to home run hitters. I know because i was a kid then and read everything i could get on baseball. The adult sports writing was even better. Those old farts noted esoteric stuff, like Forbes Field was a great place for line-drive hitters (like Roberto Clemente) but playing there hurt the home run production of Willie Stargell. And they even were willing to take on mythic heroes-- some pointed out that Yankee Stadium offered a sweet, short right field for left-handed pull hitters like Ruth and Gehrig despite the deep centerfield and power alleys.

So the pre-James dark ages were not quite so dark. And maybe the Jamesian revolution is not quite so revolutionary. Oakland has had a good three year run, but they haven't been able to make it to the World Series using Billy Beane's new methods. So Welch is reaching a bit when he pronounces

What lessons can we learn from this tale? That the pursuit of better information will eventually unearth discrepancies and irrationalities, even in a field as seemingly well-studied as baseball. That the gatekeepers of information and judgment will instinctively and defensively protect their turf, rather than question their own legitimacy. That intelligence and passion can still win in the end, especially if they take advantage of the networking power of the Web.

While it is possible to "unearth discrepancies and irrationalities" if you spend enough time at the computer, leveraging them in the real world is much harder. Long-Term Capital Management thought they could do that in the capital markets. Instead they crashed as new discrepancies and irrationalities appeared that they had not expected.

Another blogger noted that

Bill James started the whole Sabremetric deal for one reason and one reason only: so he could get an edge against his opponents in "tabletop baseball games."

Now, Strat-o-matic is a great gaming system. But it isn't real baseball. For one thing, the rules are designed to make a given "player" perform as he did over the course of the whole season. Unfortunately, the manager doesn't have a crystal ball which shows him what a player will do over the whole season. He has to make decisions based on an unfolding stream of events. In real life, each new game can rewrite the statistical probabilities of a player or team.