This post begins with a fifty-year-old memory. While I’m certain I’ve got things essentially right, I’m nearly as sure there’s some detail that’s wrong. That’s OK; the main point is the memory.
My first baseball glove had originally been my father’s. Since Dad was a kid in the 1930s, you can imagine what that was like: An uncomfortable pancake thing which didn’t appreciably improve my (negligible, to be sure) fielding prowess. Dad eventually bought me a better glove: A Ted Williams model, probably a late Wilson model before he became part of the Sears empire. Teddy Ballgame not being known as a particularly good fielder, the packaging talked mostly about his accomplishments with a bat. This is where I first encountered On Base Percentage, as the summary of his 1941 season mentioned his .553 OBP.
"What," I asked my father, "is this mystery statistic?" Dad actually had a response, which I don’t recall in detail but was certainly along the lines of "It’s what his batting average would be if you included his walks." In retrospect, that Dad knew what OBP meant is probably as delightful as the glove. Of course, I didn’t recognize that at the time.
Fact is, when I was twelve I didn’t know much about Williams. I knew he (had) played with Boston, and that Topps seemed to think he was the best ballplayer of his generation (evidence was awarding him the #1 card in 1957 & 1958, which certainly seemed like an endorsement). I discounted that endorsement as ignorant bias. Nowadays, like everyone else, I’m certain they likely had it right. The Splinter was Splendid.
But today’s essay’s about Sabermetrics, not about old ballplayers. I’ve been rereading Baseball Between the Numbers by the Baseball Prospectus staff, which was published a few years back. It’s set me to thinking about The March of On-Base Percentage (as Alan Schwarz titled a chapter in his book The Numbers Game), and about other things sabermetric.
I was slightly aware of Bill James some years before he became famous, I think because I devoured The Sporting News. When Bill’s Ballantine-published 1982 Baseball Abstract hit the newsstands, I immediately purchased and devoured that. I’ve have been collecting similar books ever since. My library now contains nearly everything James has published, all of the recognized Sabermetric classics, complete or nearly-complete runs of most James-influenced publications, and other books which resemble those. The quality’s quite variable, and there’s no book I fully agree with. But I’m definitely part of the sabermetric camp.
Anyway, the thing which strikes me about Baseball Between the Numbers is that it’s largely grown obsolete in just a half-decade. For almost 20 years, baseball management largely resisted serious statistical analysis. Management largely consisted of former players, and few were inclined to take outside analysis seriously. This was partly willful blindness–“He never played the game”–and partly statistical ignorance. But a generation later, baseball’s management’s (unexpectedly) become more businesslike, and a newer generation of baseball players–and coaches and field managers–includes a sprinkling of folks who grew up reading James, Pete Palmer, or authors influenced by James and Palmer. Some of those players have moved to front office jobs. And while fans still have blind spots, they’re generally more aware that many numbers are influenced by ballpark and batting order, and that there are legitimate reasons to debate baseball’s accepted wisdom.
The result is a new baseball culture. While people still argue about the value of statistical analysis vis-a-vis other forms of baseball knowledge, basically everyone agrees with the fundamental tenets which have driven sabermetric analysis and many managers, announcers, and fans are comfortable using some of the sabermetric toolkit. OPS and SLG are commonly understood. In the front office, it’s clear that many teams’ decisions are informed, if not driven, by Runs Created-like formulas and more realistic analyses of the context and consequences of their options. The result’s better baseball on the field, and better decision-making at all levels.
There’s no calling in life which isn’t influenced by fashion. Baseball’s new sabermetric fashion is, on the whole, a good thing, and benefits all of us who care about the game.
OK, back to Baseball Between the Numbers. Reading it has been a bit frustrating. Because I’ve followed these discussions for many years, much of what’s covered here seems pretty basic. The book has some good research, but on the whole I’m clearly not–and never was–the target audience.
There’s another thing, too. Things have moved beyond the issues discussed in this book. The cutting edge research nowadays is based on play-by-play, and even pitch-by-pitch, data. While the authors make some reference to play-by-play, on the whole the book’s based on summary statistics. So there’s a sense that they were summarizing the state of the art just as the research practice moved somewhere else. That was certainly worth doing, but clearly more valuable to a new reader than to someone who’s been following the discussion.
March 30th addition: BP’s recently published a two-volume "Best Of" book, drawn from their website; it’s heavily weighted toward recent work, so likely they’re in agreement about this. They’re also apparently working on a successor to Between the Numbers. I’m looking forward to it.
(I wrote this essay last July, filed it, and forgot it. Posting it now, with slight editing, because it seems worthwhile….)
