For those readers not privy to the conversations that go on in the special region of the worldwide web known as the Twitterverse, we had a rousing discussion with Ben Cook (aka @Skitzzo), among others, about the merits of some newer stats contra some older stats. In particular, we noted that "Whereas once batting avg. was useful (walks weren’t a big part of the game), we now know that OBP trumps it." Boy, did that rattle some cages, not only of Mr. Cook, but several others who chimed in. Cook countered "OBP doesn’t trump BA, it’s simply a different measurement. Are you gonna use OBP w/ RISP? No, you use BA." To which we challenged, "How about this? You get a team of the top 9 players in BA, and I get the top 9 in OBP and we’ll see who wins in a 162-game series."

So, to follow through on our challenge, we’ve run some numbers. With the help of David Pinto’s lineup analysis tool, we were able to determine the maximum number of runs that each of our mythical lineups would generate. We’ll start with the All-Batting-Average team from the NL as of today:

According to the lineup-analysis tool, this team (arranged in an optimal order) would generate 7.020 runs per game (talk about a fantasy team!).

Now for the OBP team. To be sure, many players who hit for a high OBP have a concomitantly high batting average; indeed, three of the players are also on the batting-average team (Pujols, Ramirez, Wright):

This team would generate 7.553 runs per game.

So over the course of a 162-game season, the OBP team would score about 86 more runs, or 8.6 more wins. Thus, in a head-to-head scenario, the OBP team would have an 85-77 record, and the BA a 77-85. We expect that the AL would yield a similar result.

This, of course, is hardly earth-shattering to anyone who has read about the importance of OBP to scoring runs. So why do people still cop to batting average (and other antiquated stats)? Several reasons, probably, but mainly because it’s "the way we’ve always done things" in baseball. It began in 1876 (or earlier) as a reliable measurement of hitting skill because it was a simple way to keep track, and, basically, it was the first of its kind. Plus, walks were a relatively minor component of batting (the NL leader, Ross Barnes, had only 20) and sacrifice flies and hit-by-pitches weren’t tracked. But even as early as 1910, FC Lane was complaining that BA was "worse than worthless" (even we haven’t said something so incendiary!) because it didn’t distinguish between singles and extra-base hits. Thus began an early form of Linear Weights.

Over the years, walks have become a more significant and strategic part of the game (and fielding has become more regular), and observers have stood on the shoulders of pioneers like Lane to continually question and improve upon the game’s statistics. If it’s any comfort to those of us who aren’t afraid of new and sometimes improved ways of looking at the game, the debate is not new. There have always been and probably always will be naysayers. In Lane’s time, according to The Numbers Game, it was writers like William Phelon of the Cincinnati Times-Star, who claimed "The present system [batting average] is about as good as any, and it seems really impossible" to improve. Sounds a lot like the curmudgeonly argument today. We’ll let Mr. Lane’s rebuttal stand for our own: "There are many men who are fond of making an unsupported statement under the apprehension that they have posited an argument … Let’s not make any effort to improve the present system which Mr. Phelon admits is grossly inaccurate. If there is any logic in this contention it escapes our feeble intellect."

This entry was posted by Pip on Wednesday, August 19th, 2009 at 4:08 pm and is filed under analysis. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.