The Futility Infielder

Salon's Allen Barra weighs in with an interesting piece on Barry Bonds. As we all know, Barry is well ahead of Mark McGwire's 1998 home run pace, having tallied 37 home runs (unless he's hit another since I had my morning coffee) in 70 games. But Barra proposes that Bonds may be on his way to the best season ever, based not on home runs but on another stat.

Barra uses a stat called SLOB, which is Slugging Percentage times On Base Percentage. He details the history of the stat, which was independently developed by two researchers in the '70s. Then he goes on to claim that SLOB "does a remarkable job of calculating actual team runs". So remarkable, to Barra at least, that he moves straight to an assertion that Bonds' SLOB of .4418 means, roughly speaking, that he has "created" .4418 runs per at bat, or 44.18 per 100 at bats.

Whoa, hoss! Let's back up. Over the past two decades, many statistical ways to accurately measure offensive contribution have sprung up. The best of these do a very good job of projecting how many runs a team will score based on certain offensive factors, such as hits, walks, and total bases.

Bill James devised a formula called Runs Created which is probably the most famous, thanks to James's role as a pioneering sabermetrician (sabermetrics, an acronym derived from the Society for American Baseball Research, is, in James's definition, the search for objective knowledge about baseball). Other formulas include Pete Palmer's Batting Runs, Paul Johnson's Estimated Runs Produced, Jim Furtado's Extrapolated Runs, Clay Davenport's Equivalent Runs, and Keith Woolner's Value Over Replacement Player.

Some of these are formulas are easier to understand than others. I've always been partial to James' system for a number of reasons: they are easy to calculate, their methods demonstrate an implicit understanding of the mechanics of offense (getting on base and advancing runners). James's published work always took the time to clarify the details of his methods. But the recent work in the field seems to have moved past James's formula, and even the man himself conceded the limitations of his work.

SLOB actually turns out to be something of a shorthand of James' formula; when you multiply SLG * OBP * AB (the number of at bats) you get a pretty decent estimate of the number of runs a team will score. Looking at the two leagues last season:

         SLOB Runs   Actual   Error
2000 AL   12039      11995     3.6%
2000 NL   12958      12976     1.7%
1999 AL   11798      11725     6.2%
1999 NL   12983      12966     1.3%

As I have a day job which I'm supposed to be doing right now, I'm not going to take this much further except to say that it works pretty damn well for a lazy person to calculate on a lunch break. It's not the most accurate estimate, just a very easy one to deal with, and it also meshes very well with another stat growing in popularity, OPS, which is OBP + SLG. OPS has entered the mainstream thanks to the work of writers such as ESPN's Rob Neyer. It's a handy metric, but this appears to be even better and no more complicated.

So anyway, Barry Bonds, at .4418, is right now second only to Babe Ruth's 1920 season, at .4506. The only other person besides the Babe whose SLOB is in the ballpark is Ted Williams, with .4049 in 1941. That's some pretty good company—the two greatest hitters ever, in fact. It's not likely Bonds can maintain his pace, in which he's hit almost twice as many home runs as singles, but he's got a clear shot at a season for the ages.