“Driving in runs has as much relevance as driving in traffic.”
—Joe Sheehan, February 1993
Some fourteen-plus years ago, I had a problem. I was in college, I had a girlfriend in another city, and long distance phone calls were really expensive. I had to find some way to keep the girl, but not be sold into indentured servitude by the phone company the second I graduated. Someone clued us in to the fact that we could communicate for free through our respective colleges’ computer systems. She got a job at her school’s computer lab, and I got my first email account.
Now, you may be asking, what does any of this have to do with baseball? (At least, those of you who aren’t wondering what a long distance phone call was.) Shortly after I discovered the Internet, my friend Joe Sheehan suggested I tune in to a usenet discussion group called rec.sports.baseball. I did, and one of the first threads I looked in on to see what my friend was doing on the Internets somehow involved a heated debate about Mark Grace and featured the above quote, which has been stuck in my mind ever since.
Back then, fans of the RBI were the enemy of sabermetric thought, a stat whose poster boy was one Joe Carter. On one end of the divide, Carter was considered a “proven run producer” who collected more than 100 RBI in a season ten times despite low batting averages and a tendency to draw walks about as often as Carrot Top made people laugh. On the other end, Carter’s detractors pointed out that his OBP only broke .320 twice in his career, that he devoured upward of 450 outs per season, and that a lot of his RBI were the result of having good table-setters in front of him in the lineup (notably Brett Butler and then Roberto Alomar). The RBI became emblematic of all the things that were wrong with traditional statistics-it gave a player credit for his teammates’ accomplishments, it didn’t provide an accurate appraisal of the player’s actual skills, and it fostered exaggerated ideas about players’ clutch ability.
I agreed with all these arguments, and enjoyed the incredible energy that was brought to bear challenging baseball’s statistical orthodoxy. But at the same time, I guarded a secret: I…love the RBI. There, I said it.
Now, before you petition Nate to have my stathead license revoked, hear me out. Like any relationship that lasts, my feelings for the Run Batted In are informed by an awareness of its limitations. Last week we talked about the prospective and retrospective viewpoints from which we can look at statistics and performance. The prime objections to the RBI are based on the prospective point of view-the idea that you can see a player’s RBI performance this year, and draw conclusions about his “talent” for batting in runs. The fact is, however, that a “talent” for accumulating RBI is heavily distorted by the number of baserunners who are on when the batter comes to the plate, and what bases those players are on.
For example, take Joe Carter’s 1997 season, which was identified in Baseball Between the Numbers as the worst 100 RBI season since 1972-a season when Carter actually performed below the replacement level (by VORP), thanks to a .234/.284/.399 stat line. If we look deeper, we can see that despite his weak rate stats, Carter’s RBI total that season was driven by large numbers of runners on base in front of him (437, good for 30th in the majors) particularly at third base (97, fourth-highest in the majors). Now, had Carter been on the Milwaukee Brewers instead of the Blue Jays, he might have found himself in the same situation as Jeromy Burnitz did that year: Burnitz played only four fewer games than Carter, but had 110 fewer plate appearances, and 69 fewer baserunners on in front of him. Burnitz did not have a 100 RBI season. Viewed prospectively, if you called Carter a “100 RBI-type” and Burnitz a “85 RBI-type”, you’d be ignoring their vastly different opportunities the two players were given to drive in runs. The following year, Burnitz led the majors in runners on base in front of him, and he collected 125 RBI.
So prospectively, the RBI isn’t useful. However, viewed retrospectively, RBI numbers have value, or at least should draw interest. The RBI isn’t as well-rounded as a number of other metrics-VORP, Equivalent Runs, Linear Weights, Runs Created, to name a few-but it does serve as a record of the events of a ballgame. It may be happenstance when and how often a batter gets the opportunity to bring his teammates home from the bases, but the event does happen-you can tell by the cheering. My own unscientific observation is that when you’re headed home from the ballpark, the first thing that friends, fans who weren’t lucky enough to attend, and passers-by will ask you is the score, quickly followed by “Who did the damage?”-who collected those RBI.
So I’ll admit that RBI are the second thing I check in the box scores each day (after the pitching lines), and I’ll ask you not to judge me too harshly. I know they’re flawed, but if lovin’ the RBI is wrong, I don’t want to be right.
- The main statistic of the report is OBI. That’s RBI less home runs-since a player has the opportunity to drive himself in from home on every at-bat, the player’s RBI on homers is taken out of the equation.
- The report breaks down the total runners on base for each batter (ROB), he total plate appearances the batter had with runners on (PA_ROB) and which bases the runners were on (R1 means runner on first, R2 means runner on second, etc.).
- The report not only shows you the total percentage of the batter’s base runners who were batted in (OBI%) but also breaks down the percentages according to which base they were on (RxBI%, where x = the base the runner was on).
- A high OBI%-usually good enough to place a player near the league lead at the end of the season-is about 20%. The highest since 1959 was George Brett in 1980; he plated 26.9% of his baserunners. Brett’s 118 RBI that season were good for third in the majors, behind Cecil Cooper and Mike Schmidt.