Anatomy of the Outfieldby Jeff Sackmann
November 22, 2006
Center fielders are better defenders that corner outfielders, right? Of course they are. While many center fielders don't have the strong arms of their teammates in right field, it's a given that the best athlete in the outfield mans center. At least in this instance, the conventional wisdom is right.
Every offseason, there are inevitably discussions of who can or should switch outfield positions; Alfonso Soriano might try his hand in center and Ken Griffey Jr. may finally move to a corner, just to take two examples. Such moves are usually decided based on visual evidence: so-and-so has the athleticism to make it in center, or so-and-so's arm is best suited to right.
It's certainly true that not every player adjusts to different positions in the same way. (More on that later.) But that doesn't mean we can't try to get a better grip on the difference in difficulty between each position.
To do that, I used a Hardball Times favorite: David Gassko's Range statistic. If you're interested in the details, follow the link; for the purposes of this article, it's enough to know that Range estimates how many plays an average fielder at each position would make and rates each player based on how many plays they made. Using Retrosheet play-by-play data from 2003-05 to look for players who split time between outfield positions, I put together a meaningful sample to work with (over 50,000 innings).
Let's See These Numbers
On average, the difference between an average center fielder and an average left or right fielder is about 15 plays; that's about 10 to 12 runs, or approximately one win. Another way to think about that is that you gain one win above average by sticking a middle-of-the-pack CF defender (2006 So Taguchi, for instance, according to David Pinto's PMR) at a corner position for 150 games. On the flip side, you'd lose a win if you put an average corner guy (say, Luke Scott) in place of an average center fielder in center for 150 games. To perhaps belabor the obvious: regardless of where in the outfield they play, Taguchi is about one win better on defense than is Scott.
Incidentally, the same translation technique suggests that there's no discernible difference in quality between average left field and average right field performance. My results were much too small to be meaningful, but they did indicate that left fielders are very slightly better.
That 15-play swing is not much less than the difference between average defense and the best or worst performance in any given year. (Because center fielders cover more ground, the difference is a little greater there than in the corners.) These numbers track intution: an average center fielder is about equivalent to the best corner outfielders. Similarly, an average corner outfielder would be near the bottom of the heap among regular center fielders.
So, How Will Soriano Do in Center?
While the results were extremely consistent from year to year, they were much less so at the player level. Part of that is a sample size problem: by definition, anyone included in this study played no more than half a year in either center field or a corner position. A full year of data can result in erratic enough results; using only 300 or 400 innings is asking for trouble. Among those players who logged at least 100 innings at each position, the correlation between their center field and corner defense numbers was very nearly zero.
However, the persistence of the same number—a difference of 2.5 to 3% of plays made relative to positional average—indicates that we can make educated guesses. We don't have Range data for 2006 yet, but using PMR as a proxy, Soriano was seven plays above average in left. If he switches, we can expect a respectable but below average -8 from him, a bit worse than Torii Hunter's '06 performance.
Translating in the other direction raises an interesting caveat. According to PMR, Coco Crisp was among the best defensive center fielders in the game last year, at +14, or about +20 per 150 games. If the relationship between center field play and corner play is linear, it would suggest that Crisp could be well above +30 in left or right field. While I suppose it's humanly possible to make 30 more catches than the average left fielder, it would be a historically great season. Put another way, it would be at least as good as Manny Ramirez is bad.
It would be reasonable to assume, then, that there's a cap on just how good a corner outfielder can be. Part of what's causing that upper bound, I think, is the presence of a good defender in center: no matter how good, say, Ryan Langerhans is, Andruw Jones will call him off at least a couple dozen times per season. Having good speed or a quick first step is only so useful when you have less ground to cover.
I doubt, however, that there's an equivalent lower bound for center fielders, at least in theory. On a major league team, there is a limit: anyone 30 plays below average (not named Ken Griffey Jr.) will find themselves playing elsewhere long before they log enough innings to lose their team two or more wins with the glove.
References and Resources
This and hundreds of other projects wouldn't be possible without Retrosheet. For 2006 defensive numbers, the best bet online is David Pinto's Probabilistic Model of Range (PMR), which he has begun to publish in the last several days.
Jeff Sackmann is the creator of MinorLeagueSplits.com. With Kent Bonham, he founded CollegeSplits.com. Jeff and Kent blog about college baseball and the draft, and you can follow them on Twitter for bite-sized snacks of minor league and college stats. Jeff also has an email address.