Across the (defensive) universe

Even if you’ve never heard the words “defensive spectrum” before, I’d bet you have a familiar grasp on the concept.

Because some positions are harder to play than others, major league teams tend to get less offensive production from those positions. That’s why first basemen and designated hitters tend to bat near the middle of the lineup and you usually find catchers at the bottom. It’s also just one reason why someone like Joe Mauer is so valuable.

The precise distribution of offensive talent across the diamond varies from year to year, but it is reasonably consistent. It doesn’t take advanced analysis to recognize that catcher, shortstop and center field are hard to play, while the others are easier (to varying degrees). Usually average offensive production from those positions reflects the difference.

The 16-year-old SS/P

Think back to Little League. The star, both offensively and defensive, is often the shortstop, who was also the best pitcher on the team. The right fielder was some kid who ended up on the debate team a couple of years later.

In college ball, we still see some of that. A disproportionate number of prospects are center fielders and shortstops. Many of those guys won’t last at those positions, of course, but because of the lower level of play, they can handle the tough positions. In time, many of those slugging shortstops become third basemen or even corner outfielders.

All of which leads to the question that sparked this article: What does the defensive spectrum look like at lower levels?

Spectra

Let’s start with a reference point. Here is the average OPS for each position in the major leagues last year. I’ve also included a percentage, reflecting how much better or worse that production is than the league average.

Position    OPS  % Diff           
C         0.717   -6.5%  
1B        0.846  +10.3%  
2B        0.752   -2.0%  
3B        0.757   -1.3%  
SS        0.721   -6.0%  
LF        0.781   +1.8%  
CF        0.749   -2.3%  
RF        0.791   +3.1%  
DH        0.782   +2.0%
Total     0.767

Rather than jumping straight to college ball, let’s work our way down. Intuitively, it seems like the differences between, say, first base and shortstop should become steadily less pronounced. If we had a big pool of Little League data (heaven help us!), we might even see some of the differences reversed.

Here’s the same data for the high minors (Double-A and Triple-A) in 2009:

Split    OPS  % Diff  
C      0.709   -3.4%  
1B     0.795   +8.3%  
2B     0.717   -2.3%  
3B     0.739   +0.7%  
SS     0.682   -7.1%  
LF     0.775   +5.7%  
CF     0.745   +1.5%  
RF     0.765   +4.2%  
DH     0.743   +1.2%  
Total  0.734

No clear trends here. There’s some movement toward the center—a little more offense from catchers (perhaps the guys who can hit but don’t have the game-calling ability required in the majors) and a little less at first base—but shortstop drops back from the pack, and the entire outfield is above average.

Continuing our journey, let’s look at the full-season Single-A leagues:

Split    OPS  % Diff  
C      0.684   -3.9%  
1B     0.756   +6.4%  
2B     0.701   -1.4%  
3B     0.711    0.0%  
SS     0.665   -6.5%  
LF     0.729   +2.5%  
CF     0.712   +0.1%  
RF     0.732   +2.9%  
DH     0.706   -0.7%  
Total  0.711

Except for the persistently low level of shortstop offense, this is starting to line up with our intuition. It’s not as odd as it might seem that DHs are below average—at these levels, many teams don’t have specialists as designated hitters. Instead, the spot is filled by a potpourri of guys in position battles who need at-bats and dinged-up hitters taking a break from the field.

One last step before we get to college. Here’s the spectrum from the 2009 rookie and short-season minors:

Split    OPS  % Diff  
C      0.682   -2.8%  
1B     0.761   +8.4%  
2B     0.694   -1.1%  
3B     0.701   -0.2%  
SS     0.681   -3.0%  
LF     0.720   +2.6%  
CF     0.684   -2.5%  
RF     0.711   +1.3%  
DH     0.697   -0.8%  
Total  0.702

Finally, some movement at shortstop toward the average. If we had to draw conclusions based purely on the data I’ve presented so far, it would seem that a large number of players who can’t hack it defensively at short are moved off the position some time between rookie ball and A-ball.

Looking at these numbers for the first time, I’m a bit surprised that there isn’t more drift to the middle. I expected more evidence that rookie and A-ball leagues are full of players who aren’t going to last at their position. That appears to be true of catchers, but the trends are more ambiguous elsewhere on the diamond.

How does college compare?

Whenever you’re comparing college and pro ball, it’s usually a safe bet to assume that college is different (if not out-and-out weird). If the defensive spectrum in the low minors is more like that of the majors that I had expected, surely it must be different in college.

Sure enough, there’s some weirdness on display. Here is the defensive spectrum for all of 2009 Division One:

Split    OPS  % Diff  
C      0.828   -1.8%  
1B     0.895   +6.2%  
2B     0.826   -2.1%  
3B     0.858   +1.8%  
SS     0.821   -2.7%  
LF     0.847   +0.4%  
CF     0.857   +1.6%  
RF     0.858   +1.7%  
DH     0.819   -2.9%  
Total  0.843

Catchers hitting better than second basemen and designated hitters? Left fielders barely above average?

Some of the usual trends are still on display, though. Shortstops are near the bottom of the list, and first basemen are comfortably camped at the top, mashing at about the same rate their full-season A-ball counterparts are. At the same time, there’s a clear drift toward Little League-style non-specialization, with center fielders above average and softer extremes in general.

But wait. If we want to start venturing conclusions based on the differences we’ve seen, we’re probably talking about the pool of draft-eligible, or even drafted, talent, and how it holds up in pro ball. Thus, using a pool as large as all of Division One is probably misguided.

What happens if we narrow the pool to the top eight conferences, the leagues from which a large majority of college draftees are drawn?

Split    OPS  % Diff  
C      0.838   -0.8%  
1B     0.915   +8.3%  
2B     0.827   -2.1%  
3B     0.847   +0.2%  
SS     0.821   -2.9%  
LF     0.868   +2.8%  
CF     0.861   +1.9%  
RF     0.854   +1.1%  
DH     0.802   -5.1%  
Total  0.845

Not much new to see here. It seems that these college powers, presumably the ones with the most recruiting strength, do a better job finding first basemen and left fielders who can mash, but fail at filling the designated hitter role. (Sometimes DH is a used as a spot in which underclassmen audition for other jobs. For a school that wants to win, maybe it shouldn’t be.)

Specialization or not?

Before putting this data dump to bed, let’s take a look at something a bit more complex.

We can interpret the drift toward the center in a couple of different ways. The first way is the simple “non-specialization” that I’ve suggested. At some level, the best players are the best athletes, and some combination of practice, interest and talent means that those “bests” are best at everything. The extent to which that is still true in college is unclear, though it’s suggestive that college center fielders hit about as well as corner outfielders, and catchers are so close to average.

The second way is to view the mass of college players as a heterogeneous group that needs to be broken down further. Perhaps if we isolated the best defenders at those demanding positions, we’d find that they are really good at everything. Or maybe specialization has set in by this point, and those great defenders are already lagging behind the pack on the offensive side.

That leads us to the final spectrum, which I’ll call “college defenders who we don’t know are bad.” Using the play-by-play metric that I’ve displayed in previous articles, I ranked all 2009 players with a low minimum of playing time and knocked out the bottom fourth. That left us with everybody who was either above average or only a run or two below average. (These numbers were regressed—given the sample size, there are a lot of college defenders in the range between -1 and +1 run.)

Here are the results:

Split    OPS  % Diff  
1B     0.890   +4.8%  
2B     0.821   -3.3%  
3B     0.855   +0.7%  
SS     0.826   -2.7%  
LF     0.837   -1.4%  
CF     0.862   +1.5%  
RF     0.850   +0.1%  
Total  0.845

I haven’t yet worked out college catcher defense, so it doesn’t shed any light on the nearly-average offensive production from catchers. Very little becomes clear for the other positions, as well.

What is most striking is how similar these numbers are to those in the previous two tables. If anything, it suggests that specialization is in its early stages in the college game. The average OPS for players above the bottom quartile of defenders is almost identical to the overall average. The top quartile is similar, as well.

There you have it. At every level, first basemen can really hit, but a shortstop who can is special. I’m guessing you knew that. But maybe you didn’t know to be impressed—or surprised, anyway—next time you see a college DH have a big night at the plate.