I’m not here to rehash the debate on how much control pitchers exert over balls-in-play.
I will, however, point you to the seminal works of this emerging sub-field of analysis. Here’s Voros McCracken’s original “Shot Heard ‘Round the Internet” piece that originally appeared on Baseball Prospectus, and here’s Keith Woolner’s counterpoint piece, which also appeared on these pages. For what it’s worth, I ascribe to the more measured, Woolnerian approach that says pitchers exert some but not much control over balls in play. Regardless of where you fall in the continuum, the preponderance of the evidence suggests that pitchers have, at best, only a modicum of influence over what happens to the ball once it leaves the bat.
This will be needlessly remedial for many of you, so pardon the following backgrounder… One of the outgrowths of DIPS (Defense Independent Pitching Statistics–a collection of metrics developed by McCracken and the practical extensions of his initial work in this area) is paying particular attention to a pitcher’s batting average on balls in play (BABIP). The upshot is that if a pitcher is wildly exceeding or falling short of the league mean in BABIP, he’s likely the victim/beneficiary of some combination of luck and defense (or the lack thereof). This is useful information because it allows us to:
- Identify performances that are out of step with key peripheral pitching indicators (strikeouts, walks, home runs allowed, groundball/flyball breakdowns), and…
- …Project how a pitcher will fare in the future given alterations in luck and defensive support.
This season, the league average BABIP is just a tad more than .300. Knowing this, we can look at a few notable performances that significantly deviate from this number. Let’s get to it…
One that stands out for me is Roy Oswalt of the Astros. On the surface, Oswalt is having a solid season: 130.2 innings, 3.65 ERA. Below the surface, however, he’s having a stellar one: 8.0 K/9, 4.3 K/BB, only seven homers allowed. The divide can be explained, of course, by his BABIP, which is a hefty .337 (second only to Mike Hampton‘s .344 among NL qualifiers).
Since arriving in the majors in 2001, Oswalt has put up seasonal BAsBIP of .297, .307 and .289 in previous years, so his mark in 2004 is not in keeping with his past performances. One explanation, as mentioned earlier, is poor defense. With Craig Biggio miscast in center for much of the year and an aging Jeff Kent at the keystone, it’s intuitive to think diminished range at a couple of key positions is possibly to blame. To explore this further, let’s see how Oswalt and his label-mates in the Astro rotation have fared in terms of BABIP, with their GB/FB ratios included:
Pitcher BABIP GB/FB Clemens .276 1.63 Miller .262 0.79 Oswalt .337 1.26 Pettitte .276 1.90 Redding .333 1.10
Nothing terribly informative here. Clemens, Miller and Pettitte have all posted comfortably better-than-average BAsBIP while having little in common in terms of groundball-flyball tendencies. Redding’s the only one with a BABIP comparable to Oswalt, and while the two do have similar GB/FB ratios, it makes little sense to suggest that an Astro pitcher is most likely to be shanghaied by his defense if he’s somewhere in the middle in terms of groundballs and flyballs.
In any event, Oswalt is not an extreme worm-burner on the mound (like, say, this guy), and the presence of Carlos Beltran will only help his cause in the second half. With a better team defense behind him and, one assumes, better luck, Oswalt is poised to post a better ERA over the season’s final two-and-a-half months.
- You might look at Tim Hudson‘s ERA of 2.98 in tandem with his basement-level K rate of 4.3 per nine and conclude he’s been uncommonly lucky this season. But that hasn’t been the case. His BABIP of .302 is actually a couple of ticks worse than average this season. While Hudson’s paltry whiff numbers are troubling on one level, he does do quite well in other areas. To wit, in 108.2 innings this season, Hudson’s given up only three homers (!) and 24 unintentional walks, and he’s posted the fourth-highest GB/FB ratio in the AL (2.39).
- Jason Schmidt is a legit Cy Young contender, and there’s certainly nothing wrong with most of his peripherals (3.2 K/BB and 9.9 K/9). However, Schmidt’s BABIP of .225 is 25% better than the league average and easily the lowest mark of any NL qualifier. A look at the BAsBIP of the other Giant starters (Hermanson: .308, Rueter: .309, Tomko: .315 and Williams: .290) suggests that’s not likely to hold up as the season deepens. Expect that ERA to rise down the stretch.
- Jose Mesa is just perfectly bizarre this year. His K/BB of 2.0 and K/9 of 5.4 are, respectively, middling and healthily sub-optimal for a closer, but that ERA of 2.45 means he’s been flushed with good fortune this season. Except he hasn’t. Mesa’s BABIP is a grisly .342. True, he’s keeping the ball in the park (one homer in 36.2 innings), but he’s also coughing up a double every three frames. Heck, I don’t get it. The only thing I do know about Mesa this season is that he should be traded to the Cubs and placed in a role of unspeakable importance posthaste. (Ed note: Had to wipe Dayn’s Cardinal-red drool off the screen before publication.)
- Better days ahead for Jake Peavy? With a 2.43 ERA, he’s probably not wont to complain about his hole cards this season, but Peavy’s BABIP of .343 tells a different story. If Peavy can continue fanning almost a batter per inning, one would think his luck would improve as the season wore on. Then again, the Padres’ poor performance covering the vast Petco Park outfield might continue to undermine Peavy. Nevertheless, he’s got an incredibly promising future.
All of this touches on why I’m rapidly losing my already withering affection for ERA as even a rough thumbnail estimate of pitcher quality. Michael Wolverton has written cogently on the pitfalls of ERA in the past. I’d go further by saying if you don’t take a pitcher’s BABIP into account, ERA shouldn’t be anywhere on your radar when making player evaluations. I’d liken ERA to the RBI–useful only at the margins, awfully prone to contextual distortions.
When we judge a pitcher’s performance, we need to grant primacy to the things over which he has the most control: strikeouts, unintentional walks, home runs and groundball-flyball tendencies. To credit or penalize a pitcher for how his defense is performing, how luck fancies him or how he wields a dubious skill of marginal influence is to do a disservice to the spirit of analysis.