April 29, 2014

Painting the Black

Singles Falling Steady

by R.J. Anderson

The easiest way to explain regression to someone is to do so in baseball terms. Batters rarely threaten .400 these days because of the upped quality of competition. The inverse works, too; pitchers with high earned run averages are replaced before long because teams have capable replacements. There are exceptions to those explanations—Neifi Perez tallied more plate appearances in the majors than Hank Sauer did, after all—but they train people to think in a certain way.

That training forms the crux for nearly all arguments about pitchers and batting average on balls in play. Any pitcher whose true-talent level leads to an inflated BABIP, the thinking goes, will not last in the majors long enough for us to be sure. There are exceptions here as well. Glendon Rusch exceeded 1,400 innings with a .331 BABIP, and Zach Duke is nipping on his heels with more than 1,000 innings to his name. Most pitchers revolve around the .300 mark, but not all, which is why it can pay to investigate potential outliers.

It's too early to call Zack Wheeler a candidate, but his 2014 has been peculiar.

Despite a stellar start against the Marlins last Friday, in which he posted the second best Game Score of his career, Wheeler's rate statistics remain at odds. His strikeout and home-run rates are stellar, yet he continues to yield more than a hit per inning—that despite pitching in front of an above-average set of gloves, per park-adjusted defensive efficiency. Of course Wheeler is not the only starter having these issues. Madison Bumgarner and Jordan Zimmermann are in the same boat. The difference is those guys have life vests in the form of longer and more decorated careers. Wheeler has 22 starts to his name, and about a quarter of them suggest he's hittable.

Darn near every time Wheeler allows a hit, an argument ensues about whether he possesses bad luck or if there's something rotten about his game. So which is it?

Let's revisit the thought about pitchers with high BABIP. These guys tend to possess some combination of a few undesirable qualities: bad stuff, poor control and/or command, lackluster deception in his delivery, unconvincing sequencing, and so on. We know for sure Wheeler does not have bad stuff. When Wheeler was recalled Mark Anderson wrote that his arsenal "includes two plus-plus offerings and two pitches with at least average grades." The other characteristics, however, carry some weight.

Wheeler's fastball command is below average and his control can waver. Even in his quality outing against the Marlins, he still posted the fifth-worst first-pitch strike rate in a start this season (min. 20 batters faced). Wheeler is a fastball-first pitcher regardless, but his heater usage tops 70 percent once he falls into hitter's counts. Add in the tendency to miss his spots, often over the plate, and you have a recipe for some prolonged innings. And let's not forget that Wheeler tipped his pitches last season.

Put it all together and you have the makings for a solid theory on why Wheeler might be more hittable than the stuff suggests. But theory is just that until it's tested.

In an effort to figure out if there's something here, each of Wheeler's 31 hits allowed were charted. Among the variables recorded were his delivery (did he go with the windup, or did he pitch from the stretch?), the count, the pitch type, the intended location (as best as one can tell from the situation and the mitt), whether he missed the spot (again, as best as one can tell), the pitch location, the hit type and location, and the amount of luck perceived in the result. The luck factor here has nothing to do with the quality of pitch, but rather if the fielder could or should have made the play.

To state the obvious, there is some subjectivity here, but it's unavoidable without access to good batted-ball data (read: proprietary data). All we can do is acknowledge it and ask to the Ghost of Colin Wyers to forgive us. Now, on to the results:

• Wheeler allowed 17 hits from the windup and 14 from the stretch. In some cases, the pitcher is more efficient from one than the other. That does not appear to be the case here, at least not with regards to allowing hits.
• Twenty of the hits came on fastballs, six on sliders, three on changeups, and two on curveballs.
• About 19 of the hits were on pitches that missed the target. Though a few others could have been included as well, depending on the leniency involved. (For instance: a pitch that was in the correct lane, but a little high.)
• Think of the strike zone as a keypad. Eight of the hits came in the middle, or on the no. 5 key, while another nine came in the 4 and 6 zones. Six hits were recorded from the 3 zone, which is down and in to right-handed batters.
• Thirteen came on ground balls. The rest were in the air, either as fly balls or line drives.
• Fifteen, or about half, came on counts where there were more balls than strikes, including full counts.
• Lastly, the luck factor. Wheeler was deemed to have a little bad luck on eight hits. Not all of the plays were easily fielded, mind you. Those that never left the infield were automatically included, as were those that hit off a glove. Lastly, there were a few plays where defensive positioning was the difference between a hit and an out.

Here are those eight plays in GIF format, thanks to the talented Nick Wheatley-Schaller:

Wheeler had some plays go against him, but so does every pitcher. What's missing from this analysis is how many at 'em balls were snagged, or deep flies caught at the wall. Funky things can (and do) occur over small samples, and it's possible Wheeler has had more bad breaks than good ones. But without going too far down the rabbit hole, and examining the butterfly effect of taking a hit here or there away, it's probably safe to suggest he might have allowed a hit or two more than he ought to have.

That doesn't mean Wheeler has a BABIP problem, however. Over the past season-plus, the Mets have about the same BABIP with him on the mound as they do with anyone else on there. In fact, if there is anything to be gleaned from his career thus far, it's not that he has a BABIP problem, but that he has a BABIP advantage. That's because Wheeler's stuff is good enough to keep batters from teeing off on mistakes. Thus far, 29 percent of his hits allowed have gone for extra bases; the league average, according to Baseball-Reference, is 33 percent.

Batters might be hitting more singles against Wheeler than you'd like, but they're not doing the damage they could be doing based on the location. You've heard people talk about the margin of error when it comes to pitchers. Wheeler has it.

Special thanks to Nick Wheatley-Schaller for your favorite part.

R.J. Anderson is an author of Baseball Prospectus. Follow @r_j_anderson
Click here to see R.J.'s other articles. You can contact R.J. by clicking here