It's 2014 and everyone is yammering about "The Shift." Watch a game broadcast, any broadcast, and keep an ear out for how frequently it's mentioned. I even explored whether or not the shift is ruining baseball in this very space two months ago. So the shift is clearly the new thing, although it's probably not new to you given that you're at Beyond the Box Score in the first place. But before it were many other "new" things. And this, perhaps, is baseball's best attribute: it's always changing. Stagnation is a downright scary thing when you think about it, and luckily, we have baseball to continually push us forward.
That introduction was probably a little misleading, however. I don't wish to talk about the shift, but rather a notion that gained popularity before everyone was adjusting their infielders - platoons. Popularized by the usual suspects among the progressive baseball crowd, platoons were created because, as someone noticed long ago but most managers never acted upon, hitters perform differently against right and left-handed pitchers. Think about it: you're standing in your batter's box, looking at the man on the mound (shelve your egos for a second and pretend that you're not a switch-hitter). The angle at which the pitch is thrown to you would have a magnitude of difference if came out of King Felix's hand as opposed to Madison Bumgarner's. Where they stand on the rubber, their arm angle and delivery are vastly different and come from obviously different places. But guess what - you didn't get to change where you were standing to receive the pitch. You don't have to be a 10-year MLB veteran to know that which hand the ball is pitched from makes a big difference as to how it is received and acted upon. You've probably thought of this before, but it's never a bad idea to visualize our principles.
Speaking of principles, it's well accepted that hitters have better results when they face opposite-handed pitchers (right-handed batters against left-handed pitchers and left-handed hitters against right-handed pitchers). This is the basis of why platoons were created in first place - to maximize the number of left-handed hitters in the lineup when a right-hander took the mound, and vice versa. By gaining the platoon advantage, teams have tried, and largely succeeded, at scoring more runs than if they had the same nine batters in the lineup each and every game, regardless of who was toeing the rubber against them. This has other benefits, too (such as resting players, getting part-time players at-bats to keep them fresh, etc.), but these are secondary to maximizing the mix-and-match of hitters versus pitchers in the hitting team's favor.
Keep that well accepted principle from the paragraph above in mind and imagine my surprise when I was researching and writing about Diamondbacks outfielder A.J. Pollock, who, at the time, had a 68-point difference in his wRC+ between left-handed and right-handed pitching. In the opposite direction than we'd expect. Pollock bats right-handed and through May 19th, had been crushing righties to the tune of a 164 wRC+ while hitting lefties at a 96 wRC+ clip. In case you don't like watching losing baseball teams play, you may have missed A.J.'s breakout campaign. Of course, this all got erased on the last night of May in which Johnny Cueto hit him with a fastball and broke Pollock's hand. Tough luck D-backs, but hey, you're used to it by now!
None of this makes A.J. Pollock's strange, reverse platoon split any less remarkable, though, and I got to wondering, how often does this happen and is it sustainable? I had an answer in mind, but for the sake of science, thought it best to take a look at the data. Several hours of spreadsheet managing and manipulating later, I had something to work with. The results were both surprising and what I expected at the same time, but first, let's establish the parameters:
- I collected data on all hitters who had at least 100 PA's against same-handed pitchers in 2012 and 2013
- I collected data on all hitters who had at least 30 PA's against same-handed pitchers in 2014
- By default, this eliminated all switch-hitters
- I calculated offensive output through wRC+ to avoid any park factors, year-to-year manipulations and to capture true offensive output
With these things in mind, let's first look at how many hitters qualified for the above parameters each year:
I felt that 100 PA's was a reasonable cutoff I needed to make sure I had enough players in the study to try to draw some kind of conclusion. This left me with the figures above, which did surprise me a little. The bulk of MLB pitchers are right-handed, so seeing a lot of right-handed-batters get 100+ PA's against fellow righties was to be expected, but I thought there would be more than 65-ish left-handed hitters who were full-time enough to received 100+ PA's against southpaws. In hindsight, this plays on the scarcity of not only left-handed pitchers, but left-handed hitters of the full-time variety as well. Think of your favorite team and there are probably no more than three left-handed hitters who play every day. This certainly limited the sample size, but I still felt relatively good about it.
Now for identifying those who had reverse platoon splits, check out the number of hitters each year in the study who were identified as having Pollock-eque splits. Also, note the percentage these players make up of their entire cohort.
|Right-handed hitters||31 (23.6%)||44 (31.9%)||42 (30.9%)|
|Left-handed hitters||9 (13.8%)||13 (19.4%)||25 (37.3%)|
*I attribute the large group of left-handed hitters with at least 30 PA's against left-handed pitchers in 2014 to the fact that we're not even halfway through the season. This should diminish as 2014 wears on.
Okay, so this does happen relatively often, at least for right-handed hitters. Between one quarter and one-third or all righties had reverse platoon splits in each year of data collected. Left-handers were another story, however. Roughly half as many left-handed hitters had reverse platoon splits in 2012 and 2013. At first, I hypothesized that left-handed hitters faced left-handed pitchers comparatively less than right-on-right matchups due to batter scarcity, pitcher scarcity, pinch-hitting, etc. This turned out to be exactly the case.
|RHB vs. RHP||68.1%||69.4%||69.5%|
|LHB vs. LHP||31.8%||29.8%||28.5%|
The second part of my question had to do with sustainability. In other words, are these reverse platoon splits something that get repeated? Sample sizes can easily wreak havoc here and I've largely held the belief that these were flukes when they did arise. In Pollock's case, I told readers to expect his strange advantage over righties to fall, and it did slightly before he went down with the injury. Looking at the pools of players I had created through my search, there were very few batters who showed a reverse platoon split in 2012 and carried it out again in 2013. Only Mike Trout and Adam Jones did more damage against same-handed pitchers than they did their pitching opposite in 2012 and 2013. For what it's worth, neither are showing a positive reverse platoon split so far in 2014. Jones and Trout are both right-handed hitters and there were no left-handed hitters that carried a reverse platoon split from 2012 into 2013.
What does this mean for reverse platoon splits? I think it tells us what I had expected: that they do crop up from time to time, but they aren't something that should be counted on. Players showing a reverse platoon split over the course of a season aren't likely to carry it over to the next season, at least insofar as this small study suggests.
This does not mean, however, that we should discount the performance of hitters showing positive reverse platoon splits for a given season. They have a reasonable chance of keeping up the trend over the course of the year as the second table above shows. For righties, there's a seemingly stronger chance, but for what reason, I'm simply not sure. If I had to guess, I'd say that it's largely due to the extra exposure (i.e. plate appearances) to right handed pitchers. Please feel free to leave a comment on why you think this may be the case in the comments section below.
So are reverse platoon splits a fluke for hitters? Well, I suppose that depends on how you define a "fluke." It's not something that we should count on to carry over from year to year from what I've observed, but a manager shouldn't hesitate to leave a player in the lineup against same-handed pitchers if he's showing a positive reverse platoon split for the given season. There's a relatively good chance that the player will continue producing at a high level against same-handed pitching, just like Pollock was, at least until Johnny Cueto shattered his hand. Here's to a quick recovery and continued mashing against righties.
. . .