The following post appeared in 2012 as part of WGN’s Stats Sunday series. Thanks to Joe Sheehan for this article.
Earlier this season, not long after Mariano Rivera’s season-ending knee injury, I was watching Rafael Soriano struggle to close out a Yankees win. Soriano wasn’t scuffling because he doesn’t have “closer mentality,” but rather, because he was facing a run of left-handed batters, and he isn’t that effective against them. I sent out a Tweet to that effect and was inundated with responses by fans with access to Baseball-Reference.com. “NO!” they cried, “Soriano is better against lefties this year.”
Well, they were technically correct. At the time, Soriano had allowed a very high OPS (OBP + SLG) to right-handed batters, and even today, he has a “backwards” platoon split: 642 OPS allowed to righties, 595 to lefties. However, there’s more to a pitcher’s abilities than 200 plate appearances in the current season — and more to his skills than OPS splits. Because of the small samples in single-season splits for relief pitchers, you need to look at multiple years in determining what any player’s true platoon split is. Soriano, in his career, has allowed a .171/.232/.284 line to righties — he owns them — and a .230/.305/.384 line to lefties. Those career stats are much more in line with what you’d expect from a hard-throwing right-handed reliever with a nasty slider.
Look deeper into Soriano’s 2012 line, and you’ll see that he’s actually wiping out righties this year, too. He’s struck out 24% of the right-handed batters he’s faced, with a 19/4 strikeout-to-walk ratio. Against lefties? a 26% strikeout rate, but a less impressive 29/10 K/BB (unintentional walks, only). Those numbers, strikeout rate and K/BB, are the clearest indication of a pitcher’s platoon skills. The OBP and SLG numbers can be influenced by a pitcher’s batting average against on balls in play, which isn’t a skill that splits along platoon lines. Soriano has allowed a .339 BABIP to righties this year, and just .260 to lefties, for no reason other than that can happen in a handful of at-bats. The fans who were yelling at me about Soriano being “better” against lefties weren’t looking at strikeouts and walks, and they weren’t looking at the pitcher’s career — and they were reaching the wrong conclusion.
This is a small point, but it’s important if you’re going to evaluate tactical decisions by managers or pickups by GMs. One of my favorite examples is from a few years back; the Red Sox had a left-handed reliever who dealt from the first-base side of the mound and threw a lot of sweeping sliders out of the reach of lefty batters. In 2007, though, he allowed a .359 BABIP to lefties, which led to an 805 OPS allowed to them…despite a strong 22% strikeout rate and an 18/7 K/BB. When the same thing happened in 2009 — a .500 BABIP allowed to lefties!– the team gave up on the southpaw. A year later, Javier Lopez landed in the Giants’ bullpen and was part of the group that allowed just three runs in September and shut down the favored Phillies and Rangers in October on their way to the World Series title. In three years since the Red Sox let Lopez go, he’s struck out 28% of the lefties he’s faced and posted a 71/25 K/BB, while holding lefties to a .173 batting average.
The Lopez story is an example of why you can’t just look at stats to make sense of a pitcher’s platoon splits. Pitchers can influence their effectiveness on hitters from one side of the plate or another with arm angle and repertoire. Soriano comes from three-quarters, throws a hard fastball and sharp slider; watch him for three innings and you can see how lefties would be much more comfortable against him. Lopez slings the ball from the side, so left-handed batters can’t find the ball but righties get a great look. There is no reason why lefties should hit .293 against him, as they did in 2007, other than, “sometimes the balls fall in.”
Loosely speaking, the closer a pitcher is to pitching over the top — from 12 o’clock — the smaller his platoon split should be. The more a pitcher works up and down in the strike zone, with 12-to-6 curves, change-ups and splitters, as opposed to with pitches that work across the plate like down-and-away curveballs and hard sliders, the smaller his platoon split should be. Pitchers who rely heavily on cutters, like Rivera or the White Sox’ John Danks, are prone to reverse platoon splits, because their pitches work away from opposite-side hitters and into same-side ones.
In determining what a pitcher’s true platoon abilities are, you want to start with what he looks like, then what his weapons are. From there, go to his strikeout and walk rates, rather than his AVG/OBP/SLG, because those numbers are more reliable indicators of skill and less influenced, even in small samples, by luck. Look beyond the current season, especially for relief pitchers, who face just a few hundred batters in a given year. With this mix of information, you’ll reach the right conclusions — sometimes more often than the managers do!
Joe Sheehan is a contributor to Sports Illustrated, SI.com and the NBC Sports Network. Follow him on Twitter at @joe_sheehan and get information about his popular baseball newsletter at http://www.joesheehan.com.