An Observation on the Fundamental Asymmetry of Platoon Splits

For his career, David Wright, a right-handed batter, has feasted on left-handed pitching. While he's hit for a .288/.364/.483 line against righties, he owns a .341/.434/.591 mark against lefties. Unfortunately for him, though not surprisingly, he's got 3623 plate appearances against right-handers, and just 1177 against left-handers. That leads to an overall performance of .301/.381/.509. If he had faced left-handed pitching as often as he's faced right-handed pitching, and vice-versa, those numbers would jump to .328/.416/.564. He would have, in the same number of plate appearances, 83 more hits, 42 more doubles, 16 more home runs, 2 more triples, giving him 178 more extra bases, 85 more walks against 168 fewer strikeouts. 58 of those extra walks would be intentional ones. He'd ground into 7 fewer double plays, while notching 2 more sacrifice flies, reaching on six more errors, and being hit by three more pitches. Okay, most of those last ones might be random statistical anomaly, as might be his five fewer runs driven in. The basic point is, though, that if lefties were as plentiful as righties and righties as scarce as lefties, David Wright would look a lot like Albert Pujols.

That's not really surprising: a righty, who naturally does so much better against lefties, would prefer for there to be more lefties and fewer righties. But the real point is that we react to David Wright as being dramatically better against lefties, but to Darryl Strawberry, say, as having been dramatically worse against lefties, rather than saying he was better against righties. Straw's OPS was .914 against right-handers, and just .763 against lefties; Wright's are .846 and 1.025 respectively. Those are very similarly-sized gaps. But because left-handers are so much scarcer than right-handers, David's numbers look overall closer to his (worse) performance vs. righties, and Darryl's numbers look overall closer to his (better) performance against righties. The imbalance of pitching creates a natural dynamic whereby a good right-hander will be nearly as good against right-handed pitching as a good left-hander, but the good left-hander will be awful against lefties compared to that right-hander.

So in a platoon dynamic, if you have two players of equal overall ability, i.e. whom you'd expect to put up similar numbers over a full season as an everyday starter, you'll basically never want to let the lefty face left-handed pitching. But letting the righty face right-handers won't be such a big deal. You'll still expect to prefer using your left-handed hitter in those situations, but having 100% rigidity on that rule won't be such a big deal. That's useful, in a sense, because otherwise you would expect all left-handed players to dominate platoon situations, getting ~70% of the playing time as ~70% of the opposing starting pitchers will be right-handed. Or perhaps it's actually a source of some good ol' market inefficiency, with managers tending to let left-handed players dominate platoon situations more than they should. My tendency would be to assume that the overall mix should align pretty well with how much you like the two players relative to each other: if they're really just as good, then give them each 50% of the time; if not, favor the better player somewhat; and if you start wanting to give the lefty starts against left-handed starters, then it's probably not really a platoon situation.