Shane Victorino and Platoon Panic

We have already run two pieces on Shane Victorino since he signed his new contract with Boston by Eno here and by Michael Barr at RotoGraphs. They are both fine pieces in their own right, but one issue that needs more discussion is Victorino’s platoon split.

As people noted elsewhere in the signing’s aftermath (and analyzed in some detail earlier this season by Jack Moore), Victorino, a switch-hitter, has a very pronounced platoon split, hitting left-handed pitching well and right-handed pitching poorly. In 2012 Victorino hit just .229/.295/.333 versus righties while facing them about in about three-fourths of his plate appearances. How much does this split really hurt his overall value?

Let me be clear up front: this post will not re-hash every aspect of Victorino’s game, such as his good base running or whether he will be even better in right field than he has been in center. This is not a general “how does this contract look in context” post, as Eno already did a good job in that respect. This post is a response to the notion that Victorino’s big platoon split in itself is something that makes him less valuable than he would be otherwise.

I am not always sure what people are implying when they mention Victorino’s (or any other player’s) big platoon split. If they mean that a big split means that his overall offense projects to be worse than any other player’s, then that can be pretty readily dispatched. As mentioned in The Book’s chapter on platoon splits, one should not project a player’s offense as if his platoon splits represented two separate players who are combined into one. A hitter’s plate-appearances against left-handers tell us something about his true talent against right-handers, and vice-versa. The way to project a hitter’s platoon split is to first project his overall true talent, then estimate his platoon split and apply that projected split to the overall projected hitting line. It is not as if Victorino has faced a different proportion of left- and right-handed hitters than the average major-league hitter. A bit more than a quarter of his career plate appearances have been versus left-handed pitching.

One might expect me to point out how much Victorino’s split can be expected to regress to the mean. However, Victorino has more than 1200 plate appearances versus southpaws during his career, and switch-hitters’ splits stabilize much more quickly than those of either right- or left-handed hitters. So while we cannot simply assume that his split from 2012 is what it will be in 2013, it is pretty big.

For an overall hitting projection, we are still waiting on systems like ZiPS and Steamer. My own crude weighted average with regression and a slight age adjustment projects Victorino for a wOBA of about .325 next season. It is nothing sophisticated, but is good enough to work with for the purposes of this post. Using that overall projection and applying projected platoon skill, I have Victorino’s true talent as .366 wOBA versus left-handed pitching and .309 versus right-handed pitching — a big split.

Does the big split make Victorino a mere “platoon player?” Some think so, but that misunderstands what a platoon player really is. Pretty much every player has a split, and would be more valuable if roster space allowed all players to be platooned. Of course, as noted, Victorino has a particularly big split. However, if a .325 wOBA is good enough for him to play full-time, it does not really matter how it is distributed. A true platoon player is generally one whose overall wOBA is not good enough to start full-time, and only projects to be an average or better overall player when having the platoon advantage. For example, if Victorino was a .305 true talent overall wOBA player with a projected .325 wOBA versus lefties, then he probably be a platoon hitter (leaving aside defense and baesrunning). However, we have Victorino at .325 overall.

Leaving aside the debate of whether or not a .325 overall wOBA is good enough given Victorino’s position and other skills, does a particular projected distribution versus left- and right-handed pitchers in itself make his offense less valuable? After all, about three-fourths of the average player’s plate appearances come against right-handed pitching. Still, in terms of traditional linear weights, this does not make Victorino less valuable. If a player has a .325 wOBA versus both left- and right-handed pitchers rather than Victorino’s .366 and .309, in terms of offensive linear weights it would be worth… exactly the same. By traditional linear weights, this is not different than a player who has a .309 wOBA for the first three-fourths of the season and a .366 for the last one-fourth. It still comes out to .325 overall. What is the problem?

This is not to deny that there are disadvantages to having a big platoon split. Late in close games, a player with a big split can be exploited by opposing managers who bring in specialist relievers. In this connection, it is interesting to note that Victorino has a noticeably lower career wRC+ (81) in high-leverage situations than in low- (108) and medium-leverage (106) situations. Without getting into the play-by-play data, this is not necessarily primarily due to facing a high proportion right-handed relief specialists in high-leverage spots, but it is a reasonable inference. Opposing managers certainly should be looking to match Victorino up versus righties in big spots.

Even granting the threat of specialist relievers in high-leverage situations is a problem for a player with big splits, how much does this really matter? Victorino has almost 4300 career plate appearances, and only 377 (under 10 percent) of them have come in high leverage situations, a seemingly fairly typical proportion. Without getting into the murky waters of how to weight things by leverage, one can grant that it does hurt his value a bit.

However, a big split also can have advantages even with a full-time player. A manager willing to adjust his batting order so that Victorino (just pulling spots out of my hat without looking closely at Boston’s projected order) hits second versus lefties and seventh versus righties could get a bit of an advantage. Over a full season, the seventh spot in the order probably gets about 100 fewer plate appearances than the second, and assuming a basic three-fourths distribution of pitchers, that would mean 75 more plate appearances on his “good side” for Victorino. Although 75 plate appearances is not very many, keep in mind that Victorino had just 78 plate appearances in high-leverage situations that year. If one is important enough to worry about, so is the other. Moreover, if one wants to quibble over a small number of plate appearances, one could also make the case that if Boston has even a mediocre left-handed bench bat, the manager could leverage Victorino’s off-days in that way, too.

Like just about every player in his thirties, Shane Victorino is not what he used to be. Still, 2012 is not the only season on his resume, so simply projecting him to repeat 2012 is bad analysis. With respect to the issue at hand, although Victorino’s big platoon split is eye-catching, when thought about carefully, it does not significantly hurt his projected value. Whether or not Victorino’s new contract is a smart one for Boston, his platoon split is not an very strong argument against it.