Platoon woes overblown for top FAs

The fact that Shin-Soo Choo struggles against lefties doesn't take away from his overall production. Andy Lyons/Getty Images

Three of the most coveted free agents on this year's market share a common weakness. Shin-Soo Choo, Curtis Granderson and Stephen Drew simply can't hit left-handed pitching, a weakness that, it has been suggested, will limit their market value. The question is how much a team should spend on a player who can contribute only about seven days out of every 10.

After all, players with huge platoon splits aren't typically considered to be in the same class as regulars, never mind stars. How should we view players who, although they may find their way into the lineup every day, have little to offer whenever a lefty steps on the mound? As it turns out, it doesn't really matter if a player has a big platoon split, which should be great news for Choo, Granderson and Drew.

First, let's backtrack. How can we forecast how well these players will perform against left-handed pitching?

Last year, Choo hit .317/.457/.554 with 21 home runs against righties, but only .215/.347/.265 without a single home run against lefties. Stat-savvy fans and readers will quickly point out that while the ability to hit lefties is demonstrably a skill, we should never take one year's worth of platoon splits at face value.

Instead, we could look at his strikeout, walk, hit and home run rates against left- and right-handed pitching from each season, weigh more recent statistics more heavily and regress all of his platoon splits to those of a typical left-handed hitter. When we do this we end up with two projections for Choo's 2014 season:

Choo vs. LHP: .235/.333/.332
Choo vs. RHP: .288/.402/.478

A full 97 percent of Choo's value (3.4 out of a total of 3.5 WAR) is expected to come against right-handed pitching. And if we look at Drew and Granderson, the numbers tell a similar story. (Note: Each of these projections is "park neutral," meaning that to the extent these players played in ballparks that favored or disadvantaged left-handed hitting their projections have been adjusted accordingly.)

Drew vs. LHP: .206/.284/.360
Drew vs. RHP: .243/.340/.471

Granderson vs. LHP: .200/.268/.309
Granderson vs. RHP: .244/.329/.406

The question then is, how much, if at all, we should count their platoon splits against them. Perhaps value is value no matter how it is dished out. Ultimately, no player's value will be distributed evenly across games, so why does it matter that 97 percent of Choo's value can be expected to come against southpaws?

Neutralized by lefty relievers?

Even if you accept the fact that distribution of performance doesn't matter all that much, you will still hear people argue that these hitters can be neutralized in late-game situations by tough lefty relievers. As a result, their production wouldn't come when it matters most.

Fortunately, we have the perfect tool to evaluate how this plays out: leverage index (LI). This measures the magnitude of each plate appearance based on how likely it is to swing the ultimate outcome of the game. An LI of 1.0 is average and higher LIs are assigned to more crucial situations. If opponents call on lefties to face Choo in big moments, his average LI will be higher in his plate appearances against lefties and lower when he faces righties, and his overall value will be diminished.

Looking at Choo's 2013 season we might believe this to be the case. His plate appearances against lefties averaged a 1.09 LI compared to an LI of only 0.87 against righties -- implying that he did face a preponderance of lefties when the game was on the line. His career numbers tell a different story, however: an average 0.95 LI against lefties and a nearly identical 0.96 LI against righties.

To get a better sense of whether left-handed regulars with large platoon splits are neutered in the clutch I looked at all left-handed regulars (defined as having more than 1,000 PAs over the past three seasons with at least 25 percent of their plate appearances against lefties) and found the 15 with the largest platoon splits. These are the hitters who might be most susceptible to lefty relievers. I dubbed this group, which includes Granderson, Drew and, of course, Choo, the Choo family.

Choo family: 1.03 vs. LHP, 1.00 vs. RHP
All batters: 0.97 vs. LHP, 1.01 vs. RHP

The Choo family averaged nearly identical LIs against lefties and righties over the past three seasons, suggesting that its members did not face predominantly lefties in crucial situations.

What if opponents' tactics were somewhat subtler than that? Perhaps opponents take advantage of the Choo family's splits, not by bringing in lefties but rather by intentionally walking or simply pitching around these hitters with righties and then pitching aggressively when a lefty is already on the mound.

One way to see whether the Choo family provides less value than its raw stats would suggest is to use a statistic specifically designed to detect in-game value: winning percentage added (WPA). This looks at a team's chances of winning before and after a hitter stepped to the plate and credits the difference to the hitter. Perhaps, due to their platoon splits and their opponents' ability to exploit them, members of the Choo family have less positive impact on games than their counting stats would suggest.

The graph to the right compares each player's in-game value (WPA) to the run production suggested by a player's stat (wRAA or weighted runs above average) over the past three seasons. Members of the Choo family are represented by red dots and other hitters with at least 1,000 PAs over the past three seasons are shown in gray. Nine of the 15 members of the Choo family, including Choo and Drew, produced more in-game value than their statistics would predict. In short, there's no evidence that their large platoon splits made them strategic liabilities.

In other words, a team attempting to choose hitters based on the number of lefties in its division or late-game strategies risks being too cute by half.

There's something to be said for simply signing the best players.