Building a Better Outfield Platoon

Thanks to Matt Foreman for passing along some Matt Diaz splits that got this post rolling...

The Braves have a luxury this year that they did not have last year: 4 starter-quality outfielders. One of the challenges that Bobby and the coaching staff will have this year is to figure out the best way to divvy up the OF playing time. There are a lot of factors to consider, obviously, but one of the most important has to be platoon splits. For instance, Matt Diaz just annihilates left-handed pitchers (.346 AVG / .383 OBP / .537 SLG career) but is below-average against right-handers (.276 / .334 / .387). In a perfect world, Diaz would get 600 PAs and never have to face a right-hander. This is far from a perfect world (if it were, we'd have a lineup of 8 Heywards), but Bobby Cox does have lot of different options.

The question I'm trying to answer here is: which configuration of outfielders gives the Braves the best chance to win against LHPs, and which group will be best against RHPs? I am going to assume that Jason Heyward will play more or less every day, so he won't be in this analysis (I don't have platoon splits for him anyway). I will examine the other main outfield options: Matt Diaz, Melky Cabrera, and Nate McLouth. I'll also look at Eric Hinske as a fill-in option. Then, I'll analyze several different platoon combinations to see which is best.

Matt Diaz

Matty was the inspiration for this post, which is kind of a backhanded compliment. He's so good against lefties, but his platoon split is ginormous. His slugging, especially, drops off the face of the Earth (his ISO drops from .190 to .111). The biggest culprit is his strikeouts, which nearly double against righties. Check out this breakdown of Diaz's plate appearances against lefties and righties:

Matt strikes out in 9% more PAs against righties, which is just a huge number. As a result, his hit and extra-base-hit rates go way, way down. Interestingly, Matty is more patient against righties, as his walk rate goes up somewhat. But that only partially mitigates the huge number of extra K's. This is one of the bigger platoon splits you'll see outside of Ryan Howard. Here's a radar graph with Matt's career platoon splits relative to the MLB averages in AVG, ISO, BB%, and K%:

Against lefties, Matt is way better than average in all categories except walk rate. Against righties, he does fine in AVG, but is not good at all in the other 3 categories. If Matt gets 300+ PAs against righties this year, and hits like these numbers would suggest, he'll be hurting the team. You could make an argument that he'll do better than this against righties, but as you'll see, we have lots of other good options.

One side note here: Matt does have a rather large amount of HBPs, which means that his OBP is not nearly as bad as his BB% would imply. The bar graph above makes this clear.

Nate McLouth

We haven't heard nearly as much about Nate's platoon splits as we have Matty's, but he does have some platoon differences. Like many left-handed batters, he struggles against LHPs (.244 / .318 / .393). He is, however, well above-average against RHPs (.265 / .350 / .474). Like Diaz, the difference in slugging is especially noteworthy (.209 ISO vs. righties, .149 vs. righties). Unlike Diaz, the problem with Nate versus LHPs is not strikeouts. Look at his breakdown:

Nate's K rate is basically the same against lefties and righties. He even hits more singles against lefties than against righties. Yet he walks less against lefties and gets far fewer extra-base hits. If he could turn a few of those singles into doubles or homers, he might be able to negate this platoon difference. Now, here's Nate's radar graph:

Interestingly, McLouth strikes out at almost exactly the league average rate against both RHPs and LHPs. Against lefties, his AVG and ISO are okay, but his BB% is well below average. Against righties, though, McLouth's ISO jumps to an excellent level and his BB% rises to a bit above average. Another way of thinking about Nate's splits is that he's an ideal cleanup hitter against RHPs, but no better than a #7 or 8 hitter against LHPs.

Melky Cabrera

The Melkman is an interesting case in this discussion, and not just because he is a switch-hitter. Throughout his career, there has been no stable relationship in his platoon splits. He'll be better against LHPs one year, then better against RHPs the next. Overall, he's been slightly better against righties (.275 / .332 / .395) than lefties (.255 / .325 / .355). Here's his breakdown:

He gets more hits but has worse plate discipline against RHPs (though his BB/K ratio is still very good). Against LHPs, he doesn't hit for a great average but walks very nearly as much as he strikes out.

One caveat to these numbers is that Melky is likely to do much better than his career numbers would suggest. To give you a better idea of what he'll do this year, here are Melky's 2009 splits: .277 / .332 / .415 vs. RHPs, .268 / .343 / .420 vs. LHPs. This is what his breakdown looked like in 2009:

In 2009, he was a bit better against lefties, with the extra walks slightly outweighing the fewer hits. Also note that he improved on his career numbers in almost every area last year. But I doubt that he's peaked. As I detailed in my Age Effects post, I'd expect Melky to be even better than this next year. Regardless, I'd expect him to have very little of a platoon difference. Here's Melky's radar graph (using his 2009 numbers):

Notice how his LH and RH splits are both similarly shaped and very nearly the same size. I'd expect just about the same from him next year, though I'd expect his AVG, ISO, and BB% to improve somewhat, too.

Eric Hinkse

Finally, let's deal with Eric Hinske. He's a great guy to have on the team because he is good in the clubhouse and can play all 4 corner spots. As a bonus, he hits well against righties (.263 / .347 / .456). He is pretty bad against lefties (.221 / .297 / .370), but for a bench guy that's just fine. Presumably Bobby will be able to get Eric most of his PAs against righties. Here's his breakdown:

Notice how his strikeout rate (already high against righties) goes through the roof against lefties. All his numbers go down against lefties, though his extra-base rate and walk rate are still just fine. The problem is that he barely hits any singles, so his OBP is only around .300. To give you an idea, Matt Diaz's singles rate against lefties is double Hinske's. Here's Eric's radar graph:

If the only things that a hitter needed to do to be successful were to hit for power and walk a lot, Eric Hinske would be a perennial all-star. Even against LHPs, his numbers are average in those categories. Of course, there's more to baseball than that (even Moneyball-era Billy Beane would say so), so Hinske is a fringe regular at best. He is, however, a great guy to have on the bench and a potentially excellent platoon partner.

Platoon Scenarios

I'm going to examine four different arrangements of our top 4 OFs, plus one "wild card" option. Here are the ground rules:

• Heyward gets 600 PAs overall. I'm just going to assume that Bobby will play him every day, more or less. For these purposes, I'm assuming that he'll hit slightly worse against lefties, but nothing dramatic. Maybe he'll turn out to have even more trouble against lefties, but there's no way to know that now.
• No player is assigned more than 600 PAs overall, more than 200 against LHPs, or more than 500 against RHPs.
• The total number of OF plate appearances is set at 2150 (the Braves had 2142 last year, and 2150 is right about average). These are further broken into 600 vs. LHPs and 1550 vs. RHPs.
• Diaz, Cabrera, and McLouth are assumed to get at least 300 PAs each. Hinske is assumed to spend most of his time backing up Glaus and Chipper, so in the first 4 scenarios, I've limited him to 100 PAs in the outfield.

For McLouth, Diaz, and Hinske, I used their career numbers for this analysis. For Cabrera, I used his 2009 numbers only. For Heyward, I used the average of his Bill James and CHONE projections.

Scenario 1: A Balanced Approach

In this approach, Bobby tries to ensure that each of his top 4 OFs gets at least 500 PAs, and then divides them up as best he can according to their platoon splits. On paper, the results would be something like this:

Pretty good, right? Let's see if any of the other options are better...

Scenario 2: Diaz & McLouth Play Every Day

Before we traded for Melky Cabrera, this was obviously the default arrangement. Many of the Melky Detractors favor this option, no doubt. And it's probably not a bad one. Of course, it's probably not the best one, either.. Melky is probably better than Diaz vs. RHPs and McLouth vs. LHPs. At any rate, here is what this scenario might produce:

Compared to Scenario 1, we have about the same number of HRs and walks, but almost 20 more strikeouts, leading to slightly worse slash lines across the board. It's not a big difference, but it is noticeable.

Scenario 3: Cabrera & McLouth Play Every Day

This is the "Bench Diaz against RHPs" option. Against LHPs, Cabrera, McLouth, and Heyward would rotate in CF and RF, with Matty in LF. Against RHPs, the OF would usually be Cabrera-McLouth-Heyward. This is probably the best defensive option, but we're not talking about a big difference. How does it rate offensively?

Compared to Scenario 1, this option ends up with a bit lower AVG & OBP, but a bit higher SLG, due to 4 more HRs. Because he isn't playing much against righties, Diaz's slash stats are much better than any of the previous scenarios. Basically, if you don't mind Diaz being a part-timer and you think this is the best defensive OF, Scenario 3 is for you.

Scenario 4: Cabrera & Diaz Play Every Day

This option is for all the people who are nervous about McLouth after his lousy spring training. In this version, Cabrera, Diaz, and Heyward play every day vs. LHPs, and most days vs RHPs. Nate become the back-up/fill-in. The obvious problem with this option is that Nate is used to being a starter and may not take kindly to a large reduction in his PAs. Also, it doesn't really make sense to ever bench McLouth versus righties unless he's really slumping (like in the preseason). Anyway, here are the numbers:

Aside from batting average, this option comes in below the other ones in every category. I see this as sort of a last resort in case McLouth keeps sucking, rather than as a real option.

Wild Card Scenario: Double Platoon!

In this scenario, Bobby plays 2 more-or-less straight platoons: Diaz and Hinske in LF and Cabrera and McLouth in CF, with Diaz and Cabrera playing versus LHPs and Hinske and McLouth playing versus RHPs. Here's what this might look like, on paper anyway:

Those numbers are clearly better overall than any other scenario, because this option makes the best use of the platoon differences. The 67 HRs are 4 to 9 more than the other scenarios, and the OBP and SLG numbers are the best of any option as well.

The flaw in this plan is that having Hinske in the lineup regularly reduces your bench flexibility. If, say, you pinch-hit for Hinske because a lefty comes in to pitch, then you've lost your main backup corner infielder. Unless Diaz or Cabrera learns to play 1B or 3B passably, I don't see this as a real option on a regular basis. But it does show that putting Hinske in the lineup against righties from time to time can be a boon to the offense.

Wrap-Up

In all of these scenarios, I'm assuming that J-Hey holds up to the MLB grind and hits at least decently, and that McLouth doesn't hit in the regular season as he did in the preseason. And, of course, I'm assuming that everyone stays reasonably healthy. A few short DL stints probably won't matter much, but any long-term injury will surely affect the allotment of plate appearances.

That being said, I'd probably go with Scenario 1 overall--the balanced approach--because I believe that Cabrera will improve (as history suggests he should) and that Diaz will play a bit better than his career numbers against RHPs. Plus, that option probably gives us the most bench flexibility on a given day and allows the top 4 guys to all get regular playing time.

Which scenario do you prefer? Vote in the poll. Or, if you have a better idea, leave it in the comments below.

Oh, one more thing. Here are the totals for Braves OFers in 2009. Compare them to any of the projected scenarios above and you'll see just how much better this outfield is. This comparison is perhaps the best reason for Braves fans to be optimistic about the upcoming season.

So, our outfielders should hit 10+ more HRs in 2010, hit around 20 points better, and slug around 40-50 points better, all while striking out less. Sounds good to me.

Thanks for reading!