Does Carl Crawford’s Platoon Split Matter?

Carl Crawford was widely considered to be the biggest prize among position players this offseason, and it was no surprise that he got the big money from the Boston Red Sox. However, historically he has had a lot of trouble with left-handed pitching. It’s one thing to point out that platoon splits can be expected to regress pretty heavily to league average. But beyond that issue, how much does his platoon split really matter, anyway?

Crawford has more than 1500 plate appearances in his career against left-handed pitching, so, in his case, Crawford’s true-talent platoon skill is probably closer to his observed split than to league average. Using his ZiPS‘ .351 projected wOBA for 2011 (with steals and caught-stealing removed because we’re concerned with the batter/pitcher matchup), I get a projected wOBA of .362 versus right-handed pitching, and .322 versus left-handed pitching. That .322 is about league-average in the 2010 run environment. If that seems high for a player with a .307 career wOBA versus left-handed pitching (and a .337 wOBA career with steals removed), keep in mind that we regress so much partly because of the year-to-year volatility (partly due to a low annual sample of plate appearances against sought paws) of splits. For example, as recently as 2007, Crawford had a .360 wOBA against left-haded pitching. He followed that in 2008 with a .289. But while regression to the mean is (as usual) important, it is really a side issue for this post.

In 2010, Carl Crawford had a .306 wOBA versus left-handed pitching and still managed to accumulate 6.9 Wins Above Replacement. In 2009, Crawford had a .313 wOBA versus southpaws and put up 5.7 WAR. Crawford’s value isn’t purely tied up in his bat: he adds value on the bases and in the field even when he’s facing left-handed pitchers. Beyond that, as bad as he can be against lefties, he hits righties very well, and (like all full-time players) sees them most of the time (more than 70 percent of his career plate appearances are against right-handed pitching). One might point out that he only had a .289 wOBA versus lefties in his disappointing 2.5-WAR 2008 season, except he didn’t exactly smash right-handed pitching (.327 wOBA) either — the problem wasn’t his platoon split, but his hitting overall. Even if Crawford’s true-talent wOBA is only about .310 versus left-handed pitching and he had to face them every plate appearance, his defense and baserunning would still make him about a league-average player. He isn’t going to hurt the team when he starts against lefties (although it would be a good idea to move him down in the order, but that’s another can of worms.). Crawford’s broad base of baseball skills and hitting ability against right-handers means that, despite his platoon issues, his value is what it is: outstanding.

That is enough to briefly answer the scattered concerns about Crawford’s platoon issues hurting his value. But might not the split still matter in another sense? Teams surely know that Crawford is a much less effective hitter against southpaws, so in crucial, high-leverage situations, they can exploit this by putting in a lefty-specialist in against him, right? Since Crawford changed teams within the division, he will be facing almost the same team opponents, so let’s see if he’s had trouble in high-leverage situations in the past. This is not to assume that there is some “clutch skill” on Crawford’s part, but rather to look at Crawford’s likely opponents’ treatment of him in the past: have they been able to exploit him in high leverage situations in the past?

During his career as a whole, Crawford has a .341 wOBA in low-leverage situations, .352 in medium-leverage situations, and a a.357 in high-leverage situations. Nothing there. Let’s looks at Crawford’s Clutch score, which quantifies player performance in terms of wins in crucial situations. Again, this is not to impute this skill (or lack thereof) to Crawford, but to see if teams have used it again him in the past. For his career, Crawford is +1.5 wins in “Clutch.” If we look more closely, we do see that his highest seasonal Clutch value came in 2007 (+1.79) when he also had his best season against southpaws (.360 wOBA). On the other hand, in his second best season against lefties, (.338 wOBA) 2006, he was -0.93 Clutch wins, and during 2008 when he wOBAed .289 against lefties, he was +1.08 wins. Without going listing every season and associated performances, it is simply worth noting that there is too little correlation with his platoon performance to conclude that the teams Crawford has faced have been trying (or at least not successfully trying) to exploit his problems against southpaws in high leverage situations.

Carl Crawford is an excellent player with a larger-than-average platoon split for a left-handed hitter. Despite that split, his abilities in the field, on the bases, and against right-handed pitching more than make up for it. Moreover, there isn’t obvious evidence most of the teams that Crawford will continue to face in the near future have been able to consistently exploit his platoon problem in high leverage situations. One ambiguous note in conclusion: there is one opponent that knows Crawford every well, although he hasn’t played against them in real competition before. He will now face them quite frequently. It’s a team known for using every bit of information they have to get that extra two percent. Carl Crawford versus the Tampa Bay Rays is a late-inning match-up I can’t wait to see.





Matt Klaassen reads and writes obituaries in the Greater Toronto Area. If you can't get enough of him, follow him on Twitter.

37 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
lester bangs
13 years ago

Mediocre Fenway stats too, which no one talks about.

lester bangs
13 years ago
Reply to  Matt Klaassen

In that case, no visiting player’s park stats should ever be mentioned on the road. The sample will never be big enough to matter. Crawford has be an in-division rival his entire career.

Then again, I expect Devil Rays (and former star Devil Rays) to constantly get the benefit of the doubt here. If Andrew Friedman burped into a tin can, there would be a hasty line to proclaim it “brilliant!”

lester bangs
13 years ago
Reply to  Matt Klaassen

Okay, Rays. I’m fining myself a dollar.

Telo
13 years ago
Reply to  Matt Klaassen

Well, sadly, most of what is mentioned or shown on screen during baseball broadcasts or in mainstream media isn’t as significant as people think it is. Some of the dumbest thing managers do is bench good hitters who are say 0 for 10 against in favor of a poorer hitter who is like 6 for 10 against a certain pitcher.

It’s like they don’t know the math!!!!!!!!!!!!!!!!!!!!!

(spoiler: they don’t)

CircleChange11
13 years ago
Reply to  Matt Klaassen

who are say 0 for 10 against in favor of a poorer hitter who is like 6 for 10 against a certain pitcher.

It’s like they don’t know the math!!!!!!!!!!!!!!!!!!!!!

That seems like a funny thng to say, given the stats presented.

Isn;t 6-for-10 better than 0-for-10? *grin*

What they don;t understand is the issue of sample size.

However, in some cases there may be more to it than just noise or sample size. A certain player may hit curveball pitchers well, and the other not or something to that effect.

This is where watching a player would have some advantage. If a guy just looks absolutely horrible against a certain pitcher, then maybe a change is due.

I always think back to the article where Mike Redmond owned Tom Glavine over 50+ At Bats. Obviously we’re talking small sample size here, and the conclusion is basically correct that it’s luck or randomness.

However, I wonder what scouting would say. Perhaps Redmond just has “changeup bat speed” or does well with lefties that live on the outside corner.

I’m just discussing at this point … but I wonder if in the near future we’ll have access to hitter’s stats on certain pitches and in certain counts? Certainly we’ll never likely accumulate enough sample size to have highly reliable data, but we’ll likely see guys that tee off on fastballs in the 92-94 range, and others that are horrific against curveballs when behind in the count and things of that nature.

We already track batter’s averages by count … the next step would be to track it by sequence and such.

Barkey Walker
13 years ago
Reply to  Matt Klaassen

OK CircleChange11, now I want to see that article!

Telo
13 years ago
Reply to  lester bangs

Probably because it’s too small of a sample to be significant enough to talk about.

Telo
13 years ago
Reply to  Telo

Sorry, Matt beat me.

Ben
13 years ago
Reply to  lester bangs

Sample size.

williams .482member
13 years ago
Reply to  lester bangs

also, he was usually facing some very good pitchers when in Fenway.

Brandon T
13 years ago
Reply to  williams .482

This is just what I was thinking…