Educate Me on Bobby Abreu
Bobby Abreu has a .339 wOBA. That figure represents the lowest of Abreu’s career in seasons that he recorded at least 500 plate appearances. The .173 ISO is right in line with his New York seasons despite the ballpark change and his walk rate is better than his final two seasons with the Bombers. He’s still hitting 45%+ ground balls; his flyball (as well as infield flyball) rates are mostly static.
Yet, his BABIP is down and not just a little. Abreu’s career BABIP is .343 and he’s never held a BABIP below .300 – heck, below .320 – during a full season. All of which is to say that seeing his BABIP hovering just above .290 is a new scene indeed. Matt Swartz of Baseball Prospectus fame asked Angels fans to jog their memory on who received the shift most often. Nobody mentioned Abreu and a handful agreed that few (if any) Angels see unkind treatment from the defense.
So, if Abreu isn’t being shifted, then what explains his .248 BABIP on balls hit to right field? Here are his BABIP to right field on an annual basis since 2002:
.257
.258
.339
.306
.372
.312
.290
.312
.248
Needless to point out, but this is his career low during that span. There is no shift in batted ball data to right field either. His groundball rates hold steadily above 60-65% throughout and even his line drive rates – say what you will about their accuracy or reliability – are mostly consistent as well. The only explanations I have are either that he’s just unlucky – which everyone hates as a reasoning but … — or that he’s hitting the ball differently, which is translating into easily fielded balls. That seems immeasurable (since what is an easily fielded ball objectively?), but that is all I can come up with.
Is there a study available on the shift itself for lefties? When did it start, who has it affected the most etc? Seems to have killed some lefties averages pretty effectively.
tdsports1,
Ted Williams’ .400 season is pretty much the reason for the shift.
Random variation?
Loss of speed?
That’s pretty much a sure thing, but I don’t think it would cost him more than 2 or 3 hits each year and it wouldn’t cost him any on balls to right field.
Yeah, that’s true. You’d be hard pressed to find someone who’s so fast they can outrun a groundball to second base.
I haven’t seen much of him this year, but he’s usually a guy that uses the whole field. Maybe the reason for the sudden drop in BABIP to RF has to do with his aging. Usually when guy age they lose bat speed, and it struck me as possible that if Abreu is cheating (clearing his hips early, etc.) and pulling balls he would have gone to LF with in the past, he’d wind up with a lot more weak ground outs and fly balls than in the past seasons. His declining LD% would be related. Just a thought.
Where can I see BABIP broken down by part of field the ball is hit to, as mentioned in the article? I never knew those numbers were available anywhere.
On the player’s Splits page, Advanced section (for BABIP), check out the rows “as … to Left/Center/Right”. Link for Abreu.
I would say a combination of [1] loss of speed, and [2] rolling over on the ball. Topspin grounders are the easiest to field, and the ball doesn’t carry.
FWIW, Joe Mauer is another guy that rolls over on balls he pulls, but stays level or slightly under balls he hits the other way. It’s actually a very interesting swing for alefty, because it’s “backwards” from the traditional lefty loop swing.
Abreau could also be pulling pitches that he used to take the other way. And if he’s beating pulled pitches into the ground, that would result in a lot of easy outs.
Guys can only get “lucky” (unlucky) to a certain degree with BABIP, etc. If we suspect that an athlete is getting tremendously lucky/unlucky, there has to be a way of seeing all of an athlete’s at bats and judging them as “should been an out” or “shoulda been a hit’. My guess is that every time trevor Cahill pitches, there are not 3 hard hit balls right at defenders or 3 diving catches, that never happen when other A’s pitchers are on the mound. I’m more comfortable with “we don’t know” than I am with “lucky/unlucky”. The latter sounds so uneducated and lazy. We could find out if we had the resources and time. As I said, if you sat down and watched all of Abreau’s at bats, my guess is that you’d be able to tell whether he was getting lucky/unlucky. We often cannot tell such things with basic offensive metrics. So, IMO, we should stop using luck/unluck as a conclusion and just say “we don;t know” and add “we don;t have the time and resources to find out”. It’s both accurate and honest.
Luck is just a shorter word for random variation. The whole point of uncertainty theory is that you have a certain set of information, and based on that information you can track other related sets of information. Of course there are always things going on that “we don’t know,” that’s why its uncertain. The point of tracking BABIP as luck is that there’s an apparently large amount of randomness in its variation that has little discernible relationship to known controllable elements of the game (at least for pitchers, the relationship is still lowish for hitters but not nearly as low). As we understand more about the controllable elements of the game (Hit F/X comes to mind, I can’t wait until something like velocity-off-the-bat bins replace or traditional LD/GB/FB batted ball types), our conclusions will of course be refined, but its unlikely that there will be some great advance made in understanding why a pitcher’s BABIP can fluctuate so wildly from year to year, yet almost always resemble the league average in extremely large sample sizes. The only reasonable conclusion is that he has little control over the number of balls in play that turn into outs. Maybe in a given year, a pitcher happens to have 20 more PAs against hitters who hit fastballs like his well. Maybe he faces 25 more who frequently swing-and-miss at like-handed sliders. These are potential bits of information that could EXPLAIN why the variation occurs, but they don’t tell us anything significant about what the pitcher is doing (or not doing, or capable of doing) to effect the results of the game.
In Abreu’s case, maybe we could understand what’s going on a bit better by sitting down and tracking every AB he’s taken and compared it to every AB he’s taken in previous years, but its not just a question of time and resource, its a question of how much information the human mind can process, store, and compare before our perception and memories become distorted. According to B-R, Bobby Abreu has “pulled” 71 balls this year. Four have left the yard, so that’s 67 balls in play to the right side. A BABIP bin with a denominator of 67 is so small that there’s virtually no mathematical conclusion that can be drawn from it without results far more extreme than these. Also, just to note, using these B-R numbers, I get a BABIP mark of “pulled” balls of .328 (26 Hits – 4 HR / 71 PAs – 4 HR). I’m assuming R.J. got his data somewhere else, but that seems to be a pretty significant discrepancy. I’ll post the question below since it’ll probably just get lost here at the bottom of this post.
“Luck is just a shorter word for random variation.”
Which is exactly why hes saying we shouldn’t use the word. A lot of the time random variation has nothing to do with it, and it is actually a change in skill/approach/etc.
“, but its unlikely that there will be some great advance made in understanding why a pitcher’s BABIP can fluctuate so wildly from year to year, yet almost always resemble the league average in extremely large sample sizes. ”
It would be interesting if that were true, but its not.
league average Babip is .302.
Tim Wakefield, for example, has a career Babip over 3000 innings of .275. In fact, the last year he had as high as the league average was 1993.
Interestingly, ‘almost always’ doesn’t mean the same thing as ‘always’. Knuckleballs are pretty different from any other pitch, though I don’t think there are really enough knuckleballers to tell if we should expect them to have lower BABIPs.
Zito has a lowish BABIP in a bunch of innings, too. Most guys, though, end up close to .300.
“:Interestingly, ‘almost always’ doesn’t mean the same thing as ‘always’. Knuckleballs are pretty different from any other pitch, though ”
Thats kind of the point. The fact that the majority of players fit a curve doesn’t mean there’s not an explanation other than luck for the guys who don’t fit the curve.
The idea that all pitcher Babip is luck is silly, and there are several extremes that prove it isn’t. The fact that Tim Wakefield, or Johan Santana, or a whole bunch of other guys have managed to keep a Babip below .280 for 2000+ innings probably means that they have some skill or method that the rest of the league doesn’t possess.
Not that they’re lucky.
Sample size, sample size, sample size!
You’re absolutely right – After 2000 IP, there may well be something going on.
One season of data? That often CAN just be explained by random variation.
I like the idea we can explain everything with what we can measure, but it’s just an idea. It’s not reality. It’s also not reality that we can explain all variation with something controllable by the player. Much of it is just “luck”, which is a word we use for variation in elements the player has little to no control over.
“. Much of it is just “luck”, which is a word we use for variation in elements the player has little to no control over.”
The problem is we assume everything we’re not sure about is luck
Simply put, its arrogant and ignorant. We assume that if we haven’t figured it out yet, there’s nothing there.
There’s definitely some aspect of pitcher Babip thats controllable, and we’d be best to stop assuming that we know better.
He’s not getting around on fastballs like he used to.
He hasn’t just lost a little speed. It may not explain all of the dip to .290, but his new pair of cement sandals have cost him more than 2 or 3 hits this season.
Hello!
I love your site, It is a pleasure to visit.
I have added your site to my site.
Please link my site to your site.
Thank you
it’s not random. it’s variation due to being on my fantasy team.
R.J., any idea why your data is so different from B-R’s? I’m looking at the “Hit Location” splits on B-R. It says he has 71 PAs where he’s “pulled” balls this year, 4 HR, 26 hits. That’s a .328 BABIP. You guys have 123 PAs where he’s hit a ball to right field, 35 hits, 6 HR, for the .248.
Actually nevermind, I see the answer, its because their definition of “pulled” is much narrower than your definition of “hit to right field”. They have him having hit 17 total HR, 13 “up the middle” and 4 “pulled” while you guys have 16 total, 10 to CF and 6 to RF. The difference appears the same last year (their defintion of RF/LF is much narrower and up-the-middle much wider than yours), when you guys have his BABIP to RF at .312, so I guess this doesn’t really help explain anything.
Though it does suggest that this is mostly a problem regarding balls hit to RCF and now towards the corner or line. His dropoff in BABIP to CF is actually pretty extreme too. In fact its an even higher dropoff from his career mark (career .362, 2010 .276 by the fangraphs definition), so maybe the problem is less about what he’s doing when he’s pulling the ball and more about what he’s doing when he’s hitting it towards the middle. Based on B-R’s splits, his BABIP on balls to center (where the bin is much wider than yours) is only .258, and oppo its .380. Looks to me like the problem is much more about how he’s using the middle of the field than the right side. Maybe its how the up-the-middle defenders are positioning themselves?
The larger strike zone affects guys like Abreu and JD the most. Abreu is forced to swing at more balls out of the strike zone, and doe not hit the ball as hard when he does, so they are fielded more easily. Until we get SOB data on each BIP, we have to live with GB and LD alone, but the harder you hit balls, the better your BIP, and vice versa.
He’s getting old. I think it’s as simple as that.
It is really a nice and useful piece of information. I am glad that you shared this useful information with us. Please keep us up to date like this. Thanks for sharing.