Xander Bogaerts Changes, Really Remains the Same
Everything Xander Bogaerts did well in his breakout season last year, he seems to be doing better this year. More power, more patience, more contact, and better defense — he’s basically equaled last year’s full-season WAR figure already, and there’s three-fifths of a season left to go. He’s leading the league right now!
Of course, WAR isn’t your traditional counting stat: Bogaerts could hypothetically put up negative wins going forward, were he to regress in one way or all of them. But since he gave us such a great preview last year, it’s tempting to believe in all of the improvements he’s made. He’s really the same guy, just a little better.
At the center of his improvement has been how hard he hits the ball, the angles (both vertical and horizontal) of those batted balls, and his defensive range. He didn’t think much had changed about those particilar variables when I asked him, though. Just a few minor tweaks.
The people around Bogaerts know that he hits the ball hard, even as they suggest that it’s not a priority. “We see power in there,” Red Sox manager John Farrell told Nick Cafardo recently. “He’s got good plate coverage and that’s what he’s striving for. That [power] will come in time.”
Only nine players added as much batted-ball velocity in the second half last year as Bogaerts, who was squarely above average and even outstanding at moments late in the season when judged by exit velocity alone. But he hits the ball into the ground a lot, too. And it’s his choice.
“This is the player I want to be,” Bogaerts told me. “Sometimes people want you to be someone else, but it’s not up to them.” So, instead of the full-scale revamp of his launch angle that has lead to power breakouts for people like J.D. Martinez, Bogaerts has chosen to stay the same. “It’s what I like, I’ve had success doing it, so why change it,” he continued. “If it comes, when it comes, that’s okay, but it’s not something I want to go start changing.”
So Bogaerts is hitting for more power this year, but it’s by hitting the ball harder, and slightly changing things, instead of big changes. He’s only subtracted 1.7 percentage points of ground balls — placing him nowhere near the top of the leaderboard in that department– and even new-fangled angle analysis says he hasn’t changed much (26.6% in ideal angles this year, 24.9% last year). The key is that he’s added 5 mph in ideal angles this year, good for the ninth-best improvement in baseball since last year.
Now Bogaerts has the 94th-best exit velocity in the ideal launch angles, better than the velocity produced by Marcell Ozuna and Matt Adams. He’s the same guy, but better.
If you wander over to his horizontal angle — his pull versus opposite field percentages — you might be tempted to say that you’ve found real change. Or, at least, he’s pulling the ball more than last year. Ends up, that might be a function of where he’s getting pitched. Take a look at his fastball (top) and changeup (bottom) heat maps, with last year on the left and this year on the right.
“Yeah a lot more curveballs, offspeed, fastballs inside,” Bogaerts confirmed. And that lines up with the map.
But it’s worth asking how much pitch location determines the outcome of the pitch. Andrew Cashner once told me it was foolish to pitch differently with the shift on behind you because “pull hitters are going to pull it,” no matter where you throw it.
Thanks to Jonah Pemstein, we can see what the league does on average. Here’s horizontal spray angle by pitch location for right-handers, with red representing more pull. Cashner is right, to some extent: there are pull hot spots all over the place. But there’s also an extreme pull tendency inside off the plate, and if Bogaerts is getting pitched more inside, he’s seeing some of those, too.
And, as it turns out, Bogaerts is actually also very skilled at… going the other way with pitches on the inside. When he says “sometimes you get pitched in, and you get jammed,” he’s intuitively referencing the fact that he’s among the five best players at going to the opposite field on inside pitches.
Pemstein will have more on that going forward, but Farrell’s comments can once again give us a guide. His player’s priority is plate coverage, and he’ll pull it or push it, but he wants to make hard contact on pitches both inside and out. This year is just more of the same — more reacting to where the pitcher is throwing it. “You prepare in the cage,” Bogaerts told me. “I can’t control if they are going to throw me in or outside, all I can do is try to be on time and ready. It depends on where he throws it.”
As for the defense, the player wasn’t going to help us find the tweak that has helped him improve. “Range is the same since I came up, but I might be a little quicker now that I know what to do, where to go,” he said, agreeing that anticipation of the ball in play is now better.
There’s one thing he has changed in the field. “I always had a problem throwing. I just had to figure out what it is I had to do to get it over there well instead of all over the place,” he said. But what that was, he wasn’t going to say. “It’s something different. I know what I do, but I don’t really like to talk about that piece. I just know what I have to do to get it over there in a good area now.”
Apologies to Xander, but I had to look. My guess is that he’s changed his arm slot. If you look at this error against the White Sox that lost the game, you can see he sidearms it into the ground. And then looks sad.
That was 2014, and even the “great play” highlights depict him throwing sidearm. Now, in 2016, his throws generally look different. On a very similar throw to the one he blew two years ago, his throws specifically look like this:
He’s over the top there. And looks like he’s more over the top in general. Who knows why he wouldn’t want to tell us this. Maybe those sliding into second would be less intimidated by a possible low sidearm throw if they thought he always goes over the top. Who knows. But it looks like a real change, even if it’s a real change that has allowed him to be the great defensive shortstop that he was supposed to be.
And that might be one of the few substantial changes we can detect. Most of what’s happening this year for Xander Bogaerts is just the result of “Working, trust and confidence,” as he put it. “It’s just me,” he concluded. In this case, “just” him is pretty glorious.
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.


In regards to his defense, I think UZR is overrating him a lot. He’s got a 0 DRS on the season, yet his UZR/150 is 12.9. His error rating is 2.9, and yet in roughly the same amount of plays and innings he’s made one error more than Lindor and Crawford who have 1.9 and 1.5 error ratings.
He already has 3 throwing errors listed which is half of his total last year. Although he is much better than 2014.
It depends on the weights. His range has been very good this year, which is probably what UZR is so excited about. His arm has alternated between spectacular and erroneous, which is likely why DRS is less excited.
Also remember that error scoring is sometimes a very deceptive thing. At least one of the throwing errors was on a play that a more experienced first baseman would have picked (and I vaguely remember one of the others being the same, but I’m not sure). Bogaerts is *trying* more edgy throws than he tried last year, and so some of them aren’t going to work out as well. I think that’s probably an overall net positive.
Am I crazy, or does your 2016 example of Bogaerts involve Jon Lester pitching to Robbie Cano?
Yah that’s weird, huh? Came up in my 2016 search, shoulda looked at the pitcher. Fixed now.
I’m a little puzzled as to why you’re using a clip from late 2013 to illustrate a point about how Xander’s throws have changed since 2014.
Weird, I looked at a bunch of 2016 video before I decided that one looked the most similar. Came up in a search that included 2016 in the term. I’ll fix the video.
After the last fangraphs piece about Xander was posted a lot was made in the comments section of his insanely high BABIP. Some commenters even went so far as to suggest that if you strip out BABIP luck from his numbers, Xander is not anything more than an average SS in MLB. It seems preposterous to suggest that bogaerts has been nothing but lucky for a year and a half and counting, but that was the pov of at least one commenter. Does that seem plausible? Is Xander just getting lucky or is there something more sustainable at play here?
The entire Boston squad has a high babip. I think at least part of this is that babip is misleading for Fenway because a large number of “balls in play” are not actually playable unless the left fielder is 30 feet tall.
That definitely part of it. But this hasn’t happened in past years and I think that it is now because of some combination of luck, skill, speed and young players breaking out. There are very few Red Sox players that aren’t very athletic and the slow guys like Papi hit the ball hard. I think they’ll continue to run a high babip but they are definitely in for some regression
I don’t have the time now to verify, but I seem to recall looking at this recently and finding that Boston has led the league in babip for several years running and that they are almost always very high.
Since Ortiz joined the Sox in 2003, they are second in BABIP, unsurprisingly to the Rockies, at .309. They are also the best offensive team by wOBA and second by wRC+ (to the Yankees) over that time period.
I know they typically run high. But they are currently stupidly high. Regression is coming but for the reasons I talked about it might not be as bad as many people think
(No Pitchers)
Team Total
2016: .335 (#1)
2015: .305 (#10)
2014: .297 (#18)
2013: .330 (#1)
2012: .301 (#11)
Current Active Roster Only
2016: .343 (#1)
2015: .301 (#15)
2014: .296 (#21)
2013: .308 (#15)
2012: .298 (#19)
are they repeating their golden 2013 run? anything’s possible, but given that that was the highest team babip since 1930 it’s unlikely.
their babip has already startee to come down in june leading to still good but not crazy good offense, and i’d guess that’s likely to be the case going forward.
The team history is irrelevant in determining how high the babip can run. That’s like saying the Cubs are unlikely to win a World Series because of how long it’s been. It may speak a little to park effects but odds are Fenway would only help babip because of the monster not hurt it. The team has a lot of players capable of running high in the category and I’d bet they stay relatively high compared to the norm with some regression mixed in
The Cubs are unlikely to win a WS in any given year, and you can derive an estimate of those odds from their success in winning WS in past years.
With no prior information and a binary outcome you would estimate the Cubs’ chances of a WS at 50%. The past seasons provide prior information about whether there is a bias in the likelihood of the outcomes. There is a bias (since there are 30 other teams currently), and the past seasons outcomes give an estimate of this bias.
Using only past seasons data gets you much closer to the truth. Of course you can improve the estimate by including information about the current team relative to the other current teams, as you point out.
“The Cubs are unlikely to win a WS in any given year, and you can derive an estimate of those odds from their success in winning WS in past years.”
Wait, what? The fact that the Cubs didn’t win the World Series in, say, 1958 (or 1964! Or 1936!) somehow informs their 2016 WS odds? That’s…that’s not at all correct.
Actually, it is correct for the reasons I stated. I’ll try rephrasing.
The distribution of past success winning the WS gives an estimate of the distribution of the likelihood of winning the WS in any one season. This is no different than using the distribution of past success getting base hits (described by batting average) as an estimate for the likelihood of getting a base hit in any one plate appearance.
I think the trouble people have with this idea is that things like batting average are thought to have some dependency between at bats that is the individual skill of the batter. And dependencies between the 1964 Cubs and the 2016 Cubs are much more difficult to see (though they may still be there, e.g. Wrigley Field). However, it is actually the case that there need not be any dependencies between past events for them to be informative for future events. The first 100 coin flips informs the likelihood of getting heads on the 101st flip, even though the events are completely independent.
You can partition the variance of the likelihood of the Cubs winning the WS into random variance and team skill level variance. The Cubs success from past years gives an estimate of the random variance component and mostly ignores the current team skill level variance. However, because the random variance is actually the biggest component of the variance, the past events will not give a terrible estimate. It just won’t be the best estimate possible.
That’s just wrong in every way
2016 Red Sox BABIP = .334
2016 Red Sox xBABIP = .330
that was one reader who shall remain nameless… what I’ve picked up on is that readers see him a tier below the top guys because he’s not viewed as the best hitter or the best fielder even though he’s good at both… but with the ways he’s playing you can argue he’s top tier as well
there is nothing sustainble about his babip.
Most of the Red Sox BABIPs will go down. JBJ, Shaw, Chris Young, Pedroia for sure. But X is definitely the type of hitter who will run high BABIPs due to a combination of high GB rates, high contact rates, hitting the ball hard, and going the other way. He’s not going to remain at .400, but he’s at .351 for his career, which is now over 1500 PA. I’ve said it many times before, but he reminds me of Jeter, and Jeter had a career .350 BABIP himself for similar reasons.
dropping to .350 babip would knock 120pts off his ops.
Leaving him with an absolutely atrocious .807 OPS, which would be 19th worst among qualified shortstops!
correct, if we are granting a 23yr old the elite babip skills of a hall of a famer, then his numbers should still look quite good.
Or if we grant him the elite skills he has displayed over the sample size of 1500 PA and the justifications that have been stated by both commenters and writers several times. The kid is phenomenal and very little recent history says otherwise
yes, there is a chance that this 23yr old is a hall of famer.
and there’s a chance he isn’t.
but i’m not sure those chances have even odds.
is this snide comment referencing my comment about his display of skills? Because non hall of fame players are perfectly capable of running a high babip. Do you remember Chris johnson? He had a career .356 babip and he has 3.3 career WAR. Just because a hall of famer has a skill doesn’t mean you have to be a hall of famer to have that skill
Where do you get 120? 60 from OBP and 60 from slugging? That’s not really how babip regression works.
how do you think it works?
It isn’t a trivial thing to calculate how a change in babip would effect OPS for a player. The extent to which babip matters for a player is a function of the proportion of balls they put in play. Babip has a negligible effect on OPS for a player that walks, strikes out or homers in 90% of their at bats. On the other hand babip will have a large effect on OPS for a player that puts 90% of balls into play.
Also, you can directly calculate how a change in babip would effect OBP, but you can’t know how a change in babip would effect SLG, because you don’t know how the lost hits would be distributed among singles, doubles, and triples.
i don’t see why you wouldn’t treat his ISO seperately. I’ve never seen anything which suggests babip impacts ISO.
It’d be a pain to explain how babip regression works instead of showing it but it’s not straight subtraction. The gap between babip and average will shrink as the babip goes down though which is why you can’t just subtract
One of the most remarkable examples of statistical ignorance consistently shown on fangraphs is that it is taken as received wisdom that for batters a high gb% contributes to a high babip but for pitchers it contributes to a low babip. In reality gb% itself has essentially no effect.
After the last fangraphs piece about Xander was posted a lot was made in the comments section of his insanely high BABIP. Some commenters even went so far as to suggest that if you strip out BABIP luck from his numbers, Xander is not anything more than an average SS in MLB.
http://www.fangraphs.com/fantasy/mike-podhorzers-2016-bold-predictions/
3. Xander Bogaerts ranks outside the top 7 shortstops
SURPRISE! Just kidding, you’re not surprised at all. You know exactly what I’m projecting for Bogaerts this year. In two full seasons, his batted ball distance has not exceeded 270 feet. Those expecting a power spike are just wishcasting. So what if that power doesn’t improve like we all figure it will (some more so than others,
obviously)? What if his BABIP regresses even more drastically than forecasted? What if the steals completely disappear and revert to his 2014 level? What if he quickly shows himself as an empty .270 batting average and gets dropped in the order, out of the favorable three hole? That’s the downside and it is clearly being ignored given his hefty draft day price.
He’s already having a power spike.
I recall the game from the first clip. It was Jackie Robinson Day, so everyone was wearing #42 and it was extremely cold, so a bunch of the players on both sides were wearing those face masks (as Xander is demonstrating). You couldn’t tell who was who.
I remember thinking “This must be what a League of Shadows intramural softball game looks like.”
Bogaerts is leading the AL in WAR but not the majors.
He’s leading mortals in WAR.
He’s leading position players.
I’m interested to know where the base stealing came from. He was never considered a threat or had much efficiency in the minors. Suddenly he became very efficient last year and now is on pace for 24/5 this year. I know the Red Sox have simply been more aggressive on the base paths as a whole. Anyone have any insight into this?
My understanding is that the Red Sox have the 1B coach (Ruben Amaro) watch a ton of film on opposing pitchers and the time it takes them to get to home plate. From this information they’re able to relay the best times to steal.
http://m.mlb.com/news/article/179574420/red-sox-lead-majors-in-stolen-base-percentage
What? Amaro is actually good at something?
I also kick ass at Parcheesi.
Mike Podhorzer’s recent article on xBABIP gives an estimate of Bogaerts of .325, as opposed to .409 actual BABIP at the time (it’s now .408). That’s the second most inflated BABIP in baseball and would knock about 70 off his batting average. He’d still be hitting .290 and has the 5th highest Def rating in the majors.
I don’t believe that takes some things like speed into account that would likely boost his. Beyond that it’s still not going to be a perfect correlation. Based on a large sample I’d be inclined to say that bogaerts is perfectly capable of running a babip of .340-.360 any given season. Obviously still plenty of regression but still a phenomenal hitter especially for shortstop. However you do seem to have had a change of heart. Last I saw you comment on bogaerts you called him average and you clearly believe differently now
He gives the formula, it does take speed into account although i don’t know if i agree with how it’s done. There is absolutely no basis for saying he may be able to have a babip in the 350 range, almost no one can do that, most obviously he hits way too many infield flies (for example about 4 times as many as derek jeter). You should learn more about xbabip. I didn’t call him average, i was just less adulatory than a lot of other readers.
Ah, there it is. My new fangraphs hobby is to search the comment sections for times that you suggest another commenter learns about babip. You’re on a roll! I’ve counted 4 so far today
And yet people do do it. Otherwise nobody would ever hit over .333.
Yet they do. It happens relatively often.
We all understand BABIP. You’re the one who doesn’t. If you’re so eager to talk down to people without getting called out for your ignorance, go to Bleacher Report. This is Fangraphs.
It is certainly possible to bat over .333 with a low babip. You just need to hit lots of home runs.
Of course people hit 333 for individual seasons. Because they get lucky. And regarding whether fangraphs readers are more knowledgeable than bleacher reports readers, check out the comment above that there is nothing about bogaerts babip which is sustainable. It got 4 likes and 13 dislikes. This suggests thst more than 3/4 of fangraphs readers think that a 409 babip is sustainable. Lol.
Yeah I guess there’s no basis for it if you opt to ignore the many reasons for it that have been explained several times in well though out articles and comments. You offered one thing and just ignore all the other evidence presented to you because it doesn’t fit your narrative. I understand that he has areas in his game to improve but it’s outweighed by strengths and he has a history of fixing flaws. And by the way I should learn more about xbabip? You should learn more about advance metrics like this that are predictive. They don’t correlate 100% because they don’t take everything into account that wouldn’t be possible. Based on the last few years it’s safe to say bogaerts is one of those points that won’t fit the line. I saw something about an xbabip equation that didn’t include speed and I thought it may have been that. It was probably another site or authors metric so I actually fully understand the metric and its uses.
please stop posting this nonsense. we get it. you dont think xbox has the skills to maintain a high babip. fine. the kid is running at .380ish over his last 1000 or so PAs and clearly to the eye test looks to have drastically improved at the plate since the start of 2015. also plenty of other readers have posted reasons why a higher babip may be sustainable for him. .350ish seems like the absolute minimum you’d project him for going forward and he did just post a .372 over a full season. i think we all agree his current babip is unsustainable but to drastically cut him down to .325 based on his contact stats not aligning with some babip-predicting formula seems stupid.
“There is absolutely no basis for saying he may be able to have a babip in the 350 range!!”
“I agree, no basis whatsoever. Especially since he’s been at .380 over his last 1000 PA”
Two Thumbs up Sooted72
Ever been in a bar where somebody keeps repeating something really dumb over and over until finally a fight breaks out?