Examining the Toronto Blue Jays’ hitting with Statcast’s xwOBA


Jays From the Couch looks at the Toronto Blue Jays offense by applying Statcast’s xwOBA




After the Blue Jays’ terrible start to the season, I dug into the team’s walk, strikeout and batted ball-contact quality rates. I wanted to see if their 1-9 start was because their plate appearances were resulting in more “bad” fundamental outcomes (strikeouts, soft/medium-hit groundballs or soft/medium-hit fly balls) and fewer “good” fundamental outcomes (walks, line drives or hard-hit groundballs/fly balls). If this was the case, hope for a turnaround in the final 152 games would seem slim. Fortunately, I found that there really wasn’t much difference between their fundamental outcomes in 2016 and 2017. In their first ten games, Jays hitters walked (10%), struck out (22%), hit hard flies and grounders (14%) and hit liners (14%) at roughly the same rate as they did in 2016.


In their last eight games, most of these measures have gotten a little worse. For the season, Jays hitters have walked in 8% of plate appearances, struck out in 24% of PA, hit hard flies and grounders in 16% of PA (the one area of improvement) and hit liners in 13% of PA. That said, these updated numbers probably aren’t all that surprising. The team has lost Troy Tulowitzki and Josh Donaldson to injury. On top of that, batters that are on a cold streak can sometimes change their process for the worse. They might grip their bat a bit tighter, swing a little harder and think too much. I’ve certainly seen it first-hand in the relatively low-pressure world of recreational softball. I can only imagine how batters react to cold streaks when millions are watching.


In light of that exercise, I thought it would be interesting to look at whether individual batters are performing “fundamentally” better or worse than 2016. I will focus on the Blue Jays’ batters as individual pitchers still have too small a sample size. And really, the Jays pitching staff has been pretty damn good. Pre-season projections had both the starting rotation and the bullpen finishing in the middle of the pack. So far, the starters have the 13th highest WAR in the majors (1.2 WAR), while the bullpen has the 10th highest WAR in the majors (0.7 WAR).


This sort of exercise might help us understand if specific batters have performed fundamentally different from 2016 and/or have been lucky/unlucky. That said, a change of approach might be useful. When comparing a team’s fundamental hitting outcomes from 2016 to 2017, I only had one set of comparisons to make.  With 14 batters getting playing time this season, I’d have 14 sets of comparisons to make. Comparing the value of different outcomes further complicates things. If a batter has hit 10% more line drives but has also struck out 10% more, should I conclude that they’ve gotten better, worse or stayed the same overall?


So I started digging and found a potential solution: a Statcast statistic called xwOBA. In a nutshell, xwOBA is a fielding independent offensive statistic. For a given batted ball, Statcast will calculate the probability that it will be a single, double, triple or home run (based on exit velocity and launch angle). So, whereas wOBA would only count actual singles, doubles, triples and home runs, xwOBA counts the expected number of singles, doubles, triples and home run caused by a batted ball’s exit velocity and launch angle. (xwOBA counts walks the same way wOBA does.) wOBA reflects both a player’s hitting talent and their opposition’s defensive talent. xwOBA only reflects a player’s hitting talent.


This methodology should help make it clearer whether much has fundamentally changed for each player from 2016 to 2017. Simply put, xwOBA should tell us whether the batter is, overall, doing more good things or more bad things at the plate than last season. As with basically all of my posts, take the analysis with a grain of salt. I’m taking a stab at examining a fairly tricky topic, separating signal from noise, talent from luck. I welcome thoughts and suggestions, as they can only help improve my analysis and our collective understanding of these interesting ideas.


For some context, wOBA and xwOBA can be interpreted like on-base percentage (OBP):

  • .290 is awful
  • .320 is average
  • .370 is great


Below are a couple of tables. The first shows player-by-player changes in PA outcomes. The second (my main focus) includes a measure of fundamental performance (xwOBA), a result-based measure (wOBA) and measures of luck (BABIP and HR/FB%). This way, we can see if a player’s results have changed and then examine if that change was the result of luck, changes in fundamental performance or both. [Note: All data current up to and including Sunday’s game against the Angels.]


Let’s start with the man, the legend, Jose Bautista. Thus far, he has seen a massive drop in his results, well examined by our site. He’s striking out a ton and it took him until his 16th start to hit his first home run (though what a home run!). The data suggests that his problem is a mix of bad luck (his BABIP and HR/FB% are way down) and a drop in performance. His xwOBA has fallen, but by about half as much as his actual wOBA has. If his luck can return to normal, his results should improve. If he can also figure out how to deal with all of the breaking pitches he’s been seeing since the ALCS, his results should improve considerably.


Kendrys Morales has been a great addition for the Blue Jays. In a season where so many unexpected things have happened, it’s comforting to have such a consistent batter in the heart of the lineup. There is little difference between the results and luck he had in 2016 with those he’s had in 2017. If anything, he’s had better plate appearances in 2017, as his xwOBA is up to .411 (19th in the majors among batters with 40+ ABs).


One of the keys to our offence in 2017 has been Kevin Pillar. I want to begin by saying that I love Kevin Pillar. He is just so good at defence (and base running) that production with his bat is just icing on the cake. Based on my back of the envelope calculations, Pillar could’ve been considered an average MLB player with a wOBA as low as .260 (!) in 2015-16. For comparison, the worst qualified hitter in baseball over the last two seasons (Alexei Ramirez) had a wOBA of .272.


While Pillar has been a bit lucky this year (his BABIP and HR/FB% are both way up), his xwOBA (7th among CF) is also higher than last season, reflecting a fundamental improvement in his hitting approach. It seems fair to say that his improved results are about half good luck and half improved hitting. A specific area of improvement for Pillar is in the quality of contact he makes on grounders and flies (more hard-hit GB/FB and less soft/medium-hit GB/FB).


The Jays’ starting middle infielders have both been the victims of bad luck thus far. Tulo’s lower wOBA seems partially the result of a much lower HR/FB%, though his xwOBA is down a decent amount from 2016. Devo’s lower wOBA is the result of both a much lower BABIP and a slightly lower HR/FB% (Sunday’s dinger helped that situation). His xwOBA has actually increased from 2016. In fact, Devon Travis might be one of the unluckiest batters in the majors so far in 2017. He has the 4th highest difference between his xwOBA and wOBA in the majors (among batters with 40+ ABs). Highlighting how unlucky the Jays have been, Kendrys has the 11th highest xwOBA – wOBA, while Jose is 24th (though some of their gap is likely due to a below-average ability on the base paths).


Justin Smoak, the other key to the Jays’ offence thus far, is raking. Fortunately, he is genuinely, legitimately raking. His BABIP and HR/FB% are stable, while his xwOBA is up (11th highest among 1Bs). He’s walking less, but he’s striking out a lot less too. In place of walks and strikeouts, he’s making more contact. Sure, he’s making more weak contact. But he’s also making more hard contact as well. This is a Justin Smoak that Jays fans can be happy with. Given his normal level of luck, we might get to enjoy this Justin Smoak for the whole season (fingers crossed).


Russell Martin is a lot like Kevin Pillar. He is so valuable to the team in non-batting ways that we should cut him some slack when he’s not hitting well. That said, his track record has shown what he can do at the plate. Fortunately, unlike Kevin Pillar this time, he seems to be another Jay dealing with bad luck. His wOBA and xwOBA are both down from 2016, which isn’t great. But his BABIP and HR/FB% are down even more so, confirming that bad luck has been partly responsible for his struggles at the plate. As they regress to normal levels, his results should return to the level we expected (.310-.320 wOBA).


Steve Pearce was a pick-up that I have very high hopes for. Things haven’t quite worked out so far (as the gif above can attest to) and, in contrast to teammates like Russ and Devo, his issues can’t all be chalked up to bad luck. Pearce’s xwOBA is down from last year, suggesting that part of his wOBA drop is based on some fundamental issues. Specifically, he has seen his walk, line drive and hard contact rates drop (replaced with a lot more strikeouts and some more weak contact). In fairness, his BABIP and (especially) his HR/FB% are also down, so a normalization of luck should help him out a bit. Like Jose, Pearce is a guy who could really use a home run or two to rebuild his confidence. His record has shown that he can produce at the plate.


Pearce’s struggles at the plate have led to more PA for Ezequiel Carrera. Zeke is a fine fourth outfielder, but he has never been a strong producer with his bat (career wOBA of .293). Prior to Sunday, he had been underperforming even those numbers, producing a .272 wOBA. A great Sunday (highlighted by a triple) pushed his wOBA up to .322 for the season. Nevertheless, his xwOBA has fallen off a bit from last year. I’ll be rooting for more Zeke triples, but I’ll also be rooting for Pearce to recover his mojo as soon as possible.


It hurts to see 35 PA beside Josh Donaldson’s name. He has (comfortably) been the second best position player in baseball over the last four seasons (his 30.5 WAR is much higher than third-place Paul Goldschmidt’s 22.7 WAR), but he’s played about half as much as we’d hoped. He was raking while he was playing, which is good. But his results seem mainly luck-driven, with a high BABIP and HR/FB% covering for a decreased xwOBA. That said, he’s Josh Donaldson. Get well soon, Josh!


The group of reserves are a mixed bag of small sample size fun. In spite of what Buck Martinez tells us every day, Darwin Barney is not a good MLB hitter. [That said, he is an elite MLB defender who produces plenty of value with his glove. Among all infielders during the advanced fielding stats era (2003-2017), Barney has the 25th most defensive runs saved (69 DRS) and the 20th highest ultimate zone rating (50.9 UZR). He’s done so while playing only the 169th most innings (5228 innings) in an infield position.] His wOBA is down, in spite of good BABIP luck. Sure, his HR/FB% is 0%, but it was never very high to begin with. Compared to last year, he’s replaced walks, weak contact and hard contact with a lot more strikeouts, resulting in a lower xwOBA.


The one they call GoGo has been playing like a guy whose job is on the line and I like it. While he’s been the recipient of some good BABIP luck (and a lot of good HR/FB% luck after Sunday’s home run), he’s also been hitting fundamentally better. He’s striking out a lot less, while walking more and hitting more line drives, which have driven a massive leap in his xwOBA. And he’s done so while facing a lot more lefties than he usually would. A .350 wOBA is a stretch for Goins, but if he can maintain one close to the .297 wOBA he produced in 2015 (rather than the .232 wOBA he produced last season), he will be a valuable piece for the Jays during their upcoming 95-49 finish to the regular season and 11-1 playoff run (I don’t think they’ll sweep all three series—that would be ridiculous).


Jarrod Saltalamacchia (it isn’t hard people—it’s spelt just like it sounds) has been striking out with breath-taking regularity. I have all the patience in the world for a player with a 14-letter Italian last name that includes “cchi”, but a strikeout rate of 66.7% is troublesome. Hopefully that regresses to his much more respectable career K-rate of 30.7%. And then maybe keeps regressing down to an even more respectable 25%. One can only hope.


As a fan of Chris Coghlan, I’m glad to see him on the 25-man roster. Unfortunately, the circumstances (Josh on the DL) obviously take a bit of the shine off. He is probably the trickiest Jay to analyze right now, with his super small sample size. His wOBA is down, so he’s definitely not producing as well as he’d like. He’s been unlucky so far (with decreased BABIP and HR/FB%), but his xwOBA is down too. He seems like a Jay we need to patiently keep an eye on as the season progresses and his sample size increases to a point where we can make more concrete observations about him.


Examining the Blue Jay hitters individually offers some hope for the rest of the season. There are a few bright spots (Morales, Pillar, Travis and Smoak), batters who are producing fundamentally good outcomes at the plate and not simply riding a lucky streak. Russ is generating similar fundamental outcomes as 2016, but suffering from a lot of bad luck. Our bench options are the (small sample) variety pack we’d expect them to be at this point. Goins has seen fundamental improvements reflected in his improved results. The rest (Carrera, Barney, Coghlan and especially Saltalamacchia) have seen their results and performances drop off to varying degrees. On less positive notes, Tulo and Josh are hurt, while Jose and Pearce’s troubles are part bad luck and performance drop. We will have to wait and see if any or all four of them are able to finish up the remaining 144 games (!) of the season healthy and productive.


A bird’s-eye-view of the team comes to similar conclusions. While the team’s wOBA is down from 2016, the team’s xwOBA is down by a little more than half as much. [Full disclosure: A bug in the Statcast system temporarily gave the Jays a .330 xwOBA on Monday evening, giving me a temporary feeling of euphoria. I’ve requested that MLB honour the “no take backs” rule and use their Statcast magic to retroactively improve the Jays’ batting performances from earlier this season. I await their response.]



On the other hand, the team seems a bit unlucky, with a lower BABIP (21st in MLB) and HR/FB% (24th). Walks and strikeouts are areas for improvement but, otherwise, they’re making roughly the same amount of good and weak contact as last season. 5-13 is a bad record. As many have mentioned, those losses cannot be undone and have put them in a potentially insurmountable hole. Nevertheless, the Jays still have a realistic shot at the playoffs (Fangraphs gives them a 17.3% chance) and they should end up with the highest final win count of any team that started the season 1-9 (Fangraphs projects us to finish with 79 wins). Our AL rivals have helped us out as most teams (11 of 15) remain bunched together with 8 to 11 wins.


Things could be better, but I’m confident that (especially if we get a good run of health soon) the Jays will keep things interesting this summer. You might have to squint some of the time, but you can still see an average (or slightly better) offence in these numbers. The big free agent signing is working out well. The pariah might finally be fulfilling his potential. Superman can hit. The young, “injury prone” second baseman is hitting better than he looks. Combined with the returns of Josh and Tulo, solid pitching and strong defence, we still have the foundation of a team that competes for a playoff spot. I think we might as well get our rally caps ready and try to enjoy the ride.




*Featured Image Credit: Arturo Pardavila III UNDER CC BY-SA 2.0








Jeff Quattrociocchi

I'm an economics professor in the GTA whose lifelong love for the Jays was reignited by that magical August of 2015 and the amazing moments since.