2017 Toronto Blue Jays: An explanation for my continued optimism


 There is still reason to be optimistic about the Toronto Blue Jays in 2017. Here’s a break down of what fuels the positive vibes


Embed from Getty Images


It’s a long one, so here’s a TL;DR summary

The Blue Jays’ underlying batting and pitching performances (in terms of xwOBA, xBA, plate discipline and contact quality) have been pretty solid this season, essentially fifth-best in the AL after the main contenders. Ultimately, these performances have not been reflected in the results for a few (likely) reasons.


On the batting side, the lack of speed is definitely an issue, but a relatively minor one in comparison to the 2016 Blue Jays. Two more important factors are driving the Jays’ poor results: their difficulty dealing with the infield shift and unexplained bad luck.


On the pitching side, the defence is a big factor, having fallen off significantly from last year. While it explains part of the bad batted ball luck Blue Jays pitchers have had to deal with, much of the bad luck remains unexplained.


Given the roles that bad luck and bad defence have played in the poor first half results, hope for the second half remains. The bad luck is something that could correct itself through regression to the mean. Similarly, the poor defence of the first half can be explained away by a few aberrations that may correct themselves in time as well.


Over his career, 300 innings of Ryan Goins at shortstop resulted in an average of 5 defensive runs saved (DRS). This year, he’s produced -5 DRS at SS. Darwin Barney’s time at second and third this season has produced -4 DRS. Based on his career numbers, Barney would have been expected to produce about 6 DRS at those positions (given his 2017 usages). We’d also expect Kevin Pillar (7 DRS in CF) to have produced an extra 5 DRS so far this year.


Even if we leave aside everyone else, these three solid defenders performing at career norms would have put the Jays’ team defence near the middle of the pack, tied for ninth in the AL with the Yankees at -3 DRS. Am I creating an alternate reality to make the Jays’ defensive situation less frightful? Yes. Is this alternate reality still well-grounded and reasonable? Yes!

If this sounds interesting, feel free to join me through the next 2000 (or so) words for a more detailed explanation.


My Baseball “Worldview”

My goal entering the 2017 season was to approach my Blue Jays fandom in a healthy, even-keeled way. In the past, I often invested myself fully into one of the teams I support and rode the day-to-day ups and downs wherever they took me. When the team was succeeding, life was great. When it wasn’t, I was left frustrated. As a quasi-lapsed fan of the Maple Leafs and Arsenal, it was a lot more of the latter than the former.


Among all sports, baseball seems to have the most dramatic swings in fortune. Even 600 plate appearances can be a small sample size, with wildly positive and negative outcomes occurring regularly. But, at the same time, it is a very quantifiable sport which allows us to understand those ups and downs better than the fans of most other sports can. In particular, we now have a pretty strong ability to think about the sources of those day-to-day, week-to-week and month-to-month variations in results, allowing us to separate changes in underlying performance/talent from fluctuations in luck.


Fluctuations in a player’s underlying performance are the key things to focus on. If an aging slugger is seeing his strikeouts increase and his barrel rate fall, it often ends up corresponding to a permanent decline in his production. On the other hand, if a player’s maintaining their normal quality of plate discipline and contact, but is seeing more of his balls in play end up going to fielders for outs (low BABIP), his poor run is likely temporary.


The diagrams below reflect the way I’ve been trying to think about baseball this year. To be clear, I didn’t invent any of this. Instead, I’ve just synthesized all of the great baseball writing I’ve read and absorbed into this sort of mindset.


Players (and teams, more generally) have control over two general aspects of offensive and defensive performance:

  • Batter/Pitcher Performance: Their approach during plate appearances, which influences things like plate discipline and contact quality
  • Base running: On the offensive side, speed and decision-making on the base paths
  • Fielding: On the defensive side, the general ability to make outs, determined by countless factors like reaction time, athleticism, arm strength, etc.



Luck captures all of the things beyond one’s direct control. On both sides of the field, luck plays an enormous role. More broadly, I’m of the mind that luck plays a massive role in all aspects of life, to an extent that we struggle to understand and get comfortable with. Accepting that we can only control so much is scary. It can also be hard to see, so we use language that removes luck and replaces it with various degrees of intent and choice.


If our luck is good, we assume we did something positive to create the good outcome. If our luck is bad, we must’ve done something wrong along the way. A batter is congratulated for an infield single off of a swinging bunt, but might be asked to “try and keep the ball down next time” after a well-struck 399-foot fly out.


The massive (apparent) role of luck in the Jays’ 2017 season

As…chance would have it, my new mindset was tested right away and has continued to be tested all season long. So far, it has helped me achieve that even-keeled fandom that I was hoping for. I think it is key as to why I’ve maintained a much more optimistic view of things than most Jays fans throughout the season.


To be clear, I’m honestly not being optimistic just to be contrarian or because I’ve drank the Blue Jays’ Kool Aid. The season has been extremely frustrating. Some games have felt like an eternity. But part of the reason they’ve been frustrating is that I’ve seen too many well-hit balls find gloves and too many good pitches end up in some open grass or in the bleachers. Ultimately, these observations have been confirmed to a large degree by the data (as I’ll explain in this post).


This outlook has led me to a counter-intuitive conclusion: the Toronto Blue Jays’ first-half performance was far, far better than their results would suggest. They still aren’t world beaters, but there’s more than enough positivity to keep a fan optimistic about the rest of the season.


Now, part of the reason I can say that they’ve outperformed their results is because their results have been extremely objectionable. In particular, the Blue Jays have been well below-average in both runs scored (14th in the AL) and runs allowed (10th), two of the most important determinants of wins and losses.


However, while runs scored/allowed are super important, they are also very noisy and do not only reflect underlying performances by batters, pitchers, fielders and runners. These stats also reflect a team’s luck on both sides of the field. In that regard, the Blue Jays stand out as arguably the (statistically) unluckiest team in the AL.



Starting with just a few numbers (we’ll get into more detail shortly) should help clarify my idea. Strong performances at the plate and poor base running have combined to produce an even poorer offensive output. The missing link? The worst offensive batted ball luck in the AL. Similarly, even stronger performances on the mound combined with very poor fielding have combined to produce a poor defensive output. Once more, the missing link is some of the worst defensive batted ball luck in the AL.


Offensive performances have been no worse than average

That doesn’t seem like a compliment, but it’s about the nicest thing Jays fans would say about the team’s offence this year. It’s also the truth. Blue Jay hitters have done a solid job of generating the exit velocities and launch angles that generally lead to extra base hits (and hits, more broadly). Not great, but solid. They have produced an average number of “good” plate appearances (those resulting in a walk, barrel, solid contact or a flare/burner) and a lower than average number of “bad” plate appearances (those resulting in a strikeout or any type of poor contact). While the team doesn’t stand out in any one stat (among BB%, Barrel%, Solid+Flare%, K% or Poor Con%), the Jays rank no worse than 10th in all of them, with that consistency key to their slightly above-average xwOBA and xBA.




The offensive performance drop-off from 2016 has been a lot smaller than it seems

Last season, the Blue Jays were not the offensive juggernaut they were in 2015, but were still comfortably above average, ranking fifth in the AL in runs scored (4.7 runs per game). In spite of their much lower number of runs per game in 2017 (4.2 so far), their underlying offensive performance hasn’t changed much relative to last year.


The team’s xwOBA, Good PA% and Bad PA% have changed for the worse, while the team’s xBA has actually improved. However, in each case, the statistic changed by relatively small amounts. If the 2017 team had replicated the 2016 edition in each of these stats, they would’ve only moved up or down by one or two spots in the AL rankings.




Offensive batted ball luck has noticeably declined

The main difference between the offences of the 2016 and 2017 Blue Jays is batted ball luck—BABIP is down, while xwOBA-wOBA and xBA-BA are up. These changes mean that compared to last year, balls in play are turning into outs more often (lower BABIP) and hits and extra base hits are not being generated as much as we’d expect given the exit velocity/launch angle of batted balls (higher xBA-BA and xwOBA-wOBA). In each of these statistics, the Blue Jays rank among the AL’s unluckiest.



The slowest baserunners in the majors, two years running

According to Statcast’s new speed metric, the Blue Jays have the slowest average foot speed in the majors (26.1 feet per second). In terms of base running ability, they seem to use their veteran cunning to outperform their speed (-11.9 Base running runs, 12th in AL). While the running issues are troubling, they didn’t stop the Jays last year. The team posted a 26.8 ft/s average sprint speed (also last in the majors) and a -7.3 BsR in the first half of 2016. So base running alone has essentially cost them only 0.05 runs per game thus far, relative to last season.


In a more general sense, relative to batting, base running is not that important when it comes to offensive production (though, obviously, it can be super important in specific situations). Last season, the difference between the best and worst base running teams in the AL (38.6 base running runs) was only 21% as large as the gap between the best and worst batting teams (184.4 batting runs).


Can the “bad luck” be explained in part by slow runners and/or the shift?

While intuition would suggest that slow runners might produce a sub-par BABIP, I found no correlation between BsR and BABIP (using data since 2002, when BsR was first introduced).



However, batters facing a defensive shift do tend to produce a lower BABIP. Since 2015, the MLB BABIP without a shift in place was .299 vs. .294 when a traditional shift was used. This seems relevant when comparing the 2016 and 2017 Blue Jays. In the first half of 2016, Jays batters faced a traditional shift in 508 PA and produced a .278 BABIP. This season batters have faced a traditional shift in 715 PA (an increase of about 40%), producing a .274 BABIP. A similar BABIP, but many more PA in 2017.


That said, the shift doesn’t explain the entire BABIP difference—the 2016 Blue Jays produced a first half BABIP of .296 without a shift in place, while the current team has produced a .285 BABIP. Thus, the possibility remains that this team has indeed had some bad batted ball luck on the offensive side.


Some evidence of true bad batted ball luck

One such example of true bad batted ball luck is the share of well-hit air balls the Jays have hit that have found opposition gloves (where a well-hit air ball is any ball that Statcast deems a fly ball or line drive and either a barrel, solid contact or a flare/burner). These are at-bats that shouldn’t be affected much by runner speed or an infield shift. The main factors here are likely the quality of the opposition’s outfield defence and the ball’s horizontal launch direction, two things generally out of the batter’s control.



The Jays rank third in the AL in the rate of outs on well-hit air balls. For comparison, in spite of a very similar number of well-hit air balls, the Yankees have generated 41 fewer outs on these balls than the Jays have. Given the high value of these sorts of batted balls (.868 wOBA across the MLB in 2017), 41 fewer outs on well-hit air balls over 88 games should have a huge impact on a team’s run production.


Pitching performances have been really good

One of the most disappointing by-products of 2017’s terrible luck is the lack of credit going towards the Jays pitching staff. Their ability to generate what should be unproductive plate appearances ranks among the top five, along with the four main AL contenders, in terms of xwOBA and xBA. In terms of the rate of “bad” and “good” plate appearances, the Jays staff ranks fourth-best. The numbers make it pretty clear that there is little separating the pitching performances of the Yankees, Clevelanders, Red Sox and Blue Jays (all following behind the Astros’ juggernaut).



There’s also little separating the performances of the 2016 and 2017 Blue Jay pitching staffs

The Jays’ 2016 ALCS run is credited mainly to a great pitching staff—in particular, an incredible starting rotation and a so-so bullpen. Last season, the Jays surrendered only 4.1 runs per game, best in the AL. While it hasn’t always been obvious, the current staff has generated even better underlying stats: a lower xwOBA, a lower xBA, fewer “bad” plate appearances and more “good” plate appearances. This is made all the more incredible given the severe injury troubles the pitching staff has faced.



While the bullpen has been particularly exceptional this season (its .288 xwOBA is well below the MLB average of .304), the starting rotation has still managed to be perfectly average (with a .320 xwOBA edging out the MLB average of .321) in spite of its constant state of flux.


Jays pitchers seem to have had a very unlucky first half

What appears to separate the two pitching staffs is luck (or something resembling it), which has led to much worse results. Compared to last season, the Jays’ pitching BABIP has soared, while the team’s xwOBA-wOBA and xBA-BA have fallen dramatically. This bears out in the 2017 AL rankings—the Jays are fourth in BABIP and dead last in xwOBA-wOBA and xBA-BA. These changes mean that compared to last year, balls in play are turning into outs less often (higher BABIP) and hits and extra base hits are being generated at a faster rate than we’d expect given the exit velocity/launch angle of batted balls (lower xBA-BA and xwOBA-wOBA).



Are these numbers the result of true bad batted ball luck or bad defence?

This is a super important question going forward. Bad luck can presumably correct itself more easily than bad defence. I tried answering this question using as much data as I had available. In particular, I wanted to see whether there was a strong relationship between defensive runs saved (DRS) and each of the three luck statistics.


In the case of DRS and BABIP, I was able to use data from 2003 to 2016 (while BABIP data is available well into the past, DRS was only introduced in 2003). There appears to be a modest, negative relationship between DRS and BABIP—the better a team’s defence, the lower its pitching staff’s BABIP. I used the trend line equation to estimate what the Jays’ pitching BABIP would based only on DRS (.299). The result implies that only part of the Jays’ high pitching BABIP (.306 vs. the league average of .294) can be explained by the team’s poor defence.



In the case of xwOBA-wOBA and xBA-BA, the samples are much smaller, as full season data is only available for 2015 and 2016. The data suggests that while teams with a good defence tend to produce wOBA/BA that are lower than expected, the relationship is fairly weak (an R-squared of only 0.09). A larger sample will be needed to make more definite conclusions.



Putting this all together, it seems like Blue Jay pitchers have been genuinely unlucky this year, even after accounting for the team’s poor defence. The team’s xwOBA-wOBA (-0.023) and xBA-BA (-0.021) are particularly notable. No team in the Statcast era (2015-16) has produced such low values over the course of a full season, even though a handful of teams have had worse or comparable defences. This strongly implies that a regression to the mean is in order.


Wrapping it all up

There are a few key takeaways from all the stats we’ve pored through in this post.

  • There is some solid evidence that the Jays’ underlying batting and pitching performances have been good enough to crack the AL’s top five.
  • There is a lot of evidence that the Jays have dealt with a lot of bad batted ball luck, in addition to bad injury luck and defensive underperformance by a few stalwarts (which might itself be considered bad luck)
  • The Jays are extremely slow base runners, but the importance of this in terms of run scored (especially compared to last year) is probably overstated
  • There is a negative relationship between team defence and batted ball luck, specifically when captured by BABIP. However, the Jays’ poor defence doesn’t fully explain their bad pitching luck.


These ideas drive my continued optimism in the Blue Jays’ 2017 season. Most importantly, they drive my continued belief that this is a team very much worth rooting for and supporting. The hole they’re in means that there’s a good chance they’ll fall short of a playoff spot. But I intend to hold onto my hope until the math tells me not to.






*Featured Image Credit: C Stem- JFtC









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.