Blue Jays’ Kevin Pillar: Caution is understandable, but he is hitting better than ever

Kevin Pillar is off to another blistering start, and Blue Jays fans can get excited because this is a whole new level for the center fielder.

Kevin Pillar is in the midst of a hot start to the season. Jays fans are understandably being very careful not to get too excited about it, as he has developed a habit of starting well and then cooling off. Countless posts have been written on his start, as Jays fans are extremely curious if this is the season in which Pillar finally produces with his bat. In 2016 and 2017, he maintained average or better hitting production through his first 34 games, only to revert to below-average production for the remainder of the season. That said, he’s never had as good a start as this.



Now, he has had 34-game runs this good in his career, in terms of wRC+. There was a stretch in June and July of 2015; another one through the end of the 2015 and beginning of the 2016 seasons; and another from mid-April to mid-May 2017.



However, in terms of xwOBA, he has never had a run this good in his entire career. After 32 days (34 games, thanks to two doubleheaders), Pillar is running a .377 xwOBA. [Note: Statcast provides data by “Game Date”, rather than by game. So, while Fangraphs provides rolling averages by game, I can only create rolling averages by day with Statcast data.] That mark is very good—he currently ranks 69th out of 225 batters with 75+ AB. His previous 32-day best was a .367 xwOBA just before he made that inappropriate slur against Atlanta last May.



These observations set the foundation for this post. On the one hand, Pillar is hitting well and doing so without typical batted ball luck, evidenced by his excellent xwOBA. On the other hand, he did pretty much the same thing last season, from mid-April to mid-May, before regressing to his career norm. So, the question is, is there any evidence suggesting that Pillar is genuinely doing something different this time? I’m going to present a bunch of data, so let me share my key points up front:

  • In 2018, Pillar has produced a higher share of good plate appearances (45%) than he did during his 2017 hot stretch (41%).
  • This has driven him to a Top 10 xBA (.325).
  • In fact, he has never produced a high xBA over 32 days than he currently is producing.
  • Compared to the hot stretch of 2017, Pillar is generating more liners, more hard contact and significantly less soft contact.
  • His high BABIP (.356) doesn’t seem to be a sign of good luck, in this case, as he is producing an even higher xBABIP (.388), the best 32-day mark of his career.
  • His xBABIP is third-best in the majors. In the Top 10, it’s basically him and elite hitters
  • Similarly, his high HR/FB% appears to be well-supported by his increased exit velocity.
  • This increased exit velocity may be the key to his improvements. Harder hit balls tend to be more productive.
  • While subtle, Kevin Pillar has improved as a batter over his career.


Given his similar wRC+ and xwOBA over these two stretches, looking at the kind of plate appearances he’s had might help us better understand what’s been different about this stretch. During his hot month in early 2017, Pillar produced better barrel, walk and strikeout rates than he has thus far in 2018. That said, the differences aren’t huge, in each case a shift of about 1%.


On the other hand, he’s improved both his solid contact and flare/burner rate by about 3%. Thus, in 2018, he’s generated some kind of good contact about 5% more often and poor contact about 5% less often, compared to his 2017 hot stretch. Since barrels are more valuable than solid contact or flare/burners, his xwOBA has been about the same in the two 34-game stretches.



The similar xwOBA—.367 in the 2017 stretch, .377 to start 2018—reflects similar overall quality. However, as a contact hitter, it’s just as useful for him to produce a lot more good plate appearances (barrels, solid contact or flare/burners) rather than a bit more very good plate appearances (barrels). In that vein, it’s important to note that he’s doing good things far more often than he ever has, including that hot stretch last Spring.



This manifests most obviously in his (currently) elite xBA. Based on the contact he’s making (as well as his propensity for striking out), we’d expect him to produce a batting average of .325. That ranks ninth in the majors in 2018 among batters with 75+ AB. His actual batting average is .310, good for 23rd best in the majors. So, a good argument can be made that he’s been unlucky, thus far.



Importantly, while his 2018 xwOBA is only slightly higher than it was during his 2017 hot stretch, his xBA is currently much higher (.325 vs. .296). In fact, he’s only produced a .300+ xBA over a 32-day stretch once in his career, for a brief time in July 2015.



Fangraphs’ batted ball stats paint a similarly positive picture. He’s producing more line drives and more hard contact than ever before—he ranks sixth in LD% and 60th in Hard%. His 2017 hot stretch doesn’t look nearly as hot anymore. His line drive rate increased only a touch above his 2015-16 level and, while he increased his hard contact rate, he also saw his medium contact rate decrease. In 2018, the increase in his Hard% has come entirely from a decrease in his Soft%. In fact, he has the 27th lowest Soft% in the majors so far (among 179 batters with 100+ PA).



There’s a lot of evidence to suggest that Kevin Pillar is hitting better than he ever has. To be clear, that statement is intended to be reflective, not predictive—I have no idea whether his genuinely hot start will continue, though the strong fundamentals he’s displayed in 2018 justifies some optimism.


One stat that has been used against him in 2018 is his high BABIP, one of the classic sabermetric stats for batted ball luck. While useful, the issue with BABIP is that it doesn’t differentiate between periods where the same old batted balls are just finding holes and periods where a batter is making genuinely better contact.


Fortunately, Statcast allows us to examine a batter’s BABIP a bit more critically. In particular, by focusing only on “balls in play” (non-home run, non-strikeout at-bats), we can now use xBA to create an xBABIP and compare that with a player’s actual BABIP. The graph below shows the 32-day rolling averages for Pillar’s BABIP and xBABIP. What we see is that Pillar’s 2018 xBABIP (.388) is actually higher than his BABIP (.356), more evidence that he’s technically been unlucky thus far.



Importantly, this graph shows us the good fortune he benefited from during his previous hot stretches. Take the stretch during June and July of 2015. He ran a BABIP over .400, but his xBABIP barely cleared .350, implying that luck was turning a good stretch into a great one. Same thing for his end of 2015, beginning of 2016 hot streak—he posted a .400-ish BABIP, but his xBABIP peaked at .336.


Finally, last season’s hot streak from mid-April to mid-May. He ran a BABIP of .364, but an xBABIP of .338—more good fortune. In fact, if we follow the graph along, we see that while his BABIP plummeted from mid-May through to the end of June, his xBABIP remained quite stable. Only after June, perhaps because he let his bad luck affect his still-effective approach, did Pillar’s xBABIP fall as well.


Pillar has been elite thus far at generating the contact that leads to a high BABIP. The quality of batter alongside him seems like clear evidence that xBABIP is meaningful, even in small sample sizes—the other nine batters have been above-average producers over their careers, with each running a career BABIP no less than .319.



Again, the key takeaway is not that he will maintain this level throughout the season. Rather, his underlying hitting ability seems to genuinely be at a higher level than it has previously been. Performance and results will certainly fluctuate as the season progresses. However, the level his performance and results will fluctuate around may be elevated, which would be a boon for the Blue Jays and their playoff chances.


Pillar has massively improved his two-strike approach

The Sportsnet article I linked to earlier included a discussion with Pillar about working on his two-strike approach: “When the heart starts beating fast, when you’ve got two strikes, you tend to try to do too much and start expanding. I’m really focused on getting good pitches to hit and if they don’t show up, then I don’t swing.” His performance with two strikes (2018 slash line of .300/.348/.550) has been compared to all-time great Tony Gwynn (career slash line of .302/.340/.401 with two strikes).


Comparing his performances with and without two strikes takes some care. His numbers are clearly better with less than two strikes. However, that’s true for every batter. Instead of looking at his stats, look at where he ranks in the majors thus far. For each stat, he ranks a bit below-average with less than two strikes and in the top five percent with two strikes against.


Thus far in 2018, he has maintained great overall production with two strikes against (wOBA in the 95th percentile), supported by a great underlying performance (xwOBA in the 95th percentile). Most impressively, he has the best xBA with two strikes against in the majors thus far (minimum of 40 two-strike at-bats).



His improvement with two strikes against is evident when we compare his 2018 performance in the split against his past performances. During his early 2017 hot streak, he had a slightly above-average xBA with two strikes against, but was below-average in terms of the other stats. In 2016 and 2017, with two strikes against, he was able to produce a similarly just above-average xBA, but a well below-average xwOBA. 2015 was the only previous season he seemed to excel in two-strike counts, producing an average xwOBA and well above-average xBA.



Kevin Pillar has never hit this well over a similar time frame in his career with two strikes against him. Prior to 2018, his best 32-day xwOBA with two strikes (.304) against him came during his aforementioned end of 2015/start of 2016 hot streak, while his best 32-day xBA with two strikes against (.273) occurred in late 2017. His current marks (.351 xwOBA, .314 xBA) blow those previous bests out of the water and are both much, much, much better than his 2015-2017 averages (.225 xwOBA, .197 xBA).



What he’s doing with two strikes against him is unsustainable— in the Statcast era, the best full-season xwOBA with two strikes against is .360 (David Ortiz, 2016) and the best xBA is .269 (DJ LeMahieu, 2016). Pillar is not Big Papi, nor will he beat the previous full-season two-strike xBA record by 45 points. Nevertheless, what he’s doing is very real, not smoke and mirrors. So, while he may not hit this well with two strikes forever, he may certainly settle at a much better level than he was previously able to maintain.


Pillar seems to be swinging the bat harder and it could be the key to his improvements

There are likely a few key reasons as to why Kevin Pillar seems like a different hitter in 2018, but a big one is that he may be swinging the bat harder than ever before. Now, at this point in the Statcast era, we do not have publicly available information on a batter’s swing speed. Thus, in order for us to even say that Pillar has increased his swing speed (let alone say that this is what’s driving his changed batting performance this season), we need to make an assumption. Fortunately, it’s an intuitive one that is supported by data: increased exit velocity is a strong indicator of increased swing speed.


Well, Kevin Pillar has definitely increased his average exit velocity, by a statistically significant amount. After 34 games, he is maintaining an 89.3 mph average exit velocity, much higher than his previous full-season marks of 84.3 mph (2015), 86.2 mph (2016) and 85.5 mph (2017). His average exit velocity over the last 34 games is higher than any 34-game stretch in his career. In fact, this is the first time he’s maintained an average EV above 89 mph over a 34-game stretch.



Swinging the bat harder as a means to generate a higher exit velocity and better outcomes at the plate makes a lot of sense. For one thing, there is a moderate correlation between a batter’s exit velocity and their overall batting production (wRC+). The graph below includes all batters who had 400+ AB in 2017 (with Pillar’s 2018 data added for illustrative purposes). The data suggests that a batter who experiences a 3.8 mph increase in their exit velocity should see a 20 point increase in their wRC+. While that’s a bit less of a wRC+ boost than Pillar’s experienced at this point, 1) Pillar was due for a better wRC+ given his exit velocity last year and 2) that would still put Pillar in above-average hitter territory, which would be great given his defence.



The relevance of exit velocity to wRC+ is also evident in Pillar’s track record. Below I’ve graphed Pillar’s exit velocity and wRC+ rolling averages (expressed as z-scores to make the two stats graph-able on the same space). It would appear that each time Pillar was producing an above-average wRC+ (for him), he was also producing an above-average exit velocity (for him). The difference this time is that his exit velocity seems to be justifying his wRC+ to a much greater degree—the three other times that Pillar produced a similar wRC+ as he is right now, his exit velocity was only a bit above-average, implying that he was getting a lot of luck.



Pillar’s increased exit velocity is also reflected in his 13.8% home run-to-fly ball ratio, one of the highest marks he’s maintained over 32 days. A big jump in one’s HR/FB% would normally set off batted ball luck alarm bells. However, batters with higher exit velocities tend to also run abnormally high HR/FB%. Based on 2017 data, we would expect Pillar’s HR/FB% to increase by 6.6% given an average exit velocity increase of 3.8 mph—his HR/FB% is up only 4.9% at this point.



Over his career, Pillar has seen his HR/FB% rise and fall with his average exit velocity, for the most part. As such, we can be reasonably sure that, as long as he can keep up his exit velocity gains, he should be able to keep turning fly balls into dingers at an elevated rate.



That really is the key to everything about Kevin Pillar’s season—if he can keep up his current underlying performance, he can keep up his current production level (139 wRC+). If he can keep hitting liners and making hard contact and maintain his two-strike approach, he’ll keep producing. The question is whether or not we should count on that. The best we can say at this point is that he is doing a lot of important things at or better than prior career-bests. In those previous occasions, he generally returned to a below-average level, unable to maintain his gains. But we might as well enjoy it while it lasts and give him kudos for making genuine improvements at this point in the season.


I’ll end with one last graph: Kevin Pillar’s career 32-day xwOBA rolling average, but this time with a trendline superimposed on it. The trendline makes an underappreciated point. Pillar has fundamentally improved as a batter over his career. His hot streaks have gotten hotter and lasted longer.



All told, the trendline estimates 2015 Pillar to have been a .280-ish xwOBA batter and 2018 Pillar to be a .315-ish xwOBA batter. That tracks pretty well with his yearly totals—he produced a .282 xwOBA in 2015, a .286 mark in 2016 and a .310 xwOBA last season. For context, in 2017, the MLB average was .321.

Over his career, Pillar’s brand of great CF defence and below-average hitting has propelled him to a career as a solid regular (2.6 fWAR/600 PA). With above-average CF defence and hitting, he is currently on pace to produce 5.5 fWAR/600 PA. Given his defence, if his offensive production ends up stabilizing around (or slightly above) the league-average, he would be a very useful player (3-4 wins above replacement) for the Blue Jays.





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.