Bullish or bearish? Bo Bichette

This short series will examine several Blue Jays with shorter track records around whose production there is significant uncertainty and the reasons to be both bullish or bearish on what they will do in 2021. Having examined Teoscar Hernandez and Rowdy Tellez, today we turn to Bo Bichette.

Projection Systems:

• Steamer — .279/.331/.466 in 673 PA, 109 wRC+ and 3.8 WAR
• ZiPS — .286/.334/.512 in 545 PA, 119 wRC+ and 4.0 WAR
• Marcel — .286/.344/.503 in 394 PA, ~120 wRC+
• The BAT X — .285/.339/.478 in 644 PA, 118 wRC+

Thus far in his career, Bo Bichette has been explosive at the plate, posting .307/.347/.549 line in 340 plate appearances, good for a 134 wRC+ that makes him an expected centrepiece of the lineup for 2021 and years to come. Of course, between the shortened 2020 and the injury that kept him out a month, that still only amounts to a track record of half a season.

And there are some cautionary notes. Most fundamentally is the .361 BABIP that has driven that standout production. A good portion of that comes from hitting the ball hard, but at the same time, it’s at the extreme historical extent of talent. In the modern integrated era, Rod Carew is the career leader at .359 (min. 2,000 PA) — and most at the top of the list were lower power contact hitters as opposed to big swinging power hitters.

Second, Bichette’s walked rate slipped to just under 4% in 2020. This is not an issue in itself, as we’ve seen with Teoscar and Tellez earlier all else equal we want players who are very productive on batted balls putting the ball in play more. But it flows directly out of a very aggressive approach (54% swing rate compared to ~46% league). It’s one thing to have an OBP only 30-40 points above batting average when the latter is above .300. But if Bichette’s BA slipped to more like .270, then it’s going to be a lot less palatable especially higher in the order.

The risk is that if pitchers can exploit that aggressiveness to manage his batted balls, then with a low walk rate and close to average strikeout rate, there’s much less on which to fall back. If I adjust his line to regress his BABIP to .320 with only a minor hit to power (remove six singles, three doubles and one home run for a .223 ISO vs. 241 actual), then his line drops to .276/.318/.498. Of course, that’s still quite good coming from a middle infielder, just not star level which for better or worse is where expectations are sitting.

Some degree of regression on batted ball output is virtually inevitable, with the critical consideration being how much. And that’s what the projections vary. They’ve all got a very similar (and perhaps generous) walk rate around 7% and a strikeout rate around 20%. By contrast, BABIP estimate range from .310 from ZiPS on the low end to a sublime .335 from The BAT on the higher end. That 25 point range makes a world of difference, between a good regular and an All-Star.

Diving into the Statcast data/outputs, some of the high levels numbers are indeed cautionary. Bichette’s average exit velocity of 89.7 MPH is fine but not standout at the 57th percentile. His “hard hit” rate is likewise at the same level. Notwithstanding that, he has excelled at finding “barrels” with a 10% rate that’s well above the MLB average of 6%, putting Bichette among the elite at the 82nd percentile.

While quality contact is a function of pairing up good exit velocities on launch angles on particular balls rather than aggregate averages, the Statcast algorithms that combine those two to predict production for each ball in play think he has been quite fortunate. This is reflected in a career .372 wOBA compared to .341 expected wOBA,

Or to put it in terms I find more much more intuitive: on his 243 career balls in play, his wRC+ is about 210, whereas the Statcast algorithm estimates it should be around 165 (and compared to league average around 135-140). Bichette has generated about 65 runs compared to about 51 that would be expected (and 42 by a hitter of average contact ability).

In that sense, Bichette has thus been quite “lucky”, or at least experienced positive variation, and seemingly due for significant regression. But while those expected values have sound methodology and are directionally correct, some players persistently out- and under-perform them. I first noticed this a couple years back with Kendrys Morales, and while the precise reasons are for another day, suffice to say that when there’s a big deviation as with Bichette I want to dig under the hood for a closer look at whether it’s a case of unsustainability or something systemic.

So I pulled Bichette’s Statcast output and for both exit velocity and launch angle created different buckets to examine in detail:

Starting with exit velocity, there’s some good news in that almost none of the performance is on balls hit very weakly, so it doesn’t look like he got lucky on some squibbers. But there is significant performance in the next bucket ofballs hit 80-90, which is still pretty weak. It effectively amounts to about four more hits than expected, but some extra bases too, so this bears drilling down on. Likewise, in each of the last two buckets there’s significant outperformance, amounting to a couple extra hits as well as more power than expected.

Switching to launch angle, we likewise see the outperformance is quite broadly distributed across different batted ball types. The exception is balls hit more than 30 degrees, which again is a good thing since we can rule out the fluke of balls hit high into the air disproportionately avoiding gloves.

The observed outperformance varies quite a bit in the other categories. In the “topped” bucket, it’s effectively beating out a couple extra singles. Maybe that’s speed, probably luck that regresses but ultimately small potatoes. In the line drive and fly ball buckets, (7 to 30 degrees), it’s a couple extra hits and additional power in each bucket. But the biggest outperformance is in the ground ball/low line drive (”gliner”) bucket, where Bichette’ piled up five extra hits, with added power.

Combining these criteria, while the outperformance is pretty widespread across batted balls, it’s particularly concentrated in the 80-90 MPH velocity range, and the -10 to 7 degree launch angle range. Courtesy of the fantastic MLB Film Room tool, I’ve strung together clips of some of these balls. Here are the five hits from the ground ball bucket with the most outperformance between the outcome and what was expected (in descending order):

There’s definitely some fortunate in there, including placement of hard hit balls, but certainly the excess power seems for the most earned (I wouldn’t be surprised if at least some of the extra singles regressed).

I was even more curious about balls in the 80-90 MPH range, since fundamentally that’s pretty weak contact.

There’s some fluky stuff there of course, but also good, if very well placed, contact. Considering this is a selection of the balls on which he most “overachieved”, there’s actually pretty good sustainability.

Overall, I’m encouraged by a couple things as regards the batted ball “outperformance”. First, it’s pretty broad nature across ranges of exit velocity and launch angle suggest not one factor driving it. Second, the areas where it doesn’t show up are the ones where it would be most fluky (very weak, balls in air).

Conclusion: Bullish. I was expecting to be (relatively) quite bearish based on ample batted ball regression, and while there will be some, not to the degree wxOBA would suggest based on what’s actually done. My base projection would be something like 170 wRC+ on contact, which with 7% walks and 20% strikeouts would work out to about 125 wRC+.