Using Fly Ball Exit Velocity

Barrel rate is an excellent tool for evaluating the true power-hitting ability of a slugger. But barrel rate does take about 50 batted balls to stabilize and most hitters do not yet have 50 batted ball events.

Even after a hitter’s barrel rate has likely stabilized, however, there are signs that it could improve or decline. One such sign that I use, particularly in the early going, is fly ball exit velocity. If a hitter has a lot of home runs supported by a great barrel rate, but he also exhibits poor fly ball exit velocity, there might be imminent regression to both his barrel and home run rates. Conversely, if a hitter isn’t hitting many home runs and has a poor barrel rate, but he hits his fly balls really hard, he might be getting unlucky.

To provide some context to early-season barrel rates, I’ve identified a few examples of both types of guys.

Context

Two years ago, I researched which Statcast metrics on the Exit Velocity & Barrels Leaderboard were best at predicting power hitting. I examined average launch angle, average exit velocity, exit velocity on fly balls and line drives (EV on FB/LD), hard-hit rate, barrels per plate appearance, barrels per batted ball event (Brls/BBE%), and FanGraphs’ Hard%. I had three key findings:

1. EV on FB/LD and Brls/BBE% were the stickiest year-to-year.
2. Brls/BBE% shared the greatest relationship with HR/FB% and ISO, followed by EV on FB/LD.
3. EV on FB/LD was more predictive of Brls/BBE% than any combination of the above metrics.

So what does this mean? Brls/BBE% is the best tool for predicting power hitting. In turn, EV on FB/LD is predictive of Brls/BBE%.

For lack of a better term, EV on FB/LD is useful for evaluating whether a hitter’s barrel rate is “earned” or due for regression. There’s just one more thing, however:

Recently, I looked at all hitters from 2015-19 with at least 50 fly balls in a season for a total of 1,228 hitters. I found the coefficient of determination (R2) between their exit velocities on fly balls and several power-hitting variables, as well as EV on FB/LD and those same variables. In layman’s terms, R2 illustrates how much variance in the sample of a dependent variable (e.g., HR/BBE%) is explained by an independent variable (e.g., EV on FB). The higher the R2, the greater the independent variable explains changes in the dependent variable (though R2 will never exceed 1).

The above table indicates that exit velocity on just fly balls, rather than on fly balls and line drives combined, shares a stronger relationship with all of SLG, ISO, xSLG, xISO, HR/BBE%, and Brls/BBE% than EV on FB/LD. As an example, exit velocity on fly balls explained 64.5% of the variance in the HR/BBE% sample and 76.5% of the variance in the Brls/BBE% sample. That’s why I prefer exit velocity on fly balls to EV on FB/LD.

And based on the strength of the relationships between exit velocity on fly balls and HR/BBE% and Brls/BBE%, eyebrows should rise if there’s incongruence between a hitter’s exit velocity on his fly balls and the rate at which he barrels the ball or hits home runs.

With that background, let’s look at some examples.

Positive Regression Candidates

An easy way to spot players due for more barrels and home runs is to derive an expected barrel rate based on fly ball exit velocity, and then determine who had the largest gap between their expected barrel rate and actual barrel rate.

I found such an xBrl/BBE% using the same sample as above — all hitters from 2015-19 with at least 50 fly balls in a season (n=1,228). In other words, xBrl/BBE% tells us what barrel rate a hitter should have based on how hard he hits his fly balls in light of barrel rates other players with similar fly ball exit velocities have managed in the past. The following are the 10 hitters with the biggest negative differential between their barrel rate and expected barrel rate (min. 10 fly balls):

Each of Rowdy Tellez, Justin Upton, Elvis Andrus, Francisco Lindor, Carlos Correa, and Marcus Semien have three or fewer barrels this year. On the one hand, for Tellez and Andrus, this is particularly surprising because their fly ball exit velocities were both top-25 among all 128 hitters with at least 10 fly balls. Tellez has 13 fly balls and hits them pretty hard, with only two barrels to show for it. Andrus at least has three barrels, but these two hitters are following a recipe that should yield more barrels.

On the other hand, none of Upton, Lindor, Correa, or Semien have fly ball exit velocities in the top-40. Semien is at least compensating for that by hitting a lot of fly balls, but the other three are struggling both to elevate and hit the ball hard. With the lowest fly ball exit velocity of this group and fewest total fly balls, I’d probably be most concerned about Correa. Still, pursuant to xBrl%, they’ve all been unlucky to a degree.

For his part, Todd Frazier has a low barrel rate given how hard he’s hitting his fly balls. His 96.1 mph EV on FB ranked 10th in MLB, sandwiched between Aaron Judge and Fernando Tatís Jr. That said, he’s not actually hitting very many fly balls, which will limit his power ceiling. Matt Olson is hitting the ball even harder, with the third-best EV on FBs in the league. His 12.5% barrel rate actually isn’t bad, but it should be better. I think that’s even more true if we consider how often he lofts the ball, which isn’t part of the xBrl/BBE% but is relevant to Brls/BBE%. In other words, he might be unlucky in terms of both Brls/BBE% and xBrls/BBE% despite eight homers to his name.

Like Frazier, Yoshi Tsutsugo and José Abreu hit the ball pretty hard, but could do with a few more fly balls. Tsutsugo was known for his power in Japan, so this isn’t particularly surprising. He’s definitely worth watching should he emerge from his platoon. And it’s a small sample thus far, but Abreu’s fly-ball rate is down 2.6 percentage points this year and his ground-ball rate is up nearly ten percentage points. He simply won’t be the same power hitter if he doesn’t revert to his former batted ball profile, which may explain the discrepancy between his Brl/BBE% and xBrl/BBE%.

Negative Regression Candidates

It is obviously impossible to have a negative barrel rate. But even if you gave Pedro Severino a 0.00 xBrls/BBE%, he’d still be vastly overperforming. He simply doesn’t hit fly balls hard at all. His 82.1 mph exit velocity ranked third-to-last among all qualified hitters. I wouldn’t be rushing to the waiver wire to add him.

There are a few players on this list with barrel rates that are simply too high no mater the fly ball exit velocity. Brandon Lowe, Bryce Harper, Dominic Smith, Corey Seager, and Mitch Moreland will struggle to maintain their 17.0+ Brls/BBE rates as the season progresses. Indeed, the highest xBrl/BBE% in the league belongs to Teoscar Hernández at 16.27%, and that’s with his league-leading 99.0 mph exit velocity on fly balls. Moreland is the only one of these five even close to that mark.

Harper actually hits a lot of line drives (33.3% line-drive rate) and hits them extremely hard (99.9 mph), thereby compensating for his low fly ball total and exit velocity. Two of his four home runs have come thus far on line drives. Seager, too, hits a lot of line drives (36.7% line-drive rate), as four of his five homers have been of the line drive variety. At least in terms of home-run hitting, it’s unclear whether a line-drive-centric approach is sustainable for either of these players. Still, Seager’s actually swatting both his line drives (102.1 mph) and fly balls (93.5 mph) harder than Harper, so I have lower expectations for Harper going forward.

Like Seager, Soler’s fly ball exit velocity of 93.2 mph actually isn’t all that bad — 31st-overall. But regression is coming for Lowe and his insane 20.7 Brls/BBE%. That’s simply unsustainable with a 91.2 mph average exit velocity across only 13 fly balls.

Smith is similarly overperforming to the tune of a 17.1 Brls/BBE%, but with an even worse fly ball exit velocity than Lowe. In fact, Smith’s exit velocity on fly balls is actually down compared to his 93.2 mph 2019 mark. With better exit velocity last season, he oddly posted a worse (but far more reasonable) 7.5 Brls/BBE%. I would be selling or, if he’s still on the wire, just don’t spend all of your FAAB expecting a slugger.

Finally, all of Christin Stewart, Mark Canha, and Kole Calhoun rank in the bottom-50 for fly ball exit velocity. My expectations are probably highest for Canha to continue outperforming his xBrl/BBE% (though to a lesser degree than he is currently), because he’s at least hitting a lot of fly balls. Stewart’s actually been pretty bad this season, with just one home run and a .190 AVG, so you’re probably not owning him anyway. But don’t expect more just because of his high barrel rate. Also, don’t be surprised if Calhoun reverts back into a pumpkin by seasons end given that he isn’t either hitting the ball with authority or putting it in the air.