What we can learn from one batted ball

Luis Robert turned on the first pitch he ever saw in the major leagues, pulling it into left field for a base hit. The swing was smooth, the power enormous. Just one pitch into his big league career, Robert had already posted an exit velocity of 115.8 mph.

To contextualize just how absurd Robert’s first batted ball is, consider this: only 21 players hit a ball at least 115.8 mph at any point during the 2019 season. That puts him in the top two percent of all players.

Really, 115.8 mph might not be Robert’s maximum exit velocity. In reality, it probably isn’t. If Robert posted an exit velocity of 115.8 mph on his first batted ball, there’s almost certainly the potential for more untapped power at a later date. At the end of the season, there’s a pretty good bet that his max exit velocity will be even higher.

Let’s say it isn’t, though. Let’s say that all we had was that one batted ball with which we can evaluate Luis Robert. Since maximum exit velocity is a good way to measure the raw power of players, how much can we learn about the success he might have in the major leagues? This is particularly important given the shortened season; we have to make judgements on players more quickly, and before things can stabilize.

We already know from that single batted ball that Robert’s power is elite. We also know from that batted ball that Robert can activate his elite power in a game. But what does that mean from a statistical standpoint?

Logically, the statistic that we can first begin to build a case around is isolated power. You would think that, on average, players with higher exit velocities tend to hit for more power. These are the Giancarlo Stantons and Pete Alonsos of the world, baseball’s big boppers, and two of the three 2019 leaders in max exit velocity.

Indeed, from one single batted ball, there is already a moderate correlation (r = 0.55) between maximum exit velocity and isolated power. On average, in each of the last three seasons, each one mph increase in exit velocity corresponds to a nine-point increase in isolated power.

Of course, there’s still quite a bit of variation here, but we can already begin to build a more legitimate case for Robert’s profile. Using the regression line, Robert would be expected to post a .230 ISO this season. While that still might not seem eye-popping for a player with 98th percentile raw power, it is a 41-point increase from his pre-season ZiPS projection (.189) and a 16-point increase from his pre-season Steamer projection (.214). From one batted ball, we can already feel decently confident that Robert’s power will already play up more this season than the leading projection systems thought.

This logic can be applied to other players, too. Teoscar Hernández, who had never hit a batted ball above 112.2 mph, took Erick Fedde for a 114.0-mph ride on July 30. Jaylin Davis, Adalberto Mondesi, Kyle Lewis, Oscar Mercado and J.P. Crawford, too, are among the players to have already experienced year-over-year gains in max exit velocity. (However, since these are individual results and not averages, they may be more susceptible to error resulting from MLB’s switch from Trackman to Hawk-Eye to track batted ball data.)

One batted ball might tell us something, but it doesn’t tell the whole story. As mentioned, there’s still quite a bit of variation in the data. Max exit velocity explains roughly 30% of the variation in isolated power — a lot, but not nearly enough to draw firm conclusions. While one can feel confident that Robert’s power numbers will come in ahead of his initial projections, we’re still not close to a firm number regardless.

Hernández is an excellent example of this. Operating under the assumption that it remains his maximum, his 114-mph homer would imply a .214 ISO this season. However, despite not posting an exit velocity above 112.2 mph prior to 2020, Hernández had a career .240 ISO. ZiPS (.226) projected a figure a touch above .214, while Steamer (.208) came in slightly below. And these expectations were made prior to his 2020 scorcher — his 112.2 mph max exit velocity prior to this year implies a .197 ISO.

What gives? For one, there’s proof that evaluating hitters after just one batted ball is nearly impossible. We can make guesses — such as the fact that it is highly likely that Robert hits for more power than we initially thought — but conclusions are hard to come by. Vladimir Guerrero Jr., for example, had the second-hardest hit ball of 2019 at 118.9 mph, implying a .258 ISO. In reality, however, he posted just a .162 ISO. That 96-point residual was the eighth-highest overestimation by the regression over the last three years — this was from a dataset that includes 838 figures.

Guerrero, of course, is the perfect example to describe the shortfalls of the model. Yes, he hit the ball extraordinarily hard last year, but extra-base hits are not purely the result of hard-hit batted balls. Doubles, triples and home runs tend to come on line drives and fly balls. So when Guerrero put nearly half of all batted balls on the ground, he was unable to turn his raw power into results.

Incorporating groundball rate into the regression certainly strengthens the correlation (up to an r-value of .72), but at what cost? We’re back to guessing on players like Robert. We don’t know what his major league groundball rate will be, but that’s an important datapoint to know in order to continue to project power. (For fun, including Robert’s 2019 minor league groundball rate, we would expect a .282 ISO, a 93rd percentile mark.)

Unsurprisingly, we don’t learn a lot from one batted ball. We can, however, adjust forecasts. And despite the fact that expectations were sky-high for Luis Robert, it’s possible that he’s already well on his way to surpassing them.

Devan Fink is a sophomore at Dartmouth College and a Contributor at Beyond The Box Score. Previous work of his can be found at FanGraphs and his own personal blog, Cover Those Bases. You can follow him on Twitter @DevanFink.