Launch Angle and Batted-Ball Volatility
How does a hitter’s average launch angle correlate with volatility in his batted-ball profile?
In my last baseball-related article, I showed you the above chart, which is a scatterplot of every batted ball from the 2015–17 seasons based upon exit velocity and launch angle. Each batted ball in the above chart is shaded one of 50 colors, ranging from blue to yellow to red, based upon expected wOBA (xwOBA). Along with the above chart, I illustrated the 2015–17 batted-ball profiles of seven 1B/DH/Corner OF on the free-agent market: Jay Bruce, Melky Cabrera, Lucas Duda, Carlos Gonzalez, Eric Hosmer, J.D. Martinez, and Logan Morrison.
In analyzing the seven batted-ball profiles, I noted that four profiles (Bruce, Gonzalez, Hosmer, and Morrison) were similar to each other. The only slight difference between the four profiles that I noticed was that Bruce had a higher percentage of batted balls in the yellow area surrounding the red area (“Yellow >95&15”), as compared to the other three profiles, but had a relatively similar percentage of total red and yellow batted balls. Here’s the summary table of the seven profiles for reference:
Consequently, I wondered how the difference in Bruce’s batted-ball profile would affect his future performance, as compared to the three other players with similar profiles. In a more general context, is all xwOBA on contact created equal as it pertains to predicting future performance? Or do batted-ball exit velocity and/or launch angle have their own predictive power, holding constant xwOBA on contact?
My hypothesis was that, holding constant xwOBA on contact, the batted-ball profiles of hitters with higher average launch angles would be more volatile in future seasons than would those of hitters with lower average launch angles. My thinking was that, in looking at the chart above, batted balls hit at lower launch angles (<15 degrees) experience a less severe drop in xwOBA as exit velocity decreases, as illustrated by little change in color. On the other hand, batted balls hit at higher launch angles (>15 degrees) experience a more severe drop in xwOBA as exit velocity decreases, as illustrated by greater change in color. In fact, at certain launch angles above 15 degrees, the color changes from red to yellow to blue within roughly 5 mph.
In order to test my hypothesis, I collected data from the 2015–17 seasons of hitters with 250 plate appearances in consecutive seasons.
First, I calculated the absolute values of the season-to-season point-change and percentage-change in xwOBA on contact for each hitter, and I found that both a hitter’s point-change (0.12) and percentage-change (0.04) in xWOBA on contact have slightly positive correlations with the hitter’s average launch angle from the first season.
Second, I calculated the average point-change and percentage-change in xwOBA on contact for hitters with above-average (>10.66 degrees) and below-average (<10.66 degrees) average launch angles. Hitters with above-average average launch angles experience a .031 and 8.1% change in xwOBA on contact from season to season, on average. On the other hand, hitters with below-average average launch angles experience a .026 and 7.6% change in xwOBA on contact from season to season, on average.
Third, I calculated the percentage of batted balls hit at launch angles of greater than 15 degrees for each hitter, and I found that both point-change (0.11) and percentage-change (0.03) in xWOBA on contact have slightly positive correlations with the percentage of batted balls hit at launch angles of greater than 15 degrees from the first season.
Fourth, I calculated the average point-change and percentage-change in xwOBA on contact for hitters who hit batted balls with launch angles of greater than 15 degrees at above-average (>43.4%) and below-average (<43.4%) rates. Hitters with above-average rates experience a .031 and 8.0% change in xwOBA on contact from season to season, on average. On the other hand, hitters with below-average rates experience a .027 and 7.7% change in xwOBA on contact from season to season, on average.
What all of this tells me is that, while the effect may be minimal, hitters who tend to produce batted balls with higher launch angles experience more season-to-season volatility in their batted-ball profiles than do lower launch-angle hitters.
Here are each free agent’s average launch angle and rate at which batted balls were hit at greater than 15 degrees from the 2017 season:
How should these results affect the value of free-agent hitters, if at all? Can Scott Boras (Hosmer’s agent) tell teams that Hosmer should be valued more highly because his batted-ball profile should result in greater consistency? Or can Matt Sosnick (Bruce’s agent) tell teams that Bruce should be valued more highly because his batted-ball profile has greater growth potential? Or does none of this matter at all?
Furthermore, how should these results matter in terms of general player evaluation? Is a high xwOBA on contact from a low launch-angle hitter more believable than from a high launch-angle hitter? Conversely, is a low xwOBA on contact from a high launch-angle hitter less cause for concern than from a low launch-angle hitter? Or maybe the answer is none of the above?
I think the overarching answer is: it depends. It depends on the team, the player, the agent, and possibly many other unforeseen factors. There are so many factors that go into evaluating a player that I would be hard-pressed if changing the perception of one part of one factor makes a world of difference. That being said, in an environment in which organizations will take any little advantage that they can, batted-ball volatility may be another piece to the puzzle.