The Intersection of Batted Ball Angles and Direction

Over the last handful of years, metrics like average launch angle and pull% have made their way into increasingly broad and casual baseball discussions. And for good reason; those metrics are quite interesting and descriptive, at the very least, while additionally being important pieces of a larger puzzle for understanding and quantifying player production.

Despite the numerous discussions of batted ball direction (pull%, center%, oppo%) and angle (avg LA, GB%, LD%, etc.) as well as the availability of those metrics (on FanGraphs’ insightful leaderboards page, for instance), it seems that the intersection of those metrics isn’t very often described. A brief glance into a player page might inform those interested that any particular hitter has a GB% of X and pulls the ball Y percentage of the time but, more granularly, it isn’t very clear how often those ground balls stuck are in fact pulled versus hit the opposite way, for instance.

This post aims to address, as well as visualize, the intersection of those angle and direction metrics versus viewing the two in isolation. To make things simple, “angle” has simply been broken down into three hit classes: ground ball, line drive, and fly balls (popups have been lumped into fly balls, for better or worse). Meanwhile, direction has been bucketed into the familiar categories: Pull%, Center%, and Oppo%.

Between 3 angle classifications and 3 direction classifications, there are 9 possible permutations of batted balls recognized here as part of a feature very simply referred to as “spray and angle”.

As for data, all batted balls have been pulled for the top 50 plate appearance leaders in 2021 from Baseball Savant. The field has been broken into thirds to recognize batted balls toward right, left or center and Baseball Savant’s predetermined hit classes have been relied on for hit type (GB, LD, FB). Again, popups got lumped in with fly balls.

To visualize the breakdown of “spray and angle” classes, below is a scatterplot of all those datapoints, color coded by classification.

The visualization above is helpful for illustrating each classification, but does a poor job describing the frequencies of each type of batted ball. To better illustrate that breakdown, the following three barplots capture the following: balls in play for (1) all top 50 hitters in 2021 plate appearances, for (2) those hitting as RHHs, and (3) finally those hitting as LHH.

Both RHH and LHH are included here and there is a fairly even split between the directions of batted balls. Still, a couple heuristics seem to apply: the most common type of batted ball is a ground ball to the left side of the infield and line drives are less common than the other hit types regardless of direction/spray. However, by splitting data by batter handedness there are more takeaways.

Here some trends are more pronounced.

RHHs are far more likely to hit ground balls to the left-hand side of the infield than the right. In fact, ground balls hit the opposite way are the least likely batted ball type among these 9 classifications. Though much less dramatically distributed, line drives are more commonly pulled as well. Fly balls, meanwhile, are more often hit the opposite way, though popups might be a confounding factor for that classification.

With LHHs, we see a virtually mirrored image of those ratios above for RHHs. Ground balls and line drives, again, are more likely to be pulled whereas fly balls are more often hit the opposite way. For all the documented value to pulling the ball in the air, I would not have expected balls in the air more often being hit the opposite way, or even up the middle. But hitting, it should be noted ad nauseam, is very hard.

So, cumulatively speaking, batters seem to put the ball in play similarly regardless of handedness. Luckily, digging deeper enables a view into the specific tendencies that make particular players unique and exciting. Below are two more barplots for two very different individuals.

Matt Chapman, whose profile in recent years has become increasingly oriented toward three true outcomes, makes for a pretty interesting case illustrated above. Chapman hit a lot more fly balls in 2021 relative to league average; he hit more fly balls than ground balls in fact (disregarding line drives), which appears fairly rare to this degree. Despite being right handed, Chapman also appears to be a solid candidate for a shift (should it not be banned going forward as part of the next CBA), considering he hit <10 balls on the ground in the direction of right field in all of 2021.

While the league as a whole had a pretty even ratio of line drives to fly balls, Chapman’s fly ball totals in 2021 dwarfed that of his line drive figures. This might be a contributing factor to Chapman’s .272 BABIP and .210 BA. Incredibly, Chapman’s average launch angle actually decreased considerably season over season from 2020 (24.1) to 2021 (18.6).

For perhaps a more balanced profile, look no further than Juan Soto.

Having looked at a handful of these barplots for various players, Juan Soto’s stands out. Soto hits a significant amount of ground balls, but puts them more up the middle than a typical LHH. His line drives, meanwhile, are hit predominantly up the middle, with equal likelihood of otherwise being pulled or struck the opposite way. Given his power to all fields, going the opposite way on fly balls doesn’t take much from Soto’s game.

This post wasn’t about establishing new information, but rather repackaging information largely available already. Still, it’s an interesting way to view the relationship of two metrics often interpreted in isolation – and sometimes looking at how two distinct metrics relate can shed meaningful light on those same numbers the next time they are interpreted on their own.