Estimating Launch Angles Without Trackman

Fly Ball = 36.5; Line Drive = 16.5; Pop Up = 63; Ground Ball = -12

This post is going to be really really short. In fact, all you need to know is in the subtitle. Here’s the methodology:

1. Determine if there was a bunt attempt. Then ignore all bunts

2. Determine the trajectory of the batted ball. Either through the “Hit Data Trajectory” field, or by parsing the description.

3. Once you have the trajectory, apply the following formula:

IF [Trajectory] = "FlyBall" THEN 36.5
ELSEIF [Trajectory] = "LineDrive" THEN 16.5
ELSEIF [Trajectory] = "PopUp" THEN 63
ELSEIF [Trajectory] = "GroundBall" THEN -12
ELSE NULL END

That’s the entire process. Let’s see how it does:

Low A

Those are R² of .89 and .94 for 2022 and 2023 respectively, minimum 100 batted ball events. That’s pretty good. What if we lower that to just 25 events?

.905 and .912, still very strong. Essentially, with a very small sample size, we can estimate a batters launch angle with very high confidence, just with stringer data.

AAA

Very similar to Class A.

MLB

Conclusion

For such a simple method, it provides an extremely robust estimate. My next project will be to estimate exit velos with stringer data, should be fun!