Analyzing the League’s Launch Angle Profile

Home runs are up across the league and everyone is searching for reasons. One assumption that would make sense is that with the feedback system of Statcast, the league gets closer together in launch angle since we know that the best hitters have an average LA of around 12-15 degrees and players that are way off that would be incentivized to correct that or else be replaced by other hitters who can do it.

First, let me say that I used at-bat cutoffs since that is what Statcast allows for. I used 250 for 2015 and 2016 and 100 for 2017 to date. That probably changes the values a little. Specifically my average LAs look higher than the usually-cited league averages, probably because bench players who hit weakly are excluded.

Looking at the average LA, the chosen group of hitters went up from 11.7% in 2015 to 15.9% in 2016 (+4.2) to 16.6% (plus another 0.7).

So what did definitely happen was an increase of the launch angles across the board. However, when looking at the standard deviations, the league did not get closer together. SD was 3.6 in 2015, 3.8 in 2016 and 4.2 in 2017. It seems like not everyone is adjusting at the same pace.

So let’s look at different subsets here.

The average of the top 20 went way up from 17.8 in 2015 to 25.1 in 2017 (+7.3). The average of the bottom 20 also went up, from 4.2 to 8.0 (+3.8) degrees. The Q25 went up from 9.4 to 13.9 (+4.5) and the Q75 from 14.3 to 19.2 (+4.9) degrees.

So LA definitely went up across the board in all groups, but if anything it accelerated more at the top than on average or at the bottom. The league is increasing LA but so far it is not getting closer together.

Print This Post

newest oldest most voted
D.K. Willardson

Thanks for more research on a great topic! I’m pulling data from BB Savant and using your same 250 PA cutoff, I’m getting 10.82 for 2015 and 11.74 for 2016 (For 2017, there is a problem with BB Savant and grouping by Team or Venue in downloading CSV files so I don’t have that data. Help from anyone on this would be greatly appreciated!). I’m not saying that my data is correct but just wanted to point out the difference.