Average Exit Velocity: May not mean what you think it means

I'm guessing this is out there already, with studies a lot more intense than what I'm about to present. That said, I ran some quick correlation studies on component average exit velocity (AEV) and component BABIP and thought I'd share the findings. Since it's too long for a tweet, I figured I'd dust off my old site column and post it here.

The impetus was writing a profile for Jeff McNeil as his Statcast levers are not favorable - at least they appeared wonky for such a consistent hitter.

The idea is learning how AEV on grounders (GB), fly balls (FB) and outfield line drives (OLD) affects their respective BABIP. I wanted a full season's worth of data, so I looked at 2018 and 2019. These make a good pair since the ball in 2019 incurred much less air resistance than 2018 and while it doesn't influence AEV, it could affect the outcomes.

I landed on 50 batted ball events (BBE) after running the data for several BBE. The number doesn't matter for this quick study since the goal is unearthing general trends to aid in profile writing as opposed to incorporation into my projection engine.

The study is straightforward. I (well, Excel) calculated the correlation coefficient between AEV and BABIP for all players with at least 50 BBE for each type of event. By means of reminder, perfect correlation is 1, perfect reverse correlation is -1 (this will be relevant) while a completely random relationship is 0.

Here are the findings:

Intuitively, I expected the correlation to be stronger. Sure, it's positive, but there is a lot of wiggle room. Looking at the GB AEV and OLD AEV and assuming because it's low, a high BABIP will regress isn't as salient an argument as I perceived. Similarly, a high GB AEV and OLD AEV and a low BABIP may not manifest in a higher BABIP.

The fly ball data may surprise some, but based on prior research, it's not alarming. The small negative correlation is saying it's not always a bad thing to hit soft fly balls. These fall into the purgatory between the outfielders and the infielders. Better struck fly balls are the proverbial cans of corn.

One more correlation was conducted, examining the relationship between FB and OLD AEV and HR%. Don't worry, this one yielded the expected results:

Sure enough, elevating the ball with velocity correlates very well to homers. The OLD data incorporates batted balls with trajectories too low be classified as FB, but with too small a launch angle to clear the fence,

Again, there isn't anything earth-shattering here, but it does demonstrate AEV is much more relevant for power than average.