Exit velocity stats have been around since 2015. Hard-hit rate and average exit velocity were the original “hip” metrics, but those proved to lack predictive power. Max exit velocity became more popular afterward, and it’s still a good indication of a hitter’s raw power. More recently, baseball minds around the internet have been using percentiles to describe a player’s raw power. Depending on the article you read, 90th percentile exit velocity is the most predictive EV-centric statistic. That said, there’s some debate as to the best percentile to use for such analysis. Tom Tango and those over at MLBAM use the 50th percentile. Some prefer the 80th percentile. That uncertainty keeps me skeptical, as I prefer metrics that aren’t so…malleable. Players only have one maximum or average exit velocity, yet they have an unlimited number of “X percentile exit velocity.” I don’t necessarily consider it far and away the “best” exit velocity metric, but it’s definitely in vogue at the moment, so let’s roll with it.

My inspiration for this article was an interview Eno Sarris did with **Zack Gelof.** Gelof had a good rookie season despite some not-so-great max and 90th-percentile exit velocity numbers. Here’s some of what he said about his production: “Maybe down the road, next year, I might unlock something in my swing a little bit and plug into a 110,” the 23-year-old thought. “But with me, I’m just okay with who I am, and I consistently get to my barrel enough. Last night, I got to a ball, 106 off the bat to right center; that’s a home run anywhere. I’m just trying to get to that consistency. If I live at 100, 102, 106 at a good launch angle, that plays to all fields. If I go for 109, it feels like I’m going out of control.” I have never said this in a baseball context before, but I love this for him. Gelof recognizes his strengths, and his comfort at the plate supersedes any exit velocity numbers he could potentially produce.

This had me wondering, though. Gelof talks about hitting a lot of balls around that 100-105 MPH range, which hurts those exit velocity numbers but also produces good production on contact. There’s lots of research on launch angle “tightness” and the impact it has on BABIP, but I haven’t seen as much research on exit velocity tightness. Tightness, in this case, is mathematically represented using standard deviations. So, is there a relationship between exit velocity “tightness” and overperforming production on contact you would expect given a player’s 90th percentile EV? To find an answer, I had to get a z-score for each metric. Think of a z-score as a way to put different numerical forms on the same scale. I subtracted the z-score of each qualified hitter’s 90th percentile exit velocity (an integer) and their wOBA on contact (decimal) to find whose production on contact most differed from their raw power. Here’s a leaderboard, just to wet your tastebuds a bit.

Before I go any further, I should point out that sd(EV) is an inherently flawed metric in a way that sd(LA) isn’t. Naturally, weaker hitters are going to have tighter exit velocities than powerful sluggers, because it’s much easier to hit a lot of balls at 80 MPH than 110 MPH. I tried to account for some of that by removing a few outliers (Stanton, Judge, Ohtani, Kwan, Kemp), but even just looking at our leaderboard from above, it’s clear that sd(EV) has a positive correlation to exit velocity and power in general. Pete Alonso and **Luis Arraez** show up twice here. Arraez, the poster boy for outperforming exit velocity numbers, is a master at hitting his batted balls right around that 80-90 MPH range. The sd(EV) gap between Arraez and the 2nd-ranked Yasmani Grandal is the same as the gap between Grandal and the 21st-ranked hitter. Many have lauded Arraez’s “barrel control” as some of the best in baseball. Think about when the shift was still around. Uneducated fans would look at a Joey Gallo plate appearance ending in a pulled groundout into the shift and ask, “Why didn’t he just hit it to the left side, it’s wide open!” The reason being, of course, it’s incredibly difficult to do that for 99% of MLB hitters not named Luis Arraez.

Now that we have an understanding of the numbers, here’s what I found. Exit velocity tightness is on the Y-axis, and over/underperforming 90th percentile exit velocity is on the X-axis.

Although not particularly strong, there is a relationship between our two variables. Arraez is the blue dot all the way on the top right of the graph. Instead of providing another graph to see if launch angle tightness does any better than exit velocity tightness, I’ll just tell you it doesn’t. Those two variables have an R-squared of .08. It makes sense to me – batted balls that are all super tight in their launch angle distribution don’t necessarily imply production on contact. Tim Anderson ran a 61% groundball rate and the 2nd tightest launch angle spread. It doesn’t matter how tight your batted balls are if they’re all on the ground. Exit velocity tightness is more intuitively linked to production, I think. For a player’s sd(LA) to mean production on contact, they have to have a decent average launch angle. That circular argument isn’t *as much* an issue when discussing exit velocity tightness.

I wanted to look to see if I could find a (even slightly) larger R-squared value, so I performed a regression analysis with sd(LA) and sd(EV) as my predictor variables and the 90th EV over/underperformance as my response variable. I added weights to each number, combined the two predictor variables, and came up with this.

A modest improvement, no doubt. If I wanted to further explain the gap and come up with a larger R-squared value, I would add in each hitter’s pull rate, groundball rate, and a ballpark adjustment. That isn’t the point, though. Finally, I wanted to see if sd(EV) had any correlation to whiff rate. My idea was that players who are good at keeping their batted balls around a certain exit velocity have a good deal of bat control and, therefore, have lower whiff rates. The R-squared between those two was only .13. Darn.

I’m not entirely moved by my findings here. As I said, sd(EV) has flaws in and of itself that make it difficult to use in this sort of analysis. Still, for a metric that is *very* sticky (R-squared of 0.58 year-over-year) to also explain why some guys aren’t as good as their 90th percentile exit velocity would indicate, it’s not nothing. I think, more than anything, this is an article about X-percentile exit velocity metrics. Consider this: The top 16 players in 90th percentile exit velocity ** all** posted a negative z-score difference (or, underperformed) between their 90th %ile EV and their wOBAcon. In other words, there isn’t a 1-1 relationship between these two metrics, and that becomes extremely apparent on the top end of the exit velocity scale. Notice how the slope of that line flattens out once you reach an exit velocity z-score around 0.5.

I started this article with a focus on exit velocity tightness, but I think the real finding is that the trendy metric that is X percentile exit velocity is not all it’s chalked up to be. I like using it more than 90% (pun very much intended) of power metrics, but it has its issues. The metric is simply a tool – the same way max exit velocity is – that does a decent job of measuring raw power. I say decent because, as Gelof pointed out, he isn’t going all out with every swing. I’m not making any wildly new discoveries here, but it’s always fun to write up your thoughts on something, come away with not much at all, and then (maybe?) find something about a topic you weren’t even thinking about.

Working online brings in $285 dollars an hour for me. My best buddy shows me how to do this and makes $29,000 a month doing it, but I never realized it was real. Visit the following link to have.

look at it………………………………………………… https://5n.gs/SGRQYnA 0.15 with wOBAcon is huge. There is so much noise that a 0.15 value is actually impressive.