The Near-Immediate Usefulness of Max EV

Maximum exit velocity (max EV) measures a player’s hardest-hit ball, typically measured within a single season and compared against other players. Our Mike Podhorzer has documented its leaders and laggards. Rob Arthur, one of baseball’s best public analysts and whom I admire greatly, wrote intelligently on the importance of max EV as a projection-buster back in 2018. Max Freeze (real name) blends extremely hard hits (114+ mph) with launch angle to look for possible power breakouts ahead of 2020.

It has been established (by Al Melchior and me, in fact) that max EV, while an effective indicator, is not the or even a superior indicator of hitter power.

That’s not to say max EV is useless, by any means. It is altogether a different breed of metric than, say, barrel rate (Barrel%, either per plate appearance [PA] or per batted ball event [BBE]) or average exit velocity (EV), both to which fantasy baseball analysts refer much more often. The latter two, and many others, are rate metrics that need large sample sizes to become reliable — or, in common parlance, to “stabilize.” (More on that here, from our former and beloved Eno Sarris.)

Meanwhile, max EV is not a rate or average but a singular data point. It can happen at any moment in time — including the very first batted ball of a hitter’s season. This makes it an intriguing addition to the ol’ tool belt insofar as it could become “reliable” (not necessarily in the statistical sense) much sooner than would barrels or EV. Potentially, we could use max EV loosely as a leading indicator of where a hitter’s barrel rate, average EV, or even weighted on-base average on contact (wOBAcon) might eventually settle.

So: what are the merits of max EV?

As noted in the posts linked in the first paragraph, max EV is a second-tier Statcast metric in terms of its correlation with traditional measurements of power such as isolated power (ISO) or home runs per fly ball (HR/FB) as well as with advanced measurements of contact quality such as wOBAcon.

Despite its relative inferiority, you can still see strongly patterned trends between max EV and wOBAcon as well as other measurements of contact quality, including hard-hit rate (Hard% or HH%), which calculates the percentage of batted ball events that exceed 95 mph. The graph below shows how wOBAcon (red dotted line) and hard-hit rate (the blue line labeled “EV95+”) increase steadily with max EV.

Click to enlarge.

With each additional mile-per-hour of max EV, league-average wOBAcon increases by roughly 11 points and league-average Hard% increases by roughly 1.8 percentage points. Keep in mind, these are league-average values. Player performance naturally varies based on myriad other factors. Max EV is but one piece of a much larger puzzle that solves for a hitter’s contact quality. But contact quality still depends on many other things — most critically, launch angle. (Looking at you, Vladimir Guerrero Jr. and Yandy Díaz, patron saints of beefy boy grounders.) But that’s another issue for another day.

I also graphed lines for every whole-number exit velocity threshold from 95 mph to 105 mph, all of which, interestingly enough (to me, at least) appear to run parallel with one another. This suggests that all exit velocity thresholds, like Hard%, increase by 1.5 to 2 percentage points with each additional mile-per-hour of max EV. (For example, if the average hitter improves his max EV from 108 mph to 110 mph, we can expect his hard-hit rate to increase from 35.1% to 38.1% — and his percentage of balls hit 105+ mph to increase from 4.5% to 7.9%.)

This takeaway is especially important because exit velocity is especially important! Increased exit velocity means improved contact quality means increased probability of superior production on balls in play. The graph below shows how the league-average wOBAcon skyrockets on batted balls in excess of 95 mph. (The X-axis is exit velocity, and the Y-axis is league-average wOBAcon.)

Click to enlarge.

In addition to a scenario in which a hitter simply improves his max EV from one year to another, let’s say a hitter establishes his max EV early in a season, but his other metrics — wOBAcon, Hard%, Barrel%, average EV, etc. — lag behind and/or have not yet become reliable because the sample is too small.

For example, let’s say I, very proficient baseball hitter, hit a ball 110 mph on my very first batted ball of the season. On average, a hitter who maxes out at 110 mph achieves a hard-hit rate of 38% and a .397 wOBAcon. However, despite this spectacular singular event, I could only muster a 31% hard-hit rate and .349 wOBAcon (the league-average values for a max EV of 106 mph) through the end of April. We know one month is hardly enough time to evaluate a player’s performance. Has my max EV already alerted us to forthcoming improvement?

The big question upon which all of this assessment hinges (of course I’ve managed to bury the lede again) is: when does a hitter typically achieve his max EV?

Trying to pin down the exact moment a hitter achieves his max EV for the season seems dubious and, above all, somewhat arbitrary. However, if we relax expectations a bit, we can look for batted balls that very closely mimic a hitter’s max EV without identifying the exact moment he achieves his max EV. Most importantly, the threshold remains not only unique but also dynamic for every hitter.

With this in mind, I reframed the question: when does a hitter typically achieve 98% of his max EV? (To be clear: any batted ball that’s 98% of max EV or better, which therefore includes max EV itself).

I limited my observations to all 694 hitters who compiled at least 250 BBE during the 2017 through 2019 seasons. (No hitters compiled 250+ BBE in 2020, hence its omission.) In 2019, the average hitter amassed 4.26 plate appearances and 2.84 batted balls per game; I will use this as shorthand for ease of mental math below. The results of my research:

• 90% eclipsed their 98% threshold by their 150th BBE (game 53).
• 80% … by their 100th BBE (game 35).
• (Per Sarris, average EV stabilizes around 80 BBEs: “what’s happening after 80+ balls in play for a player explains about half of what will happen on their future balls in play” — with emphasis on half).
• 70% … by their 65th BBE (game 23).
• 60% … by their 50th BBE (game 18).
• 50% … by their 35th BBE (game 12).
• 20% … by their 10th BBE (game 4).

Again, to emphasize a point: this doesn’t mean XX% of hitters achieved their actual max EV by game Y of the season — just that those hitters came within swinging distance (no pun intended) of their max EV, close enough to give us an idea of what they’re capable of. (Max EV can only increase, so, if anything, the 98%-threshold exercise might lead us to underestimate a hitter’s potential. That’s not necessarily bad!)

For reference, Marcus Semien (and several other A’s hitters) amassed his 35th BBE in the team’s 12th game — and in just his 50th plate appearance — of the 2019 season, consistent with my back-of-the-envelope arithmetic. Incidentally, he eclipsed the fabled 98% threshold on his 52nd ball in play — just four games later. Sixteen games into the Athletics’ season, he had a .378 wOBAcon but only a .323 expected wOBAcon (xwOBAcon). In other words, Semien achieved solid outcomes but without commensurate contact quality to support it.

If you were living and dying by xStats, you might have wagered Semien had been lucky, that he was due to backslide in the coming weeks. You’d have been wrong: Semien finished the season with not only a .391 wOBAcon — better than his first 16 games, albeit not substantially — but also a .384 xwOBAcon that fully validated his actual performance. In this instance, it appears that max EV helpfully anticipated better contact quality to come from Semien.

It’s entirely anecdotal, of course, but it (Semien’s example) makes me think this (gesturing wildly at max EV) could be helpful!

There are two large caveats in all of this:

1. As previously noted, not all hitters will perform according to the graph above because not all hitters are exactly league-average.
2. It’s impossible to know who will eclipse his 98% threshold when.

Regarding the latter, Aaron Altherr 알테어 eclipsed his 98% threshold on his very first batted ball of the 2017 season; now he’s playing in the KBO. José Abreu was one of 2019’s latest bloomers, failing to eclipse his 98% threshold until the season’s final month.

Abreu’s narrative deserves its own case study. I interpret it optimistically: Abreu was not a late bloomer as much as he was a legitimate breakout whose base skills improved, sustainably, as 2019 concluded. In fact, his improvement was so substantial that it arguably laid the foundation for his MVP-caliber 2020 performance:

the blue line depicts josé abreu's previous best 60-game rolling avg EV before 2020. he touched that peak a few times—sometimes within a season, sometimes at the end of one season into another

abreu began to consistently obliterate that peak starting in late-2019 and into 2020 pic.twitter.com/H42FCZxrnn

— Alex "Oxlade" Chamberlain (@DolphHauldhagen) October 29, 2020

A 60-game rolling average of Abreu’s exit velocity reached new levels in 2020, building upon improved performance in late-2019. It’s possible that looking at Abreu’s late-season, new-career-best max EV could have alerted us to his 2020 “breakout” (as much as a 3.5-WAR player can further break out) in which he established additional career-bests in barrels (14.3%), hard hits (53.3%), and average EV (92.9 mph), among other superlatives.

Anyway, certainly there’s more to be researched than what I’ve compiled here. In terms of major takeaways, I think max EV’s utility comes not from its absolute relationship with power and contact quality (which is weaker than other metrics) but, instead, from how quickly we can use it in-season in place of other superior metrics that are not yet reliable.

In short, perhaps max EV should be the first thing any of us take seriously in a baseball season’s early stages — historically, nearly 80% of healthy, full-time hitters have shown us their worth by the end of April. And I would look for hitters like Abreu, who, in September 2019, suddenly set a new personal-best max EV by more than 2 mph. These things don’t mean nothing, folks!