The Z Files: Applying Component Average Exit Velocity

Over the last couple of weeks, the components of average exit velocity have been examined, first on groundballs and last week flyballs. Today, several players will be evaluated, looking at their component average exit velocities in concert. The objective is to decipher whether the player can maintain the associated power or BABIP (batting average on balls in play) and if not, where regression takes it.

Before delving into some players, here's a review of some of the relevant principles. Average exit velocity unto itself doesn't tell the whole story. In general, the harder the contact, the better the chance of a hit, regardless of the batted ball type, but there is so much more involved. Each batted ball type has its own average exit velocity and corresponding average BABIP:

Scientifically, this order makes sense. A pitched ball has a downward trajectory while a swing has an uppercut, some more exaggerated than others. Physics principles instruct us the closer the swing path and ball movement align, the more energy is transferred upon contact. So not only is contact on line drives more centered, the ball and swing are often on similar paths. Groundballs and flyballs occur from non-centered contact, but the exit velocity of a fly is greater than a grounder since the upward trajectory of the swing transfers more energy even if the contact isn't centered.

The associated BABIP per batted ball make sense as well, though flyball BABIP is misleading as the BABIP formula excludes home runs. Some consider this a flaw. Depending on how flyball BABIP is applied, it can be. A low flyball BABIP may not represent back luck; it could simply mean the hard contact doesn't get captured as it does with line drives and ground balls because it generated a home run. As such, the overall BABIP of sluggers is too often deemed unlucky with the standard "expect regression and a higher batting average next season" refrain.

A lot of the time, the analysis turns out wrong because the BABIP was not luck-fueled. Sometimes it ends up correct, but for the wrong reason. If the player's home run total was lucky relative to other factors, power should drop the following season with the hard contact resulting in doubles and triples (and some long outs). The effect in such cases is almost always a higher BABIP as extra hits aids BABIP more than extra outs lower it.

Even though hit rate probability is enhanced with exit velocity, the player-by-player correlation between component average exit velocity and BABIP isn't especially strong.

The correlation with line drives is only .25, demonstrating how much luck is involved. Positioning probably influences the variance a bit, but the overriding factor is there's a lot of luck involved when a round bat meets a round ball, even for the most solid of contact.

Woah, the flyball correlation is negative, indicating the lower the velocity, the better the chance of a hit. Two reasons feed into this explanation, one of which has already been discussed with a significant portion of hard contact omitted from BABIP because the outcome was a home run. The other is weakly hit flies have a better chance of landing safely in front of outfielders than well-hit ones, which get tracked down and caught.

Luck is a big factor in groundball BABIP, as is foot speed and bunts being captured in the average exit velocity of grounders. Defensive positioning, which isn't always a shift, also contributes.

With that as a backdrop, here's a deep dive on some intriguing players.

Fernando Tatis Jr. and Keston Hiura

Tatis Jr. and Hiura are grouped since the analysis is identical. Both posted a ridiculously high BABIP in their freshman campaigns. Here's a breakdown of the batted ball date for each:

Tatis Jr.

Hiura

In both cases, the average exit velocity for line drives is high, portending a high BABIP, but the actual BABIP is through the roof! The average BABIP for batters with a line drive average exit velocity between that of Tatis and Hiura is .637, higher than league average in accordance with a higher exit velocity. Both youngsters obviously lack a track record, but it makes sense their line drive BABIP would regress. Recalculating their BABIP using a .637 line drive BABIP lowers the mark to around .380. Granted, .380 is a high number to expect, but it's a better baseline to consider regression or variance from.

Focusing on Tatis Jr., his average exit velocity on grounders is low, particularly for the .333 BABIP he recorded last season. Possessing 95th percentile speed obviously helps, as he recorded 15 infield hits, but still, repeating a .333 BABIP on grounders is unlikely. Splitting the difference between that and league average yields an overall BABIP around .370. This remains elevated but is more plausible than one north of .400. Tatis Jr.'s flyball BABIP is also high. The average BABIP for hitters with a flyball average exit velocity between 94.9 and 95.3 (+/- .2 from Tatis Jr.'s 95.1 mph) is just .103. Regressing his to the midpoint between the two renders an overall BABIP of around .340. This seems like a reasonable target for 2020.

Repeating the above for Hiura renders a final BABIP in the same range.

Yoan Moncada

Like Hiura and Tatis Jr., Moncada registered a BABIP north of .400, checking in at .406. However, the impetus was not a high BABIP on line drives:

2019

2018

2017

The first thing to note is Moncada hit the ball harder across the board, making it more likely he'll maintain his gains in the upcoming season. What stands out most is the high BABIP on grounders. He hit them harder than average, especially last season, and he does possess 72nd percentile sprint speed. Even so, Moncada's .373 BABIP on grounders last season was the highest among regulars. In 2018, it was also high. In 2017, it was elevated, but not to the same extent. However, the exit velocity was much lower, so there appears to be something about Moncada's groundballs which results in hits.

Launch angle is obviously important on flyballs, but as it turns out, it's also relevant with groundballs. The trajectory is downward so the angle is negative. The more negative, the sooner the ball hits the ground and loses velocity. It also bounces more, losing even more velocity. The league average is around -11 degrees.

Sure enough, Moncada's grounders scoot through the infield faster as they take longer to hit the grass. There may be some luck involved with his groundball BABIP, but there's another element responsible so the overall BABIP regression could be softer than some anticipate.

In 2019, Moncada's flyball exit velocity increased as did his flyball BABIP. It makes sense the BABIP went up, but that much? The flyball BABIP for players with a similar exit velocity as Moncada was .135, still above average but below Moncada's .200 level.

Regressing Moncada's groundball BABIP to a still hefty .320 and his flyball mark to .200 yields an expected BABIP of .360, which seems like a good target for his 2020 campaign.

Matt Chapman

Chapman's .270 BABIP last season feel low compared to his 92.6 average exit velocity. Here are his component data from the last two seasons.

2019

2018

Chapman's exit velocity on lofted balls rose last season while his groundball level dropped. This could be indicative of adding loft to his swing, corroborated by an uptick in flyballs, albeit at the expense of line drives. In addition, Chapman's BABIP on line drives is low compared to his triple-digit average exit velocity. Fewer line drives plus some bad luck on those he did hit help explain some of the BABIP drop from .338 in 2018 to .270 last year. Regressing the .605 BABIP on line drives to something more reflective of a 100.1 average exit velocity generates a BABIP around .280.

More line drives would improve Chapman's batting average, but they would detract a bit from his power if they replaced flies. There's a good chance the next step in his development is adding back line drives from grounders. Here, Chapman's power would remain while his batting average would increase. This is more of an upside projection, but the projected BABIP lands a bit over .300.

It's possible the difference between the two seasons is the way lofted balls were classified, as the cutoff tilted more to line drives in 2018 and flyballs last year. Regardless, the conclusion with Chapman is encouraging. His overall average exit velocity on lofted balls in particular is on an upward trend. In addition, his contact rate has improved over the last couple of seasons. Projection theory entails using previous seasons in a weighted average, which would bring Chapman's baseline below last season's numbers. However, the underlying metrics portend maintaining last season's level or even improvement.

Rougned Odor

Odor is a flummoxing player. Last season's big spike in strikeouts puts an asterisk next to any analysis. That said, his BABIP has been all over the place.

Here's his component data over the last three years.

2019

2018

2017

The initial observation is Odor's BABIP on line drives has been consistently low. It's possible this is noise, as one out of eight people will flip a coin three times and it will land heads (or tails) on all three. That is, random probability dictates 12.5 percent of players will have consistently bad luck on line drives over a three-year span. The thing is, he's also incurred bad luck on flyballs each of the past three seasons, so maybe it isn't bad luck at all?

In 2018, Odor's BABIP spiked, emanating from a big leap in groundball BABIP. Curiously, his average exit velocity on worm burners that season was extremely slow, yet he posted his lowest proportion of infield hits for his career so it's not like he was beating out slow rollers. The bottom line is Odor was lucky in 2018, at least with respect to seeing-eye grounders.

Odor's lofted ball BABIP doesn't pass the smell test, at least from component exit velocity data, as he was hitting the ball quite hard. Launch angle isn't the culprit, as Odor's levels are close to league norm. I'll keep sniffing around for an explanation, but as of now it seems like the raunchy scent of bad luck.

As hard as it is to believe, Odor has been snake bit on both flyballs and line drives the past three seasons. Devoid of another explanation, he could be in line for a batting average spike. At the cost it takes to roster him, it's worth a shot, especially since middle infield is deep so planning an exit strategy in case Odor just stinks again isn't difficult.

Tim Anderson

Last season, Anderson was the American League batting champion. Can he repeat?

2019

2018

2017

To Anderson's credit, he's maintained an elevated groundball BABIP for three seasons. Sprint speed in the 88th percentile aided 18 infield hits and he posted a -9.2 degree launch angle so like Moncada, his grounders lost less velocity and bounced fewer times. It's realistic to consider his .303 groundball BABIP a skill.

Unfortunately, the same cannot be said of 2019's lofted ball BABIP. Each component exit velocity was similar to previous seasons, but last year's flyball and line drive BABIP skyrocketed. As satisfying as it would be to contradict conventional wisdom, Anderson's BABIP is headed south, much closer to the .300 level he logged from 2017-2018 than the bloated .399 he enjoyed last season.

Let's wrap things up with a few quick hitters.

Josh Donaldson's component exit velocities have all been on the rise, so he could counteract any expected age-related decline.

Andrew McCutchen's BABIP dipped a tad before his season was prematurely ended in early June. His Statcast xBA was low, but his component BABIP suggest he was in line for an uptick if he didn't get hurt. It's mostly speculation, but McCutchen is a candidate to significantly outperform current market expectations.

There's some concern over Josh Bell's second-half swoon. If his three-year trend across the board of increasing component exit velocity is any indication, Bell could come close to matching 2019's power surge. They'll be some give-back, just not as much as some anticipate.

I invite you to apply this line of thinking to your own draft prep as you identify players with seemingly odd underlying metrics. I'm also happy to offer my dissection on any player posted in the comments.