Making the Case For Bat Speed

A few days ago, there was a question raised on Twitter that sparked a good amount of discussion:

Which is most important for a hitter?

— Jason Ochart (@JasonOchart) March 4, 2022

It’s a straightforward premise with a not-so-straightforward answer. Bat-to-ball skills came out on top, but there’s a convincing case for each option. Bat speed matters most due to a strong correlation between exit velocity and overall production. Bat-to-ball skills matter most because without a sufficient rate of contact, it’s difficult to translate raw power into in-game power. Swing decisions matter most because good ones lead to favorable counts, which in turn lead to favorable results.

This author leans toward bat speed. It’s also the option that received the fewest amount of votes. This isn’t to say one side is in the wrong — the truth is that all three skills are integral to offensive production — but I’m here to defend what seemed like an unpopular choice. As far as I can comprehend, the remaining two are predicated on, or at least amplified by, the existence of bat speed.

Onto the case! It’s a little unorthodox, though, because I didn’t start off to prove a point. Before seeing that poll, I’d been working on a metric to evaluate a hitter’s swing decisions. It didn’t turn out as hoped, so I’ve put it on hold for now, but at least all that effort didn’t go to waste, because how hitters perform according to that metric indirectly reveals the importance of bat speed. Specifically, I’m going to narrow in on 0–0 counts, for three reasons: They’re the most ubiquitous; this article can’t go on forever; and the numbers get a bit wonky in other counts. See, there’s a reason why it didn’t work out.

What a hitter can accomplish in every count — swinging at a strike, taking a strike, swinging at a ball, and taking a ball — doesn’t change, but the value of each action does. Intuitively, taking a strike in a two-strike count is far worse than doing the same in a three-ball count. Since all plate appearances require at least one pitch, the baseline wOBA for each hitter in an 0–0 count is the league-average wOBA — .314 in 2021. Using our site’s Splits Leaderboards, we can find that plate appearances through 0–1 produced a .267 wOBA last season, whereas plate appearances through 1–0 produced a .356 wOBA. Therefore, the inherent value of a strike, from a hitter’s perspective, is -47 points of wOBA. By the same method, the value of a ball is 42 points.

Figuring out the value of a swing is trickier. Let it be known that I’m ripping off Ben Clemens, who came up with the framework three years ago. To figure out what a hitter’s swing is worth, we can add up the probability that he’ll put a ball in play, foul it off, or whiff, then adjust for the strike/ball values of a given count and the hitter’s numbers on contact. I used xwOBAcon; maybe Ben didn’t, but the idea is the same. For example, when Shohei Ohtani swung at a strike last season, he was expected to add 54 points of wOBA. In contrast, when he swung at a ball, he was expected to lose 31 points. He naturally did worse when swinging at pitches outside the zone, but his stellar results against pitches in the zone made up for the occasional mistake.

Lastly, to determine the total contribution of each hitter, we multiply our swing/take values by the number of his real-life swings and takes. The final total isn’t related to actual run production; it’s merely a sum meant to represent a hitter’s performance in 0–0 counts. Yes, the math is super flimsy, but bear with me for now. Here are the ten best first-pitch hitters of last season:

So just the ten best hitters in baseball, period? It sure looks that way, but this is no coincidence. Every hitter in this table produced massive numbers when he made contact against hittable pitches. Some of them did tend to chase after pitches, but they also laid off enough balls to work towards a net positive. These are skills that are applicable to any situation, and so we end up with a bunch of superstars.

On a more depressing note, here are ten worst first-pitch hitters of last season (with a 200-pitch minimum):

These aren’t quite the ten worst hitters in baseball, but it’s close, and they’re at the bottom of the pack for a reason. What do these hitters have in common? When they made contact against pitches, even those inside the zone, the results weren’t pretty. Last season, the league averaged a .395 xwOBA on contact against in-zone pitches; the hitters listed above averaged a .291 xwOBA. As a matter of fact, none of those ten hitters accumulated a positive wOBA on their in-zone swings after factoring in whiff and foul rates. You can imagine, then, the outcomes on out-of-zone swings. That leaves only one method of emerging from an 0–0 count all smiles: taking a pitch outside the zone.

To their credit, the Fletchers and Kiner-Falefas of the world did try, in earnest, to make the best of a bad situation. Here’s a fun example: On the first pitch, Whit Merrifield took almost as many balls (284) as did Juan Soto (285) last season, which is no trivial task. He also ended up letting plenty of strikes pass by, but that’s another trait he had in common with Soto. In the realm of swing decisions, it’s merely the cost of business. My spreadsheets aren’t perfect, but they tell me Merrifield and Soto took 205 and 198 first-pitch strikes, respectively. Even with a mental margin of error, it isn’t a stretch to claim the two behaved similarly, discipline-wise.

And yet, Soto is one of the best hitters in the league, drawing comparisons to Ted Williams at a remarkably young age. Merrifield is a decent hitter at a somewhat valuable position. It’s true that Soto is adept at exhibiting patience across all counts, an attribute Merrifield lacks, but the source of the disparity is located elsewhere. You can see where I’m going with this: Soto hits the ball very hard, and Merrifeld does not hit the ball very hard.

Is that obvious? Maybe. But recall the question from earlier: Which (skill) is most important for a hitter? To answer that question, it makes sense to identify two hitters who overlap in just two aspects, which reveals the relative importance of the third. Soto and Merrifield had similar contact rates last season, and they struck out at about the same rate. They seem to have a similar approach at the plate, at least when it comes to the first pitch. But Soto has an extra trick up his sleeve, which is that he can leverage his power to work towards favorable counts. With a swing as valuable as his, taking an extra strike or two isn’t as scary; he can afford the risk. For Merrifield, the margin for error is razor thin. It’s costly to take a strike, necessitating an absurd contact rate just to keep up with Soto’s preternatural bat speed, which is an impossible task.

Making matters worse, a lifeless swing gives pitchers the green light to attack the strike zone. The ten best hitters from above had an average zone rate of 47.6%, whereas the ten worst hitters had an average zone rate of 53%. It’s not a colossal gap, but the strikes add up. Hitting is easier with a reputation for crushing the baseball; even without a deliberate approach, you find yourself ahead in the count more often than the average hitter. No such luxury is available to a lightweight like Merrifield, whose best response against a strike is about as detrimental as his worst response.

But okay, I’ve rambled on about just two hitters so far, and they’re hardly representative of the larger sample. Maybe results on contact are the difference maker for select hitters instead of the majority. So I went ahead and looked at which variables correlated the strongest to a hitter’s total output in 0–0 counts, with the same 200-pitch minimum as before to eliminate the influence of one-and-dones. Here’s what the results look like in order of correlation coefficient:

If that seems like a bunch of gobbledygook, don’t worry — we’ll go through it, one variable at a time. The strongest variable by correlation coefficient is, well, expected results on contact. Hit the ball harder, and those called strikes and out-of-zone swings are no longer as detrimental. Swings and misses also aren’t as detrimental; in fact, the positive coefficients for both in- and out-of-zone whiff rate show the massive influence of hitting for power, though contact rates still matter. In addition, hitters who put more balls into play, regardless of whether they’re strikes or balls, tend to do worse overall. Considering that sluggers are more likely to channel their might into select swings, it makes sense. After the dust settled, a hitter’s foul ball rate ended up not mattering, likely because the worst and best hitters alike can end up with similar rates, telling us how irrelevant it is.

One last quirk I’d like to point out is the negative correlation coefficient for zone take rate… and for out-of-zone rate. The former is understandable — falling behind is bad no matter who’s hitting — but what about the latter? Isn’t taking a ball beneficial no matter the situation? That’s definitely true, but there’s a hidden relationship at play here. While patient hitters often get ahead in the count, they also show a tendency to take strikes at an above-average rate. So it appears that a good decision leads to an unfavorable outcome, when really, we’re missing a crucial link. This is why we use a hitter’s out-of-zone swing rate instead, which thankfully comes attached with a positive correlation coefficient. Now you know.

What ultimately matters, though, is that xwOBAcon crushes all other variables in terms of importance. There were hitters who had poor results on the first pitch despite top-percentile numbers on contact — hitters like Mitch Garver and Mike Zunino, two hulking backstops who are passive to a fault. But I failed to identify a single hitter who had great results on the first pitch despite a bottom-percentile or even league-average xwOBAcon. In sum, it’s much more onerous to hone an approach good enough to make up for a debilitating lack of power than it is to swing a bat fast enough to make up for an ugly, all-swings approach.

Admittedly, this analysis is limited by the fact that it’s based solely on 0–0 counts. Different counts produce different payoffs that alter which factors matter more or less in generating positive outcomes. Case-in-point: in a 3–2 count, the correlation between a hitter’s in-zone xwOBAcon and total output (r=0.44) is far weaker due to how valuable a pitch outside the zone is. In that situation, hitting the ball hard isn’t necessarily a prerequisite for production. But how does one get there in the first place? By navigating through the first pitch. First-pitch success isn’t everything, and it tends to underrate certain individuals, but it’s undoubtedly the pitch that a hitter sees most often. And in answering how to deal with it, we come back to where we started: bat speed.

It’s common to regard bat-to-ball skills and swing decisions as fundamentals. They’re what hitters must master before adding bat speed, suggesting a natural order of operations. Let’s think outside the box, though. What if the inverse is true as well? Bat speed is useless if a hitter fails to make consistent contact, but there’s little value in making consistent contact with sub-par bat speed. In addition, lack of power creates an uphill battle when attempting to optimize swing decisions. The cost of taking a pitch skyrockets, and knowing this, pitchers respond with a barrage of strikes. There’s a reason why Merrifield can merely imitate rather than emulate Soto. This is the case for bat speed. It sounds blasphemous! But according to the numbers, well, maybe not.