Baseball Archives - SportRun

We don’t just care about bat speed at Driveline and Jordan Walker is evidence of that

Fri, 24 Apr 2026 16:36:28 +0000

There is no doubt possessing bat speed is a crucial trait for a professional hitter.

The math backs the core Driveline development tenet. Bat speed enjoys a strong correlation with on-field hitting performance. This is an empirical truth.

The idea that bat speed matters is hardly a new idea – Babe Ruth, Ted Williams and Hank Aaron all had plus bat speed – but we can quantify it today thanks to modern tech.

And if we can measure something, we can train it. Training and improving bat speed is a core focus in what we do to help hitters.

But that is not the only thing our coaches do.

We also train bat paths, contact points, and hone approaches. If players need to add strength or mobility, we focus on that as well. Training is comprehensive and individualized.

Consider the case of St. Louis Cardinals outfielder Jordan Walker, one of the season’s top breakouts and another Driveline success story.

When Walker arrived at our facility early last offseason to train with Andrew Aydt, now coaching with the Washington Nationals, he did not have a bat speed problem. Walker already owned elite underlying power as seen in his average 78 mph bat speed last season, which ranked at the 99th percentile.

Bat speed was not a focus with Walker, rather, it was his bat path – and how he was arriving to that suboptimal path.

At the beginning of the offseason, Walker visited our Launchpad and had a full analysis done of his swing. He wanted to understand why his underlying bat speed was resulting in below-average results including a second percentile expected batting average and 16th expected slugging last season in addition to poor traditional measures.

What Walker learned was that he suffered from a mechanical inefficiency. And once identified, he began fixing it at Driveline and, at Cressey Sports Performance where he also trains.

“It was really how forward I was coming when I was hitting and what we learned is that when I’m hitting off my backside, I’m driving the ball in the gaps way more consistently,” said Walker to reporters of what he discovered at Driveline. “(Now) I am not rolling over. I’m not getting that top spin on the ball. The focus is really how far back I’m onto my hip, and how I’m hitting on my backside rather than me focusing on launch angle.”

Interestingly, his previous focus on launch angle had led to poor results. It took an understanding of flaws to reach a better physical outcome, a better swing.

This season, Walker’s average bat speed is down a tick to 77 mph – though still very much elite – but his average attack angle changed significantly, increasing from 6 to 9 degrees. The percentage of time he is taking an optimal angle with a swing is up to 60% early this season compared to 48% last season. He’s dropped his ground ball rate by 15.5 percentage points to 32% this season, which is the eighth greatest decline among hitters with at least 300 plate appearances last year.

We want a hitter like Walker to get the ball in the air often.

The following video includes a look at some of what Walker did with our staff behind the scenes, including step backs with short bats and game bats:

Walker did not require a total overhaul. The changes and prescriptions were straightforward once identified in the lab, and they have led to big results:

But they had nothing to do with bat speed. They had to do with bat path, contact point and how his body created that.

“If I’m moving correctly, then the launch angle, and exit velo, and driving it where I want to – it will come up with it,” Walker said.

There are plenty of other stories like this.

Consider what Driveline director of hitting Tanner Stokey told me when evaluating how many clubs in a bag a hitter should carry, using the golf analogy.

“As much as we talk about bat speed and pulling the ball in the air, you’d be pretty surprised how many times we hit it hard and low to the opposite field,” Stokey said. “Backing up the point of contact, catching the ball deep, and sequencing well enough to be able to maintain posture and get the barrel on plane behind it … Hard and low to the opposite field – it’s going to help them reverse engineer a way to have a good efficient path to when they catch the ball out in front.”

Rockies outfielder Jordan Beck told me of his Driveline experience that he “learned about other avenues that I don’t know well, or don’t understand” – including that his bat path was too steep entering the zone.

“It was like I had a shorter window to get the ball in the air,” Beck said. “We worked on trying to give myself a longer window to get the ball in the air. The best guys will get it in the air deep, or, out in front.”

Consider the story of Vinnie Pasquantino working with Aydt at Driveline last year.

Pasquantino had attempted to adopt modern, optimal practices in baseball that included adding bat speed, and more often pulling the ball in the air to his pull side.

But he’d perhaps taken it too far.

“He was pulling the ball in the air with fewer ground balls but there’s no results with it,” Aydt told me in the fall. “And that was mostly just because he was pulling it too much, his spray angle and direction was too far on the pull side. There was an actual interview with MLB Network in the middle of (last) April. He talks about how (he hit) a ball 108 (mph) over there and points towards the dugout. So, his direction was just too pull happy, essentially, and so that’s one of the things we identified… He was kind of over rotating… We changed a few mechanical things, but it’s mostly about getting him to work back to the middle of the field.”

The adjustment to get the Royals first baseman to think more middle of the field played a role in his breakout season. It’s another story that had little to nothing to do with training bat speed.

Now, we care a lot about bat speed but that’s not our only focus. Hitters still have to make quality contact with a pitch, or their underlying engine is of little use. We are focused on the total package, individualized training, and for Walker it wasn’t about moving faster, it was about moving more efficiently.

The post We don’t just care about bat speed at Driveline and Jordan Walker is evidence of that appeared first on Driveline Baseball.

The post We don’t just care about bat speed at Driveline and Jordan Walker is evidence of that appeared first on SportRun.

Investigating a mystery pitch as a test case for a new (computer) vision for pitch design

Fri, 24 Apr 2026 10:42:31 +0000

Tatsuya Imai’s slider was one of the great baseball mysteries early this season before he went on the IL.

A backward slider? A pitch that darted half a foot arm side instead of glove side? How is that possible?

Driveline’s Jack Lambert was among those fascinated when he learned about the pitch Imai brought from Japan to the Houston Astros this offseason. Lambert became interested in investigating the offering when he saw a screenshot of Imai’s pitching hand at the release of the outlier offering.

“You could see with the camera angle from behind that he’s actually throwing it by getting underneath the baseball rather than on top of it, which I thought was unique,” Lambert said. “Because every backwards slider, or screwball-type pitch I’ve seen is going over the top, and kind of getting there by pronating.”

Lambert decided to jump in the Driveline lab and experiment.

While Lambert is a data scientist and baseball analyst by trade — his playing career ended after high school — he has enough throwing ability to experiment.

Could he come anywhere near replicating the Imai slider? And if he could, he reasoned, that meant others with far more pitching expertise, like professional and high-level amateur pitchers and coaches, would be able to learn and repeat it.

And rather than searching in the dark, he had some help: the next generation of pitch design being engineered at Driveline Baseball.

About a decade ago, Driveline pioneered pairing high-speed, Edgertronic cameras – which were not designed for baseball, rather, for scientific research like studying kangaroo rats – with ballistic pitch data from Trackman to create what became known as modern pitch design.

It’s allowed for smarter practice, fewer wasted reps, and far less guessing. Pitch design science has been incredibly effective in allowing pitchers to more quickly design new offerings, and sharpen existing ones, through a data- and visual-based feedback loop. It’s proliferated through professional and amateur baseball.

But as big of a breakthrough as Pitch Design 1.0 was, it did not capture all information, all effects on ball flight. It could not explain all pitch movement. While Trackman does offer observed spin axis and measures, which does get to seam-shift qualities, what it cannot do is inform quantitatively whether a movement – or how much of a movement – was created by magnus or non-magnus forces.

Outside of affiliated professional baseball, players and coaches and data analysts at places like Driveline do not have access to complete Hawk-Eye data. That matters because the public Hawk-Eye data does not include access to seam orientation or spin-based spin axis – only the observed spin axis.

This is in part why we are working to usher in the next iteration of pitch design that is aided by real world AI from our computer vision system.

And that’s where Imai’s backward slider comes in.

One way in which computer vision holds immense potential is having coaches and pitchers better understand how another pitcher is creating a shape.

“The first computer vision helper was Sam (Ehrlich) who ran (video of the pitch) through our CV model to get the spin axis, and seam orientation that Imai was creating to set as a rough target for what I should be aiming for,” Lambert explained.

In the video above, Lambert shows how the CV tool identifies the balls’ spin axis with a neon green dot and the seam orientation in other neon hues.

“I think that’s probably the most applicable use of CV at this point is that… getting some of the metrics I can’t get from Hawk-Eye,” Lambert said. “I’m sure you can imagine if my high school brother in Cincinnati is throwing a bullpen and just doesn’t have a Trackman available, if we can get some footage, get some estimates for what’s going on, we can better like adjust that process from there.”

A computer vision system learns by analyzing thousands upon thousands of labeled images – sometimes even millions like in the case of something Tesla’s early self-driving efforts – using convolutional neural networks to then identify patterns and understand spatial hierarchies. This is deep learning.

Boddy and others at Driveline did tons of labeling, heavy lifting, to train the system – labeling seams, spin axis, and pitch types of thousands upon thousands of recorded offerings. The system is still learning, it’s still getting better.

Lambert made about 50 throws earlier this month, studying the ball flight and impact of each adjustment guided by feedback from Driveline’s real-world AI effort.

He wasn’t able to perfectly replicate the pitch in one bullpen, but he was able to mimic some of its characteristics after just one grip and release recommendation from our computer vision model and some tweaking.

“What I was able to recreate was I could get the high, arm-side run that he would throw,” Lambert said. “I could not kill the spin efficiency enough to get the gyro action. I found it easier to basically create a changeup profile with a supinated (release) than to create the true gyro version of his slider. That was most of the iteration process. It actually didn’t take that long to produce some pitches with the high, arm-side run.”

Imagine what actual pro and college pitchers and coaches might be able to do with the tool?

That is one application of a computer vision model: helping coaches and players understand how to begin with a pitch.

Driveline’s pitching director Connor White explains the other great benefit of deep-learning aided pitch design.

“The speed of analysis is one of the most exciting things,” White said. “We want to keep those pens game-like. So, if that’s having to stop after every pitch and look at a bunch of metrics and consult the video, and next thing you know it’s been a minute between pitches or more it really kind of breaks that flow… The computer vision allows you to look at the observed versus like spin-based (movement), getting the closer to the ball physics of what’s happening in real time.

“The speed at which these (advancements) can be applied is just so exciting.”

Shortening the feedback loop, the understanding of what a pitch is doing, is indeed exciting.

Our computer vision model is not a finished product, but it is already having results in our gyms.

Driveline pitching trainer Grayson Liebhardt says it’s already helping him as a coach.

“It’s a really helpful tool,” Liebhardt said. “It’s earlier in development but it’s helping us bridge the gap, and understand seam orientation without any access to the data that the pro organizations have… It gives us just more context on why a pitch may move a certain way, or, how to optimize seam orientation for certain movement profiles.

“Pitch physics isn’t completely solved. There’s a lot of stuff, non-Magnus wise, like seam-shifted wake, and possibly other variables that we may not even know about, that affect ball flight,” Liebhardt said.

For instance, Liebhardt notes we know how seam-shifted wake affects ball flight but we cannot quantify how much it affects movement alongside other variables, some he notes that “we may not even currently consider.”

We don’t know everything. And what’s so exciting is computer vision will lead to more understanding.

“These tools are super helpful for utilizing the information that we already have,” he said of CV, “as well as collecting more information to be able to learn more about pitch physics.”

What’s also exciting about real-world AI breakthroughs is they keep learning, they keep getting better.

“The cool part for me is being able to have an easier way to look at seam orientation and spin axis,” Liebhardt said. “That’s just something that, historically, you’d have (study an) Edgertronic camera and try and find it and guess where the spin axis would be.”

Now, Liebhardt has a tool that cuts out more of the guessing.

He shared this clip of another Imai-like mystery pitch, this one from Driveline athlete Tony Oreb.

We can again see how the CV has learned to identify and mark the axis and seams to allow a fuller understanding of the pitch’s flight.

Liebhardt responded to Lambert’s efforts with new insights observed from the CV model: “Different orientation than yours and Imai’s. Can be done with a (four-seam) and (two-seam) orientation?”

We are already gleaning new insights and helping athletes with the next generation of pitch design. While our pioneering efforts have already helped pitchers and coaches develop scores upon scores of pitches, the new generation – a real-world AI effort – promises even more. It’s part of the Driveline process: constant iteration, and constant improvement.

The post Investigating a mystery pitch as a test case for a new (computer) vision for pitch design appeared first on Driveline Baseball.

The post Investigating a mystery pitch as a test case for a new (computer) vision for pitch design appeared first on SportRun.

To swing or not to swing, that is thee question

Fri, 03 Apr 2026 10:25:01 +0000

With the tying run on third base in the ninth inning, two outs, and the count full in the World Baseball Classic semifinal meeting last month, Geraldo Perdomo watched a wicked Mason Miller slider break well below the strike zone.

The offering should have been called ball four. Fernando Tatís should have come to the plate with runners on the corners and an opportunity to advance to the Dominican Republic to the final. But this is a game played, and umpired, by humans.

Home plate umpire Cory Blaser called a strike. Team USA celebrated. Perdomo froze in shock.

Given the leverage of the moment, it was the most costly missed call of the tournament. It was a true binary, survive-or-die, game-ending errant call, if only we had ABS technology in the WBC this spring.

As the game ended, a new debate immediately began: To swing or not to swing? Did Perdomo make the correct swing decision?

There are those who believe he ought to have been protecting with two strikes, expanding his zone to attempt and cover a large halo around the strike zone.

Did Perdomo make the right decision? It is a question worth exploring to understand probabilistic decision-making hitters are tasked with making – and doing so in milliseconds. It also raises another question: how can we help train and coach players to make better decisions and hone approaches? After all, the strike zone is the most important real estate on the field for batters as well as pitchers.

Let us begin with what Driveline’s swing decision model had to say about the decision.

It graded Perdomo’s decision as an elite decision.

That might be difficult for some to accept, but we live in a world of probability, not certainty. There will occasionally be suboptimal outcomes as byproducts of good processes. Driveline’s Jack Lambert explained the model’s logic and accounting on X:

“Sixty percent chance of swing, 40% chance of take,” Lambert wrote of the situation. “If no swing: 0.5% chance of (a called strike). If a swing, 63% chance of a whiff.”

The math was on Perdomo’s side in not taking a hack.

That pitch location is very rarely called a strike.

And even if Perdomo swung, it’s about impossible to do much damage to a Miller offering located below the zone – let alone contact it.

When examining the probabilities associated with swinging versus non-swinging in this situation not seemingly much room for debate.

(Our model is trained on MLB data, so it does not take into account WBC umpiring but Blaser is a MLB umpire).

While coaches often talk about “protecting with two strikes,” our model can help

inform batters of where their decision making might be errant and perhaps have them learn to think more probabilistically.

Professional player, and hobbyist data analyst, Robert Stock shared similar logic in thinking through a decision tree on X:

“A huge misunderstanding that people have is that an MLB batter should be protecting against every (two-strike) pitch that could possibly be called a strike. An MLB batter shouldn’t even be protecting against all actual strikes. To do so would mean increasing your chase rates to an absurdly high level. And the likelihood a batter has a positive result (foul ball or a hit) on a perfectly executed two strike pitch is very low. Couple these two facts together, and anyone arguing otherwise means one of two things – you genuinely didn’t know this until now or you have an idea floating around your head that doesn’t match reality.”

And if pro batters shouldn’t be protecting against two-strike pitches well off the plate, many amateur players should not either.

How do we think about swing decisions in our training of hitters at Driveline?

I asked Driveline lead hitting instructor Tanner Stokey about the Perdomo debate.

“On paper – on paper – (Perdomo) made the right decision. It was a ball,” Stokey said. “But the one thing for me – it’s based off the game state, the game situation – is it’s really difficult to stomach leaving the outcome of the game in the umpire’s hands. I think it was the right decision. It was a ball. But generally, I’d be pushing there to swing on the edges right there. Realistically, though, the chances of anything productive happening on that.”

This is where art starts to creep into science.

While trying to protect six inches off the plate against a Miller slider is a low expected value decision, hitters do want to be adaptable to different game states.

Hitters can benefit from the ability to expand and spoil a pitch that is an inch or two off the plate, a pitch residing in the shadow zone. The ability to fight another day, to survive to the next pitch. (The Perdomo pitch was below the shadow zone.)

Thinking about the quality of swing decisions is further complicated as Stokey notes that not all strikes are “created equal,” especially early in the count.

“I look at swing decisions differently than I look at an approach,” Stokey said. “Swing decisions in terms of models, they are just essentially grading your decision to swing or take a given pitch.”

Models do not capture all considerations, everything that matters in the calculus of an approach.

A hitter’s approach must consider game state, count leverage, and a hitter’s specific strengths, to understand when a batter should zero in on a location, and when he should expand. Our data-based tools – and training programs – can help hitters understand how to optimize their approaches.

“Especially early in the count, it’s incredibly important that you are playing offense and you’re not taking defensive, reactionary, passive swings,” Stokey said. “Early in the count, if you’re going to swing at a pitch down-and-away on the black – the chances of you doing something productive there is slim, right? Even though it’s a strike, and it’s going to put you further behind in the count, the chances of you doing something productive with that swing are not very high.

“Now, on the flip side of that, (a hitter) should damn near 100% of the time be swinging at pitches in the heart of the plate because that’s where most production happens.”

I once asked Joey Votto about how he equated the process of getting the pitch he wanted to hit as something akin to boxing.

“It’s like a boxer who is always trying to lead the guy into his straight. You have to manipulate him with your footwork. Same type of thing in baseball,” Votto said. “You have to figure out a way to funnel [the pitcher] into your hot zone. That comes with patience and that comes with accepting or realizing there will be some error on their side.

“It’s almost like as a hitter you have to be a counter puncher. The best way to be a counter puncher is just to sit and wait and absorb and then counter with whatever you think your strength is.”

Perhaps the focus on the swing-or-to-not-swing question regarding Perdomo should not have been tied to the full count pitch but, rather, the third pitch of the at bat when Perdomo watched a fastball split the center of the plate.

Granted that offering was 100.7 mph out of Miller’s hand – not the easiest pitch to turn around – but by not swinging at a center-cut fastball, the count leverage went from neutral to a 1-2 count, and severely in favor of Miller. That was also his best pitch to hit.

Hitters should all be more aggressive when they have leverage but they also ought to have individualized approaches.

For instance, Aaron Judge and Luis Arráez ought to have much different approaches.

“If you take Judge versus Arráez and they swing at a slider on the black, down and away, the chances of an Arráez doing anything good with that pitch when he puts it in play is very slim,” Stokey said. “The chances of Judge doing something productive is significantly higher because he has so much bat speed, he has so much ability to create velocity and impact on baseball.

“Low-power guys should be even more selectively aggressive. When they are swinging early, they should really be leaning into getting that ‘A swing’ off. “

This was even an issue for Mookie Betts in working with Stokey back in 2023. Betts was too worried about making contact, and having the lowest K rate possible, and it was sapping his power. One aspect tied to his rebound season is he became more selective about when he swung. Betts added bat speed but there was also a mindset change, Stokey told me for this piece last fall.

“The thing with (Betts) was ‘Man, your bat-to-ball skills are so good,’” Stokey said. “A lot of times you see players with very good bat-to-ball skills, they don’t want to swing and miss. So, while Mookie didn’t have poor chase rates he would expand on the edges and swing on balls in the shadow zones and slow himself down for the sake of putting the ball in play. He’d slow himself down and make poor, unproductive contact.”

Even the best hitters in the game might not have quite the right mindset regarding where they belong on the power-contact spectrum.

The good news? There are ways to improve. But there is the question of training economy.

There are only so many hours in the day, only so many reps a hitter can take. How much of those should be geared toward a regime that focuses on decisions over, say, bat-speed training?

“If you want to look at it through the lens of the big three: power, contact, and swing decisions… good, a good training program, a good training environment, is going to take into account all those things,” Stokey said. “You can very easily train your power, your bat speed, and your contact skills, and your swing decisions in the same environment.

“Now, you need really awesome technology, like Trajekt, or you need a pitcher that can mix pitches to you in batting practice, things like that… Say a hitter gets a 50-50 (mix of pitches) off a Trajekt, or an iPitch something like that. (The athlete) is figuring out how to get their best swing off and make quality contact in an ‘A-swing’ situation, or finding a way to hit a line drive ball flight in a two-strike situation.”

A core principle of Driveline coaching is to embrace and create environments that force athletes to adapt through implicit learning.

For example, a specific two-strike protection drill is probably a suboptimal use of time, Stokey says. Rather, we want to create difficult, game-like environments.

Beginning this season in the majors, ABS technology will begin to protect a hitter like Perdomo from the unfortunate outcome of having made a correct decision in a high-leverage spot hurt by umpire error. Perhaps someday 99.9% of ball-strike calls will be made correctly. Perhaps one day the technology will trickle down to amateur games, too.

Still, even if rogue umpire calls decided fewer fates, the art and science of swing decisions and approaches will always be important.

Not all strikes are created equal. The optimal areas to attack differ for hitters. It’s a game of probability as Perdomo found out in the WBC. There is no certainty involving any one swing decision, but we want to place our athletes in a position to succeed as often as possible.

The post To swing or not to swing, that is thee question appeared first on Driveline Baseball.

The post To swing or not to swing, that is thee question appeared first on SportRun.