Two outs, two strikes. Tying run on second. Gerrit Cole on the bump.
You’re pinch-hitting, you tell yourself in the batting cage. The crucial pitch is on the way — just as soon as it matriculates its way through that ubiquitous metal tripod, the pitching machine.
Shoop. Here it comes. A 75 mph, arrow-straight fastball, down the middle.
Actually facing Gerrit Cole is nothing like that. He could throw you a blistering four-seamer. Or a slider, or a cutter, or a changeup, or a curveball. And whatever he throws, it won’t be down the middle. Living out a version of events where you win that battle means being prepared to hit any of them.
For much of baseball history, batting practice hasn’t evolved much beyond 75 down the middle. Even in the majors.
Now, a wave of smart pitching machines are giving hitters a chance to practice against the type of major-league quality stuff they will actually have to hit in games, sparking a hush-hush race to optimize batting practice.
Powered by reams of pitch tracking data, the most advanced of these new machines can replicate the experience of facing just about any MLB pitcher, from velocity to movement to release point and how the ball looks leaving the hand. Only one MLB team was using that machine — made by the Canadian firm Trajekt Sports — in 2021, but seven clubs will have access by the end of this season.
In a league where pitchers have leveraged data to design daunting arsenals, baffle batters and stifle scoring, the advances of the pitching machine could present a real chance for hitters to strike back.
Reverse engineering an innovation for hitters
Like a lot of baseball innovations, the improvement of pitching machines can be traced back to the advent of Statcast. MLB’s tracking system started reporting public data in 2015, and pitchers were almost instantaneously parlaying detailed information on velocity, spin rate, spin direction and other characteristics into nastier, more deceptive arsenals.
“This is basically just the batters’ answer to that,” said Daniel Halleran, the lead engineer for Spinball Sports.
Noticing the proliferation of those metrics, the Illinois-based pitching machine company hired Halleran to help design a machine that could backtrack to replicate pitches based on those measurements.
“Over the last several years, with the advanced analytics and stuff, the pitchers were getting more and more of an advantage on the hitters, and pitching machine technology hadn't really changed all that much,” Halleran said.
The iPitch machine that Spinball developed allows teams, or independent coaches, to upload and organize data into pitch sequences that can be controlled from an iPad in the batting cage. So an MLB team can organize it to imitate actual pitchers’ arsenals, and can even tell the machine to throw a randomized sequence of, say, Jacob deGrom’s pitches.
Reverse engineering real pitches for practice was the first great leap forward, according to Tanner Stokey, Driveline Baseball’s director of hitting.
“The biggest change of late is the ability for machines to throw mixed pitch sequences,” Stokey said in an email to Yahoo Sports, explaining that Driveline uses the iPitch nearly every day. “The ability to not only mix pitches, but create your own custom pitches and sequences to mirror the arsenal of MLB pitchers has been incredibly valuable to the MLB hitters we work with.”
Having gained traction with trendsetting training facilities like Driveline and forward-thinking college programs, the iPitch soon broke into the majors, too.
A prominent study in 2018 — by an analyst who now works for the Tampa Bay Rays — questioned the point of the usual home run derby-style batting practice. Teams seem to agree, and now more are moving toward practice that imitates competition. Halleran says 17 MLB organizations have adopted the iPitch, and eight teams are using five or more machines. Several other companies, including Hack Attack, also produce machines catering to the idea.
It’s hard to gauge overall acceptance because individual players can decline to take part. Houston Astros hitting coach Troy Snitker chalked up some of the resistance to the fact that “players are used to practice in which they succeed.”
But the tenor of batting practice is undoubtedly moving toward a challenge. Bryce Harper, for instance, went from refusing to use the machines in 2019 to embracing them in 2020, although his comments seemed to indicate a technology improvement may have changed his mind.
The way you can tell the iPitch has become ubiquitous is that major-league coaches will talk about it by name. Some teams will use it on the field. Many travel with it, with portability being a major selling point for Spinball.
Pitching machines inspire a new MLB arms race
There are other machines that coaches and executives are more hesitant to talk about.
One leap quickly demanded more leaps. When Joshua Pope and his co-founders at Trajekt Sports set out to engineer a new pitching machine, they knew they needed to completely change a fundamental tenet of the design: how the ball spins.
“We started out building a gyro machine,” Pope said.
“Gyro” spin is the type of spin you associate with a bullet, or the spiral of a football. For a long time, it was viewed as an inconsequential — or even inefficient — byproduct of throwing a baseball. Pitching machines from your basic batting cage model all the way through the iPitch didn’t bother with it, instead using two or three spinning wheels to impart the pure directional spin that was thought to explain almost all movement.
But over the past few years, baseball’s statistical community has become increasingly attuned to the effects of seam-shifted wake. Explained and explored most prominently by a Utah State professor named Barton Smith, seam-shifted wake is a force created by the raised seams on a baseball interacting with the air.
Depending on how the seams are positioned when the ball is thrown, gyro spin can be a huge deal. The new insights make dissecting pitches far more complex, but ultimately demystify the success of pitchers like Cubs stalwart Kyle Hendricks and breakout Yankees star Nestor Cortes. They also, of course, gave pitchers new material to work with.
For a hitting coach intent on practicing against game-like pitches, the message was clear: A whole swath of the league’s weapons couldn’t really be replicated by the pitching machines in circulation. If you’re preparing for a dominant two-seamer on a machine without gyro spin, you’re not really preparing at all.
“The gyro was kind of the missing link,” said Lloyd Smith, a professor of mechanical engineering at Washington State University who works with the school’s prominent Sports Science Laboratory.
Smith’s group at Washington State, which tests bats for certification and now helps MLB test its baseballs for drag and lift, devised a cannon that creates gyro spin on a baseball by creating the effect of “fingers.” While he believes his team’s solution to the gyro spin problem is better, Smith said Trajekt also solved it and beat his more research-focused lab to market with a product that appealed to MLB teams.
Once Pope and the Trajekt team had gyro spin, and the seam positioning that dictates seam-shifted wake, they kept digging deeper into the variables that make a pitch what it is. There are arm angles, release heights, different starting locations on the rubber, deliveries with different ways of obscuring the ball, and on and on.
Pope said they determined a real pitcher has “12 degrees of freedom” of movement in throwing a pitch. He says the machine they built, the Trajekt Arc, can recreate 11 of them.
The Arc, which is actually branded as a “pitching robot” on the Trajekt website, is the machine MLB teams don’t want other teams to know about. Complete with a projector to simulate the optical experience of batting, it will show you a real pitcher’s delivery from the hitter’s point of view and fire a replica of his pitches timed up to the visual.
Pope wouldn’t comment on Trajekt’s team clients, and was careful not to reveal any identifying information other than to confirm that seven MLB clubs have placed orders for a total of nine machines. Six machines have been deployed so far, and one team — which he called the company’s “early adopter” — was using the technology last year.
Pope also declined to comment on how much they cost, but explained that teams lease the machines and a custom software package from the company.
They work much like a souped-up iPitch. Teams load pitch tracking data into a tablet or computer, along with the batter’s eye video footage. On a day they plan to use the machines, a staff member will load the pitchers they want to practice against into the Arc and fire off four sample balls for each pitch type. That takes about a minute, Pope said, so preparing the machine to imitate a typical starting pitcher takes about five minutes. Built-in systems provide instant feedback on how closely the pitches mirrored their real-life targets, and can also track a batter’s performance metrics against the machine.
Load up 100 balls into the Arc and it can fire a new one every eight-to-12 seconds, either from a preset sequence or chosen live by a coach on a tablet. One drawback of its immense capabilities? The Arc does not travel.
One newly deployed Trajekt machine lives in the bowels of Citi Field, in one of the batting cages behind a curtain. The New York Post initially reported that the New York Mets had purchased an innovative pitching machine from a Canadian company earlier this season. And last week, sources confirmed to Yahoo Sports the Mets recently installed an Arc.
Their experience is still early and limited, and some players are hesitant to speak publicly about the new technology. Francisco Lindor said the machine “throws fuego” but echoed a widespread concern for the toll on hitters’ hands from facing game-speed pitching in practice.
Mishitting a ball in competition is one thing. Mishitting dozens of them in practice is a sacrifice most aren’t willing to make.
Spinball engineers have heard about teams using softer “dimpled” balls for machine practice because they save hitters’ hands. Pope said the Trajekt machine is currently only compatible with real balls, but the company is studying ways to work with other types, like Driveline’s Smash Factor ball, a softer model built to behave like a baseball.
And although the machines are supposed to give hitters a chance to catch up, the Mets have also found a way for their pitchers to benefit as well. Manager Buck Showalter expressed an interest in having pitchers stand in against themselves to get a batter’s-eye view of their own arsenal. At the very least, ace Max Scherzer has done so. He reported that he thought the Arc replicates his fastballs accurately, but not the breaking pitches.
New York is one of the seven teams that either has or will get a Trajekt machine this year. The other six have been even quieter about their foray into the latest hitting technology, in an effort to preserve whatever competitive edge they’ve been able eke out.
Asked for confirmation, one executive of a team suspected to have an Arc balked, saying, “I don’t know what you’re talking about.”
When pressed, he countered, “I don’t think we’ve said anything about it publicly,” and “Who told you that?”
Why new technology could change the game for hitters
Baseball is the rare sport where the offense is always on the defensive.
Pitchers control the action. They have the ball. They can do things intentionally. Hitters must react to what’s thrown their way. That gap has always existed.
“We're not solving the same task,” said Brewers manager Craig Counsell, whose club has an iPitch. “Hitting is such a different task than pitching. There’s so many more inputs to solve a hitting thing. The reaction causes so many more problems in creating an answer.”
The imbalance has been only exaggerated by the proliferation of data that allows pitchers to quantify and fine-tune their craft. Declining scoring and soaring strikeout rates have been a persistent concern for MLB, even with a sometimes-juiced ball buttressing the top-line numbers. The league has limited sticky substances and the size of pitching staffs to give hitters a better chance, and the pitch clock could soon mark another attempt.
If Trajekt’s technology proves effective in mimicking major-league at-bats, it could represent hitters’ best chance yet at organically turning the tide. It could lend a tiny bit of agency to the players who are always reacting.
The potential power of the machines is best illustrated by a bedrock principle that guides modern pitcher usage. Hitters get better the more they see a pitcher in the same game. Pitchers facing a lineup for the first time in 2022 strike out 23.3% of batters and hold them to a .236/.310/.378 line. They have a 3.59 ERA. But pitchers facing a lineup for at least the third time have an ERA of 5.50. They strike out only 18.6% of batters and allow a .267/.326/.443 slash line.
This is known as the Times Through the Order penalty.
“The effect exists,” Pope said. “Our whole premise was what if you could see 100 pitches before the game started?”
You can argue about how much of the penalty stems from increased familiarity and how much stems from the pitcher’s fatigue and an increased understanding of a specific game’s circumstances and strategy, but there’s no doubting its significance. Stokey, the Driveline hitting director, said he would encourage anyone he is coaching to use the technology.
"Hitters are almost always more successful against a pitcher their second, third or fourth time facing them in a game, and that’s becoming more and more rare in today’s day and age for exactly that reason," Stokey said, nodding at the league's increased use of bullpen arms to reset the penalty and throw fresh looks at hitters. “So it makes perfect sense that you’ll be more prepared to do damage vs. a pitcher your first at-bat if you spent the afternoon prior to the game getting your timing right vs. a machine closely mirroring their arsenal.”
It’s too early to test the effectiveness of the Trajekt machines. It's not known exactly who is using it, or when. Pope said the company is working on a statistically significant study, but for now he is happy to know some players are buying in.
“Anecdotally, players have used it game in, game out.” he said. “And that’s enough for us to be really confident in its effect.”
There will undoubtedly be resistance to — and questions about — the new training methods. Virtually every major leaguer got to where they are using the same old pitching machines and same old batting practice as previous generations.
Success in at-bats that matter, though, will keep getting harder to attain without adaptation.
“In training, anything you can do that is more ‘task specific’ or closer to the game itself, the higher the likelihood of retaining those improved skills/changes transferring to the game itself,” Stokey said.
The dynamic between pitchers and hitters has always favored the guys on the mound, leaving batters desperate for ways to close the gap. That’s obvious just through decades of purported sign-stealing efforts, including the 2017 Astros scandal that put the sport on edge about the ways technology could be illicitly exploited.
The commissioner’s office is aware of the growing use of technologically enhanced pitching machines, fielding inquiries from interested clubs and curious executives. The legality of such machines is governed by existing rules that prohibit the use of electronics during games, except for the dugout iPads controlled by the league. The result is that teams are not allowed to use any pitching device, like the iPitch or Trajekt, that relies on computer technology once the game has started — unless there is a rain delay, which constitutes a suspension of regulation.
Policing of this policy is handled by the three “gameday compliance monitors” stationed at every game. Thus far, they have not raised any concerns about impropriety.
As teams race to stay on the leading edge, there are big incentives to solve those barriers to entry.
“I think we’ve applied so much more expertise on the pitching side,” Minnesota Twins general manager Thad Levine said recently, speaking broadly about whether hitters could ever utilize tech to catch up to pitchers. Citing the sheer number of pitches thrown in a game compared to an individual hitter’s at-bats, he called advancements in pitching “lower-hanging fruit.” He noted that half a decade ago, as pitching coaches became more analytically minded, hitting instruction remained focused largely on the mental side. That’s changing.
Virtually every team understands that part of a modern pitching coach’s job is analyzing and translating the vast array of analytical information available to his staff’s benefit. Now that the technology available to aid hitters is closing the gap, so soon will the training methods.
“So I would expect,” Levine said, “that we see massive gains in the next two-to-four years.”