Excerpt. © Reprinted by permission. All rights reserved.
1.
deCervo
"How Can You Think and Hit at the Same Time?"
There was no indication that anything unusual was taking place on an early Saturday morning in August at the Hilton Garden Inn, of Avondale, Arizona, other than the piece of loose-leaf paper taped to the wall by the elevator bank. On it was scribbled in black Sharpie: decervo testing room 307. The room number was underlined. The tone was "no trespassing." Still, the housekeeper knocked on the door of Room 307 at 8:15 with an armful of fresh towels. No one answered, so she used her key to enter. When she did, she did a double take. The furniture in the dumbbell-shaped suite had been rearranged completely. The beds were still made and the blinds were drawn. Two scrawny, acne-pocked Latino teenagers in T-shirts and sandals were seated at matching desks on opposite sides of the room staring unblinking at laptop screens. Each wore a sort of thin metallic hairnet, with wires snaking down the back of their necks. A pile of plastic syringes and two padded briefcases lay scattered on the floor. The only sound came from soft, intermittent taps on the laptop keyboards. Neither of the men looked up to see the housekeeper quickly drop the towels off and go.
In the everlasting war for even the slightest competitive advantage in Major League Baseball, the battlefields have come to look a lot different than the playing fields. They have left the playing fields behind. This new terrain was once thought to be impregnable. Now, suddenly, held captive on Saturday mornings in suburban hotel rooms, it was spilling its secrets. When other teams learn of this, they will undoubtedly try to do the same. "Moneyball" was that way; once the data-driven revolution started, it became difficult to contain, until every team started using advanced analytics to discover new players or rediscover old ones. Then the battle had to be moved someplace else. Those teams that were late to that data revolution now had a chance to get ahead in this one. This data revolution required a new type of radar gun, one that could measure in milliseconds.
At 8:25, there was another knock at the door of Room 307. A third baby-faced teenager appeared: Manny, a shortstop, wearing a gray T-shirt and sandals, his eyes puffy and reddened. The boys, they were really just boys, had played in a doubleheader the day before, in the searing Sonoran heat, as the playoffs neared. This being a rookie league team, below Single-A, even below Low-Single-A, every player had recently been drafted or acquired from overseas. It was their first taste of professional American baseball. They remained years away from a whiff of a chance at the Majors; most will never even get that. But as the newest and youngest parcels of a Major League Baseball empire, they are handled delicately. They reside in the hotel, a short drive from a hulking, concrete-and-glass Spring Training complex, where they relax and train in uniforms that bear the familiar colorway of their big-league parent club. They are currently chaperoned by Frank, the organization's director of sports science, who popped in and out of Room 307 with a list of the telephone numbers to each of the players' rooms, in case any of them tried to sleep in. A stocky man with soft blond hair, reddish cheeks and bright eyes, he is friendly, but with a no-bullshit mien, like a waiter at the end of his shift. Frank did not seem to care that 8:00 a.m. for an 18-year-old on the Saturday morning after a late doubleheader is a considerable, if not downright malicious, request. But there was a lot to get done. Jason Sherwin and Jordan Muraskin were only in town for two days. Their sort of expertise is not easily replaceable. The club paid $2,000 to fly them out there. As the ballplayers tapped on their keyboards, and Manny waited on the couch, Jason and Jordan shushed about, adjusting the hairnets. They chatted idly with Frank about the upcoming fantasy football season, but there were giveaways that they were not members of a typical athletic entourage. Noticing the colorful symbol on the front of Manny's T-shirt, Jordan asked him, "Is that a Google shirt?" "No," he replied sheepishly. "World Baseball Classic."
A spot opened up at one of the desks after the first player finished. Manny took a seat and waited as Jason prepared the laptop and Jordan readied the headgear. He used an alcohol swab to dab behind the player's ears and fitted the strange translucent swim cap-an EEG headset-over Manny's short hair. Then he grabbed a syringe and squirted a pale creamy substance into the seams around the nine spots where the sensors were expected to maintain the closest contact with the skull. The cream, the consistency of toothpaste, is a conductive salve for the electrodes. "You remember this?" he asked. Manny nodded. He quickly typed his username and password into the system and the screen went dark, with only a small rectangular box appearing in the center. A moment later Jason signaled the program was ready.
"It'll take about 40 minutes," Jordan said. "Do you want any practice?"
"No," Manny said. "I'm good."
The simulation began. "And we're off," Frank said.
First came the orthopedists. They came to baseball in the late 1950s and early 1960s, transforming how pitchers were assessed and treated. The psychologists followed. Then the optometrists, strength coaches, massage therapists and nutritionists. The economists and sabermetricians. The Zen masters and sleep doctors and yoga instructors. And finally, at last, there came the neuroscientists, fresh from school, brandishing doctorates and peer-reviewed papers and exactly nothing of any value that mattered to a baseball executive other than their answer to the question, ÒCan you make my team better?Ó To which, Jason and Jordan, cofounders of a startup called deCervo, would answer unequivocally, ÒMaybe.Ó They were not sure. They were scientists, and they had no background in business. They seemed to have no business in baseball. That a sport moored to tradition-where managers still wear uniforms in the dugout and make calls to the bullpen using a landline phone; the last major league to adopt on-field instant replay-had any interest in doing business with them was also unclear. But they wanted to help. There is a saying about baseball that, even after 170 years, you can still see something new in any game. Outside of cheating, though, there was not much new for improving the act of hitting. Hitters can so often seem besieged by so many factors-mounting velocities, defensive shifts, the unyielding constraints of our visuomotor system-that reaching base safely even on occasion is widely considered a paragon of skill. Ted Williams once called hitting a baseball Òthe most difficult thing to do in sport.Ó Some say that the hands need to be ÒcoordinatedÓ well with the eyes, which can be deceiving if you are one to believe that hitting, like a lot of athletic endeavors, is mainly a rote exercise based on muscle memory, a term coaches use often. Coaches also say things like Òwatch the ball hit the batÓ and Òslow the game down.Ó In 1921, psychologists at Columbia University designed a battery of sensory-motor tests for Babe Ruth, under the guise of empiricism, to explain his prodigious hitting ability. After the tests, the researchers declared (misleadingly) that RuthÕs vision, reaction time and coordination were far and away better than othersÕ. It is unknown if any other ballplayers were assessed. The study was headline news at the time. A hypothesis for what it takes to be a great hitter emerged: Be Babe Ruth.
But almost a full century later, two other researchers from Columbia University began to try a different approach, this time with techniques adopted from a place called the Laboratory for Intelligent Imaging and Neural Computing. Their assessments probed deeper than the cursory physical examinations performed on Ruth. They went hunting for data. Brain data. With the EEG and a carefully tailored video simulation, Jason and Jordan believed they had landed upon a novel collection method for the type of information previously left to guesswork. They could transport their services anywhere, taking their exploratory findings out of the confines of the laboratory and bringing them into the dugouts of Major League Baseball. It was relatively quick, entirely painless, and no more invasive than using an Apple Watch to measure your heartrate.
When the two researchers were first testing the headgear on players for the baseball team at Bradley University, in the fall of 2014, they finally got the full scope of the concept they were developing and what it could mean for baseball. The coaches wanted to know the results of one player in particular, who seemed to struggle at the plate despite obvious athleticism and a picturesque swing. Jordan, then 29, slender, with a boyish face, trimmed dark hair and strong eyebrows, patiently explained the metrics he had recorded during the player's neural assessment, which they categorized with academic patois like Neural Decoding Performance, Decision Position Metrics, and Neural Discrimination Strength. On Jordan's laptop, the readout was filled with line graphs, column graphs, data tables, and heat maps. He explained to the coaches that the player's "neuronal curve has shifted backwards": He was late recognizing certain pitches and therefore late in deciding whether to swing.
There was silence on the other end of the line. Finally, one of the coaches said, "We never understood why he's not the best player on the team." Now they had a clue. "It was like, 'Yes, yes, yes!'" Jordan told me later.
What Jason and Jordan were showing was a baseball version of what is known as rapid perceptual decision-making, which is obviously quite different than the kinds of decisions we mull over (What should I have for lunch today?) or expressly calculate (Which exit off the highway should I take?). But the ability to hit is often mistaken for reaction time, which is virtually the same for everybody. We make fast decisions all the time on a day-to-day basis. They can be reduced, like most things related to the brain, to the patterns of spatially and temporally distinct and interdependent neuron activations. Baseball players, the really good ones, produce or respond to these activations in ways different from other people. The result is they can recognize certain pitches the same way automobile enthusiasts can recognize the make and model of a car as it disappears out of sight, or the way bird-watchers can detect an instantaneous flash of color or flight pattern. It is similar to the way a chess master can quickly visualize and interpret movements on the board. We have always known this, more or less tacitly, from quotidian statistics such as batting average or on-base percentage, which have been used to assign value to players for decades. But these, deCervo likes to point out, are ad hoc variables. They come only after the player has finished his at-bat. They don't consider how much luck is involved in inflating or deflating those statistics, such as whether the ball skipped off a fielder's glove or the wind shifted direction to rescue a fading bloop. There are advanced analytics that help factor in some of that, but they are intensive and complicated. They are not always computed in the Minor Leagues and rarely get weighed by scouts scouring high schools or overseas for future big-league talent.
But from their earliest trials with the varsity teams from Bradley, Brown and Columbia, deCervo could produce graphs that pinpointed when the batter decided to swing versus when he decided to take, along the time line of the pitch, down to the millisecond. A hitter stands at the ready, sees a 90-miles-per-hour slider come toward him, and makes no movement of the bat. DeCervo could still delineate the moment he made that choice to look at the pitch, rather than go for it. It registered as activity on the EEG. It registered as tiny explosions of neural action. After more testing, they had graphs that showed the spectrum of response times based on different pitches; graphs that assessed the batter's concentration level (based on eye movements and the corresponding flutter of brain activity) before the pitch is thrown; graphs that correlated to the part of the brain that is firing when decisions are made. After a year, they looked at the traditional batting statistics the players had produced and compared them to their neural metrics. They showed them to the coaches. "It was dead-on," Bradley's coach, Elvis Dominguez, told me. He wound up organizing his bench around whose readouts showed a better capacity for laying off pitches, which, he believed, contributed to a higher on-base percentage.
Jason and Jordan cast out for bigger fish. They published a few academic papers and opened a Kickstarter to raise funds. They leased space at a shared desk in Columbia's subterranean startup incubator. They attended the Sloan Sports Analytics Conference, hosted by the Massachusetts Institute of Technology, and handed out business cards. They got a few short write-ups in newspapers and baseball blogs, and their Twitter account grew to 137 followers (it is now 263). They outsourced a couple web designers from Nepal and Brazil to improve their simulation and construct an app. A few Major League teams slowly began showing interest, always with an eye over one shoulder. One executive agreed to meet with Jason only if they could speak at a Chipotle across the street from the conference they were attending, so nobody would notice.
DeCervo drew its name from French, de cerveau, meaning "of the brain." "There are a lot of companies that say they're doing neural," Jordan says. "They're not doing neural." He was referring to the cognitive gaming companies, most of them modeled after Lumosity, which claimed (in some cases, deceptively) to improve mental performance through an app. But Jason and Jordan did not want to claim any performance benefits from utilizing their system, or direct teams how and why they should use it. "We wanted to be the first company to measure the impact" of a decision to swing, Jason said, "and relate that mental side into performance outcomes." In essence, they were a data company-they had the means to quantify something seemingly thought to be incalculable: how and when hitters decide to swing. They thought that information could be of value to the teams who knew what to do with it. Jordan said, "I found a quote from Paul DePodesta," the former front office assistant with the "Moneyball" Oakland A's and the New York Mets, "that said 'the problem isn't with scouts or scouting. The problem is that it is based on a metric that is subjective, and not data-based.' What we're trying to do is go right into there and say, 'We're scouting purely on the stats.'"
