The MVP Machine

by Ben Lindbergh

  Bannister makes a medical analogy to describe how the front office and field staff function. The analysts upstairs are the radiologists who dissect the data and relay it to the surgeons (coaches), who operate on the patients (players). Bannister, who’s inhabited all three roles, likens altering players without using technology to performing surgery without ordering an MRI.

  To stick with the surgical theme: years passed between Joseph Lister’s nineteenth-century discoveries about the infection-fighting effects of sterile surgery and the advent of universal, routine handwashing in the operating room. In the time it took for Lister’s ideas to be adopted, lives were needlessly lost. Baseball decisions aren’t life or death, but teams are still trying to reduce that delay. In the absence of better information, experience and a history of trial and error were competitive advantages. Now they can be constraints, unless they’re accompanied by a healthy humility. “People upstairs are learning faster than you can ever learn, and so you can’t fight it,” Bannister says. “You have to be the conduit.”

  Coaches who can stomach that status get something back in the bargain: they don’t have to worry that doing their jobs will cost them their jobs. As a former No. 1 pick, Floyd knows better than most that tampering with top prospects has historically been a career-jeopardizing proposition. “Most pitching coaches are fearful that if they tweak anything and they hurt you or make you even less effective, then all the eyeballs come looking at them,” he says. “I saw that a lot.” The consequence of cover-your-ass coaching was that players with the most talent received the least instruction. Today, the wisdom of any tweak a coach wants to make can be tested and backed by data. That doesn’t guarantee the tweak will work, but it ensures the coach won’t be crucified if something goes south.

  Naturally, technology has changed the type of tweaks that teams tend to make. In the 2011 interview where he verbally brushed back Bannister, McClure said, “You don’t have to have great stuff in order to be a good pitcher. You just need good command and good feel.” Bannister’s research has led him toward a different emphasis. “I’d rather have a pitcher with a 70 or 80 [pitch on the 20–80 scouting scale] that really has no idea where it’s going than a pitcher with a 50 pitch that can throw it exactly where he wants to every time,” he says. “I’ll always err on the side of pitch quality.”

  A tweak as subtle as reorienting a slider’s spin axis by 15 percent may translate to more efficient movement, more whiffs, and the difference between foul after foul and a true putaway pitch that can replace a weaker weapon. “It’s just almost like Christmas when you see a guy with all the right physical characteristics to throw a 70 or 80 pitch, and then you realize he’s gripping the ball wrong, and you’re just like, yes!” Bannister says. “That’s the final puzzle piece, that he doesn’t know that that’s making his pitch a 50 or a 60 instead of a 70 or 80.”

  There’s an art to that insight, but it’s nothing ineffable. “Pitching is not mysterious, it’s just physics,” Bannister says. “[We’re] trying to get away from, ‘This pitcher has this quality that we can’t teach anybody else.’… It’s not magic. He’s just doing something better than everybody else.” Once a team deciphers that “something” and identifies the components of pitch quality, it can automate much of the process of finding players who are ripe for repackaging. A program combining machine learning and basic artificial intelligence can comb through data on mechanics and pitch characteristics and flag anything that seems suboptimal: a great pitch that’s barely being thrown, or an underperforming pitch with high spin but poor spin efficiency, which Bannister compares to a car with a powerful engine but bald tires. Then the pit crew of coaches and communicators can come in and devise the best strategy for fixing the flaw.

  In his current role, Bannister says, “99 percent of the work I do is standing out in the outfield during batting practice or during a bullpen session, holding my phone and showing it to a pitcher, showing him what the data says and then telling him why I think he should make an adjustment and backing it up on the spot.” Before Bannister, no one on the Red Sox staff needed to deploy data that way. Boston’s old internal information repository dated back to before the iPhone-iPad era, and it wouldn’t work on mobile devices. After joining the Red Sox, Bannister learned SQL (Structured Query Language)—a programming language that’s become a prerequisite for front-office work—to retrieve information from the database more quickly, but the queries he created weren’t powerful enough for his purposes.

  Enter a new application called PEDRO (pitching, evaluation, development, research, and optimization), a nod to Red Sox legend and onetime Bannister teammate Pedro Martínez. PEDRO, which was built by R&D analyst Spencer Bingol, a former baseball blogger hired by Bannister, functions as a “sandbox of ideas on the player-development and scouting side.” It also enables Bannister to do with one click what once would have taken him hours or days, applying his custom pitcher evaluations on a “mass scale” and allowing him to exchange computer time for face time with players.

  Much of that time also takes place in the company of cameras. Bannister’s affinity for photography isn’t just a rich source of analogies. It’s also a skill he draws on daily as he works with a suite of optical tools that includes Edgertronic, Rapsodo, and KinaTrax, a six-figure markerless, long-distance motion-capture system whose eight to sixteen high-speed cameras mounted in the stands from the first-base side to the third-base side record the movements of twenty-five joints on athletes’ bodies and enable its clients (including four MLB teams in 2018) to dissect the mechanics of players who appear in their parks.

  Bannister has used that technology to satisfy some lingering curiosity about his own career. Although he didn’t throw hard, he knew there was more to his fastball’s ineffectiveness than velocity alone. After diagnosing himself the same way he diagnoses new pitchers, the mystery was solved: his fastball suffered from low spin efficiency. “Knowing what I know now, I would never have pitched the way I did,” he says. Instead of compensating with command and his cutter, the only measures he could take at the time, he “would have completely redesigned my biomechanics and how my arm worked.”

  Bannister isn’t crying over spilt spin efficiency. “You only get one short career, and we just didn’t have the data available,” he says. The latest technology arrived too late to help him, but not too late for him to help others.

  During games, Bannister typically sits in the clubhouse, consulting Statcast to monitor pitch selection, spin, and speed changes that might indicate injuries. If a pitcher encounters command problems, Bannister checks KinaTrax for mechanical misalignments, and when other issues arise, pitchers (or pitching coach Dana LeVangie) walk through the tunnel to ask for advice. But much of the magic occurs before games, when Bannister turns the outfield or the bullpen into an interactive pitching lab, setting up cameras and tracking devices and inviting pitchers to ask each other questions, share information, and experiment in a more focused way than they can while playing catch casually. “I’ve found that kind of coffee-shop environment builds a lot of culture, because you get pitchers talking about their craft more openly,” he says, adding, “There’s all kinds of little miracles that happen with guys that pump their game up a little bit.”

  One of the most obvious outward manifestations of Bannister’s influence is the rate at which the Red Sox throw their four-seamers in the upper third of the zone or above. Shortly after bringing Bannister on, the Sox became the kings of the elevated four-seamer. “It was a fun way to take an analytical concept and exploit it and surprise teams,” he says.

  After a twentieth-place showing in park-adjusted ERA in 2015, Boston’s MLB ranks climbed to fifth, then third, then second from 2016 to 2018. But any advantage the Sox derive from their onslaught of high heat won’t last long. Bannister believes Boston is about two years ahead of public awareness on analytical topics, but no team ever builds a big lead on its most progressive rivals. Tactics that work well soon inspire copycats, and tea
ms that get lapped by the field resort to a policy of “If you can’t beat ’em, hire ’em.” In late 2016, the once defiantly traditional Diamondbacks hired Hazen as their new GM; Hazen, in turn, brought Sox executive Amiel Sawdaye with him as an assistant GM and also hired Boston bench coach Torey Lovullo as Arizona’s manager. Not long after, a newly slider-reliant Corbin became the club’s ace.

  PERCENT OF FOUR-SEAMERS THROWN HIGH, 2015–2018

  Year: 2015

  MLB: 40.0

  Red Sox: 39.4

  MLB Rank: 16

  Year: 2016

  MLB: 39.9

  Red Sox: 48.6

  MLB Rank: 2

  Year: 2017

  MLB: 45.3

  Red Sox: 59.1

  MLB Rank: 1

  Year: 2018

  MLB: 45.9

  Red Sox: 59.1

  MLB Rank: 1

  To cope with that cutthroat environment, Bannister takes a cue from the stock market, which he watches closely. “I can’t beat the league long term, but if I find an idea, I can beat it in the short term, and that’s all I’m looking to do,” he says. By the time a team could test a good, data-driven idea at one affiliate or with a small group of minor-league pitchers, assess the results, and get the green light from leadership, the advantage would have dissipated. Bannister believes it’s better to employ people who’ll go rogue but not reckless, Jack Bauers of baseball-skill acquisition who can cut the red tape, take ideas directly to the big leaguers who are most likely to benefit, monitor the effects, and expand the approach to the rest of the roster as confidence in its efficacy increases.

  The second he thinks he’s beaten other teams to a developmental tactic, the clock starts counting down to the day it slips away. “You always have to come up with the next thing, and then it’s like, ‘Aw, we could have been way ahead of everybody if we just did this last year, but I didn’t think about it that way.’ It’s a never-ending cycle. The rabbit hole goes very deep.”

  Boston’s conduit count is now up to two. While Bannister travels with the big club, one of his past pitching opponents, Dave Bush, serves as his counterpart in the minor leagues. The more modern development becomes a way of big-league life, the more minor leaguers want to mimic it. “They want to look like a big leaguer, they want to act like a big leaguer,” Bush says. “They’re gonna want to use data like a big leaguer too.” Especially if the people telling them to have been big leaguers themselves.

  Given Bannister’s concerns about the league catching up, it’s not surprising that other clubs have hurried to hire their own conduits. At least eight non-Boston teams now employ cerebral, recently retired players in positions that task them with funneling information back and forth between the front office and the field. With the exception of Haren, who was one of the best pitchers in baseball for a few years in his twenties, there are no former all-stars in the group. Most of the hitters had short careers as part-time players, and most of the pitchers threw fastballs that sat below 90 mph: eventually even Haren morphed into a homer-prone finesse pitcher in his thirties and now tweets from the handle @ithrow88. To a man, they all had to think about what they were doing rather than relying on pure power and speed. “There’s definitely a class of players that ha[s] to work very hard and be very good at other parts of their game in order to stick around,” Bush says. “And I’m not surprised that those are the kind of guys that are gravitating toward this type of job.”

  It also helps that, Haren and the Rangers’ Brandon McCarthy aside, none of the players made bank by big-league standards, which makes them more willing to keep working. “I’m sure a lot of it is correlated to what your bank account looks like,” says former outfielder Sam Fuld, whose account was boosted by about $7 million over an eight-year MLB career that ended in 2015. Fuld joined the Phillies in 2018 as major-league player information coordinator, a nebulous label for a broad role. “I’m trying to figure out as simple a question as, am I a coach or a front-office member?” he told us in May. The answer is obviously “both.” Some days Fuld spends a few hours upstairs and then changes into uniform like Clark Kent and dashes down to the field. On different days, it’s the other way around.

  When Fuld was five, his father gave him The Complete Handbook of Baseball, an annual tome stuffed with stat lines that he studied obsessively. After majoring in economics at Stanford and being drafted by the Cubs in the tenth round in 2004, he discovered Moneyball in 2005, his first professional season. “That was when I got so much confirmation of these things that I had thought about,” he says. “It felt really comfortable and cool to know that there were other people out there in the world that thought like that.”

  As a former fringe player who had to scrap for roster spots, Fuld understands the thought process of a middling major leaguer who’s afraid to jeopardize a big-league gig by making a change that could help him in the long run. “‘I’m making seven figures. I don’t want to make $50,000 at Triple-A next year.’ How can you blame a guy?” he says. One of a conduit’s duties is trying to coax players past that sticking point. In spring training and at the All-Star break, Fuld joins manager Gabe Kapler, GM Matt Klentak, and other coaches and front-office members in individual, data-driven meetings with every player, designed to reinforce what’s working well and offer concrete goals for improvement. “It’s almost emotional support, knowing there’s somebody there to explain things to you,” he says. Over the 2018–19 offseason, the Phillies hired two more recently retired players, Ed Lucas and Rob Segedin, as player information assistants for the minor leagues.

  Kapler, a conduit himself, is a former outfielder and farm director. Exposure to so many aspects of progressive development makes conduits sought-after manager material. Gone are the days when much of the minor-league system was outside the skipper’s purview; a modern manager is expected to be plugged into player development. “When I got called up [by the Cubs], [Manager] Lou Piniella had no idea who I was,” Fuld recalls. “He was like, ‘Can you play left field?’ [I thought], ‘Sure, but why don’t you know this?’” In late 2018, Fuld interviewed for the Blue Jays managerial job and reportedly impressed, but out of satisfaction with his current role and a reluctance to relocate his family, he withdrew from the running.

  Advanced development, and its conduit accomplices, are even starting to spread overseas. In Nippon Professional Baseball, Japan’s major leagues and the second-highest-level league in the world, conduits have come to the Tohoku Rakuten Golden Eagles. Rakuten was the first NPB team to develop its own internal database, installed TrackMan in 2014, tested a Statcast-like system in 2018, and has acted aggressively in enlarging its strategy office to ten members. One of those members is Norihito Kaneto, a lefty reliever who retired in 2017 after a ten-year career, the last half of which he spent with the Eagles.

  Kaneto wasn’t stat-savvy as a player, but since he joined the strategy office, he says (with director of team strategy Shingo Murata acting as interpreter), “my perspective has changed 180 degrees.” Murata adds, “He regrets that he didn’t really use [data] before. But now he doesn’t want the younger guys to make the same mistakes. He wants to let them know that there are all kinds of resources available to them. I call him the sales guy of the strategy office.”

  In NPB, player development may be even more vital than it is in the majors. Players don’t reach free agency until after their eighth year in the twelve-team league, which restricts player movement, and organizations control far few players. Aside from several draft picks per year and some shuffling of each club’s complement of (at most) four foreign players, rosters are mostly static, so teams have to enhance what they have.

  Japanese culture’s hierarchical relationship between senpai (elders) and kōhai (junior colleagues) promotes deference to seniority, which makes it harder for young players to buck tradition. “Whatever coaches say is pretty absolute,” Murata explains. It’s crucial, then, that some of Rakuten’s coaches are saying the same things as Murata. It’s not just Kaneto, who�
��s insisted that players meet with Murata personally in the locker room, but also Tatsuya Shiokawa, a former Eagles infielder who retired in 2011 and worked in the strategy office before returning to the clubhouse as a strategy/infield coach. Twenty-eight-year-old Eagles ace Takahiro Norimoto, who led Pacific League pitchers in WAR in 2018, was the first player to meet with the analytics team and adopt TrackMan as a tool to experiment with his pitches. “His goal is to throw one pitch that confounds us in each game,” Murata says. Eagles closer Frank Herrmann, a former major leaguer who gained an appreciation for the finer points of player development while in the Indians’ system, says the Eagles “have a blueprint” and are “taking the proper steps.” He credits Murata with helping him trust his slider, which he threw almost five times as often in his second season abroad as he did as an NPB rookie.

  Dr. Tsutomu Jinji, a biomechanics expert who worked for Rakuten, left the team in 2016 to join a company called Next Base, the Driveline of Japan. Founded in 2014 by IT executive and former varsity player Shinichi Nakao, Next Base offers a web application called BACS (Baseball Analytics and Coaching System) that recommends adjustments based on TrackMan data. The company also conducts motion analysis, imaging almost 70 pro pitchers so far. Jinji, who’s now Next Base’s director, says the company’s philosophy is that “tracking data is useful for the development of players, so the players themselves should use it.” Although Jinji acknowledges that few NPB players do, he’s worked closely with one of Japan’s top pitchers, Yusei Kikuchi. Kikuchi, a Seibu Lions starter who signed with the Mariners after the 2018 season, sought Next Base’s help with pitch design, pitch selection, and tunneling in preparation for his move to MLB.