The Only Rule Is It Has to Work Read online

  “I don’t know about Rand,” I say, to see what his response will be as much as anything else.

  “Oh, I already saw him, he’s legit,” Fehlandt says.

  “But everywhere he’s pitched, he’s been bad.”

  “That all depends. He threw strikes. He made adjustments and put it where he wanted to. His presence on the mound, he knew what he needed to do to get ready. I could see where his four-seamer could get him in trouble, but if he can throw his two-seamer for strikes, it runs all over the place.”

  “At some point, though, you’d like to see he had success,” I respond. “At some point his skills would have shown themselves. Nothing so far says he’s been good at pitching. Maybe he looks like a good pitcher, maybe he is a good pitcher, maybe he hasn’t even gotten a chance, but—”

  “That’s what I’m saying: You can’t just go off numbers,” Feh counters. “You’re saying he hasn’t pitched good, but he might not have been put in a position to get a chance to do what he’s capable of.”

  “But how does he get to this point in his life without anybody else seeing this, you know? How come no better team gave him a chance, no other league gave him a chance?”

  That’s the question. That’s always the question. Why is he available to us? Finding the right “whys” is what is going to separate us from the other teams in the league.

  As it turns out, Jon Rand is a charming kid with a pretty good “why.” In high school, he threw 75 mph. He went to junior college, threw only four innings, got stronger, went to a summer league in Missouri, and got the attention of a Division I pitching coach. That pitching coach was so impressed he offered him a full scholarship to Division I Central Arkansas.

  “And in Arkansas, at least on the team I was on, you had a victory party win or lose,” he says when I meet him in person. He shows me his driver’s license, where his picture shows a kid who was at least thirty pounds heavier than he is now. He attributes his old weight to beer and McDonald’s.

  “I was more excited that I was playing D1 baseball than I was actually appreciating the opportunity,” he says.

  “You were excited to be there because it was a big deal, but you didn’t actually realize what a big deal it was,” I respond.

  “Exactly. I wasn’t prepared for how serious they took it,” he admits. “And I was mentally checked out and picking fights with coaches. My mom always put it like this: I’m a summer kid. When it came to balancing out school and baseball, I had to be doing well in both, because if I do bad in school then I do bad in baseball, and if I do bad in baseball I do bad in school.”

  Jon dropped out of Central Arkansas just short of graduating, and declared his retirement on Facebook: “I’m hanging up the cleats. Baseball never brought anything good to me.” But his junior college coach saw that post and talked him into driving down to San Jose to pitch in a men’s league. He cut the beer from his diet, got into P90x, and struck out 55 batters (and walked 2) in 29 innings for the San Jose Colt 45s. I have no idea what to do with Sunday men’s league stats, but I start to see what Fehlandt saw: This is an athlete, he’s in shape, he’s right in front of me ordering the chicken breast and iced tea instead of the bacon burger and a beer, and he’s a nice kid. He’s also kind of quirky: His entrance music, he says, is an acoustic ballad by a Swedish sister duo called First Aid Kit—a far cry from the country, hard rock, or hip-hop that most players choose.

  “I can’t believe you of all people would be picking fights with your coaches,” I tell him.

  “Never trust the nice guys,” he says. “Those are the scariest people. You never know what will set them off.”

  I can work with this “why.” At our first flashpoint, I’ve come to agree with Fehlandt, and I start to trust him. We hire Fehlandt, and our team, including Jon, is now up to five. And it has a manager.

  3

  MODERN BASEBALL

  One thing Sam and I don’t have to deal with, in our daily lives as baseball analysts, is not knowing what happened in any given baseball game. If anything, we’re overloaded with information: At the major league level, every event is almost perfectly preserved. As I write these words, I’m also working on an article about the length of the leads that runners take from first base in major league games. The data I’m drawing on, which was provided by Major League Baseball Advanced Media, tells me the average length of the primary and secondary leads each runner takes and each pitcher allows, both overall and in every individual pitcher/runner matchup, out to three decimal places—as in, Ichiro’s average distance from first base before he takes his secondary lead is 13.006 feet, or 13.032 feet against right-handed pitchers. “Down to the millimeter” would be less precise than this.

  Modern baseball analysis depends on this ever-more-detailed data: As the unexploited advantages available to teams shrink, researchers sift for still smaller ones. Sam and I are spoiled, having had an endless supply of stats at our disposal since we started at Baseball Prospectus: the speed, movement, location, and type of each pitch; the approximate angle and landing point of each batted ball; the result of every pitch and play, preserved with perfect accuracy back to before we were born. All of this information is easily retrieved, sorted, and exported, either for free or for a small fee, at websites such as Baseball Prospectus, Baseball-Reference, and FanGraphs. In seconds, we can dig up estimates of what any player was worth at any point in the past, or how he’s projected to do today, for the rest of the season, or over the next decade. We can see spray charts that display the locations of a hitter’s batted balls over any span of games, or heat maps that show where a pitcher tends to work, or results for any player on pitches in any part of the strike zone. And anything that doesn’t come precalculated can be sliced and diced in minutes, direct from a database, by someone smarter than we are who’s usually happy to help. It’s an incredibly rich resource that’s hard to appreciate until it’s taken away.

  This is more than most fans want to know, and much more than they need to know to appreciate baseball’s basic pleasures—the human drama and simple aesthetics that make us obsessed with the sport in the first place. For us, though, the game we learned to love as kids was a gateway drug to baseball’s unseen structure, where the interactions of complex forces lead to wins in ways that are often imperceptible to people who’ve been watching, playing, and even teaching the game for years.

  As the information at our fingertips expands at an exponential rate, it’s getting tougher to touch bottom, no matter how deep we dive. Lead length is one of many new measurements made possible by Statcast, a system installed in every major league park for the first time in 2015. Statcast combines a Doppler radar array that takes two thousand readings per second with a network of high-definition cameras that capture images thirty times per second, producing a three-dimensional record of every action on the field: every player’s position at every instant, as well as the speed, spin, and trajectory of every thrown and batted ball. The system makes it possible to study previously unquantifiable aspects of fielding and base running such as player positioning, speed, acceleration, and route efficiency; to identify pitchers whose stuff seems faster and nastier than its velocity would suggest because of its spin or release-point proximity to the plate; to find hitters whose stats undersell their abilities because their hard-hit balls have happened to find fielders. Each deep gulp of a game produces terabytes of data, more than generations of statisticians and official scorers generated throughout the sport’s first dozen decades.

  That gulp is only the beginning. The digestion is more difficult. Statistical studies in sports are plagued by the same problems that pop up in any field’s peer-reviewed research: errors in coding and data collection; publication bias; “statistically significant” results that are really random effects masquerading as real ones. Still, when we try to answer questions about big league games, we take it for granted that some useful information exists. That’s not a safe assumption in the Pacific Association, where the stats are abou
t twenty-five years behind the bleeding edge.

  Pacific Association records are kept by a company called Pointstreak, which handles data entry for many professional, semipro, and amateur leagues across the country. For a seasonal fee of $1,400—it should probably be more, but the Pacific Association cuts costs by buying one account for everyone instead of one per team—Pointstreak keeps track of pitch-level results for every game, compiling full-season stats that are accessible by anyone through a web portal. In theory, these results are accurate and complete. In practice, they’re imperfect.

  Pointstreak’s data is tracked not by a network of cameras, computers, and radar installations, but by a well-meaning (if not completely caring) high school kid working without pay in the unlikely event that Pointstreak experience is a prerequisite for his dream job. Sometimes in 2014, the Pointstreak stringer wouldn’t show up, or would enter the stats so carelessly that they’d be almost useless. If that high school kid mistakenly marked a called strike instead of a swinging strike, then as far as the stats are concerned the batter simply didn’t swing. The only way to fact-check Pointstreak would be to listen to an archived Internet radio broadcast, and even that might not be available. Video isn’t an option: Using my computer, I can call up any major league pitch from the past several seasons, but only one Stompers game in 2015 will be televised.

  On top of the accuracy issues, there’s the matter of how much information is missing. Although Pointstreak’s software has the capacity to record hit locations, that function is optional, and unpaid operators aren’t inclined to do more than the minimum. With no means to track them, pitch speeds and locations are lost to posterity. And while the standard seasonal stats are easily found, something as vital and simple as platoon splits—a player’s performance against righties or lefties alone—must be manually calculated with a laborious slog through the game logs. We’re not that far removed from Allan Roth, the first full-time statistician in baseball, who was hired by Branch Rickey of the Brooklyn Dodgers in 1947 (the same year Jackie Robinson debuted) to compile and dissect stats by hand.

  “Knowledge is power” applies in the Pacific Association, particularly to Sam and me. Information is our sole source of authority, the only handhold we’ll have in a dugout environment that might not be hospitable to statheads. Everyone else in the Stompers’ sphere will wear uniforms, outward indications that they’ve earned the right to be there based on baseball skill or coaching experience. But why are we here, with our non-baseball backgrounds and nonconforming clothes? We don’t get outs or drive in runs, which makes us wastes of space until proven otherwise.

  No Stomper has heard of Baseball Prospectus, so our writing résumés won’t help: In the time it takes to explain what we do, the players’ interest will ebb away. And while we have the owner’s and the general manager’s permission to be backstage, their reflected authority isn’t enough: The appeal to a higher power is the last recourse of the flustered substitute teacher, whose only source of respect is the threat of a trip to the principal. We want to be self-made Stompers who aren’t just pretending to be part of the team. And that means we have to be baseball oracles with the power to tell players things they don’t know, about both themselves and their opponents. So if we want to run the Stompers in a forward-thinking way while cementing our standing in the clubhouse, Pointstreak’s level of detail won’t do. We will have to close the information gap between the Stompers and the majors, and widen the information gap between the Stompers and their Pacific Association competitors. And we’ll have to do it fast.

  * * *

  Modernizing a throwback baseball team in next to no time and on next to no budget demands some dependence on the kindness of strangers. In our case, two of these strangers have cold, corporate names: Sportvision and Sydex.

  Sportvision has done as much as any other entity to drag baseball into the information age. In late 2006, the Bay Area company began installing PITCHf/x in major league parks. The system, which triangulates each pitch’s location to within approximately one inch, was tracking every game by 2008, and websites soon sprang up to make the data more manipulable. Before long, one could find a pitcher’s release point, average velocity, movement, and pitch-type percentages with a two-second search. Suddenly, anyone with Internet access knew almost as much—or maybe even more—about what happened after the ball left the pitcher’s hand as an experienced scout sitting in the stands with a radar gun. In later seasons, Sportvision used its cameras to roll out HITf/x and COMMANDf/x (measurements of batted-ball trajectory and pitch accuracy, respectively) before being supplanted in 2015 by Statcast’s combination of cameras and radar.

  As accustomed as Sam and I have become to instant feedback on pitcher performance, the prospect of a PITCHf/x-free season is more dismaying than any other information deficit we face. But it doesn’t take long for our grief to enter the bargaining phase: Maybe we don’t have to have a PITCHf/x-free season. Sportvision’s main office is in Fremont, only seventy-five miles from Sonoma, and I’m on good terms with the company’s staff. Research requests aren’t the same as full system installations, but we have little to lose by asking.

  In late February, I email Ryan Zander, Sportvision’s senior vice president and general manager, to tell him about our plan to take over a team in the nearby Pacific Association (which, we find out later, no one at the company has heard of). Unbeknownst to Zander, I also email someone from TrackMan, the rival company whose tech powers Statcast. Combining two long shots might make for a medium shot, and we’d be thrilled to work with either. We’re equal-opportunity moochers.

  It takes TrackMan two weeks to respond. Sportvision answers in only two hours, which gives it the inside track on our entirely unremunerative business. Zander’s promising reply spawns a multimonth megathread that periodically surfaces in everyone’s inbox, punctuated by the occasional conference call. The upshot of the extended exchange is that while Sportvision doesn’t typically hand out pitch-tracking systems to anyone who asks—the system costs thousands of dollars, enough that only well-financed teams with the potential for seven-figure insights can afford it—our proximity to Fremont, the potential for publicity, and the ability to offer our data to MLB clients (who are increasingly interested in indy leaguers) might make it worthwhile.

  Our slow Sportvision seduction culminates in a conference call with Zander and the company’s baseball analytics specialist, Graham Goldbeck, which makes the arrangement official. “We’re happy to just participate with you guys without really any sort of monetary contributions,” Zander says. “Just being a part of this project is something we’re excited about.” Jackpot. Not only will we have PITCHf/x, but we’ll also have HITf/x and COMMANDf/x—all for free. All we have to do is rent a boom lift to open our eyes in the sky: one on the first-base side, one on the third-base side, and one in center field. We also enlist the aid of Dan Brooks and Harry Pavlidis, two Baseball Prospectus colleagues who operate a consulting company called Pitch Info, which enhances and analyzes PITCHf/x and TrackMan data for major league teams. For baseball nerds, this is better than bottle service at an exclusive club, and all it costs us is some Stompers caps and T-shirts. We were worried that the transition to the Pacific Association would mean living with less ball-tracking data than we’re used to, but once Sportvision installs our system we’ll have even more.

  In Arnold Field, at least. At the other three parks, we won’t have any high-tech equipment, which means we’ll be blind in half of our games. With only seventy-eight games in the season, we’ll have to capture as many plate appearances as possible to extract any significance when we sift through the stats, so we need a road solution. It has to be something we can use from the stands, but also something we can collate quickly into reports about player tendencies. In other words, a scorebook won’t suffice.

  My intern experience is helpful here: A few times during my year with the Yankees, I was asked to fill in for another intern whose primary job was to chart games off of discs that
were mailed to the Bronx from the organization’s minor league affiliates. To do this, I used a program called BATS (Baseball Analysis and Tracking System), which would import the video and allow the user to set timestamps for each pitch, tag its type and location, and mark the outcome of each at-bat. The result would be a collection of neatly organized clips that could be viewed on command, as well as associated stats that functioned like a low-tech, less-precise PITCHf/x. Theoretically, anyway. I was the worst at this—without radar readings, I couldn’t reliably recognize pitch types, and I’d spent so little time with the software that I never knew where to click. After my first failed attempt, I tried to be busy with other tasks whenever it looked like I might draw BATS duty (or, failing that, stall until the other intern returned). But I knew BATS would do what we wanted if we could wheedle a license.

  Sydex, which distributes BATS, claims that it’s used by twenty-nine of the thirty major league organizations. Because the company has virtually cornered the market on charting baseball games via video—an extremely specialized task—it can charge thousands of dollars per license, which would be beyond the Stompers’ means. But when I send an exploratory email, I get the same response I received from Sportvision: Cool project. How can we help? A couple of calls later, BATS is installed on my laptop, which brings back traumatic memories but also fills one of our most pressing needs. We won’t be blowing baseball’s collective mind—as innovative as it is in the indy leagues, BATS is old hat in affiliated ball—but we need to catch up on the basics before we can break new ground.

  We also need help from humans. To help us straighten up our sloppy Pointstreak stats, we draft John Choiniere, a thirty-year-old research scientist in the Department of Pharmaceutics at the University of Washington. We’ve never met John, nor even talked to him, but he’s been a fixture in our inboxes for almost two years. A regular listener to our podcast, John emailed us out of the blue in 2013 with an offer to track the bets, predictions, and drafts that the two of us make in the course of the show. We were relieved to have an official podcast scorekeeper, especially once John turned out to be meticulous, pouring far more effort into chronicling our competitions than we did in preparing for them.