Big Data Baseball Read online

  The Astros, under the new leadership of general manager Jeffrey Luhnow, are now one of the most cutting-edge analytical teams and set a record for shift usage in 2014.

  By 2013, 1 percent more of batted balls were being turned into outs compared to 2007 levels across baseball. Over a season that’s 2,540 fewer base runners in baseball. Such numbers are fueling a rapid increase in the use of shifts.

  Consider the team-by-team increase in shifts from 2013 to 2014 according to Baseball Info Solutions:

  Said Baseball Info Solutions’ John Dewan, “In baseball when one team does something and they are successful, other teams try it out.… Shifting was way up last year [in 2013]. Our analytics are showing it has value.”

  Some teams, however, are still resistant to fielding more smartly aligned defenses. The Rockies had not shifted once in the first month of the 2014 season. The Phillies and Padres were also resistant to moving away from tradition and were some of the worst clubs in the sport in 2014.

  While the majority of baseball was catching up to the Rays, Brewers, Pirates, and other shift-heavy teams, advanced analytical teams were already employing the next generation of shifting. In their first series against the Cardinals in 2014, the Pirates weren’t just shifting their defense per player, they were shifting alignment per pitch. For example, against Cardinals third baseman Matt Carpenter, who is one of the best contact hitters in the league, Pirates infielders were in near-constant motion. On various counts—the number of balls and strikes on a batter during a plate appearance—they altered their alignments to Carpenter’s tendencies, overloading one side of the infield for one pitch, then balancing into a more straight-up alignment for another.

  “There’s still a lot left out there,” says John Dewan. “Even the most aggressive shifting teams are not shifting often enough, that’s the first thing. There are other elements that are very, very important. For example, the count. When the count is in favor of the hitter, almost every hitter becomes a pull hitter. Vice versa, when the count is in favor of the pitcher, almost every hitter becomes a nonpull hitter. That’s very, very important. Also, the type of pitch. The second baseman and the shortstop need to read the sign from the catcher. If it’s a fastball, it’s less likely to be pulled. It’s almost common sense, but not every player does this: read the sign and anticipate where the ball is going to go. I think the more progressive teams are starting to do some of that.”

  The Pirates also began more aggressively aligning their outfielders in 2014, with left fielder Starling Marte sometimes within thirty feet of the left-field foul line, an unusually extreme position.

  However, outfield shifts were trickier to employ. When you see data clusters of batted balls in certain areas of the infield, the data points meant that they were mostly either ground balls or short line drives: they were often hit at similar speeds and were on the ground or low to the ground. But in the outfield, optimizing fielding position was more difficult since the cluster points included batted balls with different trajectories, hang times, and speeds—with the need also to try to calculate the range of an outfielder along with the dimensions of the field to determine optimal positioning.

  “The outfield was difficult for us because where you see clusters isn’t always where you optimize fielding,” Fitzgerald said. “For balls in the outfield you might see a huge cluster in left-center field. I may see twelve balls to left-center, but the four balls down the line end up being more important [because they could be extra-base hits].”

  Pirates analysts used visuals to plot the different types of balls in play and where they suggested positioning each outfielder, and Fitzgerald said, “Visually it jumped out to each of the guys.”

  Cincinnati Reds slugger Joey Votto is one of the most cerebral and feared hitters in the game. The Pirates, however, found Votto had specific tendencies on fly balls. When Votto came to bat, the Pirates had Marte hug the left-field line as if Votto were a pull-heavy right-handed hitter.

  “Outfield shifts have made a difference. I’ve lost some hits. I’ve lost quite a few more than I’ve gained,” Votto said. “The Pirates have always shifted heavily. There’s been numerous occasions … when I’ve hit to left field, and down the left-field line … those [batted balls], which are usually hits, have been outs against the Pirates. They are taking away hits.”

  Another consequence of big data is that it has tilted the playing field toward the defense since most big data entering the game has been tied to maximizing run prevention. The proliferation and increase in shifting and enhanced scouting information has played a role in the decline of offense. Shifting, along with the increase of strikeouts, has endangered traditional batting average standards and warped the run-scoring environment. In 2013, the major league batting average fell to .253, the lowest in baseball since 1972. For decades, a .300 batting line was the gold standard of a good hitter, but that standard is now in jeopardy.

  Of course the batting average decline was tied in part to strikeouts. To isolate the effectiveness of shifts, consider batting average on balls in play. From 2006 to 2008, the major league batting average on balls in play hovered between .303 and .300, around the historical average. The number fell to .295 in 2011 and to .293 in 2013. Ten points in average might not seem significant, but that is hundreds of what would be hits against traditional defenses being turned into outs.

  The ultimate proof of runs being cut down due to big data is the scoreboard. The average number of runs scored by a team per game has declined every year since 2006, from 4.85 then to 4.2 in 2013, the lowest rate since 1992. In 2014, scoring fell again, to 4.07 runs per game, the lowest since 1981. The major league batting average has fallen every year since 2006, from .269 then to .251 in 2014. And this is with only about a quarter of the league significantly shifting, including the Pirates, who in 2013 were ahead of the curve. They, along with such teams as the Brewers and the Rays, enjoyed a considerable competitive advantage—for the moment.

  The Pirates learned early in 2013 that the shifts were working, and their success influenced the rest of the sport in 2014. But for Hurdle to keep his job, and for Huntington and Fox to remain in the front office, the Pirates would need to do more than smartly align their defenders.

  7

  ATTRITION

  In 2012 the Pirates were badly in need of quality pitching, so they sent several of their promising prospects—including Robbie Grossman and Colton Cain, who were each signed to $1 million bonuses as draft picks under Huntington—to the Houston Astros for veteran, left-handed pitcher Wandy Rodriguez. Rodriguez was under contract for two and a half more seasons. The Pirates were banking on Rodriguez to be a stabilizing force in their 2013 rotation. Other than A. J. Burnett, Rodriguez was the only Pirates pitcher to have logged 200 innings in a major league season. In the four previous seasons, Rodriguez had pitched at least 191 innings in each season.

  On a balmy afternoon at Turner Field in Atlanta, Georgia, on June 5, 2013, the Pirates’ season fell into a precarious position. On his fourteenth pitch of the game, Rodriguez threw a fastball that seemed to slip out of his hand, traveling high and inside to the left-handed hitting Freddie Freeman. The errant pitch, an uncharacteristic throw for a man who rarely misplaced his pitches, hit Freeman in the shoulder. Rodriguez motioned to the visitors’ dugout with his gloved hand, his right hand. The motion transcended any language barrier—Rodriguez was summoning the training staff. Russell Martin was the first to reach Rodriguez. As Martin reached the mound, Rodriguez motioned to his left elbow and muttered something. Pirates trainer Todd Tomczyk was second to arrive, rushing to the center of the diamond, where he was flanked by a mass of gray road jerseys as the entire Pirates infield congregated. Rodriguez again raised his left elbow. Tomczyk looked at it and consulted briefly with Rodriguez before they slowly walked off the field to the Pirates dugout. The crowd in the middle of the infield wore long, serious faces as Rodriguez was examined. Players and training staff knew it was rarely a minor issue when a pitcher c
omplained of elbow pain. It was the last time Rodriguez pitched in 2013.

  This was just one in a string of injuries that had been plaguing Pirates players recently. Three days earlier, starting pitcher Jeanmar Gomez, who had begun the year as the Pirates’ long man out of the bull pen, had to leave a game with a right forearm strain. Gomez had replaced another starting pitcher, James McDonald, who went on the disabled list May 1 with a shoulder strain. McDonald had pitched like an All-Star in the first half of 2012, but then struggled in the second half of the season along with the rest of the team. McDonald’s shoulder began to bark in late April 2013 as his velocity and performance declined. Like Rodriguez, he missed the rest of the season.

  To make matters worse, earlier in the season Jonathan Sanchez, another member of the team’s Opening Day rotation, had been so ineffective that he had been let go on April 30. In a thirty-five-day span, the team had lost 60 percent of their Opening Day rotation and their first reserve starter in Gomez.

  The Pirates were in trouble.

  They could little afford and had little control over injuries. They could design the most elegant defensive plan ever and extract more hidden value from the free agent market than any other team, and their front office and coaching staff could pull all the right levers, but none of that would matter if they lost too many key players to injury or underperformance. And beyond star Andrew McCutchen, they could least afford injuries to their starting rotation.

  This left the Pirates with little room for error. Entering the season, the club possessed the game’s fourth-lowest payroll at $66 million, and nearly one-fifth of that, tied to injured pitchers, was now dead weight. The Pirates could not take on bad contracts or sign free agent stars to white-out misfortune. When the Pirates lost a player, they often had to look internally, not externally, for help. And they needed help badly.

  The night Rodriguez got injured, the team lost to Atlanta, 5–0, their third straight loss, and they had fallen 3½ games behind St. Louis in the National League Central Division, and one game behind Cincinnati. They had surprised many observers with a strong first two months of the season, winning 35 of their first 60 games, but the injuries threatened to end the feel-good story. Perhaps they were at a high-water mark.

  Another problem of attrition was haunting the Pirates organization: the erosion of interest in Pittsburgh in its baseball club. For two decades, layer after layer of trust had broken down between the Pirates and their fans. Despite the early-season success, PNC Park stood half-empty during the first two months of the season. A sea of empty deep blue seating remained in the upper deck. Many in the city viewed those first two winning months as a fluke. Pittsburghers were teased by the previous two seasons when the Pirates towed winning records into the second half, only to see them evaporate into consecutive losing seasons number nineteen and twenty. Though the Pirates thought they were onto something new and promising with their defensive plan and free agent acquisitions, they couldn’t discuss their plans in public and thus lose their secret advantages. Even if they did broadcast their plans, since most of it wasn’t plainly observable, it would be a tough sell: an extra batted ball caught here and an extra ball turned to a strike there didn’t have the splash of signing a big-dollar slugger or an ace pitcher. The value didn’t show up on the back of a baseball card and couldn’t be sold on a billboard. The Pirates had hoped wins would boost attendance, but the fans had not come.

  A number of Pirates ownership groups throughout the last two decades hadn’t trusted that increased spending would make fans show up. This lack of trust was perhaps reinforced as attendance fell from 2 million in 1990 to nine hundred thousand in the strike-shortened 1995 season. Attendance had crept above 2 million only once in the last two decades—in the inaugural season of PNC Park in 2001. Only a last-minute funding agreement in 1998 had ensured a new stadium would be built on Pittsburgh’s North Shore and that baseball would remain in Pittsburgh. Support for the club had fallen so quickly since the breakup of the Barry Bonds–led teams of the late 1980s and early 1990s. Ownership groups were reminded again and again that Pittsburgh was a football town, and when it wasn’t a football town, it was a hockey town.

  Even though Bob Nutting had only owned the team for a handful of years and had little role in the organization’s losing streak, he was perceived as more interested in his other investments, such as the Seven Springs Mountain Resort, which became a frequently cited property on local sports-talk radio. Fans joked that since Nutting wasn’t spending money on players, he must have been spending cash on ski lifts. Pirates management spoke often of small-market limitations, but the city didn’t hear the Steelers or Penguins making similar excuses, though the NFL and NHL employed salary caps to promote parity. Fans sensed that Nutting would not spend on a big trade or any significant free agents. Even a former longtime minority owner of the club, Jay Lustig, doubted that Nutting would ever spend enough to bring winning baseball to Pittsburgh. Lustig sold his stake in the Pirates in 2012. Interviewed by the Pittsburgh Tribune-Review in April 2013, he said, “If you are a small-market franchise, if you want to win, you have to be willing to lose [money].… [Nutting’s] problem is, he is a rational owner in an irrational business. People say he is a cheap owner, but nothing is further from the truth. He allocates the money properly. He wants to make enough money to keep us from going into the red.

  “Now did Bob Nutting and I have personality clashes? Absolutely. I’ve talked to him several times and tried to convince him that now is the time to sell to a multibillionaire who is willing to come in here and spend more money and see if he can make the Pirates win. [But] Bob Nutting told me something that his grandfather told his father and his father told him. And that’s when the Nuttings own something, they own it forever. I just felt I’m too old to find out how long forever is.”

  The public’s confidence was further tested when Pirates management’s competence was questioned after revelations of their Navy SEALs training practices and after the image of a poster outside the minor league players’ spring training clubhouse in Bradenton with the phrase Embrace the Suck was leaked to the press. It was a military phrase designed to encourage subordinates to embrace their trying work. The team was urging their minor leaguers to embrace hard work as beneficial. Instead, the phrase was represented by some in the local media as a snappy, humorous slogan to sum up twenty years of losing.

  The empty seats in PNC Park felt like a boycott, expressing the erosion of trust and the desire for change. Clint Hurdle saw this broken trust when he came to Pittsburgh late in 2010. He often spoke of his mission to rebond the city with its baseball team. He had first known the Pirates as a great organization, the “We Are Family” Pirates of the 1970s that had won a pair of World Series with stars such as Dave Parker and Willie Stargell. Pittsburgh had not always been a football town. It had not always been a place where the hockey team had been the number two sport of choice. Pittsburgh had once been a baseball town. Before the Steelers were winning Super Bowls and the Penguins were winning Stanley Cups, the Pirates had won championships.

  To be an ambassador for the Pirates, to be the face of the franchise, Hurdle thought he should not be a part-time Pittsburgh resident. So he bought a home and lived year-round in the city, even through its snowy and slushy winters. He made daily stops at a Starbucks, where his name was written on his cup without a barista’s asking. You could find him at North Hills grocery stores and barbershops near his suburban home. He noticed when he was out in public that most kids wore Ben Roethlisberger Steelers jerseys and Sidney Crosby hockey sweaters. He saw far fewer kids adorned in Pirates gear. It was as if they were embarrassed by their baseball club, and he knew that had to change if he was going to stick around. This was a business. It was about selling tickets, jerseys, and compelling people to turn on their televisions. The Pirates had to win.

  What the public did not see were the data-driven, smart decisions that were making a real difference in the season’s first half. Hurdle believed i
f the Pirates could continue this play into the second half, it would be impossible to ignore, even in a town obsessed with its more successful franchises, the Steelers and the Penguins. He felt the Pirates could capture the city’s imagination if they kept winning. He also knew he could not afford another second-half slide. He already had two strikes and couldn’t afford a third. The injuries to the pitching staff were threatening to guarantee Hurdle’s third strike and a twenty-first consecutive losing season. The Pirates and Hurdle had to find a way to overcome the loss of 60 percent of their starting rotation through just the first third of the season. Such a rash of injuries would have threatened the prospects of even the game’s most talented and deep-pocketed teams.

  Baseball is often called a game of attrition. Hurdle knew it was a specific type of attrition: pitching injuries. Pitchers broke. It was nearly a scientific law in the game. The teams that won either had better luck in keeping their pitchers healthy, more pitchers in reserve, or the money to acquire additional pitchers via free agency or by taking on large contracts in trades. Los Angeles Dodgers head trainer Stan Conte was one of the first medical officials in the sport to study injury data. Conte found a starting pitcher has a 50 percent chance of being injured in any given season. A team without pitching depth is therefore highly unlikely to succeed. Moreover, starting pitchers are the game’s highest-paid players, with major league teams devoting a disproportionate share of payroll to pitchers. From 2008 to 2013, teams spent $1.3 billion on pitchers disabled by injuries, according to journalist and sports-injury expert Will Carroll. No type of injury has been more costly than ulnar ligament tears in the throwing elbow, which require a reconstruction of the ligament, a procedure known as Tommy John surgery, which that can take a year to eighteen months to rehab.