Big Data Baseball Read online

  “Just sitting in the video room in spring training and having players come in and out, that’s an opportunity when you’re able to say, ‘Hey, how do you approach this particular aspect?’ Players are usually more than happy to share the information they have,” Fox said. “Through some of those conversations in the spring, we provided some different information as part of the scouting reports and [game strategy] recommendations on other teams.”

  Hurdle knew there would be plenty of questions beginning with Why that spring; that’s in part why he wanted Fox and Fitzgerald around. He couldn’t afford a revolt against the plan by his players. Not now. This was a last chance.

  Hurdle occasionally thought of the story of Hernán Cortés, the Spanish conquistador, who upon reaching the shores of the New World burned his ships to keep his men motivated. He gave them no alternative but to go forward. That was one way to avoid a mutiny: eliminate an option. Hurdle wanted his players to understand why they were doing this, and that’s why the presentations to the players that spring were so methodically planned. But he also realized the need for other motivations so there would be no going back to the team’s old, unsuccessful ways. The pitchers could buy in and throw where they were directed—more on this concept in upcoming chapters—or they could watch their ERAs and hit totals inflate as they went against the plan and opposing hitters took advantage of the often-undefended opposite field. The Pirates infielders understood that if they resisted the shift, then they would not only be going against a mountain of convincing data but be giving up even more hits since the pitching staff was adopting a specific philosophy. Hurdle was burning his ships the best he could.

  6

  SHIFTING THE PLAYING FIELD

  Spring training is a time to renew hope. Maybe this player will bounce back, maybe this prospect will shine, maybe this will be the year. But for twenty consecutive years, building hope and optimism as a Pirates fan was a fool’s endeavor. At the end of March 2013 the Pirates’ spring training ended, and they left the sunshine and Gulf breeze of Bradenton for the unpredictable weather of Pittsburgh. On April 1, Opening Day, there were snow flurries. The winterlike weather was in a stark contrast to the warmth of spring training and helped hammer home that the hope and goodwill of March was over. The weather gods seemed to be saying, “Welcome back to reality.” Now was the time for doing. This was the beginning of the last chance for many in the organization.

  None of the Pirates, from management to the players, had much sense of how this season was going to play out. Would their radical plan work? No other franchise in the league had ever turned around its fortunes based almost entirely upon a run-prevention plan like the one they were going to implement. The team knew that they were going to be a guinea pig. The playing field had been financially tilted against small-market clubs such as the Pirates for years. They couldn’t afford evaluation mistakes, and they couldn’t attract stars like the Yankees, Red Sox, or Dodgers. Instead, the Pirates were literally going to attempt to shift the playing field in their favor.

  Clint Hurdle’s favorite time during game day is thirty minutes before the first pitch. He likes to come out to the dugout and sit there alone. He likes that the field is in immaculate condition, freshly lined and watered. He likes to watch the crowd file in, bubbling with energy. He likes that anything is possible. There’s always a different energy on Opening Day. It was one of the few days you could count on a sellout crowd in Pittsburgh. It was one of the few days the city did not associate a negative feeling with its baseball club. There’s anticipation before the first pitch, a joyfulness, a release for Pittsburghers, who were finally enjoying an outdoor activity after enduring western Pennsylvania’s typically harsh winter. In these pregame minutes, music thumps from the sound system, and the aroma of grilled meats wafts about. The beer vendors make their rounds and calls as the crowd settles in. Hurdle soaked it all in on this frosty day and tried not to think about it potentially being his last Opening Day as a manager.

  Hours before the first pitch, Nick Leyva met with the infielders in the cramped video room. There Leyva showed the players where they should position themselves against the Cubs hitters. They focused on the projected starting lineup and he supported the positioning with data. Each player was given the same printouts that Leyva had prior to the series. He showed them the Cubs regulars’ spray chart versus the Pirates’ Opening Day starter, A. J. Burnett, and right-handers similar to him. A particularly dramatic spray chart was that of Cubs first baseman Anthony Rizzo, a left-handed pull hitter. On the left side of the infield was nothing, a few stray lines. On the right side was something resembling the NBC peacock logo, a variety of warm and cool colors expanding out from home plate, color-coded to denote where ground balls were most concentrated. Any ball Rizzo hit on the ground was incredibly likely to go to the right side of the infield.

  After the meeting, the players went back to their lockers to grab their gloves and bats for batting practice. Now Hurdle and Leyva would have to wait several hours before the first pitch to find out whether the players would actually follow the alignments Leyva had mapped out.

  The home dugout at PNC Park, along the third-base line, is about four feet below the playing surface, with several steps leading up to the field. To watch the experiment play out during the game, Hurdle took his usual spot in the Pirates dugout. A three-foot-tall nylon net rises at field level around the dugout to protect the players and coaches from line drives. The netting also makes it difficult to see the field from the dugout bench, so Hurdle occupied the top step at the corner of the dugout nearest home plate. He usually stood with his folded arms resting on the slim layer of green padding that rests atop the rail supporting the netting. He could get a better view of the field there. He would take that position and chomp on bubble gum for seemingly the entire length of a standard three-hour game. He chews so often a Twitter account with the handle #ClintHurdlesGum has been created. Gum chewing is his iconic in-game characteristic, but on that Opening Day in 2013, perhaps he chomped with more nervous vigor as he stood and wondered, Will they move? Would his infielders accept the alignment plan?

  In the top of the first inning, Hurdle watched his infielders as the public address announcer introduced the first batter of the season, the left-handed David DeJesus. Would they shift? Would they move? Would there be a mutiny? When shortstop Clint Barmes began moving to his left, Hurdle could exhale.

  Barmes took the first step toward second base and Walker followed his lead, swinging closer to the right-field line and moving deeper into right field. Barmes slid into place just to the left of second base. While they were not in a Ted Williams Shift—when three infielders are placed to one side of second base—they had significantly altered their alignment plan. But they couldn’t shift high enough. Two batters later, Rizzo launched a towering fly ball off Burnett that went over the center-field wall for a home run, also scoring Starlin Castro, who had reached on a single. The Cubs’ early 2–0 lead was all the support Cubs starter Jeff Samardzija needed to cruise to a 3–1 victory.

  Fans muttered, “Here we go again,” as they filed out of PNC Park on the cold, overcast day. Despite the loss, what was most important was the Pirates’ infielders had shifted, they had followed Leyva’s direction, and they were giving the plan a chance. Now all his team needed was some evidence, some positive rewards, for their unusual alignment movements. They would soon have it.

  After a day off, the Pirates returned to play on Wednesday, April 3, another cool night. The hitters must have felt as if they were swinging at rocks with their numb and frozen hands. Pirates starter Wandy Rodriguez had not allowed a run in over 6 innings, and McCutchen and Starling Marte had produced RBI hits to give the Pirates a 3–0 lead going into the top of the ninth inning. The stadium was half-empty, and those that were still witnesses in the grandstand seemed to be frozen to their seats when Rizzo came to the plate to lead off the inning against Pirates first-year closer Jason Grilli. Barmes again moved to the right of sec
ond base, while Walker plotted himself twenty feet to the right of where he would traditionally be stationed, then moved several steps back deeper into the right-field grass. Walker was positioned perfectly.

  Rizzo blistered a one-hop ground ball just to the left of Walker. Walker snagged the ball and threw to first for the out. It was just one play of thousands in that season, but it delivered the first positive reinforcement for the shifted infield. Had Walker been in his traditional position, Rizzo would have reached first base on an infield hit.

  “There are certain times you feel like you’re in no-man’s-land as a middle infielder,” Walker said of shifts. “You’re in certain spots that when you catch the ball, you kind of feel like you don’t know where you’re at, and that’s kind of a strange feeling.”

  Sure, teams had been shifting on select lefty power hitters for a number of years, but that season the Pirates began shifting on nearly every lefty hitter to different degrees, and many right-handed hitters. Against every hitter, at least some slight variation from the traditional defensive positioning was made. And in the season’s first month, a mountain of evidence showed that the shifting was turning hits into outs.

  “There were more instances [of shifts working] than you care to count,” Hurdle said. “There were two dozen of those that were electric eye-openers early in the season where the ball is hit and the initial thought is ‘There’s no way it’s an out,’ but then you have a guy right there. I just continue to remind them I played in an era where a hard-hit ground ball up the middle was a base hit nine out of ten times. Now a hard-hit ground ball up the middle might be a hit two out of ten times. The game is changing. Sometimes you need to change with the game or it will pass you by.”

  On April 10 at Arizona, Barmes was positioned deep in the six-hole, closer to third base than the usual shortstop position, and Walker was shifted to nearly behind second base. The Pirates weren’t in a Ted Williams Shift by definition—where three infielders are to one side of second base—but it was still a shift away from traditional base defensive lineup. The Pirates were tailoring defensive positioning to each batter. The infielders always moved prior to the beginning of an at bat. The defense never looked the same for any two opposing batters. Arizona slugger Paul Goldschmidt smashed a ground ball into the hole, which would traditionally have been a base hit, but Barmes was there to field the ball and throw across diamond.

  “There was talk between infielders: ‘Why are we giving a right-hander the four-hole [between the traditional alignments of the first and second basemen] and the whole right side? Why do we have Walker stand straight up the middle in situations where they can drive runs in? That was a tough one for all of us to swallow at times,” Barmes said. “They are big league hitters, too. They can jam themselves and hit a ball weakly through the four-hole and it scores a run from second. Why are we giving them that?”

  Still, the evidence for why to do it grew.

  Early in the season against the Detroit Tigers, left-handed hitter Don Kelly bounced a sharp ground ball up the middle of the infield, which for the first hundred-plus years of the game’s history would have rolled into center field for a single. Instead, Barmes was there to gobble up the ground ball, step on second for one force out, and throw to first to complete the double play. Through experiences like that the team strengthened its resolve to enforce the new game plan. But it wasn’t just those experiences that strengthened the belief.

  One white-haired coach on staff was a bit different from the others and because of that acted as something of a liaison between the analytics team and the rest of the staff. Dave Jauss, fifty-five, with piercing blue eyes and gray stubble, looked like the other coaches. But he had never reached the major leagues. The son of a Chicago Tribute sportswriter, Jauss had graduated Amherst with a degree in psychology and later earned his master’s in sports management. He had captained the baseball team at Amherst and became a longtime coach and minor league evaluator under Dan Duquette with the Montreal Expos, where he met Huntington. Jauss suggested to Fitzgerald that atop the data-based scouting reports given to players prior to each series should be an anecdotal example of the data’s working in a recent game. This important communication, a positive reinforcement, helped build acceptance.

  Not only were the players beginning to believe more and more through circumstantial evidence; the coaches, schooled in twentieth-century baseball thought, were starting to believe as well. Said the sixty-year-old Leyva, whose coaching career began with rookie-level Johnson City of the Appalachian League in 1978, to the Pittsburgh Tribune-Review, “When I first came over to the Pirates, you could consider me as an old-school guy. But numbers don’t lie. Halfway through 2012, I was probably using maybe fifty, sixty percent of what I was getting from stat guys last year. [In 2013] I was close to one hundred percent.”

  The Pirates would increase their use of shifts by nearly 500 percent in 2013. For baseball, being such a conservative game with a culture that is often slow to accept change, this was a stunning pivot in strategy. In the first year under Hurdle, in 2011, the Pirates shifted 87 times. In 2012, the Pirates shifted only slightly more often, 105 times. In 2013, the Pirates shifted 494 times. This remarkable increase was a previously unheard-of one-year change in philosophy.

  Just a few years earlier, only the Brewers and the Rays were shifting more than 100 times per season. The Pirates advanced from being in the bottom third in baseball in shift usage to sixth in 2013, and that helped them to a winning first month. The Pirates were 15-12 at the end of April. They had not fallen out of the gate. The infielders had seen more examples than not of shifts turning hits into outs, and their acceptance was growing. Pirates players had paid close attention to the anecdotal results in the first month of the season. More wins and more anecdotal evidence of the shifts’ success piled up in May, another winning month for the Pirates.

  The Pirates’ aggregate improvement was from a collection of individual defenders who were more efficiently converting batted balls to outs. According to BIS, Pirates second baseman Neil Walker improved from -4 defensive runs saved in 2012 to +9 in 2013, a 13-run improvement equal to 1.3 wins. The dramatic improvement was tied to simply getting to more balls in play. Walker made 32 more plays in zones outside traditional areas played by second basemen in 2013 than in 2012. Being positioned more intelligently, Pirates third baseman Pedro Alvarez improved from costing the Pirates 5 runs in 2012 to saving them 3 runs in 2013 and was involved in 71 more defensive plays despite the same amount of playing time. Pirates first baseman Garrett Jones was worth -5 defensive runs saved in 2012, then improved to a league average value in 2013, or 0 defensive runs saved. Despite playing in 400 fewer innings due to injury and the emergence of rookie Jordy Mercer, shortstop Clint Barmes’s out-of-zone plays remained the same, as did his total defensive runs saved—telling of more efficient defensive play.

  The Pirates also altered outfielders’ positioning, though less dramatically. Pirates center fielder Andrew McCutchen improved from costing the Pirates 5 runs in 2012 to saving 7 runs in 2013. It all added up to more wins with the same players.

  “I think [shifts] are something that’s going to be universally kind of implemented in the game of baseball,” Walker said prophetically after the 2013 season. “But if you ask me, it’s going to take some time, because most people don’t want to give in to it.’’

  It took much less time for baseball to evolve than even Walker could imagine.

  Two thousand fourteen marked the year of the shift in baseball. The use of shifts skyrocketed across baseball in part because clubs witnessed the one-year jump in shift usage—and resulting defensive improvement—by the Pirates. Baseball saw 7,461 total shifts in 2013, according to BIS, but by July 1, 2014, halfway through that year’s season, there had been 8,800 total shifts. According to BIS, teams shifted 13,294 times in 2014, a threefold increase from 2012’s 4,577. More and more teams were signing on to the concept, seeing how they could cut their runs allowed without adding a
dollar to payroll.

  The St. Louis Cardinals had employed little shifting in 2013, with Cardinals manager Mike Matheny citing his starting pitchers’ discomfort with them. However, the Cardinals announced plans to gradually increase shifting at the major league.

  “We have shifted and I think that’s going to continue to happen as people get more information,” Matheny said. “Statistics might not tell the whole story, but they don’t lie.”

  In 2013, shift-heavy Tampa allowed 230 fewer hits than traditional-minded, anti-shifting Colorado. Shift-heavy Milwaukee allowed 154 fewer hits than shift-resistant Philadelphia. That’s roughly just 1 hit per game, a seemingly trivial number; however, hits add up to runs, and runs add up to wins, and the industry has taken more and more notice of this. After the 2013 season, the Reds replaced the anti-sabermetrics Dusty Baker with Bryan Price, a manager more open to shifts and analytics.