Phils getting into the numbers game
Roberto Hernandez lugged a 5.00 ERA into the middle of last August, and that is when Tampa Bay demoted the righthander to their bullpen. The Rays signed the man formerly known as Fausto Carmona for $3.25 million with the hope he could recapture his effectiveness. He did not.
There is nothing extraordinary about Hernandez, a 33-year-old whose English seems limited to the phrase, "I need to keep the ball down." The Phillies, however, offered him a raise - $4.5 million - to be the fifth starter. This move did not excite fans; Hernandez's 5.03 ERA since 2008 is the second highest for any pitcher with at least 800 innings.
On a Wednesday afternoon in December, general manager Ruben Amaro Jr. explained his reasoning in the middle of the Phillies clubhouse. A TV reporter baited him: "Sign any more old guys?" Amaro eluded the topic.
"We think Roberto Hernandez will help us," Amaro said. "Our scouts and our analytics people looked at the middle-of-the-road, back-end starters and we felt like he would be a good choice for us."
This, perhaps the first public instance of Amaro's citing advanced metrics to justify an acquisition, was not a watershed moment. They applied that data, compared it to scouting reports filed during the season, and identified Hernandez as a target. This is a practice adopted by many front offices. Some value analytics more than observations. Others don't.
Most employ some sort of mix. Until now, the Phillies did not.
"We have to try to look to improve," Amaro said this spring. "There are things we are missing and ways we can get better in certain areas. We owe it to the organization to do that."
This increased attention should not be mistaken for a full plunge into advanced metrics. The Phillies, Amaro said, are at least willing to think in different ways. They will remain one of the game's more conservative front offices. But the general manager said this curiosity about analytics is not a hollow gesture.
"The whole analytics thing is just another support for the information we are getting from our scouts," Amaro said. "Again, I still believe in our scouts as the most important evaluators. But we also have to look at other areas to give us support. We would do ourselves an injustice if we don't try to add some of the analytics to the decision-making process."
The commissioner's office loaned Scott Freedman, who worked in Major League Baseball's labor relations department, to the Phillies last November on a quasi-internship. Two months later, the Phillies hired Freedman to a full-time position as manager of baseball analytics. They permitted him to add an analytics intern, and the position was advertised on sabermetric-friendly websites such as Baseball Prospectus and FanGraphs. Responsibilities: "Design, test, implement, and maintain proprietary tools and metrics."
Freedman is not a decision-maker in the front office. He will support Amaro's lieutenants. He had a seat at the team's war-room table during the winter meetings in Lake Buena Vista, Fla.
The Phillies declined to make Freedman available for an interview. He spoke at length with a fan blog called The Good Phight in February, but revealed little about how the Phillies will implement advanced metrics.
"Much of what I've been doing is ad-hoc analysis for decision-making," Freedman said. "I shouldn't discuss specifically what these analyses have been, but they've been related to offseason decisions and general strategy."
Amaro said the club's first task is sifting through what metrics are useful and those that are, as he called them, "minutiae."
"I just think it's important for us to get all of the information," Amaro said. "There are a couple of reasons. It could give us an advantage as to how other clubs are valuing players."
He used Hernandez as an example. The Phillies liked Hernandez's 53.2 percent ground-ball rate, the seventh highest in the majors last season. They examined his propensity for allowing home runs - 24 of his 115 fly balls permitted landed as homers - and pegged him as a candidate for regression. That 20.9 percent rate was the highest for any starter since at least 2002, according to FanGraphs. (The major-league average was 10.5 percent.) They believe Hernandez will be luckier in 2014.
Brad Lincoln, a hard-throwing reliever acquired from Toronto for a backup catcher and minor-league pitcher, was another.
"Years past," Amaro said, "we looked at some of the analytics as far as he is concerned."
The Phillies saw a pitcher misused as a starter and one who rolled grounders in shorter stints as a bullpen arm.
The most visible application of analytics this season could be in how the Phillies shift their infield for certain hitters. Manager Ryne Sandberg said he approved of the idea, and he employed it against hitters such as Freddie Freeman, Mark Teixeira, and Chris Davis this spring. The Phillies will use information gleaned from video analysis, spray charts, and advance scouting to formulate their shifts.
This, of course, is no new phenomenon. For years, teams have shifted against Ryan Howard because he is predisposed to pull the ball. But the trend gained traction last season; teams shifted 8,134 times last season, according to Baseball Info Solutions data. That was up from 4,577 shifts in 2012 and 2,358 in 2011.
Amaro said the Phillies discussed the idea this winter. They proposed it to Sandberg and Larry Bowa, the team's bench coach and infield coordinator. Bowa, 68, "seemed open to it," Amaro said. Clint Hurdle, Pittsburgh's 56-year-old manager, said analysts in the Pirates front office convinced him the benefits of shifting were significant.
The Pirates shifted with more aggressiveness than any National League team in 2013. Pedro Alvarez, their third baseman, played 55 more innings than in 2012 but fielded 94 more chances. That sold Hurdle.
"The game has evolved," Hurdle said. "We showed them the statistical analysis and the reasons why."
The Phillies were slower to acknowledge that evolution. They will embrace analytics in earnest, and Amaro promised it will no longer be ignored.