FanGraphs was well represented at the MIT Sloan Sports Analytics Conference, as this intrepid reporter trailed Daves Appelman and Cameron through the halls of the Boston Convention and Exposition Center in Boston this weekend. Unfortunately, baseball in general was not as well represented as it could have been. Maybe that’s not surprising for a conference started by Daryl Morey, GM of the Rockets, and scheduled during the beginning of spring training.
The baseball highlight of the conference – other than some question-spawning research projects about featuring eight starting pitchers or refining pitch-type values – was the Baseball Analytics panel. On the panel, moderated by SBN baseball editor Rob Neyer, were Sportvision Director of Baseball Products Greg Moore, Red Sox Director of Baseball Information Services Tom Tippett, San Diego Padres Vice President of Strategy and Business Analysis John Abbamondi, Arizona Diamondbacks Scout Joe Bohringer, and noted Author and Roustabout Jonah Keri. Strong group.
The biggest “news” of the round table was the long-rumored Command f/x tracking system from Sportvision. Some in the Pitch f/x community have bemoaned the inability to track the catcher’s glove – perhaps made difficult by the relative lack of contrast between the glove and a dark uniform. Moore said that his team is now able to track the glove so that analysts can identify if a pitcher is hitting his spots and if a catcher is framing well. Along with the possibilities of Field f/x, which will focus on defensive positioning and ball trajectory, Moore felt that we are getting closer to “encapsulating the game.”
Tom Tippett agreed that the two new data streams were great advances because teams can now take a step back from results, focus on process a little more and find “quality pitches” by isolating what the player can control. After lamenting the fact that he feels that maybe he has to “throw away everything” he’s done in the last 20 years, Tippett went on to get excited about perfecting the “inexact science” that is figuring out the degree of difficulty of a defensive play.
John Abbamondi agreed that these are exciting times, but also mentioned that it’s difficult to check our answers. If we get a surprising result from one of these systems, we don’t have a second source of data that we can use, and sometimes teams are left unsure of how much impact the result should have. Jonah Keri, always thinking, suggested we install chips in the player’s brains… no, uniforms. Then he asked if that was something that would need to be collectively bargained. Moore thought that because players “have their own space and then return to their space after the play,” camera-based tech was fine for baseball, and didn’t invite the questions a chip might.
John Bohringer, who often helped bring the conversations to a practical, scouting-based point, mentioned that many of these tools only help in the “shallow end of the pool.” The minor leagues, college, international baseball makes up the “murky” deep end. He used a “see-saw” analogy to describe the relationship between data-based analysis and eyes-first analysis, saying that the deep end often needs the scouts (a.k.a. “professional guessers”) because of the incomplete data coming out of the minor leagues. Tippett agreed but felt that he could go further: “It’s all scouting.” Video, interviews, pitch f/x data, raw stats – these are all part of the “different ways we can talk about a player’s future.”
A conversation about market inefficiencies spawned some repartee. Jonah Keri talked about the Rays, as is his wont, and pointed out that the team had a 2.5-year stretch in which no starters hit the DL. Yes, the team filled the roster with young, able-bodied players, but they also did a lot of work on injury prevention. Their stadium houses more after-game concerts than anyone. They’re looking for that edge anywhere. When Tippett felt like the possible inefficiencies at play were greater than 2% (‘when a player is overpaid, we don’t say he was overpaid by 2%”), Abbamondi essentially disagreed by pointing out that because of personnel movement around the game, inefficiencies last less time and shift quicker. He said execution was the key because of the fact that ideas often travel with management from team to team.
Keri talked about how many teams missed out on a pudgy Dominican catcher that is now the best player in the game, and wondered what the Rays would be like today had they picked Albert Pujols. In one universe, perhaps Vince Naimoli is still the owner and the team re-signed some of their veterans to try to supplement their youth? After a shudder passed through most of the crowd, Tippett brought the laughs by pointing out that the Rays wouldn’t have had all those top-three picks, either.
After talking about ticket sales and dynamic pricing (Abbamondi – don’t put up the cheaper think-ahead ticket prices at the walk-up ticket booth), the panel turned to some questions from the audience. One attendee wanted to know about how teams treat the data from the minor leagues, given baseball’s unique structure of progression from one level to the next. Tippett had a good point that though major league teams have to be results-oriented, they also want to know why something happened. The minor league teams are working on “developing players, not on winning,” which makes some of the low-minor data less reliable.
Red Sox Beacon blogger Patrick Sullivan asked about Carl Crawford‘s contract. Specifically, does the deal mean that there’s “a lot of confidence in baseball about defense now?” Tippett acknowledged that the ‘story of the Red Sox going into 2010’ was that it was all about defense and run prevention, but that the team was just looking for valuable players. The Red Sox “have some confidence in defense and the importance of defense in winning baseball,” but if you add up Crawford’s home runs and triples, he’s in the thirties. “He has power.” Keri wondered if the value of his defense could be stunted by the green monster, even if Crawford can play the obstacle well, but Tippett said that all he could say is that they “spent a long time looking” at that issue, and don’t think his defensive value will be lessened by the monster.
All in all, the panel was an excellent general overview of the state of analytics in baseball. The panelists were well chosen, thoughtful, and had insights into the role of data analysis in the production of a winning baseball team. The only complaint (other than the fact that it took too long to get Keri a working microphone) might be that it would have been nice to see more baseball infused into the rest of the conference. It seems that the organizers are aware of the issue, though, and potential changes may even be in the works. Kudos to MIT Sloan for organizing a great conference and bringing together the best in sports analytics on a cold March day in Boston.
Print This Post