This is the second of a series of features on innovators who work behind the scenes in the sports world, leading up to our coverage of the 2010 MIT Sloan Sports Analytics Conference on Saturday, March 6, in Boston. Last week we talked to Red Sox general manager Theo Epstein, and on Thursday we’ll be speaking with Houston Rockets GM Daryl Morey. Today FootballOutsiders.com creator and stat guru Aaron Schatz drops by the studio to answer our questions about the MIT Conference, his panel on Emerging Analytics, and some other general football issues.
WBRU Sports: Can you tell us a little bit about the conference?
Schatz: Well interestingly enough, you know, I’m a football guy, but it’s not actually very big for football. It’s really huge for basketball. It started three years ago, and it’s run by two MIT Sloan business school graduates; Jessica Gelman, who works in the front office for the Patriots on the business side, and Daryl Morey, who is the general manager of the Houston Rockets. So it’s very split in two. You have a lot of panels which are related to business things: internationalization of sports, how to deal with labor relations, how to increase attendance for minor leagues, things like that. And then you have another set of panels that are analytical for front offices, and the biggest of those really is the basketball panel. Basically anybody who does analytics for an NBA franchise shows up to this thing. You also have a big baseball panel, you’ve got what they’re calling an Emerging Analytics panel, which is basically sports we didn’t know what else to do with; that’s me and Paraag Marathe from the 49ers, and two soccer guys from England. They’re having one on Coaching Analytics this year, how to use it in coaching with Jon Gruden and Avery Johnson, who used to coach the Mavericks. And then a couple of just big overall- how fans relate to the game today, where they have people like Bill Simmons and Mark Cuban. It’s a really good thing, it’s really not that technical. People hear analytics, they think it’s gonna get all technical on them. It’s not, but it’s actually more business related than a lot of people who I think are just fans of the game and they hear about it from Bill Simmons. I think it’s more business related than people realize.
WBRU Sports: On the topic of emerging analytics, other than the fact that baseball is more of a purely statistical sport than football, what do you think has caused football and the rest of these sports to be so far behind?
Schatz: Well, that’s the biggest, is that baseball is based on stats and always has been. Box scores have been in the paper for a hundred years. So people were playing around with baseball stats long before they were playing around with stats in other sports, it was the biggest sport in America for a long time, now the second biggest, behind football, but it was still a lot bigger than basketball, which, you know, has a lot more stats than football does. So baseball has always been the leader in this kind of thing and basketball and football have been catching up, and then other sports are even harder to do particularly any kind of a game analysis because you don’t have discrete events. Soccer flows. It’s really hard to mark in some kind of numerical way whether a certain soccer player is good or bad, because so few of them result even in shots on goal, let alone scores. Hockey can be somewhat similar. It’s hard to do these sports, though, I think actually the next sports- the next single sport in particular for this kind of revolution to spread to is cricket. But somebody has to pick up the ball and that somebody probably has to be in England, ‘cause it’s not going to be me, but that’s another sport that has sort of a lot of numbers to it, like basketball, baseball, and football do.
WBRU Sports: Do you think that emerging statistical evaluations in football serve to ultimately start to replace more conventional scouting or just to complement them more?
Schatz: Always to complement. Numbers do not replace scouting, numbers replace numbers. Ok? Everybody who looks at football- whether they are a fan on the barstool, or a general manager- uses both numbers and their own eye in order to figure out what they think about the game. Now, obviously your own eye, if you’re a general manager, is going to be a lot better than your own eye if you’re on a barstool. We call that scouting. And as far as numbers- the numbers we do with Football Outsiders or some of the other folks out there do, like Advanced NFL Stats or KC Joyner at ESPN- those numbers are a lot better than the conventional “let’s just add all the yards together as if they’re all the same thing” stats that the league does. But they’ll never replace scouting. They just complement scouting much better than traditional stats do.
WBRU Sports: Daryl Morey has become kind of a statistical hero in that he’s had a statistical background, he went to MIT and now he’s a general manager of an NBA team. Do you think that could happen in football? Do you see it happening in the future?
Schatz: It somewhat already is. Football teams have always accepted statistical analysis more than teams in other sports for a couple of reasons. Teams were family owned for a long time so they were used to somebody who never played the game coming from a business background to run the show. They had a salary cap before other sports so they needed people who could find economic efficiency. So the thing is that football teams have always been very quiet and proprietary about their analytics but you already do have people in some of the front offices who do think a lot like Daryl Morey. A good example is Paraag Marathe with the 49ers who is their Director of Football Operations. Now he’s not the number one guy, he’s the number two guy. But you have a lot of teams that have number two, three type guys who have a lot of this background. You just don’t have anywhere where the person who is the face of the franchise like Morey is, the face of the front office at least, has that background.
WBRU Sports: But could that happen in the future? That someone from MIT would come up…
Schatz: Not only do I think it will happen in the future, I’m pretty sure that within ten years you’re going to have someone who kind of grew up reading my work and Football Outsiders at the head of an NFL team.
WBRU Sports: When you’re evaluating projections for a player, how do you find a solid balance between statistics and intangible qualities?
Schatz: My projections are all going to be based on statistics. Now there is some subjectivity into that because you have to be subjective in trying to figure out the player’s role on the team. For example, when you’re doing fantasy football projections, huge swings in value come from the whims of coaches, and sometimes those whims are not based on which player actually is better on a per-play basis, they are based on the whim of the coach, or some intangible, like a guy’s got a drinking problem or had a car accident in the offseason — probably related to his drinking problem. So we have to look at that and include that subjectively, but normally what we want to do is use just the stats and then we subject those projections to a little bit of a smell test based on common sense. I try to get across to people that we don’t ever use just numbers, we try to subject everything to a common sense filter before we put it out there.
WBRU Sports: In the case of a guy like Tim Tebow, is there anything to be said for his perceived leadership or winnability? Do you think that the team that will draft Tebow is going to get an NFL ready player?
Oh, not an NFL ready player. They’ll get an NFL ready leader if they want him to lead chapel on Sundays. You know, what I say often is skills like leadership, character, these things exist. But you cannot measure them; that’s why they’re called intangibles. So I don’t attempt to measure them. For someone like me who’s a stat analyst to put together some kind of ridiculous leadership quotient number would just be the dumbest thing on earth. I trust that teams will take these things into account along with player abilities and things that can be measured numerically. And Tebow also is an interesting case because it’s specific things with his style that caused a problem for him in translating from college to the pros. It’s not something you have to deal with for most players, particularly at other positions. It just so happens that the particular way he played quarterback in college is not conducive to success at the NFL level.
WBRU Sports: How much emphasis should we put on the results of the combine when evaluating players?
Schatz: The most important parts of the combine are the parts you do not see on television. The medical reports, and the player interviews that are done in the evenings, in the hotel suites by teams, which are usually a combination of personal questions and asking players to diagram plays, try to explain plays- those things are far more important than the drills that you see on the NFL Network. Now there are certain things with certain drills that are predictive. One thing we have found is that you can do a lot to predict a running back if you look at just his 40 time at the combine and his weight. We used that to create a stat called Speed Score. For example, the top Speed Score this year was Ben Tate because, even though he only ran slightly faster than a guy like C.J. Spiller, he’s heavier, and stronger, and therefore, his speed means more. That’s why, for example, we have Toby Gerhart now, in Speed Score, is higher than C.J. Spiller. Now that’s not the only thing teams should look at, but it is important in looking at how they’re gonna do. For the most part, though, these drills are kind of pointless. If scouts are doing their due diligence, what matters is what they see on film of these players and how they play. It’s not necessarily the quality of their play at the college level, but it’s certain specific things that translate to the NFL because, in college, the best players have a much larger advantage over an average player than they do in the NFL, so you don’t want to be tricked by a guy making a great play solely because he’s going up against a guy who’s never gonna play in the pros. You have to look really- they look much more for guys with raw skill sets.
WBRU Sports: Transitioning to the Patriots just for a second, what do you think Julius Peppers brings to a team? Do you think he fits into the Patriots’ system at all? What do you think the chances are that he ends up there?
Schatz: You would be surprised how much the Carolina Panthers dropped Julius Peppers into coverage on zone blitzes, so he does have experience doing that, if you wanted to pick him up as an outside linebacker in the 3-4. I think the thing about Peppers is that his performance has not lived up to his notoriety over the last 2 or 3 years, but at the same time, it’s been very good. It just so happens that he’s thought of as a “Superstar,” and what he has been is a “Star,” a good player, maybe even a very good player, but not the kind of player where 2 or 3 of these guys transforms your team. I think he would be a really good addition to the Patriots. The question is, do they want to spend the amount of money it would take to get him? It’s definitely a hole that they have, an outside linebacker.