M.I.T. Sloan Sports Analytics Conference Video

At the start of the month we noted the news that Colts President Bill Polian was in attendance at the Mecca of sports statgeekdom, the Sloan Sports Analytics Conference. ESPN has posted video of the panel Polian participated in, which also featured Rockets GM Daryl Morey, Mavericks Owner Mark Cuban, Patriots executive Jonathan Kraft, Bestselling author of Moneyball, The Blindside and many other non-sports titles, Michael Lewis and ESPN's Bill Simmons. 18to88 and a rather offended Aaron Schatz from FO have already commented and the full 80 minute video is embedded after the jump.

My thoughts and some highlights:

• 4:15  Bill Simmons complains about stats being hard with all the division and multiplication and such.
• 8:35  Jonathan Kraft says that the main application for statistical analysis in Football is how to most effectively use the resources limited by the salary cap, because the subjective measures of a player are a much bigger part of football due to player performance being very interrelated with teammate performance. He expresses some frustration about attempts by the public to apply statistical analysis to the NFL without the required knowledge of the ins and outs of the salary cap.
  
• 11:00  Bill Polian divides Football into Personnel Management and On-Field Management. He praises Moneyball for it's message about the key to Personnel Management being finding undervalued assets and the extreme usefulness of statistical analysis as a tool for finding them.  Polian says that the Colts have certain "inviolate" baselines for measurables and if a player doesn't meet their minimum the player is not on the draft board no matter what their scouts say.
• 14:00  Polian calls everything he's seen in the field of statistically analysis for game management, strategy and tactics, worthless, because of the key role of teamwork and matchups as well as the totally different systems, schemes and techniques used by different teams. He doesn't feel that situations are very comparable between matchups, game situations, etc, which leaves you with too small a sample for reliable, meaningful, analysis.

• 24:00  Cuban and Lewis talk about the future and current role that evaluating players ability to learn and retain information, as well as other psychological factors, play in sports like Football and Basketball (but not baseball as Lewis says he's yet to hear an anecdote of a player too stupid to play baseball).
• 27:00  Kraft points out the numerous layers to building a successful football team, finding talent that fits the scheme, signing them in a way that leaves you room in your cap/budget to go out and get more, coaching up that talent once it's found and signed, all before they even hit the field and the layers that adds.
• 29:40  Polian thinks the new frontier of Football analysis is finding a way to measure players accounting for the scheme they play. The example given is Vince Wilfork. Wilfork's assignments in New England are totally different than what Indy demands from it DTs. How to evaluate Wilfork in the light of how he'd play in a different scheme, is what Polian would like to see advance.
• 34:55  Polian says that Football Coaches are as advanced and accepting of statistical analysis as you will find in sports. That adjusting for matchups and tendencies based off stats has been a part of the game for as long as he's been a part of it. Kraft agrees and says the only real difference in coaches as far as stats usage is their abilities to understand and use them, instead of differing levels of acceptance.
  
• 37:00  Kraft says that the three coaches he's worked with since buying the Pats, Pete Carrol, Bill Parcells and Bill Belichick have all been pretty good coaches, but had dramatically different intellectual capacities.....  and refused to go further into their individual "strengths and weaknesses."
• 38:50   4th and 2 talk.    Polian makes the case for the Pats going for it based off the game situation. 2 Pats rushers injured, a 3rd gimpy, playing 2 non-starter CBs due to injury and the game having been heavily passing focused up to that point, meaning the rushers were worn down. In addition the Pats have a history of success on 4th down in the Brady era and Kevin Faulk is one of their best at gaining short yardage. Polian says that all the statistical analysis goes out the window based off game situation elements that aren't factored in.
• 42:50  Bill Simmons admits to having written a column against the 4th and 2 attempt. Uses "we" talking about the Pats repeatedly, makes fun of himself for it, then accidentally does it again. Thinks the 4th and 2 attempt was "panicky" and "not from a position of strength"
• 45:15  Polian and Simmons agree that a blanket statement that a team should always go for it on 4th at x yardline with y down and distance is wrong because of all the game situation factors that aren't included in that analysis.
• 46:50  Polian talks about the widely varying amounts of different penalties called by different crews and the role of statistical analysis in discovering and utilizing those differences in ref tendencies.
• 48:00  Simmons repeats the "rules changed after 2003" BS to Polian's face. Polian corrects Simmons on what the penalty in question was, and there's a laugh all around and an awkward silence.
• 49:40   Daryl Morey reveals that in order to help keep their statistical systems secret they sign their interns to long term contracts and try to keep their FO staff together.
• 51:10  Mark Cuban  talks about teams playing "protect the moron" with their players, in order to pass them off on other teams.
• 54:00  Kraft talks about how deeply they go into psychological evaluation of players after they determine the player has the base athleticism needed to play in the NFL. The Colts and the Patriots are among 3-4 teams that use the same independent company that does psych evals of prospects.
• 60:00  Polian talks about the opportunity for an edge for teams that can find good ways to find personality traits that are important for the sport and to the organization and nurture those traits once they are in the organization. Polian says he absolutely believes in clutch performance.
  
• 62:00  Daryl Morey and Mark Cuban disagree about paying extra money for a player with a history of clutch performance. Morey says he'd like to have it in a player, but won't pay for it when it hasn't been shown to be consistent, while Cuban says he will and has paid for a player he felt had clutch ability (Jason Kidd).
• 63:20  Polian admits that the best talent evaluators hit about .550, that a little bit of an edge in developing a player, or finding a just slightly more talented guy makes a significant difference in wins.
• 69:10  Answering a question about the importance having played the game has to player eval, Polian says that coaching experience is what really makes a different. He believes having played the game is a hindrance when it comes to negotiating contracts because the culture difference. A agent coming out of law school respects the others argument while players have learned a "my way or the highway" approach.
  
• 73:35 Polian thinks the best video replay officials would be ex-scouts.
• 75:40  Cuban compares owning the Pittsburgh Pirates to having a job that consists of being yelled at and hated by a whole city, but that pays 15-20mil a year.
• 78:30  Polian calls the Patriots the "Team of the decade" and praises their and the Red Sox' ability to pool their talents to address all the different aspects of running a successful franchise.
  

In all a very interesting panel that gives some insight into how the Colts are run and how the Colts utilize statistics.