Saturday, March 2, 2013
Football leftovers from Sloan panel
By Mike Reiss
A few leftover thoughts from the football analytics panel at the MIT Sloan Sports Analytics Conference:
1. What makes football unique when it comes to analytics? When moderator Andrea Kremer asked panelists if there were key measurables that correlate to performance, the discussion turned to what made football different from other sports in that regard. "There is so much in play with respect to the team's system ... there is so much covariance in a football game because there are so many other factors," said Rams executive vice president of operations/chief operating officer Kevin Demoff, who later added this is one reason that trades often times don't work out because the projection into a new system is a challenge. That point was echoed by former Patriots vice president of player personnel Scott Pioli. "What's seemingly different when you talk to people in different sports, one of the things we bump up against all the time in trying to be predictors through analytics, is that football is based on all these interdependent relationships," he said. "There are so many variables, whether it's playing surface, the weather, the environment. You're talking about 11 interdependent relationships, where a person's success on one play is so dependent on what someone else is doing, or not doing. It makes it really difficult to say, 'X, Y and Z are going to equal this.'"
2. Knowing your own free agents & context of outside-the-meeting-room film evaluation. Pioli shared his opinion on why he feels that one of the most important parts of free agency is a team's evaluations of its own players. "That's the critical part because you are not going to know any player better than your own players; you're going to know what's great about them, and you're going to know all of their warts. It's an interesting thing when you read some of the analysis of people who are looking at tackles, for instance. They're saying 'they've only given up X number of sacks.' But truly, the only people that know how many sacks that tackle gave up are the coaches or the people who sit in the meetings and hear. Or the number of mental errors they make. Anyone who is just watching the game, they have no idea -- 'Was that a game plan where they were sliding things a certain way? Was it supposed to be a chip from the back? Was the tight end supposed to be in one place; did the tight end make a mental error?'"
3. Wholesale vs. retail and the Patriots' replenishing the system. San Francisco 49ers chief operating officer Paraag Marathe delved into some of the salary-cap based analytics of building a team, saying: "Free agents and veterans are retail -- it's name your price. The draft is where it's wholesale. So that's why the more good players you can have on wholesale -- the more draft picks you can accumulate -- the better off you'll be. ... The difference between wholesale and retail is so wide that you have to figure out which players to keep and which players to let move on and churn out, because you have to continue to replenish the system. That's what Scott did so well at New England. They were the kings at replenishing and that's why they always stayed competitive and won Super Bowls despite the existence of a hard salary cap. When everyone else has ebbs and flows, the Patriots were consistently good."
4. How having a head coach with solid job security benefits the Patriots. There was a good discussion on trades, and teams that deal a future draft pick in a higher round for a current-year pick. This has been an area in which the Patriots have been able to capitalize in Bill Belichick's tenure, and the reason was highlighted by the panel. Pioli touched on the patience that the Patriots practiced and how they were sometimes rewarded for it -- turning a seventh-round pick into a mid-round pick a few years later. While Pioli noted that point of view, which was rooted in Belichick knowing his job was secure, others on the panel pointed out a contrasting view: Sometimes a desperate team is running out of time. "One of the things that you hear come up today constantly is how long it takes, the arc of time for decisions to come true," Demoff said. "Trading down is great in theory, but it doesn't work when you have a coach in the next-to-last year of his contract, or when teams are starting to look at things and you feel like you're in your window. That's the great thing about trades; everyone has a different vantage point of what they need now. ... Ultimately if the time frame for decision-making and how things play out for head coaches and GMs keeps shrinking the way it has over the last few years, it's harder and harder to convince them to make good decisions. To a team that just missed the playoffs because they don't have a return man or they need speed, it looks like a lot different. That's where the trading teams find value."