Defense wins championships?
Well that depends on how you measure defense. Across the football statistics community, there's a lot of varying defensive measures out there and a lot of people who claim to possess the right defensive statistics; many that are severely flawed, but there are some that do have some shreads of predictive power.
Several weeks ago, on Jan. 13 during the divisional championship weekend, I made my Superbowl prediction that the San Francisco 49ers would square off against the Baltimore Ravens in the Superbowl. Most people thought I was crazy.
Meanwhile, Nate Silver predicted that the Seattle Seahawks would play the New England Patriots in the Superbowl [link]. Both teams can now both be seen on the golf course. What's even more frightening is that this second prediction was published by Nate Silver offering wisdom about predicting the Superbowl between the Ravens and the 49ers, tilting toward a 49ers win [link], but giving a hint at some ways you can approach analyzing defenses. The one decent token of wisdom here is "that Defense-adjusted Value Over Average (D.V.O.A.), accounts for a team’s success or failure on every play it ran during the year and not just on final scores".
The dominant and widespread amateur forecasts rely on metrics using final scores. More accurate forecasts rely on metrics using plays. This is not rocket science. More granular data is more content rich.
When you look at D.V.O.A. Ravens actually come out with much better special teams metrics and roughly the same defensive ranking as the 49ers (part of not a bad reason to forecast a 49ers-Ravens superbowl).
Does Special Teams Win Championships? Probably not. Though for these reasons and a few more:
My prediction is that the Baltimore Ravens will win the Superbowl XLVII (2013).
No comments:
Post a Comment