Brian Burke, ESPN Sports Data Scientist, discusses the Football Power Index (FPI) and creating confusion against offensive lines. The podcast also delves into personnel's impact on team strategies, calculating uncertainty in football analytics, analyzing priors and game probabilities in sports models, and the importance of cognitive abilities for quarterbacks.
Simulated pressure in football defenses can create confusion and lead to one-on-one matchups and wasted blockers, increasing the chances of generating sacks and reducing yards per attempt for the opposing offense.
ESPN's Football Power Index (FPI) incorporates offensive, defensive, and special teams ratings based on net EPA per play, along with a quarterback component, and even though last year's version performed below expectations, they are currently reevaluating and refining the model.
Deep dives
Discussion on the effectiveness of simulated pressure in football
In this podcast episode, Brian Burke, a sports studio scientist at ESPN, discusses the concept of simulated pressure in football defenses. Simulated pressure involves lining up like a blitz with multiple defenders on the line of scrimmage, but only rushing four, dropping the others into coverage. This creates confusion for the offensive line and can lead to one-on-one matchups and wasted blockers. Burke explains that more exotic pressure schemes, like simulated pressure, can be highly effective in generating sacks and reducing yards per attempt for the opposing offense.
Explaining the Football Power Index (FPI) and its updates
The podcast also delves into ESPN's Football Power Index (FPI). Brian Burke provides insights into the model, explaining that FPI involves offensive, defensive, and special teams ratings based on net EPA per play. The model also incorporates a quarterback component to account for the impact of the quarterback on the offense and overall game. While there is uncertainty baked into the model, the priors play a significant role even as the season progresses. Burke acknowledges that last year's version of FPI did not perform as well as expected due to ambitious changes, but they have reverted to the previous version and are currently reevaluating and refining the model.
Analyzing the performance of rookie quarterbacks and the New York Jets
The podcast wraps up with a discussion on the performance of rookie quarterbacks in the NFL, with a specific focus on Zach Wilson of the New York Jets. Brian Burke suggests that the Jets should adopt a defensive-focused strategy to minimize the risk given Wilson's struggles. He mentions that Wilson is currently performing below average, but the team could compensate for this with a strong defense and special teams. The conversation also touches on the challenges of accurately evaluating and projecting quarterback abilities, highlighting the cognitive and psychological aspects that are harder to gauge prior to actual game experience.
Wharton's Eric Bradlow, Cade Massey, Adi Wyner, and Shane Jensen speak with Brian Burke, ESPN Sports Data Scientist, about the start of the NFL Season, College Football, and the Football Power Index (FPI).