
NFL Big Data Bowl:
Using Player Tracking Data to Rank Gunners, Vises, and Punt Return Degree of Difficulty
For the 2022 NFL Big Data Bowl, contestants were provided with player tracking data from Next Gen Stats (NGS) and scouting data from Pro Football Focus (PFF) for all special teams plays from the 2018–2021 NFL seasons. Contestants were tasked with utilizing the tracking data to derive actionable insights. I focused my analysis on punt coverage, setting out to understand which factors in punt coverage have the biggest impact on return yardage. I built Random Forest Regression and Classification Models, which helped to highlight key features, mostly related to getting close to the punt returner. I used these key features to rank special teams coverage players, specifically punt gunners and vises (defenders playing opposite the gunners). Given the importance of closing distance in punt coverage, I used tracking data to rank punt gunners in their ability to get downfield, both in absolute yardage and relative to the position of the vise. I also ranked vises by their ability to prevent gunner penetration. Finally, I developed a Degree of Difficulty metric used to rank punt returns.
Read. Watch. Review.
-
Watch the powerpoint presentation.
-
Read the full project write-up on Medium.
-
Review the competition workbook on Kaggle.
-
Review the project code at my Github.