NFL Data Bowl Winner Announced

The annual NFL Big Data Bowl serves as a pipeline for innovation, with over 75 participants having been hired into data and analytics roles in sports. Winning projects from previous years have introduced metrics that are now used by NFL teams and in live game broadcasts, analyzing everything from special teams to pass rush plays. The competition challenges the analytics community to use NFL's Next Gen Stats to develop new insights into the game. This year's competition focused on analyzing player movement while the ball is in the air. Ferraz's "Ghostbusters" project introduced a novel method for evaluating defenders in pass coverage by utilizing "hypothetical distributions of ghost defenders" to analyze their movement. This approach moves beyond traditional metrics to quantify a defender's positioning and path to the ball. Traditional methods of evaluating pass defense often rely on statistics like interceptions, pass breakups, and completion percentage allowed. While valuable, these stats don't always capture the nuance of a defender's performance on plays where they aren't directly involved in the outcome. More advanced analytics have introduced metrics like Defensive Value Over Average (DVOA) and Expected Points Added (EPA) to assess a defense's overall efficiency. The NFL's own Next Gen Stats have also developed "Coverage Responsibility" models to better identify which defender was covering a receiver on any given play. Ferraz's "Ghostbusters" concept and the resulting SLIME scores offer a new layer of analysis by focusing on the quality of a defender's movement and positioning, regardless of the play's result.