My name, Kevin Baer, written in a fancy logo font
  • About
  • Projects
  • Shiny Apps
  • Data Viz Gallery
  • Resume

Swarm Score: Quantifying Rallying to the Football

Football
Big Data Bowl
Paper
Published

January 8, 2024

Big Data Bowl 2024 Entry (Link)

Swarm Score is a quantitative assessment of the age-old football maxim of rallying to the football and swarming the ball carrier. The statistic is formed by comparing each defender’s bivariate normal spatial influence probability density function against the ball carrier’s spatial influence model to determine how many defenders are swarming the ball carrier to maximize tackling efficiency. Swarm Score proves to be an effective success measure of a variety of tackling statistics such as YAC allowed, YAC over expected allowed, seconds from catch till tackle, EPA allowed, and many more. Finally, Swarm Score is shown to be a much better predictor of future YAC allowed than prior YAC allowed performance, demonstrating it’s stability and effective value.

To read the full paper, click here!

To see the code, click here!