This research aims to analyze recent injuries that occurred in the National Football League (NFL) to identify patterns and variables that increase injury incidence. Our analysis of injuries spans from the 2019-2020 season to the 2020-2021 season. Throughout the semester, we have been pulling, scraping, and analyzing NFL data from multiple sources. We have been manually encoding injuries as contact or non-contact, injury area, and player role by watching 1000+ plays of NFL Film. Our primary research goal is to analyze the difference between injuries that occur on turf, grass, and hybrid fields. We hypothesized that more lower body injuries would occur on turf fields due to its rigidity and stiffness. Football is traditionally a very contact heavy sport, and analyzing why these injuries are occurring and in what contexts may be helpful to design games to reduce injury. We will analyze over 2000 plays of data throughout this research project and use this information to create a Poisson regression model. This model will show what predictors are significant to injuries, such as field type (turf or grass), weather, stadium type (dome, open, etc.), player position, days since last game, and more. Based on our analysis, we found that most injuries occurred on turf fields, to defensive backs, and affected the lower body. Our model suggests that field type is not strongly correlated to increased rates of injury. This research has been extremely relevant to our team as we have been able to apply techniques that we learned in our business analytics courses to solve a real-world problem.
Author(s): Samantha Erne, Business Analytics Major
Jacob Holroyd, Business Analytics Major
Peter Walsh, Business Analytics and Finance Major
Brendan Beattie, Information Systems Major
Advisor(s): Fadel Megahed, Department of Information Systems and Analytics