Description

This project provides a review of nflWAR in the area of sports analytics, provides context and motivation of the approach, and demonstrates the approach on data. nflWAR is an innovative statistical framework designed to revolutionize the way we evaluate individual player contributions in the National Football League (NFL). The incentive of nflWAR is rooted in addressing the need for a reproducible and robust evaluation method within the realm of professional football. Building upon the foundation of Wins Above Replacement (WAR), a metric traditionally used in baseball to assess a player’s overall impact on team success, nflWAR adapts and refines these concepts specifically for football. Emerging as a solution to often-limited statistical frameworks available for a complex team sport like football, this advanced metric allows analysts, coaches, and fans to quantify how many additional wins a player provides over a replacement-level counterpart. By taking into account the unique roles and responsibilities in football, nflWAR provides a more standardized way of evaluating players across different leagues and seasons, offering a comprehensive tool that enhances player evaluation and strategic decision-making in NFL. Moreover, it brings a level of precision and insight to player evaluation that goes beyond traditional statistics, enabling a deeper understanding of a player’s contribution to a team’s success. The data used in this project is the 2017 NFL season play-by-play data, scraped using nflscrapR by the authors of nflWAR. The data contains 45,241 plays and 103 variables.