Primary Submission Category: Applications in Physical Sciences, Engineering, Environment and Miscellaneous Applications
Is Cross Country a Team Sport? A Bayesian Hierarchical Analysis of Attached vs. Unattached Runners
Authors: Jared Fisher, Nathan Sandholtz, Sam Lee, Brylee Wilcox, Garritt Page,
Presenting Author: Jared Fisher*
Cross country races are scored as team competitions yet experienced primarily as individual efforts, raising the question of whether team affiliation measurably affects individual performance. We investigate this question by comparing collegiate runners’ race times when they compete “attached” to a team versus “unattached.” Using more than a decade of results from the Track and Field Results Reporting System (TFRRS), we develop a directed acyclic graph (DAG) to formalize the causal structure governing performance and to identify covariates required to block relevant backdoor paths. Guided by this DAG, we fit Bayesian hierarchical models that adjust for runner ability, race distance, course, year in school, and time-in-season, i.e. factors that jointly confound team status and individual performance. Across specifications, we find consistent evidence that runners perform better when racing as part of a team, improving by roughly seven to thirteen seconds depending on race distance. These findings provide quantitative, causally interpretable evidence that team affiliation enhances individual performance, suggesting that psychological or motivational mechanisms associated with shared identity play a meaningful role even in a minimally interactive sport.