Fair Qualifying Times Across Age and Gender Categories for the Boston Marathon

• Activity Number: 167 - Sports Analytics Outside the Big Four
• Type: Topic Contributed
• Date/Time: Tuesday, August 4, 2020 : 10:00 AM to 11:50 AM
• Sponsor: Section on Statistics in Sports
• Author(s): Richard Smith*
• Companies: University of North Carolina at Chapel Hill
• Keywords: Running performance; Age and gender; Mixed-effects models; Bayesian statistics; Boston Marathon

Abstract

All runners get slower as they age, and in the vast majority of events, women are slower than men. But how should one quantify the differences? One widely used method is age-graded times, but being based on world record performances they may not correspond to performances by ordinary runners. Furthermore, many large races (especially, the Boston Marathon) impose qualifying times for guaranteed entry, but the standards are not based on detailed comparisons between age groups. This study (based on data from the Boston Marathon) aims to quantify the age-gender discrepancies based on typical runners’ performances. A mixed effects model is proposed to estimate the time-age curves for both men and women using random effects to account for differences among runners. The results show marked discrepancies from both age-graded curves and from the age-gender relationships that are implicit in the Boston Marathon standards.