Missing data patterns in runners' careers: do they matter?

TL;DR

This paper introduces a latent class matrix-variate state space model to analyze missing data patterns in runners' careers, demonstrating that accounting for these patterns improves performance prediction accuracy.

Contribution

The paper presents a novel modeling framework that explicitly incorporates missing data patterns in athletes' career data, enhancing predictive performance.

Findings

Accounting for missing data patterns improves prediction accuracy.

Missing data patterns provide valuable information for performance forecasting.

The proposed model outperforms traditional methods in out-of-sample predictions.

Abstract

Predicting the future performance of young runners is an important research issue in experimental sports science and performance analysis. We analyse a data set with annual seasonal best performances of male middle distance runners for a period of 14 years and provide a modelling framework that accounts for both the fact that each runner has typically run in three distance events (800, 1500 and 5000 meters) and the presence of periods of no running activities. We propose a latent class matrix-variate state space model and we empirically demonstrate that accounting for missing data patterns in runners' careers improves the out of sample prediction of their performances over time. In particular, we demonstrate that for this analysis, the missing data patterns provide valuable information for the prediction of runner's performance.