ABSTRACT
According to legend, a Greek warrior ran about forty km to proclaim the victory of a battle and died after his proclamation. This legend was taken up in 1896 at the first Olympic Games of the modern era. A long-distance run of about forty km was created and called a marathon. Nowadays, 42.195 km is the official distance of a marathon. The marathon finish time can then be extrapolated using the finish time of the shorter distance competition. Predicting the goal time of a marathon can help athletes to prescribe their pacing strategy. This should lead to exhaustion at the finish, and not earlier. A model for predicting marathon goal time was presented by Pradier. They made use of Bayesian nonparametrics (BNP) and the hierarchical Dirichlet process (HDP) to analyze running patterns out of split times during marathon competition. Subsequently, goal times can be predicted by given marathon split times during the first kilometers.
