Fig. 2

Distribution of KsA values for neighboring distances (in meters; A/B), KsA values for middle-distance runners (C), and the trajectory of theoretical KsA values as a function of distance (D). (A, B) Data points represent KsA values (median, interquartile range) derived from annual best times of German female runners (AB) or personal best times of female runners from regional to world-class level (PB). n denotes the number of KsA values. p-values: Mann-Whitney U test. Datasets: Tab. S01–S14. (C) Data points represent KsA values (median, interquartile range) derived from personal bests of female middle-distance runners, categorized by absence (-) or presence (+) in the 3,000 m rankings. n denotes the number of KsA values. p-value: Mann-Whitney U test. Dataset: Tab. S15. (D) The two-phase exponential function was derived by successive multiplying median KsA values (Table 1, annual best times of German female runners). Inset: Same data with log-transformed distance. Arrow indicates a breakpoint in curve trajectory. Respective data for male runners are shown [5]