M. Moinat, O. Fabius & K. S. Emanuel (2018) Data-driven quantification of the effect of wind on athletics performance, European Journal of Sport Science, 18:9, 1185-1190, DOI: 10.1080/17461391.2018.1480062
The relationship between wind and athletics performance has been studied mainly for 100 m sprint, based on simulation of biomechanical models, requiring several assumptions. In this study, this relationship is quantified empirically for all five horizontal jump and sprint events where wind is measured, with freely available competition results. After systematic scraping several elite and sub-elite results sites, the obtained results (n=150,169) were filtered and matched to athletes. A quadratic mixed effects model with athlete and season as random effects was applied to express the influence of wind velocity on performance in each event. Whether this effect differs with performance level was investigated by applying the model on subgroups based on performance level.
In the fitted quadratic model, the linear coefficients were significant (p<0.001) for all events; the quadratic coefficients were significant for all events (p<0.001) except long jump (p=0.138). A 2.0 m·s-1 tail wind provides an average advantage of 0.125, 0.140 and 0.146s for the 100 m, 200m and 100/110m hurdles, respectively, and an advantage of 0.058 and 0.102m for long jump and triple jump, respectively. Performance level had a significant effect on the wind influence only for 100m (p<0.001). Amateur athletes (~13s) benefit 69% more from a 2.0 m·s-1 tail wind than elite athletes (~10s). Practical formulas are presented for each event. These can easily be used correct results for wind speed, allowing better talent scouting and championship selection. This study demonstrates the efficacy of answering scientific questions empirically, through freely available data.
|Input performance||Predicted performance|
|Event||at m/s||at 0.0 m/s||at +2.0 m/s|
How to use this calculator
For a given result, input the the performance in one of the boxes. Select the wind speed with the slider. Then, you can see what, according to the model proposed in the article, the result would have been at a wind speed of 0.0 m/s and at 2.0 m/s (maximum allowed wind assistance). For example, a 200m of 21.63 with a wind speed of +0.2, tells you that it would have been 21.65 without wind and 21.51 with 2.0 m/s. For each event, you can input a performance which you want to study.