Molenaar, L. (Luuk) (2020) Performance of general classification riders on a final climb in professional road cycling: what can be predicted and what does influence performance? thesis, Sport Sciences.
Full text available on request.Abstract
Introduction: Professional road cycling is one of the most physically demanding sports, with the three Grand Tours (the Giro d’Italia, the Tour de France, and the Vuelta a España) as the most challenging events. The outcome of the Grand Tours general standings is mostly determined during the uphill finish of mountain stages. This study aimed to reveal the most important factors that could influence performance during a final climb. Furthermore, this study aimed to build an optimal prediction model for different performance outcomes: relative power output, time gap, and ranking. Methods: Power output (PO) data of four professional cyclists were collected and a total of 132 races were analyzed. One-way analysis of variance was used to determine the differences between the four riders. Stepwise Multiple Linear Regression (SMLR) analysis was used to identify which factors influenced the performance outcome. Finally, a machine learning approach with an ElasticNet regression was used to build the optimal prediction model. Results: The SMLR analysis conducted in this study showed that the relative PO was the best performance measure on the final climb. The machine learning model showed that by using an ElasticNet regression a prediction model can be created with an explained 76 % of the variance of the PO·kg-1 during a final climb. A high TSS·km-1 before the climb and a higher average gradient of the climb resulted in a higher relative PO during the climb. The PO on the climb was negatively influenced by a higher elevation gain before the climb and a longer duration of the climb. Conclusion: A machine learning model can be used to predict performance outcomes on a final climb with PO·kg-1 as the most valid performance outcome measure. The prediction model showed that the most important factors that influence performance during a final climb were the load before the climb, short-term fatigue, and the characteristics of the climb. Keywords: professional road cycling, final climb, relative power output, machine learning
| Item Type: | Thesis (UNSPECIFIED) |
|---|---|
| Supervisor name: | Kempe, dr. M. and Groot, S. |
| Faculty: | Medical Sciences |
| Date Deposited: | 20 May 2022 09:21 |
| Last Modified: | 20 May 2022 09:21 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/3368 |
Actions (login required)
 |
View Item |