<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Periods of intensified training may increase athletes’ fatigue and impair their recovery status. Therefore, understanding internal and external load markers-related to fatigue is crucial to optimize their weekly training loads. The current investigation aimed to adopt machine learning (ML) techniques to understand the impact of training load parameters on the recovery status of athletes. Twenty-six adult soccer players were monitored for six months, during which internal and external load parameters were daily collected. Players’ recovery status was assessed through the 10-point total quality recovery (TQR) scale. Then, different ML algorithms were employed to predict players’ recovery status in the subsequent training session (S-TQR). The goodness of the models was evaluated through the root mean squared error (RMSE), mean absolute error (MAE), and Pearson’s Correlation Coefficient (r). Random forest regression model produced the best performance (RMSE=1.32, MAE=1.04, r = 0.52). TQR, age of players, total decelerations, average speed, and S-RPE recorded in the previous training were recognized by the model as the most relevant features. Thus, ML techniques may help coaches and physical trainers to identify those factors connected to players’ recovery status and, consequently, driving them toward a correct management of the weekly training loads.</p>
</abstract>

Periods of intensified training may increase athletes’ fatigue and impair their recovery status. Therefore, understanding internal and external load markers-related to fatigue is crucial to optimize their weekly training loads. The current investigation aimed to adopt machine learning (ML) techniques to understand the impact of training load parameters on the recovery status of athletes. Twenty-six adult soccer players were monitored for six months, during which internal and external load parameters were daily collected. Players’ recovery status was assessed through the 10-point total quality recovery (TQR) scale. Then, different ML algorithms were employed to predict players’ recovery status in the subsequent training session (S-TQR). The goodness of the models was evaluated through the root mean squared error (RMSE), mean absolute error (MAE), and Pearson’s Correlation Coefficient (r). Random forest regression model produced the best performance (RMSE=1.32, MAE=1.04, r = 0.52). TQR, age of players, total decelerations, average speed, and S-RPE recorded in the previous training were recognized by the model as the most relevant features. Thus, ML techniques may help coaches and physical trainers to identify those factors connected to players’ recovery status and, consequently, driving them toward a correct management of the weekly training loads.

Analysis of Relationship between Training Load and Recovery Status in Adult Soccer Players: a Machine Learning Approach

Department of Movement, Human and Health Sciences

University of Rome “Foro Italico” Rome

University of Coimbra, Research Center for Sport and Physical Activity

Computer, Control and Management Engineering Department

International Journal of Computer Science in Sport

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Periods of intensified training may increase athletes’ fatigue and impair their recovery status. Therefore, understanding internal and external load...

Analysis of Relationship between Training Load and Recovery Status in Adult Soccer Players: a Machine Learning Approach

Pubblicato online: 17 gen 2023

Pagine: 1 - 16

DOI: https://doi.org/10.2478/ijcss-2022-0007

Parole chiave
SOCCER, RECOVERY, MACHINE LEARNING, PREDICTION, PERFORMANCE

© 2022 M. Mandorino et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Analysis of Relationship between Training Load and Recovery Status in Adult Soccer Players: a Machine Learning Approach

Pubblicato online: 17 gen 2023

Pagine: 1 - 16

DOI: https://doi.org/10.2478/ijcss-2022-0007

Parole chiaveSOCCER, RECOVERY, MACHINE LEARNING, PREDICTION, PERFORMANCE

© 2022 M. Mandorino et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Parole chiave
SOCCER, RECOVERY, MACHINE LEARNING, PREDICTION, PERFORMANCE