Back to Search
Start Over
Football Movement Profile–Based Creatine-Kinase Prediction Performs Similarly to Global Positioning System–Derived Machine Learning Models in National-Team Soccer Players.
- Source :
- International Journal of Sports Physiology & Performance; Sep2024, Vol. 19 Issue 9, p874-881, 8p
- Publication Year :
- 2024
-
Abstract
- Purpose: The relationship between external load and creatine-kinase (CK) response at the team/position or individual level using Global Positioning Systems (GPS) has been studied. This study aimed to compare GPS-derived and Football Movement Profile (FMP) –derived CK-prediction models for national-team soccer players. The second aim was to compare the performance of general and individualized CK prediction models. Methods: Four hundred forty-four national-team soccer players (under 15 [U15] to senior) were monitored during training sessions and matches using GPS. CK was measured every morning from whole blood. The players had 19.3 (18.1) individual GPS-CK pairs, resulting in a total of 8570 data points. Machine learning models were built using (1) GPS-derived or (2) FMP-based parameters or (3) the combination of the 2 to predict the following days' CK value. The performance of general and individual-specific prediction models was compared. The performance of the models was described by R<superscript>2</superscript> and the root-mean-square error (RMSE, in units per liter for CK values). Results: The FMP model (R<superscript>2</superscript> =.60, RMSE = 144.6 U/L) performed similarly to the GPS-based model (R<superscript>2</superscript> =.62, RMSE = 141.2 U/L) and the combination of the 2 (R<superscript>2</superscript> =.62, RMSE = 140.3 U/L). The prediction power of the general model was better on average (R<superscript>2</superscript> =.57 vs R<superscript>2</superscript> =.37) and for 73% of the players than the individualized model. Conclusions: The results suggest that FMP-based CK-prediction models perform similarly to those based on GPS-derived metrics. General machine learning models' prediction power was higher than those of the individual-specific models. These findings can be used to monitor postmatch recovery strategies and to optimize weekly training periodization. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 15550265
- Volume :
- 19
- Issue :
- 9
- Database :
- Complementary Index
- Journal :
- International Journal of Sports Physiology & Performance
- Publication Type :
- Academic Journal
- Accession number :
- 179020508
- Full Text :
- https://doi.org/10.1123/ijspp.2024-0077