Football performance results from the dynamic interaction between physical, tactical, technical, and psychological dimensions—each of which also influences player well-being, recovery, and readiness. However, integrated monitoring approaches remain scarce, particularly in youth and sub-elite contexts. This systematic review screened 341 records from PubMed, Scopus, and Web of Science, with 46 studies meeting the inclusion criteria (n = 1763 players; age range: 13.2–28.7 years). Physical external load was reported in 44 studies using GPS-derived metrics such as total distance and high-speed running, while internal load was examined in 36 studies through session-RPE (rate of perceived exertion × duration), heart rate zones, training impulse (TRIMP), and Player Load (PL). A total of 22 studies included well-being indicators capturing fatigue, sleep quality, stress levels, and muscle soreness, through tools such as the Hooper Index (HI), the Total Quality Recovery (TQR) scale, and various Likert-type or composite wellness scores. Tactical behaviours (n = 15) were derived from positional tracking systems, while technical performance (n = 7) was assessed using metrics like pass accuracy and expected goals, typically obtained from Wyscout^® or TRACAB^® (a multi-camera optical tracking system). Only five studies employed multivariate models to examine interactions between performance domains or to predict well-being outcomes. Most remained observational, relying on descriptive analyses and examining each domain in isolation. These findings reveal a fragmented approach to player monitoring and a lack of conceptual integration between physical, psychological, tactical, and technical indicators. Future research should prioritise multidimensional, standardised monitoring frameworks that combine contextual, psychophysiological, and performance data to improve applied decision-making and support player health, particularly in sub-elite and youth populations.

Optimizing recovery is crucial for maintaining performance and reducing fatigue and injury risk in youth football players. This study applied machine learning (ML) models to classify mental fatigue in U15, U17, and U19 male players using wearable signals, tracking data, and psychophysiological features. Over six weeks, training loads were monitored via GPS, psychophysiological scales, and heart rate sensors, analyzing variables such as total distance, high-speed running, recovery state, and perceived exertion. The data preparation process involved managing absent values, applying normalization techniques, and selecting relevant features. A total of five ML models were evaluated: K-Nearest Neighbors (KNN), Gradient Boosting (XGBoost), Support Vector Machine (SVM), Random Forest (RF), and Decision Tree (DT). XGBoost, RF, and DT achieved high accuracy, while KNN underperformed. Using a correlation matrix, average speed (AvS) was the only variable significantly correlated with the rating of perceived exertion (RPE) (r = 0.142; p = 0.010). After dimensionality reduction, ML models were re-evaluated, with RF and DT performing best, followed by XGBoost and SVM. These findings confirm that tracking and wearable-derived data are effectively useful for predicting RPE, providing valuable insights for workload management and personalized recovery strategies. Future research should integrate psychological and interpersonal factors to enhance predictive modeling in the individual long-term health and performance of young football players.

The aim of this study was to test a machine learning (ML) model to predict high-intensity actions and body impacts during youth football training. Sixty under-15, -17, and -19 sub-elite Portuguese football players were monitored over a 6-week period. External training load data were collected from the target variables of accelerations (ACCs), decelerations (DECs), and dynamic stress load (DSL) using an 18 Hz global positioning system (GPS). Additionally, we monitored the perceived exertion and biological characteristics using total quality recovery (TQR), rating of perceived exertion (RPE), session RPE (sRPE), chronological age, maturation offset (MO), and age at peak height velocity (APHV). The ML model was computed by a feature selection process with a linear regression forecast and bootstrap method. The predictive analysis revealed that the players’ MO demonstrated varying degrees of effectiveness in predicting their DEC and ACC across different ranges of IQR. After predictive analysis, the following performance values were observed: DEC ($\overline{\mathrm{x}}$_predicted = 41, β = 3.24, intercept = 37.0), lower IQR (IQR_predicted = 36.6, β = 3.24, intercept = 37.0), and upper IQR (IQR_predicted = 46 decelerations, β = 3.24, intercept = 37.0). The player’s MO also demonstrated the ability to predict their upper IQR (IQR_predicted = 51, β = 3.8, intercept = 40.62), lower IQR (IQR_predicted = 40, β = 3.8, intercept = 40.62), and ACC ($\overline{\mathrm{x}}$_predicted = 46 accelerations, β = 3.8, intercept = 40.62). The ML model showed poor performance in predicting the players’ ACC and DEC using MO (MSE = 2.47–4.76; RMSE = 1.57–2.18: R² = −0.78–0.02). Maturational concerns are prevalent in football performance and should be regularly checked, as the current ML model treated MO as the sole variable for ACC, DEC, and DSL. Applying ML models to assess automated tracking data can be an effective strategy, particularly in the context of forecasting peak ACC, DEC, and bodily effects in sub-elite youth football training.

The aim of this study was to analyze the match-to-match variation in high-intensity demands from one Portuguese professional football team according to playing positions. Twenty-three male outfield professional football players were observed during eighteen matches of the Portuguese Second League. Time–motion data were collected using Global Positioning System (GPS) technology. Match running performance was analyzed based on the following three playing positions: defenders (DF), midfielders (MF), and forwards (FW). Repeated measures ANOVA was utilized to compare match running performance within each position role, and seasonal running variation. Practical differences were assessed using the smallest worthwhile change (SWC), coefficient of variation (CV), and twice the coefficient of variation (2CV). Significant differences were found among playing positions in total distance covered (F = 15.45, p < 0.001, η² = 0.33), average speed (F = 12.79, p < 0.001, η² = 0.29), high-speed running (F = 16.93, p < 0.001, η² = 0.36), sprinting (F = 13.49, p < 0.001, η² = 0.31), accelerations (F = 4.69, p = 0.001, η² = 0.132), and decelerations (F = 12.21, p < 0.001, η² = 0.284). The match-to-match running performance encompassed TD (6.59%), AvS (8.67%), HSRr (37.83%), SPR (34.82%), ACC (26.92%), and DEC (27.85%). CV values for total distance covered ranged from 4.87–6.82%, with forwards and midfielders exhibiting the greatest and smallest variation, respectively. Midfielders demonstrated the highest match-to-match variation for all other analyzed variables (8.12–69.17%). All playing positions showed significant variation in high-demanding variables (26.94–37.83%). This study presents the initial analysis of match-to-match variation in high-intensity demands within a Portuguese professional football team. Thus, the position’s specificity and context can provide a helpful strategy for evaluating match-to-match running performance, and for recommending individualized training exercises based on the peak and high-intensity demands for each player’s role within the game.

Utilizing techniques for reducing multivariate data is essential for comprehensively understanding the variations and relationships within both biomechanical and physiological datasets in the context of youth football training. Therefore, the objective of this study was to identify the primary factors influencing training sessions within a standard microcycle among young sub-elite football players. A total of 60 male Portuguese youth sub-elite footballers (15.19 ± 1.75 years) were continuous monitored across six weeks during the 2019–2020 in-season, comprising the training days from match day minus (MD-) 3, MD-2, and MD-1. The weekly training load was collected by an 18 Hz global positioning system (GPS), 1 Hz heart rate (HR) monitors, the perceived exertion (RPE) and the total quality recovery (TQR). A principal component approach (PCA) coupled with a Monte Carlo parallel analysis was applied to the training datasets. The training datasets were condensed into three to five principal components, explaining between 37.0% and 83.5% of the explained variance (proportion and cumulative) according to the training day (p < 0.001). Notably, the eigenvalue for this study ranged from 1.20% to 5.21% within the overall training data. The PCA analysis of the standard microcycle in youth sub-elite football identified that, across MD-3, MD-2, and MD-1, the first was dominated by the covered distances and sprinting variables, while the second component focused on HR measures and training impulse (TRIMP). For the weekly microcycle, the first component continued to emphasize distance and intensity variables, with the ACC and DEC being particularly influential, whereas the second and subsequent components included HR measures and perceived exertion. On the three training days analyzed, the first component primarily consisted of variables related to the distance covered, running speed, high metabolic load, sprinting, dynamic stress load, accelerations, and decelerations. The high intensity demands have a high relative weight throughout the standard microcycle, which means that the training load needs to be carefully monitored and managed.

INPLASY202550010 Shaping Training Load, Tactical-Technical Behaviour and Well-Being in Football: A Systematic Review Keywords: Football; Training load; Well-being; Monitoring; Performance; Sub-elit…