Data collection in soccer encompasses a variety of methods, ranging from simple daily ques tionnaires that monitor subjective well-being to more sophisticated approaches such as wearable sensors that track external load during training and matches. While subjective measures like wellness questionnaires and Rate of Perceived Exertion (RPE) offer insights into players’ inter nal states, they fall short of providing comprehensive assessments of players’ physical fitness and performance. Consequently, the use of wearable sensors has become prevalent, offering detailed metrics such as distance covered, sprints, accelerations, and energy expenditure. These metrics enable coaches to better manage training loads, plan sessions and monitor the return-to play (RTP) process after injuries. However, traditional methods that aggregate data across entire matches often hide the vari ations in players’ performance throughout the game. This dissertation addresses this issue by implementing a minute-by-minute analysis of player physical intensity using wearable data, re vealing more details about players’ fitness levels. By examining how physical performance fluctuates within different game contexts, the study provides a more accurate reflection of a player’s condition and readiness. The relationship between cognitive load (CL) and physical performance is another critical aspect explored in this work. WhileCL’simpactontacticalandtechnicalaspects is well-studied, its effects on physical performance, particularly in real-world settings, remain unexplored. Us ing data fromtheNASA-TLXquestionnaire, thisresearchinvestigates how varying levels of CL influence physical metrics such as total distance, high-speed running, and deceleration. More over, the study applies machine learning (ML) algorithms to classify players into motivation clusters, offering insights into how different motivational factors might affect performance un der cognitive stress. Injury recovery is a crucial area in sports, and the RTP process is crucial in ensuring players return to competition safely. This dissertation advances the understanding of injury manage ment by developing ML models to predict recovery duration, using data collected by medical staff. These models are compared with traditional expert-based predictions, demonstrating that the integration of expert input with ML techniques enhances the accuracy of recovery time es timates. In addition to injury recovery, the research focuses on developing individualized fatigue and recovery profiles. These profiles are essential for conditioning coaches to tailor training regimens and ensure that players return to their pre-injury fitness levels. By providing a de tailed, data-driven approach to monitoring physical recovery, this dissertation contributes valu able tools for optimizing player performance and reducing the risk of re-injury. This work provides a comprehensive framework for soccer performance and recovery anal ysis, integrating wearable sensor data, cognitive load assessments, and ML techniques. The findings offer practical applications for coaches and sports scientists, helping them to better un derstand and manage the complex dynamics of player fitness and performance throughout the season.