This chapter examines the important role of machine learning in sports predictive analysis and play evaluation. It covers a spectrum of techniques, from traditional result analysis to advanced machine learning approaches, addressing key areas such as game result, event, and trajectory prediction, as well as action and space evaluation in team sports. The chapter introduces various datasets and methodologies, highlighting the evolution from rule-based systems to deep learning models. It explores how these techniques are applied to classify plays, cluster similar behaviors, extract meaningful features, and learn complex representations from sports data. The discussion extends to counterfactual analysis, providing insights into hypothetical scenarios and their potential impacts. By presenting cutting-edge research and future directions, including posture analysis and real-time analytics, this chapter offers a comprehensive view of how data-driven approaches advance sports analytics, enhancing our understanding of individual and team performances, and informing tactic decision-making in sports.