The proliferation of artificial intelligence (AI) technologies is significantly enhancing the precision and efficacy of decision-making in basketball. This study focuses on players from the men’s basketball team at School A, employing AI to model their sports postures, collect athletic performance data, and analyze physiological metrics in real time. By utilizing fuzzy gray correlation analysis, the research aims to identify the factors influencing basketball game outcomes, thereby informing tactical decisions for coaches. Moreover, this paper compares the proposed analysis method with other correlation techniques to evaluate its relative superiority. The findings indicate that post-training, the serum levels of all ten athletes exceeded 400 U/L. Yet, the creatine kinase (CK) levels remained below 6.8 mmol/L, and the urine specific gravity surpassed 1.05, suggesting that the training intensity was high but within the athletes’ physiological tolerances. With a correlation coefficient of 0.4, the study confirms a positive impact of game outcomes on the coach’s tactical adaptations. This underscores the utility of integrating AI systems into basketball to facilitate timely and effective decision-making by coaches.