Enhanced VisionPlay Research_ A Comprehensive Framework for Real-Time Decision Support and Tactical Optimization in Sports Analytics
Anand Subramani, Prathamesh Pednekar, Adnaan Shaikh, Nikshay Shingarvelu Arunthathiyar, Omkar Parulekar · Zenodo (CERN European Organization for Nuclear Research) · 2025
The report details comprehensive decision support frameworks for real-time sports analytics, including a six-stage closed-loop system for performance optimization that covers objective definition, data acquisition, integration, analytics, decision support, and implementation. It describes AI-powered tactical optimization systems like Hierarchical Deep Reinforcement Learning (H-DRL) for real-time adjustments and Markov Decision Process-based engines for substitution optimization. The infrastructure enabling this relies on edge computing and streaming frameworks to achieve sub-100ms decision cycles. The frameworks emphasize human-AI collaboration through interpretable AI (e.g., SHAP values) and are applied in sport-specific cases like basketball fatigue monitoring and soccer tactical dashboards.