Published Paper Details:

Kabaddi is a fast-growing indigenous team sport India, but most school, college and club teams still depend on manual scoring sheets and subjective judgment to evaluate performance. This limits objective player assessment, tactical learning and data-driven talent identification, despite the sport's increasing professionalization through the pro kabaddi league. This paper presents an ai-driven performance analytics and decision support system for kabaddi that integrates structured raid-wise data capture, player and team profiling, machine learning models, and natural language reporting into a unified web and mobile platform. The proposed methodology defines a kabaddi-specific data schema, a layered system architecture, and a set of models for match win-probability estimation, expected raid points (raid) and unsupervised player role clustering, together with template-based natural language generation for automatic match summaries and profile insights.