Prediction of Shot Selection of a Batsman in the death over of a T20 Cricket Match using Machine Learning

Cricket has seen remarkable changes since the beginning of the 21
^st
century. T20 cricket has gained immense popularity, captivating a global fan base with its fast-paced action and continuous excitement. This study delves into the intriguing dynamics of the final over in T20 cricket, aiming to predict a batsman's shot selection during this crucial phasethat can drastically influence the match's outcome through the implementation of machine learning algorithms. Leveraging a comprehensive dataset obtained from sources like ESPN Cricinfo and Cricsheet.org, which includes ball-by-ball commentary, delivery outcomes, ball types, shot types, required run rates, and more. The proposed model used Ensemble techniques to show noteworthy improvement in the performance metrics. SHAP, an explainable AI tool, was employed to study the impact of different features on the prediction. This research seeks to provide valuable insights for cricket teams, both established and emerging, to enhance their game analysis and strategic planning during the last over, as strategic planning and execution are paramount, particularly in the second innings of a match.