A Comparative Study of Machine Learning Algorithms for Run Chase Prediction in IPL

Abstract

Machine learning has evolved as a potent tool for predicting outcomes in numerous sports, including cricket. This study investigates the potential of machine learning for run chase prediction in Indian Premier League (IPL) matches. We explore the effectiveness of five algorithms - Random Forest, Logistic Regression, Gradient Boosting, K-Nearest Neighbors and Decision Tree Classifier - in developing models to predict the success of a team chasing a set target in the second innings. Historical data on batting/bowling teams, target score, wickets lost, and other factors was used to train various models. The Random Forest model achieved the highest accuracy (99.81%) in predicting win/loss compared to other algorithms (80.06% - 98.73%). Our research emphasizes the potential of machine learning, particularly Random Forest, for accurate IPL run-chase prediction. This offers valuable insights for cricket fans, analysts, and potentially even strategists. Key Words: Random Forest, Logistic Regression, Machine Learning Algorithms, Model Performance, Classification.