A Study on Predicting the Performance of Bangladesh Men’s ODI Cricket Team Using Machine Learning

Abstract

The proposed study is an example of machine learning to analyze and predict the success of the Bangladesh national cricket team in the ODI International matches. The study uses a structured dataset with batting, bowling and match level characteristics to model match results using pre-processing, feature engineering and supervised learning methods. Three machine learning techniques, namely, Logistic Regression, Random Forest and XGBoost, were trained and tested. Out of them, XGBoost has the best accuracy of 0.7556, which was higher compared to Logistic Regression and Random Forest. A measure of evaluationand analysis of features importance statistics indicated that the run rate, wick- ets taken,economy rate, and power-play efficiency have been the main determinants of the ODI success in Bangladesh. The results have indicated that machine learning can help to capture complex performance trends and provide useful data-driven decision-making in cricket.In general, the research presents the opportunities of predictive analytics in helping the strategic planning and performance measurement of ODI cricket.