Advancing Cricket Analytics: A Machine Learning Approach to Predicting Players and Optimizing Lineups

Statistics have been crucial in shaping the structure and format of cricket as it has recently developed into one of the most popular international mainstream sports. Data on cricket matches and individual players are fast expanding due to novel forms like Twenty-Twenty (T- 20) cricket and an increase in the number of matches over time [1]. The demand for big data analytics is increasing, as are the possibilities for using this large data successfully in many advantageous ways, such as team player selection, predicting match victors, and many more future forecasts using big data or machine learning models [2]. Without access to massive data sets or the big data platform Spark ML, machine learning linear regression models are employed to forecast team scores [3]. This thesis contains research articles and a non-statistical essay about the decision problems of when to declare during the 2nd innings of an ipl cricket match, the choice of the best team lineups in Twenty20 cricket, and an analysis of the value of fielding in cricket. The article also explores the use of machine learning techniques to forecast cricket match outcomes using historical IPL match data [4][5].