Since cricket matches are recorded using multiple independent variables within a historical data set and a dependent variable (the result of the match), this problem can be solved through predictive analysis (classification methods) within machine learning. Jhanwar & Pudi conducted an experimental study to predict the outcome of ODI cricket matches using data mining techniques. This article investigates machine learning technology to address the problem of predicting the results of cricket matches based on historical data from IPL matches. To take into account potential future growth, the International Cricket Council (ICC) commissioned a global market study that revealed that cricket has more than 1 billion fans around the world, with significant growth potential.
When applying these methods to predict the outcome of the match, it was discovered that the model derived by Naïve Bayes offered a prediction accuracy of around 64% in the data set used. This research attempts to evaluate different machine learning techniques for the problem of predicting the outcome of IPL cricket matches. While no statistically significant difference could be found in the estimations of the parameters, the class difference (when a developing cricket country hosted an established cricket country) decreased (29.6 races against 26% in Kampakis). Thomas conducted a study to predict the outcome of cricket matches in the format of twenty sachet.
Predict the outcome of a T20 cricket match while the match is in progress (doctoral thesis, BRAC University).