Cricket Forecast: Unraveling Future Matches’ Outcomes

Identifying important factors that frequently affect a team’s success is very important, that leads to the usage of mathematical models in the sports especially in cricket, in order to predict which side might win a ODI or a T20 match. Many people have started to forecast the outcome of these cricket matches that ultimately adds an extra layer of amusement and entertainment to the game. The project aims to develop a mathematical model that analyses pre match conditions and players previous performance and finally gives which team has a better chance to win. The model using RandomForestClassifier achieved an accuracy of 63% and better precision, recall and f1 score on pre match conditions and RandomForestRegressor acquired r2 score of 88% and 79% for ODI batting, T20 batting and 97% and 98% for ODI bowling and T20 bowling is achieved for different formats of the game based on players performance. Hyperparameter tuning is used to increase accuracy. Cross validation is used for performance evaluation.