Cric Win: Cricket Bigdata Processing with a Deep MLP Classifier to Predict Direct Win in Live Games

Cricket is one of the most popular open-air and unpredictable sports, producing a large volume of quantifiable data. As the popularity of hundred-ball cricket (HBC) games grows, the most difficult challenge is predicting the results of cricket matches, where the high precision score prediction is still anticipated. Traditional methodologies and algorithms for forecasting the probable outcome of a match are restricted in terms of dependability and timeliness, particularly in the case of HBC. In this paper, we study cricket game events and related big datasets and propose CricWin, a web-based approach to predict the outcome of cricket matches for both innings. To predict a match winner, we use a deep learning approach that has yet to be utilized in cricket, particularly in HBC. We propose a Multi-Layered Deep Learning Perceptron (MLDLP) network-based classification model in the deep learning approach to construct different models for predicting the outcome of both innings of a match. We construct a web application to do predictive analysis on a live HBC tournament to predict the game's outcome before the game begins and/or on ongoing matches. The evaluation result demonstrates that CricWin can predict outcomes with up to 100% accuracy, showing that the proposed MLP strategy is dependable and perfect.