PSL Eye: Predicting the Winning Team in Pakistan Super League (PSL) Matches

Abstract

Pakistan Super League (PSL) is a well-known T20 cricket league with millions of viewers. With this large viewer base, predicting the outcome of PSL matches opens a new research avenue for academic researchers.  In this paper, we collect PSL data from relevant sources and generate a validated data set for machine learning experiments. We implement the “PSL Eye” solution which employs Neural Networks (NNs) to predict the match winning team. We preprocess the dataset to eliminate the extra variables then we tune the hyper parameters of NN. After acquiring the optimal values of hyper parameters, we run our NN based PSL Eye to obtain the final results. The overall accuracy of PSL-Eye with testing data set is 82% which is very promising and shows the importance of NN in predicting PSL match outcome.