IPL Win Prediction System To Improve Team Performance using SVM

Abstract

Machine learning and data analytics are trending field currently with rise in huge amount of data
generated and sophisticated algorithms developed. One of the fields of machine learning is predictive
analytics, where probability of a particular outcome in the future is predicted, based on their
historical data. Cricket is one of the most popular sports globally, especially in India. Huge amount of
resources, effort, and time is spent to organize cricket matches and leagues, such as IPL (Indian
Premier League). Thus, with large stakes involved, it creates enormous pressure on players, coaches
and team management to perform well. Hence, in this paper, we worked on creating a win prediction
system, which predicts the probability of a team winning in a particular match, through various
parameters (features), based on their historical matches. The input parameters are changed, so that
the highest probability of winning can be obtained, in that particular match. Thus, it helps team
captain, coaches, and management to choose those parameters (players), in that match, to increase
their win probability. Also, the strengths and weaknesses of the order of the bowling and batting of
the team are obtained to improve team performance.