Player Recommendation System for Fantasy Premier League using Machine Learning

Abstract

Before the rise of popularity of Fantasy Sports, people were restricted to the passive consumption of sports via television and print media. With the rise of this new age industry, people are more involved with their stakes on their selected players. This aims to enable an average interested person to make informed decisions on which players to choose and invest in based on visualizations, statistical measures, and analytics. In the past, parameters like Return of Investment (ROI) were used as a metric, but that alone is insufficient to make decisions. We attempt to solve the favoritism bias (people tend to choose from their favorite teams) and generate actionable insights using Statistical Analysis and Data Science. We use the data extracted from Fantasy Premier League (FPL) API and test against the English Premier League 2021–22 (Soccer).