Predicting Soccer Results in the English Premier League
There were many displays of genius during the 2010 World Cup, ranging from Andrew Iniesta to Thomas Muller, but none were as unusual as that of Paul the Octopus. This sea dweller correctly chose the winner of a match all eight times that he was tested. This accuracy contrasts sharply with one of our team member’s predictions for the World Cup, who was correct only about half the time. Due to love of the game, and partly from the shame of being outdone by an octopus, we have decided to attempt to predict the outcomes of soccer matches. This has real world applications for gambling, coaching improvements, and journalism. Out of the many leagues we could have chosen, we decided upon the English Premier League (EPL), which is the world’s most watched league with a TV audience of 4.7 billion people [1]. In our project, we will discuss prior works before analyzing feature selection, discussing performance of various models, and analyzing our results.
Research Paper Link: Download Paper