Lee Booth Undergraduate Dissertation 2014/15

Investigating use of algorithmic approaches in automated team selection for fantasy football league

Supervised by P.McMinn

Abstract

Fantasy Football is a hugely popular online game, in which players of the game must select a team of footballers that maximise point scoring on a weekly bassis based on real world footballer performances, whilst abiding by constraints such as transfer budget and football players per team, in each implementation of the game, e.g. Fantasy Premier League, The Sun Dream Team etc. 

This project will investigate and implement algorithmic approaches into automated team selection within the domain of Fantasy Football, with the aim of optimising team selection on a weekly basis using statistical analysis of real world player performances. 

The optimisation proces of automated team selection can be represented in several ways, for example a heuristics problem, a deterministic selection based or search problem. However, this project will investigate approaching the problem using the latter, search based approach. Therefore a genetic search based algorithm will be implemented and investigated during this project, in order to try and determine the best algorithmic approach towards automated team selection.