In this talk, Chris will explain the history of, and methodology behind, his ‘Rate My Team’ Fantasy Premier League points projection model, which he first built in 2011 using the statistical programming language ‘R’.
Chris will explain the three main components of updating this model, as well as how they all tie together:
- the team model (how strong each team's attack and defence are, as well as how big home advantage is)
- the player rates (how likely each player is to score a goal, assist a goal, get a yellow card etc., assuming that player is on the pitch)
- the playing probabilities (how likely each player is of playing each week)
Mathematical and Economic concepts such as Monte Carlo Simulation, Bivariate Poisson Distributions and the Efficient Market Hypothesis will be touched upon, as well as more simplistic concepts like Expectation and Conditional Probability; concepts that Fantasy players have intuitive notions of, even if they are not comfortable with the formal language/notation behind the ideas.