Dr Ian McHale, senior lecturer in statistics at the University of Salford, discusses in IET this month how mathematical models of football matches are used in the gambling industry – and sportingly puts his neck on the line by supplying his own predictions for the World Cup 2010.
Predicting football results is a rapidly growing area of academic interest. Economists use models to assess the efficiency of betting markets, operational researchers use models to experiment with the various effects of tournament design, and statisticians showcase their proficiency with advanced statistical techniques by modelling the intricacies of football data.
It is not, of course, just academics who are mining the archives of football scores. Bookmakers live and breathe football prediction models - as do the more committed flutterers. Mistakes cost money and jobs, whilst finding a small advantage can carry great rewards.
Betting markets
In academia, the most common application of football forecasting models is to test for betting market efficiency. The Efficient Markets Hypothesis (EMH) is a cornerstone of financial theory and, in its simplest form, states that an investor should not be able to consistently obtain returns above the average. Finding a forecasting model of football that can generate better-than-average - or even positive - returns usually results in a publication for the academic as an example of a violation of the EMH, but the proprietary nature of the models means that the published ones rarely (if ever) represent the very best models, and even less often generate positive returns consistently.
The best performing models are the reserve of the gambling industry. It is paramount for a bookmaker to set odds at a value that realistically represent the probabilities of a match being won, drawn or lost. If the bookmaker fails to do this, it will risk huge losses.
For instance, Asian bookmakers would think nothing of taking an individual bet of US$200,000 - and regularly receive bets of $400,000 - and a typical weekend in the English Premier League typically attracts $500m turnover in Asia. With such tides of cash being wagered, it is not surprising that bookmakers make use of every possible tool at their disposal - one of them being mathematical models.
A mathematical model is not typically used on its own to set odds. An expert odds-setter is employed to adjust the model-generated odds given any extra information. For example, a…
