Improving TrueSkill for football match prediction
A fun experiment in applying Bayesian inference to football match prediction. I extended Microsoft’s TrueSkill
Despite the sophisticated Bayesian approach, Johnny Cash surprisingly lost money over the course of the experiment. This outcome reinforces what many have discovered before: betting markets are remarkably efficient, with bookmakers’ odds being highly accurate predictors of match outcomes. The house edge and bookmaker margins make it extremely challenging to achieve consistent profits, even with advanced statistical models. This project served as an interesting lesson in both Bayesian modeling and the efficiency of betting markets.
Note: This was a fun academic exercise. Remember to gamble responsibly!