DEPARTMENT OF COMPUTER SCIENCE
MONASH UNIVERSITY

Clayton, Victoria 3168 Australia


TECHNICAL REPORT 96/297


Information-theoretic football tipping

D L Dowe, G E Farr, A J Hurst and K L Lentin

ABSTRACT

This paper described a football tipping competition based on the estimation of probabilities of victory (rather than simply an estimation of which team will win), and its connection with information theory and gambling. It also describes a football tipping competition for which entrants tip the mean and standard deviation they perceive on the margin of the game. The optimal long-term strategy in both competitions is the minimisation of the expected Kullback-Leibler distance from the "true" probability (distribution) to that tipped by the entrant.