A Statistical/Algorithmic Framework for Modeling Fixed Odds Games

N. Glinos, Y.C. Stamatiou, and M. Vamvakari (Greece)


Fixed-Odds games, statistical modeling, linear program ming, distribution fitting.


Fixed odds forecast games attract the attention of both game operators and players as for the former they offer an attractive game product with potentially large profit mar gin and the for the latter they offer a trade-off of certainty versus volume of winnings that, in turn, challenges them to developing their knowledge about the events in the coupon. However, from another perspective, fixed odds forecast games abound in interesting statistical and combinatorial problems that can be of a practical interest to the game or ganizers in order to monitor, adjust and optimize some of the game parameters. For example, one problem of statisti cal nature could be the following: “Do the odds assigned to the outcomes of event x fit in players’ preferences?”. In this paper we propose a theoretical framework for addressing such questions in the context of general fixed odds games based on elements of statistical inference and elements of linear optimization. This framework has been implemented and successfully applied to the modeling and the analysis of the Greek fixed-odds forecast game. Part of this analysis is also presented in this paper.

Important Links:

Go Back