Open Many98 opened 3 years ago
Current betting strategy seems to be trying to maximize expected profit but it does not take into account minimizing risk of ruin.
I have tried very simple approach to bet on opps which have positive E[profit] i.e. E[profit] > 1.05 but at the same time satisfying the condition E[profit] < some_limit. Using Elo as base predictive model this approach yielded 8000 final profit. The Poisson model was terminated at (2009-07-09: available: 6110.61, invested 0.00, total 6110.61) because the more data there was, the longer the model took to fit.
First proposed strategy is based on work of H.Rue and O.Salvesen Predicting and Retrospective Analysis of Soccer Matches in a League.
Second one is Kelly criterion.