Hi,
I'm looking at racing in Hong Kong and would like to apply the Kelly criterion to my probability estimates of the races.
I've seen a few algorithms for solving the kelly optimized bet percentages for mutually exclusive outcomes, but was never able to get something that looked correct.
To start with, all of the algorithms calculate a "score" and then sort things by that "score". I've see the score represented as "edge", as "odds * implied_odds", as "expected value", etc. Not sure which is ideal.
Is there a good resource, sample code, or good algorithm outline somewhere?
Thanks
I'm looking at racing in Hong Kong and would like to apply the Kelly criterion to my probability estimates of the races.
I've seen a few algorithms for solving the kelly optimized bet percentages for mutually exclusive outcomes, but was never able to get something that looked correct.
To start with, all of the algorithms calculate a "score" and then sort things by that "score". I've see the score represented as "edge", as "odds * implied_odds", as "expected value", etc. Not sure which is ideal.
Is there a good resource, sample code, or good algorithm outline somewhere?
Thanks