We've all heard at one time or another of a MIT professor that was able to successfully apply informatics to gambling. Of course this refers to Claude Shannon's work which was also applied to the stock market with a great deal of success at the time. It's pretty much the basis of all digital communications. They had an equation for the sizing of bets. Obviously gambling might actually be easier to apply the equation to as the odds can more accurately be determined than with investing. Does any one have any examples of applying the Kelly Criterion that they developed to gambling and how one would do it? Could you share it with us?
The Kelly Criterion was about sizing the bets logarithmically to account for the possibility of losing the bet so that you would still have capital for subsequent bets. It required expectations and the concept of surprisal which were odds expressed in bits of surprisal ie.: log(odds)/log(2) where odds = 1/probability not the gamblers bit of odds = 1/probability - 1
Odds are usually displayed as 1
dds so it would've been more correct for me to say odds against - sorry.
The Kelly Criterion was about sizing the bets logarithmically to account for the possibility of losing the bet so that you would still have capital for subsequent bets. It required expectations and the concept of surprisal which were odds expressed in bits of surprisal ie.: log(odds)/log(2) where odds = 1/probability not the gamblers bit of odds = 1/probability - 1
Odds are usually displayed as 1
dds so it would've been more correct for me to say odds against - sorry.