It differs from year to year, but overall, I have benefited from Kelly. This year, if I used a higher fraction, I would be up a lot more. Since my medium and lower bet plays (small sample) are actually slightly outperforming the bigger plays that I have been posting. For this reason a % of bankroll method with a starting bet size commensurate to Kelly, would be outperforming Kelly. But I do not expect that to hold up.
What's your line on Fla/Col for tomorrow. I am leaning Rockies, but I stay away from Greg Smith because i don't know how to valuation him with all the sample size or what kind of match-up favors etc. Anything i should know about his pros and cons? Everything else favors Rockies at home. I don't think Ricky matches up well with this stacked Rockies squad at home.
-102/102. Market seems to be in the process of correcting an inflated Marlins open. HFA is magnified in this one. I am not big on Smith, and Coors is a horrible home park for his pitching style. Despite having past success in Coors and against the Rockies, Nolasco's flyball propensity can catch up with him int his park.
Kelly is optimal for those that have logarithmic utility, seeking to optimize bankroll growth, and for those who can accurately quantify their edge to inflection point where betting larger sums on games with the perceived edge is greater than those with smaller edges outweigh the variance increase.
