Quote:
Originally Posted by Franco Santiago
Concur, Mr. Schwartz.
As we know, Kelly is a mathematical function. If you absolutely know the probabilities and the return on investment, it works, just assuredly as the Titanic sinking once it was holed.
Therefore, the only conclusion we can draw is that, in horse race handicapping, the probabilities of winning are actually lower than the handicapper estimates.
And the truth is, the public is a much better handicapper than the great majority of individuals. So when an individual says the probability of winning is 25% and he/she thinks she has an overlay at 5 or 6-1 (and begins to salivate like Pavlov's dog), well, the fact of matter is that the probability of that horse winning is really probably closer to 15 or 20%, creating an underlay. And thus, when this is repeated over and over again, Kelly "seems not to work". But it ain't Kelly that ain't workin'...it's the handicapping that is deficient.
(Disclaimer - Dave knew all this, but didn't wanna say it. LOL)
|
Franco,
Actually, I did say it in an earlier post that was VERY LONG.
For the most part, I agree with you.
The only statement I disagree with is in bold. (And I know you meant it somewhat metaphorically.)
The issue isn't that WE overestimate probabilities. The issue is that we overestimate sometimes, underestimate other times, and cannot tell the difference.
Many players make a line on a horse - say 3/1 - and (logically) think that if they double that to 6/1 they just HAVE TO be profitable.
Adding a margin of error will not overcome weak handicapping!
I actually had a coaching client years ago who said that he'd be better off playing against his own picks, so I told him to try it. He reported back a week later that his results improved. (Still lost money but not as badly.)
In my opinion, the single thing that prevents most players from winning is that they do not actually have a simple core handicapping approach.
Most players view each race as a unique problem to be solved.
It is much easier to be competitive (in terms of bankroll growth/deterioration) if you find commonalities between races and develop a "core method."
Viewing each race - and each horse in a race - as a unique puzzle to be put together leaves you with reinventing your approach in every race.
What I am saying is that when "puzzle players" develop a
consistent core method, it is like providing them with a picture of the box showing the edges and corners. It gives them a logical place to begin.
Dave
