Mike H: Kelly Betting fractions and SCORE

[Mike H]

Thanks for your patience with me on this topic. There's really no way to maximize SCORE because to get SCORE you must follow all the preconditions. The only way to change SCORE for a specific game is to change count systems. If you change to some fraction of Kelly you no longer have SCORE, but you still have c-SCORE or fractional-Kelly-SCORE. If a fractional-Kelly-SCORE is higher than the real SCORE, would that betting strategy be more attractive? If so is there a fraction of Kelly that generally outperforms the others? I think I'm beginning to understand but it's late and my brain is fried.

[Don Schlesinger]

> Thanks for your patience with me on this topic.
> There's really no way to maximize SCORE because to get
> SCORE you must follow all the preconditions.

Given a particular set of rules, pen, etc., a game can have only one SCORE. So, yes, in that sense, there is nothing to maximize.

> The only
> way to change SCORE for a specific game is to change
> count systems.

Right.

> If you change to some fraction of Kelly
> you no longer have SCORE, but you still have c-SCORE
> or fractional-Kelly-SCORE.

Right.

> If a fractional-Kelly-SCORE
> is higher than the real SCORE,

Impossible. Except for rounding to reasonable real-world bet sizes, changing the Kelly fraction does nothing to the SCORE. You bet half as much, you win half as much, the variance is half as much, and SCORE doesn't change.

> would that betting
> strategy be more attractive?

See above.

> If so is there a fraction
> of Kelly that generally outperforms the others?

You keep asking the wrong question.

> I think I'm beginning to understand but it's late and my
> brain is fried.

I'm beginning to notice! :-)

Don