Strategy having an edge

[Toby]

Some time ago I read a post by bayes who stated that you could lose even having an edge(short or long run).

What IÃ,´d like to analize is how fluctuations damage a BR when you are a regular gambler or when you are an AP.

My first analisis was  to measure the SDs using the 68-95-99.7 rule that can be applied to fluctuations too.

We supose to flatbet at table limits(to eliminate any other subject).

I sure find an expert math-member who can clarify the extremes of a rush(good or bad).


Best Regards
toby

[iggiv]

i am not an expert. But from my observation this conclusion may help u. Let's say you have a method which basically wins frequently. What happens then sometimes, u start playing and u start winning. after a while u lose your winnings and more than that. Many people use hit-run strategy. But what may help u also -- there could be certain "gaps", some number of spins it is better to skip and wait. You can analyze your strategy in RX, and sometimes what looks like loser may come up as a winner if u for example win a little, wait 100 spins and then come back.

Because of fluctuations u stay too much playing and u get them, and they destroy your winnings.

i may be wrong, not everybody agrees with that for sure

[Bayes]

mr.ore posted a formula for how your bankroll will fluctuate:

standard deviation - determine your possible profit

What is your possible profit if flat betting constant amount of money(WAGER) on an option with probability p and PAYOUT from {1,2,5,8,11,17,35} with unit size = UNIT?

BOUNDARY(x) = -WAGER/37 + x*(PAYOUT+1)*sqrt(WAGER*p*(1-p)*UNIT)

To compare results of a regular gambler vs an AP, just modify the PAYOUTS for the AP such that you have an advantage. For example, if you were betting the dozens, then instead of using 2 for the payout (which gives the standard negative expectation assuming 37 slots), try using a number > 2, like 2.2.

If you then compare this with using the standard value of 2 (which the regular gambler is subject to), then if all other values in the formula are the same, you will get a comparison of how the fluctuation differs.

[Toby]

I read the formulae and understand almost everything.

Supose The advange become the hit rate from 1/37 to 1/34 for strait bet.

What must be modified?

Sorry, but sometimes I get confused.

[Bayes]

I read the formulae and understand almost everything.

Supose The advange become the hit rate from 1/37 to 1/34 for strait bet.

What must be modified?

Sorry, but sometimes I get confused.

In that case you need to change 'p' in the formula from x/37 to x/34, and keep the payouts the same. Also the first term needs to be -WAGER/34.

It doesn't matter whether you change the payouts or the probability of a hit, because it's the relation between the two which gives you your edge (or lack of it). So with the chance of a hit being 1/34, and the payouts remaining the same, your expectation is:

(1/34) âÅ"• $35 + (33/34) âÅ"• (-$1) = (1/34) âÅ"• ($35 − $33) = 2/34 or a 5.9% advantage.