Looking at results to date . . .

[UK-21]

Visited the local house of chance recently, playing for the first time in many months, and logged my 14 unit win on my spreadsheet record as usual.

The spreadsheet I've rustled up shows that my game - playing all with a 1-8 spread (ENHC) - to have a long term +EV of 0.46%. Overall my results from the modicum of play I have on the slate (twenty four and a quarter hours) puts my actual win result to date way above three Std Devs - I've popped in a graph that shows +EV and 3 Std Devs either side (using one of the functions in Microsoft Excel) together with actual results. I've put the results down to the fact that I've been playing with a much greater advantage than the 0.46% I've assumed - sometimes, where I can, I do leave the tables when the count nosedives or spread greater than 1-8 where I'm having a good session and can afford to.

If I adjust the assumed advantage to 1.5%, my actual result sit just on +3 Std Devs. But when I do this, the 3 Std Devs limit on the neg side just about rests on the win/loss line.

Question: how many hands to I need to play in order to hit a point where the volatility is such results fall within the +/- 3 Std Dev range 99.7% of the time? (assume this to be the theoretical point N-Zero). Is is possible to achieve this in just 1,700 hands?

I've put the actual results to date down to the fact that the relatively little play I've put in has significantly skewed the results away from the EV.

Thanks in advance.

 

[blazin22]

I've put the results down to the fact that I've been playing with a much greater advantage than the 0.46% I've assumed - sometimes, where I can, I do leave the tables when the count nosedives or spread greater than 1-8 where I'm having a good session and can afford to.

Your spread has no effect on the advantage.

Question: how many hands to I need to play in order to hit a point where the volatility is such results fall within the +/- 3 Std Dev range 99.7% of the time? (assume this to be the theoretical point N-Zero). Is is possible to achieve this in just 1,700 hands?
.

Have you got CVDATA yet? It will answer all of those questions.

 

[rrwoods]

Your spread has no effect on the advantage.

Pardon? It absolutely does.

Have you got CVDATA yet? It will answer all of those questions.

I was going to say the same thing here

 

[UK-21]

Your spread has no effect on the advantage.

It must do, as the greater the spread the greater the theoretical win rate.

No, I don't have any simming software. I did look at one (can't remember which) when I first started playing, but didn't have the time/inclination to learn how to use it (I reckon at work I have to learn how to use a new software package about every two months - which is a drag when you only need to use it about once every six. I'm new software'd out! ! ! ).

 

[blazin22]

It must do, as the greater the spread the greater the theoretical win rate.

I'm sorry but this is not true, Its true the win rate increases but the advantage you're playing with is the same as it is with a lower spread.

 

[Jack_Black]

There is no increase in advantage by changing the spread. Changing your technique will change your advantage. Hole carding, ace tracking, straight out cheating, etc.

Question: how many hands to I need to play in order to hit a point where the volatility is such results fall within the +/- 3 Std Dev range 99.7% of the time? (assume this to be the theoretical point N-Zero). Is is possible to achieve this in just 1,700
hands?

Too general a question to ask. But if I was to take a guess at your game given the few details you stated, I will say 30,000 plus hands to overcome 1SD. You really should get cvdata, it doesn't take long to learn it. I never read the manual for it. It is pretty self explanatory while filling in the fields, as well as selecting options and reading reports.

If that still doesn't suffice, post all details of the game you are playing, including pen., rules and I or someone will sim it.

 

[rrwoods]

Unless we're using "advantage" differently, changing your spread changes your advantage. I am taking "advantage" to mean your total expected win/loss ratio. Thus if you have no spread (you are flat betting), your advantage is equal to -HE.

It's true that changing your spread does not change your advantage for any given hand, but it does change your advantage for the game as a whole. The context of UK-21's post leads me to believe he's using "advantage" the way I am here.

 

[UK-21]

Unless we're using "advantage" differently, changing your spread changes your advantage. I am taking "advantage" to mean your total expected win/loss ratio. Thus if you have no spread (you are flat betting), your advantage is equal to -HE.

It's true that changing your spread does not change your advantage for any given hand, but it does change your advantage for the game as a whole. The context of UK-21's post leads me to believe he's using "advantage" the way I am here.

Correct assumption.

Too general a question to ask. But if I was to take a guess at your game given the few details you stated, I will say 30,000 plus hands to overcome 1SD.

Can you explain what you mean by overcoming one Std Dev? I don't understand this comment. How does one overcome a standard deviation, when Std Devs are simply the expression of a measure of probability?

Are you saying that 30,000 hands should yield results that will be within 1 Std Dev of the EV?

I think my >3 Std Dev pos results are largely the combination of a relatively small sample size coupled with the fact that a %age of my winnings relate to dealer payout errors and cash comps and arre not purely the product of applying my betting spread and playing.

Thanks for your responses guys.

 

[Jack_Black]

you are correct in stating that your 3SD result came from the small sample size, as well as your bonuses. So how does one know if they are playing a winning game? they must track their winnings and losses and plot it against their calculated SD to see if the results show consistent and correlated growth with the calculated EV. You should plot a chart with winrate/hr expressed as a linear function. then you can highlight the area around your EV to express the calculated 1SD. Your bankroll growth should have wild swings along the winrate line, but generally staying in the calculated SD area. You will have positive and negative variance but over all, your growth can be measured by the EV as a base line.

Or you can study the function of N0(zero) the amount of hands required to overcome the luck factor of SD and show actual bankroll growth.

N0 = SD^2/EV^2

Here’s a great post by Sonny that will explain it better.

http://www.blackjackinfo.com/bb/showpost.php?p=28936&postcount=3