Standard Deviation... Does anyone really know they're a winner?

[Dirk Gently]

After 31 139 hands and two months of playing one limit I'm up 449BB. If having 300BB at a certain limit is supposed to effectively eliminate your risk of ruin why does winning 449 BB @ 1.38 BB/100 produce a confidence interval that is still well into the negatives? My standard deviation is around 14BB/100 (6 max).

Is 300BB really not enough? It seems I'll have to play 100,000 hands before my confidence interval converges to even within 1/2 a BB of my poker tracker value. That's six months of very consistent play required to find out if a strategy is working or not... This game is insane, isn't it?

[jba]

[ QUOTE ]
This game is insane, isn't it?

[/ QUOTE ]

yes

as someone just posted in another thread, a person can play 1,000,000 hands perfectly and lose every single one.


insane



edit to add: my mom always said I was a winner and I've always believed her

[pzhon]

[ QUOTE ]
If having 300BB at a certain limit is supposed to effectively eliminate your risk of ruin why does winning 449 BB @ 1.38 BB/100 produce a confidence interval that is still well into the negatives? My standard deviation is around 14BB/100 (6 max).

[/ QUOTE ]
The figure of 300 BB assumes you are winning at a higher rate. As you decrease your win rate, you need a larger bankroll to maintain a low risk of ruin.

The risk of ruin is not eliminated. It's just low, as is the chance of running 2 standard deviations below average.

[Dirk Gently]

Yeah, I understand that providing enough trials the risk of ruin is always 100%... Which of course means, the risk of winning 3BB an hour for your whole poker career while limping preflop on every hand is also possible.

But goodness... Six months of playing four hours a day and still not knowing for sure if your true win rate is 1 BB an hour or 50% less than that is insanity. If your true win rate was 50% less that (not that unlikely) it means you have an equally good chance of playing the next six months completely break even... Scary thought.

I don't think 99% of people who play poker understand the scope of standard deviation on the marginal win rates most of us shoot for. 100,000 hands is essentially meaningless when you're considering something like whether or not you're going to have enough money to cover a car loan for the next five years... I don't know how the pros do it.

[frappeboy]

[ QUOTE ]
don't think 99% of people who play poker understand the scope of standard deviation on the marginal win rates most of us shoot for.

[/ QUOTE ]

I agree, most players don't realize how much luck is involved, even over a long period of time. It's why I prefer playing online so much more than B&M.. You simply don't get in enough hands in live play to ensure a victory at the end of each month.

Understanding this is one of the ways I am able to keep my sanity during downswings. Although sometimes I lose it anyways /images/graemlins/smile.gif