Hi,

Last week, I played online poker which is something I had not done for many years.

I played 8.800 hands 6max NL100.

My winrate was 12 EV bb/100.

I was wondering how to understand such a result.
The questions behind that are :
-basically, does this mean I am a winning poker player in this field?
-does this mean that if I continue playing the same way in the same players pool, etc.. I'd still get a positive winrate?
-what might be the effect of variance with such a sample?

I found a poker variance calculator at pokerdope. I don't know if I can paste the link, so i won't, but feel free to add it if it's allowed.

So I filled the form :
Winrate in BB/100 : 6 (I think 12 EV bb/100 means 6 EV BB/100, right?)
Observed winrate in BB / 100 (optional) : I didn't change anything because I don't know what they mean
Standard deviation (»?«) in BB / 100 : 38 (found in HEM)
Number of hands to simulate : 8800

So they say :

70% confidence interval (»?«)
[172 BB, 884 BB]
[1.95 BB/100, 10.05 BB/100]

95% confidence interval (»?«)
[-185 BB, 1241 BB]
[-2.10 BB/100, 14.10 BB/100]


So I assume this means that if the environment factors remain the same (my game doesn't change, the field doesn't change, etc...) :
- there's 70% chance that my long term EV winrate is between [1.95 BB/100, 10.05 BB/100], so [3.9 bb/100, 20.1 bb/100]
- there's 95% chance that my long term EV winrate is between [-2.10 BB/100, 14.10 BB/100], so [-4.40 bb/100, 28.20 bb/100]

Am I right?

I believe that's for 8800 hands only? Try changing it to 880000 hands and see what you get.

Edit: I checked it out and it looks like the graph extrapolates predictions over a large number of hands. I think, though, that 8800 hands isn't enough to give the calculator the sample size it needs to give you meaningful results. Try bumping the thread below with your question.

http://forumserver.twoplustwo.com/32...u-run-1648950/

Hi OP.

You have posted this in the wrong forum. I have moved it across to the beginners forum. You will get better answers there.

But variance is 1/3rd of what we do here!

Isn't that what the confidence interval is for?

Yes, and it's going to give you a very wide range of possibilities if you feed it a small beginning number of hands, was my rather clumsily-made point.

I think the Pokerdope variance simulator uses BB to mean big blinds, not Big Bets (or 2x bb).

Was your standard deviation really only 38bb/100? It is almost certainly going to be a fair bit higher once you've played a larger sample.

If your numbers turned out to be accurate (12bb/100, SD of 38bb/100), then Pokerdope says the 95% confidence interval is 3.9 - 20.1bb/100. Just from experience though, I know that this sample size is largely meaningless. I've had 10k samples where I won at over 20bb/100, but I'm closer to 5bb/100 in reality, and I've come across several bad players that were winning for 5000 hands that I'd expect to lose at -10bb/100 in the long term.

I thought it was pretty much impossible but I guess it might be correct if he is using EV bb/100 which has a lower standard deviation.

it's 38.5 EV BB/100
so 77 EV bb/100

You're right. They mention it in the comments.
If I replace all the BB input by the bb input, I get :

70% confidence interval
[334 BB, 1778 BB]
[3.79 BB/100, 20.21 BB/100]

95% confidence interval
[-389 BB, 2501 BB]
[-4.42 BB/100, 28.42 BB/100]

I still don't really understand these numbers.

They say : "Your actual results over the simulated amount of hands will be within this interval 70% of the time. The first interval shows absolute numbers, the second translates those into BB/100, showing the 70% confidence interval for your winrate."

This is not clear at all imo.

I think your understanding of the meaning of the numbers was correct in the first post. If you play another 8800 hands, there's a 95% probability that the result will be somewhere between a loss of 3.89 buy-ins and a profit of 25 buy-ins. Your "true" winrate is very likely to be between minus 4.42bb/100 and plus 28.42bb/100. Clearly, that's a huge range of possibilities (since you might turn out to be a small loser, or a massive winner), which goes back to the original point that there is a helluva lot of variance in poker, and the short term is relatively meaningless.

You cannot really figure it out. You assume that yours and your opponents skill level remain constant, but that's not true in reality.