http://www.svenskpoker.com/math.php

figure out:

1. your true winrate (based on your current winrate, sample size / number of hands, standard deviation, and confidence interval).

2. your true bankroll needs (based on your current winrate and standard deviation).

great website, definitely bookmarked for me.

aseem

aseem, you might want to change the link that you posted as it alread has parameters for someone's play entered in

thanks for the link, though

1) If you are running bad you normally play worse. Your winrate goes down.
2) When you feel you easily can beat one level you get lazy and dont look for soft tables and dont focus when you play. Your winrate goes down.
3) A certain limit doesnt stay the same all the time. If it - for example - gets more aggressive (or if you now prefer to play against aggressive but bad players, somethinf u didnt like before) your Standard deviation goes up and that means your swings will get bigger.
4) When you get better you normally play more hands and that increases your Standarddeviation. Swings can be bigger.
5) If you multitable your winrate goes down.
6) You tilt less, your winrate goes up.

PREACH ON BROTHER!

What's a typical standard deviation figure for low limit hold 'em?

18 BB/100.

aseem

Indeed, when I tried my stats with a standard deviation of 20 BB/100 and my true stats, I found it to be pretty darn accurate. Its high range was a value I reached once within the first 10,000 hands, and its low mark is one that I've never seen after 70,000.

Interestingly enough, my bankroll only has to be 225 BB. Who knew?

Uhm, there's something odd about their bankroll requirements.

It gives you a figure of what your maximum BB loss would be for a given standard deviation and win rate, but it doesn't say with what certainty that will hold true. Is there a default percentage that it calculates all bankrolls for?

i noticed that too. i think it's whatever the commonly accepted "risk of ruin" is, at least the same one that mason malmuth uses.

there was an interesting link to a thread where a 2+2'er argues that mason's bankroll number has to be DOUBLED because his risk of ruin isn't low enough.

that is, for example, that if the bankroll the webpage calculates (using mason's formulas) leaves a 5% risk of ruin, doubling the bankroll would make the risk of ruin around 0.1%.

aseem

We could probably calculate the figures ourself using a one tailed test at, say, 1% if we wanted to find out what bankroll would give us 99% certainty. It would probably only require an additional 100 or so big bets.

95% certainty is pretty bad, if you've got a lot riding on your bankroll. Especially if every time you move up in the limits you're exposed to that same risk.

If you follow the links from that page a couple of levels you get to this page, where you can enter the ROI:
http://support3.com/poker/bankroll/

Also, here is the argument by the 2+2'er. I think this is the post you're referring to, Aseem.
http://archiveserver.twoplustwo.com/showfl...&fpart=all&vc=1

Aseem where did you get 18 from? From the Standard Deviation.. It's a pretty complicated statistic... ussually it runs from 15-20 for people. Remember it's not always 15-20, and this isn't something that you should take to serious.

I can't dig it up right now but one of the 2+2 guys wrote a large post about why this is way to conservative of a bankroll managment.

Then again I'm saying that after averaging -4.00 BB/100 over my last 3,000 hands at 3/6... Thats NOT within my Standard Deviation lol.

Anyways this is a pretty cool tool and it's fun to toy around w/.

I just dug up this post trying to figure some things out about my own risk of ruin.

First of all it's an excellent link. Thanks for posting it in the first place.

Here's my question:

The ranges for typical standard deviations you guys mentioned were 15-20. Pokertracker shows my SD to be 13 BB/100 after about 30,000 hands.

Does this mean I could be playing too conservatively? Or is it more a product of the type of games I'm playing in? Does it even matter at all?
I'm just curious. Thanks in advance if you decide to respond.

it could mean you're playing a little weak/tight, as the hallmark of weak/tight play is usually low variance. however, it might just be the games you are in. i'm really not sure, but i lean towards thinking you're not pushing enough marginal edges.

aseem

jayson, can you sticky this?

aseem

Thanks aseem. I think it is possible that previous bouts with variance have made me pull back on some edges I pushed previously. Not quite weak/tight, but maybe a tendency in that direction.

I dont get it? too much math and stuff.