PDA

View Full Version : Calculate your winrate and bankroll requirements with this

kiddo
08-18-2005, 04:24 AM
Every week there is a new post about winrates and bankroll requirements. Maybe this can help some of those posters.

Me and Snow (another swedish 2+2er) have created a page where u can calculate how big bankroll you need (given your BB/100 and Standard deviation). And calculate how much your true, long term, winrate can differ from the one you got right now (given your number of hands, your winrate, your standarddeviation and how sure u want to be (1-100%)

The math behind it is taken from Malmuths Gambling Theory and Other Topics . A book you should read if u are interested in these sort of things.

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

BruceZ
08-18-2005, 04:42 AM
[ QUOTE ]
Every week there is a new post about winrates and bankroll requirements. Maybe this can help some of those posters.

Me and Snow (another swedish 2+2er) have created a page where u can calculate how big bankroll you need (given your BB/100 and Standard deviation). And calculate how much your true, long term, winrate can differ from the one you got right now (given your number of hands, your winrate, your standarddeviation and how sure u want to be (1-100%)

The math behind it is taken from Malmuths Gambling Theory and Other Topics . A book you should read if u are interested in these sort of things.

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

[/ QUOTE ]

Malmuth's formulas cannot be used to compute bankroll. They will underestimate bankroll requirements by more than a factor of 2. This thread has the correct bankroll formulas (http://archiveserver.twoplustwo.com/showflat.php?Cat=&amp;Number=207100&amp;page=&amp;view=&amp;sb=5&amp;o =&amp;fpart=all&amp;vc=1) along with a discussion of the problem with the Malmuth formulas. Here is a derivation (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&amp;Number=683150&amp;page=0&amp;view=ex panded&amp;sb=5&amp;o=14&amp;fpart=2#Post682045683150) of the correct formulas. If you want a tool that uses the correct formulas, someone just posted one here (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&amp;Number=3171070&amp;page=0&amp;view=c ollapsed&amp;sb=5&amp;o=14&amp;fpart=1) earlier today.

I see that your webpage does not give a risk of ruin, but I can tell you that if you implemented Malmuth's "3 sigma" formulas, they will give a risk of ruin of 1.1%, (not 0.13%). So if anyone wants a 1.1% risk of ruin, then OK they can use this calculator, but realize that this is what the numbers mean.

Your win rate and confidence calculation should be OK if you implemented Malmuth's formulas correctly.

kiddo
08-18-2005, 06:56 AM
Malmuth says in the thread you linked:

[ QUOTE ]
By the way, after many years of using my published numbers, I have found them to be very accurate (at three standard deviations).

[/ QUOTE ]

Do you know pokerplayers that have played for many years that think he is wrong? And do they instead say you need twice as big bankroll as Malmuth says?

In my experience, when pokerplayers that have played for years talk about their biggest downswings Malmuths numbers are pretty ok.

If you play poker for a living, what you really want to know is how bad is an extremly bad downswing? You want to know this because

1) If u know you can lose a lot being unlucky but still playing good it makes u feel a little better when u actually loose.

2) You want to know how much money you need to have in your account when you play online.

I have played for 3 years about 700 hands/day (+700K hands). I still havent lost much more then Malmuths formulas says. And actually, 2 of those 3 times when I lost that much I didnt play really good poker so I think my winrate was lower then normal.

Isnt Malmuths numbers good for practical purposes? How often will a pokerplayer loose twice the amount Malmuth says.