Bankroll

[adios]

Linky to Bankroll Formula Links Posted By BruceZ

This link should provide you with your required bankroll given you know your win rate, standard deviation, and risk of ruin.. I think 300 BB is probably way to small for online play FWIW.

This forum might be of interest. Interesting to read about the downswings and upswings.

BBV Forum

These horror stories may be of interest:

1200 BB Downswing

1500 BB Downswing

Downswing reaches 2100 BB after original post of 1500 BB Downswing


Downswings Suck

1000+ BB Downswing

[adios]

If you drop down in limits say after losing 150 BB you can reduce your bankroll requirement quite a bit.

[Vern]

I wrote a sim several years ago that took Homer's variance post and put it in a graphical program format. I was amazed when I let it run to see that 10K really is a small sample size for WR 2BB/100 @ SD 15BB and I mean a really small sample size, several were in the negative. Only after about 250K hands did the line start to look truly consistantly straight, but the program also analyzed the highest and lowest point differential over every 10K hand period and any sample size larger than 250K, while converging to the true win rate pretty accurately, also shows natural variance swings of 400-500BB down swings over a 10K period in the sample. The program is crude and probable needs to be re-worked. I don't have anywhere to host it at the moment, but let me check a couple of places and see what I can do.

I just ran it for the old 10Billion hand cycle, 2BB/100 @ 16BB SD and the apparent win rate was 1.99 BB/100 and the largest upswing was 1007BB and the largest down swing was 590. So for all of you that want to play a 10 Gigahands, be fore warned [img]/images/graemlins/wink.gif[/img]

Also, it calculates the time need to play, at 60 hands an hour playing 40 hours a week, 10 Gigahands takes 8012+ years.

[Leader]

[ QUOTE ]
These horror stories may be of interest:

1200 BB Downswing

1500 BB Downswing

Downswing reaches 2100 BB after original post of 1500 BB Downswing

[/ QUOTE ]

These guy don't look like they're playing there best games or really even close. The first guy says he plays 6-10 tables. Josh says this:

[ QUOTE ]
No pots were won (4 tabling) while writing this post. I just reread the post. There is one small exaggerations in it that i don't feel like correcting.

[/ QUOTE ]

So he's playing 4 tables and surfing 2+2, which isn't good. It's really impossible to play your A game while playing too many tables or playing distracted.

[milesdyson]

was there any conclusion from josh's "are poker winrates normal distributions?" thread?

[kapw7]

[ QUOTE ]
If you drop down in limits say after losing 150 BB you can reduce your bankroll requirement quite a bit.

[/ QUOTE ]

I'd say 150BB is too conservative for short-handed 5/10+ games. It also costs money to drop down when you don't have to, i.e. you are still a winning player.
However it's better to be conservative than go on and lose 2100BB because you think you are the best.

[adios]

One thing I've noticed is that these horrendous downswings do seem to occur as a result of multi tabling. If you're playing badly, maybe tilting, then it's multiplied by that many tables. I'm fairly certain though that a bankroll of 300 BB won't cut it for win rates of 1.5 BB/100.

[Vern]

[ QUOTE ]
was there any conclusion from josh's "are poker winrates normal distributions?" thread?

[/ QUOTE ]
This program was made before that, it is just a normalized distribution probability sim based on an assumed true win rate and SD value for a series of hands. While there is an argument to be made for skill reducing the magnitude of probability down swings, I think the variance calculator is a nice visual tool to see how chance can affect an outcome and that chance is reduced as the sample size increases.

Edit: The biggest thing I got from that thread and others is it would be impossible to play a large sample size at the same underline win rate because personal skill and opponent skill levels change over the sample. Any probability sample large enough to have trends override chance, is too large to believe anybody besides a bot could play that consistantly the same and only against similar bots would the competition not change during the sample period.

The variance calculator was developed just to see whether the claims of probability causing massive down swings was an excuse or plausable. I think after viewing the results I think it is plausable, but will not pin the entire fault of any massive down swing on probability. It did help me stop caring about any downswing less than 100BB. At 100BB I start to make sure I didn't change or react differently and contribute to the down swing, spend more time reading and reviewing. That helps me mentally and it helps my play just in case my play level was falling off.

[Guruman]

if you don't have to make periodic withdrawls for living expenses you should probably try to cap your roll in terms of bb and start moving up and down more aggressively.

I could continue my entire poker career on 100 bb if I decided to move down every time I dropped to 50. BR management is really move down management. you only go busto if you want to.

believe that.

[kahntrutahn]

As a high stakes pro, 1kBB+ for shorthanded limit is essential. I offer as evidence (and I feel that not one bit of tilt entered into my play)....