noblepoker do them, and actually do 3 different structures, normal, turbo and superturbo. The last one is 1 minute a blind, so is nutso.

Say I have 5k or 50 BI's for 100NL. Anything above I could take a shot @200NL (lord willing), anything below that, do I go back to 50NL with 4.9k that's 98 BI's for 50NL or do I keep playing 100NL until I (god forbid) drop to 2.5k for 50BI's @50NL?

Did I understand correctly?
Anybody use muad'dib on HU cash yet?

Is this a typo? Playing $10.50 at 100-200 and $21 at 200-250? Shouldn't it be $5.25 at 100-200 and $10.50 at 200-250?

For those who are mathematically inclined, you may decide on using the Kelly criterion. This approach allows you to quantify when to move up or down in stakes by comparing the rate of growth for the games of comparison.

This thread may be of interest.

http://forumserver.twoplustwo.com/15...-poker-424693/

The generalized Kelly bankroll for strategy games with a known win-rate and standard deviation is

B = s.d^2/(wr*k)

Typical standard deviation numbers for the 180 mtt's can be around 6 buy-ins per tournament. Of course, this depends on your playing style and to a lesser extent, your win-rate.

If your roi is 50%, and you're content playing at half-Kelly, then you require about 144 buy-ins. At full Kelly, it is simply half of that.

You will often hear people say that you need at least 100 buy-ins for mtt's. However, this depends on the size of the tournament as well which is often omitted.

Nice post diebitter. It sounds like a sound approach. Any feedback on whether or not people find this method to be effective (particularly in the long run)?

im playing turbos with a similar br strat. i just have the rule:

1) dont play a game thats more than 5% of ur br

started at the 22$ earlier this week and im soon good to go at the 88.

you just have to prepare yourself to move down.

How does this work compared to Kelly? Say you have 50BI for stake A, you get up to 60BI, and you decide to take a shot at stake B, which is 10x the stakes of A. If you lose, you start to play A again. Assume your ROI is like 5-10%, doesn't matter as long as it's improper to play at stake B according to Kelly.

Basically, "MuaD'ib" will have you move up less quickly than Kelly simply because Kelly is proven to be the fastest way to grow a bankroll, right? So ensuring you're rolled by the Kelly Criterion should get you up faster than following this approach.

But look at it this way, which system by name alone seem better..some Kelly who or Muad'Dib

+ some1 asks you what bankroll management you follow to which you reply Muad'Dib..you get 100 cool points (or nerd ones)

Hey DB: there is a gaping hole in your strategy: you cannot beat 1 or 2 dollar. I didn't read all this, but if you wonder why no one cares...

Thanks for the advices ! I'm gonna use that wisely

Hey

This is really interesting and i'm going to try and use it for PLO HU SNG but instead of 50BI i'll be using 100BI.

I'm a big fish at understanding things, but I THINK I get it, maybe I have misunderstood it so if anybody could let me know if I have done anything wrong please let me know, many thanks in return


Say our roll is $525 and we're using twice the requirements..


Seed Bankroll - $525

100 BI @ $5.25

(I play $5.25 until I have a bankroll for Step1?)


Step 1: Gamble Bankroll - $735

20 BI @ $10.50

Step 2: Gamble Bankroll - $945

20 BI @ $21



Seed Bankroll #2 - $1050

100BI @ $10.50

(I play $10.50 until I have a bankroll for Step1 or continue playing $21s?)


Step 1: Gamble Bankroll - $1260

20BI @ $21 .. and so on


Step 2: Gamble Bankroll - $1680

20BI @ $31.50


And so on..

Flawed thinking. In-game variance doesn't impact game-to-game swings. You could actually play PLO SNGs at a smaller bankroll since you can expect a bigger edge.

Yes it does? Hence you see all those insane -ev/+ev graphs at omaha cash. You get in alot more 55%/45% situations in omaha then in holdem; losing 0.45^x is a higher probability then losing 0.25^x.

holy mother of all bankroll nits!

100 buyins for $5s?

wtf?

see you in the hundreds in a few years...

oh wait, nm. my bad. 20 buyins isn't that bad.

Wait, no, you really do play to 100 buyins at 5s.

I have no idea what the general rule of thumb is for HUPLOSNG, there is no information anywhere on this and numerous people have told me anywhere from 40 - 150 is good.

Anyway, technically i'm using 20BI using this method unless i've hugely misunderstood the concept.

Well at SNGs if you have a 7% ROI and SD=1, you'll need the same bankroll restrictions, no matter whether the game played within the SNG is holdem, omaha, Chinese poker or darts. The in-game, chip variance may have to do with your ROI, but it won't affect the outside variance (the one that really counts).

Yeah, but I think the std deviation of Omaha is higher then Holdem. higher std deviation --> Need a bigger BR.

True, but that doesn't make a difference. Standard deviation of HUSNG is always 1 (you can either win one or lose one).

i didnt know DB actually played poker..

A higher ROI gives you less variance in a game though?
I'm having difficulties comprehending why omaha gives a higher ROI if it has more variance. How can a 30% ROI have a very variance ingame style yet his winrate is super high?

Just wondering whats the math behind your statement "husng is always 1", ofc 1 husng in practice, you can only win/lose, but what would the equation be like in theory? I thought it worked out like this: Taking a 60% win chance.
binominal, take n=1 , 1*0.6*0.4 = variance.
std deviaton = sqrt(0.6*0.4) = 0.489

When I put it into Excel, it gives me 1.05 for 50/50 and 1.03 for 60/40 over a sample of 10 games, which has obv nothing to do with the binomial formula, but it makes more sense (you obv can't lose or win 0.5 BI in one game unless you chopped).

BTW Omaha players can have bigger ROI since their villains are on average much worse than their holdem counterparts

Well the thing I'm wondering about.

Imagine person A: plays a non-variance style against donk C

Person B: plays a variance style against donk C


Shouldnt person B have a lower ROI? What defines a variance style? Does a style become a "variance style" because you go after low edges (barreling alot)? But going after low edges would decrease your ROI since well, going after 55%/45% situations would decrease your ROI instead of going after 80/20% situations.

cool I been looking for something 2 do with abit of gamble in it might give it a go post results in a weeks time.

is it possible to make this into a excel so i could put in my BR and it would tell me what levels to start and output the whole future roadmap like the chart we see in OP.

also voted this 5 stars