I am just wondering what a good HU bankroll is for micro thru mid stakes! I am not a great HU player yet, but I feel I am good enough that I can hold my own and win a good amt of the time! I came up with this BR chart, how does it look?

Heads up BRM:
NLHE$2 thru NLHE$25 - 20x Buy Ins at 100BB/BI
NLHE$50-NLHE$100 - 25x Buy Ins at 100BB/BI
NLHE$200-NLHE$600 - 30x Buy Ins at 100BB/BI
HU SnGs - 30x Buy Ins

My goal is to start with $200 and withing a period of time get to where I can play NLHE$200 Regularly HU and SH!

I understand the rake at NLHE$2-NLHE$25 is pretty high, but it still can be micro profitable! Especially playing on AP/UB being able to get pts for cash and rakeback!

I don't play HU but those figures look way too low to me.

I mean, 20-30 BI is standard for 6-max, so you're surely going to need more for HU.

I'm guessing that if you treble the number of BI's at each level, you might be nearer the mark.

(Unless you're being uber-aggro with your bankroll and can easily reload if you end up in bustoville.)

Be interested to see what those who do play HU have to say.

Good Luck

EDIT: I will say that I'm a bankroll nit, but even so, your figures look too low to me.

Well all i play is HU SnG's with 20-40 buy ins,

Put $25 in my account, played $2 games till $100.
Then from $100+ i started $5 HU shootouts, which i got up to $400 and today ive started $10 games which i will play until i get over $600 then i will move to $20.

So far its going all good, in 1 1/2 weeks ive got my roll over $500 from $25.

For me 20-40 is fine but im crushing my levels so it might be different for other people as my ROI on $5 is 45%.

So again it depends on how much your winning, your swings etc.
Ive been lucky so far that 20-40 roll has been pretty perfect for me, im guessing though the higher i get the bigger i might have to make my BRM.

Thats my experiance so good luck.

20-30 is fine for HUnl - but you need experiance, this question all depends on that

40 Buyin for HU cash. I wouldn't do anything below that.
25-30 for SnG's before you move up.

at the lower stakes ..(50s and lower) you are actually losing a lot of money if you wait to make 25 buyins per level before moving up. up to about 30s you should start taking shots at 10 buyins ..as long as you move down as soon as you drop below 10.

i pretty much went from 2s and moved up at 10 buyins.. even with a few sick downswings it worked out pretty well. took a few months to move up to 30s

obviously im assuming the player is competent and understands basic husng strat



EDIT: however this should depend on how comfortable you feel.. theres no point in taking 10buyin shots if its gonna make you play bad or scared. keep in mind also that to take a aggressive BRM like what i did you need to know that you can replenish your bankroll.. when your bankroll is too big to replenish then going to as much as 50 buyins is fine.

Im using 20-40 buy in rule and im at $10 level atm with my $5 level at 45% ROI and ive won 6 out of 8 of my first $10 games.

So you think i shouldnt sit around and grind up to 20+ BI's and take shots instead once ive won 10+ BI.

To be safe, I'd go with at least 35 BI's. That's my opinion. HU can be totally brutal with swings.

nothing wrong with taking shots as long as you are mentally prepared to lose a small portion and grind it back.

once you have about 200-250 then i would suggest just give it a go at 20s .. honestly the play is the same as 10s ..

its all really a matter of what you are comfortable with.

HU is one form of poker I never play and I often hear this stated but it just doesn't make sense to me. All things being equal you should generally need more buy ins as the field of players grows so I would think the lowest amount of buy ins would be required for heads up while the maximum amount of buy ins would be needed if you played mtt's.

You may be right, Rek. You usually are. (And I mean that sincerely.)

I must admit, though, I've always assumed that more players -->> smaller bankroll requirement.

Like I said, I've never played HU but I've always assumed that you need to be prepared to get the money in with much thinner edges, otherwise you're likely to get run over.

HU, if I have bottom two pair on the river and villain shoves, folding every time is almost certainly going to cost us money.

In a 6-max game, calling is not quite such an easy decision.

In a FR game, we can probably lay it down, no problem.

We'd generally expect players in a FR game to enter pots with stronger hands than in 6-max, and 6-max players to do so with stronger hands than when they play HU. (These are all horrible generalisations but you get my point.)

So, HU, if we need to get it in thinner, the variance is going to be worse than in 6-max/FR. Hence, we need a bigger bankroll.

(Similar thing happens in Limit hold'em. You're constantly pushing thinner edges, so downswings can be way more brutal.)

As always, it's going to be villain dependent, of course, but as a general principle, seems to me that the more opponents we have, the bigger edge we want before we start putting our stack in the middle.

On a completely different note, some of the guys replying to OP are saying that if he can reload easily, then 20 BIs (maybe even less) is sufficient. Let's face it, if you can reload easily then 1 BI is enough.
Then we're back on to the old argument about "if you can reload easily, it's not a bankroll, it's just the amount you choose to have online at any particular time."

I repeat, I've never played HU, so these are just the idle ramblings of someone who's interested in this stuff.

no rek is wrong - hes confusing the higher varience of bigger more top heavy tourney structures to cash games which have fixed payouts and thus varience depends on how light one gets money in.

OK, well that makes sense.

Bigger the tournament field -> longer downswings -> bigger bankroll needed.

I think Rek plays a lot of tournaments, so maybe that's the reason we seemed to be disagreeing.

As always, never having played a tournament in my life, I'm looking at the world wearing my cash-game blinkers.

I'm just *starting* to think about this question, and I'll probably think about it more on the subway ride home. I'm sure someone here already has the answer all figured out, and knows all the related math.

Anyhow, here's my first instinct. I think there may be contrary forces at work. On the one hand:

-Heads up should have less variance than full ring because *every swing occurs one hand at a time.* That is, swings come by winning hand after hand after hand, or losing hand after hand after hand. In full ring, your hands are contested multiway, (especially if you consider the preflop game to be part of the contest). Those hands are all basically long shots with big payoffs. Whereas a heads up contest is closer to even money betting at even odds. On the other hand:

-Heads up should have more variance than full ring if the structure of the game fundamentally forces players to stake large volumes of money, relative to stake size, on closer decisions, rather than giving an opportunity for large edges. That is, *if* your hand per hand ev is +0.4% in typical heads up contests, but it's +2.5% in a ring game (purely fabricated numbers), or if the pots tend to routinely involve more money because there are fewer folds preflop/postflop, then we expect big swings.

Overall, though, I tend to expect the first factor to be more important than the second factor... that is, if you're per session ROI is roughly similar, then I instinctively expect that heads up should be less variance.

Does this make any sense at all? Maybe not... just silly ramblings perhaps.

its better to be imperical, just compare a 30k graph of a hu player, and a 30k graph of a fullring player / question solved.

Yes I was

Yes I must have skimmed the OP and he mentions at the end about HU sng's and in my mind I was thinking he was talking about sng's only.

My bad

Makes sense.

Still I'm hung up on the theory now. Also, I stopped myself to come back and correct myself. I think I conflated the ideas of "variance" and "risk of ruin."

Now, I think it's more correct that "Risk of Ruin" is proportional (exponentially) to one's expectation, but that "Variance" is a factor of the game structure and the style of play employed. Right?

So anyway, I then found myself trying to model some weird visualization of risk of ruin in my head. I imagine a bubble, floating some distance above the ground. As time passes, the bubble grows at a rate dictated by the variance. At the same time, it is rising or falling at a rate dictated by the player's expectation. The more the bubble intersects the ground, the greater chance of ruin at any given time. If a player has negative expectation, eventually the whole bubble passes through the ground and he's surely broke. If his expectation is positive enough, the bubble eventually rises high enough to make ruin an impossibility... unless the bubble is growing faster than rising, in which case there is never a 100% safe time, just reduced risk.

The starting altitude, of course, is the size of the bankroll relative to stakes.

Crazy... and believe it or not, I'm one of those prudes who keep telling players not to get high when playing.

Sorry to hijack, but obviously my bubble model needs refinement. Perhaps closer to correct but harder to visualize:

-The bubble must always be expanding faster than the center is moving. In this way, there will always be a part of the bubble above the floor, and a part below.

-Maybe some gravity needs to be introduced. So a person with a 0 ev will have a falling bubble.

Reasons: The chance that you will broke at some time in the future must always be at least fractionally higher than any time previous. This chance approaches some sort of limit as we move towards the infinite future. But I imagine that limit, for most evs is 100%. Certainly it should approach 100% for 0 ev.

Aww, hell... it's still flawed. I just realized that a 0 ev AND 0 variance should look like an unmoving point. Damn, I should have studied math or something.

Stop the presses!!! My bubble is getting stranger. (And I'm sure you're all eagerly awaiting the latest installment of my insanity).

No Gravity! Scratch that idea. Bad idea.

A positive ev is rising, a negative ev is falling, 0 ev is standing still.

The bubble's expansion is a factor of variance and (the absolute value of ev)+1. So, if the variance is zero, the bubble is always a point that doesn't grow. If there's variance, the bubble's radius increases faster than the the center moves. So the "leading edge" of the bubble moves relatively quickly, and the "trailing edge" retreats releatively slowly.

Touching the floor is bad! So, the chance of going broke is a factor of how much of the bubble has passed through the floor, multiplied by the square root of the time.

"Time" is itself a complicated factor. I figure it should be measured in bets made, divided by the number of bets in the initial bankroll. So, if a person starts with a $1000 bankroll, and makes $1 bets, then after 1000 bets the time is "1," and after 2000 bets, the time is "2." But any time less than 1 is zero, because it's impossible to bust a $1000 roll when you've only bet $999 of it.

I'm sure there's some other constraint, but I don't want to get TOO abstract (however, if I did, I'd have to imagine the bubble distorting in shape, starting as a vertical line but bulging more around the middle over time... and now it's time to stop thinking and start playing poker (real formula for success there)).

Note also, this bubble is only expanding as we look deeper into the future. In the present, it's always a point. And Boetheus would argue that, to God, everyone's bubble is reduced to a point at all times, past present, and future. DUCY?

P.S. This is not a level.

What's all this talk of bubbles? What I want to know is:

1. how big is a bubble?
2. what colour is a bubble?
3. where does a bubble go when it bursts?

btw I love bubble and squeak.

The third question is most philosophical, but I reckon the answer is "somewhere close to Hell." As for question 2, the answer is "all of them." Question 1.... calculating... calculating... alright, I'll get back to you in a month.

So do you guys think this would work? To be honest I do pretty well in microstakes HU cash and SnG on Tilt and UB/AP, small stakes I do good in also! Its when i get to midstakes when i start having trouble with variance, so what about this:

Heads up BRM:
NLHE$2 thru NLHE$50 - 25x Buy Ins at 100BB/BI
NLHE$100-NLHE$200 - 30x Buy Ins at 100BB/BI
NLHE$400-NLHE$1000 - 35x Buy Ins at 100BB/BI
HU SnGs - 30x Buy Ins

I really think i will not experience too many swings at NLHE$2-NLHE$50, but with NLHE$100 to NLHE$200 is where i do experience most variance and swings, so I feel maybe upping it 5-7 Buy Ins may be good? Then Obv with NLHE$400-NLHE$1000 im gonna want 35-40 buy ins! I also do pretty good on HU SnGs on tilt so i think 30-35 BIs would do great!

I wish you would keep your comments relevant to the thread. We are talking about bubbles here.

I hate to spoil the spoil, but that's correct.

Though what I said earlier may seem rather silly, if you do need a bigger bankroll for higher stakes, it's because your expectation is less positive there, not because of increased variance. That is, lower expectation means greater risk of ruin. And if your expectation is negative, no bankroll is big enough.

It is not necessarily that my expectation is negative and will also be negative, it is because the better players are more alert/logical and have better strategy, tricks and move! I feel that once i move up and get my practice vs the better players then i will be good enough! Sorry about the mistake between variance and swings! My swings are bigger in mid stakes games then in micro-low stakes, so i feel the few extra BIs will get good!