Bankroll Definition and Recommendation for No-Limit Hold 'em

[Matt Flynn]

Hi all,

This debate could get interesting. I will start.


I have strong bias about the definition of bankroll. My bankroll is what keeps me in action. That is entirely money that I've earned playing poker. A lot of money flows irrevocably each month from bankroll to house funds. In addition, bankroll has loaned the house money - specifically to pay down a HELOC when the HELOC was in use. House gets the interest savings for free as part of the flow from bankroll to house. But that's a loan. If I go bust, I write a check from House the bankroll to cover the loans.

If I go poker broke, I would not take more than maybe $500 a month out of income to play poker. I would treat it just like a golf hobby. After a few months of losing I would likely quit save for the occasional small-stakes foray. Actually not quite, because several people would stake me. That's important, because I could rebuild far faster with staking.

So in return I ask what is a pro? In effect I have a partitioned financial life: working man and poker pro. Like any pro, if I lose my bankroll, I face a full rebuild. Unlike a traditional no-job-having pro, I do not risk my livelihood or have to get a job and change careers if I go poker broke.

So for me, bankroll is that money which keeps me in action plus the goodwill/respect that would get me staked. I would not borrow, so that sector's out.

For a regular poker pro, I would define bankroll as whatever keeps you from having to get a job. That's clearly whatever funds you would expend and loans you would take before getting a job. Add the goodwill/respect of backers and potential backers.

Some have mentioned the difference between liquid bankroll and illiquid / non-bankroll money. That's less interesting to me. Everything's convertible anyway, the only question is the transaction cost in time and money.

The loan question is very interesting. How much would you borrow? How much stress would that put you under? How much would you give up in acceleration of rise in limits to avoid that stress?


Does that match your definition of bankroll?



Dean asked what my bankroll recommendations were. I am not at the recommendation stage. I'm at the musing stage. Before I go into those musings, let me warn you that I'm in the conservative camp. Your average successful poker player imo underestimates variance, overestimates his advantage, and does not adjust his bankroll requirments enough for tilt and other negative tendencies.

No one can "know" what a player's variance is. EV and variance vary with every game you play - and in some cases quite widely from game to game and circumstance to circumstance. There is no mathematically correct answer to the bankroll question. We're all making guesses and assigning comfort thresholds. However, we might rationally attempt to bound variance and hence bound bankroll requirements. But not in this post.

btw, in such situations, it pays to attack the other guy's assumptions rather than present your own. Hence the paucity of discussions of bankroll requirements for no-limit.

One way to approach it is to declare you need X number of buyins for the level you are willing to drop down to once you go "broke" at the higher level. So first we have to define your willingness to move down. If you are 100% able to move down and will continue to move down to each lower level until you hit zero, then assuming you are a decently winning player and can expect to win more per 100 as you drop down (I'm making a lot of easily violated assumptions here), a hard floor of 20 buyins for the lower level should be enough unless you habitually play heads up, hyperaggressive / LAG, etc. Even then you should be ok, but you must be a winning player and not screw that up by getting in several heads up games with better players, tilting, not quiting when you lose your concentration, etc. The higher your burn rate / depletion for expenses relative to your bankroll, the higher that number of buyins should be adjusted. Never play with the rent money.

If levels double, a hard floor of 20 buyins gives you 10 buyins at each level. So you might instead choose 30 to give you 15 buyins at each level.

I suspect a player who does not tilt, always drops down, avoids negative EV situations well, and quits when playing badly will do better with 20 buyins as a hard floor than an average successful poker player would do with 30 buyins as a hard floor. So I'm encouraged to advocate 30 as a hard floor and stress repeatedly that tiltlessness, concentration, control, ability to quit, gambool aversion, and similar "soft" skills have a radical affect on bankroll requirements.

I further suspect those who do best with those soft skills will likely "like" 30 as a hard floor, while those who don't will prefer 20 as the hard floor and face order-of-magnitude greater chances of busting.


In internet terms at 1PTBB/100 and 500 hands/hour we're talking 200 hours to move up each level with the 20 buyin hard floor. Two months to double for a pro. $1-$2 to $25-$50 in less than a year if you learn quick. A year and a half for 30 buyins assuming no withdrawals and no big rushes. If you win more, you get to move up quicker.


As an aside, I'd recommend jumping up right after a big win and strongly recommend taking a few days off immediately after going level-broke at any level. They are the cheapest vacation days you will ever take.



Your turn.


Matt