Most players who quit poker did not lose because they were bad at it. They lost because they played a limit their bankroll could not support, ran into a normal losing stretch, and had nothing left when it ended.
Variance in poker runs longer than intuition allows. A winning player can lose across a sample that feels far too long to be luck, and the only defence against that is having enough behind you to still be playing when it turns.
How many buy-ins you actually need
The requirement changes with the format, and the differences are large enough that one number cannot cover them.
For no-limit hold'em cash at soft low limits, around 40 buy-ins is a workable floor. That is not conservative advice, it is close to the minimum at which a normal downswing does not end you.
Pot-Limit Omaha needs considerably more, and 100 buy-ins is not excessive. Equities run much closer in PLO, so pots go to showdown more often and the swings are bigger even when the field is worse than a hold'em field at the same stake.
Tournaments need the most by a distance. Cashing in roughly one event in five is normal for a winning player, and the money is concentrated in rare deep runs. A hundred buy-ins is a starting point rather than a comfortable cushion.
Short-stacked or shallow formats need less, since your maximum loss per hand is capped. Deep-stacked games need more for the same reason in reverse.
What moves the number up or down
Two things beyond format.
The first is how soft the game is. A high winrate shortens downswings, so a genuinely soft field lets you play a little closer to the line. The catch is that most players overestimate their winrate, and a bankroll built on an optimistic estimate fails exactly when you needed it.
The second, and it is the one people ignore, is what happens if you lose it. A player with an income who can rebuild in a month can run thinner than a player whose bankroll is their rent. That is not a poker question, but it changes the correct answer more than any variance calculation.
Rakeback changes the arithmetic
At low and mid stakes, the deal you play on is a real component of your winrate rather than a bonus on top of it.
Two players with identical strategy can finish a year on opposite sides of zero because one was getting a share of their rake back and the other was not. Since bankroll requirements scale with winrate, a better deal genuinely lowers the number of buy-ins you need for the same level of safety.
This is why serious low-stakes players care more about their arrangement than about their preflop chart, and why it is worth asking what your deal actually is before concluding the limit is unbeatable.
Moving up, and moving back down
Move up when you have the buy-ins for the next level and a sample large enough to mean something. A good week is not a sample. A month at reasonable volume is closer.
Shot-taking works if the shot has rules written before it starts: how many buy-ins you are risking at the higher limit, and at what point you return. Without those two numbers, a shot is just playing too high with extra steps.
Moving back down is the harder discipline and the more important one. Dropping a limit after losing part of your roll is not a failure, it is the mechanism that keeps you in the game. Players who refuse to move down are the ones who stop playing entirely.
A worked example
You have $600 and want to play NLH cash. At $10 buy-ins that is 60 buy-ins, comfortably above the floor, so this is a reasonable starting point in a soft field.
You run badly and drop to $400. That is still 40 buy-ins at the same limit, so nothing needs to change yet. The plan survives a normal downswing without a decision.
You drop to $250. Now you are at 25 buy-ins and below the floor, and the correct move is down a limit rather than a change of strategy. Play $5 buy-ins until you are back above $400, then return.
The point is that all three decisions were made in advance, when nothing was at stake. That is the entire value of a bankroll rule.
What this looks like in our clubs
Across the apps we run clubs in, low limits play loose and passive, which raises winrates and shortens downswings compared with a tougher pool at the same stake. That genuinely allows a slightly thinner roll than a textbook aimed at a harder game.
The complication is chip value. Club apps price chips differently, and a stake that reads as small can be larger than it looks once converted. Confirm what one chip is worth in real money before you size a bankroll around it, and confirm it in writing.
The players who get into trouble in our clubs are almost never the ones who ran badly. They are the ones who sized a roll for one limit and then sat at a bigger game because it had an open seat.
Frequently asked questions
How many buy-ins do I need for cash games?
Around 40 as a floor for no-limit hold'em at soft low stakes, more if you tilt or plan to move up quickly. PLO needs roughly 100 because the swings are much larger.
Is 20 buy-ins enough?
Only if losing the lot costs you nothing and you can reload immediately. At 20 buy-ins a normal downswing takes a meaningful share of your roll, so you will be moving down often.
How big a bankroll do tournaments need?
A hundred buy-ins is a starting point rather than a safe number. Tournament results are concentrated in rare deep runs, so the stretches between them are long enough to end an underfunded player.
Should I withdraw winnings or leave them in?
Leave them in until you have comfortably cleared the requirement for your limit, then withdraw on a schedule rather than after good sessions. Withdrawing when you feel rich and reloading when you feel poor is how rolls quietly shrink.
Does rakeback really change how much I need?
Yes, because the requirement scales with winrate and rakeback is part of your winrate at low stakes. It does not make a losing player safe, but it measurably shortens downswings for a winning one.
Where to start
Work out how many buy-ins you currently have at the limit you play. If the answer is under 30, the useful change this week is the limit, not the strategy.
If you want to know what your rake and rakeback actually are before you size anything around them, message @supremaunion, or compare the apps we run clubs in first.
18+. Play within your means.