Bankroll Management: Never Go Broke

Why winning players still go broke

Here is the least intuitive fact in poker: you can make consistently good decisions and still lose for weeks. Short-term results are dominated by variance — the statistical noise of cards falling how they fall. A genuine winner playing stakes their bankroll can't absorb will, sooner or later, hit a routine downswing that wipes them out before their edge has time to show up. Going broke isn't proof you played badly; playing stakes your roll can't survive is the bad play. Bankroll management is how your skill gets the sample size it needs.

The buy-in rules

Your bankroll is money set aside only for poker, measured in buy-ins for your stake (one cash buy-in = 100 big blinds). Widely used guidelines, sized to survive routine downswings:

• •Cash games: 30-50 buy-ins. At NL10 (a $10 buy-in), that's a $300-$500 roll. Downswings of 10-20 buy-ins are routine for winners — see the downswing article below for the math.
• •Tournaments: 100+ buy-ins. MTT variance is far more brutal: most tournaments you enter, even played well, pay nothing.
• •Learning a new format? Use the conservative end. Your winrate is unproven, and unproven winrates might be negative.

What a downswing actually looks like

'Routine downswing' is easy to say and hard to believe until you see the numbers. We simulated 100,000 careers of 100,000 hands each — 70 billion hands — for win rates from 1 to 10bb/100, drawing every hand from a normal distribution with a standard deviation of 100bb/100, which is typical for 6-max no-limit cash. For a genuine 3bb/100 winner the results are sobering: 81% hit a 20+ buy-in downswing somewhere in that career, 39% hit 30+ buy-ins, and the median player's worst stretch runs 27 buy-ins deep. Now re-read the buy-in rule with those figures in mind — 30-50 buy-ins isn't caution, it's roughly the minimum that survives the median outcome for a *winning* player. The full dataset is linked below.

Moving up — and moving down

The rules work in both directions. Reach 40-50 buy-ins for the next stake and you've earned a shot at it. But set the retreat trigger before you move: if the shot drops you back to ~30 buy-ins of the higher stake, move back down. No shame, no tilt, no 'one more session to get even'. Every long-term winner has moved down; the players who refused are the cautionary tales. Moving down in stakes and rebuilding is the skill — going broke with pride intact is not.

Poker money is not rent money

The bankroll must be fully separate from life money — a dedicated amount you could, in the worst case, lose entirely without touching rent, bills, or savings. The moment poker money and life money mix, every decision degrades: you play scared with money you can't lose (which is unprofitable), or you chase losses with money that was never for poker (which is dangerous). One account, one number, one rule: poker plays only against the poker roll.

Track everything, honestly

Every session goes in the record: date, stake, hours, result. Including — especially — the ugly ones. Players who 'forget' losing sessions aren't protecting their confidence; they're corrupting the only dataset that can tell them whether they're winning at all, whether that stake is beatable for them, and when a downswing is deep enough to trigger the move-down rule. Your tracked results are the instrument panel for every bankroll decision in this lesson. Fly by instruments, not by feel.

Drill: set your plan in the tool

Open the bankroll manager and set your actual numbers: current roll, current stake, and the buy-in thresholds for moving up and moving down. Write the two trigger numbers somewhere you'll see them mid-session. A plan set while calm is the only thing that reliably beats a decision made while stuck three buy-ins at midnight.