Players must be 18+. Real-money play carries financial risk. Nothing here is legal or financial advice, and offshore access sits in a legal grey zone you weigh yourself. Read the responsible-gambling page.

poker bankroll management

Poker Bankroll Management: The Math That Keeps You Playing

Last updated November 2026

---

Poker Bankroll Management: The Math That Keeps You Playing

Poker bankroll management is the difference between a losing streak that stings and a losing streak that ends your game. A bankroll is money set aside for poker, separate from rent, bills and savings, sized so that normal variance cannot break you. Skill decides whether you win at poker. Bankroll decides whether you stay in the game long enough for skill to show.

This is the maths page of our strategy hub, and it earns that label. The rules below are not folklore; they come from how losing streaks actually distribute across hands and events. We work one full example step by step, then turn it into practical gates for moving up and moving down. Everything assumes you play online, in general at micro or low stakes, in Texas Hold'em. If you are still choosing a game format, the cash game strategy page covers stakes from the table side first.

What counts as a bankroll

A bankroll is a dedicated pot of money with one job: buying action at your stakes.

  • Separate from living money. If losing it changes your rent, it is not a bankroll.
  • Sized in buy-ins, not in dollars. "$200" means little at NL25 and plenty at NL2.
  • Replenished by skill or deposit, never by chasing. Money deposited to recover yesterday's loss is not strategy, it is tilt with a payment method.
  • Held in something stable if you play on crypto rooms. Volatility in the coin you hold is a second bankroll problem stacked on the first. For players using USDT, our USDT poker deposits guide explains why most bankroll-focused players prefer it.

The buy-in rule

The standard advice reduces to one habit: carry enough buy-ins that a normal downswing stings, never breaks. The floors below are conventional ranges used by serious players, in general:

FormatAbsolute minimumComfortable standardWhy
Cash games (NLHE, 6-max)20 buy-ins40–50 buy-insModerate swings; downswings of 15–20 buy-ins happen to skilled players too
Multi-table tournaments100 buy-ins150–200 buy-insVery wide swings; variance commonly runs 10–15 buy-ins per event
Sit-and-gos50 buy-ins75–100 buy-insSwingy, but narrower payouts than MTTs
Freerolls00Practice only; no full buy-in to lose

Read the cash row carefully before arguing with it. Twenty buy-ins is the survival floor for a proven player who withdraws regularly. Forty to fifty is where the risk of ruin drops to tolerable levels for everyone else, most of all for players still proving themselves at their stake.

How the rule scales in practice

Applying the standard to a cash-appropriate ladder:

StakeBlindsOne buy-in (100bb)Bankroll at 40 buy-ins
NL2$0.01/$0.02$2$80
NL5$0.02/$0.05$5$200
NL10$0.05/$0.10$10$400
NL25$0.10/$0.25$25$1,000
NL50$0.25/$0.50$50$2,000

For MTTs the same logic runs on entry fees. A player grinding $2.50 events needs an effective base of $250–$375 at the 100–150 buy-in floor (and, in general, a share of entries can come from cheap satellites to stretch the roll).

Worked variance example: why the multiplier exists

Here is the arithmetic behind every table above. The model: a player grinding NL5, blinds $0.02/$0.05, buy-in $5, bankroll $250, or 50 buy-ins.

Assumptions, stated plainly so you can swap your own:

  • Win rate: +12 bb per 100 hands. Treat this as the range solid winners post at micro stakes, generally speaking, not a promise.
  • Standard deviation: 85 bb per 100 hands. A commonly cited ballpark for 6-max cash play sits around 80–100 bb/100, so 85 is representative.

Step 1: expectation over sample. A busy month of 25,000 hands (about 60 hours at four tables) carries an expectation of 12 bb/100 × 250 = 3,000 bb. At NL5, one bb equals $0.05, so the expectation is $150.

Step 2: the spread around it. Standard deviation grows with the square root of hands played. Over 25,000 hands: 85 bb × √(25,000/100) = 85 × 15.8 ≈ 1,344 bb ≈ $67.

Step 3: what normal months look like. Around two thirds of identical-skill runs will land within one standard deviation of the $150 expectation, roughly between $83 and $217. A genuinely cold stretch, two standard deviations below expectation, swings around $134 below the mean. That swing alone, at 27 buy-ins, can flat-out consume a 20-buy-in bankroll even while the player is running exactly as skilled as ever.

Step 4: how quickly the sample shrinks the risk. Using the same model, the approximate chance of finishing a stretch below break-even:

Hands playedApprox. chance of a losing stretch
5,000~16%
10,000~8%
25,000~1%

The pattern is the whole lesson: skill bends the curve slowly, and volume compresses the variance. Neither step ever reaches zero, which is why every stake ladder above builds in a cushion rather than betting the roll can survive a two-standard-deviation quarter.

Two honest caveats: this normal approximation assumes a stable win rate and independent stretches, while real play drifts with tilt, table choice and changing opponents; and rake, which the model leaves out, eats thin margins at micro stakes. The example measures risk, not promise, and that is exactly how a bankroll rule should read.

Variance in tournaments, briefly

MTT swings dwarf cash swings, and the reason sits in the payout structure: most entries return nothing, so results concentrate in the top few per cent of finishes. Standard deviation in buy-ins per event, in general, commonly lands around 10 to 15 depending on field size and structure. Over a few hundred events, that compounds into swings of tens of buy-ins even for players posting solid ROI. That is the arithmetic behind the 100+ buy-in floor for MTT play, and why our tournament strategy page tells you to judge results in hundreds of events, not dozens.

Moving up: the gate check

Use three gates. All three, not any two.

  • Bankroll gate: at least 40 buy-ins for the new cash stake (or 150 for tournaments), counted without touching the funds that keep your current stake playable.
  • Sample gate: results at your current stake hold steady over a meaningful sample, in general several thousand hands of cash play or a few hundred events.
  • Skill gate: you can name what your last betting exercise taught you. If your study habits stopped when your graph rose, the sample gate is lying.

A common compromise when a gate is half-met: take shots, not leaps. Keep your base stake fully funded and set a defined shot capital of two to five buy-ins at the higher stake, held separately. Lose the shot capital, drop the shot, replenish the base roll first. Our cash game page covers stake ladders from the table side.

Moving down: triggers and attitude

Moving down is a repair step, not a verdict on your skills. Recognise the triggers:

  • Bankroll falls below roughly 30 buy-ins at cash stakes (around 60–75 buy-ins for tournaments, recalculated against the lower format's floor).
  • A downswing that keeps grinding after a fortnight of careful play, which usually signals leaks or tilt, not cards.
  • Life changed: hours down, budget down, focus elsewhere.

The procedure is boring by design: drop one or two stakes, rebuild the sample and roll, and treat the smaller games as a reset for your sizing discipline. Games two rungs below your peak, in our experience, expose exactly how much rust sits in your game; play them with full attention.

Withdrawals: skim or compound

Two standard policies, both legitimate:

  • Skim policy: keep the bankroll at target size and move out periodic surplus. Common practice is to withdraw about half of a good month's profit above the target roll.
  • Compound policy: leave profits in the roll and move up stakes on schedule faster.

Skimming protects you from re-losing profits and forces discipline against stake creep. Compounding raises your earning ceiling faster but also raises your exposure to variance. A reasonable blend: bank 40–50% of profits in general, play the rest, and review the blend quarterly.

Common bankroll mistakes

  • Playing with money that was never poker money. Rent and bankroll never share an account.
  • Skipping the sample. Twenty sessions tell you about cards far more than they tell you about skill.
  • Treating a win streak as a licence for stake jumps, right before variance gets its turn.
  • Chasing losses upward. Bigger stakes do not feed a tilted mind, they feed it.
  • Ignoring rake and fees. Micro-margins consumed by costs demand a bigger cushion, not a smaller one.
  • Holding bankroll in a volatile coin. The roll's job is stability; convert accordingly.

Your bankroll setup in five steps

  1. Pick your format and stake (freerolls, then NL2 or NL5, in general).
  2. Move in a deposit-sized cushion that covers the buy-in floor with room, using the ladder table above.
  3. Separate the money: dedicated wallet or account, living money untouched.
  4. Write your two rules on screen: minimum buy-ins and maximum session loss per day.
  5. Review monthly: hours, hands, results, and one leak to study. Then adjust stakes by the gates, not by mood.

The safest first steps pair a funded, correctly sized roll with rooms that suit our region's payment reality; our review of the best poker sites explains how we rank rooms for deposits, withdrawals and Arabic support.

FAQ

Q: What is the minimum bankroll for cash games? A: Twenty buy-ins is the conventional survival floor for proven players, while the comfortable standard runs 40–50. At NL10, that means roughly $200 if you play at the floor, and $400–500 for the comfortable standard.

Q: Why do tournaments need a much bigger bankroll? A: Because variance in MTTs, in general, sits far wider: most entries finish outside the money, so swings across hundreds of events commonly dwarf anything seen in cash play. The 100+ buy-in floor absorbs those swings.

Q: What is a buy-in? A: The amount of money you bring to one sitting: your full stack in cash play, or one tournament entry. Rules are quoted in buy-ins because that makes them portable across stakes.

Q: How much can I withdraw from a winning bankroll? A: Under a skim policy, in general about half of the profit above your target roll each month, leaving the base cushion untouched. Exact rates are a personal discipline decision, not a formula.

Q: When should I move up in stakes? A: When all three gates pass: bankroll covers the new stake's buy-in floor with breathing room, your sample holds, and your study habits are still running. If any gate fails, hold or move down.

Q: Does this guide cover bankroll in volatile coins? A: Yes, in one line plus context: volatility in the coin is an extra tune on the top of poker variance, so most bankroll players prefer stable-value coins. Our USDT poker deposits guide goes into the mechanics.

---

Last updated: November 2026 · Editorial review: Saudipoker team

Online poker is for adults 18 and over. Poker carries real financial risk, and no bankroll system removes it; never play with money you cannot afford to lose. If play stops being entertainment, our responsible gambling resources list practical tools.