Why bankroll management matters (the math behind it)
The single biggest reason competent poker players go broke is not bad play. It is treating a 5BB/100 win rate as if it were guaranteed income, when the standard deviation around that win rate is roughly 100BB/100. That gap — the difference between expected value and what you actually realize over a finite sample — is variance. Variance is huge. Even genuinely +EV players lose for months at a time.
The point of bankroll management is not to maximize profit. The point is survival. A winning player with too little bankroll will go broke during a normal downswing, with mathematical certainty. A winning player with adequate bankroll will ride out the same downswing and compound. Most of the difference between players who make it and players who do not has nothing to do with skill — it has to do with whether their bankroll was big enough to absorb variance long enough for skill to assert itself.
Here is the loose framework everything else in this guide flows from. If your standard deviation is X over a session, and your win rate is Y per session, the probability you go broke given a starting bankroll of B drops exponentially in B/X. Doubling your bankroll relative to your standard deviation does not halve your risk of ruin — it cuts it by an order of magnitude. This is why "how many buy-ins" matters so much more than "what is my win rate."
How many buy-ins do you actually need?
These are working numbers used by professional and serious recreational players. They are starting points, not laws. Adjust up if your win rate is unproven or your format is high-variance. Adjust down only if you have a large realized sample (50k+ hands cash; 1,000+ tournaments) confirming a strong edge and the ability to move down without ego.
Live cash: 20–40 buy-ins
For $2/$5 with a $500 max buy-in, that is $10,000–$20,000. 20 buy-ins is aggressive. It works if you are confidently winning, willing to move down quickly during downswings, and not relying on poker as your only income. 30 buy-ins is the moderate baseline — recommended for most live cash players. 40+ buy-ins is professional-grade conservative and appropriate if poker income pays your rent.
Online cash: 30–50 buy-ins
Online cash plays more hands per hour, so variance compounds faster. 30 buy-ins is the floor; 50 is comfortable. NL100 ($100 max buy-in) → $3,000–$5,000. NL200 → $6,000–$10,000. Add a buffer if you are six-tabling, since multi-tabling increases variance per hour.
Live tournaments: 100+ buy-ins
Tournament variance dwarfs cash variance because most of your equity is in the top three finishing positions, and you reach those positions in roughly 1 in 10 events. For a $300 live MTT circuit grinder, that means a $30,000+ bankroll. For a $1,500 series-trip player, $150,000+. Players new to MTTs systematically underestimate this and go broke chasing big scores during inevitable five-month dry streaks.
Online tournaments: 200–300 buy-ins
Online MTTs have larger fields, higher rake, and much higher variance than live MTTs. 200 buy-ins is the absolute floor for anyone playing online MTTs as a primary game. 300 buy-ins is professional-grade. For a $33 online MTT regular, plan on $6,600–$10,000. For a $109 mid-stakes regular, $22,000–$33,000.
Mixed games, PLO, and PLO5: higher variance, more buy-ins
Pot-Limit Omaha has variance roughly 2× No-Limit Hold'em. PLO5 is higher still. Mixed games (H.O.R.S.E., 8-game) carry the variance of their highest-variance component (usually PLO or stud variants). Add 30–50% to the cash-game buy-in baseline for these formats. 30 buy-ins of NLHE becomes 40–50 buy-ins of PLO at the same stake.
Conservative vs aggressive bankroll strategies
The conservative-aggressive axis is real and usually misunderstood. Aggressive bankroll management is not "play scared at higher stakes than your roll allows." It is willingness to move up faster after proving an edge, with a plan to move back down immediately if you fall below threshold. Aggressive bankroll management requires more discipline, not less.
Conservative bankroll management means thicker buffers, slower moves up, and more tolerance for losing months at the same stake. It is appropriate for players whose poker income is essential, whose mental game suffers at higher stakes, or who are still building a realized sample at their current level.
Most players err on the aggressive side without realizing it — they call themselves "conservative" but in practice play stakes their realized win rate cannot support. The honesty check: if your bankroll dropped 25% tomorrow, would you move down? If no, you are aggressive. If yes, you are conservative. Both can work. Neither works without an explicit move-down rule.
When to move up in stakes
Three conditions should all be true before you move up:
- Proven win rate over a meaningful sample. For cash, that means 50,000+ hands at the current stake — not 10,000. For tournaments, 1,000+ events of the same size and structure. Smaller samples produce false positives. (See sample size.)
- Enough buy-ins for the new stake, with a 1–2 buy-in buffer. If $2/$5 requires 30 buy-ins ($15,000) and you have $14,000, you are not ready. If you have $17,000, you have a 4-buy-in buffer — fine. The buffer matters because the first downswing at the new stake always lands harder than expected.
- Emotionally ready to lose at the new stake without tilt. The "scared money" effect is real. If $1,000 swings feel different than $400 swings did at the previous stake, you will play worse. Some players move up and play down to a lower edge purely because the money feels bigger.
When to move down in stakes (the hardest decision in poker)
Moving down is harder than moving up because it punctures ego. The rule is simple and unforgiving: if you drop below the minimum buy-in count for your current stake, you move down. Not "after one more session." Not "after this MTT." Not "after I run good for a week." Immediately.
Example. You play $2/$5 cash at 30 buy-ins ($15,000). After a $3,500 downswing, your bankroll is $11,500 — 23 buy-ins at $2/$5, but 38 buy-ins at $1/$3 ($300 max). Drop to $1/$3 today. Rebuild to 30+ buy-ins of $2/$5 before moving back up. The expected hit to your hourly is real, but small. The expected hit to your bankroll from not moving down is ruin.
Players who reliably move down on schedule outlast players who refuse to. This is the single highest-EV bankroll decision available, and it costs nothing but ego.
Risk of ruin explained
Risk of ruin is the probability you go broke given your win rate, your standard deviation, and your current bankroll size. The formula approximation (for symmetric distributions and constant stake):
RoR ≈ exp(−2 × WR × BR / SD²), where WR is win rate in BB/100, BR is bankroll in BB, and SD is standard deviation in BB/100.
Three practical takeaways:
- A 5% risk of ruin is the typical professional target. Lower than that, you are over-rolled and leaving EV on the table. Higher than 10%, you are dangerously under-rolled.
- Doubling your bankroll relative to SD cuts risk of ruin by an order of magnitude, not by half. Small bankroll changes have huge risk-of-ruin effects.
- SD is the variable most players ignore. A 5BB/100 winner with 70BB/100 SD is in totally different territory than a 5BB/100 winner with 130BB/100 SD (PLO, mixed games, big-field MTTs). Calculating risk of ruin without measuring your actual SD is fiction.
StackEdge measures your realized SD automatically from your session log and updates your risk-of-ruin estimate every time you end a session. Try the variance calculator.
Bankroll management for professional vs recreational players
Professional players need more buy-ins, not fewer, because their poker income pays rent and they cannot easily replenish from outside sources. A professional cash regular should run 40+ buy-ins minimum and ideally keep 60+ during stable periods. Professional MTT players should run 300+ buy-ins online and 150+ live, with explicit move-down rules at 70% and 50% of starting bankroll.
Recreational players have an external income source, which paradoxically lets them run with fewer buy-ins — because going broke means a re-deposit, not unemployment. 20 buy-ins cash, 100 buy-ins tournaments is fine for a recreational player who can replenish from salary. The catch: if you redeposit too easily, you stop respecting bankroll management, and your edge erodes from emotional decisions you would not make if the bankroll was the real constraint.
The 5 most common bankroll management mistakes
- Moving up after a hot run, refusing to move down during the downswing. This is the most common, most expensive mistake. The hot run was variance, not skill growth. The downswing is the regression. Move down.
- Treating tournament bankroll like cash bankroll. 30 buy-ins of $300 MTTs ($9,000) is bankroll-suicidal. The same player would never play $20/$40 cash on $9,000.
- Not having a separate poker bankroll at all. Mixing poker money with rent or savings is how careers end. Use a separate account, separate cash, separate mental ledger.
- Underestimating variance in PLO and mixed games. PLO at the same stake as NLHE needs roughly 50% more buy-ins. Mixed games inherit the highest-variance component.
- Ignoring sample size. 8,000 hands at a stake is not a sample. 30 tournaments is not a sample. Trusting a "+10BB/100 winrate over 5k hands" leads to moving up on a player who is actually a small loser.
Bankroll management with sold action and staking
Sold action changes the math of bankroll management. If you sell 50% of a $1,000 tournament with 20% markup, your backers pay $500 × 1.2 = $600 for 50% of the action, leaving you a net cost of $1,000 − $600 = $400 for 50% ownership. Bankroll management applies to your net cost ($400), not the gross buy-in ($1,000).
This lets professional MTT players play larger fields than their raw bankroll could support, by selling some action and keeping the rest. It is also how some players end up extremely over-leveraged — by treating every tournament as a discounted purchase rather than capital at risk. The honest bankroll calculation is: your net buy-in × the same 200–300 buy-in tournament multiplier. If a $400 net buy-in needs 200 buy-ins of bankroll, that is $80,000 — even though you "only" play $1,000 tournaments.
StackEdge models this automatically in the sold-action module. Per-backer ROI calculates from net invested, and your bankroll updates against your net cost per session.
How to actually track your bankroll
A good bankroll system requires three things, in this order: a session log, an auto-updating balance, and risk-of-ruin awareness. Most people stop at the session log (usually in a spreadsheet) and never close the loop.
The minimum manual system:
- Log every session immediately after you end it.
- Update your bankroll total after every session.
- Recompute risk of ruin at least monthly.
- Track your realized win rate and standard deviation by stake.
- Have a written move-up and move-down threshold.
StackEdge automates all five. The live timer captures the session, the bankroll updates instantly, risk of ruin recalculates with every new data point, your metrics break down by stake automatically, and the app surfaces stake-up and stake-down recommendations from your own data — not from generic guidelines.