41 Bankroll Management & Risk of Ruin
In Chapter 9 we established the uncomfortable truth at the heart of poker: even a clearly winning player will endure long, vicious downswings purely from variance. A solid online cash-game grinder beating their stakes for, say, 5 big blinds per 100 hands will still run into stretches of 100 buy-ins of negative variance over a career. A good tournament player can — and statistically will — go six months or more without a meaningful cash. None of this means they are playing badly. It means the variance is real, and a single buy-in is a rounding error against it.
Bankroll management (BRM) is the discipline that lets your edge actually express itself before the variance bankrupts you. It is not a sexy topic. It wins no pots and earns no respect at the table. But it is, statistically, one of the highest-leverage skills in the entire game, because the alternative — going broke — sets your expected lifetime winnings to exactly zero no matter how good you are. A genius who busts is, financially, indistinguishable from a fish.
This chapter turns the variance math from Chapter 9 into concrete rules: how many buy-ins to keep, by format; what “risk of ruin” actually means; when to move up, when to move down, how to take shots; and — because this is ultimately a psychological discipline — how to keep your roll and your head separate from each other.
41.1 What a bankroll actually is
Your poker bankroll is money dedicated to poker and only poker. It is not your rent, not your emergency fund, not money you would feel sick to lose. The single most important rule in this chapter is the cleanest: separate the life roll from the poker roll.
The reason this matters is not moralistic. It is mechanical. The buy-in math below assumes that each buy-in is fungible and replaceable from the roll, and that losing one does not change how you play the next. The moment your poker money is also your grocery money, two things break at once:
- Your decisions distort. Money you cannot afford to lose plays “scared.” You fold the thin profitable river call because the loss would hurt your life, not just your roll. Scared money systematically forfeits expected value (EV) — you stop making the close +EV plays that are the difference between a winner and a breakeven player.
- Your tilt amplifies. A normal downswing becomes an existential threat, which feeds the emotional spiral we cover in Chapter 30.
So before any numbers: wall off a sum you can genuinely afford to lose entirely, keep it physically and mentally separate (a dedicated account, a separate online cashier balance), and treat it as the only fuel poker is allowed to burn.
41.2 Risk of ruin, intuitively
Risk of ruin (RoR) is the probability that you lose your entire bankroll before variance turns in your favour — assuming you keep playing the same way at the same stake. It is the number BRM is designed to push toward zero.
You do not need to compute it by hand (calculators and the formula derived from Chapter 9’s distribution exist), but you must internalise how its three inputs trade off:
- Win rate (your edge). The bigger your edge, the faster cumulative winnings outrun the swings, and the lower your RoR. Edge is the single most powerful lever — but it is also the one you can least control or honestly measure in the short run.
- Variance (standard deviation). Higher variance means wider swings around your trend line, which means deeper drawdowns are more likely, which raises RoR. This is format-driven, and it is why a tournament roll must be so much larger than a cash roll.
- Bankroll size (in buy-ins). More buy-ins means more room to absorb a drawdown before zero. RoR falls roughly exponentially as you add buy-ins.
That last point is the one that makes BRM tractable. Because risk of ruin decays exponentially with bankroll, you do not need an enormous cushion to make ruin negligible — you need a sufficient one. The relationship, for a typical solid winner, looks roughly like this:
| Bankroll (buy-ins) | Approx. risk of ruin (winning player) |
|---|---|
| 10 | very high — ruin is likely |
| 20 | meaningful — uncomfortable |
| 30 | low single-digit % |
| 40 | low |
| 50+ | approaching negligible |
These numbers assume you are a winning player. Risk-of-ruin math is brutally asymmetric: if your true win rate is zero or negative, no bankroll is large enough — RoR climbs toward 100% as you play more hands. BRM buys time for an existing edge to show up. It cannot manufacture an edge you do not have. If you are not beating a stake, the answer is study and moving down, not a bigger roll.
Two more realities that the clean formula hides:
- Win rate is uncertain. You never actually know your true win rate; you have a noisy sample estimate. Over-estimating your edge is the most common way players quietly run a far higher RoR than they think. Build your roll as if your win rate is at the pessimistic end of your confidence interval, not the optimistic end.
- Ruin is rarely a clean “lose it all.” In practice almost nobody plays their last buy-in. The real failure mode is being forced to move down (good — that’s the system working) or, worse, reloading from the life roll to stay at a stake the poker roll can no longer support (bad — that’s the system failing). Treat a forced move-down as success, not shame.
41.3 Recommended buy-ins by format
The right number of buy-ins is set by variance, and variance is set by format. Lower-variance formats (deep-stacked cash where your skill edge compounds every street) need fewer buy-ins; higher-variance formats (large-field tournaments where you must win a coin-flip lottery to cash at all) need vastly more.
The guidelines below are starting points for a solid, studied player. Multiply them up if you are risk-averse, still learning the format, or play higher-variance styles; you can run a little leaner only if your edge is large and proven and you are comfortable moving down.
Cash games
For a standard 100bb full-ring or 6-max cash game, one buy-in = 100bb.
| Risk tolerance | Recommended bankroll |
|---|---|
| Aggressive (large, proven edge; willing to move down) | ~20–30 buy-ins |
| Standard | ~30–50 buy-ins |
| Conservative | ~50+ buy-ins |
So a player attacking $0.50/$1.00 (a $100 buy-in) should hold roughly $3,000–$5,000 for that stake under the standard guideline. Cash variance is lowest because you realise your edge continuously and you can reload to keep playing deep-stacked, where skill matters most.
Adjustments within cash:
- Heads-up and very loose/aggressive tables are higher variance — push toward the conservative end.
- Deep-stacked games (200bb+) swing harder in absolute dollars; size your roll on the effective buy-in, not the nominal blind level.
- Live cash has a lower hands-per-hour rate, which means variance per session feels smaller but your edge also realises more slowly; the buy-in counts above still apply, just over a longer calendar.
Tournaments (MTTs)
Tournaments are a different universe of variance. The prize structure is top-heavy: you cash perhaps 12–18% of the time, and the bulk of your EV sits in the rare deep runs and final tables. You can be a strong, clearly winning MTT player and still not cash for dozens of buy-ins in a row. The bankroll requirements reflect that.
| Field type | Recommended bankroll |
|---|---|
| Small-field / sit-and-gos (single table, ~6–9 players) | ~30–50 buy-ins |
| Standard MTTs (a few hundred entrants) | ~100–200 buy-ins |
| Large-field / huge online MTTs / big live series | ~200–300+ buy-ins |
The headline number: cash games, tens of buy-ins; multi-table tournaments, hundreds. If you take one thing from this chapter, take that the order of magnitude differs by format. Rolling for MTTs like they are cash games is the single most common way good tournament players go broke.
Tournament-specific notes:
- Re-entry and rebuy events raise your effective cost per tournament. Count a likely re-entry into your per-event buy-in when sizing the roll.
- Satellites have their own bubble-driven variance (lumpy, all-or-nothing payouts) — treat them like high-variance MTTs.
- PKOs (progressive knockouts) are higher variance than vanilla MTTs because a chunk of every prize pool is locked in bounties you only collect by winning all-ins; push toward the upper end of the range.
- Field size is the master variable. A 50-runner local nightly and a 5,000-runner online major can both be “MTTs,” but the second needs roughly double the buy-ins of the first because the variance scales with field size.
Other formats, briefly
- Spin-and-gos / hyper-turbo lottery sit-and-gos: extreme variance from the random prize multiplier plus short stacks. Treat as 100+ buy-ins minimum despite the single table.
- PLO (Pot-Limit Omaha) cash: equities run much closer than in Hold’em, so variance is meaningfully higher than NLHE cash. Add roughly 50% to your Hold’em cash buy-in count.
41.4 Moving up and moving down
A bankroll is not static — its whole purpose is to let you climb. The rules for when to move are where discipline earns its money.
Moving up
The textbook approach is to set a clear, pre-committed buy-in threshold for the next stake and only move when you hit it. For cash, a common standard is: have a full roll for the higher stake before you sit there full-time. If $0.50/$1.00 wants 40 buy-ins ($4,000) and $1/$2 is a $200 buy-in, you want $8,000 (40 × $200) before treating $1/$2 as your home game.
Two refinements:
- Move up only when you are beating your current stake over a meaningful sample, not after one heater. A 30-buy-in upswing over a weekend is variance, not a promotion.
- Move up only when you are mentally ready for the swings in dollar terms. The cards play identically at $1/$2 and $2/$5, but a 10-buy-in downswing is now five times the money. If that number makes you play scared, you are not ready regardless of what the roll says.
Moving down
This is the rule people have and then don’t follow, and it is the most important one. Set a move-down trigger before you ever sit at a stake, and obey it without negotiation. A standard trigger: if your roll falls below the minimum buy-in count for your current stake, you drop to the stake your remaining roll can support, and you grind back up.
Moving down is not a demotion or an admission of being a losing player. It is the mechanism that makes your low risk of ruin actually low. The math earlier assumes you do move down when the roll shrinks; a player who refuses to drop and instead reloads has thrown the whole model out and is now running a genuine risk of ruin.
Refusing to move down to “win it back at the stake where I lost it.” This is loss-aversion and ego, not strategy. The cards do not know or care where you lost the money; expected value is identical whether you grind it back at the current stake or the one below. The lower stake simply lets you do it with a vastly lower chance of going broke. Insisting on the bigger game while short-rolled is exactly how a temporary downswing becomes permanent ruin.
Shot-taking
A shot is a deliberate, bounded attempt to play a stake your roll does not yet fully support — a way to test the higher game and accelerate your climb without committing the whole roll to it.
The structure that keeps a shot disciplined:
- Pre-allocate a fixed number of buy-ins to the shot — say 3–5 buy-ins for the higher stake — set aside from the main roll.
- Define the failure condition in advance: if you lose those buy-ins, you drop straight back to your normal stake. No re-deciding mid-session, no “one more.”
- Define the success condition too: if the shot puts you up enough that you now have a real roll for the new stake (e.g., you win some buy-ins and your total clears the move-up threshold), you stay; otherwise you bank the win and return, having proven nothing yet about your edge there.
- Keep the rest of the roll untouched — the shot money is the only money at risk.
A worked example:
You are a $0.25/$0.50 cash player ($50 buy-in) with a $2,800 roll — a healthy 56 buy-ins, well into conservative territory. $0.50/$1.00 ($100 buy-in) is the next stop; a full standard roll there would be ~40 × $100 = $4,000, which you do not have. Instead of waiting, you carve out a 4-buy-in shot = $400. Rules: you sit $1/$2… no — you sit $0.50/$1.00 with $400 of shot money. If you drop the full $400, you are back to a $2,400 roll at $0.25/$0.50 and you return there immediately to rebuild. If instead you run the $400 up such that your total roll clears $4,000, the shot has “graduated” and $0.50/$1.00 becomes your new home stake. Either outcome is a success of process: you got controlled exposure to the bigger game while risking at most ~14% of your roll, and your downside was a clean, pre-agreed retreat rather than a panic.
Note what makes this safe: the shot was a small, fixed fraction of the roll; the move-down was automatic; and the rest of the bankroll never moved. A shot that violates any of those — chasing losses with “just one more buy-in,” dipping into the core roll, or staying up at the new stake after a single winning session — is no longer a shot. It is just being under-rolled with extra steps.
41.5 The psychology of playing within your roll
The numbers in this chapter are the easy part. The hard part is that BRM is a behavioural discipline practised under emotional load — usually mid-downswing, usually tilted, usually exactly when you least want to follow it. Three ideas to anchor it:
Being correctly rolled is a strategic weapon, not just insurance. A player who knows that no single buy-in matters can make every thin +EV call, every well-timed bluff, every close river hero-fold on its merits — because the money at stake is genuinely trivial against the roll. The under-rolled player cannot; their EV silently leaks out of every marginal spot through scared, results-oriented decisions. Proper BRM is what unlocks the aggressive, fearless A-game the rest of this book teaches.
- Decouple session results from self-worth and from your roll’s health. One session is a sample of essentially zero. The roll exists precisely so that no session can hurt you enough to matter; let it do that job. If a normal buy-in loss genuinely rattles you, that is reliable evidence you are playing too high for your psychological bankroll, which is often smaller than your financial one.
- Withdraw and protect. Once your roll comfortably exceeds what a stake requires, skimming profits into your life roll is not a leak — it is locking in the work. Many strong players keep their poker roll capped at the buy-in count their current stake needs plus a shot cushion, and sweep the rest out. This also quietly reinforces rule number one: the life roll and the poker roll are different accounts, and money flows out of poker far more easily than it flows back in.
Do this in writing, before your next session, for the stake you currently play:
- State your true bankroll — money you can lose entirely without touching rent, savings, or anyone else’s money. Write the figure down.
- Convert it to buy-ins for your stake and format. Where does it land against the tables above — aggressive, standard, conservative, or under-rolled?
- If under-rolled, write the stake you should drop to so that you are at least at the standard count, and commit to it for your next session.
- Write your two triggers as single sentences: “I move down to ___ if my roll falls below .” and ”I move up to when my roll reaches ___ and I am beating this stake.”
- If you intend to shot-take, write the fixed buy-in count for the shot and the exact stake you retreat to when it fails.
Keep the page where you can see it. The entire value of BRM is that these decisions are made now, in calm, in writing — so that the tilted, downswing version of you who shows up later has nothing left to negotiate.
Bankroll management is the bridge between the variance theory of Chapter 9 and a poker career that survives long enough for skill to win. The edge you build everywhere else in this book is only worth what your discipline lets you keep on the felt. Pick your buy-in counts honestly, separate your money, move down without ego, take shots with a seatbelt on — and let the long run actually arrive.