Risk of Ruin: The Math of Going Broke
Risk of ruin is the probability that your bankroll hits zero before you reach your goal. It is the one piece of gambling math that is genuinely about survival rather than profit, and it delivers an uncomfortable punchline: in a negative-EV game, ruin is not a risk you manage down to zero — it is a destination whose arrival time you control. The controls, in order of power: bet size, bet size, bet size, then variance, then edge.
Play within a plan →What risk of ruin actually measures
Risk of ruin (RoR) answers a precise question: starting from a given bankroll, playing a given game with a given bet size, what is the probability I go broke before hitting my target (or before I choose to stop)? It is not the same as expected loss. Two players can face identical EV and wildly different ruin probabilities, because RoR lives at the intersection of three forces: the edge (which way the tide pulls), the variance (how big the waves are), and the bet size relative to bankroll (how close to the waterline you are standing). Classic gambler's-ruin analysis — one of the oldest problems in probability, studied since Pascal and Fermat — makes all three explicit. Casinos run this math on themselves obsessively; players mostly run it never. That asymmetry is worth fixing.
The three drivers, ranked
Edge matters, but less than intuition says: the difference between a 0.5% and a 1% edge changes how fast the tide pulls, not the direction. Variance matters more: at the same edge, a high-variance bet (dice at 10% win chance, 9.9x payout) produces drawdowns roughly three times deeper than an even-money-style bet, so the same bankroll is effectively three times smaller. But bet size relative to bankroll dominates both, because it enters the math exponentially rather than linearly. In the classic even-money model, ruin probability behaves like a number less than one raised to the power of how many betting units your bankroll contains. Double your units — by halving your bet — and you roughly square that factor. No system, progression or game selection produces effects of that magnitude. Halving your unit does more for your survival than any strategy ever sold.
The honest math of negative-EV games
Play any negative-EV game forever and ruin is certain — probability one, no exceptions, no system immune. That sounds bleak but it is actually clarifying: since the destination is fixed, the only real variables are speed and path. Speed is your expected loss rate: edge times stake times bets per hour. Path is variance: whether the road down is smooth or swings through exhilarating peaks. Every choice you make — game, bet size, win-chance slider, session length — is a choice about speed and path, never about destination. The practical conclusion is not to despair; it is to reframe. Set a stop point (a loss cap, a time cap, a target) so you are never in the play-forever regime where certainty applies. Ruin is only guaranteed for the player with no exit condition. Be the player with an exit condition.
Worked numbers: doubling 100 SUI at the dice table
Concrete gambler's-ruin arithmetic, dice at 49.5% win chance, 2.00x payout, goal of turning 100 SUI into 200. Betting 50 SUI a roll, you double up about 49% of the time — two units deep, nearly a fair coin flip, because so few bets touch the edge. Betting 10 SUI a roll: about 45%. Betting 5 SUI: about 40%. Betting 1 SUI a roll, grinding 100 units: roughly 12% — and about an 88% chance of ruin before the target. Same game, same goal, same edge; only the unit changed. The lesson cuts two ways. If you insist on an all-or-nothing target in a -EV game, bold play is mathematically correct: fewer, bigger bets expose less total volume to the edge. If instead you want playtime and entertainment, small bets buy vastly more of it — at the cost of near-certainty that you will not double up. What you cannot do is grind small and expect the big score. The math has already priced that dream at 12 cents on the dollar.
How professionals use RoR — and how you should
Players with a genuine edge — poker pros, advantage players — use RoR as a solvency constraint: choose a bankroll-to-stake ratio that pushes ruin probability below a personal ceiling, commonly under 5%. That is why serious poker bankrolls run to dozens of buy-ins even for winning players: a positive edge with a thin bankroll still busts with ugly frequency, and a busted pro earns nothing forever after. Casino players should borrow the discipline while inverting the framing. You do not have an edge, so your version of RoR management is session budgeting: your 'bankroll' is the amount you have decided, in advance and in cold blood, that you can lose with zero consequence. Ruin of the session budget is an acceptable, priced-in outcome — that is what the budget is for. Ruin that reaches past the budget into money that mattered is the only kind that counts, and it is fully preventable by one decision made before the first bet.
The Kelly connection
Risk of ruin and the Kelly criterion are two views of the same mountain. Kelly asks: given an edge, what fraction of bankroll maximizes long-run growth? The formula — edge divided by odds — has a famous property: a full-Kelly bettor never faces certain ruin when the edge is real, because stakes shrink with the bankroll, though the swings are brutal (most pros bet fractional Kelly for exactly that reason). But feed Kelly a negative edge and it returns a negative fraction: the growth-optimal stake in a -EV game is zero — or, read literally, take the other side, which is precisely what the house does. That is the deepest way to understand casino economics: the house is the Kelly bettor, sized comfortably against millions of hands, and provably-fair platforms like Llamabet at least let you verify the terms of that trade — the committed seed hash on Sui proves the edge you face is the one advertised. Knowing the trade is the whole game. Fund your entertainment like entertainment, and keep serious bankroll thinking for arenas where your edge can actually be positive.
Frequently asked questions
What is risk of ruin in gambling?
It is the probability that your bankroll reaches zero before you hit your goal or stop. It depends on three things: the game's edge, its variance, and — most powerfully — your bet size relative to bankroll. In negative-EV games, playing indefinitely makes ruin certain; bet sizing and stop conditions control how fast and by what path.
Does betting smaller reduce risk of ruin?
It depends on your goal. Smaller bets make your bankroll last far longer and smooth the ride, but in a negative-EV game they expose more total volume to the edge — grinding 1 SUI bets to double a 100 SUI bankroll succeeds only about 12% of the time, versus roughly 49% with two 50 SUI bets. Small bets maximize playtime; big bets maximize the chance of hitting an all-or-nothing target.
How much bankroll do poker players keep to avoid going broke?
Winning players typically size bankrolls so their risk of ruin falls below a chosen ceiling, often under 5% — in practice dozens of buy-ins for cash games and more for tournaments, because even a solid edge busts frequently on a thin roll. The principle transfers to casino play as session budgeting: predefine an amount whose total loss is acceptable, and never let ruin reach past it.