The Gambler's Ruin: Why Infinite Play Always Ends in Bankruptcy
A gambler with finite wealth faces a casino with infinite wealth. Even with a fair coin, the gambler is mathematically certain to go broke. The proof is exact and merciless.
Imagine a gambler sitting down with $100 at a roulette table. They will play until they have doubled their money to $200 — or until they are bankrupt. The wheel is fair (ignore the house edge for now). What is the probability they reach $200 before $0?
The answer is exactly 50%. But the more interesting question is: what happens if the target is not $200 but $1,000, and the casino has no limit to its bankroll? The gambler's ruin theorem answers this, and the answer is one of the most clarifying results in probability theory.
The formal setup
Player A starts with $k. The casino (Player B) has effectively infinite wealth. On each bet: A wins $1 with probability p, loses $1 with probability q = 1−p. Play continues until A reaches $N (wins) or $0 (bankrupt).
For p ≠ 0.5 (unfair game): P(ruin) = 1 − [(1 − (q/p)ᵏ) / (1 − (q/p)ᴺ)] For p = 0.5 (fair game, finite N): P(ruin | start at k, target N) = 1 − k/N For p = 0.5 (fair game, N → ∞): P(ruin) = 1 — ruin is certain even with a fair coin
With an infinite casino, a fair coin, and infinite time, the gambler is certain to eventually go bankrupt. Not likely — certain. The proof follows from the Optional Stopping Theorem for martingales.
The unfair game
In a real casino, p < 0.5. Consider European roulette: betting on red/black gives p = 18/37 ≈ 0.4865 and q = 19/37 ≈ 0.5135. Starting with $100, targeting $200:
q/p = (19/37) / (18/37) = 19/18 ≈ 1.0556 P(ruin before $200 | start $100, N = 200): = 1 − [(1 − (19/18)¹⁰⁰) / (1 − (19/18)²⁰⁰)] ≈ 1 − [small number / (small number)] ≈ 0.999... (virtually certain ruin) Starting with $1,000 targeting $2,000: probability of ruin > 99.999...%
Why doubling your bankroll is not 50/50
| Starting amount | Target (double) | Probability of success (European roulette) |
|---|---|---|
| $10 | $20 | ~49.3% |
| $100 | $200 | ~46.8% |
| $1,000 | $2,000 | ~23.0% |
| $10,000 | $20,000 | ~0.5% |
The longer the session must last, the more opportunities the house edge has to work. Trying to double $10,000 requires on average thousands of bets — and the cumulative effect of a negative edge over thousands of bets is devastating.
The one case where the math improves
The gambler's ruin probability decreases dramatically if you concentrate bets. If you need to double $100 and bet it all on a single roulette spin, your probability of success is exactly 18/37 ≈ 48.6% — the best mathematical outcome possible. Splitting it into 100 smaller bets gives the house edge 100 opportunities to work rather than 1. Bold play minimizes expected loss when loss is the likely outcome.
This is the mathematical basis of the "bold play" theorem in gambling theory — a deeply counterintuitive result that optimal strategy in a losing game requires maximum bet sizes, not conservative spreading.