Risk of Ruin Calculator
Even winning bettors can go broke if stakes are too large relative to edge. This tool applies a textbook gambler’s-ruin style approximation: you specify win probability, the net odds you receive (decimal minus one on a win), and how many bankroll units you are willing to lose—then see a rough ruin probability.
Assumptions: independent bets, fixed stake per unit, no borrowing, no tipping the dealer from the roll. Real betting has correlation, stake changes, and tilt—treat the output as an order-of-magnitude lesson, not a guarantee.
Ruin Probability Estimate
Requires positive expected value per trial.
Table of Contents
How to Use This Calculator
Enter win probability (0–1), net odds received on a win as decimal-minus-1 (e.g. 0.91 for −110 style payoff), and units of bankroll you plan to risk as sequential flat bets. The tool outputs an approximate long-run ruin probability.
Win probability
Your estimated chance to win each independent bet at the posted odds.
Net odds b
Profit per $1 risked if you win—decimal odds minus 1.
Units
How many fixed one-unit bets you simulate before stopping—larger n increases bust risk if edge is thin.
No edge case
If expected edge ≤ 0, the tool warns—classic ruin formulas assume positive edge.
What Risk of Ruin Means
Risk of ruin is the probability your bankroll hits zero before stopping, under simplified flat-betting assumptions. It is a thought experiment for bankroll sizing—not a guarantee.
Model Notes
The implementation uses a standard gambler’s-ruin style approximation with positive edge. Real betting has correlation, stake changes, and replenishment—adjust expectations accordingly.
Examples
Example 1 — Defaults
p = 0.55, b = 0.91, units = 100. Edge positive; ruin probability tiny but nonzero—check the live readout.
Example 2 — No edge
If win probability implies negative EV at your b, the calculator reports that ruin framing is undefined—fix inputs.
Example 3 — Bankroll planning
Compare ruin estimates at 50 vs 200 units to see how longer play stretches thin edges.
Frequently Asked Questions
No—Kelly sizes stakes; ruin probability answers how safe a flat unit is over many trials.
With negative EV, eventual loss approaches certainty—ruin isn’t the interesting limit.
American −110 is roughly decimal 1.909; b ≈ 0.909 (profit per $1 risked).
Here one unit is fixed. Changing stake fraction of bankroll requires a different model.
The math assumes independence—real series can be correlated (same league, same model).
