How to read the results
The simulator plays your settings as many sessions as you choose, each starting with the same bankroll. Then it summarizes what happened across all of them.
Chance of going bust is the share of sessions unable to fund another bet at the selected fixed stake. A positive balance below that stake still counts as bust. Read it together with the ending-balance distribution instead of judging a setting from one lucky path.
Median vs mean ending balance. The mean is the average of all ending balances. The median is the balance of the session right in the middle when you line them all up. Rare large payouts can lift the mean above the median. This relationship depends on the table, settings and sample; it is not a rule that every run must follow. Read both values and the distribution rather than treating either as a prediction.
Chance to finish ahead, lose 25%, lose 50%, double up - straightforward shares of sessions. Read them together: a setting can have a decent chance to double and a much bigger chance to lose half.
Drawdown is the largest fall from an earlier balance peak during a session. Here it is expressed relative to the starting bankroll. Median drawdown summarises those maximum falls across sessions; worst drawdown is the largest observed in this run. These are sample statistics, not a bound on future losses.
Longest losing streak shows how many drops in a row paid less than the bet. On High risk this number gets long, and that’s expected, not a sign of anything wrong.
Convergence. The convergence table compares a separately sampled return at increasing drop counts with theoretical RTP. Larger samples tend to reduce relative variation, but the gap can widen between checkpoints. Rare large payouts can keep a sample away from theory even after many drops. This table is separate from the finite-bankroll sessions.
Fixed bets, bankroll size and expected loss
This simulator uses a fixed bet and stops each session when its balance cannot cover the next stake. If the stake is close to the whole bankroll, many sessions can stop after only a few below-stake payouts. Their balance paths overlap, so a tight group of sharply falling lines can be a valid result rather than a chart error. Try comparing smaller fixed stakes with the same starting bankroll to isolate that effect.
For a fixed table, expected net loss is expected actual turnover times the house edge. Planned drops and completed drops can differ when sessions stop early. Changing stakes after a win or loss cannot remove the negative expected value of the next independent bet. This tool does not currently simulate Martingale or other progressive staking strategies; read the strategy guide for their limits.
Reproducible runs
Every run has a simulation seed and records the engine version and the multiplier table version. Share the link and anyone gets exactly the same result. The simulator uses a fast statistical random generator, not the casino’s provably fair algorithm: it models millions of drops, it doesn’t replay real ones. To check a real drop, see how provably fair works.
| Metric | Meaning | Common mistake |
|---|---|---|
| Ruined | The balance cannot fund the next fixed stake before the planned end. | Assuming the balance must be exactly zero. |
| Median final | The midpoint of sampled final balances. | Treating it as a guaranteed personal result. |
| Mean final | The average, including rare large sampled returns. | Ignoring how far it can differ from the median. |
| Drawdown | Decline from a session peak, reported relative to the starting bankroll. | Reading it as a percentage of that peak. |