What martingale is
Martingale is a system where you double the bet after a loss and return to the base bet after a win. The idea: the first win pays back everything lost plus the base bet. On Dice that is a 48% chance (2× multiplier); on Limbo it is a 2× target (also a 48% chance).
Where the system breaks
You cannot double forever: there is a bankroll and a maximum bet. Take a $0.10 base bet and a $12.50 maximum (as in the example below). The streak's bets: $0.10 → $0.20 → $0.40 → $0.80 → $1.60 → $3.20 → $6.40. The next one, $12.80, is already over the cap. So you have 7 attempts.
| What we count | Result |
|---|---|
| Chance of losing 7 times in a row at 48% | 0.52⁷ = 1.03% |
| Chance of losing 10 times in a row | 0.52¹⁰ = 0.145% |
| Profit of a streak that ends with a win | +$0.10 |
| Loss of a streak of 7 losses in a row | −$12.70 |
| Average result of one streak | 98.97% × $0.10 − 1.03% × $12.70 = −$0.032 |
| Average turnover of one streak | $0.79 |
| Result as % of turnover | −4.000%, exactly the house edge |
In short: martingale turns "many small wins" into "a rare big loss". The average result stays the same: −4% of everything wagered.
Simulation: 200,000 players
Each player starts with $20 and plays martingale at a 48% chance: $0.10, doubling after a loss. If the next bet would exceed the $12.50 cap, the streak counts as lost and the player starts over at $0.10. They stop when the bankroll reaches $40, or when there is not enough money for the next bet of the streak.
| Outcome | Share of players |
|---|---|
| Doubled the bankroll to $40 | 26.5% |
| Could not afford the next bet | 73.5% ($2.82 left on average) |
On average a player ended with $12.66 of the $20. They lost $7.34, the same 4% of roughly $183 turnover. Martingale did not make the game better. It only changed the shape of the result: a quarter of players double their bankroll, three quarters lose almost everything. What RTP and house edge mean is in the glossary.
What about flat bets, no system?
Chance of being in profit with flat bets on 2× (48% chance):
| Bets made | Chance of being in profit |
|---|---|
| 10 | 32.9% |
| 100 | 30.8% |
| 1,000 | 9.7% |
| 10,000 | 0.003% |
Over a short run you can win, and that is the game. Over a long run the math wins: a 4% edge on every dollar of turnover.
So what should you do?
- Decide in advance how much you are willing to spend, and treat it as the price of entertainment.
- Pick the risk for your budget, not "for a system". A high chance means a long session, a low chance means rare big wins.
- Check fairness: any Dice or Limbo roll can be recomputed in the verifier.
- Set limits. How to do it is on the responsible gaming page.
How we counted
- Game formulas are from the X3NA Dice and Limbo specifications. Dice multiplier = ⌊96,000,000 / (chance × 100)⌋ / 10,000; Limbo chance = 96 / target.
- Exact numbers follow from the formulas above.
- Simulation: 200,000 players, NumPy generator, seed 20261021, the streak restart rule described above.
- Flat-bet odds: binomial distribution.
NewDiceYou choose the chanceRTP 96%from $0.10
NewLimboTarget multiplier, instant resultRTP 96%from $0.10up to 500×Sources
- Dice specification v1 (19.10)
- Limbo specification v1 (19.10)