Dice guide
Space Dice: bankroll size and the risk of going bust
The Space name is a tag for this page alone; the game is the usual 0.00-99.99 dice. The question here is about a bankroll: if you bet one unit at a time, how likely are you to run out within a set number of rolls?
The setup
The calculation below is exact. It tracks the probability of every possible balance after every roll and adds up the probability of reaching zero. Each bet is one unit. Two settings, both with a 1% edge, are compared: a 49.5% chance paying 2x, and a 33% chance paying 3x. Both multiply to 0.99, so both lose 1% per unit on average.
| Starting bankroll (units) | P(bust within 300 one-unit bets) |
|---|---|
| 5 | 81.01% |
| 20 | 30.15% |
| 60 | 0.09% |
| Starting bankroll (units) | P(bust within 300 one-unit bets) |
|---|---|
| 5 | 85.85% |
| 20 | 45.39% |
| 60 | 1.72% |
What the two tables show
With 20 units at the 49.5% setting, the chance of busting within 300 bets is 30.15%. At the 33% setting it is 45.39%. The edge is the same, but the 3x bet swings more, and those swings carry a small bankroll to zero more often.
A larger bankroll helps a lot in the short run. With 60 units at 49.5%, the bust chance falls to 0.09%. That is because 300 one-unit bets have a standard deviation of only 17.3 units, and 60 units is more than three of those away.
The long run is different
Stretch the horizon and the picture changes. With 20 units at 49.5%, the bust chance within 300 bets is 30.15%, but within 3,000 bets it rises to 84.20%. With a negative edge and no stopping rule, ruin becomes more and more likely the longer play continues. A big bankroll only delays it.
Practical reading
- Size the bet to the bankroll. Halving the unit size doubles the bankroll measured in units, which cuts short-run ruin sharply.
- Lower-chance targets raise ruin risk at the same edge, because of their higher variance.
- A stop-loss or a time limit is the only rule that truly caps the loss. No choice of target does.
Expected loss along the way
Even when the bankroll survives, it is expected to shrink. Over 300 bets at a 1% edge, the expected loss is 3 units. That is a small number next to the spread, which is why the risk of ruin, not the average, is the better guide to how a session can end. For the broader idea of how margins work, see the house edge guide.