Spunk Odds Reference

Crash guide

Egyptian Crash: how bankroll size changes the chance of going bust

The theme in this address is a tag and changes no rule. This guide is about money management in the plain sense: with a fixed bankroll and flat bets, how likely is it to run out before a session ends?

Setting up the question

Assume a starting bankroll of B whole units. Every round you stake exactly one unit on a 3.00x target. A win adds two units and a loss removes one. The session lasts 300 rounds unless the bankroll reaches zero first, which we call ruin, or bust. In the site model at a 1% edge, each round wins with probability 33.00%.

The figures below come from an exact calculation. It tracks the probability of every possible balance after each round, and moves probability into the ruined column whenever a balance hits zero. No random numbers are involved.

Chance of bust within 300 one-unit bets at a 3.00x target (1% edge), exact
Starting bankroll (units)P(bust within 300 one-unit bets)
585.85%
1558.02%
4011.93%

What the table says

A 5-unit bankroll is very fragile at this target. The chance of losing five rounds in a row right at the start is already 13.50%, and over 300 rounds there are many more ways to drift down to zero. Fifteen units buy much more room, and 40 units more again, but no row reaches zero. With a negative edge, the chance of eventual ruin climbs toward certainty as the session gets longer.

The expected loss over 300 rounds is 3.00 units no matter which bankroll you start with. A larger bankroll does not reduce that cost. It only makes it less likely that a bad run ends the session before the 300 rounds are up.

Target choice and ruin

The same bankroll behaves differently at a different target. At 2.00x a win adds one unit and the swings are smaller. The table runs the same three bankrolls over the same 300 rounds.

Same bankrolls at a 2.00x target (1% edge), 300 rounds, exact
Starting bankroll (units)P(bust within 300 one-unit bets)
581.01%
1544.58%
403.09%

Compare the two tables row by row. The two targets have the same expected loss per unit staked, yet their bust chances differ because the size and frequency of wins differ. Ruin depends on the variance of the bet and on the bankroll, not only on the edge.

Using this as a budgeting tool