Crash guide
Crown Crash: what a simulation of 1.50x rounds shows
Crown is only what this page is called; there is no crown edition of the game. The page runs a reproducible computer simulation of crash rounds at a 1.50x target and explains how to read it without fooling yourself.
Why simulate when the formula is known
The exact expected value of a 1.50x target in the site model is easy: the chance is 0.99/1.5 = 66.00%, and the expected net per unit is -0.0100. What the formula does not show well is texture: how sessions look from the inside, how deep they dip, and how long the bad patches last. A simulation gives concrete sample paths to look at.
The simulation below uses Python's built-in pseudo-random generator with a fixed seed, 1771. A fixed seed means anyone who runs the same code gets the same table. It also means the table is one sample among countless possible ones. Change the seed and every number changes.
| Session | Rounds won | Net (units) | Lowest point | Longest losing run |
|---|---|---|---|---|
| 1 | 91 | -13.50 | -13.50 | 4 |
| 2 | 99 | -1.50 | -3.00 | 3 |
| 3 | 95 | -7.50 | -9.50 | 5 |
| 4 | 99 | -1.50 | -5.00 | 4 |
| 5 | 106 | 9.00 | -3.50 | 5 |
| 6 | 97 | -4.50 | -8.50 | 4 |
| 7 | 102 | 3.00 | -7.00 | 4 |
| 8 | 105 | 7.50 | -4.50 | 4 |
Seeded pseudo-random simulation (Python random.Random, seed 1771), 8 sessions of 150 one-unit rounds. Average session result -1.12 units. This is a simulation, not a record of real play; a different seed gives different numbers.
How to read each column
- Rounds won
- How many of the 150 rounds reached 1.50x. The model expects 99.0 on average, with a standard deviation of about 5.8.
- Net (units)
- Each win adds 0.5 units and each loss removes 1. The theoretical average for a 150-round session is -1.50 units.
- Lowest point
- The deepest the running total fell below the start during the session. It shows how much money was at risk, which the final net hides.
- Longest losing run
- The most losses in a row. At a 1.50x target each loss costs as much as two wins, so even a run of three is a setback.
What the sample does and does not prove
With only eight sessions, the average of the table can sit well above or below the theoretical value. That is expected: the standard deviation of one 150-round session is about 8.70 units, much larger than the expected loss of 1.50. A few lucky sessions can make a losing bet look profitable.
A much longer run settles the question. Using the same simulator with seed 1771 over 200,000 rounds, the average net per round comes out at -0.0066, against the exact -0.0100. The more rounds, the closer the two get. That is the law of large numbers at work, and it always favors the side with the edge.
Pitfalls when you run your own
- Do not tune a betting rule on one seed and then report its results. It will fit the noise of that seed.
- Report the seed, the round count and the model, so others can repeat the run.
- Compare the simulated mean with the exact value. If they disagree badly over many rounds, the code has a bug.
- Never present a simulation as a record of real play. This one is labelled as a simulation on purpose.