In crypto Dice you pick a target number and bet that the roll lands under it or over it. Lower win chances generally pay larger multipliers. The house edge is the gap between the amount wagered and the expected return. Provably fair operators let you replay disclosed results from seeds and a nonce; that check does not establish a profitable game.
How Dice works
How this Dice model works
The illustrative model has 10,000 equally likely integers, displayed from 0.00 to 99.99. BCH.games maps the first 32-bit HMAC result to that grid, with slightly different probabilities. Roll Under wins strictly below its target; Roll Over wins strictly above it. A result exactly equal to the target loses. Captured BCH.games rules are dated October 25, 2026. They do not establish the rules of other operators or versions. The verifiers for each operator explain their exact method.
Probability and payout are separate
In the illustrative model, win chance is the winning-outcome count divided by 10,000, and gross payout is (1 − house edge) / that probability. BCH.games uses its captured nominal-chance calculation and rounds the gross multiplier to four decimals; its calculated RTP uses the actual 32-bit mapping. Gross payout includes the returned stake; net profit subtracts that stake. A qualifying result can still return less than the wager when the multiplier is below 1x.
A simulation is not a prediction or a replay
The demo uses browser cryptographic randomness: rejection sampling for the illustrative model and the captured float mapping for BCH.games. The simulator uses a repeatable pseudorandom sequence for statistical exploration. It runs fixed stakes and stops a session if the next bet cannot be funded. Neither tool establishes how an operator generated a real result.
| Nominal win chance | Payout at 1% edge | Payout at 2% edge |
|---|---|---|
| 10% | 9.9× | 9.8× |
| 25% | 3.96× | 3.92× |
| 49.5% | 2× | 1.9798× |
| 50% | 1.98× | 1.96× |
| 75% | 1.32× | 1.3067× |
| 90% | 1.1× | 1.0889× |