Dice: The Cleanest Game In The Casino
How the Dice slider works, why the odds are transparent, what the RTP means in practice, and how to use dice as a bankroll management training tool.
Dice is a single-number RNG game with a slider. Move the slider, see the exact win probability and payout, hit roll. That is the entire game. The RTP is 99% at quality crypto casinos, the highest available on standard casino formats. There is no skill involved. The reason it is worth knowing is that the transparent odds make it the best game for understanding how casino math actually works, and the cleanest format for running controlled bankroll experiments.
What it is
The screen is minimal. A number line from 0 to 100 with a draggable threshold marker. You choose to bet either "roll over" or "roll under," move the slider to set your win probability, and the payout multiplier updates in real time.
Set the threshold at 50: you win 50% of the time and collect just under 2x. Move it to 90: you win 90% of the time and collect roughly 1.10x. Move it to 5: you win 5% of the time and collect roughly 19.8x. Every configuration at 99% RTP yields the same expected payout per unit bet: 0.99.
You can set a specific target number rather than using the slider. You can enable auto-roll and set it to run for a fixed number of rounds, or until a profit or loss threshold is hit. The game supports bet-change-on-win and bet-change-on-loss automations, which is how Martingale strategies get implemented, and usually how they fail.
Dice is available at every major crypto casino as an original casino game. Stake's implementation is the reference, but the format is essentially identical everywhere. The simplicity is the point.
The math
The payout for a given win probability W at RTP R is:
Payout = R / W
At 99% RTP and 50% win probability: 0.99 / 0.5 = 1.98x. At 99% RTP and 10% win probability: 0.99 / 0.10 = 9.9x. At 99% RTP and 1% win probability: 0.99 / 0.01 = 99x.
The formula is linear. The house extracts its 1% regardless of slider position. This is the cleanest possible demonstration that changing variance does not change expected value.
| Win chance | Payout (at 99% RTP) | Expected value per unit | Loss rate per 100 rolls |
|---|---|---|---|
| 95% | 1.042x | 0.99 | ~0.95 units |
| 50% | 1.980x | 0.99 | ~1.00 units |
| 25% | 3.960x | 0.99 | ~0.75 units |
| 10% | 9.900x | 0.99 | ~0.90 units |
| 2% | 49.50x | 0.99 | ~0.98 units |
The expected loss rate per session is 1% of total handle, not 1% of bankroll. If you bet 100 units across 1,000 rolls at 50% win chance, you expect to lose approximately 10 units. The variance around that expectation depends on the win probability selected.
Standard deviation per roll at win probability W is: sqrt(W x (1-W)) x payout. At 50%, this is roughly 0.99. At 5%, the standard deviation per roll is about 4.4x your stake, meaning wild swings are expected even over hundreds of rolls.
Strategy
Dice has one genuine strategic feature: visible, exact odds. Most casino games hide the math. Dice displays it openly, which makes it useful for understanding variance management in a controlled environment.
What low variance buys you. Setting the slider to 90-95% win chance means you win nearly every roll at a small profit. Your balance decreases slowly and smoothly. This is not a winning strategy, the 1% edge never disappears. But it gives you far more rounds per bankroll unit than low-probability bets. More rounds means a smoother experience and better intuition for what 1% house edge actually feels like at volume.
What high variance costs you. At 1% win chance, you can easily go 150+ rolls without a hit. Even with a 99x payout on success, a 150-roll drought at 1 unit per bet costs 150 units to recover from a single 99-unit win. The math still says EV is 0.99, but the standard deviation is large enough that most players experience either a quick bust or an early exit before variance averages out.
Martingale does not work here. It does not work anywhere with a negative-EV game, but Dice makes the failure spectacularly clear. Doubling after a loss at 50% win probability seems safe until you hit a seven-loss streak (1.56% probability per 100 sets), at which point you need 128x your original stake on the eighth bet. The auto-bet feature makes it easy to watch this happen very quickly.
Kelly Criterion adaptation for fixed-edge games. Kelly tells you to bet f = edge / odds, where edge is the expected return and odds is the payout minus 1. At -1% edge (which is what Dice gives you), Kelly says bet zero. Practically speaking, this means keep unit bet size to 0.5-1% of your bankroll if you want to survive variance and not grind through your entire stake in one session.
Use auto-roll responsibly. Auto-roll at 95% win chance for 500 rounds is a laboratory for the 1% house edge, not a money-printing machine. The final balance after 500 rolls at 1 unit per roll should be around 495 units on average. It will not be, because variance, but the expected value is there in the data.
Common mistakes
- Thinking low win chance gives better odds. Every win probability on the slider has identical expected value. Lower win chance means higher variance, not better math.
- Automating a Martingale and forgetting about it. The auto-bet feature can burn through a large bankroll faster than any manual strategy. Set a loss limit before enabling it.
- Interpreting a winning session as skill. Dice has no skill component. A profitable 200-roll session is variance, not technique.
- Playing Dice on platforms without published RTP. The 1% house edge exists at quality crypto casinos like Stake. Some dice implementations run at 2-4% without disclosure.
- Chasing losses at higher variance. Moving the slider toward lower win chances after a losing streak does not improve expected value. It amplifies the loss risk.
- Forgetting that speed matters. Dice rolls fast under auto-bet. 1,000 rolls at 1 unit each is 1,000 units of handle. At 1% house edge that is an expected 10-unit loss, but the session ends in minutes.
Where to play it
The 100% RTP Dice demo on this site is genuinely the most educational free tool we have. Set any win probability, run 500 auto-rolls, and watch variance at work in real time. It is exactly the same format with no house edge applied.
For real-money dice at competitive RTP:
- Stake runs Dice at 99% RTP with full auto-bet support, roll history, and provably fair verification.
- Roobet offers Dice as a core Originals title with the same slider format and clean session stats.
- Shuffle carries Dice with transparent odds and fast round processing.
- Rainbet includes Dice in its original casino games suite with published 99% RTP.
---
FAQ
Q: What is the house edge on Dice? A: At Stake, Dice runs at 99% RTP, meaning a 1% house edge. This applies regardless of which win chance you select on the slider.
Q: What does roll over versus roll under mean? A: Roll over means the result must be higher than your chosen threshold. Roll under means it must be lower. Both modes produce the same expected value at the same win probability.
Q: Is Dice provably fair? A: Yes. The roll result is derived from a server seed and client seed combination that you can verify after each roll. It is one of the most audited game types in crypto casino history.
Q: What win chance gives the best odds? A: All win chances carry the same 99% RTP. The slider changes variance, not expected value. Higher win chances (95%+) give very small payouts per win. Lower win chances (1-5%) give large payouts rarely.
Q: Can you use Dice to beat the house? A: No. No betting system, including Martingale and Paroli, overcomes a negative expected value. The 1% edge applies to every roll regardless of sequence.
Q: What is the maximum payout on Dice? A: Maximum payout varies by platform and bet size, but win probabilities approaching 0.01% are technically available. Practical bet limits usually cap real-money payouts well below the theoretical maximum.
Q: Why do players use Dice over other casino games? A: Transparency. The slider shows you exact win probability and exact payout before every roll. No other game format makes the math this visible.
FAQ
What is the house edge on Dice?
At Stake, Dice runs at 99% RTP, meaning a 1% house edge. This applies regardless of which win chance you select on the slider.
What does roll over versus roll under mean?
Roll over means the result must be higher than your chosen threshold. Roll under means it must be lower. Both modes produce the same expected value at the same win probability.
Is Dice provably fair?
Yes. The roll result is derived from a server seed and client seed combination that you can verify after each roll. It is one of the most audited game types in crypto casino history.
What win chance gives the best odds?
All win chances carry the same 99% RTP. The slider changes variance, not expected value. Higher win chances (95%+) give very small payouts per win. Lower win chances (1-5%) give large payouts rarely.
Can you use Dice to beat the house?
No. No betting system, including Martingale and Paroli, overcomes a negative expected value. The 1% edge applies to every roll regardless of sequence.
What is the maximum payout on Dice?
Maximum payout varies by platform and bet size, but win probabilities approaching 0.01% (100,000x multiplier) are technically available. Practical bet limits usually cap real-money payouts well below the theoretical maximum.
Why do players use Dice over other casino games?
Transparency. The slider shows you exact win probability and exact payout before every roll. No other game format makes the math this visible.