Casino Originals House Edge Database: Every Game at Every Platform

BC.Game, Stake, and Duelbits each run all five casino originals (crash, dice, limbo, plinko, mines) at a 1% house edge. The BitPartners group (LTC Casino, ETH Casino, Crypto Casino) matches that 1% on every game they offer. Shuffle sits at 1% across crash, dice, and limbo but does not offer plinko or mines. The Dama N.V. casinos (7Bit Casino, Katsubet, Mirax Casino) charge 3% on crash via Aviator but run dice at 1%. Roobet, Gamdom, and Rollbit are included below based on published claims, but we have not independently verified their hash chains; those rows are marked with an asterisk.

Master Table: House Edge by Casino and Game

* Roobet, Gamdom, and Rollbit house edges are based on published RTP claims and community data. We have not independently verified their hash chains. All other casinos were verified by analyzing 10,000+ game rounds per platform.

Crash Game House Edge

The difference between a 1% house edge and a 5% house edge sounds small. It is not. If you bet $10 per round and play 200 rounds per hour, a 1% edge costs you $20 per hour in expected losses. A 5% edge costs you $100 per hour. Over a 4-hour session, that is $80 versus $400. The casino you choose for crash games is one of the most consequential decisions you can make as a player, and most people never look at the actual numbers.

I spent three weeks verifying the house edge on every major crypto casino's crash game. Not by reading their marketing pages or trusting their stated RTPs. I pulled game hashes, reverse-engineered the formula constants, and calculated the true mathematical edge from the algorithm itself. Here is what I found.

Why House Edge Matters More Than You Think

Most crash players focus on strategy. They obsess over when to cash out, whether to auto-bet, whether the 2x strategy or the 1.5x strategy is better. None of that matters nearly as much as the house edge of the game they are playing.

Here is the math. The expected loss per bet is:

Expected loss = bet size x house edge

The expected loss per hour is:

Hourly cost = bets per hour x average bet x house edge

A typical crash player places between 120 and 300 bets per hour depending on the round speed and whether they use auto-bet. Let me use 200 bets per hour as a reasonable middle ground.

A player at a 5% edge casino loses five times more than a player at a 1% edge casino, betting the same amounts. Over a month of moderate play, that is $2,000 versus $400. The difference is $1,600 that you either keep or hand to the casino, determined entirely by which platform you chose. Use the house edge calculator to plug in your own numbers.

How I Verified Each Casino's House Edge

I did not trust self-reported numbers. Every casino claims to be "provably fair," but the actual edge is buried in the algorithm. Here is the methodology I used.

Step 1: Obtain the Hash Chain

Every provably fair crash game generates results from a hash chain. The casino publishes either the full chain or a verification tool that lets you check individual game results. I collected at minimum 10,000 consecutive game results from each casino.

Step 2: Reverse-Engineer the Formula Constant

The standard crash game formula is:

crash_point = max(1, floor(C / (1 - value)))

Where C is the formula constant and value is derived from the game hash. If C = 99, the house edge is 1%. If C = 97, the house edge is 3%. If C = 95, the house edge is 5%. The relationship is simple: house edge = (100 - C) / 100.

For a detailed breakdown of this formula, see crash game math.

Step 3: Count the Instant Crashes

The easiest way to verify the edge is to count how many rounds crash at exactly 1.00x. In a 1% edge game, approximately 1% of rounds crash instantly. In a 3% edge game, approximately 3% do. I counted the 1x crashes across my sample of 10,000+ rounds for each casino and compared against the expected percentage.

Step 4: Statistical Validation

I also calculated the average crash point across the full sample and compared it against the theoretical mean for each house edge level. For a 1% edge game, the theoretical average crash point converges to approximately 100 (the mean of the distribution). For a 3% edge game, it is approximately 33.33. If the observed mean matches the theoretical mean within a reasonable confidence interval, the stated edge is confirmed.

The Crash Ranking: Every Major Casino's House Edge

After verifying each platform, here are the results sorted from lowest to highest house edge.

Notes on Key Casinos

Stake (1% edge). Stake uses the standard provably fair crash algorithm with C = 99. Their hash chain is publicly verifiable, and the 1.00x crash rate in my sample of 14,200 rounds was 1.02%, consistent with the 1% theoretical rate. This is the lowest edge I found on any major platform.

BC.Game (1% edge). BC.Game also runs at C = 99. I verified this across 12,500 rounds and found a 1x crash rate of 0.98%. The math checks out. BC.Game ties with Stake for the best crash game house edge.

Shuffle (1% edge). Shuffle runs at the same 1% edge as Stake and BC.Game. My analysis of their game results showed a 1x crash rate consistent with C = 99 across 10,000 rounds.

7Bit Casino, Katsubet, Mirax (3% edge). These Dama N.V. casinos do not offer a native crash game. They run Aviator by Spribe, which has a fixed 3% house edge (97% RTP). You pay three times more per bet compared to a 1% platform.

Roobet* (3% edge). Roobet's crash game appears to run at approximately 3% based on community data and published claims. We have not independently verified their hash chain.

Gamdom* (5% edge). Gamdom runs the highest house edge in this database based on published information. A Gamdom crash player loses five times more per bet than a Stake crash player.

How Rakeback and VIP Programs Reduce the Effective House Edge

Raw house edge is not the complete picture. Most crypto casinos offer rakeback programs that return a percentage of losses (or wagers) back to the player. This effectively lowers the house edge you actually experience.

The formula is straightforward:

Effective house edge = house edge x (1 - rakeback rate)

If a casino has a 3% house edge but gives you 10% rakeback on losses, your effective edge is 3% x 0.90 = 2.70%. If they give you rakeback on wagers instead of losses, the math changes slightly, but the principle is the same.

Notice something important. Even with generous rakeback, Gamdom's effective house edge of 4.25% is still more than four times higher than Stake's effective edge of 0.90%. Rakeback helps, but it cannot overcome a fundamentally higher base edge. The casino with the lowest raw edge plus rakeback will always win the math. Check the bonus calculator to see how rakeback and bonuses affect your expected value across different platforms.

Calculating Your Real Hourly Cost at Each Casino

Now let me combine everything into a single practical number: what it actually costs you per hour to play crash at each casino. I will use 200 bets per hour and a $10 average bet, applying the effective house edge (after typical rakeback).

The annual difference between playing crash on Stake ($4,320) and playing the same game on Gamdom ($20,880) is $16,560. That is not a rounding error. That is a used car. Same player, same strategy, same bet sizes. The only variable is which website they opened.

These numbers scale linearly with bet size. If you bet $50 per round instead of $10, multiply everything by five. The Gamdom player at $50 average bets would lose $104,400 per year versus $21,600 on Stake.

For a deeper dive into managing your bankroll around these numbers, see bankroll management.

The Hidden Edges: Games That Claim One Thing but Deliver Another

Not every casino is transparent about its crash game house edge. During my analysis, I found several patterns worth highlighting.

Roobet's variable edge. Roobet's crash game edge appears to fluctuate between 3% and 3.5% depending on when you sample. This could indicate different game versions, A/B testing of edge levels, or slight variations in their algorithm. Either way, the edge is consistently higher than the 1% that the top platforms offer.

Third-party crash games. Some casinos do not build their own crash game. They license it from a game provider. When this happens, the house edge is set by the provider, and the casino takes a cut of that edge. The result can be a higher effective edge than what appears in the game's stated RTP. I found instances where a game listed 99% RTP (1% edge), but the actual return was closer to 96-97% after the casino's additional margin.

"Boosted" crash games. A few platforms run promotional crash games with supposedly lower house edges or boosted payouts. In every case I analyzed, the base game still ran at the same edge. The "boost" came from a separate bonus pool that had its own wagering requirements, making the effective improvement far smaller than advertised.

Non-provably-fair crash games. Some smaller casinos offer crash games without any provably fair verification. Without access to the hash chain or game seeds, there is no way to independently verify the house edge. I excluded these from my ranking because verification was impossible. If you cannot verify it, assume the edge is higher than claimed.

Dice House Edge by Casino

Dice is the most standardized game in crypto casinos. Unlike crash where house edges range from 1% to 5%, virtually every major platform runs dice at exactly 1% house edge (99% RTP). The game mechanic is simple: you pick a number, choose over or under, and the payout adjusts automatically based on your win probability minus the house edge.

The reason dice is so uniform comes down to its simplicity. The house edge is calculated directly from the payout formula, and players can immediately spot any deviation. If a casino offered dice at 2% house edge, the payouts for common targets like "over 50" would be noticeably lower than competitors, and players would leave. This competitive pressure has standardized dice at 1% across the industry.

The verification process for dice is straightforward. The payout for any target is calculated as: payout = (100 - house_edge) / win_probability. At 1% house edge, betting "over 50" on a 0-100 range gives a 50% win chance and pays 1.98x (not 2x, because 1% goes to the house). If the payout on any casino's dice game does not match this formula at 1%, the edge is higher than claimed.

Dice is one of the best games for clearing bonus wagering requirements because of its 1% edge, low variance at conservative targets, and fast round speed. For players focused purely on the math, dice offers the most consistent and predictable cost structure of any casino original.

Limbo House Edge by Casino

Limbo is mathematically identical to crash. You set a target multiplier before the round, the game generates a result, and you either hit your target or lose. The key difference is that limbo removes the manual cashout timing. There is no rising multiplier to watch; you commit to your target upfront.

Because the underlying math is the same, limbo house edges mirror crash house edges at most casinos. If a platform runs crash at 1%, their limbo game is also 1%. The formula constant C works the same way.

Not every casino offers limbo as a standalone game. Duelbits, ETH Casino, Crypto Casino, Cloudbet, BitStarz, and the Dama N.V. group (7Bit Casino, Katsubet, Mirax Casino) do not have limbo in their originals lineup. If limbo is important to you, Stake, BC.Game, Shuffle, and LTC Casino are the verified options, all at 1%. For more on how limbo compares to crash mechanically, see casino originals compared.

Plinko House Edge by Casino

Plinko drops a ball through a pyramid of pegs, and the landing slot determines your payout. Most platforms offer three risk levels (low, medium, high) and varying row counts (8 to 16 rows). The critical point: the risk setting changes variance, not the house edge. High-risk plinko has bigger swings in both directions but the expected loss per dollar wagered stays at 1% on platforms that run it at 1%.

Shuffle, Cloudbet, BitStarz, and the Dama N.V. casinos (7Bit Casino, Katsubet, Mirax Casino) do not offer plinko. Gamdom and Rollbit do not have a plinko game in their originals lineup either.

The verification method for plinko is more involved than crash or dice. Each peg outcome is determined by the provably fair hash, and the payout table maps slot positions to multipliers. I verified plinko house edges by calculating the expected value across all possible slot positions weighted by their probabilities, and confirmed each platform returns 99% (1% house edge) regardless of the risk setting chosen.

Mines House Edge by Casino

Mines is a grid-based game where you reveal tiles while avoiding hidden mines. Each safe tile increases your multiplier, and you cash out before hitting a mine. The house edge is built into the payout formula: the multiplier for each revealed tile is slightly less than the mathematically fair payout based on the remaining probability.

Shuffle, LTC Casino, Cloudbet, BitStarz, and the Dama N.V. casinos do not offer mines. Gamdom and Rollbit also do not have mines in their lineup.

The mines house edge stays constant regardless of how many mines you place on the grid. Placing 1 mine on a 25-tile grid gives higher win frequency but lower multipliers. Placing 24 mines gives a tiny win chance but massive potential payouts. In both cases, the expected return is 99% (1% house edge) because the payout formula adjusts proportionally. The number of mines you choose affects variance, not the edge.

The Bottom Line

The crash game algorithm is nearly identical across every platform. The game looks the same, plays the same, and feels the same. The only meaningful difference between casinos is that one number in the formula: the constant C. That single digit determines whether you lose $4,320 or $20,880 per year on the same bets. Check it. Verify it. Choose accordingly.

For dice, the choice barely matters since every casino runs at 1%. For crash, limbo, plinko, and mines, stick to platforms verified at 1%: Stake, BC.Game, Duelbits, Shuffle, and the BitPartners group (LTC Casino, ETH Casino, Crypto Casino).

Use the house edge calculator to run these numbers with your own bet sizes, or read the full crash game math breakdown to understand exactly how the algorithm works.

Changelog

• April 2026: Expanded from crash-only to five-game database (crash, dice, limbo, plinko, mines). Added 7 casinos (Duelbits, LTC Casino, ETH Casino, Crypto Casino, 7Bit, Katsubet, Mirax). Reclassified Roobet, Gamdom, and Rollbit as unverified. Updated Shuffle crash edge to 1%.
• March 2026: Original crash game house edge comparison published with 8 casinos.