House edge is the part of every bet a game keeps on average: one minus the RTP. A game that returns 97% has a 3% house edge, so each 100 units staked cost 3 units in expectation. No cash-out target, bet pattern or timing removes it; over many rounds it is the price of play.
What is the house edge?
The house edge is the game's built-in margin. It is the mirror image of the return to player:
House edge = 1 − RTP
| RTP | House edge | Expected cost per 100 units staked |
|---|---|---|
| 99% | 1% | 1 unit |
| 97% | 3% | 3 units |
| 96.5% | 3.5% | 3.5 units |
| 92.13% | 7.87% | 7.87 units |
The last row is not hypothetical. Classic perya Color Game, with three dice and the traditional 1:1, 2:1 and 3:1 payouts, returns exactly 199/216, or about 92.13%, which leaves a house edge of 17/216. You can check that arithmetic in the Color Game odds tool or on the Color Game page.
For crash games, the stated figures are provider claims. Spaceman, for example, is stated at 95.5%, a 4.5% edge, and Aviator at 97%, a 3% edge. The lab counts rounds to test those claims; see Understanding RTP for how.
How does the edge become a cost?
On a single bet, the expected cost is simple:
Expected cost of a bet = stake × house edge
A 10-unit bet at a 3% edge costs 0.3 units on average. You will never lose exactly 0.3 units on one round: you either lose the stake or win a multiple of it. But over many rounds the average cost per bet settles at that figure.
Over a session, the stakes add up:
Expected cost of a session = total staked × house edge
The total staked is not your deposit. It is every stake you place, including winnings you bet again. If you start with 100 units and bet 5 units on each of 200 rounds, you have staked 1,000 units, ten times your starting balance. At a 3% edge, the expected cost of that session is 30 units. At a 1% edge it is 10 units.
Illustrative numbers, not lab measurements.
How does it add up per hour?
Because crash rounds are short, cost per hour is the most useful figure for a player:
Expected cost per hour = rounds per hour × average stake × house edge
| Rounds per hour | Average stake | House edge | Expected cost per hour |
|---|---|---|---|
| 60 | 5 units | 3% | 9 units |
| 120 | 5 units | 3% | 18 units |
| 120 | 5 units | 1% | 6 units |
| 240 | 5 units, two bets a round | 1% | 24 units |
Illustrative numbers, not lab measurements.
The last row shows why the RTP alone is a poor guide. A game that returns 99%, played fast with two bets a round, costs more per hour on average than a 97% game played at half the pace with one bet. Crash Game Economics works through this from the other side: why a thin edge can still earn a lot.
Why can't a strategy remove the edge?
Because the edge sits inside every single bet, and a strategy only rearranges bets.
- Cash-out targets. In a typical crash design, the chance of reaching a target falls in proportion to the target, so the return comes out about the same at 1.5x, 2x or 10x. A target changes how often you win, not what each bet costs on average.
- Doubling after a loss. Within a session, doubling after a loss usually ends in a small win, but occasionally a losing run outgrows the bankroll or the bet limit, and that one loss wipes out the many small wins. Every bet keeps the same house edge. The expected cost per unit staked does not change, but doubling raises the total staked, so the expected cost of the session rises. Run it yourself in the martingale simulator, or read the martingale check.
- Reading the history. When rounds are independent, past results do not change the next one. History describes behaviour. It does not predict the next round. See patterns and hot hours.
- Predictor apps. They cannot see a result that is fixed by a secret seed or a certified generator. See predictor bots.
How does the house edge relate to what a game earns?
What a game keeps from players, before costs, is called gross gaming revenue (GGR): stakes minus winnings. Over many bets it settles at about:
Expected GGR ≈ total staked × house edge
That is the same formula as the player's expected cost, seen from the other side of the table. It explains why a game with a thin edge can still earn a lot: if the total staked is large because rounds are fast and players repeat them, a small percentage of a large number is a large number. Crash Game Economics explains this with illustrative arithmetic, from a player's point of view.
What the house edge does not tell you
- Not how a session will go. Variance dominates the short run. A 1% edge game can still empty a balance in minutes.
- Not whether the stated edge is true. A provider's figure is a claim until it is measured. Game pages show the claim and, once a game passes the minimum sample, the measurement with its interval. Start from the games list.
- Not what is right for you. The edge is a price. Whether the price is worth paying, and for how long, is a personal decision. If gambling is costing you more than money, the help resources at the foot of every page are a good place to start.
Was this page useful?
Anonymous: no account and no cookie.