basics
The house edge by game in one table, from 0.5% blackjack to the 14% baccarat tie bet, with the arithmetic showing why the cost accumulates with turnover rather than with the hours you spend at the table.
The house edge is the price of the product. Everything else about a casino, the interface, the bonuses, the payment rails, sits on top of one number: the fraction of each bet the operator keeps in expectation.
Knowing that number turns gambling from a hope into a purchase with a visible price tag. This guide sets out the figures game by game and then shows the part most people get wrong, which is what the price is charged against.
The house edge is the expected loss per unit wagered, expressed as a percentage.
A 2.7% edge means that for every $100 you bet, the operator expects to retain $2.70 across the long run. You get $97.30 back on average. That is the same statement as a 97.3% return to player, viewed from the other side.
The edge comes from the gap between the true odds of an outcome and the odds the game pays. European roulette has 37 pockets. A single-number bet wins once in 37 attempts. A fair payout would be 36 to 1. The table pays 35 to 1. That one-unit shortfall, spread across 37 outcomes, is 1/37, or 2.7%.
American roulette adds a second green pocket. Now there are 38 pockets, the payout is still 35 to 1, and the shortfall is two units across 38 outcomes. That is 2/38, or 5.26%. The wheel looks almost identical. It costs nearly twice as much to play.
| Game | House edge |
|---|---|
| Blackjack, basic strategy | ~0.5% |
| Stake Originals | 1% |
| Baccarat, banker | 1.06% |
| Baccarat, player | 1.24% |
| Craps, pass line | 1.41% |
| European roulette | 2.7% |
| Slots | typically 2% to 6% |
| American roulette | 5.26% |
| Baccarat, tie | ~14% |
The spread from top to bottom of that table is roughly twenty-eight fold. The tie bet at a baccarat table costs about twenty-eight times as much per dollar wagered as blackjack played with basic strategy, and it sits on the same felt, one chip placement away.
A note on blackjack. The 0.5% figure assumes basic strategy played correctly and reasonably standard rules. Deviations cost money quickly. Rules vary between tables, and a table paying 6 to 5 on a natural blackjack rather than 3 to 2 raises the edge substantially. Basic strategy lowers the price of the game. It does not remove it, and no playing pattern turns a negative expectation into a positive one.
The same applies to every system you will encounter. Doubling after a loss, waiting for a colour to repeat, tracking hot and cold numbers: these change the shape of the distribution of outcomes. They leave the expected cost per dollar wagered exactly where it was, because the edge is a property of each individual bet and not of any sequence of them.
Here is the point the table above cannot show you.
The house edge is charged against turnover. It is not charged against your deposit, and it is not charged against your time. Two players at the same edge, seated for the same three hours, can face wildly different expected costs, because expected cost is edge multiplied by total amount wagered.
Turnover is stake size multiplied by rounds played. Both halves matter, and speed is the half people ignore.
Consider two players, both at a European roulette table with a 2.7% edge, both playing three hours.
Player one bets $5 a spin at a live table running 40 spins an hour. Turnover is 5 x 40 x 3 = $600. Expected cost is $16.20.
Player two bets $20 a spin on a software table running 100 spins an hour. Turnover is 20 x 100 x 3 = $6,000. Expected cost is $162.
Same game. Same edge. Same three hours. Ten times the cost. Nothing about the wheel changed, only the rate at which money passed across it.
Because the meter runs on turnover, the honest way to compare games is cost per hour at a realistic stake and speed. The figures below use ordinary stakes and typical round rates.
| Game | Stake | Rounds/hour | Turnover/hour | Edge | Expected cost/hour |
|---|---|---|---|---|---|
| Blackjack, basic strategy | $10 | 70 | $700 | 0.5% | $3.50 |
| Craps, pass line | $10 | 30 | $300 | 1.41% | $4.23 |
| Baccarat, banker | $10 | 60 | $600 | 1.06% | $6.36 |
| Stake Originals | $1 | 900 | $900 | 1% | $9.00 |
| European roulette | $10 | 50 | $500 | 2.7% | $13.50 |
| Slots | $1 | 600 | $600 | 4% | $24.00 |
| American roulette | $10 | 50 | $500 | 5.26% | $26.30 |
| Baccarat, tie | $10 | 60 | $600 | 14% | $84.00 |
Look at the Stake Originals row against the European roulette row. The originals carry a 1% edge, less than half of roulette's 2.7%, and they still cost two thirds as much per hour. The reason is speed. A crash round, a dice roll or a limbo result resolves in seconds and needs no dealer, so the round count per hour is many times higher than any table game manages.
This is the practical lesson. A low edge is genuinely cheaper per dollar wagered, and the RTP simulator will show you that clearly. But the cost you actually pay is edge multiplied by turnover, and a fast game with a low edge can easily outrun a slow game with a high one. If you want a cheap hour, you need a low edge and a modest rate of play. Either alone is not enough.
Turnover is not a one-off measurement. Money recycles, and each pass through the bankroll is charged again.
Start with $100 on a 4% edge game. Bet the whole balance through once and you expect $96 left. Bet that through and you expect $92.16. The expected remaining balance after n passes is your starting figure multiplied by 0.96 to the power n.
| Passes through the bankroll | Cumulative turnover | Expected balance remaining |
|---|---|---|
| 1 | $100 | $96.00 |
| 5 | ~$462 | $81.54 |
| 10 | ~$838 | $66.48 |
| 25 | ~$1,599 | $36.04 |
Erosion is geometric because each pass takes its slice of what survived the last one. This is the sense in which the edge compounds, and it is why a bankroll can vanish on a game whose advertised return is 96%. The 96% was true on every single bet. It was applied twenty-five times.
Now hold the turnover fixed and change the price. That same $1,599 of turnover costs $63.96 at a 4% edge and $15.99 at a 1% edge. Identical play, a quarter of the cost. That is what a lower edge buys you, and it is the entire practical argument for checking the number before you sit down.
The edge is not a prediction about your session. Over a few hundred rounds, variance dominates completely, and results land far from expectation in both directions. A 5.26% game can pay you handsomely for an evening, and a 0.5% game can take everything you brought.
The edge is not a schedule either. It does not become due. A game that has taken money for an hour is not about to return it, and one that has paid out is not now depleted. Each round is drawn fresh.
And the edge is not something a strategy dissolves. Bet sizing, stop-losses and staking plans are budget tools. They control how much turnover you generate and how long your bankroll lasts, which genuinely matters. They do not change the price per dollar wagered.
Treat the edge as a ticket price and the arithmetic becomes simple. Decide what an evening is worth to you. Divide that figure by the edge of your chosen game, and you have the turnover your budget supports. Divide by your stake and you have the number of rounds. Play those, and stop.
At $30 of intended spend on a 2.7% European roulette table, that is about $1,100 of turnover, or 110 spins at $10. On a 14% tie bet, the same $30 buys around 214 dollars of turnover, roughly 21 bets. The tie bet is not more dangerous in some vague sense. It is simply ten times more expensive per bet, and the table tells you so before you place it.
Read RTP explained for the same number from the player's side, and set deposit and session limits before you start using the guidance on our responsible gambling page.