Roulette payouts are easy to memorize and easy to misread. A straight-up number pays 35:1, a split 17:1, a street 11:1, a corner 8:1, a six-line 5:1, a dozen 2:1, and red or black 1:1 on standard layouts. Those figures tell you what the table pays after a win. They do not tell you what a perfectly fair game would have to pay. The difference between posted payout and fair true odds is where the casino price lives.
Begin with one straight-up number
Take a standard single-zero wheel with 37 pockets. If you choose one number, one pocket wins and 36 pockets lose.
That means the true odds against winning are 36:1. A zero-edge game would need to pay 36 units of profit when the number hits in order to balance the 36 losing outcomes.
Standard roulette pays 35:1.
That missing unit is not a dealer mistake. It is the pricing mechanism.
The same concept becomes more expensive on a standard double-zero wheel. One number still pays 35:1, but now there are 37 losing pockets instead of 36. The payout stays the same while the probability gets worse.
Standard payouts are built around a 36-number core
The familiar schedule has a useful internal pattern:
| Standard bet | Numbers covered | Posted net payout |
|---|---|---|
| Straight up | 1 | 35:1 |
| Split | 2 | 17:1 |
| Street | 3 | 11:1 |
| Corner | 4 | 8:1 |
| Six line | 6 | 5:1 |
| Dozen / column | 12 | 2:1 |
| Even-money | 18 | 1:1 |
Regulated rules such as the Maryland standard roulette rules publish these payout relationships directly. The structure is essentially priced as though the game contained the numbered 1–36 field; zero, and double zero where present, add losing outcomes without proportionally improving the ordinary payout.
This is why many standard bets on the same wheel share the same percentage house edge even though their volatility is completely different.
True odds change when the wheel changes, even if the payout does not
Compare a straight-up wager:
| Wheel | Win probability | Losing pockets | Fair true odds | Posted payout |
|---|---|---|---|---|
| Single zero | 1/37 | 36 | 36:1 | 35:1 |
| Double zero | 1/38 | 37 | 37:1 | 35:1 |
The payout sign does not need to change for the game to become more expensive. Adding 00 reduces the chance of winning while leaving the 35:1 prize unchanged.
That is the core reason ordinary double-zero roulette is roughly 5.26% while ordinary single-zero roulette is roughly 2.70%.
Fair payout is not the same thing as “how much I get back”
Roulette descriptions usually quote net payout. A 35:1 straight-up win means $35 of profit for each $1 wagered, with the winning $1 stake also returned.
Players sometimes mix up three different numbers:
- the amount staked;
- the net profit paid by the table;
- the total chips returned after the winning stake is included.
For a $10 straight-up win at 35:1, the net win is $350 and the total returned is $360. When comparing to true odds, use the net payout convention consistently.
This settlement language matters in real disputes, not just in formulas. A player and dealer can agree on the result and still argue if they are using “pays,” “returns,” and “to one” differently.
A dozen shows the same shortage in a less dramatic form
On a single-zero wheel, a dozen covers 12 of 37 pockets and loses on 25.
The table pays 2:1. If the game contained only the 36 numbered pockets, 12 winners against 24 losers would make 2:1 exactly fair. The zero adds a twenty-fifth losing outcome without increasing the payout.
That produces the same familiar single-zero house edge as the straight-up bet.
On a double-zero wheel, the dozen still pays 2:1 but now loses on 26 pockets. The second zero makes the same posted wager more expensive.
Even-money bets reveal why “almost 50/50” is not enough
Red/black, odd/even, and high/low each cover 18 of the 36 numbered pockets.
On a single-zero wheel:
- 18 pockets win;
- 19 pockets lose because zero is included;
- the payout remains 1:1.
The bet looks like a coin flip because it is close to one. It is not a fair coin flip. The green pocket breaks the symmetry.
On a standard double-zero wheel, 18 pockets win and 20 lose. The payout is still 1:1.
Special rules such as La Partage can change this analysis for eligible even-money wagers by returning part of the stake on zero. That is a real rule improvement because it changes settlement, not because the probability of red suddenly changes.
Percentage edge and payout size answer different questions
A straight-up win pays much more than a red win, but that does not make the straight-up bet a better value on a normal wheel. The large payout compensates for a much lower hit probability.
Likewise, a 1:1 outside bet is not automatically “safe.” It is simply a lower-variance way to buy exposure to the same basic wheel price under standard rules.
This is why roulette house edge should be read beside roulette variance. One describes long-run price; the other describes how violently session outcomes can move around that price.
Derive the house edge directly from a posted payout
For a $1 straight-up bet on single-zero roulette:
$$EV=(1/37\times35)+(36/37\times-1)$$
$$EV=-1/37\approx-0.0270$$
So the player expectation is about -2.70% of the amount wagered.
On double-zero roulette:
$$EV=(1/38\times35)+(37/38\times-1)$$
$$EV=-2/38\approx-0.0526$$
The same 35:1 sign now produces roughly twice the percentage cost.
The five-number American bet shows how a different payout can be worse still
Most ordinary standard wagers on a double-zero wheel share the familiar 5.26% edge. The American five-number basket—0, 00, 1, 2, 3—is the classic exception because its payout schedule is shorter relative to the probability.
This is a useful reminder not to assume every bet on a wheel has the same edge simply because many do. The exact wager definition and payout always come first.
For any unfamiliar layout, calculate from outcomes and settlement rather than borrowing a headline percentage from another bet.
From the casino side, payout accuracy is a control function
Roulette’s mathematics only works operationally if wagers are settled according to the approved schedule.
A dealer must identify the winning number, clear losing chips, preserve winning positions long enough to pay them, and apply the correct payout multiplier. A floor supervisor resolving a dispute needs to distinguish a straight-up chip from a split or corner placement and verify the amount actually booked before betting closed.
Surveillance evidence can support the sequence, but the payout table remains the pricing rule. A “lucky number” theory is irrelevant to whether an 8:1 corner was paid correctly.
This is why roulette disputes and mispaid bets belongs next to payout mathematics even though one page is operational and the other mathematical.
A three-step check works for any roulette wager
When you want to know whether a payout is fair, do this:
- Count the total wheel pockets under the actual rule set.
- Count how many pockets make the wager win.
- Compare the posted net payout with the fair odds implied by winning versus losing outcomes.
If a wager has $A$ winning pockets on a wheel with $N$ total pockets, then the probability is $A/N$. The true odds against winning are $(N-A):A$. The posted payout can then be compared directly with that fair benchmark.
The roulette odds calculator can automate the arithmetic, but the wager definition still needs to be correct.
The missing payment is the price of the game
Roulette does not need a complicated commission box to create a casino edge. The price is embedded in the relationship between outcome count and payout.
A 35:1 prize sounds generous until you notice that single-zero true odds are 36:1 and double-zero true odds are 37:1. A 2:1 dozen feels clean until zero adds an extra losing outcome. A 1:1 red bet looks fair until the green pocket breaks the 18-versus-18 balance.
Once you can see that gap, roulette pricing becomes much easier to compare. Continue with roulette probability basics, roulette expected value, roulette payouts, and why most roulette bets have the same house edge.