The wheel does not remember your last spin
You are standing at a roulette table, looking at a layout of 36 numbers plus a zero, and the question in front of you is which numbers to put chips on. The honest starting point is that no number on the layout is more likely to appear on the next spin than any other. A European wheel has 37 pockets and a single zero; an American wheel has 38 with a double zero. Each pocket has the same physical chance of catching the ball on any given spin, and the wheel has no memory of what came before.
What changes from bet to bet is not the probability of any individual number appearing. It is the payout the table offers against that probability, and the gap between those two things is where the operator's advantage lives. The rest of this article works through that gap, what it means for the numbers you pick, and why short sessions so often look nothing like the long-run average.
Every bet on the layout carries its own edge
A roulette table is not one game with one house advantage. It is a collection of separate bets, each with its own payout and its own probability, and those two figures determine the edge on that bet. On a European wheel, a straight-up bet on a single number pays 35 to 1, but the true odds against that number appearing are 36 to 1 because there are 37 pockets. The one-unit gap between the payout and the true odds is the operator's margin on that bet.
The same arithmetic applies to every other bet on the layout. A split bet covers two numbers and pays 17 to 1; a street covers three and pays 11 to 1; a corner covers four and pays 8 to 1. In each case the payout is slightly short of what the number of covered pockets would justify. The edge is built into the payout schedule itself, not into which particular numbers you happen to select.
| Bet type | Pockets covered (European wheel) | Payout | What the gap means |
|---|---|---|---|
| Straight-up | 1 of 37 | 35 to 1 | Payout is one unit short of the 36 to 1 true odds |
| Split | 2 of 37 | 17 to 1 | Payout reflects 2-in-37 chance, but is rounded down |
| Street | 3 of 37 | 11 to 1 | Payout is short of the 3-in-37 true odds |
| Corner | 4 of 37 | 8 to 1 | Payout is short of the 4-in-37 true odds |
| Six line | 6 of 37 | 5 to 1 | Payout is short of the 6-in-37 true odds |
The American layout makes the same point more sharply. The extra double-zero pocket adds one more losing number to every bet without increasing any payout. The house edge on the American layout is higher than on the European layout for the same bet type, because the payout schedule stays the same while the probability of winning falls. If you have a choice of wheel types, the wheel with fewer zero pockets carries the smaller edge on every bet you can place.
RTP and house edge are two sides of one calculation
Return to Player (RTP) is a theoretical percentage of total wagered money a game returns to players over a very large number of spins or hands. It is a long-run statistical average, not a per-session guarantee. House edge is the mathematical advantage the operator holds on a given game, expressed as a percentage of each wager. House edge and RTP are complements: they sum to 100% for the same game.
This relationship only holds for the same bet under the same rule set. A straight-up bet on a European wheel and a straight-up bet on an American wheel do not share the same edge, and a straight-up bet and an even-money bet on the same wheel do not necessarily share the same edge either. When you see a single RTP figure attached to a roulette game, it is a composite or a representative figure, not a promise that every bet on the layout returns that percentage.
As an example, suppose a European wheel straight-up bet has an expected return of 97.3 units for every 100 units wagered over a very large number of spins. The house edge on that bet is then 2.7 units per 100, because the two figures sum to 100. If a different bet on the same wheel carries a different edge, its RTP is different too. The number you pick does not move this calculation; the bet type and the wheel type do.
Why a short session looks nothing like the average
The long-run average is a property of a very large number of spins, and a single session is almost never large enough to reveal it. Suppose you place 100 straight-up bets in an evening. The expected number of wins is a little under three, but the actual number can easily be zero, or five, or eight. The payout on a straight-up win is large enough that one extra win or one missing win swings the session result far away from the long-run average.
This is volatility, and it is a separate property from RTP. Volatility describes how payouts are distributed over time. High volatility means rarer but larger payouts; low volatility means more frequent but smaller ones. Volatility does not change a game's RTP. A straight-up bet is high volatility because most spins lose and a few pay 35 to 1. An even-money bet is lower volatility because wins and losses arrive more evenly. Neither bet has a better long-run return because of its volatility; they simply distribute the same edge differently across time.
What this means for number selection is that no set of numbers is "due" to appear. A number that has not appeared for 200 spins is no more likely to appear on spin 201 than any other number. The wheel's physical structure is the same on every spin, and the past sequence of outcomes does not alter it. A machine or wheel being described as "due", "hot", or "loose" is a description of past variance, not a prediction about the next spin.
Decisions per hour multiply your total exposure
The house edge is expressed as a percentage of each wager, which means the total amount you are expected to lose over time depends on how much you wager in total, not just on the edge itself. A roulette table can resolve dozens of spins per hour, and every spin is a fresh wager against the same edge. The faster the game moves, and the more chips you put on the layout per spin, the larger your total wagered amount becomes.
As an example, suppose you wager 10 units per spin and the table resolves 40 spins per hour. Your total wagered amount is 400 units per hour. If the edge on your chosen bet is 2.7 units per 100, the expected loss over that hour is about 10.8 units. If you wager 20 units per spin instead, the expected loss doubles, even though the edge has not changed. The edge is a rate; your total exposure is the quantity that rate applies to.
This is where bet selection and session length genuinely vary under your control. You can choose a bet with a smaller edge, you can choose a slower table, you can wager smaller amounts, and you can stop earlier. None of these choices turns a negative-expectation game into a positive one, but they change the total amount of money exposed to the edge. Number selection, by contrast, changes none of these quantities.
Progression systems redistribute outcomes, not the edge
A common response to the house edge is to adopt a progression system: double the bet after a loss, or increase it after a win, or follow some sequence of numbers that dictates the next stake. The appeal is that a win after a run of losses recovers the earlier losses plus a small profit. The problem is that the system does not change the probability of any individual spin, and it does not change the payout schedule.
What a progression system changes is the distribution of outcomes. Most sessions under a doubling system end with a small win, because the first win after a losing run recovers the accumulated losses. A small number of sessions end with a very large loss, because the losing run continues until the player reaches the table maximum or runs out of bankroll. The expected value of the system is the same as the expected value of flat betting the same amount on the same bets: the edge is still there, multiplied by the total amount wagered.
As an example, suppose you start with a 1-unit bet and double after every loss, with a table maximum of 100 units. You can double six times before you hit the cap. If you lose seven spins in a row, you have wagered 127 units and cannot double again to recover. The probability of seven consecutive losses on an even-money bet is small, but it is not zero, and when it happens the loss is large enough to erase many prior small wins. The system has redistributed the losses into a rarer but larger event; it has not removed the edge that produces them.
What genuinely varies under your control
If the numbers themselves do not matter, what does? The variables that change your expected result are the ones that change the edge, the amount wagered, or the time spent at the table.
Wheel type is the first variable. A European wheel has 37 pockets and a single zero; an American wheel has 38 with a double zero. The extra pocket increases the house edge on the American layout, so choosing the wheel with fewer zero pockets reduces the edge on every bet you can place. Some tables also offer rule variations such as surrender on even-money bets when the zero appears, which changes the edge on those bets specifically. The table rules are posted at the table or in the game information panel, and they are the document that carries the exact figures for that table.
Bet type is the second variable. Different bets on the same table carry materially different edges. The even-money bets on a European wheel carry a smaller edge than the straight-up bets in many rule sets, but the exact figures depend on the table rules. The edge on each bet is not printed on the layout, but it can be derived from the payout schedule and the number of pockets, and the game information panel or the operator's game rules page is where the specific figures for that table are stated.
Stake size, game speed, and session length are the third set of variables. These do not change the edge, but they change the total amount wagered, and the total amount wagered is what the edge applies to. A smaller stake, a slower table, or a shorter session all reduce the total exposure to the edge. None of these choices is a way to win; they are ways to lose more slowly, and that distinction is the honest one.
FAQ
Is there a mathematically best number to bet on in roulette?
No. Every number on the layout has the same probability of appearing on any given spin, and the payout for a straight-up bet is the same regardless of which number you choose. The edge on a straight-up bet is the same for every number. What differs is the bet type and the wheel type, not the specific number.
Do some numbers come up more often than others?
Over a very large number of spins, the physical properties of the wheel and the ball produce a distribution that should approach uniformity, but any short sequence can show clusters and gaps. A number that has appeared often in the recent past is not more likely to appear on the next spin. Past frequency is a record of variance, not a predictor of the next outcome.
Can a betting system overcome the house edge in roulette?
No betting pattern changes the built-in edge of a negative-expectation game. Progression systems such as doubling after losses redistribute outcomes: most sessions end with a small win, and a few sessions end with a large loss that erases many prior wins. The expected value of the system is the same as flat betting the same total amount on the same bets.
Gambling involves risk. Please only gamble with funds that you can comfortably afford to lose. While we do our best to provide good quality and useful information for our readers, we cannot guarantee its accuracy. Gambling is not a way to solve financial difficulties. If you need support, contact a responsible gambling helpline such as the National Gambling Helpline at the helpline printed in the footer below or GamCare. Please check the footer for additional resources.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.