Every bet is a small lottery with published tickets
When you put a chip on a roulette number, you are buying one ticket in a tiny lottery that resolves in under a minute. The wheel has a fixed set of pockets, the ball lands in exactly one of them, and the payout is printed on the table. The probability distribution of that bet is simply the complete list of what can happen, how likely each result is, and what each result pays. The distribution is not hidden and it does not change from spin to spin. What changes is which slice of it you buy.
A probability distribution sounds abstract, but on a roulette layout it is concrete. Pick a bet, and you can write down every outcome. For a single-number bet on a European wheel, there are two outcomes: the ball lands on your number, or it lands on one of the other 36 pockets. The probabilities come straight from the wheel's construction. What surprises most players is not the probabilities themselves but what happens when you multiply them by the payouts.
Reading the wheel before reading the payouts
The first thing a probability distribution needs is the set of possible results, and in roulette that set is fixed by the wheel. A European wheel has 37 pockets: the numbers 0 through 36. An American wheel has 38, adding a double zero. Every bet on the layout is a wager that the ball will land in some subset of those pockets. The size of the subset, divided by the total number of pockets, is the probability of winning that bet.
This is why the wheel type matters before anything else. The same bet placed on a European and an American wheel has a different probability, because the denominator changes while the winning subset does not. A single number is one pocket out of 37 on one wheel and one pocket out of 38 on the other. The payout printed on the felt is typically the same, which means the bet's distribution changes while its advertised reward does not. The extra pocket does not create a new way for you to win; it creates a new way for the ball to miss every bet on the layout.
| Bet | Winning pockets (European, 37 total) | Winning pockets (American, 38 total) | Typical payout |
|---|---|---|---|
| Single number | 1 | 1 | 35 to 1 |
| Split (two numbers) | 2 | 2 | 17 to 1 |
| Street (three numbers) | 3 | 3 | 11 to 1 |
| Red or black | 18 | 18 | 1 to 1 |
| Even or odd | 18 | 18 | 1 to 1 |
| Dozen | 12 | 12 | 2 to 1 |
The table shows the first half of the distribution: how many pockets win. The second half is what those wins pay, and the mismatch between the two is where the operator's edge lives.
Why the payout does not match the probability
A fair bet would pay exactly according to the odds against winning. On a European wheel, a single number has 36 losing pockets and 1 winning pocket, so the odds against it are 36 to 1. A fair payout would be 36 to 1. The table pays 35 to 1. That one-unit gap, applied to every spin over a very large number of spins, is the house edge on that bet.
The same arithmetic runs through the whole layout. A red or black bet on a European wheel wins on 18 pockets and loses on 19, because the zero is neither red nor black. The odds against winning are 19 to 18, but the payout is 1 to 1, which pays as if the odds were 18 to 18. The difference is the edge. No bet on the layout escapes this structure. Some bets have a smaller edge than others, but every bet carries one, and the edge is built into the printed payout rather than added later.
The relationship between house edge and RTP is worth stating precisely because it is often blurred. For the same bet under the same rule set, house edge and RTP sum to 100%. If a game returns 97.3% over the very long run, the house edge is 2.7%. That does not mean you will lose exactly 2.7% of your session. It means that if the bet were repeated an enormous number of times, the total returned would trend toward that percentage of the total wagered. Any single session is a small sample, and small samples wander far from the average.
Building the distribution with example arithmetic
Suppose you place a single-number bet on a European wheel. As an example, imagine a 1-unit wager. The distribution has two outcomes. The ball lands on your number with probability 1 in 37, and you receive 35 units in profit plus your original unit back. The ball lands elsewhere with probability 36 in 37, and you lose the unit. The expected value of the bet is the sum of each outcome multiplied by its probability.
The arithmetic runs like this: (1/37 × 35) + (36/37 × −1). The first term is 35/37, or about 0.946. The second term is −36/37, or about −0.973. Add them together and the expected value is −1/37, roughly −0.027 units per unit wagered. That is the house edge expressed as a fraction of the bet. It is not a fee deducted from your balance; it is the gap between the payout and the true odds, realised over repetition.
Run the same example on an American wheel and the arithmetic shifts. The win probability becomes 1 in 38 and the loss probability 37 in 38. The expected value is (1/38 × 35) + (37/38 × −1), which simplifies to −2/38, or about −0.0526 units per unit wagered. The payout did not change, but the added pocket nearly doubled the expected loss per spin on that bet. This is why the wheel type is the first fact to check, not the last.
What short sessions do to the distribution
A probability distribution describes the long run. A session of fifty spins is not the long run. In fifty spins of a European wheel, a single-number bet has a reasonable chance of never hitting at all, and a small chance of hitting two or three times. Neither outcome contradicts the distribution. The distribution is the weighted average of all possible fifty-spin sequences, and most individual sequences sit nowhere near that average.
This is the gap between expected value and experienced value. Expected value is a property of the bet. Experienced value is a property of one short sequence. A player who hits a number on the third spin is not beating the house edge; they are drawing a favourable sample from a distribution whose average is still negative. A player who misses for eighty spins is not being cheated; they are drawing an unfavourable sample from the same distribution. The edge does its work invisibly, across thousands of spins, while any single session is dominated by variance.
Variance is the word for how spread out the outcomes are. A single-number bet has high variance: most spins lose the whole unit, and rare spins return 35. A red or black bet has lower variance: wins and losses are nearly balanced in frequency, and each result moves the balance by a small amount. Both bets can carry the same house edge on the same wheel, but they feel completely different in a short session. The distribution explains why: the same average is built from very different shapes.
Decisions per hour change exposure, not the edge
The house edge is quoted per bet, but your total expected loss depends on how many bets you make. A roulette table resolves a spin every minute or faster, and a player who bets every spin is multiplying their exposure. If a game carries an expected loss of 0.027 units per unit wagered, then wagering 100 units across an evening produces an expected loss of roughly 2.7 units. Wagering 1,000 units produces an expected loss of roughly 27 units. The rate did not change; the volume did.
This is why game speed matters as much as game choice. A slow table with fewer spins per hour exposes the same bankroll to fewer resolutions of the same distribution. A fast table, or a player who covers many numbers per spin, multiplies the number of tickets bought from the same negative-expectation lottery. None of this changes the probability distribution of any individual bet. It changes how many times you sample it, and therefore how quickly the long-run average asserts itself.
Bankroll sizing is a method for controlling how long you can keep sampling, not a method for changing the sample's average. A player who brings a fixed amount and bets small units will, on average, play longer than one who bets large units, but both face the same edge on every spin. Stopping rules work the same way. Deciding to leave after a win or a loss changes when you stop sampling the distribution; it does not change the distribution itself.
Why progression systems redistribute outcomes
A progression system is a rule for changing bet size after wins or losses. The most famous is doubling after a loss, on the theory that a win will eventually recover all previous losses plus one unit. The system does not alter the probability of any spin. The wheel does not know your bet size, and the ball does not owe you a red because the last six spins were black. What a progression does is reshape the distribution of your session outcomes.
Under a doubling progression, most sessions end with a small win, because most losing streaks end quickly. A minority of sessions end with a very large loss, because the required bet size grows past the table limit or the player's bankroll before the streak breaks. The average of all those sessions is still negative, and it is still governed by the house edge on the underlying bets. The progression has not removed the edge; it has concentrated it into rare, severe losses while sprinkling frequent, small wins. The player experiences a different shape of outcomes, but the same expected value.
The same logic applies to any betting pattern. Covering more numbers per spin changes the probability of winning that spin, but it also changes the payout when the win arrives, and the edge on each covered number remains. There is no arrangement of chips on the layout that turns a set of negative-expectation bets into a positive-expectation portfolio. The distribution of each bet is fixed, and the sum of fixed negative expectations is still negative.
Where the numbers are printed
Every figure a player needs is published somewhere on the table or in the game's information panel. The wheel type is visible on the wheel itself. The payouts are printed on the felt. The table minimum and maximum are posted at the table or in the game lobby. The game's RTP, where an online game publishes one, appears in the information or help panel, and it applies to the specific rule set in use. A player who wants the exact house edge on a particular bet reads the wheel type and the payout, then does the arithmetic shown above.
What no document will show is a guarantee about the next spin. The probability distribution is a complete description of the bet's long-run behaviour, and it contains no information about which pocket the ball will find on any given spin. That is the point of a distribution: it describes the whole set of possibilities, not the next draw from it.
FAQ
What is the difference between probability and payout in roulette?
Probability is the chance a bet wins, based on how many pockets it covers out of the wheel's total. Payout is what the table pays when it wins. The payout is always slightly lower than the true odds against winning, and that gap is the house edge.
Does the probability distribution change if I bet on more numbers?
Betting on more numbers changes which distribution you are sampling, not the distribution of the wheel. A bet covering 18 numbers wins more often than a bet covering one, but it pays less, and each covered number carries its own edge. The total expected loss scales with the total amount wagered.
Can a progression system change the probability distribution of roulette?
No. A progression changes the size of each bet, not the probability of any spin. It reshapes the distribution of session outcomes, creating many small wins and rare large losses, while the average across all sessions stays negative and governed by the house edge on the underlying bets.
Gamble responsibly. If you or someone you know has a gambling problem, call the helpline printed in the footer below for confidential support.
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.