Start with the wheel, not the stake
A 100-unit bet on roulette does not have one answer. The amount you stake matters less than where you put it, because every position on the layout pays at odds that are slightly worse than the true odds of the wheel. The wheel itself is the fixed part of the equation. A European wheel has 37 pockets, numbered 0 to 36, with a single zero. An American wheel has 38 pockets, because it adds a double zero. That extra pocket is not a detail; it changes the arithmetic on every bet you can make.
When you place 100 units on a single number and it hits, the payout is 35 to 1, so you receive 3,500 units in winnings plus your original 100 back. That sounds generous until you count the pockets. On a European wheel, the true odds against any single number are 36 to 1, because there are 37 pockets and only one of them wins. The casino pays 35 to 1. The gap between 36 and 35 is the house edge, built into the payout schedule rather than applied as a separate charge.
How the edge is built into each bet
Roulette does not take a commission. The edge sits quietly inside the difference between what a bet pays and what the wheel's geometry says it should pay. For a single-number bet on a European wheel, the payout is 35 to 1 while the true odds are 36 to 1. For a split bet covering two numbers, the payout is 17 to 1 while the true odds are 17.5 to 1. The pattern repeats across the layout: every payout is set just below the true odds.
House edge and RTP are complements. They sum to 100% for the same bet under the same rule set. If a game returns 97.3% to players over a very large number of spins, the house edge is 2.7%. That is the long-run statistical average, not a per-session guarantee. A single spin has no average; it has an outcome.
| Bet type | Payout | True odds (European wheel) | What the gap means |
|---|---|---|---|
| Single number | 35 to 1 | 36 to 1 | The payout is one unit short of the true odds |
| Split (two numbers) | 17 to 1 | 17.5 to 1 | The payout is half a unit short |
| Street (three numbers) | 11 to 1 | 11.33 to 1 | The payout is a third of a unit short |
| Red or black | 1 to 1 | Slightly worse than 1 to 1 | The zero pocket means the even-money bet is not truly even |
The American wheel makes every one of these gaps wider, because the double zero adds another losing pocket without changing any payout. The same 100-unit bet on the same number faces 37 losing pockets instead of 36.
What a 100-unit bet returns
Suppose you place 100 units on red. The payout is 1 to 1, so a win returns 100 units in profit plus your stake. On a European wheel, there are 18 red pockets, 18 black pockets, and one green zero. The zero is the reason the bet does not pay at true 50/50 odds. On an American wheel there are two green pockets, which moves the odds further against the bet.
As an example, imagine 37 spins of a European wheel, each with 100 units on red. If every number came up once, red would win 18 times and lose 19 times. The 18 wins return 1,800 units in profit. The 19 losses cost 1,900 units. The net result is a loss of 100 units across 3,700 units wagered. That is the edge made visible, and it is exactly why the payout schedule is built the way it is.
For a single-number bet, the same thought experiment works differently. One win in 37 spins pays 3,500 units. The other 36 spins lose 100 each, for 3,600 units lost. The net is a 100-unit loss across 3,700 units wagered. The edge is the same rate on both bets under the same wheel, even though the payouts and the variance are completely different.
Short sessions look nothing like long-run averages
The long-run average is a mathematical property of the wheel, not a promise about your evening. A 100-unit bet on a single number can return 3,500 units, nothing, or anything in between, and none of those outcomes tells you the edge. The edge only emerges over a very large number of spins, far more than anyone plays in a single sitting.
Variance is what you feel in a short session. A single-number bet has high variance: most spins lose, and the rare win is large. An even-money bet has lower variance: wins and losses come in a rough alternation, and the zero pocket slowly drains the difference. The edge is the same rate on both, but the experience of playing them is not.
This is why a player can leave ahead after an hour and still have been playing a negative-expectation game. The short run is dominated by luck. The long run is dominated by the payout schedule.
Decisions per hour change total exposure
The edge is expressed as a percentage of each wager, which means the total amount you wager matters as much as the percentage. A 100-unit bet placed once is a 100-unit exposure. A 100-unit bet placed every spin for an hour is a much larger exposure, because the edge applies to every unit that crosses the layout.
Roulette is a fast game. A busy table can produce a spin every minute or faster, and an online game can run faster still. If you bet 100 units per spin and the wheel spins 40 times in an hour, you have wagered 4,000 units. The edge applies to that total, not to the 100 units you sat down with. The same principle applies to any game: decisions per hour multiply exposure, and exposure is what the edge acts on.
Slowing down does not remove the edge. It reduces the total amount wagered in a given time, which reduces the expected loss in absolute terms while leaving the rate unchanged. A player who bets less often is not beating the game; they are simply exposing less money to it.
Why progression systems do not change the edge
A progression system is a rule for changing your stake after wins or losses. The most famous is the martingale, where you double your bet after every loss so that a single win recovers all previous losses plus one unit. The arithmetic looks convincing on paper, and it collapses in practice for two reasons.
First, the edge is unchanged. Doubling your bet does not alter the payout schedule or the number of pockets on the wheel. A red bet on a European wheel still faces 18 winning pockets and 19 losing ones, no matter how large the stake. The progression changes the size of the wager, not the odds of the wager.
Second, progressions redistribute outcomes rather than remove them. Most of the time, a martingale session ends with a small win. Occasionally, it ends with a very large loss, because a losing streak forces the stake up until it hits the table limit or the player's bankroll runs out. The small wins and the rare large loss balance out to the same edge as flat betting. The system changes the shape of the outcomes, not their average.
As an example, suppose a player starts with a 100-unit bet on red and doubles after each loss. After four losses in a row, the next bet would be 1,600 units. A fifth loss costs 3,100 units in total across the streak. A win at that point returns 1,600 units, leaving the player down 1,500 units overall. The progression has not recovered anything; it has simply concentrated the loss into a single larger bet.
What the player controls
No betting pattern changes the built-in edge of a negative-expectation game. What a player does control is narrower, and it is worth being precise about it. Bet selection matters, because different bets on the same table carry different edges, and on roulette the wheel type sets the baseline. A European wheel carries a smaller edge than an American wheel on the same bets, because the American layout has one more losing pocket.
Stake sizing matters, because the edge applies to every unit wagered. Session length matters for the same reason. Stopping rules matter, because they determine when a session ends, but they do not change what the session would average over many repetitions. None of these are winning strategies. They are ways to control how much money is exposed to a game whose edge is fixed.
If you want to know the exact payout for a specific bet, the table layout itself carries the answer. The payout schedule is printed on the felt or shown in the game information panel, and it is the same document the casino uses to settle every spin. The wheel type is also stated on the layout or in the game rules. Those two pieces of information are enough to work out the edge on any bet you are considering.
FAQ
Does a 100-unit bet on a single number pay more than 100 units on red?
Yes, when it wins. A single-number bet pays 35 to 1, so a 100-unit stake returns 3,500 units in profit. A red bet pays 1 to 1, so the same stake returns 100 units in profit. The single-number bet also loses far more often, because it has many more losing pockets.
Can I calculate the expected loss on a 100-unit roulette bet?
You can, using the edge for the wheel and the bet. The edge is the percentage of each wager the house keeps on average over a very large number of spins. Multiply the edge by the amount wagered to get the expected loss for that wager. The result is a long-run average, not a prediction for any single spin.
Does betting 100 units on two different spots change the edge?
It changes the variance and the chance of winning on a given spin, but not the edge. Each spot carries its own payout and its own true odds, and the gap between them is the edge. Covering more of the layout reduces the size of the swings but does not remove the house advantage.
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.