Roulette Mechanics

Why Every Roulette Bet Carries the Same Edge

A hypothetical stake shows why most standard roulette bets share one expected loss rate, so switching bets changes the swings, not the average cost.

On this page
  1. A cautionary note first
  2. The claim
  3. Checking it with a hypothetical stake
  4. What changes when you switch bets
  5. Why people believe some bets are cheaper
  6. The exceptions
  7. Total cost depends on amount wagered

A cautionary note first

Roulette is a game of chance with a built-in house edge. No strategy, system, pattern or timing changes the odds in the player’s favor over time, and most players lose money over time. Nobody should gamble money they cannot afford to lose. The point of this article is that picking a different bet does not make the game cheaper, not that any bet is a good one.

The claim

On a single-zero wheel, almost every standard bet has the same expected loss: 1 unit for every 37 units staked, about 2.70 percent. A straight-up bet, a corner, a dozen and a bet on red all cost the same on average. What differs is how that average is reached: how often the bet wins, how big the wins are, and how widely results swing along the way.

Checking it with a hypothetical stake

Use a stake of 10 units on a single-zero wheel, one bet per spin.

Straight up (35 to 1). It wins 350 units on 1 pocket and loses 10 on 36 pockets.

  • Total over all 37 pockets: 350 − 360 = −10.
  • Average per spin: −10 ÷ 37 ≈ −0.27 units.

Corner (8 to 1). It wins 80 on 4 pockets and loses 10 on 33.

  • Total: 4 × 80 = 320 won, 33 × 10 = 330 lost, so 320 − 330 = −10.
  • Average: −10 ÷ 37 ≈ −0.27 units.

Dozen (2 to 1). It wins 20 on 12 pockets and loses 10 on 25.

  • Total: 240 − 250 = −10.
  • Average: ≈ −0.27 units.

Red (1 to 1). It wins 10 on 18 pockets and loses 10 on 19.

  • Total: 180 − 190 = −10.
  • Average: ≈ −0.27 units.

Every row works out to −10 over the 37 equally likely pockets. This is no coincidence. Each bet’s payout is set so that winning pockets times (payout plus 1) equals 36, as though the wheel had 36 pockets. Because the wheel actually has 37, every bet falls short by exactly one stake’s worth over a full cycle of pockets. Because the payout ratios are all built from the same 36-pocket rule, the shortfall is the same everywhere.

What changes when you switch bets

If the average is fixed, what changes is the spread of results. Standard deviation measures that spread. For a 10-unit stake on a single spin:

  • Straight up: outcomes of +350 or −10. The standard deviation works out to about 58 units per spin.
  • Red: outcomes of +10 or −10. The standard deviation is about 10 units per spin.

Both bets lose about 0.27 units per spin on average. The straight bet gets there through rare, large wins and frequent small losses, so a short session is more likely to end well down, and occasionally far ahead. The red bet moves in small steps, so results stay closer to the average over a short stretch, though they are still uncertain. Neither is a better choice in terms of cost. This is the content of statistical variance: it changes the experience, not the price.

The law of large numbers means that as the number of spins grows, the average loss per unit staked settles toward the edge, whatever bets were used. The spread in total money does not vanish; it grows more slowly than the amount wagered, so the average cost shows up more clearly over a long run.

Why people believe some bets are cheaper

Several habits of thought make particular bets seem better.

  • Win frequency. Red wins nearly half the time, which feels like a better deal. But the smaller payouts per win offset the frequency.
  • Big payouts. A single number paying 35 to 1 feels generous, which is a framing effect. The chance against it is 36 to 1.
  • Memory. A few remembered wins can outweigh many forgotten losses, a form of selective memory.
  • Favorite numbers. Lucky numbers and rituals can feel as if they make a difference. They are among the gambling superstitions that have no effect on the result.
  • Mixing bets. Spreading chips over several bets does not average the edge away. A mix of bets that each cost 2.70 percent still costs 2.70 percent of the total staked. It also does not work as hedging bets in any sense that removes the cost.

The exceptions

The shared edge holds for the standard bets on a single-zero wheel with normal payouts. A few things break the pattern:

  • The double-zero wheel. All standard bets rise together to 2 in 38, about 5.26 percent.
  • The five-number bet on double-zero layouts. Covering 0, 00, 1, 2 and 3 typically pays 6 to 1, which gives (5 × 6 − 33) ÷ 38 = −3/38, about 7.89 percent, noticeably worse.
  • Partial-return rules. Surrender or en prison lowers the edge on even-money bets only, to roughly 1.35 percent on a single-zero wheel, while other bets stay at 2.70 percent.
  • Side bets and special formats. Added bets and altered payouts have their own edges, which are often higher.

Tables also differ in minimum and maximum stakes, so a variation in conditions is not the same as a variation in the edge.

Total cost depends on amount wagered

Because the edge is shared, the thing that determines how much is lost on average is the total amount staked: stake size times number of spins. Doubling the number of spins doubles the expected loss. A player who bets 10 units on 100 spins has staked 1,000 units, so the average loss is 1,000 × 1/37 ≈ 27 units on a single-zero wheel, whichever standard bets they used. The actual result could be far better or far worse in any one session.

For that reason, the useful decisions are about how much to spend and for how long, not which bet to place. Treat the money as an entertainment cost you can afford to lose, and stop when it is gone. If gambling is causing you worry, a local support service can help.

The complete guide to roulette mechanics

How Roulette Works: Wheel, Layout and Bets
Read the guide

Keep exploring

Guides from other topics

All topics →