Roulette’s zero is not an extra red-or-black result. It is the pocket that causes most even-money bets to lose while those bets pay only 1:1. That mismatch creates the standard house advantage. A second zero increases the mismatch, while special rules such as La Partage can reduce the loss on eligible bets when zero appears.
The basic layout is explained in GambleRoad’s roulette rules and odds guide. This article isolates the zero because many systems discuss colour streaks or payout patterns without identifying the single rule that determines the long-run cost.
Single-zero roulette
A European-style wheel has 37 pockets: numbers 1 through 36 plus zero. Eighteen numbered pockets are red and eighteen are black. An even-money colour bet therefore wins on 18 pockets and loses on 19. The payout remains one unit for one unit.
The expected result on a one-unit colour bet is (18/37 × 1) minus (19/37 × 1), which equals minus 1/37, or approximately 2.70%. The same standard edge applies to most other ordinary bets because their payouts are set against 37 possible outcomes.
Double-zero roulette
An American-style wheel adds 00, creating 38 pockets. A red wager still covers 18 pockets but loses to 18 black numbers, zero and double zero. The expected loss becomes two units across 38 equally likely one-unit outcomes, or approximately 5.26%.
| Wheel or rule | Pockets | Red wins | Red loses | Approximate edge |
|---|---|---|---|---|
| Single zero | 37 | 18 | 19 | 2.70% |
| Double zero | 38 | 18 | 20 | 5.26% |
| Single zero with La Partage | 37 | 18 | 18 plus half loss on zero | 1.35% on eligible even-money bets |
| Triple-zero example | 39 | 18 | 21 | 7.69% on standard even-money terms |
Why the payout table does not compensate for zero
A straight-up number pays 35:1 rather than the fair 36:1 on a 37-pocket wheel. A two-number split pays 17:1, a four-number corner pays 8:1 and a dozen pays 2:1. The omitted unit across the payout schedule is another way to see the single-zero edge.
Adding 00 without increasing ordinary payouts doubles the disadvantage on standard bets. This is why the wheel type should be checked before stake size or betting pattern. A visually attractive table with faster dealing can still be materially more expensive per unit wagered.
The five-number bet is a special American-wheel trap
The basket covering 0, 00, 1, 2 and 3 is commonly paid at 6:1. It covers five of 38 pockets but returns less than a fair price, creating an edge of about 7.89%. It is worse than the usual 5.26% American-wheel edge.
A rule summary that says “all bets have the same edge” is therefore incomplete. Most standard bets share the main edge, but special combinations can be worse. Check the paytable rather than assuming a familiar shape has a standard price.
La Partage and En Prison
Under La Partage, an eligible even-money bet loses only half its stake when zero appears. That reduces the standard single-zero edge to about 1.35%. GambleRoad’s La Partage guide explains the settlement in detail. En Prison can hold the stake for a subsequent spin instead, depending on the table rules.
These rules do not improve straight-up, dozen or column bets. They also may be unavailable online or restricted to particular tables and stake levels. Confirm the help screen and do not infer the rule from a “French roulette” label alone.
Zero does not become due or avoidable
A run without zero does not increase its next-spin probability on a properly operating independent game. A recent zero does not make another zero less likely. Betting around zero, switching colours after zero or pausing because zero is “hot” does not remove the edge.
The scoreboard can show the observed frequency, but a small sample is not a reliable wheel audit. Physical wheel bias is a specialist measurement problem involving thousands of observations, equipment control and conditions that online players generally cannot verify.
How zero changes progression systems
A progression on red or black often treats zero as an unlucky interruption. Mathematically it is not an exception; it is part of the losing probability. Doubling after zero increases the next stake without changing the chance of the colour. Table limits and finite bankroll eventually prevent indefinite recovery.
Systems that place a separate hedge on zero pay for that hedge on every spin. The added bet can change volatility and the distribution of outcomes, but it does not remove the combined house advantage unless a genuine promotional or rule exception is quantified.
Expected loss should be translated into turnover. At a 2.70% edge, $1,000 of total standard single-zero wagers has a theoretical average cost of about $27 over a very large sample; at 5.26%, the equivalent average is about $52.60. A single session can finish far above or below those values, but the comparison shows why wheel selection matters repeatedly.
Online interfaces can obscure the wheel type by emphasizing table name or speed. Open the rules and inspect the numbered layout rather than relying on a thumbnail. Live-dealer tables may state the zero rule in a help panel, and automated games may offer several versions with similar graphics. The exact paytable is the controlling evidence.
Promotions can alter effective cost temporarily, but only after contribution, maximum-bet and withdrawal conditions are included. A bonus that excludes roulette or counts it at a low rate does not neutralize the zero. Keep game mathematics and promotion value as separate calculations, then combine them only when every rule is verified.
Zero also affects insurance-style side rules differently across tables. Some games return half the stake automatically, some imprison the bet and some offer no concession. The label and interface can be similar, so the exact written rule should be read before wagering. A rule that applies only to red, black, odd, even, high and low should not be extended to columns or dozens.
Game speed multiplies the zero’s effect. A rapid automated table can process many more wagers per hour than a physical table, so the expected cost per hour can be higher even with the same edge and stake. Comparing only minimum bets misses this turnover effect. Record spins per hour or total amount wagered when choosing between formats.
The zero is also central to marketing claims about “low house edge.” That phrase may refer to one eligible bet under one rule, not the entire table. Side bets, special combinations and double-zero variants can have different costs. Verify the exact wager rather than applying the best advertised percentage to every option.
Use the zero to compare tables before playing
- Identify whether the wheel has one, two or more zero pockets.
- Check for La Partage, En Prison or other zero-settlement rules.
- Avoid unusually costly special combinations unless the entertainment value justifies them.
- Calculate expected cost from total turnover, not only the initial chip value.
- Use a fixed stake and session limit rather than a zero-triggered progression.
- Retain the paytable when disputing an online settlement.
The clearest table choice is usually a single-zero game with favourable rules, assuming the game and operator are otherwise equivalent. GambleRoad’s roulette variants comparison places zero structure alongside layout and settlement differences. No choice eliminates variance, but avoiding additional zeros reduces the mathematical cost on every qualifying unit wagered.