Five reds in a row do not make black due. A doubling system cannot turn a negative-expectation bet into a guaranteed profit. And a roulette wheel that seems to favour one number is not necessarily defective. The common thread behind most roulette myths is confusing something that looks meaningful in a short sequence with something that actually changes the probability of the next spin.
The useful distinction is between chance, game rules and genuine irregularities . Chance produces streaks. Rules determine payouts and the house edge. An actual wheel defect or compromised game requires evidence, not a betting hunch. For New Zealand players, a further distinction matters: a website accepting NZ dollars is not proof that it holds a New Zealand online casino licence.
First, what are the actual roulette odds?
In standard European roulette, 18 pockets are red, 18 are black and one is green zero. A fair wheel therefore gives red an 18/37 chance (48.65%) on every spin, not 50%. American roulette adds a second green pocket, double zero (00), giving red an 18/38 chance (47.37%) . A zero loses a conventional red-or-black wager.
| Wheel or rule | Pockets | Chance of red | House edge |
|---|---|---|---|
| European single-zero | 37 | 48.65% | 2.70% |
| American double-zero | 38 | 47.37% | 5.26% |
| Triple-zero variant | 39 | 46.15% | 7.69% on standard bets |
| Single-zero with La Partage | 37 | 48.65% | 1.35% on eligible even-money bets |
The house edge describes the expected share of total amounts wagered retained by the casino over many plays, not a fee deducted from each winning wager and not a prediction of an individual session. On a fair European wheel, a NZ$10 red bet has an expected loss of approximately NZ$0.27 per spin; the actual result is normally either a NZ$10 loss or a NZ$10 profit. This difference between expected value and variance explains why a lucky evening does not invalidate the underlying maths.
The 15 most persistent roulette myths
1. “After several reds, black is due”
Verdict: False for a fair wheel. This is the gambler's fallacy : believing that independent events must compensate for one another immediately. After seven red results, the chance of black on the next spin of a European wheel remains 18/37. The wheel has no memory of the sequence.
The law of large numbers says proportions tend towards their theoretical probabilities over a very large number of trials. It does not say that a short streak must be corrected by the next outcome. A roulette history board can accurately show past results while providing no predictive advantage.
2. “Hot numbers are more likely to land again”
Verdict: False under normal random conditions. A number that appeared four times recently is a hot number only as a description of history. It still has a 1/37 chance on the next fair European spin. Likewise, cold numbers are not overdue.
Pattern recognition is useful in daily life but misleading when applied to independent random outcomes. The hot-hand fallacy and confirmation bias can make a few memorable hits feel like strong evidence. Roulette statistics and number-frequency charts are descriptive, not a forecast.
3. “The Martingale system guarantees a win”
Verdict: False. Martingale doubles an even-money stake after each loss, seeking to recover losses plus the original unit after the eventual win. Starting at NZ$5, successive wagers are NZ$5, NZ$10, NZ$20, NZ$40, NZ$80 and NZ$160. Losing all six means NZ$315 has gone, and the next required stake is NZ$320.
A limited bankroll, maximum table limits and the possibility of an extended losing run stop the progression from being risk-free. The method often produces small wins, but occasional large losses can overwhelm them. Increasing stakes never removes the zero pocket or changes the underlying house edge.
4. “Other betting systems beat the house edge”
Verdict: False as a general claim. Fibonacci, D'Alembert, Labouchère and the Paroli or Reverse Martingale systems change how wagers grow or shrink. The James Bond strategy combines several betting areas. None alters the probabilities or payouts of a fair wheel.
Positive progressions increase stakes after wins; negative progressions increase them after losses. They create different volatility, drawdowns and bankroll requirements, not a reliable positive expected value. Keeping a fixed stake is easier to understand financially, but it does not transform roulette into a profitable investment.
5. “A skilled player can predict the next number”
Verdict: Not in ordinary fair play. Roulette is not a skill game in the way poker can be. Choosing a straight-up bet, split, street, corner, dozen or column affects hit frequency and payout size. It does not enable a player to steer a fair spin.
Some historical studies of physical roulette explored whether ball speed, wheel speed and the release point could give partial information about landing sectors. That does not establish a reliable method for contemporary casino play, where deflectors, bouncing and operational controls create considerable uncertainty. A claim of perfect prediction, especially for online RNG roulette, warrants scepticism.
6. “A croupier can make the ball land on a chosen pocket”
Verdict: Unsupported as a routine casino practice. A live dealer or croupier controls when and how the roulette ball is released, but not a precise pocket after multiple collisions with diamonds or deflectors. So-called dealer signatures claim that consistent spins cluster around certain wheel sections. Ordinary variation can produce apparently convincing clusters without proving control.
Suspected manipulation needs an evidence-based investigation of the equipment and procedures. Picking numbers based on a dealer's hand movement does not eliminate uncertainty.
7. “Every roulette table has the same odds”
Verdict: False. Wheel layout and special rules matter more than the colour of the felt or the size of the minimum bet. A single-zero European wheel has a smaller standard house edge than an American double-zero wheel. Some US venues also use a triple-zero format, which is worse again for conventional payouts.
French roulette sometimes offers La Partage , returning half an even-money stake when zero wins, cutting the edge on those wagers to 1.35%. An En Prison rule instead holds the stake over for a subsequent spin; its precise effect depends on the table's treatment of repeat zeros. Check the rules displayed for the actual table rather than assuming every game labelled “French roulette” includes the same protection.
8. “Betting red and black together guarantees profit”
Verdict: False. Betting both colours in equal amounts generally cancels the winning and losing payouts when a colour lands. If green zero lands, both standard wagers lose. The same issue undermines simplistic hedging arrangements across odd/even or high/low.
Covering more pockets can increase the probability of collecting some payout, but that is different from earning a net profit. Payouts are calibrated to the number of covered outcomes and preserve the casino advantage.
9. “Inside bets are worse than outside bets”
Verdict: Usually false when comparing expected value on the same standard wheel. A straight-up number generally pays 35 to 1 but hits infrequently; an even-money red-or-black bet hits far more often but pays 1 to 1. On a standard single-zero wheel, both have the same 2.70% theoretical house edge.
The distinction is risk profile . Inside bets such as splits, streets and corners tend to produce more volatile results than broad outside bets. A La Partage rule creates a meaningful exception because it can improve the value of eligible even-money wagers without improving ordinary straight-up bets.
10. “Online roulette is always rigged”
Verdict: Too broad to be true. A properly implemented random number generator (RNG) is designed to select outcomes without using a player's previous wins or losses. Live dealer roulette, by contrast, relies on a physical wheel and video stream. Neither format is automatically trustworthy merely because a site calls it fair.
Look for independently documented game testing, clear rules and regulatory oversight that can be verified with the relevant authority. A displayed return-to-player (RTP) figure is a theoretical model, not evidence that a specific short session will return that amount. Unverified claims about “AI prediction software”, hidden algorithms or special winning hours should be treated as marketing, not proof.
11. “A wheel can never be biased”
Verdict: False in principle; detecting it is difficult. Real equipment can have physical imperfections, including pocket wear, levelling problems or mechanical inconsistencies. Historical cases of wheel bias explain why the idea exists.
However, several apparent repeats are not sufficient evidence. To demonstrate that a specific wheel deviates from expected behaviour would require a substantial and carefully collected sample, suitable statistical testing and checks for changing operating conditions. Modern maintenance and inspections make casual exploitation an unreliable assumption. Physical wheel bias should not be confused with the normal appearance of streaks on a fair wheel.
Myth 12: Lucky numbers and rituals change outcomes
Verdict: False as a probability claim. Birthdays, the number seven, a particular seat, lucky charms and placing chips in a familiar order may carry personal meaning. They do not alter the physical or digital selection process.
The illusion of control can make a ritual feel effective, particularly after a memorable win. The same applies to the belief that changing tables after a loss resets the odds. A new table can have different rules, but a different chair or ritual does not confer a mathematical advantage.
13. “A winning streak proves your strategy works”
Verdict: False. Short-term results are noisy. A player can win several spins using a losing strategy because random variation occasionally favours them. This is not evidence of a long-term edge.
A valid evaluation would need a clearly specified rule, accurate records of all wagers, net results rather than selected wins, sufficient observations and an honest assessment of chance. Even a genuine recent profit says nothing definite about the next spin. The tendency to remember hits and forget misses is another form of confirmation bias.
14. “Near misses mean a win is getting closer”
Verdict: False. A ball stopping one pocket away from your chosen number can feel more significant than landing across the wheel. Yet a near miss is still a loss, and pocket adjacency offers no carry-over benefit for a future independent spin.
Wheel order and the numerical layout on the betting grid are not the same thing. Neighbour bets cover physically adjacent wheel pockets, but they remain ordinary combinations of wagers with real costs and no built-in predictive power. A close-looking result should not be a reason to chase losses.
15. “More spins will eventually recover your losses”
Verdict: False. More play creates more opportunities for both wins and losses, while also increasing cumulative exposure to a negative expected value. The law of averages is not a repayment mechanism.
Suppose a player places 200 separate NZ$10 bets on a standard European wheel: NZ$2,000 is wagered in total, even if money from previous wins is recycled. At a 2.70% house edge, the theoretical expected loss across those bets is about NZ$54. Actual outcomes vary, but the expectation becomes more negative as turnover increases. No loss-chasing strategy makes the casino owe you a recovery.
Which roulette claims actually change a decision?
Most myths fail because they try to extract a future prediction from past outcomes. A smaller group points towards something worth checking: whether the game uses one, two or three zeros; whether La Partage or En Prison genuinely applies; whether payouts match the stated rules; and whether the operator's credentials are independently verifiable.
| Claim or observation | What it really tells you | Useful check |
|---|---|---|
| “Black is due after red” | Nothing about the next fair spin | Compare the claim with the unchanged 18/37 probability |
| “This system wins most sessions” | Win frequency alone hides potential large losses | Include total stakes, worst drawdown and table limit |
| “French roulette has better odds” | Only particular rules and bet types may improve the edge | Read the zero-handling rules for that game |
| “The site is licensed” | A statement, not independent evidence | Check the appropriate regulator's records |
| “The wheel is biased” | A testable physical hypothesis, not a conclusion | Demand reliable data and equipment checks |
What changes for roulette players in New Zealand?
Roulette mathematics is the same in Auckland, Wellington and Christchurch as anywhere else: NZD stakes do not affect pocket probabilities. The relevant local difference is the legal and consumer-protection environment.
New Zealand's Online Casino Gambling Act 2026 came into force on 1 May 2026. The Department of Internal Affairs (DIA) states that the licensing regime is being introduced in stages and is not expected to become fully operational until 2027. During the transition, an overseas gambling site accepting Kiwi players or showing NZ$ amounts should not be described as holding a New Zealand licence without verifiable proof. The broader Gambling Act 2003 also remains part of New Zealand's gambling law.
Where an online roulette product is offered, check who operates it, which authority regulates it, whether the licence can be independently confirmed, how complaints are handled and what safeguards are available. The 2026 Act provides for an age threshold of 18 for licensed online casino gambling. A badge, an unfamiliar testing logo or a promise of “guaranteed fair wins” is not a substitute for verification.
A better response to roulette superstition
Randomness can be frustrating precisely because it supplies no reliable short-term signal. Clear rules and financial limits are more useful than theories about the next spin. If you choose to play, decide your maximum affordable loss before starting, keep track of net spending rather than headline wins, avoid increasing stakes to recover losses, and stop when the planned limit is reached. Deposit limits, time-outs and self-exclusion are safeguards, not betting techniques.
Gambling should not be used to solve debt or financial pressure. If play is becoming difficult to control, Safer Gambling Aotearoa provides information and support. New Zealand's free 24/7 Gambling Helpline is available on 0800 654 655 or by free text to 8006 .
Roulette myths: frequently asked questions
Can roulette results be predicted from previous spins?
No reliable prediction follows from the previous results of a fair, independent roulette game. A string of reds, repeated zeros or an unusually frequent number may look notable, but it does not change the probability assigned to the next spin.
Is there any roulette strategy that guarantees a profit?
No betting progression guarantees profit in standard roulette. Martingale, Fibonacci, D'Alembert and similar strategies change bet sizing rather than the house edge; table limits and bankroll limits make promised recovery particularly misleading.
What is the safest type of roulette bet?
No wager eliminates gambling risk. Even-money outside bets hit more often and tend to be less volatile than straight-up numbers, but they can still lose, including when zero lands. If an eligible La Partage rule applies, it reduces the expected loss on those even-money bets.
Is European roulette better than American roulette?
For conventional bets with standard payouts, European single-zero roulette has a lower house edge (2.70%) than American double-zero roulette (5.26%). This is a comparison of theoretical cost, not a promise that a particular player will win more often in a short session.
Do online roulette algorithms change after a player wins?
A properly tested RNG game should not adjust probabilities in response to an individual's wins. But fairness should be assessed through verifiable regulation, independent testing and transparent game rules rather than relying on either an operator's promises or conspiracy claims.
Does roulette have a memory?
A fair roulette wheel does not. Previous outcomes are historical observations, and each new spin starts with the same pocket probabilities. A genuine physical defect is a different proposition and requires evidence independent of short winning or losing streaks.
Are online roulette casinos legal for New Zealand players?
New Zealand introduced the Online Casino Gambling Act 2026 and is moving towards a licensed online casino regime, with staged implementation extending into 2027. The rules affecting operators and players are not identical. Check the latest guidance from the Department of Internal Affairs rather than assuming that an accessible website is locally licensed.
The underlying rule is simple: a fair spin cannot be made more predictable by a story about previous spins. The meaningful differences are in game rules, expected cost, operational integrity and the limits you set for yourself.
