Published: March 24, 2026 | Updated: 3 October 2026 | Reading time: 7 minutes
Almost everyone who has gambled remembers a streak: a run of reds at the roulette table, a slot that paid three times in ten spins, an evening when nothing went right. Streaks feel like they mean something. They seem to say that the table is “hot”, that a change is “due”, or that it is time to bet more.
This article looks at streaks from two sides: the maths of how often runs actually happen, and the psychology of why our brains read so much into them. All numbers below are pure probability, and each example states its assumptions.
About GanaWin: the casino this site was built around has closed. Its own domains (ganawin.com, ganawin.bet) now show only a page recommending a different brand, MangoDorado. This article is general information about how gambling works — it is not an invitation to play at GanaWin or anywhere else. What the closure means for former players is explained on our homepage.
What a streak really is
In casino games such as roulette or slots, each result is designed to be independent of the previous one. A roulette wheel has no memory, and a slot’s random number generator (RNG) does not track what it produced a minute ago. A streak is simply a stretch of similar results that happened to line up. It is a description of the past, not a signal about the future.
The trouble is that independent random sequences contain far more runs than most people expect. When we see one, it feels like a pattern, and the brain is very good at finding patterns, including ones that are not there.
How rare is a run on a single try?
Assumptions: European roulette with 37 pockets (18 red, 18 black, one green zero), a fair wheel and independent spins. The probability of red on one spin is 18/37 ≈ 48.6%.
The chance of several reds in a row is that probability multiplied by itself:
| Reds in a row (from a given spin) | Formula | Probability |
|---|---|---|
| 3 | (18/37)³ | ≈ 11.5% |
| 5 | (18/37)⁵ | ≈ 2.7% |
| 7 | (18/37)⁷ | ≈ 0.64% |
| 10 | (18/37)¹⁰ | ≈ 0.074% |
Five reds starting from a particular spin is a roughly 1-in-37 event. That sounds rare, and on a single attempt it is.
Rare per spin, common per session
A session is not a single attempt. Every spin is a new starting point for a possible run, so over many spins the chance of seeing at least one long run grows quickly. The table below was calculated exactly (not simulated) under the same assumptions.
| Spins in the session | At least one run of 5+ reds | 5+ of the same colour | 8+ of the same colour | 10+ of the same colour |
|---|---|---|---|---|
| 37 | ≈ 40% | ≈ 66% | ≈ 9.7% | ≈ 2.2% |
| 100 | ≈ 77% | ≈ 96% | ≈ 27% | ≈ 6.8% |
| 200 | ≈ 95% | ≈ 99.8% | ≈ 47% | ≈ 14% |
| 500 | ≈ 99.9% | > 99.9% | ≈ 80% | ≈ 31% |
In other words, a run of five reds somewhere in 100 spins is the normal outcome, and a run of eight of the same colour turns up in about a quarter of 100-spin sessions. Nothing unusual has to be going on for this to happen.
Slots behave the same way. Example, assumed values: if a slot pays something on 25% of spins (a hit frequency of 25%), the chance of 10 losing spins in a row from a given point is 0.75¹⁰ ≈ 5.6%. Over a long session, dry spells of that length are close to certain.
Two opposite thinking errors
The gambler’s fallacy
The gambler’s fallacy is the belief that after a run of one result, the other result is “due”. After five reds, black feels more likely. It is not: on the next spin, red still has a probability of 18/37 ≈ 48.6%, black the same, and zero 1/37. The wheel does not balance the books. The law of large numbers describes long-run proportions; it works by dilution over thousands of spins, not by correction in the next few.
A useful check: on a fair wheel, the sequence red-red-red-red-red is exactly as likely as red-black-black-red-black. Both have probability (18/37)⁵. The first one just looks more “special”.
The hot-hand belief
The opposite error is to believe that a streak will continue because the game, the table or the player is “hot”. The term comes from sport. In a well-known 1985 paper, Gilovich, Vallone and Tversky analysed basketball shooting and concluded that the hot hand people perceived was largely an illusion. In 2018 Miller and Sanjurjo showed that the original method contained a subtle statistical bias, and their re-analysis suggested that a modest hot-hand effect in human shooting may exist after all. The debate continues for skills performed by people.
For a roulette wheel or an RNG, though, there is no debate to have. A machine does not get confident, tired or warmed up. Whatever the truth about basketball players, a casino game designed with independent outcomes cannot be “hot”.
Why streaks change how people bet
The real danger of streaks is not the run itself but what it does to bet sizing:
- After wins, money that was just won can feel like “the casino’s money”, and stakes creep up. A larger stake does not change the odds; it only increases the expected loss per bet and the size of the swings.
- After losses, the urge to “win it back” leads to chasing: bigger bets, longer sessions, or skipping the limit that was set earlier.
Progressive systems such as the Martingale (double after each loss) are built around the idea that a losing streak must end soon. Example, assumed values: a 10-unit base bet on red at European roulette.
| Losses in a row | Next stake | Total lost so far |
|---|---|---|
| 1 | 20 | 10 |
| 3 | 80 | 70 |
| 5 | 320 | 310 |
| 7 | 1,280 | 1,270 |
Losing seven in a row on a red bet (where zero also loses) has a probability of (19/37)⁷ ≈ 0.94% from a given point, but over 200 spins the chance of at least one such run is about 60%. At that point the system requires a stake of 1,280 units to win back 10, and table limits or the bankroll may not allow it. Every bet in the sequence still has the same negative expectation: on European roulette an even-money bet loses on average 1/37 ≈ 2.70% of the stake, about 0.27 units per 10 staked. You can read more about why systems cannot change this in our article on what actually affects outcomes.
Practical limits that work with the maths
- Decide before you start: a loss limit, a time limit and a stake size. Write them down if that helps.
- Keep stakes flat. Do not raise them because of a streak in either direction; the next result does not know about the last one.
- Expect runs. Long streaks of losses are a normal part of variance, not a sign that a win is coming.
- Take breaks, especially after an unusual run, when emotions are strongest.
- Treat the cost as entertainment. Casino games have a negative expected value; the house edge or RTP sets the long-run price.
More practical habits are collected on the tips page, and terms such as RTP, variance and house edge are explained in the glossary. If you are curious why lucky rituals feel so convincing, see the article on gambling rituals and the illusion of control.
Frequently asked questions
What is the chance of five reds in a row on European roulette?
From a given spin it is (18/37)^5, about 2.7%, assuming a fair single-zero wheel. Over a 100-spin session, however, the chance of seeing at least one run of five or more reds is about 77%.
After a long run of red, is black more likely?
No. Each spin is independent, so red and black each keep a probability of 18/37 on European roulette. Believing otherwise is the gambler’s fallacy.
Is the hot hand real?
For human skills such as basketball shooting, researchers still debate whether a modest effect exists. For roulette wheels and random number generators it does not apply, because the outcomes are designed to be independent.
Should I change my stake after a streak?
Changing stakes does not change the odds of the next result. Raising them after wins or losses only increases the money at risk, so a flat stake and a pre-set loss limit are the safer approach.