The gambler's fallacy is the mistaken belief that a random outcome becomes more or less likely because of what happened before. It often appears in coin flips, dice rolls, roulette, sports streaks, and lottery choices. The idea matters because it can lead people to make risky decisions based on patterns that are not real.
Good statistical thinking separates short-term streaks from true changes in probability.
For independent events, the probability of the next outcome does not change after a streak. If a fair coin lands heads 6 times in a row, the probability of heads on the next flip is still 1/2. The law of large numbers says long-run averages tend to get closer to expected probabilities, but it does not force future outcomes to cancel past streaks.
The hot-hand fallacy is a related error where people assume a streak means success is now more likely, even when there is no evidence the probability has changed.
Understanding Statistics: The Gambler's Fallacy
A useful way to see the error is to list whole sequences rather than focus on the ending. In seven coin flips, the sequence heads heads heads heads heads heads heads is one possible sequence. So is heads heads heads heads heads heads tails.
Before the seventh flip, neither sequence is favoured over the other. Each complete seven-flip sequence has the same chance, one out of one hundred twenty-eight. People often notice that a run looks unusual, then treat the next result as if it must repair the run.
Randomness has no target total for a small set of trials. It produces uneven patches naturally.
The law of large numbers can seem confusing here. Over many flips, the fraction of heads tends to move nearer one half. That movement can happen in several ways.
A long run of tails may be followed by a mixed set of results with roughly equal heads and tails. It does not require an extra run of heads. For example, after ten heads, another ten flips could contain five heads and five tails.
The total would then be fifteen heads out of twenty flips, which is still above one half. The average became closer without any special correction on individual flips.
The key skill is deciding whether repeated events are truly independent. A physical system can carry information from one event to the next. Cards dealt from a deck without replacement are a clear example.
If many hearts have already been dealt, fewer hearts remain, so the chance changes. A biased coin may land differently depending on how it is flipped. In weather, today’s conditions can help predict tomorrow’s weather.
In sports, a player’s injury, fatigue, opponent, or changing strategy may affect later performance. Past results alone do not prove dependence. Evidence about the process is needed.
Casinos benefit when players mistake a streak for information. A roulette player may keep increasing a bet because a number or colour seems due. The size of previous losses does not make a future spin more likely to pay back those losses.
Lottery players can make a similar mistake by avoiding numbers that appeared recently. In data work, students should separate a striking pattern from a tested pattern.
Record enough observations, compare them with a sensible baseline, and consider possible causes. A streak is evidence worth checking, not proof that the underlying chance has changed.
Key Facts
- Independent events do not affect each other: P(next outcome | past outcomes) = P(next outcome).
- For a fair coin, P(heads) = 1/2 and P(tails) = 1/2 on every flip.
- For a fair six-sided die, P(rolling a 6) = 1/6 on every roll.
- The probability of 6 heads in a row is (1/2)^6 = 1/64.
- After 6 heads in a row, the probability of another head is still 1/2, not less than 1/2.
- The law of large numbers describes long-run relative frequency, not a short-run correction: observed proportion approaches expected probability as n becomes large.
Vocabulary
- Gambler's fallacy
- The mistaken belief that a random event is due to reverse after a streak, even when each trial is independent.
- Independent events
- Events are independent when the outcome of one event does not change the probability of another event.
- Conditional probability
- Conditional probability is the probability of an event given that another event has already occurred.
- Law of large numbers
- The law of large numbers states that the average result of many independent trials tends to approach the expected value.
- Hot-hand fallacy
- The hot-hand fallacy is the mistaken belief that a person or process is more likely to keep succeeding just because of a recent streak.
Common Mistakes to Avoid
- Thinking tails is due after several heads is wrong because a fair coin has no memory and each flip still has P(tails) = 1/2.
- Confusing a rare streak with an impossible streak is wrong because unlikely sequences can happen, especially when many trials are observed.
- Using the law of large numbers to predict the next trial is wrong because it describes long-run averages, not a forced correction on the next outcome.
- Assuming every streak means a changed probability is wrong because a streak can occur by chance unless there is evidence that the underlying process changed.
Practice Questions
- 1 A fair coin lands heads 5 times in a row. What is the probability that the next flip is tails?
- 2 A roulette wheel has 18 red spaces, 18 black spaces, and 2 green spaces. If black occurs 4 times in a row, what is the probability that the next spin is red?
- 3 A basketball player makes 6 shots in a row. Explain what information you would need before deciding whether this is evidence of a hot hand rather than random variation.