The Gambler's Fallacy: Why Streaks Do Not Predict the Next Result

Picture a roulette board that shows red, red, red, red, red. Many people feel strongly that black must be due. That feeling is so common that it has its own name: the gambler's fallacy. Understanding why it is a mistake is one of the most valuable lessons in basic probability, and it has nothing to do with being unintelligent.

What the fallacy says

The gambler's fallacy is the belief that after a streak of one outcome, the opposite outcome becomes more likely. It sounds sensible because we know that over many trials, results tend to balance out. The mistake is to think that the balance is achieved by the system correcting itself on the next turn.

Independent events

In most casino games, each round is independent. The wheel, the dice, and the shuffled deck do not remember what happened before. A fair coin has a probability of one half for heads each time, whether the last ten flips were heads or tails.

Here is a check with roulette. On an American wheel, red has 18 chances in 38, about 47.4 percent. The probability of five reds in a row is that number multiplied by itself five times, roughly 2.4 percent. That is a small chance, so seeing it feels remarkable. But once four reds have already happened, the probability of a fifth is simply 47.4 percent again. The earlier spins are in the past and cannot influence the next one. For more on how these odds are set up, see our guide to American and European roulette wheels.

Why streaks fool us

Several habits of thinking combine to make the fallacy persuasive:

We expect randomness to look mixed. True random sequences contain surprisingly long streaks, but people tend to expect frequent switching.
We remember dramatic runs. A long streak stands out, and a sequence that looked ordinary is quickly forgotten.
We confuse the long run with the short run. Results balance out over very large numbers, but they do so by dilution, not correction. A few extra reds in the first hundred spins become negligible after a hundred thousand.

The reverse mistake: the hot hand

The opposite error is believing a streak will continue because the person or table is hot. In games of pure chance, that is just as unfounded. The dice do not have favorites.

In skill-based contexts, such as sports, streaks can sometimes reflect real changes in form, but even there people tend to see patterns that are weaker than they appear. In casino games with fixed randomness, neither a cold nor a hot streak tells you anything about the next result.

How the fallacy leads to bigger losses

The danger is practical. A player who believes black is due may raise wagers to take advantage. If red keeps landing, losses grow faster than expected. Because every spin has a negative average result, increasing wagers after losses simply increases the amount of money exposed to that average. It does not improve the chance of winning.

This is closely related to the idea of chasing losses, and the habit of trying to win back what was lost can quickly turn a planned session into a harmful one.

A mental toolkit

To stay clear of the trap, try these steps:

Ask whether each event is independent. If the wheel or deck resets, past results are irrelevant.
Ignore recent-results boards when making decisions. They are decorations, not predictions.
Remember the long-run math. The odds stay the same, as we describe in our article on expected value.
Decide your budget before play, and do not change it based on a streak.

Gambling is for adults only. If you notice yourself reasoning that a result is due and raising your wagers, treat it as a signal to pause.

Frequently asked questions

Does a long streak ever signal a rigged game? In rare cases a game might be unfair, but a streak alone is not evidence. Long streaks are a normal feature of randomness.

Do results eventually even out? In proportion, yes, over a huge number of trials. But there is no mechanism that forces a correction on the next spin.

Is the gambler's fallacy the same in lotteries? Yes. Each draw is independent, so numbers that have not appeared recently are not more likely to appear.