🎲Gambler's Fallacy Test

After six heads in a row, does tails feel like it has become more likely?

Take this free Gambler's Fallacy Test to challenge your intuition about independent random events.

You will see four coin-flip streaks. After each one, choose what is most likely on the next flip:

• ✓the streak continues;
• ✓the streak breaks;
• ✓or Heads and Tails remain equally likely.

For a known fair coin with independent flips, the next outcome is always 50% Heads and 50% Tails, regardless of the previous streak.

How to Take the Gambler's Fallacy Test

1. 1Press Start Test.
2. 2Study the coin-flip streak.
3. 3Decide what is most likely next.
4. 4Choose streak continues, streak breaks, or 50/50.
5. 5Repeat for all four sequences.
6. 6Review your answers.

Do not ask:

“What sequence would look more random?”

Ask:

“What is the probability of the next independent flip?”

Those are different questions.

What Is the Gambler's Fallacy?

The gambler's fallacy is the mistaken belief that after a streak in one direction, an independent random process becomes more likely to produce the opposite outcome.

For example:

Heads, Heads, Heads, Heads, Heads...

followed by the thought:

“Tails is due.”

If the coin is fair and each flip is independent, that reasoning is incorrect.

The next flip still has:

P(Heads) = 0.50

P(Tails) = 0.50

The coin does not compensate for earlier outcomes.

Why Five Heads in a Row Does Not Make Tails More Likely

Independence means the probability of the next outcome does not depend on previous outcomes.

For a fair coin:

P(Tails on next flip | five Heads already occurred) = 0.50

The streak may be unusual as a complete historical sequence.

But once those five heads have already happened, they do not change the physical probability of the next flip.

This distinction between:

probability of a whole sequence before it occurs

and

probability of the next event after part of the sequence is known

is essential.

“HHHHHT” and “HHHHHH” Are Equally Likely Beforehand

For six fair independent flips, every exact sequence has probability:

(1/2)^6 = 1/64

That includes:

HHHHHH

and

HHHHHT

and

HTTHHT

Each exact six-flip pattern is equally likely before the experiment begins.

People often feel that a mixed-looking sequence is “more random” than a streak.

But random processes regularly produce clusters.

Randomness does not promise local balance in every short run.

The Law of Small Numbers

Tversky and Kahneman's 1971 research described a belief in the law of small numbers.

People often expect small samples to resemble the larger population more closely than probability theory guarantees.

For a fair coin, the long-run proportion of heads tends toward 50%.

But people may incorrectly expect even short sequences to contain roughly equal numbers of Heads and Tails.

A long streak then feels like an imbalance that needs to be corrected quickly.

The gambler's fallacy is one consequence of that intuition.

The Law of Large Numbers Does Not Mean Immediate Correction

The genuine law of large numbers says that as the number of independent trials grows, the sample proportion tends to approach the true probability.

It does not say that a streak creates a debt the process must repay.

Suppose a fair coin produces:

10 Heads, 0 Tails

After 10 flips, Heads = 100%.

If the next 990 flips happen to contain approximately half Heads and half Tails, the overall proportion moves much closer to 50% even without any special excess of Tails.

Long-run balance emerges because the original streak becomes a smaller fraction of the growing sample—not because probability forces an immediate reversal.

Gambler's Fallacy vs. Hot-Hand Belief

These are opposite sequential beliefs.

Gambler's Fallacy

After a streak, expect a reversal.

Example:

five Heads → Tails is now more likely

Hot-Hand Belief

After a streak, expect continuation.

Example:

five successes → another success is now more likely

For a known fair independent coin, both predictions are wrong if they assign a probability other than 50/50.

But they are not the same bias.

This matters because the test offers:

• ✓streak continues;
• ✓streak breaks;
• ✓50/50.

Choosing streak breaks is specifically gambler's-fallacy reasoning.

Choosing streak continues reflects positive-recency or hot-hand-like reasoning.

The correct answer for the fair-coin task is 50/50.

Real Sports Are Not Fair Coins

The hot-hand example requires caution.

Basketball shots, tennis serves, sales performance, and financial returns may not be independent identically distributed coin flips.

Player skill, fatigue, opponent quality, confidence, strategy, injuries, and environment can change from trial to trial.

Therefore:

“A streak never contains information”

is false as a general statement.

The correct rule is narrower:

past outcomes do not change the probability of the next outcome when the process is genuinely independent with fixed probabilities.

Your browser test deliberately uses a fair independent coin model.

Drawing Cards Is Different

Suppose you draw cards from a standard deck without replacement.

Now previous outcomes do matter.

If four aces have already been removed, the probability of drawing another ace is zero.

The trials are dependent because the composition of the deck changes.

This is not gambler's-fallacy reasoning.

It is valid probability updating.

Before applying the 50/50 logic, always ask:

Are the events actually independent?

Roulette Is Not Exactly a 50/50 Coin

Roulette is often used to explain the gambler's fallacy because gamblers may expect red after a run of black.

But a roulette wheel includes green zero pockets.

On European roulette there is a single zero.

American roulette traditionally has both 0 and 00.

Therefore red versus black is not exactly 50/50.

The important point remains that previous independent spins do not make red “due.”

The exact probability simply depends on the wheel design.

For mathematical clarity, a fair coin is the cleaner example.

Why Random Sequences Look “Too Streaky”

People often expect randomness to alternate more than it really does.

A sequence such as:

HTHTTHTH

looks random to many people.

A sequence such as:

HHHHHTHH

may look suspicious.

But fair random processes naturally produce runs.

In fact, as a sequence gets longer, seeing at least one noticeable streak becomes increasingly unsurprising.

The misconception comes from expecting small samples to look perfectly balanced and irregular.

True randomness often contains local patterns.

What Does Your Score Mean?

The site presents four streak scenarios.

If you answer 50/50 on all four, you correctly applied the independence rule in every scenario.

If you choose “streak breaks,” you made a response consistent with the gambler's fallacy.

If you choose “streak continues,” you made the opposite sequential prediction.

Four questions are not enough to diagnose your general probability reasoning.

The result is a compact demonstration.

Can Someone Know the Math and Still Feel Tails Is Due?

Yes.

Intuitive judgment and explicit knowledge can conflict.

You may know perfectly well that the correct answer is 50/50 while still feeling that six Heads in a row makes Tails psychologically compelling.

That gap is one reason cognitive-bias demonstrations are memorable.

The goal is not to shame intuition.

It is to recognize when intuition conflicts with the mathematical structure of the problem.

Gambler's Fallacy vs. Conjunction Fallacy

The Conjunction Fallacy Test concerns the rule that A-and-B cannot be more probable than A alone.

The gambler's fallacy concerns mistaken negative dependence between independent sequential outcomes.

Both are probability errors.

But the mathematical structures are completely different.

How to Avoid the Gambler's Fallacy

Before predicting the next event, ask:

1. 1Is the process genuinely independent?
2. 2Is the probability fixed from trial to trial?
3. 3Has the underlying mechanism changed?
4. 4Does the previous result physically alter the next trial?
5. 5Am I expecting short-run balance merely because the sequence “looks wrong”?

If the process is a known fair independent coin, the solution is immediate:

ignore the streak when predicting the next flip.

Frequently Asked Questions

A Streak Does Not Create a Debt

Randomness does not promise:

“I gave you too many Heads, so now I owe you Tails.”

For a fair independent coin, every new flip starts with the same probabilities.

That is the simplest defense against the gambler's fallacy:

separate the history of the sequence from the probability of the next independent event.