← The collection

Gambler’s Fallacy.

A streak can feel like it makes the opposite result due, even when trials are independent.

Interactive experimentintuitiveField note ·
Preparing the experiment…
THE SHORT VERSION

Gambler’s Fallacy, explained.

The gambler's fallacy is the belief that an independent random process must soon compensate for a recent streak, making the opposite outcome more likely next.

01 / THE MECHANISM

Why it happens

A long-run balance does not create a short-run debt. For an independent fair coin, the mechanism generating the next toss has no memory of earlier tosses. More observations can dilute a past imbalance without requiring a corrective streak.

Independent random events do not remember the sequence before them.

Read the result

After a streak, distinguish the probability of the next toss from the probability of a whole specified sequence. Earlier outcomes are already known when you predict the next one.

02 / FOLLOW IT THROUGH

A worked example

Five heads in a row

  1. A fair coin has produced five heads. You now choose whether the next toss will be heads or tails.

  2. The chance of tails on that next independent toss is still 50%.

  3. Six specified heads were unlikely before the sequence began, but that does not make tails more likely after five have happened.

OPTIONAL DEEPER DETAILGo deeper: inside the model

Inside this model

The fair-coin chance of heads on the next flip remains 50% after any specified streak. The plot contrasts that with the declining chance of having seen the whole streak beforehand, (1/2)^n.

03 / BEYOND THE EXPERIMENT

Where this idea is useful

A practical use

Avoid treating a losing run in a fair game as evidence that a win is owed.

CHECK YOUR INTUITION

A common misconception

THE TEMPTING CONCLUSION

“A streak proves the opposite result is due.”

THE MORE USEFUL DISTINCTION

Independence means previous outcomes do not change the next probability. A suspected biased coin is a different question about the mechanism.

What this explanation leaves out

  • If the device is biased or conditions change, past results can be informative; independence is an assumption.
ONE MORE QUESTION

Can a streak ever be useful evidence?

Yes, if the probability is unknown or outcomes are dependent. It may inform a hypothesis about bias or changing conditions; it still does not create a compensating obligation.

TAKE THE IDEA WITH YOU

Does the process remember its past, or are you supplying the memory yourself?

Associated thinkers

Further reading

Explore the original research or the teaching reference behind this experiment.