Gambler's Fallacy

28 Gambler’s Fallacy

On August 18, 1913, at the casino in Monte Carlo, black came up on the roulette wheel a record twenty-six times in succession. There was a near-panicky rush to bet on red, beginning about the time black had come up a phenomenal fifteen times. Players doubled and tripled their stakes, believing that after black came up the twentieth time there was not a chance in a million of another repeat. In the end, the unusual run enriched the casino by some millions of francs.

The gambler’s fallacy is committed when someone reasons that in a situation of pure random chance, the outcome of a future event can be influenced by the outcomes of previous, independent events. It confuses independent, non-causal occurrences with genuinely causal ones: a roulette spin does not “remember” what came before, and the wheel has no obligation to restore some cosmic balance. Despite this, the fallacy is so compelling that it regularly appears in two opposite forms. After a streak, people either believe the opposite result is “due” (the classic gambler’s fallacy) or that the streak itself will continue—the hot hand fallacy. Both predictions are equally unfounded when each event stands alone.

The error mimics valid probabilistic reasoning. It is true that over a very large number of random trials, outcomes tend to even out in percentage terms. But that happens simply because a long series of fair coin flips yields roughly equal counts, not because the coin compensates for past deviations. Consider this case: I plan to flip a fair coin 20 times and you must predict the total number of heads. Before any flips, your best guess is 10. Then I flip and get six heads in a row. If I let you change your bet with 14 flips remaining, you should now predict about 13 heads (the six already seen plus half of the remaining 14, i.e., seven). If I were flipping 100 times, after those first six heads you’d expect 6 + 47 = 53 heads. For a billion flips, it’s 6 + 499,999,997 = 500,000,003—practically 50 %. As the number of trials grows, the percentage of heads drifts toward 50 %, but the raw number of extra tails needed to “cancel” the early streak never materializes. Luck evens out only in the sense that a few early deviations get diluted by an enormous number of later 50–50 splits. It does not even out by producing an opposite run.

Thus, the belief that a streak of heads makes tails more likely on the next flip is unfounded. In roulette, the fact that red 21 just came up has no bearing on whether it will come up on the next spin. The same flaw underlies predictions that a streak will continue: if a coin lands heads ten times in a row, the eleventh flip remains exactly 50 %. The hot hand feeling is just the mirror image of the classic gambler’s fallacy, and it is equally fallacious when events are genuinely random and independent.

Everyday life provides innumerable illustrations. Lottery players often think a ticket with the numbers 1, 2, 3, 4, 5, 6 is less likely to win than any other combination, even though all specific sets have identical odds. A family who has had two sons may convince themselves the next child is certain to be a girl; a sports fan may declare that the Lakers have to win tonight because they have lost five in a row; or someone who just dated three disappointing partners in a row may insist the next one can’t possibly be a loser. These examples highlight the fallacy’s split personality—the same mistaken reasoning can fuel predictions of change or predictions of continuity.

The fallacy does not apply when the outcome of one trial genuinely changes the odds of the next. If you draw two aces from a deck without replacement, betting against a third ace is rational because the remaining deck now contains fewer aces. Similarly, when a long streak might be a clue that the underlying process is not actually random—for instance, a weighted coin, an unbalanced roulette wheel, or a skilled player genuinely performing better—past outcomes can legitimately indicate future results. The key is learning to distinguish true independence from situations where probability updates are warranted.

To avoid the gambler’s fallacy, remind yourself that each independent chance event stands alone. Past flips, spins, draws, or sports outcomes do not bend future probabilities—they are non-causal events. Familiarizing yourself with the basics of probability and statistics helps. And if you choose to gamble, do it for entertainment, not as a way to make money: never wager more than you can comfortably afford to lose, and remember that in casino games the odds are designed against you.