Glossary
Gambler's fallacy
The mistaken belief that a random event is less likely after a streak of the same outcome, as if chance must even out in the short run.
The gambler's fallacy is the belief that past outcomes of independent random events change the chances of the next one. After five heads in a row, many people feel tails is "due". A fair coin has no memory: the chance of tails is still one half.
The best-known example took place at the Monte Carlo casino in 1913, when a roulette wheel came up black many times in a row. Gamblers bet heavily on red as the streak lengthened and lost large sums.
Its opposite
The hot-hand fallacy assumes a streak will continue: a shooter who has made several baskets "can't miss". In 1971 Amos Tversky and Daniel Kahneman traced the gambler's fallacy to a belief that small samples should look like the long-run average. Later work, starting with a 1985 study of basketball shooting that Tversky co-wrote, linked the hot-hand belief to the same root.
In markets
- "Due for a correction." A market that has risen several years in a row is not more likely to fall next year for that reason alone.
- "Due for a bounce." A stock that has fallen for ten days has no built-in tendency to rise on the eleventh.
- After a losing streak, a trader may increase position size, expecting a win to arrive.
Where it gets subtle
Market prices are not coin flips. Mean reversion and momentum are real tendencies in some assets over some periods, measured on large samples. The fallacy lies in expecting a reversal or continuation because of a streak alone, with no evidence that the process has such a tendency.
Limits
Deciding whether events are truly independent is itself hard. Some streaks do reflect a change in the underlying process, such as a strategy that has stopped working.
