A lottery hasn’t hit the number 7 in 100 draws. Someone bets heavily on 7, thinking it’s ‘overdue.’ The probability is still 1 in (whatever), but the bias makes it feel like 7 ‘should’ come soon.
The Original Discovery
Observed in Monte Carlo, 1913, when roulette landed on black 26 times in a row. Gamblers bet heavily on red, believing red was ‘due.’ Red didn’t come (lost millions). Formalized in probability theory as a cognitive bias.
How It Works in Real Life
The Gambler’s Fallacy isn’t a rare phenomenon—it’s everywhere once you start looking:
- A stock is down 10 days in a row. An investor assumes ‘it’s due for a rally.’ But stock price movements are largely independent. The down streak doesn’t make an up day more likely.
- A job applicant has been rejected 5 times. They feel the next application is ‘due’ to succeed. But each application is independent. Past rejections don’t improve odds.
- A person flips a coin and gets heads 4 times. They bet on tails next. The probability is still 50%, but the streak makes tails ‘feel’ more likely.
Why This Matters to You
Gambler’s Fallacy is why you should not ‘average down’ on bad decisions just because they’ve been bad. If an investment thesis was wrong 5 times, that’s data that your thesis is wrong—not evidence that you’re ‘due’ for a win. In hiring, a rejection doesn’t increase odds of the next hire. In relationships, a string of bad dates doesn’t make the next one more likely to be good. Break the pattern by addressing root causes, not by assuming probability will equalize.
See It in Action
Play Mind Traps to see if you can recognize the Gambler’s Fallacy in the wild. The quiz forces context-based recognition—the hardest and most useful form of learning.
Related Reading
The Opposite Error: Hot Hand Fallacy
Gambler’s Fallacy says a streak of losses makes a win more likely. Hot Hand Fallacy says a streak of wins makes another win more likely. Both are errors — but they point in opposite directions.
In truly random events (roulette, coin flips, lottery), neither pattern holds. Each outcome is independent. But in skill-based domains, the Hot Hand may actually be real: a basketball player who has made five shots in a row may genuinely be in a better state — warmer muscles, higher confidence, better feel for the court. The streak carries information about current capability.
The key question is: does this domain have memory? Random events don’t — past outcomes carry no information about future outcomes. Skill-based events do — past performance reflects real capability. Confusing the two leads to both fallacies.
Where It Costs Real Money
Averaging down on losing investments. If a stock you own has dropped 40%, Gambler’s Fallacy whispers that it’s “due for a recovery.” But the price action carries no obligation. The question isn’t whether it has fallen — it’s whether the underlying business is sound. Many investors hold losing positions longer than they should, not because of analysis, but because a loss streak feels like it creates entitlement to a win.
Doubling down at casinos. Casinos are built on Gambler’s Fallacy. Players on losing streaks increase bets, believing they’re “due.” The casino’s edge doesn’t care about recent history. Each spin is independent. The house wins in aggregate not because it cheats, but because it has infinite time and players have finite money.
Job applications and rejection. After several rejections, it feels like the next application “should” succeed. But rejections don’t create odds. If the rejections share a root cause — a weak resume, a mismatch in experience, a market that’s closed — the streak is data, not bad luck. Fix the root cause rather than betting that probability will balance out.
Frequently Asked Questions
If a coin lands heads 10 times, isn’t tails slightly more likely? No — assuming a fair coin. Each flip is independent. The probability of tails on flip 11 is exactly 50%. What the streak should update is your belief about the coin: 10 heads in a row is unusual enough that you should consider whether the coin is fair. But if the coin is fair, the next flip is 50/50.
How is Gambler’s Fallacy different from regression to the mean? Regression to the mean is real: extreme values in any dataset tend to be followed by values closer to the average. But this happens because extreme outcomes usually involve some luck, and luck doesn’t persist — not because the universe is “correcting” anything. The mechanism matters. Regression is statistical; Gambler’s Fallacy imagines a balancing force that doesn’t exist.
Does understanding this bias make me a better investor? It should help you avoid one specific error: holding bad positions because they’re “due” for a recovery, or selling good ones because they’ve “run too far.” But most investing mistakes come from other sources. Bias awareness is necessary but not sufficient for good investment decisions.
From My Own Life
After several losing months in the stock market, I found myself thinking ‘I’m due for a win.’ As if the market kept a score of what it owed me. It doesn’t. Each investment decision stands alone. Past losses don’t create future entitlement. Understanding this fallacy didn’t make me a better stock picker, but it stopped me from increasing position sizes after losing streaks, which probably saved me significant money.
See all 40 psychology laws: Complete Index →
Leave a Reply