The Gambler's Fallacy Explained
Black has come up seven times in a row. Surely red is "due"? It's one of the most natural thoughts in the world — and it's wrong, in a way that has cost players for as long as wheels have spun. The gambler's fallacy is the belief that past random results change the odds of the next one. They don't. Here's why.
Independent events have no memory
Two events are independent when the outcome of one tells you nothing about the other. A roulette wheel, a fair coin, a crash round, a slot spin — each result is independent of every result before it. The wheel does not know it just landed on black seven times. It has no mechanism to "remember," no ledger it feels pressure to balance.
On a European wheel, the chance of black on any single spin is 18 in 37, about 48.6%. After seven blacks, the chance of black on the eighth spin is still 18 in 37 — about 48.6%. The streak is already in the past. The next spin starts from scratch, every time.
The trick your brain plays
The confusion comes from blending two different questions. It's true that a run of eight blacks in advance is unlikely: roughly 0.486 multiplied by itself eight times, under 0.4%. But once seven blacks have already landed, those seven are locked in — probability 1, they happened. The only open question is the eighth spin, and it's just one fresh spin: 48.6%.
| Question | Probability |
|---|---|
| 8 blacks in a row, predicted up front | ~0.4% |
| Black on spin 8, given 7 blacks already landed | ~48.6% |
| Red on spin 8, given 7 blacks already landed | ~48.6% |
The rare-streak intuition is real, but it only applies before the spins happen. Plucking the last spin out of a finished streak and treating it as if it's still part of the unlikely run is the whole error.
Key idea: a fair coin has no memory. After ten heads, the next flip is still 50/50. Nothing is ever "due." The odds reset completely on every independent trial.
The hot-hand fallacy: the same error, flipped
There's a mirror-image mistake. The gambler's fallacy says a streak must break ("red is due"). The hot-hand fallacy says a streak must continue ("black is on a roll, ride it"). Both assume past independent results predict the next one. For a roulette wheel or a crash curve, both are equally baseless — the streak is just a pattern your brain found in noise after the fact.
(In genuine skill activities, a real hot hand can exist, because confidence and form aren't independent. But a slot machine has no form, and a wheel has no confidence.)
Large numbers, not the "law of averages"
People defend the fallacy with the "law of averages" — the sense that things must even out. There is a real theorem nearby, the law of large numbers, but it says something subtler. As you flip a fair coin more and more, the proportion of heads converges toward 50%. It never promises the running count of heads and tails will balance — in fact the raw gap between them typically grows.
Averaging happens not through a corrective force that "owes" you red, but through dilution: a streak of seven blacks becomes a smaller and smaller fraction of the total as thousands more spins pile up around it. The universe doesn't fix the imbalance; it simply drowns it out. No spin is ever paying back a debt from a previous one.
On Riskr you can chase "due" numbers and ride "hot" streaks all night and watch independence shrug it all off — with fake chips and a leaderboard as the only thing on the line. It's a risk-free place to feel the fallacy fail in real time.