Money Myths
The wheel has no memory, and a famous night in Monte Carlo proved it expensively
The fallacy is named after a run of one colour at a casino in 1913, when the crowd bet ever larger sums on the correction that a spinning wheel had no way to deliver.

There is a short answer about the gambler's fallacy and a useful one, and they are not the same. What follows is the useful one.
The short version
- Independent events carry no record of previous outcomes.
- A long run of one result is unremarkable across enough trials.
- The same reasoning error appears in investing and in lottery number choice.
The reasoning that feels irresistible
After a run of one outcome, the other starts to feel overdue, as though the universe were keeping accounts and would settle up. The intuition draws on genuine experience, since most real processes do have memory and do correct themselves.
A machine that has not failed for a long time really is more likely to fail, because parts wear out. Applying that reasoning to a device specifically engineered to have no memory is where it breaks down. The error is not stupidity but the misapplication of a rule that works nearly everywhere else.
The night the fallacy got its name
The standard account describes a roulette wheel at a Monte Carlo casino in 1913 producing an extraordinarily long run of one colour. As the run extended, players reportedly bet increasingly heavily on the opposite colour on the grounds that it must be due. The run continued well beyond the point at which most people had committed serious money to its ending.
The episode gave the reasoning error the alternative name by which it is still known in psychology and statistics. The details of the story come to us through repetition rather than through a contemporary audit, which is worth noting.
What long runs actually mean
The chance of any specific long run is small, but the chance of some long run appearing over many thousands of spins is not. People underestimate how streaky genuine randomness looks, which is why fabricated random sequences are usually too evenly mixed.
Checked against the record, statisticians can often identify an invented sequence precisely because it lacks the clusters that real randomness produces. A wheel producing a long run is therefore evidence of nothing except that a lot of spins have taken place. The only circumstance in which a run is informative is when it suggests the equipment is biased, which is the opposite conclusion.
The mirror-image error
The same misunderstanding produces the belief that a streak will continue because something is running hot. Both versions assume the outcomes are connected, and they simply disagree about the direction of the connection.
Where outcomes genuinely are connected, such as in games involving skill or exhaustible decks, tracking them is rational. Card counting works precisely because a dealt deck does have memory, which is why casinos object to it so strongly.
Knowing which situation you are in is the entire skill, and the two look identical from the outside.
Where it costs real money
Investors sell holdings that have risen because they are due a fall, and hold falling ones because they must be near the bottom. Lottery players avoid recently drawn numbers, which changes nothing about the odds and slightly increases the chance of sharing a prize with fewer people.
Choosing unpopular numbers is in fact the only rational adjustment available, and it affects payout rather than probability. Doubling a stake after a loss, in the belief that a win must come, is the fallacy formalised into a system. That system fails against table limits and finite bankrolls, which is precisely why table limits exist.
Why it persists after being explained
Understanding the principle intellectually does very little to weaken the feeling, which operates faster than the reasoning. The feeling is generated by a pattern-detection system that is extremely useful in almost every other context. Being told a wheel has no memory does not stop a run of eight looking meaningful, because it looks meaningful to a system that does not read.
Trace it back and the practical defence is a rule set in advance, since rules made before the run are not subject to the feeling. That is a general principle for decisions under uncertainty, and it applies well beyond a casino floor.
The takeaway
The wheel keeps no ledger, and everyone who thought otherwise was funding the correction they were waiting for.
The satisfying version of a story is the one that travels, which is the whole problem.
Questions readers ask
Is a long run evidence of anything?
Only of many trials, or of faulty equipment. It never indicates that the other outcome has become more likely on the next attempt.
Does the fallacy affect investing?
Yes. Assuming a rise or fall must reverse simply because it has continued is the same reasoning error applied to prices.
Also by Leela Mathews
- Renting is not throwing money away, and neither is most of a mortgage paymentMoney Myths
- Checking your own credit file does not damage itMoney Myths
- Carrying a credit card balance does not build anything except interestMoney Myths
- Einstein never called compound interest the eighth wonder of the worldMoney Myths





