What it means
The fallacy takes its nickname from a night at a Monte Carlo casino where a roulette wheel reportedly landed on black many times in succession, and players lost heavily backing red on the reasoning that it had to come up soon. The wheel had no memory, and neither does any other genuinely independent process.
Underneath the error is a confusion between two different probabilities. The chance of a long run happening from the start is genuinely small, but the chance of the next event given that the run has already happened is unchanged, because the earlier results are already settled facts.
In business the fallacy rarely announces itself so plainly. It appears as "we have lost four bids in a row, so we are due a win", or "this stock has fallen five days running, it must bounce", both of which assume a self-correcting force that does not exist in independent events.
The complication is that not every business process is independent. Sales pipelines, credit defaults and machine failures often have real serial dependence, so a run of losses can genuinely signal a changed win rate, and the correct question is whether the events are independent rather than whether a run has occurred.
There is a mirror-image error worth knowing. The hot hand fallacy assumes a run will continue rather than reverse, and both mistakes come from the same source: reading patterns into small samples of a random process.
In practice
Real-world examples.
Example
A trader watches a currency pair fall for eight consecutive sessions and doubles a long position on the reasoning that a bounce is overdue. The move has no statistical basis, and the position is sized on a feeling about fairness rather than on any signal.
Example
A retail investor avoids a fund that has beaten its benchmark for three years, believing a poor year must be coming. The relevant questions are fees, strategy and whether the outperformance came from skill or from one concentrated bet, not from any notion of turn-taking.
Example
A credit committee approves a marginal loan because the portfolio has had an unusually low default rate for two years and "the numbers have to even out". Defaults are driven by borrower quality and economic conditions, not by a quota that must be filled.
Formula
Calculation
For independent events, the probability of the next outcome does not depend on what came before:
P(next outcome | previous outcomes) = P(next outcome)
Consider a fair coin that has just landed tails six times in a row.
Probability of the seventh flip being tails = 0.5, or 50%. The six previous flips change nothing.
Probability of seven tails in a row, assessed before any flip = 0.5 to the power of 7 = 1/128 = 0.78125%.
Both statements are true at the same time, and confusing them is the whole fallacy. Before you start, a run of seven is unlikely; once six have happened, the seventh is a coin flip like any other.
The same arithmetic applies in business. A sales team with a genuine 30% win rate on independent competitive bids has a 70% chance of losing any single bid, so the probability of losing six in a row is 0.7 to the power of 6 = 11.76%. That is uncommon but far from remarkable, and it certainly does not make the seventh bid more winnable: the chance of winning it remains 30%.Case study
Seen in the real world.
Halcyon Bridge Capital is a fictional trading firm invented to illustrate the fallacy and its consequences. One of its illustrative traders ran a strategy that doubled the position size after every losing day on a stock index, on the theory that a string of down days made an up day increasingly likely.
For fourteen months the approach appeared to work, because most losing streaks were short and each recovery repaid the accumulated position. Then the index fell for nine consecutive sessions, the doubling rule pushed the position to over six times its normal size at the worst possible moment, and a single week wiped out two years of gains.
The post-mortem found the flaw was never the trader's market view but the arithmetic behind the sizing rule. Daily index moves are close enough to independent that no number of down days makes an up day more likely, and the illustrative firm rewrote its risk policy so that position size could never be increased on the basis of prior losses alone.
Watch out
Common mistakes.
- Believing an outcome is "due" after a run of the opposite result. Independent events have no memory, so nothing accumulates that must be paid back later.
- Applying the fallacy label to processes that genuinely are dependent. If a run of losses reflects a real deterioration in win rate or credit quality, treating it as random noise is the opposite error and just as expensive.
- Increasing position size after losses in the belief that a reversal must be close. This turns a harmless misunderstanding of probability into an active risk that grows exactly when losses are mounting.
Questions
People also ask.
Is regression to the mean the same as the gambler's fallacy?
No, regression to the mean is a real statistical effect about extreme observations being followed by less extreme ones, whereas the fallacy claims a self-correcting force in independent events, which does not exist.
How do I tell whether events in my business are independent?
Test whether the outcome rate after a run differs from the overall rate, and look for a plausible mechanism linking one event to the next, such as a shared customer, market condition or process.
Does the fallacy apply to long-term investing?
Not in the same way, because asset returns show some mean reversion over long horizons, but that is a valuation effect rather than a rule that a bad year must be followed by a good one.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%