Learn & Understand

The Law of Large Numbers and the Gambler's Fallacy

The companion calculator computes the probability of getting a certain number of heads in a series of coin flips using the binomial distribution. Behind those probabilities lies a pair of ideas that are constantly confused: the law of large numbers, which says a fair coin's results even out over many flips, and the gambler's fallacy, the mistaken belief that a coin is somehow "due" to balance out. Understanding the law of large numbers, why it does not imply a coin has memory, and how the gambler's fallacy misapplies it turns a coin-flip calculation into an appreciation of one of the most misunderstood truths in probability.

The Law of Large Numbers

The law of large numbers is a genuine and important result: as the number of flips grows, the observed proportion of heads tends to approach the true probability, one-half for a fair coin. Over a small number of flips, the proportion of heads can deviate substantially from half, but as you flip more and more times, the fraction of heads gets closer and closer to one-half, so in the long run the results even out to reflect the true probability. This is why a fair coin flipped an enormous number of times will show very close to half heads, and it is the reason casinos and insurers can rely on probabilities working out over many events. The law of large numbers is about the long-run proportion converging to the true probability as the number of trials increases. Understanding the law of large numbers correctly is essential, because it is true and powerful, but it is also the source of a common misconception when misapplied to the short run. It says the proportion evens out over many flips, not that any particular flip is influenced by the past, a distinction that is the crux of the gambler's fallacy.

The Coin Has No Memory

The crucial fact that the gambler's fallacy ignores is that each coin flip is independent: the coin has no memory of previous flips, so the outcome of the next flip is completely unaffected by what came before.

No matter how many heads have come up in a row, the probability of heads on the next flip is still one-half, because the coin does not remember or "know" its previous results, each flip is a fresh, independent event with the same odds. A run of heads does not make tails more likely on the next flip; the coin is not "due" for a tails to balance things out. This is the gambler's fallacy: the mistaken belief that past outcomes change the probability of future independent events, that a streak in one direction makes the opposite outcome more likely to restore balance. It feels intuitive, after many heads, tails seems overdue, but it is simply wrong, because independence means the past has no bearing on the next flip. Understanding that the coin has no memory is the antidote to the gambler's fallacy: each flip stands alone at the same probability, so no sequence of past results makes any future result more or less likely. The calculator's probabilities assume exactly this independence, computing each outcome as if the flips do not influence one another, which is the reality.

How the Two Are Reconciled

The apparent tension, the results even out in the long run (law of large numbers), yet the coin has no memory and is never due (independence), is resolved by understanding how the evening-out actually happens. The long-run proportion approaches one-half not because past imbalances are corrected by compensating future outcomes, but because they are overwhelmed and diluted by the sheer number of subsequent independent flips. A run of extra heads early on is not cancelled by extra tails later; instead, as more flips accumulate, that early imbalance becomes an ever-smaller fraction of the total, so the proportion drifts toward one-half even though the absolute difference between heads and tails need not shrink. In other words, the coin evens out in proportion through dilution over many independent flips, not through any tendency to compensate. This reconciles the two ideas: the law of large numbers describes the long-run proportion converging by dilution, while independence ensures no individual flip is influenced by the past, so both are true simultaneously without contradiction. Understanding how the two are reconciled dispels the confusion at the root of the gambler's fallacy: the evening-out is real but comes from dilution over many trials, not from the coin correcting past imbalances, so it never makes any single flip more likely to go one way. The proportion converges; the individual flips remain stubbornly independent.

Why This Matters Beyond Coins

The gambler's fallacy and the correct understanding of independence and the law of large numbers matter well beyond coin flips, because the same misconception appears throughout gambling, decision-making, and reasoning about chance. People wrongly believe a lottery number that has not come up is due, that a roulette wheel showing red repeatedly makes black more likely, or that after a run of bad luck, good luck must follow, all instances of the gambler's fallacy applied to independent events. These beliefs can drive poor decisions, betting more because an outcome seems overdue, when in reality the odds are unchanged. Recognizing that independent events have no memory protects against these errors, while understanding the law of large numbers correctly, as long-run proportional convergence by dilution, prevents the opposite mistake of expecting short runs to reflect the true probability. The calculator's coin-flip probabilities embody the correct model: each flip independent, outcomes over a series governed by the binomial distribution, with no "due" outcomes. Understanding why this matters beyond coins reveals the practical value of getting these ideas right: they guard against a pervasive fallacy that leads people to misjudge chance in gambling and life, and they clarify what randomness actually does, even out in proportion over the long run while remaining unpredictable in each independent instance. The calculator computes the probabilities; understanding the law of large numbers and the gambler's fallacy is what reveals why the coin is never due and why the results nonetheless even out over many flips.

Understanding Coin-Flip Probability

Use the calculator to compute coin-flip probabilities, and understand the ideas behind them: the law of large numbers says the proportion of heads approaches one-half over many flips, but each flip is independent so the coin has no memory and is never "due," and these reconcile because the long-run evening-out comes from dilution over many trials, not from compensation. The gambler's fallacy is the error of ignoring this independence. The calculation assumes independent flips; understanding the law of large numbers and the gambler's fallacy is what reveals why past flips never change the next.