Small samples mislead us about the world. Human stories born from the fluctuations of probability. The law of small numbers refers to the human tendency to believe that results from few observations or small samples represent overall trends or universal laws. The smaller the sample, the greater the variation in results, making extreme, random biases more likely to appear. This is a natural statistical phenomenon, yet counterintuitive.

Small samples mislead us about the world. Human stories born from the fluctuations of probability. The law of small numbers refers to the human tendency to believe that results from few observations or small samples represent overall trends or universal laws. The smaller the sample, the greater the variation in results, making extreme, random biases more likely to appear. This is a natural statistical phenomenon, yet counterintuitive.

For example, when rolling a die, theoretically each face should approach one-sixth probability. Yet in the first ten or twenty rolls, it's not uncommon for a specific face to appear repeatedly or not at all. Despite this, people often interpret such skew not as random fluctuation, but as the die having a bias, bad luck today, or some underlying cause. Here, the psychological urge to assign meaning to few results comes into play. At the heart of this error lies the representativeness heuristic. People tend to feel that even a small number of cases should adequately represent the whole. In reality, however, the smaller the sample size, the more unstable the average or proportion becomes; it only stabilizes as the sample size increases. In statistics, it is known that the magnitude of variation decreases as the sample size increases, and the volatility of small samples is an unavoidable property. This concept was sharply pointed out by behavioral economist Daniel Kahneman. Toget
her with Tversky, he experimentally demonstrated that humans struggle to correctly understand random phenomena and tend to draw excessive conclusions from small samples. This argument was introduced to a general audience in Chapter 10 of his book Thinking, Fast and Slow, leading to the widespread recognition of the term "law of small numbers." Why are people drawn to the law of small numbers? Behind this lies anxiety about uncertainty. People find it more reassuring to assign causal relationships or narratives to chance events than to leave them as mere coincidences. Extreme outcomes tend to be more memorable and create a sense of having an explanation, thereby reinforcing false certainties. This tendency influences research and societal judgments. Flashy results from small-scale surveys or experiments are sometimes overvalued. However, effects observed in small data sets often revert to the mean in subsequent observations. Discussing causation without understanding this can
spread errors into policy decisions, news reporting, and investment judgments. Addressing the law of small numbers is simple, but it requires an attitude that defies intuition. Verify the sample size, look at the denominator, be aware of error and variance. And, especially with small samples, avoid rushing to conclusions; take a step back and question. This attitude is the practical wisdom needed to avoid being swallowed by the stories created by chance.