Ellenberg explains a common misunderstanding of the Law of Large Numbers. The principle does not mean that future outcomes actively compensate for unusual results in the past. Instead, as more observations are added, earlier deviations make up a smaller share of the total, so the overall average can move closer to the expected value.

For example, an unusually high proportion of one outcome in a small sample does not make the opposite outcome more likely next time. What changes is the weight of the earlier data as the sample grows. New observations gradually reduce the influence of an initial imbalance. The quotation therefore distinguishes statistical convergence from the idea that chance somehow corrects previous results. Its central point is that long-run regularity emerges through accumulation of data, not through a balancing mechanism acting on individual future events.