Toss a fair coin a few times and the results are erratic; toss it many times and the fraction of heads settles near one half. Averages of many trials converge toward the underlying tendency. Order emerges from repetition.
The claim
The law of large numbers states that the average of many independent trials converges to the expected value as the number of trials grows. Short-run fluctuation gives way to long-run regularity. The many are more predictable than the few.
Beneath the surface
The mechanism is cancellation. Random deviations in individual trials tend to offset one another when averaged, so the noise shrinks relative to the signal. Aggregation tames chance.
A reframing
The turn is that regularity is a property of the many, not the one. No single trial becomes more predictable, yet their average does, which is a subtle and often misunderstood point. Predictability lives in the aggregate.
The trade-off
The law is easily misapplied. It does not say that past outcomes must be balanced by future ones, and the gambler who expects a due result mistakes its meaning. Independence forbids such compensation.
Where it breaks
The implication underlies statistics and insurance. Because averages of many are stable, one can predict aggregates that individual cases defy. The unpredictable becomes manageable in bulk.
The larger point
The law of large numbers says averages of many independent trials converge to the expected value as noise cancels. Regularity belongs to the aggregate, not the individual, and implies no compensation between trials. It makes the unpredictable manageable in bulk. Seen this way, the concept is less a fact to be filed than a lens through which other facts arrange themselves. Its worth lies not in any single application but in the pattern of thought it makes available. To hold it clearly is to see a whole family of problems as variations on one theme, and to recognize the same shape recurring where it was not expected.
