law of large numbers

Statistical concept that larger the sample population (or the number of observations) used in a test, the more accurate the predictions of the behavior of that sample, and smaller the expected deviation in comparisons of outcomes. As a general principle it means that, in the long run, the average (mean) of a long series of observations may be taken as the best estimate of the 'true value' of a variable. In other words, what is unpredictable and chancy in case of an individual is predictable and uniform in the case of a large group. This law forms the basis for the expectation of probable-loss upon which insurance premium rates are computed. Also called law of averages. See also law of small numbers and law of statistical regularity.