To estimate the areas of the fields to be discovered in a mature play, we use a model that conceptually divides the play (or a portion of it) into hexagons of "unit area," loosely defined as the minimum area that an oil field should have to be commercial. The individual hexagons may contain oil, and they are supposed to be spatially independent. The distribution of field sizes is defined by just one parameter (the quotient cells with oil in the area of study/total number of cells in the area of study). Samples taken from this discrete distribution look as if they had been drawn from continuous log-normal distributions. The paper considers the hypothesis that all samples drawn from a distribution of field sizes are biased, overestimating the true areas of the fields existing in the play.