The author points out that missing plot value formulas provide genuine estimates of the missing data. If the estimate is meaningless (e.g. a negative yield), examination of its confidence interval may determine whether the discrepancy is a random one or can be attributed to an inadequate model.