@Bix

As long as ther's a way for sample counters of any kind to cook the stats, even unconsciously, by imposing their idea of where the central point on the curve *ought* to be and what results should be considered outliers, the results will be influenced by preconceived notions.