A similar calculation produces a closeness measure of 5,402,013 for the high-low line. Thus, the scatterplot line fits the points better than the high-low line. This outcome supports the earlier claim that the use of judgment in the scatterplot method is superior to the high-low method. In principle, comparing closeness measures can produce a ranking of all lines from best to worst. The line that fits the points better than any other line is called the best- fitting line. It is the line with the smallest (least) sum of squared deviations. The method of least squares identifies the best fitting line. We rely on statistical theory to obtain the formulas that produce the best-fitting line.