Sum of Squares (SS) is a statistical measure of deviation from the mean (Total) or from a model's predictions (Residual). Squaring the differences ensures that deviations don't cancel out and penalizes larger errors more heavily, a core principle in least-squares regression.
