The best linear prediction rule is the one that has the least

A) error when predicting from the mean.
B) squared error when predicting from the mean.
C) error when predicting using that rule.
D) squared error when predicting using that rule.

D