Abstract: In this paper, we give a rule to compare a loss function with another. To single position perameter distribution class. We obtained that square loss function and convex loss function are equivalent about linear estimates. To i.i.d. sample. Wo gained a class of linear admissible estimates under convex loss function.