On the growth "models" of NS stats