Im think people place far too much emphasis on standard deviation.I find that the loads that produce the best groups for me seldom have the lowest standard deviation.I judge loads on group size and no longer care about standard deviation.