“Behind every great point estimate stands a minimized loss function.” Rasmus Bååth ( Jan 8th, 2015 )