On Meritocracy in Optimal Set Selection