It is not enough to show that drug A is better than drug B on the average. One is invited to ask, 'For which people ("& why") is drug A better than drug B, and vice versa? If drug A cures 40% and drug B cures 60%, perhaps the right choice of drug for each person would result in 100% cures.'