Randomised clinical trials
Clinical trials are carefully designed to judge the medicine on a level playing field. To do this, researchers have two groups of people who are the same in every way (or as close they can get to it), but who are given different treatments. A good clinical trial will have similar people in each group to avoid any unfairness.
The best way to make the groups similar is to randomly put people into groups (e.g. by drawing names out of a hat). So randomised clinical trials are considered better than non-randomised trials.