You have 300 people with disease X, which causes a 100% chance of death. Of the 300, half have genetic marker A and half have genetic marker B. You have 100 doses of a drug. For people with marker A it cuts the risk of death to 10%. For people with marker B it only cuts it to 50%.