Cherry-picking caused by selection/survivorship bias, with a multiple-testing problem. If the target was defined only after seeing the results, it is also the Texas sharpshooter fallacy.