A heuristic that occurs when a visible successful subgroup is mistaken as an entire group, due to the imperceptible failure subgroup. The bias’s name derives from the error an individual makes when a data set only considers the “surviving” observations, without considering those that didn’t survive.