These biases include "survivorship bias" and "incubation bias." Survivorship bias, which also affects mutual-fund data, occurs when dead or closed funds are dropped from the database and are no longer included in long-term performance averages. Many funds die because their performance was awful, so once such funds are stripped from the data, the average performance looks better than it really was. Incubation bias, meanwhile, results from the practice of hedge-fund companies starting multiple funds internally and reporting results only for the ones that do well (while closing the ones that do badly).