A bit like publishing bias where only positive results end up in scientific papers.