Data and findings are not always shared openly in science. Sometimes, researchers hoard those data and provide little to no access to colleagues from their field. These were the results of a survey of 563 researchers from 64 countries conducted by scientists at Martin Luther University Halle-Wittenberg (MLU) and the German Center for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig. The so-called "Gollum effect" hampers collaborations and harms the careers of less established researchers, claims the team.