For every surprising result you read about, how many failed tests never got published?
In 1979, psychologist Robert Rosenthal named the 'file drawer problem': studies that fail to find an effect are far less likely to be submitted or accepted for publication than ones that do, leaving journals stuffed with false positives. He calculated how many null results would need to be sitting unpublished in researchers' drawers to cancel out a given published finding. It's now a central reason psychology and medicine both push researchers to pre-register studies before running them.