The Error of Crowds is the idea that under many scoring rules, following an average of the various beliefs will result in a belief less useful than randomly selecting one. A related idea is that a popular belief is likely to be wrong because the less popular ones couldn't maintain support if they were worse than the popular one.