糖心TV

Drazen Prelec (MIT Economics)

Finding truth even if the crowd is wrong

I will discuss the problem of finding the true (or best) answer to a question in situations where the individuals with the best evidence are possibly a small minority, and no independent measures of competence are available, i.e., there is no track record on similar questions. However, instead of past performance, we can measure how well individuals can predict what others will say about the unique question at hand. Is it reasonable to treat success at this 'meta-knowledge' game as evidence of genuine expertise?