Herman Cain’s actual vote share in the 2012 Republican primaries will be significantly lower than his averaged national poll numbers at the time (as indicated by standard measures of statistical significance).

Created by NathanMcKnight on 2011-11-11; known on 2012-08-30; judged right by gwern on 2012-09-08.

• NathanMcKnight estimated 80% on 2011-11-11
• gwern estimated 75% on 2011-11-11
• JoshuaZ estimated 85% on 2011-11-12
• Nic_Smith said “Why is everyone so confident that the polls will even go over rather than under? Am I missing something about this claim?” on 2011-11-12
• NathanMcKnight said “@Nic: See “The Bradley Effect” (http://en.wikipedia.org/wiki/The_bradley_effect))” on 2011-11-12
• HonoreDB estimated 40% on 2011-11-12
• fork estimated 95% on 2011-11-12
• Nic_Smith estimated 65% and said “Thanks.” on 2011-11-13
• Nic_Smith said “How to deal with this in light of Cain’s exit from the race? Will he still be listed as an option in polls?” on 2011-12-05
• gwern said “I think we could morally close this as right – if no one is polling for him, one could expect the vote share being even lower” on 2011-12-05
• gwern   judged this prediction right on 2012-09-08.