I have expressed a lot of reservation about the feasibility of prediction of large, important outcomes in the social world (link, link, link). Here are a couple of observations drawn from these ear…