A Hacker News user questions whether prediction markets lived up to their reputation during the recent midterm elections. The author argues that when compared directly to traditional polls or Nate Silver's FLIPR model, prediction markets offered no significant advantage. Specifically, in races that were closely contested or experienced polling flips, prediction markets showed no predictive edge—what the author calls 'no alpha.' The post suggests these markets may be more entertainment than reliable forecasting tools and even proposes taxing them for their poor advertising.
The discussion invites others to reflect on whether this underperformance was obvious or surprising.
Source: Hacker News · Summarized by HeadlinesBriefing