In reference to https://www.predictit.org/Contract/1863/Will-Obama's-Gallup-approval-be-47-or-higher-for-January-12-14#data
Political betting draws its foundation from the idea that a market of people investing money will have a better grasp of political outcomes than other predictive institutions. If that's the case, where does the (seemingly, feel free to correct me) consistent undervaluation of Obama's popularity or ability to increase his approval ratings come from? If the comments on the linked thread are any indication (not to mention the bets themselves), then it seemed as though the SOTU would trash Obama's approval ratings. In fact, quite the opposite happened and the SOTU shot Obama's approval up from 47% to over 50, according to Gallup.
What accounts for PredictIt's market failure to see this coming in advance?