A year in politics is a long time. Predicting political races is inherently problematic, but when taking ALL the information from polls and historical context into account, along with the what for Republicans will be the dramatically positive effect of the Tea Party, then the most likely scenario is that Obama loses in an electoral college landslide. Even if the popular vote winds up being somewhat close, Obama will lose state after state, notably swing states like Florida and Ohio, because people now know how incredibly twisted and destructive Obama is and any one of the Republicans save possibly Paul or Huntsman could probably trounce him.
Unlike the wildly biased and disconnected from reality "analysis" which BN provided, mine actually is defensible and will probably come to pass! But I must express thanks to BN for once again doing an excellent job of illustrating the WRONG way to analyze situations! He's a natural at that!