via Why Ames Actually Matters by Nate Silver
Avert your eyes
The gist of Nate’s argument is that Ames is more predictive (of Iowa caucus results) than people think.
This does not mean that Ames is “important” per se: that a candidate will do well in the caucuses because he did well in the straw poll. But Ames does seem to pick up on a variety of “intangible” factors that don’t show up well in other variables, and therefore serves a useful role as a leading indicator.
In English: If Bachmann wins today, she’ll probably come in 1st or 2nd in the caucuses in January. If Paul wins, no one knows what the fuck it means, but he’ll be better positioned for January than before, presumably.
In reality: None of the results today matter if Palin decides to run, other than perhaps a Paul victory. 5 months from now the race could very well be Palin/Perry/Romney/Paul with Bachmann and Pawlenty already dropped out. Romney could even be dropped out by then if Palin and Perry drown him out.
– A Bachmann win has short term consequences: a collective groan by establishment Republicans as well as increased pressure and attacks on her. Long term I doubt it changes anything at all.
– A Paul win will mean very little short term, but I suspect it could enable the Paulites to recruit many more national libertarians to their cause, boosting fundraising, viral media presence, volunteer network, and having the media in general take him a little more seriously.
If you’ve never read or heard of Nate Silver, take a look at the top link. He is the single best (non-campaign worker) electoral statistician analyst around, making a name for himself with his advanced state prediction models during the 2008 election and going from running a small private site to being hired by the New York Times. He’s always quality.