auditor0007
Gold Member
Amazing
Completely opposite of how it will turn out
RealClearPolitics - 2012 Election Maps - Electoral Map No Toss Ups
Hmmmm, let's review.
On the one hand, academic, peer reviewed, research that has successfully predicted every election since 1980.
On the other, polls.
I know which I lean towards.
The problem with the model is that it was set up after the fact, not before. What they do is find a set of circumstances that works for all past elections, and then they say it is a guaranteed predictor of future happenings. We see this with trends in sports gambling all the time. On Monday night football, the home underdog who lost their last three games on the road has not lost in this situation for the last ten years, so we can now predict that a very bad team is going to win or cover the spread on a Monday night game. The problem with it is that it's just a trend, and sooner or later that trend doesn't pan out. Just ask all the sports gamblers who have lost betting on trends.