Originally Posted by patteeu
The question I'd ask about any predictor's claim of accuracy would be how accurate how far in advance? All the pollsters polls get more accurate as the election becomes imminent, for example, but 2 months out they are less so. I know that Silver has a good reputation, but it's not obvious to me that his reputation is built on two month advance predictions. That's all I'm saying.
True Dat. It's still a probability outcome prediction, not a certainty. And those probabilites change with time due to impacts from the variables in the algorithm.
I don't see those variables changing. Debates are less of a factor because gaffe's or "moments" are no longer campaign killers or propel the candidate to victory ie. Nixon's 5:00 shadow or regans are you better off than you were 4 years ago moment.
the only thing that can change the variables enough to create a different outcome is a "real" scandal directly connected to Obama. I don't see anything connected to a possible escalation with Iran that would bring Obama down.
It will be damn close. But, without taking 2 out of 3 of Florida, Pennsylvania and Ohio there is simply no path for Romney to win. Not saying that Romney cant take 2 out of those 3. It's just the chance % in Silver's algorithm.