mynameisjim

That's a study of only 155 polls and published in 1989. Not exactly comprehensive. There have been 155 polls in the last few weeks probably in the current election. So studying 155 polls from decades ago is hardly definitive. It's a study, but nowhere near conclusive or even close to that. That's the problem with the internet, you can cut and paste a study of almost anything. Just Google "undecided voters break for incumbent" and you'll get studies that contradict the ones you posted.

There is no proof either way and you will never prove it either way. We are essentially talking about gambling here. Just like stocks, you can use the numbers up to a certain point, but at some point it gets fuzzy and the numbers can't help you anymore. That's where the skill and the art of something comes into play. The stock market is a collection of people, not a collection of numbers. Just like the electorate, you can use numbers to analyze them up to a certain point, but the final one or two percent is where the magic happens.

I guess we'll see on election day

I only see one of two things happening:

1) Obama wins a close election.

2) If Romney does win, it will be a landslide nobody saw coming that leaves everyone in shock. Basically he rolls the table.

__________________