sam vest Wrote:
-------------------------------------------------------
> Bubble popped Wrote:
> --------------------------------------------------
> -----
>
> > But he's still voting for him.
> >
> > As will about 51% of voters.
>
>
>
> From the Gallup website:
"Each
> seven-day rolling average is based on telephone
> interviews with approximately 3,050 registered
> voters"
>
> 3,050 people is nothing.
>
>
http://www.gallup.com/poll/150743/Obama-Romney.asp
> x
Not talking about Gallup. You can call it roughly even now by the time that you factor in error.
I'm not going to dig out the studies to support this but they're easy to find if you want to look. In an incumbent Presidential election, independents and undecided voters break decisively to the challenger about 8 out of 10 times. Or said another way, if they're not counted in the incumbent's totals at this point, then the probability is that they're voting against him. Unless something major event happens within the next two weeks, they will again. This year, the number of those voters amounts to about 4 or 5 points in the polls. The split will be probably 3:1.
So, the general rule is that for every point under 50% that the incumbent has in the polls going into the election, the likelihood of his winning decreases by about 20%. At 49% he's OK. 48% it's an even bet. 47% he's done.
Obviously, the electoral votes come into play so you'd need to extend that on a state-by-state basis. And every rule is made to be broken, but that's what history says.