Pollster.com

there's gotta be a typo in the rasmussen nat numbers. i think you mean clinton 47 mccain 43

And there goes Obama, bouncing back up again after a few days below water. This is becoming comical at this point.

Lets make a guess, whichever candidate has a good news week will gain 5-6% in the polls to a make of a 10% lead for Obama and 5% for Hillary with a base of +2-3% for Obama. Just as the country has voted. If it's a good for Obama he jumps to 10% if it's a bad week he is down by 3-4%. But on average it's a 2-3% lead, which is why he is just a bit ahead in the popular vote. These polls tell us nothing when an electorate is already so evenly split.

What is really frustrating is Gallups silly attmpts to interpet the data. When Obama was down a couple of points he was tanking. When he is in the lead by 5 points, it is a statistical tie. Gallup uses it's daily poll not to provide information but to gnerate publicity for itself by over interpeting the data. Rasmussen at least aalmost always interpet change as 'statistical noise" until the pattern has existed for several days or more. Gallup helps cause the low regard polls are held in by it's overinterpatation of data.

You guys must not be reading the same poll narratives that I do... which have been consistent with their analysis.

Note that the Obama being "down a couple of points" was not the story, but rather that Obama had a ten point swing from being up around 8 points.

I admit to feeling some pleasure at seeing Obama's negatives climb up - not because I'm a Clinton partisan (though I am) but because from day 1 I thought the "her negatives are so high / Obama's aren't" argument was inexcusably naive.

Yes, I am a Somberby devotee.