[BigEdLB]

A lot of these polls used to generate the map use REGISTERED VOTERS. That is a plus 5 Dem advantage. Understanding that most undecideds go for the challenger as well makes a difference too. And, of course, any incumbent under 50% even though leading is in trouble. Plus marstegreg's point about certain polls oversampling Ds is a good one.

Some polls, which should be called push polls,anything with NBC, CBS, ABC, and CNN attached, is suspect, on its surface.

Pennsylvania was put in the lean Obama column by a Quinnipiac Poll, which had Obama +8. But a poll of Registered Voters, not Likely voters that is 47-39 Obama with 14 undecided is almost useless. Look at this poll this way - Registered to Likely - D -2.5 R+2.5 So now it is 44.5-41.5 Obama. 14 undecided which splits 3-1 for the challenger (Romney) That means 4.7 more for Obama, and 9.3 more for Romney. Therefore you end up with 49.2 for Obama and 50.8 for Romney. And RCP says that this fresh poll, finished 5/1 is a poll that makes the state 'lean Obama'. RCP is run by good people, McIntyre, is a decent fellow, but with what polls they include in their averages, they do not mean much.

14% undecided? LOL

[marstegreg]

14% undecided? LOL

I agree!!! With that many undecided, he is toast. Also, when I look at the “lean” Obama states, some of those are just wishful thinking (and a lot of spinning). I wonder what the polls for Jimmy Carter looked like. I am guessing that he was doing better than Obama and look what happened to him. Even with the voter fraud, he isn’t going to win.