.comment-link {margin-left:.6em;}

Born at the Crest of the Empire

Saturday, November 01, 2008

One of those "insidery" things on polling averages

Most of the polling averages use a week for their sample which means that the polls that drop out the bottom (after one week) can have as much effect as any new polling released today.

By the way, Gallup tracker Obama +10 in both models?


  • Obama gained in both "likely voter" models and stayed at +11 in the RV model.

    This is "tightening"?

    The infomercial seems have stopped any potential McCain gains.

    Gallup trad LV model of O +8 leaves only 6% undecided. That means that McCain has to win all undecided voters and still pick up most of the MOE on top of that.

    From Gallup:
    The vote choices of these early voters -- all of whom are included in the likely voter pool since they are definite voters -- skew more toward Barack Obama than the sample average. Thus, more and more of these Obama-oriented voters' choices are being "locked in" to the likely voter pool through early voting, benefiting Obama.

    Sure, polling isn't perfect, but McCain hasn't led in LV trad since it was instituted, and that is the strongest model for McCain. More than two dozen national polls by Gallup have failed to detect the alleged "hidden support" that is supposed to be the secret weapon for McCain. And the idea that Obama's support is too lazy to turn out has been disproven by early voting.

    And the national trackers are the good news for McCain. The state polling is downright grim for Republicans.

    McCain leads MT by only four, and ND by only one in R2K (10/28). That same polling shows McCain up by only 1 point in Arizona. The magic of MOE works both ways.

    By Blogger Todd Dugdale, at 3:11 PM  

  • That's kinda why I mentioned it. I have a couple of non-commenting readers, two of whom I ran into this morning who saw that ridiculous Drudge/Zogby thing and were freaking out, so I thought I'd mention the Gallup because that was part of my verbal, "stop freaking out."

    More specifically, the traditional at +10 really surprised me. It's been moving up quickly, but that's alot from where it was a week ago.

    I'm wondering to what degree this Gallup movement is real, noise, or some sort of early voting feedback affecting their models.

    .....Huh. If I''d just read your comment a little further.....

    And you may know this, shouldn't I be seeing a last round of the set piece polls, NBC, NYTimes, McClatchy, etc? Tomorrow? Monday?

    I figured I'd hold off the official predictive post until then.

    By Blogger mikevotes, at 3:45 PM  

Post a Comment

<< Home