Skip to main content

View Diary: The 2012 polling hall of shame (269 comments)

Comment Preferences

  •  A lot of your data in this post is really stale. (2+ / 0-)
    Recommended by:
    Lawrence, NM Ward Chair

    Some of it excerpted from already stale data on election night. There has been a lot of vote counting since then, and there is a lot more to come in Cali, Washington State, Arizona, Nevada Ohio, Florida.

    If we're going to write the gospel on polling error, let's at least be data-driven and get the numbers right — otherwise we're just as sloppy as the GOP hucksters.

    As I type, Obama just ticked up to 50.5% nationally, a symbolic mark that will be rounded to a 51-48 victory. His margins in Iowa, Nevada, New Hampshire are 6.6, 5.6, 5.8 respectively (or 7, 6, 6 in typical reportage).

    Once the remaining 10% of the vote rolls in from the Dem counties of Ohio, we're likely to see a 3 pt spread or better.

    Once the popular vote is finalized, Pew is likely to once again become the gold standard. And the RAND survey, which pegged its final spread at +3.32 Obama, may look like one of the best estimators of the cycle.

    I encourage you to be patient and repeat this analysis when all the data is in. Premature takes like this, when published, tend to be recycled for years to come — and that's the bad journalism that communities like this were built to correct.

    'Fie upon the Congress' - Sen Bob Byrd

    by Maxwell on Thu Nov 08, 2012 at 09:02:32 AM PST

    [ Parent ]

    •  Don't judge Rand by the final number (1+ / 0-)
      Recommended by:
      NM Ward Chair

      I am a big fan of their methodology, not so much their weighting procedure. We will have a much better read on Rand once the release all the data and we can try more sensible weighting.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site