Skip to main content

View Diary: Best public pollster of 2012: Public Policy Polling (42 comments)

Comment Preferences

  •  Sorry, but this is wrong... (0+ / 0-)

    Using your rules of public pollsters that polled four or more states within the last four weeks of the election, We Ask America isn't in here and they should be, b/c they would have beat out PPP. And if the the screen would have been
    three instead of four, Quinnipiac would have beat out PPP as well.  Here's why:

    CO: WAA was 2.4 off, Qpc didn't poll in last 2 wks
    FL:  WAA was 1.9 off, Qpc was .1 off
    IA: WAA was 3.8 off, Qpc didn't poll in last two wks
    NV: Neither polled in last two weeks
    NH: '' ''
    NC: '' ''
    OH: WAA was .9 off, Qpc was 2 off
    VA: WAA was 2.9 off, Qpc was 1.9 off
    WI: WAA was .1 off, Qpc didn't poll in last two weeks

    That puts WAA's average error at 2.0  and Qpc at 1.3

    •  Q didn't poll enough states (0+ / 0-)

      So it's not "wrong", you just disagree with the methodology. You can make your own chart. Personally, getting three races correctly isn't as impressive as getting nine or 10. But YMMV.

      As for WAA, I added them to the chart. My spreadsheet calculates an error margin of 2.12. You got all the WAA margins wrong. Not sure where you got them, or whether you're looking at the most updated vote counts.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site