Skip to main content

View Diary: Most accurate popular-vote pollsters: ABC/WaPo, Pew Research (12 comments)

Comment Preferences

  •  Be careful of the practise of "herding" (5+ / 0-)
    Recommended by:
    CalvinV, winsock, DigDug, buckstop, blueoasis

    which is when pollsters change their methodology as the election comes closer to bring themselves in line with other polls. Gallup and Rasmussen are famous for that, as they attempt to set a narrative with outlier polls until the election is just a few days away. Hence for some pollsters just looking at the last poll could be very deceiving.

    There is no such thing as an off year election. Every election effects each other. We need to work as hard in 2014 as we did in 2012.

    by pollbuster on Wed Nov 21, 2012 at 10:02:57 PM PST

    •  right--from this perspective both Gallup and Raz (3+ / 0-)
      Recommended by:
      Lawrence, buckstop, pollbuster

      look far worse, because they both kicked toward the median in the final week.  Raz still missed at the end by, what, 4-5 pts?  (R+1 vs D+3or4)  Actually I think 4-5 pts was about their typical lean this cycle, so maybe they stayed consistent to the end.  Gallup....heh...they were still showing R+5 LV until Sandy came along to save their asses by shutting down nat'l polling; when they finally were able to poll again they were only at R+1.  But right through Sept and Oct Gallup was giving us these crazy R+5-7 leads.  I dunno what WaPo was saying then...  

      •  Yes, Gallup was truly (1+ / 0-)
        Recommended by:
        Lawrence

        the worst of all of them. And you are completely correct, Sandy did "save their asses" by allowing them to shut their unabashed crap down.

        There is no such thing as an off year election. Every election effects each other. We need to work as hard in 2014 as we did in 2012.

        by pollbuster on Thu Nov 22, 2012 at 08:33:19 AM PST

        [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site