Skip to main content

View Diary: More accurate than Nate Silver (308 comments)

Comment Preferences

  •  Ohio (4+ / 0-)

    Does anyone find it suspicious that Ohio is the only state whose actual was lower than predicted, and substantially so? I remember poll after poll showing 4, 5 point leads, few less than 3, yet it seems to be the only heavily polled state where Obama significantly underperformed the aggregate polling.

    On top of that, Ohio was the most heavily polled state, so in theory should have been the most accurately aggragated. Yet, check this graph from Votamatic:

    There could be any number of explanations other than active vote tampering. Yet, in 2004, Bush well outperformed exit polls. And then there was Karl Rove, disputing Ohio. Could all be a bunch of coincidences. My concern is, if only 1-2% of votes can be "counted" a certain why in a marginal election, and OH polling was much narrower than it was, well... that's why the experimental polling software really had me worried.

    •  There are 300k provisonals left. (3+ / 0-)
      Recommended by:
      Vicky, Angie in WA State, Onomastic

      Say 200k get counted. They're likely to break ~2/1 D. You've upped the margin a factor of 1.6, to over 3%. I'm not saying that there was no cheating on the count, but simple visible vote suppression, some of which will be undone by the remaining count, is the obvious explanation for most if not all of the discrepancy.

      Michael Weissman UID 197542

      by docmidwest on Wed Nov 14, 2012 at 10:43:15 AM PST

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site