Skip to main content

View Diary: Pure 190 Proof Schadenfreude! Nov. 5th 'ROMNEY WILL WIN' (167 comments)

Comment Preferences

  •  I wonder if those commenters know (10+ / 0-)

    that the University of Colorado people they like to quote (that have the model like Nate's only predicting a Romney win) have never predicted a Presidential election until this year? All of the other "predictions" were "postdictions"; that is, they looked back and they SAY their model would have accurately predicted those other elections. So they are exactly Zero in actually predicting an election outcome.

    Friends don't let friends vote Republican.

    by OhioNatureMom on Wed Nov 28, 2012 at 05:04:49 AM PST

    •  Interesting. I did not know that. (2+ / 0-)
      Recommended by:
      OhioNatureMom, Eric Nelson
    •  Oh, they were right (1+ / 0-)
      Recommended by:
      OhioNatureMom

      They just mixed up the candidate names during the copy and paste editing.  ;-)

    •  Nate Silver tried to clear things up by tweeting: (3+ / 0-)
      Recommended by:
      OhioNatureMom, MRA NY, Lefty Coaster
      Also, it’s false advertising to claim CU model has predicted the last 8 elections right. It’s a new model. Hasn’t predicted anything yet.
      And Dr. Berry (one of the CU professors) responded with this:
      Mr. Silver and others confuse a prediction with an estimate. Our model was developed after the 2008 election. The only election that we forecast is the 2012 election. When we populate the model with data from each of the election years from 1980 through 2008, we correctly estimate the winner in each of those elections.
      (emphasis mine)

      Many media outlets, if not most, reporting about the model, claimed that it had been in use since 1980 so it's no wonder that right wingers grasping at straws paid no attention to Nate (or to Dr. Berry's response).

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site