Wednesday, November 07, 2012

Predicting the Election

This graph from the NY Times showing the shift in presidential votes between 2008 and 2012 is I think my favorite graphic:

Screen Shot 2012 11 07 at 5 38 10 PM copy

I think this chart is the most telling by far of what happened:

11 6 2012 majority copy

WonkBlog of course has The 2012 election in charts. I don't think there's anything surprising in them other than that of those people listing health care as the most important issue, 74% were Democrats and only 25% Republicans, that's not a mandate to repeal Obamacare.

xicd got it right again:


Nate Silver hasn't posted a summary yet but he seems to have gotten every state right for the presidential election. I think more importantly he's been right about the election the whole year.

Screen Shot 2012 11 07 at 5 33 27 PM

Paul Glastris wonders about The Mystery of Why Republicans Were So Sure They’d Win. "One of the more interesting questions of this election is how and why so many Republicans, who are certainly just as capable as Democrats at reading polls, chose to ignore the overwhelming statistical evidence that a Romney win was unlikely. I suppose one could say it is not much of a mystery, and that this mass refusal to accept politically inconvenient facts is of a piece with, say, conservative denial of global warming. And maybe that’s all it is. But I suspect that there was something else at play, too.

Alexis Madrigal reminds us that Data Doesn't Belong to the Democrats and that "The 2012 election was a triumph of Democratic data over Republican data."


Karl said...

I like the NYT plots too. Particualrly the margin chart which I think shows some data not available on the typical charts. The vector chart seems to have missed a chance though using both magnitude and phase to encode a single variable.

Howard said...

I'm not sure what you mean. It does use magnitude to show how much of a change. It uses both direction and color to show the same thing, the change of party direction.