My Photo

Welcome to Mannionville

  • Politics, art, movies, television, books, parenting, home repair, caffeine addiction---you name it, we blog it. Since 2004. Call for free estimate.

The Tip Jar

  • Please help keep this blog running strong with your donation

Help Save the Post Office: My snail mail address

  • Lance Mannion
    109 Third St.
    Wallkill, NY 12589

Save a Blogger From Begging...Buy Stuff

The one, the only

Sister Site

« Fugue for Tinhorns | Main | Control freak out »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Ralph H.

This is a bit generalized, reductionist, pessimistic, in my opinion. Let's set aside the question of whether Republicans exclusively drink their own Kool-Aid & talk about us liberals, progressives, what have you. We in the RBC as often as not are hardly coming into these issues out of the blue. Re. Silver, I made a point of reading up on his methodology -- which he shared widely -- & also happened to know something about Bayesian analysis, so his predictions made intuitive sense. On other issues, particularly involving national security, I try to read what the much-despised "mainstream media" reports -- just the facts, ma'am, as Joe F. used to say -- & amplify that with selective use of the blogosphere (which can be much better than many people believe), & further draw on my long background in military & intelligence issues, which I admit is a blessing for guys like me & the lack thereof a handicap for others of our broad political persuasion. But the point is that my opinions nearly always evolve, and are not reflexive. I'd like to think there are others in the RBC like me.

Ken Houghton

My faith in Mark Bowden is significantly lower than yours.

And the reason many of us trusted Silver (or, even more, Sam Wang at Princeton who had the courage of his data, unlike Silver who flipped VA [iirc] in his final EC prediction) was that you could see most of the inputs--the polls. And they stayed stable or went more for BarryO as time went by.

Which meant that it was more difficult to come up with a reasonable narrative as the election neared. You had to assume that a Presidential election would be relegated to an off-year or midterm mix of voters. You had to assume the "low-information" voters would all break the same way (which has, to be nice about it, never happened on a national scale over multiple states)--when the "undecideds" themselves were trending Obama with each poll.

What fit the data: that Obama was likely to be re-elected and that the chances were getting better as the election came closer.

Does that mean certitude about Silver's prediction of a 75% or 78% or 80% (let alone another decimal place) prediction? No. The faith is in the story, because it fits the data. Given the same chance I got in 1992, would I have taken Obama at even odds? Absolutely. (You owe me a sandwich, I believe.) Would I have given 2:1 to a Romney supporter? Probably; almost certainly by the last week of October. 3:1 or 4:1? Maybe, but I wouldn't swear to it even the day before, even though Nate Silver said that range was Fair Value.

And, as Ralph H. notes above, Silver didn't make a secret of his methodology, unlike the Unskewed Polls gaggle. So you really had to make several assumptions that didn't fit with 2004, let alone 2008, to decide Silver was wrong, as opposed to "right in the abstract."

Because the story fits the data--not that the story exists and the data has to be culled to fit it.


The thing about Silver, as I understand it, is that he set up his methods and simply followed them. He made no judgement about what the data said, but simply reported it. And remember, he stated his results in terms of probabilities. Saying someone has an 80 percent probability of winning an election is not predicting that he will win the election. Predictions may be made on the basis of the probabilities, but the probabilities themselves are not a prediction.

Norma Cartwright

My head aches and will probably explode. My ego is bruised; I infer that you are saying that I'm NOT infallible...and that is just unacceptable.

Dammit, I'm a Democrat. A liberal! Even a Dirty-Fukkin-Hippie, fer gawd's sake! How can I be wrong? Or, is it that I've taken for granted that my worldview, superior as it is to RightLandia's, has morphed into a superiority-complex. Oh, the Hubris! It burns!

Actually, I think the headache comes from two sources: one, my mind's door opened so fast, I can only compare it to a "home invasion", complete with miscreants wielding assault rifles and Super-sized Magazines loaded with RealityBites.

Second (and the lesser of the events), the flashback of Rummie's performance-art/artful-dodge...

The comments to this entry are closed.

Data Analysis

  • Data Analysis


April 2021

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  

Movies, Music, Books, Kindles, and more

For All Your Laundry Needs

In Case of Typepad Emergency Break Glass

Be Smart, Buy Books

Blog powered by Typepad