So how did Nate Silver do?
His final prediction for the popular vote was 50.8 to 48.3. The result as of now is 50.4 to 48.1.
His probability calculations corresponded with the actual outcome in all 51 Electoral College races.
He said the most likely Electoral College vote count would be 332 to 206. The final tally: 332 to 206. However, his model predicted a final outcome of 313 to 225. (Not sure how that would happen or why his prediction differs from his most probable result.)
Not too shabby.
More anomalies in the Senate races. I guess he had less reliable data to work with. He had the probability of a Republican victory in Nebraska at 92.5%, but the Democrat won by 1 point. Montana was 62.5% likely to go Republican, but the Democrat won by almost 4 points.
He had the chance of the Democrats taking the 8-seat lead they achieved, at 17%, his third most likely outcome. (He had a 5 or 6 seat lead as the most likely outcome at around 30%) His prediction was a 5-seat lead.
|by Anonymous||reply 20||11/07/2012|
He predicted all 51 electoral races correctly.
|by Anonymous||reply 1||11/07/2012|
OP, it is clear you don't understand statistics and are the last person to evaluate Nate Silver,
But to answer your question, Nate Silver, as did all the other quant jocks that applied statistical analysis to the polls: Pollster.com, TPM Poll Tracker, Sam Wang, etc.
In addition to Obama's win, this will be known as the election when Moneyball came to politics and the geeks crushed the pundits.
|by Anonymous||reply 2||11/07/2012|
And yet, R2, you do not identify a single error in my post.
And Nate Silver kicked butt in 2008 too.
|by Anonymous||reply 3||11/07/2012|
Pay me $500 per credit and I will educate you, OP. That does not include remedial math class.
|by Anonymous||reply 4||11/07/2012|
Your pitch is rather feeble, R4. I think I'll pass.
|by Anonymous||reply 5||11/07/2012|
[quote]He had the probability of a Republican victory in Nebraska at 92.5%, but the Democrat won by 1 point.
Bob Kerrey, the Democrat, LOST by 16 points.
|by Anonymous||reply 6||11/07/2012|
r2 why are you reading OP's post as anti-Silver when he is obviously commending Silver for acing it?
|by Anonymous||reply 7||11/07/2012|
Do you suppose it's just a coincidence that the figures I mention are correct for North Dakota?
|by Anonymous||reply 10||11/07/2012|
I think it is silly for pundits to be so upset by statistics. The fact that they seem to be so ignorant on the subject just blows my mind. If they are going to even look at polls, they should have some basic "Cliff' Notes" idea of what they are looking at. Their blind defensiveness is actually disconcerting. It really is like creationists trying to defend against evolution in that there is a lot of sputter and posturing, but no idea of what they are arguing against. These pundits like to think that they can predict the outcome, but they are really only discussing some of the assumptions that may apply. The fact that they seem not to understand this makes me wonder about their ability to interpret anything.
Some of these same people just fell apart when Paul Krugman won the Nobel prize. Economics relies a lot on statistics and modelling, however, and Krugman does that in his work. It gives him the basis for making the commentary and performing the analysis he does in his appearances and columns. Some people, like George Will, assume he is making it up as he goes along. Just because he does not bring out the white board when he talks to someone like Will doesn't mean he hasn't thought many times about the math before he discusses the issue.
|by Anonymous||reply 11||11/07/2012|
There are no coincidences.
|by Anonymous||reply 12||11/07/2012|
[quote]OP, it is clear you don't understand statistics and are the last person to evaluate Nate Silver,
Uh, no, the last person would be Dean Chambers or someone else of that ilk.
[quote]But to answer your question, Nate Silver, as did all the other quant jocks that applied statistical analysis to the polls: Pollster.com, TPM Poll Tracker, Sam Wang, etc.
You might try adding a verb and perhaps an adverb or two to this sentence if you'd like to comment on what Nate Silver and the other quant jocks did, and how well they did it.
|by Anonymous||reply 13||11/07/2012|
What really infuriated me, R11, was the tendency of most networks to treat early returns are significant. "With 3% of the returns in, candidate A is leading 54 to 46" they'd report. Some of them even have graphics that changed from red to blue as the totals are counted regardless of which areas had been counted. A lot of people here (including the self-identified 2012 Poll Troll) would treat these early returns as significant.
At least MSNBC did not play that game.
You don't need to understand statistics to recognize that partial vote counts mean nothing without understanding which areas have been counted. Even KKKarl Rove was complaining about declaring Obama the winner in Ohio when all the outstanding votes were from Cleveland. (I can't imagine he didn't know better.)
|by Anonymous||reply 14||11/07/2012|
R14 Are you at all familiar with the concept of EXIT POLLING? The networks don't call races based on raw votes.
|by Anonymous||reply 15||11/07/2012|
I never said they did, R15. I said they reported early raw votes as if they were significant. You are very quick to anger. Perhaps some tea?
|by Anonymous||reply 16||11/07/2012|
He's going to get so much dick.
|by Anonymous||reply 17||11/07/2012|
Should we lick his taint now?
|by Anonymous||reply 18||11/07/2012|
He saved me a huge amount of stress and made me look like a political genius at an election party last night.
|by Anonymous||reply 19||11/07/2012|
FOR THE LOVE OF GOD WOULD SOME PLEASE SHIT IN MY MOUTH??!!!!
|by Anonymous||reply 20||11/07/2012|