NJ: Parties Wonder Which Side's Polls Reflect Reality

[brandymac27]

Just glancing at the map of his prediction and looking at the results map on CNN it would appear he only missed 1 senate race. He slightly favored the Republican in Montana and it actually ended up tipping to the Democrat.

Edit: Actually North Dakota hasn't been called yet, he favored the Republican there and with 93% of the vote counted the Democrat is leading 160,752 to 157,758.

Damn, that's pretty good. Do you know how he does this? I know he won't give out his "formula" lol, but there has to be some way he's doing this with such accuracy.

[G.A.C.O.L.B.]

http://isnatesilverawitch.com/

As of Wednesday, November 7th, 2012, Nate Silver is probably a witch.

Lol

[greenspandan]

Damn, that's pretty good. Do you know how he does this? I know he won't give out his "formula" lol, but there has to be some way he's doing this with such accuracy.

his "secret sauce" isn't that complicated. he does research on past elections to figure out the "house effect" (how far left or right) a particular poll tends to lean. then, after adding in the house effect for each poll, he averages all the polls together. there are also other details where he incorporates some economic factors, and uses national polls to detect movement, but for the most part if you just account for house effect of polls and average them together you will come out real close to what silver does.

(personally, i feel like economic factors are already baked into the polls themselves, and he's just sort of adding noisy variables by trying to incorporate them seperately, but hey he's clearly a smarter person than i am)

[China]

http://isnatesilverawitch.com/



Lol

And what do you do with witches?

[PeterMP]

Thank you for the explanation. This guy seems to be a genius with kind of stuff. But ultimately, he couldn't make his predictions without the use of the other polls if I'm understanding you correctly. I wonder if, over time, he eliminates his use of polls that are way off so he can get even more accuracy out of his own predictions?

Knowing that people fail, especially if the fail in the same manner consistently, is useful information.

[brandymac27]

his "secret sauce" isn't that complicated. he does research on past elections to figure out the "house effect" (how far left or right) a particular poll tends to lean. then, after adding in the house effect for each poll, he averages all the polls together. there are also other details where he incorporates some economic factors, and uses national polls to detect movement, but for the most part if you just account for house effect of polls and average them together you will come out real close to what silver does.

(personally, i feel like economic factors are already baked into the polls themselves, and he's just sort of adding noisy variables by trying to incorporate them seperately, but hey he's clearly a smarter person than i am)

Thank you for the explanation. This guy seems to be a genius with kind of stuff. But ultimately, he couldn't make his predictions without the use of the other polls if I'm understanding you correctly. I wonder if, over time, he eliminates his use of polls that are way off so he can get even more accuracy out of his own predictions?

[Monte51Coleman]

It's hard to image that his predictions could get much more accurate.

Plus, kudos to the folks at Public Policy Polling (PPP). They took a lot of flack from right-wingers for their supposedly ridiculous left-wing slant and were also highly accurate.

[DjTj]

Knowing that people fail, especially if the fail in the same manner consistently, is useful information.

I wonder if he just assigns weights based on bias or whether he has a real model of the particular sampling errors associated with each poll that allows him to adjust based on specific issues or accounts for correlations between different polls.

There are definitely many ways to do this, and I wonder if there will be more Nate Silver competitors in future elections.

[Monte51Coleman]

http://projects.newsobserver.com/und...2012_elections

newsobserver.com blogs

Public Policy Polling most accurate predictor of 2012 elections


Fordham University has ranked the polls that predicted the outcome of the presidential election, and Raleigh-based Public Policy Polling came in first—and second.

One of the polls conducted by the Democratic-leaning firm was independent and the other in tandem with Daily Kos and the Service Emplyoees International Union.

Here are the full rankings:
1. PPP (D)*
1. Daily Kos/SEIU/PPP*
3. YouGov*
4. Ipsos/Reuters*
5. Purple Strategies
6. NBC/WSJ
6. CBS/NYT
6. YouGov/Economist
9. UPI/CVOTER
10. IBD/TIPP
11. Angus-Reid*
12. ABC/WP*
13. Pew Research*
13. Hartford Courant/UConn*
15. CNN/ORC
15. Monmouth/SurveyUSA
15. Politico/GWU/Battleground
15. FOX News
15. Washington Times/JZ Analytics
15. Newsmax/JZ Analytics
15. American Research Group
15. Gravis Marketing
23. Democracy Corps (D)*
24. Rasmussen
24. Gallup
26. NPR
27. National Journal*
28. AP/GfK

[BRAVEONAWARPATH]

Click on the link to read the rest.
http://www.csmonitor.com/USA/Electio...de-in-US-media

Surprise at Obama’s victory illustrates growing partisan divide in US media




The big media story emerging from President Obama’s reelection is the fact that so many on the right were so stunned by the results.

Social media were abuzz with shock and dismay at what many conservatives felt was a last-minute reversal of the prolific positive predictions they had been hearing.

More than a few conservative commentators, including prominent pundits such as George Will, had been predicting that Mitt Romney would take more than 300 electoral votes in a landslide election on Tuesday.

At the same time, statistical blogger Nate Silver at The New York Times and survey aggregator Real Clear Politics were citing polls that showed Mr. Obama with a clear lead.

But, rather than the purportedly surprising election results reflecting some national subversion of the voting process, many political scientists and other analysts say this right-wing upset is dramatic evidence of a growing partisan divide in our media.

Increasingly, the public consumes media that reinforce personal views rather than give actual information about the world, says University of San Francisco political scientist Corey Cook.


“The biggest story of this election is the stories that were being told about the election,” says Professor Cook, adding, “the two sides had very different views heading into the election night.”

[DjTj]

http://www.csmonitor.com/USA/Electio...de-in-US-media

Conservative media really failed in this election. I wonder if it will hurt their ratings, or if will push them to favor a little more truth over truthiness?

On the biggest political story of the year, the conservative media just got its ass handed to it by the mainstream media. And movement conservatives, who believe the MSM is more biased and less rigorous than their alternatives, have no way to explain how their trusted outlets got it wrong, while the New York Times got it right. Hint: The Times hired the most rigorous forecaster it could find.

It ought to be an eye-opening moment.

But I expect that it'll be quickly forgotten, that none of the conservatives who touted a polling conspiracy will be discredited, and that the right will continue to operate at an information disadvantage. After all, it's not like they'll trust the analysis of a non-conservative like me more than the numerous fellow conservatives who constantly tell them things that turn out not to be true.

http://www.theatlantic.com/politics/...d-file/264855/

[AsburySkinsFan]

Conservative media really failed in this election. I wonder if it will hurt their ratings, or if will push them to favor a little more truth over truthiness?

http://www.theatlantic.com/politics/...d-file/264855/

The Right was quick to blame every one/thing but themselves, forget that they got it wrong forget that they lied about and manipulated the numbers in their favor, forget that they slammed the real researchers as voodoo shamen...who is to blame? The self entitled minorities who want stuff.

So no I don't believe for a moment they'll admit their misske and change, in fact I believe they'll just double down again.