NYT Nate Silver Doubles Down: 75% Chance Obama Wins

He does - but he also looks at their historical biases and accounts for them. So if Gallup is always 4 points more Republican than every other poll, he will account for that in his model. Or if Gallup is always wrong, but not in one particular direction, it will be weighted less in the model.

This is why the model is much better than just averaging all the polls together.

Absolutely - as long as the model processes the data correctly. For example, the polls still showed Romney winning VA and FL, but the model knew there was a bias there and adjusted for it. Just averaging it all together wouldn't do that.

 

OK fair enough.

But, Nate Silver embarrassed these guys pretty bad in this election. Silver needs these guys to get the data. But, he's shown that they aren't doing a good job interpreting it -- or else they are intentionally misinterpretting it. I'm betting the Romney campaign had data internally that informed them they would lose despite the public polls saying they had a chance. The reputations of Gallup and Rasmussen took a big hit.

 

I think they will try to confuse and mislead people by coming up with similar "models" that predict the outcomes they want.

 

lol, you are pathetic..seriously how can someone have a mindset like this. I do not get it!

 

Nate Silver is like Google News in that he depends on the work of many others (i.e, the countless state and national polls) to gather data while he aggregates, interprets and gives the data to the public in easy to consume form. His credibility comes from his track record of being right.

 

Fox News

 

and his methodology is open to everyone. its very simple and easy to understand you just have to take time to read it.

http://fivethirtyeight.blogs.nytimes.com/methodology/

here's an outline of his steps:

Stage 1. Weighted Polling Average

* Recency. More recent polls receive a higher weight.
* Sample size. Polls with larger sample sizes receive higher weights.
* Pollster rating. Lastly, each survey is rated based on the past accuracy of “horse race” polls commissioned by the polling firm in elections from 1998 to the present.

Stage 2. Adjusted Polling Average

* The trendline adjustment. An estimate of the overall momentum in the national political environment is determined based on a detailed evaluation of trends within generic congressional ballot polling.
* The house effects adjustment. Sometimes, polls from a particular polling firm tend consistently to be more favorable toward one or the other political party. Polls from the firm Rasmussen Reports, for example, have shown results that are about 2 points more favorable to the Republican candidate than average during this election cycle.
* The likely voter adjustment. Throughout the course of an election year, polls may be conducted among a variety of population samples.

Step 3: FiveThirtyEight Regression

A state’s Partisan Voting Index
The composition of party identification in the state’s electorate (as determined through Gallup polling)
The sum of individual contributions received by each candidate as of the last F.E.C. reporting period (this variable is omitted if one or both candidates are new to the race and have yet to complete an FEC filing period)
Incumbency status
For incumbent Senators, an average of recent approval and favorability ratings

Step 4: FiveThirtyEight Snapshot

This is the most straightforward step: the adjusted polling average and the regression are combined into a ‘snapshot’ that provides the most comprehensive evaluation of the candidates’ electoral standing at the present time.

Step 5. Election Day projection

It is not necessarily the case, however, that the current standing of the candidates – as captured by the snapshot — represents the most accurate forecast of where they will finish on Election Day.

Step 6. Error analysis

Just as important as estimating the most likely finish of the two candidates is determining the degree of uncertainty intrinsic to the forecast.

The error is higher in races with fewer polls
The error is higher in races where the polls disagree with one another.
The error is higher when there are a larger number of undecided voters.
The error is higher when the margin between the two candidates is lopsided.
The error is higher the further one is from Election Day.

Step 7. Simulation.

Knowing the mean forecast for the margin between the two candidates, and the standard error associated with it, suffices mathematically to provide a probabilistic assessment of the outcome of any one given race. For instance, a candidate with a 7-point lead, in a race where the standard error on the forecast estimate is 5 points, will win her race 92 percent of the time.

 

Just a few miles from Charlotte, Davidson College visiting professor Josh Putnam, whose blog FrontloadingHQ has been a hit with political junkies from both parties, said more than a month ago that Obama could carry every swing state he carried the last time with the exception of Indiana and possibly North Carolina. That would mean, he said, 332 electoral votes for Obama and 206 for Romney. Putnam had this same prediction on his blog in July.


http://obsdailyviews.blogspot.com/2012/11/davidsons-putnam-correctly-called-obama.html#storylink=cpy

 

You are still not seeing the difference. Polls are like box score keepers. And Silver is a stat geek using it to predict future success. They are inherently different things. Now, if Gallup or any other poll uses their poll number to prove a point, then perhaps you are right. But raw data itself isn't wrong. If anything, polls are vindicated this election, because only with accurate info comes accurate predictions. And the info provided by various polls seem right on the money.

 

and it's the electorate is young and dumb that is the reason.

because what he predicted about the electorate is so far opposite of what actually happened, it's the electorate that is young and dumb and the reason for the Romney loss...

 

No, I understand the difference. Their polling doesn't seem problematic. But, Rasmussen and Gallup are also providing interpretation of the data and making calls about which way states will swing. And they were pretty wrong about several swing states.

 

I certainly will be paying attention to his forecasts in the future. I sincerely had not heard of the guy prior to this election but you can't argue with facts.

The man had it right.

 

I started paying attention after '08, but still thought there could have been luck involved, as well. 2010 put paid to that idea. His numbers were what I followed during the run up to the election. The guy is uncanny!

 

Shouldn't the pollsters get the credit since the RCP average called 49/50 states correct?

Silver just poached their data and got the same result.

 

Nope.

 

Why aren't you posting tweets anymore Commodore?

 

Yes, but he uses their historical accuracy to give them each a different weight. For example, if Rasmussen tends to be inaccurate, but PPP tends to be more accurate, more weight would be given to PPP polling than Rasmussen. He still factors in Rasmussen since it is still a valid poll, but it's an understanding by looking in past history that he shouldn't give it as much credence.

It's called Bayesian forecasting (we do it at my company) and it's pretty damn awesome. It eliminates poll bias (in this case) quite effectively by looking at the historical accuracy. Once you figure out the weights, which are really the most crucial differentiator between his models and the typical ones, the rest is simply statistics and probabilities.

 

which pollsters who got all 50 states correct?

 

Well, I don't think there's a pollster that bothers to poll all 50 states because a lot aren't competitive.

That said, PPP was the only pollster I could find that got all the competitive states correct. YouGov had Romney +1 in Florida but got everything else right. In the November polls, PPP, Mellman, and NBC/Marist got Florida right but only PPP did every other swing state. Mellman only did a handful of other states but it got right what it did.

 

Garbage in.......garbage out.