PDA

View Full Version : General Nate Silver Backtracking on Twitter




RonPaulFanInGA
11-04-2012, 01:14 PM
https://twitter.com/fivethirtyeight/status/265092500291739648


We have Obama as ~80% likely to win Electoral College if popular vote is a tie. 98% if it's O+1. 30% if it's R+1.

So it drops almost fifty points if the national electorate moves one whole point, from a tie to R+1? Gallup, Rasmussen and Pew all have the national electorate as R+1 to R+3.

69360
11-04-2012, 01:34 PM
I think Silver is realizing he screwed up with his margin of victory for Obama. It's really close. I give FL, NC and VA to Romney. I think CO and NV are likely Romney. It should come down to OH, MI and PA deciding the election. I still think Obama by less than 10 EC votes.

sailingaway
11-04-2012, 02:25 PM
https://twitter.com/fivethirtyeight/status/265092500291739648



So it drops almost fifty points if the national electorate moves one whole point, from a tie to R+1? Gallup, Rasmussen and Pew all have the national electorate as R+1 to R+3.

wow, that really IS a back track. I've been reading his blog and he has been very comfortable with the electoral college for Obama up to this point.

MozoVote
11-04-2012, 02:41 PM
Adjusting his prediction model down from 85% to 80% is still pretty strong in Obama's favor though.

I don't see Romney winning VA. Not with the weight of the DC suburbs against him, and Johnson + Goode + RNC backlash from people who lean towards Paul or know Mr Blackwell.

sailingaway
11-04-2012, 02:48 PM
Adjusting his prediction model down from 85% to 80% is still pretty strong in Obama's favor though.

I don't see Romney winning VA. Not with the weight of the DC suburbs against him, and Johnson + Goode + RNC backlash from people who lean towards Paul or know Mr Blackwell.

30% if the popular vote is R+1 though is a big deal. My understanding is the popular vote is very much in question. I'm just speaking to his shift of position, not who will win, though.

RonPaulFanInGA
11-04-2012, 02:49 PM
Adjusting his prediction model down from 85% to 80% is still pretty strong in Obama's favor though.

I don't see Romney winning VA. Not with the weight of the DC suburbs against him, and Johnson + Goode + RNC backlash from people who lean towards Paul or know Mr Blackwell.

Pfft, Romney will win Virginia, Florida and North Carolina. His problem though is that he needs another high electoral vote state, and OH/PA/WI/MI/MN all look pretty iffy, at best.

MozoVote
11-04-2012, 02:59 PM
I assume Romney will win NC like everyone else does, since it was such a narrow Obama win 4 years ago. But for what it's worth, there was a big Democratic surge here in Mecklenburg County yesterday, on the last day of early voting. The Dem ground game is still active.

I doubt Romney will break 40% in Mecklenburg but he might do a little better than McCain.

thoughtomator
11-04-2012, 03:05 PM
I think all the predictions are going to be way off this year. None of their polling representation assumptions work in the age of ubiquitous cellphones, and traditional polling techniques skew to Democrats to begin with.

For example, Virginia is listed as a tossup. I don't think Obama has a prayer in hell here; they're not sinking resources in to win the state, that's for sure. My mailbox has a Romney flier in it daily; I get 5 pro-Romney calls on my phone per day compared to one (tops) for Obama. Romney has the support of a multitude of outside groups which are pressing the point hard, whereas Obama's campaign seems to be the only organization on the Democratic side. When the Kaine campaign (Dem Senate nominee) reaches out to me, they don't mention Obama at all.

Clearly one side thinks this state is winnable and worth an investment of significant resources, and the other does not. Yet the state is called a toss-up? Something has to be wrong with the public polling, Obama's private polling must be telling him that VA is a lost cause or Dem forces would be hitting the state hard.

Nate Silver really stepped in it with his over-the-top Obama boosterism (I do not for a minute believe that his highly-favorable-to-Obama reporting is innocent nor accidental). I'm not surprised he's backtracking now, as to persist further would be a reputation killer.

brandon
11-04-2012, 03:18 PM
What did he backtrack on? His website is showing Obama with 85% chance to win now.

RonPaulFanInGA
11-04-2012, 04:05 PM
What did he backtrack on? His website is showing Obama with 85% chance to win now.

But 30% if R+1. And probably about 0% if R+2. There's a better chance of that happening than 15%. Silver has become nothing but a comfort blanket for those who want an Obama win to cling to.

The New York Times stands to lose a lot if Romney wins. Silver and Krugman have gone all-in on the 'Obama will win' thing. They'll have no credibility left if Obama loses.

MozoVote
11-04-2012, 06:00 PM
Silver was pretty good with his calls in 2008 on Obama and 2010 on the Tea Party wave. I don't think he's a shill for Obama.

ronpaulfollower999
11-04-2012, 06:08 PM
But 30% if R+1. And probably about 0% if R+2. There's a better chance of that happening than 15%. Silver has become nothing but a comfort blanket for those who want an Obama win to cling to.

The New York Times stands to lose a lot if Romney wins. Silver and Krugman have gone all-in on the 'Obama will win' thing. They'll have no credibility left if Obama loses.

Well, TBH, Krugman really doesn't have any credibility left if Obama wins with 538 electoral votes either.

supermario21
11-04-2012, 08:36 PM
Silver's problem is that he has invested himself too much into the polling. His science and statistical analysis isn't wrong, it's just that he might be analyzing junk data. I think most of the polls dramatically overstate the electorate in favor of Democrats. And with regards to Ohio, one of the reasons O won was because he did well in SE Ohio, Cincy, and Cleveland. Turnout is already drastically down in EV in Cuyahoga County, the War on Coal has made Obama extremely unpopular in the SE, and the evangelicals are back on the ground this time for Romney, after many stayed home 4 years ago.

devil21
11-05-2012, 05:26 AM
But 30% if R+1. And probably about 0% if R+2. There's a better chance of that happening than 15%. Silver has become nothing but a comfort blanket for those who want an Obama win to cling to.

The New York Times stands to lose a lot if Romney wins. Silver and Krugman have gone all-in on the 'Obama will win' thing. They'll have no credibility left if Obama loses.

Both party mouthpieces are engaging in this stuff leading up to the election. You must not have read Rove's and Morris' "Romney will win in a landslide" pieces, among others.

fletcher
11-05-2012, 08:57 AM
https://twitter.com/fivethirtyeight/status/265092500291739648



So it drops almost fifty points if the national electorate moves one whole point, from a tie to R+1? Gallup, Rasmussen and Pew all have the national electorate as R+1 to R+3.

You aren't understanding what he is saying. He isn't talking about the makeup of the national electorate. He is talking about the actual popular vote. If the popular vote is a tie, there is an 80% chance that Obama wins the electoral vote. If Romney wins the popular vote by 1%, there is still a 30% chance that Obama wins the electoral vote. He isn't backtracking at all.

supermario21
11-05-2012, 10:50 PM
Actually Rove isn't predicting a landslide. I think he has Romney at 285.

devil21
11-06-2012, 12:56 AM
Actually Rove isn't predicting a landslide. I think he has Romney at 285.

Krauthammer then. Just swap in one neocon for another. Does it matter?

brandon
11-07-2012, 07:16 AM
Nate silver called it perfectly. He picked every single state correctly.

http://www.washingtonpost.com/blogs/ezra-klein/wp/2012/11/05/pundit-accountability-the-official-2012-election-prediction-thread/