Rassmussen vs Nate Silver

[HomerJS]

Who will emerge as the most accurate? Fox News and the right have made a habit of always quoting Rassmussen claiming others underpoll Republicans. It will be interesting

As of now popular vote...

Rasmussen

Romney 49%
Obama 48%
Others 2%

Nate Silver

Obama 50.8%
Romney 48.3%

 

[Todd33]

Rassmussen is what is is so it can be quoted by Fox News, the cart and the horse IMO.

 

[hal2kilo]

It a race between Rasmussen and Gallup for the tard award this year.

 

[buckshot24]

They are all short on the Romney margin of victory.

 

[First]

Fox will likely continue voting Rasmussen despite their middling-to-poor polling accuracy.

 

[AstroManLuca]

Rasmussen has Romney winning CO, IA, and VA, with WI and OH as exact 49-49 ties.

Rasmussen is also the only polling firm that has recorded anything other than a lead (albeit a slight one) for Obama in Ohio since mid-October.

http://www.realclearpolitics.com/epolls/2012/president/oh/ohio_romney_vs_obama-1860.html#polls

They called it a tie 3 times in the past month and gave Romney a slight lead a week ago. Every other poll has Obama up by at least a small amount.

Virginia appears to be the closest, with RCP averaging out the polls to give Obama just a 0.3% lead. Anyone's game there.

Wisconsin has been reported as slightly in favor of Obama in every poll on RCP except Rasmussen, which is calling it a tie. The RCP average is Obama +4.2.

Iowa and Colorado appear to be close. More polls are showing Obama with lead than Romney, but it's not like Ohio and Wisconsin where Rasmussen is completely alone.

But even if you give Colorado, Iowa, and Virginia to Romney, it won't be enough. If Obama takes the consistently blue (except for Rasmussen) states of Ohio and Wisconsin, he wins anyway, by a narrow 275-263 margin.

 

[buckshot24]

Fox will likely continue voting Rasmussen despite their middling-to-poor polling accuracy.

52-46 vs 53-46 was such a miss in 2008. Obvious hack.

 

[Charles Kozierok]

Silver is an analyst, not a pollster. It would be more accurately titled "Rasmussen/Gallup versus Everyone Else".

And the "popular vote" number is both meaningless and irrelevant. What matters are the states, where Rasmussen is an outlier, and where he will likely to be shown to be just as wrong as he was in 2008.

 

[AstroManLuca]

Nate Silver has actually discussed Rasmussen several times on his blog, and it's not all negative. He tends to hold them up as a better-than-average pollster, and the accusations of bias against them are often overblown. I do find it hard to believe that Wisconsin is a tie, against what every other polling firm is predicting. We won't know until tonight or tomorrow.

There have been a few cases where they were way off in favor of the Republican. Notably, a senate race in Hawaii where Rasmussen called a 13% lead for the Democratic incumbent and in reality wound up with a 53% lead.

EDIT: Also I agree, state polls are more important. Popular vote doesn't matter. I'd be interested to see the final Rasmussen state-by-state predictions for 2008, but I can't find them, only their prediction of the national popular vote (which was very close).

 

[First]

Silver is an analyst, not a pollster. It would be more accurately titled "Rasmussen/Gallup versus Everyone Else".

And the "popular vote" number is both meaningless and irrelevant. What matters are the states, where Rasmussen is an outlier, and where he will likely to be shown to be just as wrong as he was in 2008.

Yup, see here for those interested in documented Rasmussen bias/inaccuracies. They were particularly bad in 2010, but also off 1.2% on the national level in 2008 (12th overall, which is middling, not bad) but worse on the state level that year, particularly Ohio where their final dozen polls or so had McCain winning slightly, but ended up being off by nearly six points! Every other aggregated poll was far more accurate for Ohio that year, btw.

 

[werepossum]

One thing no one on the left has addressed is the outsized Democrat representation in all polls except Rasmussen and Gallup. In an election where everyone agrees Obama has less support than 2008, how does one justify assuming a heavier Democrat voting base? This is especially true since early voting has been up 14% in counties McCain won and down 4% in counties that went Obama?

Also, I think every pollster has Romney winning Ohio's independents, whereas McCain lost them. How then is Obama winning?

I suspect that Romney will be ahead in Ohio tonight when the polls nominally close, but Obama will pull it out on provisional ballots and after-hours votes from heavily black districts. It's the wave of the future.

 

[First]

One thing no one on the left has addressed is the outsized Democrat representation in all polls except Rasmussen and Gallup. In an election where everyone agrees Obama has less support than 2008, how does one justify assuming a heavier Democrat voting base? This is especially true since early voting has been up 14% in counties McCain won and down 4% in counties that went Obama?

Also, I think every pollster has Romney winning Ohio's independents, whereas McCain lost them. How then is Obama winning?

I suspect that Romney will be ahead in Ohio tonight when the polls nominally close, but Obama will pull it out on provisional ballots and after-hours votes from heavily black districts. It's the wave of the future.

You can't and shouldn't adjust party ID, it's merely a snapshot representation of the electorate's willingness to identify themselves as D, R or I. Their models say nothing of the likelihood of those D/R/I %'s actually being that way on election day. What matters more than anything else is actual aggregated poll results, as the sampling error is very low with all that extra data.

 

[Charles Kozierok]

One thing no one on the left has addressed is the outsized Democrat representation in all polls except Rasmussen and Gallup. In an election where everyone agrees Obama has less support than 2008, how does one justify assuming a heavier Democrat voting base? This is especially true since early voting has been up 14% in counties McCain won and down 4% in counties that went Obama?

I don't see an "outsized Democrat representation" in most of the other polls.

I do see Rasmussen predicting an R/D split far more favorable to the GOP than the 2010 election was, which is beyond insane.

Also, I think every pollster has Romney winning Ohio's independents, whereas McCain lost them. How then is Obama winning?

I don't think that's actually the case.

I suspect that Romney will be ahead in Ohio tonight when the polls nominally close, but Obama will pull it out on provisional ballots and after-hours votes from heavily black districts. It's the wave of the future.

Urban areas nearly always report later than rural ones. Happened during the GOP primaries as well.

 

[AstroManLuca]

Ah, found it:

http://www.rasmussenreports.com/pub...ohio/election_2008_ohio_presidential_election

The day before the election, Rasmussen had Ohio as an exact 49-49 tie (sound familiar?), but Obama ended up winning 52-47.

There were other states where Rasmussen may have called the winner, but mistakenly predicted it to be much closer than it ended up being.

Nevada: Prediction 50-46, actual 55-43 (8% difference in spread)
Colorado: Prediction 51-47, actual 54-45 (5% difference in spread)
Wisconsin: Prediction 51-44, actual 56-42 (7% difference in spread)

I'm not saying they'll be as off this time around but it wouldn't surprise me. This is not the same electorate that elected Obama in 2008, but all indications are pointing to a relatively comfortable win.

 

[Charles Kozierok]

The day before the election, Rasmussen had Ohio as an exact 49-49 tie (sound familiar?), but Obama ended up winning 52-47.

And note how many prior polls had McCain in the lead.

Rasmussen is a hack.

 

[First]

Ah, found it:

http://www.rasmussenreports.com/pub...ohio/election_2008_ohio_presidential_election

The day before the election, Rasmussen had Ohio as an exact 49-49 tie (sound familiar?), but Obama ended up winning 52-47.

There were other states where Rasmussen may have called the winner, but mistakenly predicted it to be much closer than it ended up being.

Nevada: Prediction 50-46, actual 55-43 (8% difference in spread)
Colorado: Prediction 51-47, actual 54-45 (5% difference in spread)
Wisconsin: Prediction 51-44, actual 56-42 (7% difference in spread)

I'm not saying they'll be as off this time around but it wouldn't surprise me. This is not the same electorate that elected Obama in 2008, but all indications are pointing to a relatively comfortable win.

Wow, that's an epic sort of bad. I knew intuitively they were bad based on the daily polls I've watched for a while now, Rasmussen always favoring red being the common theme. Yet more evidence they need to be taken in context for what they are.

 

[buckshot24]

I don't see an "outsized Democrat representation" in most of the other polls.

I do see Rasmussen predicting an R/D split far more favorable to the GOP than the 2010 election was, which is beyond insane.

He doesn't use the +6 R model his party affiliation poll is showing in his polls. He uses a +2 D model for his national tracking poll.

 

[Charles Kozierok]

He doesn't use the +6 R model his party affiliation poll is showing in his polls. He uses a +2 D model for his national tracking poll.

Not sure where you got this from. But if true, it doesn't exactly make a lot of sense.

 

[Moonbeam]

What does it matter how the people vote when it's how the vote is counted that matters. How many key states have Republicans in charge of the elections....

It didn't matter at all that Florida voted for Gore.

 

[HomerJS]

Ah, found it:

http://www.rasmussenreports.com/pub...ohio/election_2008_ohio_presidential_election

The day before the election, Rasmussen had Ohio as an exact 49-49 tie (sound familiar?), but Obama ended up winning 52-47.

There were other states where Rasmussen may have called the winner, but mistakenly predicted it to be much closer than it ended up being.

Nevada: Prediction 50-46, actual 55-43 (8% difference in spread)
Colorado: Prediction 51-47, actual 54-45 (5% difference in spread)
Wisconsin: Prediction 51-44, actual 56-42 (7% difference in spread)

I'm not saying they'll be as off this time around but it wouldn't surprise me. This is not the same electorate that elected Obama in 2008, but all indications are pointing to a relatively comfortable win.

but......but.....but....fair and balanced??!!??

 

[buckshot24]

Not sure where you got this from. But if true, it doesn't exactly make a lot of sense.

No, not really.

 

[lotus503]

One thing no one on the left has addressed is the outsized Democrat representation in all polls except Rasmussen and Gallup. In an election where everyone agrees Obama has less support than 2008, how does one justify assuming a heavier Democrat voting base? This is especially true since early voting has been up 14% in counties McCain won and down 4% in counties that went Obama?

Also, I think every pollster has Romney winning Ohio's independents, whereas McCain lost them. How then is Obama winning?

I suspect that Romney will be ahead in Ohio tonight when the polls nominally close, but Obama will pull it out on provisional ballots and after-hours votes from heavily black districts. It's the wave of the future.

Couple of reasons

Republicans don't like Romney
Democrats don't like Romney

 

[buckshot24]

Silver is an analyst, not a pollster. It would be more accurately titled "Rasmussen/Gallup versus Everyone Else".

And the "popular vote" number is both meaningless and irrelevant. What matters are the states, where Rasmussen is an outlier, and where he will likely to be shown to be just as wrong as he was in 2008.

The popular vote is not meaningless. If one candidate wins by 2 points nationally they will almost always win the electoral college. If somebody wins by 10 points then they are guaranteed the EC votes to win. Of course you know this but you keep saying that it is meaningless when it is not.

 

[First]

lol

 

[Charles Kozierok]

The popular vote is not meaningless. If one candidate wins by 2 points nationally they will almost always win the electoral college. If somebody wins by 10 points then they are guaranteed the EC votes to win. Of course you know this but you keep saying that it is meaningless when it is not.

It's meaningless. Not only is it less predictive than looking at the state results that come out at the same time, it confuses people into drawing conclusions that are usually invalid.

And to top it off, what people call the "national popular vote" really isn't an accurate representation of what that would really be in an election conducted on that basis.