NJ: Parties Wonder Which Side's Polls Reflect Reality

[nonniey]

Personnally my money would be on the Gallup and Rassmussen models.

http://www.nationaljournal.com/colum...ality-20121101

"......There's no question that President Obama's 2008 campaign, which focused on turning out new low-propensity voters in minority communities, helped inflate nonwhite voters' influence in the electorate. In Virginia alone, the nonwhite share of the electorate spiked from 21 percent in 2006 to 30 percent just two years later, a virtually unprecedented leap. The question is how many of those voters come back to the polls in 2012.

Republicans and Democrats alike believe the African-American vote is unlikely to change between 2008 and 2012. But they differ dramatically on the number of Hispanic voters who will show up at the polls—a key factor in critical battleground states like Colorado and Nevada. Republicans believe turnout will be down, depressed by Obama's failure to pursue immigration reform during his first term. Democrats think the booming number of Hispanic residents means their share of the electorate will only increase.

The same argument applies to younger voters. In 2008, 18 percent of the electorate was made up of voters between 18 and 29 years old. That's higher than the percentage has been in recent presidential years, when the youth vote has made up around 15 or 16 percent. Republicans believe the younger share of the electorate will slide slightly, and that Obama will win fewer of those voters anyway.

The manifestation of these disagreements is evident in polling weights. Most Republican pollsters are using something close to a 2008 turnout model, with the same percentage of white, black, and Hispanic voters as the electorate that first elected Obama. Most Democratic pollsters are a little more bullish on minority turnout, which helps explain some of the difference between the two sides......"

[skinsfan_1215]

Comparing 2006 turnout to 2008 is stupid... Presidential election vs non-Presidential election is apples to oranges.

I don't disagree turnout will be less for Obama than it was in 2008, but to what degree is what will decide the election.

Regardless, Rassmussen and Gallup are outliers right now. The results Tuesday will make some pollsters look really bad. The only question is who.

[Jumbo]

I see confirmation bias and it's cognitive cousins playing a very dominant role in most people's politics.

[Unforgiven]

I don't like really focusing on individual polling companies. I've lately been focusing more on people like Nate Silver who step back and look at everything and then crunch the numbers.

Personally, based on looking at all the different polling I think there are 3-4 states the Rassmussen lists as 'toss-up' states that should be filed under 'Leans Obama'. I think come Wednesday morning they won't be in the top 5 as far as accuracy goes.

Like it's been posted above, it's not fair to really compare off years to Presidential elections but Rassmussen in 2010 were pretty inaccurate and biased toward Republicans and I don't think they corrected for that since then.

[twa]

I'm looking forward to reality intruding, no matter it's flavor.


http://maggiesfarm.anotherdotcom.com/

unforgiven...how did Ras do in the POTUS race last time?
Nate plays games on weighing for state races in which polling in fewer states seems to increase his scores
it's like comparing GM to Telsa

[PeterMP]

Rassmussen's latest result has it an even tie, which seems right in line w/ most of the other polls.

http://www.realclearpolitics.com/epo...bama-1171.html

The lastest Gallup has it +5 Romney, but that was several days ago now and before the Hurricane, and I'm not sure what their margin of error is.

Nationally, all of the public polls are essentially saying the samething, except for MAYBE Gallup.

Rasmussen has the electoral college 237-206 for Obama (the rest too close to call).
CNN puts it at the SAME exact thing.

Gallup doesn't do an electoral map that I can find.

[twa]

Rassmussen's latest result has it an even tie, which seems right in line w/ most of the other polls.
.

yep, I don't understand why some claim it as biased
btw the MOE is usually 3 or less for Ras

They are working with a D+3 weighing(national)......should be interesting

[Unforgiven]

I'm not really concerned with national popular vote numbers. It's all about the electoral college. If we're talking strictly popular vote then Rassmussen is much closer to reality than Gallup.

If you look at their EC board Rassmussen has it tied up in Ohio and Wisconsin, Romney up by a point in Iowa, up by 2 in New Hampshire and up by 3 in Virginia and Colorado.

I'm aware of margins of error and all that but those numbers are out of line with the majority of pollsters. We'll have to wait till election day to see who was right and who was wrong, but those numbers just seem out of whack.

[PeterMP]

yep, I don't understand why some claim it as biased
btw the MOE is usually 3 or less for Ras

They are working with a D+3 weighing(national)......should be interesting

I've walked you through this before w/ respect looking at the data from the 2004 election.

Months and even weeks out, Rasmussen tends to be biased with respect to the other polls for the Republican. The last month or so, it changes (not by much), and then he ends up with the right answer the last week or two before the election.

Even a few weeks ago, he had Romney leading something like +7 (which was more than any other poll, including Gallup).

http://www.rasmussenreports.com/publ..._tracking_poll

Since then, it has been adjusted down and come in line with the other polls.

[nonniey]

I'm not really concerned with national popular vote numbers. It's all about the electoral college. If we're talking strictly popular vote then Rassmussen is much closer to reality than Gallup.

If you look at their EC board Rassmussen has it tied up in Ohio and Wisconsin, Romney up by a point in Iowa, up by 2 in New Hampshire and up by 3 in Virginia and Colorado.

I'm aware of margins of error and all that but those numbers are out of line with the majority of pollsters. We'll have to wait till election day to see who was right and who was wrong, but those numbers just seem out of whack.

Out of Whack? Only if you believe Obama is going to outperform his numbers from 2008.

[Unforgiven]

Out of Whack? Only if you believe Obama is going to outperform his numbers from 2008.

Not at all, my point was that on those states Rassmussen is 2-5 points more in favor of Romney in each state than the average poll.

Take Ohio for example, it's 49-49 according to Rassmussen while 95%+ of other pollsters in the past week+ have Obama up by 2-5 points. It's pretty much the same thing state by state, they're giving Romney 2-5 points per state more than the average. In the tossup states he has Romney winning by 2-3 points and in the states that lean Obama by 3-4 points in other polls he has them tied up or even +1 for Romney.

I realize these are within margins of error and he's not giving them to Romney but if someone that didn't know better looked at the Rassmussen EC map they would think this is a dead heat that Romney could easily win.

The eight states he lists as 'toss-up' states are Florida, Virginia, New Hampshire, Ohio, Wisconsin, Iowa, Nevada and Colorado. Out of those eight states only one single state does he give Obama better than a tie (Nevada). Five out of the remaining seven are listed as favoring Romney by 1-3 points. Again, yes...MoE.

If you look at Nate Silver's website he has Obama as the statistical favorite in every single one of those states with the exception of Florida.

[nonniey]

Not at all, my point was that on those states Rassmussen is 2-5 points more in favor of Romney in each state than the average poll.

Take Ohio for example, it's 49-49 according to Rassmussen while 95%+ of other pollsters in the past week+ have Obama up by 2-5 points. It's pretty much the same thing state by state, they're giving Romney 2-5 points per state more than the average. In the tossup states he has Romney winning by 2-3 points and in the states that lean Obama by 3-4 points in other polls he has them tied up or even +1 for Romney.

I realize these are within margins of error and he's not giving them to Romney but if someone that didn't know better looked at the Rassmussen EC map they would think this is a dead heat that Romney could easily win.

The eight states he lists as 'toss-up' states are Florida, Virginia, New Hampshire, Ohio, Wisconsin, Iowa, Nevada and Colorado. Out of those eight states only one single state does he give Obama better than a tie (Nevada). Five out of the remaining seven are listed as favoring Romney by 1-3 points. Again, yes...MoE.

If you look at Nate Silver's website he has Obama as the statistical favorite in every single one of those states with the exception of Florida.

Again you have to go back to original premise of the article. Who is right? The polls you and Silver reference assumes that Obama will out perform his 2008 numbers (all of them). If that premise is correct Obama will win and win the majority of those states. If the other premise is correct that he won't outperform his 2008 numbers then it comes down to Ohio as the toss-up. And if Gallup is correct and the Republicans outperform the Dems then Obama will lose.

[Duckus]

The article is right. On election night one side will be wrong and look silly.

The probabilities are on Obmama's side, but the Redskins were overwhelming favorites when we scored against the Giants late in the 4th. We know how that ended.

Romney has a chance. Right now stats says 1/5. We will see Tuesday.

---------- Post added November-3rd-2012 at 04:54 PM ----------

Again you have to go back to original premise of the article. Who is right? The polls you and Silver reference assumes that Obama will out perform his 2008 numbers (all of them). If that premise is correct Obama will win and win the majority of those states. If the other premise is correct that he won't outperform his 2008 numbers then it comes down to Ohio as the toss-up. And if Gallup is correct and the Republicans outperform the Dems then Obama will lose.

Nate's model actually accounts for the polls being off or "biased" (in a statistical way). He talks about it a lot. Hence he gives Romney a 18% chance. He calculates that 18% is roughly the chance the polls are all completely off. If polls were perfect, he would give Obama 100% chance.

[AsiaticSkinsFan]

If you look at Nate Silver's website he has Obama as the statistical favorite in every single one of those states with the exception of Florida.

woah, woah, woah.

The Right have decided that Silver's figures are bunk now because it doesnt agree that Romney is winning.

And Nate Silver is gay, so we cant take his opinion seriously.



[/sarcasm]



its the GOP's war on facts. Silver's numbers have no political bias, but the Right doesnt care.

[twa]

Months and even weeks out, Rasmussen tends to be biased with respect to the other polls for the Republican. The last month or so, it changes (not by much), and then he ends up with the right answer the last week or two before the election.

.

They use a likely voter model in almost all their polls, many of the others switch close to the election

They appear biased early since they weed out registered voters unlikely to vote

add
The only time I see a 7 point lead was in january, in the interim other polls have shown larger leads for Romney than Ras
http://www.realclearpolitics.com/epo...bama-1171.html