Reid leads Angle by 11 points 43% to 32%

[grantcart]

That's right the screaming headlines show Angle up 2 points.

Oh the media wants a horse race.

So faced with a Reid landslide in registered voters they apply standards to show a tied race among 'likely voters'.

Currently Angle is supported by only 32% of Nevada voters. Its unlikely that will climb much.

Is it POSSIBLE that she will win. Of course. If people get lazy.

But the Democrats and the union GOTV phone campaign hasn't even started yet.

So when the polls show a result that is difficult to believe read a little deeper.

In Nevada, 42% of likely voters say they would vote for Angle, 40% say they would vote for Reid, and 7% say they would vote for third party candidate Scott Ashjian. If Ashjian were out of the race, Angle maintains the same two point lead, 47% to 45% over Reid. Reid's biggest problem appears to be voters' enthusiasm. He holds a strong lead over Angle among registered voters, who support him 43% to 32%. That represents a 13-point deficit attributable either to disconsolate Democrats or motivated Republicans ahead of November's election.

Read more: http://www.time.com/time/politics/article/0,8599,2024012,00.html#ixzz11t5dEdoN

Of course Time fails to explain that the other possible explanation of a 13 point difference between 'registered' and 'likely' voters - a really high margin of difference - could be their polling techniques and projection of who is really 'likely' to vote.

[Tatiana]

1. All these polls are oversampling conservative voters

and undersampling moderates and liberals. On election day, I think we will see a surprisingly strong showing on the part of Democratic registered voters.

[Arkana]

2. So if those "unlikely" voters show up to vote on November 2

then Reid will win by like 10 and Angle's Idiot Brigade will be wondering what happened.

[Qutzupalotl]

19. Right. So we have to motivate our side, and a good way to do it is

by letting Sharon Angle speak for her crazyass self, and ask voters: is this the kind of representation you want?

Fear works, especially if the boogeyman is real and on video.

[Zax2me]

3. And the good news is this applies to all races

Where reps are "leading" in polls. We are being fed a lot of speculation, inaccurate speculation. Not complete polling results as your link article spells out.
Not the first time, either.

[Mefistofeles]

4. So you predict Reid will beat Angle by approximately 11%?

?

[grantcart]

8. Of course not. Try and not be obtuse.

The Time article notes that a 13 point disparity is unusual but they don't ask the obvious point - that their methodology in paring down from registered to likely is faulty.

It is particularly faulty in a state where SEIU has one of the best GOTV machines in the country.

Now here is my question to you. When was the last time a candidate won a state wide election with only 32% of the registered voters.

[Mefistofeles]

9. You think "SEIU has great GOTV" is serious criticism against poll data?

Interesting.

Now, let's see. You and I know that final results traditionally disfavor the democratic party relative to RV polls. That's why very few pollsters cling to RV as the election approaches. So based on that, you and I agree that the margin will be less than Reid +11. Correct? I think it will be a nail biter, with the winner winning by 1-3%. What do you think?

Obviously you think it's not close, as you criticized the media for concocting a horse race. If you don't think it's a horse race, then how far ahead of Angle do you think Reid is, approximately? 5%? 8%?

[grantcart]

10. Your reading comprehension is fractured, and you take statements out of context

to make absurd conclusions that have nothing to do with my OP.

For example your statement "So you predict Reid will beat Angle by approximately 11%?"


I am not predicting who is winning and I even said that Angle might win.


The OP isn't about predicting who is going to win the election.


The OP is about the media using polling results to create headlines that are not consistent with the results of the poll.


The most accurate headline for this poll is

Reid ahead in polls but election will be determined by who gets the vote out.

Now for SEIU's impact on the election.

Half of Nevada's population lives in Clark County and 40% of that population is non white.

SEIU and the Culinary Unions have 100,000 members in Clark County. So yes if the poll has under represented non whites in Clark County it is seriously under representing 'likely voters'. Does this mean that they will get the vote? Of course not. But where the SEIU has gotten energized it has been very effective in increasing the non white vote turn out. Clark County is particularly good ground for GOTV because the unions are so concentrated on the 'strip' making it easier for union organizers to have an impact. Such details don't impact national polls but they are significant in jurisdictions with relatively small populations, like Nevada.

[Mefistofeles]

13. Unlike you, pollsters are aware of our disadvantage when it comes to getting out the vote

In midterm elections. Pollsters don't sit and wait until people show up. Voters traditionally behave in certain ways during midterms. When a Democrat is President they behave one way. When a Republican is President they behave a different way. And am I wrong to state that you believe the SEIU's awesome GOTV will result in an advantage for Democrats?

[grantcart]

17. So you think a poll that shows no liberals or non whites as likely

voters is a reliable poll?


So research the poll and tell me how many 'non whites' and 'liberals' were included in this poll?


http://i2.cdn.turner.com/cnn/2010/images/10/06/topstate5.pdf


So they don't give the tabs on registered voters.

They don't break down the tabs on the overall sampling.

They admit that 13% discrepancy between registered and likely voters is a huge discrepancy.

They register NO non whites in their Nevada tabs.

They register NO liberals in their Nevada tabs.

And you still defend their results which conveniently help their meme that it is an absolute horse race which helps their ratings.





Now you may have confidence that pollsters are able to figure out who 'likely' voters are, but the pollsters don't share that confidence:


http://www.huffingtonpost.com/2010/10/05/likely-voters-how-pollsters-choose-them_n_751560.html

Needless to say, those numbers can't all be right. The enormous variation has left a lot of reporters and readers emailing to ask, just how do pollsters identify a likely voter? More important, does anyone know what the heck a likely voter is anymore?

That question is arguably the most important polling story of the year. The answer is not obvious, but the basic issue is simple. There were roughly 213 million eligible adults in 2008, but only about 68% of those adults told the U.S. Census that they were registered to vote and only 62% turned out to vote (up from 60% in 2004). More important, turnout is typically much lower in midterm elections. The turnout among eligible adults was 40% in both 2006 and 2002.

So conceptually, this is a simple problem: It doesn't make sense to interview a sample that represents all registered voters, when somewhere between a third and half of those who say they are registered voters will not vote.

The hard part is figuring out who the true "likely voters" really are. Part of the problem, of course, is that some of those who will actually vote don't know yet that they will (and vice versa). To paraphrase my friend, political scientist Joel Bloom, the 2010 electorate is "a population that technically does not yet exist." It is, rather "in the process of becoming one."




Pollsters also will concede that the smaller the population and the smaller the sampling and the less frequent the polling company has conducted the same poll will also undermine results.


This is CNN's only poll of a small population and sampling of only 789.

[Imajika]

5. Likely voter screens are always used in polling...

...and they are fairly accurate coming from reputable polling organizations. This is nothing new. The same methodology was used in 2006 and 2008, only then it showed Democrats much more likely to turn out. I've no doubt the right wingers were busy decrying the poll results then too. I remember the likes of Hannity and Limbaugh claiming the polls in 2006 and 2008 were just propaganda, released only to depress GOP turnout, etc. They just did not want to accept the results.

The arguments over polling on DU in this cycle is no different, it's just reversed. Many people just don't like the results so they choose to attack the messenger.

Unlikely voters don't turn out to vote. They just don't. You can hope they will. You can wish they will. But they simply don't vote. As we get closer to the election there will be some natural poll tightening, and some of those unlikely voters will become likely voters - but there just doesn't seem to be much doubt about the fact that the right wing is extremely fired up for this cycle. Even if we achieve a 50/50 split on election day, they will still pick up a huge number of seats since we have so many more than them right now.

[grantcart]

12. That's not the issue.

No one is arguing that there shouldn't be 'likely voter screens'.


The point here is that the disparity between them here is so unusual that the media shouldn't be leading headlines all over the place that polls show a tied race.


The article admits that 13 points is an unusual spread between likely and registered.


It is also known that polls tend to under estimate non whites in lower incomes.


SEIU and the Culinary Unions have effected high voter turnout in the past - if the poll doesn't factor that in then that alone could account for such a large discrepancy between likely voter and registered voters.

[Imajika]

14. Yeah, I'm not buying it...

Your OP reads as if Reid were really up 13 points and that Angle isn't actually tied or ahead with likely voters.

Most polls show this race tied, some with Angle up very slightly.

Reid is apparently pretty much hated in Nevada. I think the likely voters screens are probably pretty accurate. I have no idea who will win, but I am confident Reid will not win by 13. If Reid pulls this out he probably wins by no more than 1 or 2 points - which is pretty much inline with most polling likely voter screens.

[grantcart]

16. Your correct that I showed the results of the registered poll

And no I don't think that it represents likely voters.


My point in doing so was to counter the multiple headlines and threads that show Angle up 2% in 'likely' voters but bury the registered voter results.

It would highly unusual if a candidate that was only supported by 32% of the registered voters won a contested statewide contest.

The tabs for the poll raise further questions.

They don't have any tabs for registered voters.

Tabs for 'non white' and 'liberal' show only N/A response, the only categories so affected. They have numbers for white, conservatives and moderates but none for liberals or non white.

http://i2.cdn.turner.com/cnn/2010/images/10/06/topstate5.pdf

Based on the poll results its hard to justify an Angle lead. I probably could been clearer in my OP. The subject line was, intentinaly misleading to shock the widely held assumption that polls showed conclusively that Angle has a clear lead.

[stevenleser]

21. Angle is also hated in Nevada. Both of their negative ratings are through the roof...

I normally defend polling techniques, and these may in fact be correct but there are a couple of additional things to consider.

1 - I am not sure that the disgust that many voters feel, particularly Democrats, with the crop of Republican wackjob candidates has been factored in. Sure, you might be less energized, but does that mean you will sit there and let an Angle or a Paladino or Buck get elected because of that? Would you let them be in the leadership of the senate because of that even if they are not the candidate in your district or state?

2 - The Likely voter polls can be viewed as a worst case scenario because it is hard to imagine that the disconnect between the full electorate and those who turnout could be worse. An even small dissonance in the likely voter model would produce a big difference in 8-10 states.

3 - The message you should give if you are trying to motivate voters to GOTV is exactly what I noted, we actually have more people who want the Democratic candidate in virtually every senate race, we just need to make sure Democrats get out and vote.

[Bitwit1234]

6. I still say they fudge the results by interviewing more republicans

and the excuse is republicans are the "likely" voters. I just don't trust Razzypoop's results.

[mzmolly]

7. The "likely" voter meme in this election is interesting isn't it?

How does Time etc. determine who falls into the likely voter category? Isn't a likely voter, generally someone who votes in every election? If so, wouldn't younger people automatically be excluded?

[BrklynLiberal]

11. oops

[SoCalDem]

15. Media loves the horserace angle (no pun intended)

and of course if there are "manipulation" of the electronic votes in "just the right places", coupled with LOW turnout, the public will "buy" a squeaker loss by Reid..

Portraying a race as razor-thin, also is a way to nudge recalcitrant voters TO the polls, so it cuts both ways.. If you think "your guy" is several points ahead, you might just not even vote, and if enough people feel the same way, the race could be lost too.

Reid is not the most charismatic person (understatement of the decade), and off-year elections skew older, so the tea party nuts may be more enthusiastic voters, so it's possible that she could win if enough dems stay home.

Republicans who don't like her, will probably vote anyway, even if it not to elect her...there are other things on the ballot too. They CAN vote "no thanks" to all candidates running for the senate.

All that said, if Harry squeaks by, I HOPE he is NOT re-elected as senate leader (if we maintain the majority)

[D23MIURG23]

20. This is completely off topic, but that swimming cat picture is great. n/t

[w4rma]

18. Great. Now lets work to keep Sen. Feingold in the Senate. (nt)

[applegrove]

22. Woohoo!

:woohoo: