Virginia 2009 Megathread

[Umengus]

How large was each group?

Rasmussen doesn't release those numbers.

Big shocker

Since the Presidential Election I haven't been able to find Rasmussen group numbers.  Anyone who trusts them is making a big mistake.  Some of there polls are far outliers to the right  like the Presidential Approval rating.  They've had Obamas approval rating tiedwith his dissaproval rating for 2 monhs.  All other polls ahve Obama's approval rating 8-24% above the dissaproval rating.  That means that they have at least a 16% difference with all polls.

Rasmussen polls Likely voters, others firms poll adults. It's a reason to explain the difference (at least 3 points, maybe plus). Rass push also the undecideds.

CBS and ipsos polls are joke because the party id of their poll is not conform to the reality (euphemism).

Idem for gallup poll wich was terrible in 2008.

Some polls institutes confirm +- rasmussen: zogby (ok zogby is bad), R2000 (bad also) democracy corps (good reputation), susa (idem), PPP.

Conclusion: Rasmussen is not an outlier and historic is there to confirm this conclusion (sorry it it's not conform with your liberal biais)

[Umengus]

Party ID:

D: 37 %
R: 29 %
other: 34 %

what a joke... who thinks that it will be the party id of the election day ? it's manipulation to make good story for medias.

If with a D+8 sample deeds is behind, he is in very bad position.

[Umengus]

The party id of the susa poll:

R: 37 %
D: 32 %
I: 29 %

My opinion is that the race is pretty stable for one month. The changes are due to pollsters...

McDonnell can not lose if he wins the independent vote by great margin and all polls say that it's the case for now.

[Umengus]

First, Rasmussen is a very good survey research firm, whereas Quinnipiac is a mediocre pollster and  R2K is a lousy pollster.

Second,  timing is very important.  The latter poll I cited had time for the public to react to Deeds equivocation on increasing taxes.

Third, no, most Democrat candidates are opposed to choice, except for abortion. 

Fourth, so are you saying you will support any candidate nominated on the Democrat party line irrespective of his/her position on abortion?

Your full of it.   Republicans only think Rasmussen is good because they show more positive results than other polls.   That's why Don Rasmussen gets air time on Fox.  They are no better than any other polling station.  Actually I"ve seen CNN do better than them on some elections. 

Okay, now that that is out of the way.  RAsmussen just came out with a poll showing Mcdonnel up by 9 so I don't think Survey USA is too much of an outlier. 

I think Demcorats are for for choice and the REpublicans are against choice. 

I don't vote on abortion.  ACctually I"m on the fence when it comes to that issue.  By the way, if Republicans support choice, why are they against a big government law like the DEfense of Marriage Act, which takes away the choice of gay Americans to get married and why are they against gays joing the military?  Doesn't sould like they are the party of choice.
[/quote]

Historic proves that Rasmussen is a top pollster: cfr 2004 and 2008 (at national level). Even top pollster can make mistakes.

[Umengus]

Umengus,

First, Rasmussen has a proven track record.

Second, CNN retains ORC to do polls for it (it does not do its own surveys).  Admission: I worked for ORC many years ago.

Third, I adressed my point about abortion to another poster whose moniker implied he was for life but supported Deeds.

Dem4Life

PEW commissions Princeton Associates to do its polls.  Not suprisingly, many of the people at Princeton Associates are former ORC employees.

To both of you,

The key Virginia poll (other than the actual election) to watch is Mason-Dixon.

I'm completely agree with you. Due to a technical mistake, only the last sentence of my post was made by me.

Ironically, the rasmussen poll with mcdonnell +2 was probably a outlier, considering the last ras poll.

[Umengus]

I'd say the most telling thing right now is that McDonnell has dumped the pretense of running a positive campaign and is going 100% on Generic Republican Attack Mode: "THE DEMOCRATS WILL RAISE YOUR TAXES!!!11". If he were really up by 14 like SUSA says, he wouldn't need to be doing that.

and deeds is doing a positive campaign... who has begun to be negative ?

and Mcdonnell is true: deeds wants to increase taxes and it's the good strategy to make: voters must know that.

[Umengus]

Umengus,

First, the prior Rasmussen survey you cited, which indicated a slight lead for McDonnell was probably an outlier.

Second, given the margin of error, the McDonnell lead could well have been five or more.

Third, remember that one time out of twenty the MoE will be exceeded.

Fourth, before Deeds put his foot in his mouth about taxes, there is some reason to believe that he was reducing McDonnell's lead.

Fifth, there is still more than a month until the election and unforseen events could significantly impact the race.

Finally, absent such unforseen events significantly helping Deeds, McDonnel is a likely winner.

the notion of the margin of error doesn't existe when polls are reweighted.

It would be interessant to reweighting all polls with a gop +2 advantage. And I think that the result will be always +- the same.

[Umengus]

The fact that McDonnell has now gone negative shows that he's worried; or at least he was worried when he decided to go negative.

In the end, I believe there's a good chance this race isn't as close as the polls show it to be. A 58-42 McDonnell win isn't out of the question. The Democrats should just be happy they will hold the Senate to redistricting next year. Once again, I will be happy to be proven wrong.

McDonnell goes "negative" because deeds did a strong mistake with tax increase. Even with a 20+ lead, you make the ad and you kill him.

It's sure for me that the major reason to explain Mcdonnell loss in some polls (wapo and ppp polls especially) is due to the modification in the party id samples (gop oversampled in the first time, dem oversampled in the second time). When you look at internals, McDonnell stays strong with independents. No move there. The increase of the dem turn out  is based on fantasy and is contradicted by the last polls..

[Umengus]

Based on the Exit poll:

- party id: Rep: 37 % dem: 33 % I: 30 %

No surprise...

- Obama job approval: 48 %

And not 57 % like wapo poll gave one week ago... it's very similar to the RAS daily track poll.

-full time working woman (28 %): have vote at only 51 % for McDOnnell. Probably the incidence of the thesis...

-16 % black. It's a little more than predicted. And Mc Donnell is weak with them: only 9 % of the vote. I would be curious to see the latino vote but the poll doesn't give this infirmation.

[Umengus]

Based on the Exit poll:

- party id: Rep: 37 % dem: 33 % I: 30 %

No surprise...

- Obama job approval: 48 %

And not 57 % like wapo poll gave one week ago... it's very similar to the RAS daily track poll.

-full time working woman (28 %): have vote at only 51 % for McDOnnell. Probably the incidence of the thesis...

-16 % black. It's a little more than predicted. And Mc Donnell is weak with them: only 9 % of the vote. I would be curious to see the latino vote but the poll doesn't give this infirmation.

The Rasmussen tracker is not the poll you want to use.  Rather, it's SUSA which nailed this race to a tee.

That being said, as I noted above, I see probably a couple of points swing to Republicans in Virginia against the national mean.

I just wanted to say that the job approval in Virginia matches with the rasmussen tracking poll. (48 % at national  level, 48 % in VA). And so the ras tracking poll is very accurate.

But indeed, survey USA did a very good job. I'm more nuanced with PPP, considering the great variation of the party id of their samples during the campaign. 

[Umengus]

Based on the Exit poll:

- party id: Rep: 37 % dem: 33 % I: 30 %

No surprise...

- Obama job approval: 48 %

And not 57 % like wapo poll gave one week ago... it's very similar to the RAS daily track poll.

-full time working woman (28 %): have vote at only 51 % for McDOnnell. Probably the incidence of the thesis...

-16 % black. It's a little more than predicted. And Mc Donnell is weak with them: only 9 % of the vote. I would be curious to see the latino vote but the poll doesn't give this infirmation.

Also worth noting that the reported presidential vote of VA voters yesterday according to that same exit poll was 51-43 McCain. Obviously a lot of Obama voters were not inspired enough by Deeds (imagine that) to come out and vote this year. Anyone assumes those same voters will stay home when Obama is back on the ballot in 2012 does so at their own peril.

Also HIGHLY worth noting is that 1 in 5 McDonnell voters approve of the job Obama is doing, compared to only 1 in 20 Deeds voters who disapprove of the President's job. (The numbers for NJ, FWIW, indicate over 1 in 4 voters who approve of Obama's job performance didn't vote for Corzine.)

One of the most insightful comments I've read on the forum is Spade's note that all 50 governors offices can be considered "open seats" in terms of party competitiveness. Federal races obviously are much, MUCH different in that regard. At the risk of over-extrapolating too much in the other direction of the MSM's common wisdom this morning, I see Owens's upset win in NY-23 as far more indicative of 2010 midterms than 2 governors races (but still admittedly only a tiny glimpse).

vote in 2008 is not a good indicator in my opinion. I prefer largely party id. It's clear for me that the party id in 2010 (2012 is different) will be +- similar to the party id in 2009 and hence, gop will win big in VA..

[Umengus]

Concerning NY 23, Hoffman with the "R" next to his name would have win without problem. And Owens, a blue dog democrat, has won without to get 50 % + (in 2008, Obama won with 52 %.

In ca-10, the win of the dem is so weak compared to 2008 et 2006.