Posted on 09/22/2008 1:51:32 PM PDT by 2ndDivisionVet
There seems to be quite a bit of gloom over a SurveyUSA poll of Virginia that puts Obama up by 6 percent.
(I note SUSA has Obama ahead, 49 percent to 45 percent in "Central Virginia," a sub-sample of about 200 voters.)
No offense to the fine folks at Survey USA, but I decided to go back and check their last poll in Virginia's 2006 Senate race. They had Jim Webb ahead of George Allen, 52-46. The final results were Webb's 49.6 percent to Allen's 49.2 percent, a margin of 9,329 votes out of 2.37 million cast. SUSA had the state's marriage amendment polling at 42 percent support,and it ended up passing with 57.1 percent of the vote.
I note this item on 538.com, talking about the upside of Obama's massive voter registration drive in the Commonwealth:
Obama campaign strategists believe that, with their massive months-long, grinding-it-out-every-day registration plan, that 80 percent of those new registrations would vote for Obama, and that 75% of the newly registered voters will turn out. If 75% of an 80-20 split on 300,000 new registrants turns out, that’s Barack Obama adding 135,000 bonus votes to his total in Virginia alone. Organizers in Obama’s Virginia campaign offices have been sternly instructed to focus on those numbers by spending long, exhausting days recruiting volunteers instead of spending their limited time worrying about whether there are enough yard signs to go around.
An additional 135,000 votes would indeed be impressive. But it is worth noting that Bush beat Kerry by 262,217 votes in 2004. And the above accounting takes into account that that other 20 percent, are McCain voters (okay, some would be third party) adding about 45,000 votes into the McCain (or non-Obama) pile.
We will hold it this time but the demographic trends longterm are bad for us in Va. NOVA is exploding in population and is blue. The hispanic population is exploding and they tend to go blue.
If it only takes one cheez-whiz poll to make you gloomy, you need to get a life.
“Central Virginia” is the code word for Richmond - a very black city. I do not feel that the numbers are indicative of the rest of the state at all - in my county, for example, we vote over 80% Repub... (And we’re only 50 miles or so north of Richmond)...
If by "fishy" you mean "flawed, uncertain, and inconsistent", then I have to agree with you.
By the way, the polls showing McCain up by several points are ALSO flawed, uncertain, and inconsistent.
There is one, and only one, poll that counts. It will be taken on November 4th of this year. Only "highly likely voters" will be polled, and its results will be dead accurate. I'm waiting for that one. Meanwhile, enjoy the show...
Obama’s assumption is that these new registrants will actually vote. I doubt the turnout for these people will be as high as expected.
This poll of a few hundred in a restricted geographic part of VA is not controlling of VA. It would be the equivalent of me saying that I have spoken to the VAST majority of my fellow club members (of a private club in New Orleans) and they overwhelmingly support X. Bad sample, bad information and certainly not a prevailing interpretation of how the vote will go.
To be clear on this, don’t look as much at the polls, especially the wild state polls, as what the campaigns are doing. McCain-Palin opened up 50 offices in CA for goodness sake! The only reaosn to make such a huge commitment is if your the micro-story on the ground was very favorable.
Obama closed out his offices in GA, which, by inference, means he can’t be doing as well in NC as some polls have him. Obama also closed out his offices in ND and sent the staff to WI. If he were as close in VA and FL as these polls and the voter registration anecdotes say, he would be in Richmond or Ft. Lauderdale rather than in Green Bay.
Something struck me yesterday when I looked at the internals from a Florida poll (showing McCain up only with 1%).
McCain had the lead among whites by 53 - 38, among Hispanics 48 -42 (IIRC), but he was down among blacks by (4 - 88).
Now, if you can decide your polling result by actually only looking at the prospective respondents (sure a lot of the polls are made via telephone) it is just to easy to skew your results. Obviously choosing black neighborhoods or not in your polling (direct interviews or telephone interviews) will probably determine the outcome.
Polling firms like Rasmussen and Gallup balance their samples (or mathematically correct them) according to party affiliation and gender (I assume). In this election they certainly must balance the samples also for race. Do they do that? Anyone who knows?
Virginia will go McCain, theres been a concerted effort over this last week to bring McCains percentages down, PPP is hugely oversampling and so are SUSA, don’t believe a word of these polls, i just posted a new poll from PPP on the threads has Obama up 11 in NM, however, they hoped we didnt notice they used a 52% dem sample compared to a 32% republican sample.
Yet he is ahead?
How many grave yards are there in Richmond? There may be enough dead voters to win the state for Obama.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.