Polls find Democrats ahead in 3 vital states (Mason-Dixon, DeWine -2, Burns -7, Corker -1)

The State ^ | 1 October 2006 | Steven Thomma

[okstate]

WASHINGTON — Democrats are slightly ahead of Republican incumbents in three election battleground states that will help determine control of the Senate, a series of polls released Sunday showed.

In Montana, Democrat Jon Tester had the support of 47 percent of registered voters, while incumbent Republican Sen. Conrad Burns had the support of 40 percent.

In Ohio, Democrat Rep. Sherrod Brown had 45 percent of registered voters, while incumbent Republican Sen. Mike DeWine had 43 percent.

In Tennessee, Democratic Rep. Harold Ford Jr. had 43 percent, and former Chattanooga Mayor Bob Corker, the Republican Senate nominee, had 42 percent.

Democrats probably must win all three races if they’re to take back control of the Senate on Nov. 7. They need to gain six seats overall, and these three are among the six seats held by Republicans that are considered most vulnerable.

Another Republican incumbent, Sen. George Allen of Virginia, was locked in a 43 percent to 43 percent dead heat with Democratic challenger James Webb, according to a poll released Friday by McClatchy Newspapers and MSNBC.

The surveys underscored how much these states are up for grabs and how much rides on the final five weeks of campaigning. The work of both major parties to get their supporters to turn out on Election Day could prove decisive. One in 10 voters remain undecided in Montana and Ohio, 12 percent in Virginia and 14 percent in Tennessee.

The polls were all conducted by Mason-Dixon Polling & Research Inc. for newspapers in each of the states — Lee Newspapers in Montana, the Cleveland Plain Dealer in Ohio and the Memphis Commercial Appeal and Chattanooga Times Free Press in Tennessee. Each state poll was of 625 registered voters and had an error margin of plus or minus four percentage points. The polls were taken between Monday and Thursday.

Mason-Dixon conducted a broader series of polls for McClatchy Newspapers and MSNBC in seven other closely fought Senate battleground states. Virginia results were released Friday, and the other six will be released Monday.

[okstate]

Who calculated the 240000:1 odds?

And to stain Gallup's accuracy based on events of the 1940s is ridiculous. Do you think the same people running the company then run it now?

[LS]

Yes, polls usually are off by that much. In 1996, every one of them was on the very low end of the MOE---but more important, they ALL . . .ALL were tilted toward the Dem. Randomness suggests that if the polling problems were truly "accidental," some would be off in the Dems' favor and some in the Republicans' favor. But the odds, as calcualted by one Freeper at the time, were 240,000 to 1 that ALL would be off in the direction of the Dems.

It seems logical to say they wouldn't be off, but the evidence says otherwise . . . THIS YEAR. We have an analysis of Ipsos-Reid that shows it to be oversampling Dems by a whopping 10%. Now, to put that in perspective, that's enough to essentially give EVERY SINGLE "VULNERABLE" REPUBLICAN the victory, plus elect Steele and Kean. Moreover, we have evidence from last month of a pollster admitting IN COURT that she had "re-labeled" Dems as Republicans merely to get her quotas on her numbers.

Trust me. I'm working for the Blackwell campaign, and I can tell you the polls are nowhere close---not in the same universe---as what is happening.

[DontBelieveAugPolls]

he needs to run hard on it, if it resonates in MT, if it doesnt your strategy is a flwed one.

MT is libertarian and far from any potential terrorist threats. I wouldnt expect support for the Patriot Act to be as high in MT as say in TX.

Fact the majority of americans, support it could be irrelevant to a MT Senate race.

[DontBelieveAugPolls]

""Yes, polls usually are off by that much. In 1996, every one of them was on the very low end of the MOE---but more important, they ALL"'


I agree in 1996 no poll showed Dole above 40% or Clinton under 50%, but I am asking name some polls which failed to predict the winner. Other than Smith in NH in 1996.

[LS]

1) Spencer Abraham (MI), as I recall, was down deep double digits in the last poll of his last victory and won. 2) They called the winner right in 1996 . . . but were ALL massively off, and all in Clinton's favor. Strange, no? Randomness would suggest one of them just by accident would be off in Dole's favor.

2) I don't carry around the polling results of every poll on some note card from previous races. I think they missed a ton. But better yet, I'll give you MY predictions, made right here on FR, before the 2004 and 2006 elections for the key races (I didn't bother with those that were "sure thing."): in 2002, I called every single Senate race correctly (including Allard, which, I think, all the pollsters had losing) except Thune. He lost by 500 votes.

In 2004, I called the electoral count at 300 for Bush (I missed only PA). I called every single contested Senate seat correctly except Salazar in Colorado (again, a very, very close vote). So out of all the contested elections in two election cycles, I've been right on all but two races. Match that against Sabato or Charlie Cook or most pollsters.

And I'm telling you the GOP will gain at least one seat each in the House and Senate.

[LS]

And you're dreaming if you think the Dems can do ANYTHING with this. ANYTHING.

Try this little test. Go to the supermarket and ask ten people who Bob Foley is. I rest my case.

[okstate]

a) Ipsos-Reid has already had its internals analyzed and was found to be oversampling Dems by a whopping 10%. That alone would give victory to EVERY "VULNERABLE" REPUBLICAN, PLUS KEAN, PLUS STEELE and would make Kennedy and Harris quite competitive.

Point 1: Ipsos-Reid was a national poll. It applies in no way to analyzing the accuracy of state polls

Point 2: You keep referring to this one poll, not the at least 150 state Senate polls released in the last two months.

Point 3: "Oversampling" Democrats by 10 percent still wouldn't give Santorum a victory.

b) A month ago a pollster pleaded guilty to manipulating data vis-a-vis labeling "Democrats" as "Republicans" to achieve pre-set quota numbers.

Point 4: Said "pollster" is DataUSA or Viewpoint USA. I've never seen a single poll ever released by that corporation. Never seen any here, on pollingreport, on RCP, on Cook, or on any blogs. Nowhere. Besides that, Viewpoint USA was not a media or mainstream pollster -- they specialized in INTERNAL polling, which I don't happen to put a lot of stock in anyway. The ironic thing is that me, the "poll guy," doesn't believe internal polls. On the other hand, you, the tinfoil hat media-conspiracy guy, believe only what campaign workers (and Zogby Interactive -- as long as he shows Blackwell losing by less than 10) tell you while completely denying all independent and nonpartisan evidence to the contrary.

[DontBelieveAugPolls]

"") Spencer Abraham (MI), as I recall, was down deep double digits in the last poll of his last victory and won.""


He wasnt running against an incumbent in 1994. Don Reagal resigned.

Ill have to remember your screen name for Nov 8th to chastise you.

[LS]

Perhaps, but it also means that the Dems are in deep doo.

For them to have known about this and not saved it for election week tells me they are screwed. But, believe what you want. In 10 days, if not sooner, Terrill Owens' suicide will still be getting more press than Foley.

[LS]

You better remember it because I'll expect homage.

[LS]

Shows you really don't know much about polls. Bigger companies outsource to these smaller organizations.

Ipsos, as a national poll, is PRECISELY why this is so serious. No one bothers to get inside the internals of these crap state polls, and you know it. That's why you hide and obfuscate about them.

And the last I saw on Santorum, he was six down. Yep, that would do it.

[okstate]

"In 2004, I called the electoral count at 300 for Bush (I missed only PA)."

Yet another lie. Here's what you said the night before Election 2004.

"To: Bonaventure What I'm hearing from NJ is that it is a TIE. Watch for NJ to go early to us. 4 posted on 11/01/2004 9:04:40 AM CST by LS"

Another one:

Posted by LS to slowhand520 On News/Activism 10/27/2004 11:26:05 AM CDT · 3 of 7
Yep. Kerry is toast, and the race has been over for 10 days, a I said 10 days ago. The internals show Kerry that no matter what he has done, he hasn't been able to pull even in either FL or OH, and has drifted further back in IA, WI, NM and now is losing HA and even NJ."

I assume "HA" is Hawaii, no? Another one:

"Posted by LS to NYC Republican On News/Activism 10/27/2004 7:23:26 PM CDT · 154 of 190
You answered your own question. ARG is as reliable as a poll of DNC. BTW, I was polled tonight by Mason-Dixon about OH. Very professional poll. Bush, in my assessment, is up about 4 in OH, and likely will go up 5 or so after the undecideds shake out."

Well you were only off 3 points on the Ohio result. Not as good as most pollsters, but it's within the margin of error! ;) Funny also that you praise Mason-Dixon while bashing them on this very thread.

Stinks when people can search the archives, huh?

[GregH]

'MT is libertarian and far from any potential terrorist threats. I wouldnt expect support for the Patriot Act to be as high in MT as say in TX. '
While levels of support vary, I think Patriot act is popular right across the country including places like CA, WA etc. Thats why many Dems in congress voted for it, as they would know very well they dont want to appear on the wrong side of the issue. While Libertarian streaks in Montana are there, I still think the patriot act has majority support there and this can help Burns cast Tester as the left wing kook he is. Dems weak issue is fighting terrorism and Repubs like Burns should be strategising on getting high political mileage on that issue and exploit the opening available.

[DontBelieveAugPolls]

keep in mind that after Reagans 1980 landslide, MT had two liberal DEM Senators

[okstate]

Santorum has not been within six since August 21st, the last Keystone Poll. http://uselectionatlas.org/POLLS/SENATE/2006/polls.php?fips=42 Bigger companies outsource to these smaller organizations.

Where have I disputed that? The point is that national polling is not relatable to state polling. It's not like Ipsos hires 50 different companies to poll each state and then merge the results together.

No one bothers to get inside the internals of these crap state polls, and you know it.

Survey USA, one of your favorite targets, posts detailed crosstabs of all demographic breakdowns for every poll they conduct. www.surveyusa.com

As for these polls, check out post 67. Internals for M-D's Senate polls are posted there at the link.

The main state pollster that doesn't post good internals is Rasmussen, mainly because I don't have premium membership there so I can't access the crosstabs. I bet you can get them if you pay enough.

If you would actually read the threads on which you post this would be a lot easier.

[tcrlaf]

Competitive has an effect, but at mid-terms the GOTV has a HUGE effect....

Simply put, the sheeple WILL NOT turn out in droves for this, and the Party Machines, not the "nutroots" will rule the day...

It's an incumbant world now...

[Redmen4ever]

Here are my rules for deciding how "the undecideds" will break:

1. INCUMBENT RULE ... If the incumbent has better favorables than his number in the beauty contest, the difference between his favorables and his number in the beauty contest will break toward him.

2. HISTORICAL VOTING TENDENCY RULE ... After allocating any excess in the favorables over the beauty contest, allocate a fraction as high as half of the remaining undecideds to the candidate of the party that the voters of the state/district historically vote for, and the remainder to abstaining.

3. COAT-TAIL RULE ... A strong candidate at the top of the ticket can drag in voters for down-ticket races (and off-set the non-voting part of the prior rule) (which can really help down-tickets Republicans in red states and down-ticket Democrats in blue states).

4. THIRD-PARTY RULE ... If there is a third-party candidate in the race pulling anything significant, at least half of the indicated support for this candidate will break to the major party candidate on the same side of the political spectrum.

Because of Rule 4, I always thought Allard was going to win (and that the third-party candidates in that race weren't going to do as well as was being indicated).

[GregH]

'keep in mind that after Reagans 1980 landslide, MT had two liberal DEM Senators'

also keep in mind, GOP won the state in the last 2 Presidential elections.

[SolomoninSouthDakota]

I've witnessed the same phenomena having researched it in 1996 or 2000. The pre-election polls were nearly always skewed to the liberal side--almost without exception if I am remembering correctly. I believe it is 3 to 4 percentage points you can credit to the Republican candidate right off the top. I think things did improve in the polling world though after the 2000 debacle.

Great to hear your views. I recall your post over time and though your message is not always a comforting one it does prove accurate. This gives me even greater confidence in a GOP maintained Congress.

[1035rep]

Great to hear your views. I recall your post over time and though your message is not always a comforting one it does prove accurate. This gives me even greater confidence in a GOP maintained Congress.

I'll second that!