Talk Elections

Just saw the report on the channel, looking for the update on the website.

Consider the source!

Why would an otherwise seemingly intelligent person post such a mind numbingly stupid comment?

Considering it makes me all the more likely to believe it.  FOX was one of the top two firms for the 2000 result I believe.

The final Opinion Dynamics/Fox news poll of 2000 predicted a dead heat between Vice President Gore and then Governor Bush.

A "fairly" good call I believe :)

Your right.  Fox/Opinion Dynamics called the race dead on in 2000.  This poll must be accurate.

I wouldn't go that far.  Even the best poll can be a victim of statistics, but it certainly isn't biased.

Whats the % undecided etc? Hang on a minute... I'm commenting on a national poll... sh*t... breaking me own rules... I'll go now...

No... it means Opinion Dynamics is a good firm with relatively few structural flaws and biases in their polls.

The "probability god" can smack them around just as badly as any other poll.

Alright, fine.  I know that, I was just refuting his assertion that the poll was biased.

Source aside, polls do contain error.  But regardless what the polls currently say, Bush is indeed in very serious trouble..and...and...and...his underlying support is weakening.

And Bush's poor communication skills (esp. when he is under fire) is preventing him from launching an offensive.

Bush has no voice.  If Kerry could learn to talk without pointing his finger...

...I still think Dean was the Dems best bet.  At least he had some kind of a message.

I only glanced the poll, so I did not get the full details.  I am fairly certain that the breakdown was

Bush 47
Kerry 40
Nader 3
Undecided 10

Don't hold me to that though.

Dick Morris should be in hog heaven after this poll so he can go back on TV and say he knew all along it would be a Bush rout.  47-40 is quite good considering the constant influx of negativity from Kerry and the plastering of Abu Gharaib(sp) on the front cover of each and every newspaper.

Come on in... National polls are fun... :)

What was the % of undecided then? ;)

()

Look up.  

Or here.

Bush 47
Kerry 40
Nader 3
Undecided 10

Not 100% certain on Nader and undecided.


EDIT:  They just ran it, I was right.  Numbers are accurate.

I hate posting right after myself, but here is the pdf on the Fox poll.

http://www.foxnews.com/projects/pdf/062404_poll.pdf (http://www.foxnews.com/projects/pdf/062404_poll.pdf)

Their story on it:

http://www.foxnews.com/story/0,2933,123679,00.html (http://www.foxnews.com/story/0,2933,123679,00.html)

Edit:  THey have some state polling data in there I will pull out :

FLorida:

Bush: 48
Kerry: 38
Nader: 3
Undecided: 10

Michigan:

Bush: 42
Kerry: 40
Nader: 5
Undecided: 12

Ohio:

Bush: 45
Kerry: 41
Nader: 4
Undecided: 9

Pennsylvania:

Bush: 46
Kerry: 41
Nader: 3
Undecided: 9

These were 4 seperate polls in those states.

Your numbers are right.  June 22-23.  900RV.

This poll completley contradicts Gallup, completley Contradicts two ARG polls and a Quinnipiac poll, has Bush up 43-42 in blue states.  (I thiink their are some real serious issues with this poll.  Is every other pollster on the national and state level wrong??  I highly dount that, and the fact Bush is doing slighlty better in blue states 43-42 than battlegrounds 43-43 is the real kicker.

This is 5 different polls.  The main one is where the blue states come from.  AN error in there woudl not effect the 4 state polls, unless there was a systemic problem.

The polls do also support a number of recent polls too, Survey USA in FL, the Harris and Tipp nationals, and carry the trend for MI, PA and OH.  

It is likely the method is overpolling Bush a bit, but not by a whole lot.

Well there seem to be no internals anywhere in the lengthy file, except a brief paragraph indicating a sample of 900 RV and and the MoE.

Hiding internals is a disturbing trend.  

Then again, times are better now than they were ever before.  We now get confidence reports with polls, something that was unheard of before.  Hopefully, in a few years, putting ll the relevant internals up with your poll will be a standard practice.

Harris poll was over a week ago.  Since then ALL the other polls have showed it close the ABC Poll had its issues the other direction.

7 of the 8 previous PA polls showed Kerry ahead exception Rasmussen by 1 9 of the last 10 MI polls including 5 in a row have showed Kerry ahead in MI.  Ohio polls have gone back & forth the entire way.  ALL FL polls have been close, Survey USA poll that showed Bush up 7 was shown to be flawed with an overampling of Reps  (the poll had something like Reps identity +7 in its sample, when the state is actually Dems +2)

So you are telling me to ignore a poll that is a week old in the national and pay attention to several month old polls on the state level?  See the contradiction?

ALso, since Harris has been a TIPP/IBD poll that had Bush up.

The Survey USA poll used self identification, they pointed this out and said it may be due to the Reagan funeral that they were skewed toward Bush.

http://www.realclearpolitics.com/bush_vs_kerry.html (http://www.realclearpolitics.com/bush_vs_kerry.html)

Look there and tell me where the national trend lines are.

I was going back further for the state polls because you said you didn't trust the newer Ohio polls that showed Kerry ahead.  The bottom line is as far as the Ohio pokks its been back & forth.

The +2 for Dems was also in the 2000 exit polls for FLA, the Reagan so called boost could impact the results to a point, but its not going to have a 9 point swing in identity.

TIPP had Bush up 3 with Nader, tie without Nader.  Gallup has Bush +1 both with & without Nader, Democorocy Corps has Kerry +1 without Nader and even with Nader.  So the TIPP poll, Gallup & Democrocy Corps all show it really close, meanwhile ABC & Fox don't

A few things.  First, the Republicans have had a better registration effort in FL the past 4 years.

Second, this is self identification.  You can have 100% ID themselves as republican and not be wrong, if that is what they all say.   During the Reagan funeral a huge swing to the Republicans seems about right.  They did the right thing by pointing out that this had occured.

Finally, on all the polls being all over the place, ain't it fun?  

Yet another reason never to trust the lying Faux News Channel.

Like they lied in 2000 when they NAILED the result perfectly?

That proves that they knew what the result was going to be and that the whole "election" was a fraud.

Brit Hume needs to get his smug face off the airwaves.

Yeah.  That sure proves it alright.

Right, because the vast right wing conspiracy decided the best thing for them would be to lose the popular vote and win on a tight recount in Florida.  Yeah, that was the plan.

Or how about the time Fox found out about Saddam Hussein's plans to kill his 2 son-in-laws that had defected to Jordan and did not tell anyone so they could keep their Baghdad bureau open.

Wait, that was CNN, America's most trusted news network.  My bad.

Let's pretend for a moment that you aren't in denial and everything you just said is true.  This poll would become MORE accurate because they know what the result will be again..

...thus making the whole system even more corrupt than I thought.

I'd like to know why ABC had the result of the 1998 election for Texas governor up on its site a day before the so-called election occurred.

(Remember, we're still playing pretend)

Oh, it's just about as corrupt as you thought.  Fox News cheated to predict the election and they would have just cheated again.

It still disproves your original argument that we shouldn't trust the poll because of the source.

ABC is now part of the right wing conspiracy.

Bhahahahahahahahaha

All forms of media are in his opinion.  He argued the LATimes was just the other day.

I notice you folks haven't denied that ABC did in fact post the results of a so-called election before the election even took place.

Nope.  I don't.  The election was fixed.  Everyone knows that in Texas, a Republican couldn't possibly win a statewide election unless the election was fixed.

I guarantee that ABC was in on it.  That closet conservative Peter Jennings was no doubt the ring leader.

I thought James Carville was the closet Conveervative...?

Did I miss the last meeting of the vast right wing conspiricy?

e-mail me the minutes ... ok ...?

I'll have the UFO drop it off for you in a few hours, if you can wait that long.

I am actually in a hurry...

If I take off my tin foil hat can you use the UFO to beam it directly into my brain using the pentagon's new mind control beam?

Lay off Brit Hume....he is great.

Brit Hume is a dour man with the credibility of a dried booger.

There's also this one reporter on one of the other networks who sort of looks like Brit Hume except he constantly smirks and grins at stuff that isn't funny. He gives me the creeps too.

I used the time machine.  Its in your mailbox as of 4 hours ago.

not credible like you, eh bandit?

Well it tends to show inaccuracy.  The Republicans have made gains in vopter drives thats true, but Florida has also seen an increase in women's population and Non-Cuban Hispanics.  So the overall what was added was preety equal.

Anyway point is when the Voter ID in the polls is so far off from the overall id of the state that shows the poll has problems.  The LA Times poll which was inaccurate to the left at a Voter ID of +14 to the dems (+10 over the Nat'l average of +4 dems) this poll had voter ID of +7 to the Republicans which is +9 in favor of the Reps over the state ID of +2 Dems.  So both polls are off about the same.  I'm not saying the Survey USA poll was biased, what I'm saying is its accuracy is a problem because the sample was a much heavier representation of Republicans than the state actually has

We interupt this thread for a "I hate the LA Times, they are a bias Yellow Journalism Rag" interuption

Re: LA Times VS SUSA.

The "correct" National voter ID is +4 to the Dems (2000 Exit polls) or +2 (Rolling average of National polls)

If we use +4, the LA Times missed by 9%

IN a poll of 1230 sample size, the chance that this is random error is about 7 chances in 10,000.

The LA Times AVERAGE party ID over their last 8 polls has been +11 to the Dem side.

I don't think it is random. - I invite anybody to argue it is :)

Survey USA had +4 to the GOP side, about 7% too high based on current registrations. - With their smaller sample size, about a 2% chance it was just random.

Also - SUSA clearly published the voter ID, the LA times went to extra-ordinary lengths to bury it.

SUSA TWICE on their release said "Hey - Reagan just died - All bets are off on this sample!" (or words to that effect) - The LA Times went to great lengths to DEFEND +13 as normal.

A hell of a moral and intelectual honesty related difference! (I hope Dems and GOPers alike will agree?)

Survey USA is also even handed - they miss left as often as they miss right. - A claim the LA Times cannot make.

A HUGE difference between the conduct of the two firms IMHO.

If you want to argue that the LA Times survey and the SUSA Florida were both off, hey I agree with you.

To argue that the moral integrity of the two outfits is even CLOSE to the same, I think is something that is a hard argument to make.

We apologize for this interuption - please return to your thread

Wait a minute.

Does anyone like ABC?

Well based on your picture, Ill take Hume for a TV face