It's Time to Cut to the Chase.

[TruthIsAll]

It's time to cut through the crap.
It's time to cut through the fog.
Let's get back to basics.

We know Mitofsky adjusts the Exit poll to the Actual Vote.
But what if the vote counts are wrong?

What if votes are corrupted by BBV and forever lost in the Ether of Cyberspace?
Wouldn't we want to know?

We know there is no audited paper backup.
How can we assume that fraud did not occur?
Especially when out of 88 touchscreen incidents, 86 Kerry votes registered for Bush?

Would it not be prudent to assume that fraud may very well have occurred?
Especially knowing that Republicans control the machines and count the votes.

Could that be why Oregon uses paper ballots and the exit poll perfectly matches the vote?

Why would anyone want to mix corrupted votes with a pristine exit poll?
Does that make any sense?

Wouldn't it make more sense to check the votes with exit polls like they do in other countries?
Rather than to adjust the exit polls to legitimize stolen votes?

Naysays, answer those simple questions.
They are so basic but you never address them.

You would rather waste our time with crap like:
"Republicans don't like to talk to pollsters."

[Old and In the Way]

Why not have 2 types of BBV machines?

Red ones that are owned by Republicans and Blue Ones that are owned by Democrats. The voter can choose which machine he chooses to have his vote registered on.

Seems like a fair compromise.

[Hugin]

2. Wouldn't work...

You'd have massive stuffing of both sides.

Outrageous outcomes... I've never claimed the blues are
completely honest. O8)

Also, what about 3rd party candidates?

BTW... I like your user name. :)

[Old and In the Way]

4. Just thinking outside the ballot box.

Welcome aboard Prag....and thanks for the comp!

[Hugin]

5. I never object to thinking! :) n/t

[davidgmills]

42. The blue ones would come with a paper trail and the red ones wouldn't

And we would have to count first

[Mistwell]

1. Well, for once, some signs of sainity

Well TIA, I gotta say, for once, I see some signs of logic in your post.

If your point is that we should have better exit polls, I agree. We need to take much larger samples, like they do in other countries, and cross check the data in mulitple ways, like they do in other countries. The point of exit polls show no longer be to just check demographic data after the election. It will take some money, but it's worth it as a good check on the system.

If your point is that we should use paper ballots and more secure forms of voting, I agree.

If your point is that fraud may have occurred and we should investigate it, I agree.

Of course, we both know where our areas of disagreement are...but for once, I think we can leave that out of this thread.

[TruthIsAll]

3. But you don't agree with this...

13,047 respondents,
randomly selected,
1.0% MOE


               HORIZONTAL		WEIGHTED		
PARTY ID	
     MIX	Bush	Kerry	Nader	Bush	Kerry	Nader
Dem 	38%	9%	90%	1%	3.4%	34.2%	0.4%
Rep 	35%	92%	7%	0%	32.2%	2.5%	0.0%
Ind	27%	45%	52%	2%	12.2%	14.0%	0.5%
	100%				47.77%	50.69%	0.92%
				
Probability: Poll(47.77%) to vote (50.73%): 0.00000000329447

         ********* 1 in 303,538,508 ************

Prob = 1 - NORMDIST (.5073,.4777, .01/1.96, TRUE)

[Mistwell]

9. Indeed

I believe that there is an inherant, basic flaw in one of the assumptions you use to analyze that data.

You think the Margin of Error is reliable for using that data to detect election fraud. I do not. Nor does the person who made that exit poll. Nor do ALL past exit polls. ALL of them were off. Almost all were off by well beyond the margin of error, sometimes by a LOT more than 2004. And given THAT empirical data, one cannot logically conclude that the MOE for 2004 is meaningful for detecting election fraud. And to date, you have never actually put forth a logical explanation for why that might be.

[davidgmills]

41. I'll take these exit polls over the count any day

You never answer my questions.

Which of course is that even these "flawed" exit polls are in all probability far more accurate than the count.

You always attack the accuracy of the polls and never attempt to attack the accuracy of the actual tabulation.

You always assume that doing something hard is more accurate than doing something much easier.

Maybe when you and TIA's critics start proving to the rest of us non-statistical and non-scientific types that there is very legitimate reason to assume the tabulation is accurate, then we will start listening to you naysayers. But until then, we won't.

So get to work and prove that the counts of yesteryear have been accurate and by what percent, then we can have a meaningful discussion of whether the exit polls were accurate enough to warrant our trust or not.

Until then, you guys sound like bags of wind.

[Mistwell]

54. Answers

David, how many times are you going to accuse me of "never answering your questionsd" when I answer ever single response to every single post I make every single day. If you actually want an answer, reply to my post, and every time I will answer you (unless you ask a rhetorical question).

"Which of course is that even these "flawed" exit polls are in all probability far more accurate than the count."

The flawed exit polls have historically been off by more than 8%, and off in every single year they have ever been done.

There is simply no evidence at all that the actual count is off by that much. We may think it is off due to fraud, but there is not solid evidence that it is off by more than 8%.

"You always attack the accuracy of the polls and never attempt to attack the accuracy of the actual tabulation."

I do attack the accuracy of the actual tabulation. However, I just make it plain that the exit polls are simply not a useful tool to do that. The exit polls are MORE flawed than fraud at the actual polls, and at best leave an ambiguous result.

"You always assume that doing something hard is more accurate than doing something much easier."

I don't assume thatr at all. I just make it clear that the exit polls are hopelessly flawed the way they are done right now, which is why they are specifcally not supposed to be used for the purpose you want to put them to. I am all for getting an accurate actual vote count...I just don't think exit polls, as they are done right now, in any way help with that.

"Maybe when you and TIA's critics start proving to the rest of us non-statistical and non-scientific types that there is very legitimate reason to assume the tabulation is accurate, then we will start listening to you naysayers. But until then, we won't."

Dude, how many times do I have to tell you that I do NOT think the actual tabulation is correction...that I JUST think the exit polls are even more hopelessly flawed and therefore by focusing on exit polls we hurt our own cause because the exit polls are soooo much easier to refute than us pointing out the actual flaws at the actual poll!

"So get to work and prove that the counts of yesteryear have been accurate and by what percent, then we can have a meaningful discussion of whether the exit polls were accurate enough to warrant our trust or not."

Yeah, I don't have to prove any such thing. If every actual poll varies from every exit poll by about the same amount, and that same amount is almost always beyond the margin of error, and there is NO evidence of fraud in those prior elections given the fact that it wouldn't have possibly changed the results in any of them (since the Reps LOST 3 of the 4, and won the 4th by such an overwhelming margin that even the flawed exit poll showed they still would have won no matter what), I have already shown enough to shift the burden back to anyone claiming fraud in those prior years. If you cannot show any evidence of fraud, and all the motives go the opposite direction, and all the results go the opposite direction, that's way more than enough to meet my burden of proof.

"Until then, you guys sound like bags of wind."

Oh yeah, I am the wind bag when TIA posts EVERY SINGLE DAY the SAME DAMN FLAWED CRAP. And he, every time, refuses to respond to the same flaw in his same crap. Gimme a break...stop marching that goose-step and start to think for yourself. TIA's analysis is the equivelent to him claiming he has proof the sun is hauled across the sky on the chariot of a roman diety because he sees it move across the sky every day at the same speed that diety is alleged to fly his chariot. The analysis is hopelessly flawed based on one of his most basic assumptions being wrong, and he is afraid to even address the issue. And you guys encourage him by mindlessly droning to his defense without giving any deep thought to the flaws in his claims.

[davidgmills]

55. Well you finally answered

"The flawed exit polls have historically been off by more than 8%, and off in every single year they have ever been done.

There is simply no evidence at all that the actual count is off by that much. We may think it is off due to fraud, but there is not solid evidence that it is off by more than 8%."

......

You are like arguing with a rock.

The first statement obviously shows your bias in presuming it is the polls and not the count that is wrong.


If that were not so, what you really would have said is that there has been a 8% discrepancy between the polls and the count for every single year they have ever been done. That might be accurate. I don't know (there seems to be an argument about this). But, assuming you are correct with that statement, that could very well mean the count was off five and the polls three. But that is not what you say or even imply.

I think it is a combination of the two, although I do not buy the arguments advanced so far as to why exit polls favor democrats.

I can easily believe that the Democrats have consistently been undercounted for all these years. I believe there is ample evidence to support that notion.

By the way, TIA is not the only one. Just a few days ago, a statistician from Britain posted on DU. He said he had forty years of experience in statistics. He supported TIA's views and encouraged him to press on. So where does that leave those of us who are not statisticians?

If I can find his post, I will pass it on.

[Mistwell]

57. response

"The first statement obviously shows your bias in presuming it is the polls and not the count that is wrong."

If it walks like a duck, quacks like a duck, swims like a duck, looks like a duck, then I say it is a duck. If every single Presidential election the exit polls overweight Democrats in the pre-fully-weighted form, and studies show why that might be happening, and the maker of the poll admits it is happening and writes a formula to try to fix it, and every other expert agrees that is what is happening, and the people directly impacted by it like Clinton and Kerry all also agree that is what is going on, and to assume different would mean a monumentally huge Republican conspiracy dating back to pre-electronic-voting days where despite big risk of criminal fraud and thousands of people NOT EVER talking about it they all committed massive fraud all to no avail because they still always either lost the popular vote or would have won it anyway...then yeah...I assume it is the polls and not the count that is wrong.

If you want to call a stack of evidence supporting my belief that is taller than me "bias", then you go right ahead. Might want to look in the mirror first though.

"By the way, TIA is not the only one. Just a few days ago, a statistician from Britain posted on DU. He said he had forty years of experience in statistics. He supported TIA's views and encouraged him to press on. So where does that leave those of us who are not statisticians?"

Expertese in exit polls is different than general expertise in statistics. If one is not aware of the epirical data supporting that particular sub-field, then ones assumptions can be flawed before the statisitical analysis is even started. And that is exactly the case here. TIA, and others, start with the premise that the margin of error is accurate for measuring fraud. And yet, that is a basic, flawed assumption. They guy who runs the exit poll says the MOE is not for fraud detection or the actual vote, but just for the demographics data. All the empirical evidence for the same exit poll shows that it is always off for the thing you are trying to measure. So I disagree with you that some anonymous british statistician is highly useful in determining if this sub-field of study is helpful for what you are trying to prove.

[brindis_desala]

59. But people have spoken about it

you just choose not to give them credence. Black disenfranchisement did not begin with GW nor is it exclusively the modus operandi of Republicans. The truth is that undoctored exit polls have probably been closer to the truth than most U.S. elections. Time to get it fixed, don't you think...

[Mistwell]

61. I agree

I agree there is and was fraud. I agree it is time to get it fixed. I just disagree that exit polls are a useful tool for doing that. And, I think eventually you will see that as well. Exit poll stuff is so inconclusive and awash with experts saying that your position is wrong that it presents our weakest claim as the front line. If you want it fixed, focus on stuff people will believe...

[RaulVB]

6. Well, we are all in the same page here (n/t)

[joevoter]

7. One correction

Oregon does not have exit polls because of the mail in ballot system.

And yes, The issue is the vulnerability for fraud in our election processes. Why should we trust more government/industry sponsored "voting reform" when suspects in this election are not being investigated? We need official investigations now! Exit poll discrepancies are an indicator that something went wrong.

[ottozen]

11. The FBI are WIMPS!

They work for CHIMPS!

[qwghlmian]

8. For crying out loud, do you even read what you post, TIA?

"Could that be why Oregon uses paper ballots and the exit poll perfectly matches the vote?"

Oregon has a mail-in ballot. There are no exit polls. There cannot be - there is nothing to "exit" from. Where do you come up with this stuff?

You have gotten answers to these questions at least a dozen times. Yet you pretend you didn't hear them. The exit polls are not "pristine" - the data you keep touting is collected in a way that is not random, and needs adjustment to the election results in order to be useful. You have been given links to professionals who explain this. You ignore these links.

You have been shown that the kind of exit polls that are conducted in the US are NOT the same as the kinds of exit polls that are conducted in some other countries in order to validate election results. You have been shown opinions of professionals that state outright that the US exit polls, by design, should not be used to validate election results. You ignore this and claim that no one ever answered you.

As for "Republicans don't like to talk to pollsters" - here is a research paper that shows that the non-responders in exit polls skew it toward Democrats.

http://www.duke.edu/~mms16/non_response2000.pdf

Feel free to ignore this just like you ignored other stuff and pretend that no one answered your questions.

[Mistwell]

10. Interesting find

"Our results also lead us to conclude that exit polls are likely to over-represent the opinions of younger and non-white voters. Because non-white voters tend to vote for Democrat candidates, over-representation of this social class will skew an exit poll’s results in that direction."

Okay. So that helps a bit.

[liam97]

12. Mistwell, can you explain why

a random sample would "necessarily" give skewed results? That seems to be your argument. It might over or underestimate the actual votes, but why/how can we automatically assume that there are overrepresentations of particular class of voters? How would we "know" that 3 months later - the demographic data was collected during the exit polls and posted on election night. How can the demographic data and therefore the nature of the sample change later?

[Igel]

17. Because Mitofsky's big "breakthrough" was ensuring

a random sample.

If you interview every 5th voter (1, 6, 11, etc.), you're stuck with the non-respondents. Prior to Mitofsky, they acted as if the non-respondents didn't count, so the actual voters could be # 1, 6, 12, 16 (if #11 bailed). This pretends non-respondents don't exist, and makes the poll non-random--it includes only those who wanted to answer.

Mitofsky has his people make a note of who didn't answer, and continue as if that person had responded. It means at the end, you wind up with a bunch of non-respondents to somehow "resolve".

If non-respondents were random, it wouldn't matter. But they're almost never random non-responses.

If Bushies don't answer, you get more Kerryite participation. If Kerryites don't answer, Bushies are over-represented. Until Mitofsky waves his hands and pulls off his patented resolve-the-non-respondent magic. He says that in strong Kerry areas he wound up overstating Bush support: it almost has to be because of the proportion of rep/dem non-respondents he assumes for those areas.

[Old and In the Way]

13. Look at the correlation between exit poll/paper ballots and

exit polls/BBV ballot states. I believe there's plenty of data to show that exit polls mirrored the actual results within the MOE (in both directions), but in states with BBV, the actual results always favored Bush outside the MOE when compared with the exit poll data.

[qwghlmian]

14. Example - New Hampshire

fully paper ballot state, I believe. Huge discrepancy between exit polls and actual election results - so big that Ida Briggs has convinced Nader to go and recount some precincts, with DUers sending contributions to Nader to do this. That was the first recount post-election, if you remember. Result of the recount: no problems found, no discrepancies between the recount and the actual election results.

Question: do you believe the exit polls were correct in NH and that Kerry should have won it by a landslide, like the exit polls indicated, instead of 1-2%, which was the actual election result?

[Old and In the Way]

21. Did they recount the entire state?

I thought it was only a selected small %. I do know this about New Hampshire, there was only about 400,000 votes and it took most of the night before a winner was called. Republicans control that state's political machinery, too.

But, OK, let's say we throw out NH....I still think we have a pretty strong correlation that shows the actual results and exit polls match up pretty damn well with states that have paper ballots, but not so in BBV states....and all deviations favor Bush.

[buzzard]

27. No, not even close if I find the link I will pass it on. n/t

[dbDESIGN]

35. NH Briggs links

http://www.invisibleida.com/Press_Release_1.htm

http://www.invisibleida.com/index.html

http://www.invisibleida.com/New_Hampshire.htm

The first link explains how they identified the most anomalous precincts to recount given limited resources.

The suspicion was not aroused by exit poll discrepancies but by analysis of voting trends and extrapolation from them.

I thought we had them here but the limited recount did not pan out.

Anybody know what the exit polls in NH showed?

[qwghlmian]

30. No, they did not have the money to recount

the whole state. They took the precincts that Ida Briggs told Nader to recount, because those looked the most suspicious to her. They said that if they found problems with those, they would raise funds to recounts more. Since no problems were found, no more precincts were recounted.

[FogerRox]

48. IIRC NH used paper ballots scanned by Diebold scanners-thusly

providing a legal basis to then ask for a full recount in Ohio if the NH recount showed a problem. NH used Paper, scanned by Diebold. Just like part of Ohio.

[Igel]

18. Mitofsky addresses this in his report.

Believe the numbers or not.

[Igel]

15. Actually, there are Oregon exit polls.

Mitofsky said they under(over?)stated Kerry by 1.2, and Bush 0.2.

All by phone. No preliminary numbers released. Just the final ones.

[Nederland]

24. No there were not

see post #23.

[Igel]

25. You may not want to call them exit polls,

because they didn't actually collar people as they left their precincts; you're right, the Oregon survey wasn't an exit poll. But the Oregon survey served the same purpose: to help call the state and to provide demographic breakdowns of who voted how before the actual returns were in.

In that case let's call them "polls taken for the purpose of assisting NEP clients". Exit polls are one kind, another is the absentee telephone survey Mitofsky did. For most states, the phone survey was an adjunct to the "main" exit polls; in Oregon, it was the only survey possible.

See the much bally-hooed report, p. 72, where Mitofsky compares the absentee telephone survey estimate with actuals.

Mitofsky overstated Kerry support by 1.2, and understated Bush support by 0.2.

In any event, TIA isn't exactly right: Mitofsky's figures for Oregon didn't match the certified results, but I can't say if they were within the calculated error.

[Nederland]

33. Thanks

A well reasoned response that clears things up.

[TruthIsAll]

19. No exit poll in Oregon? Wrong.

Seems to me, you are new to this board.

Your talking points are boring and have no credibility.
When you have something relevant to say, I will respond.
Meantime, don't waste my time by trying to engage me in your prefabricated web of spin.

You're just repeating others who preceeded you. You have nothing new to offer. DU has heard all those worn out talking points before.

FYI, Oregon was one of the EIGHT states which deviated to Kerry- by 1.60%, well within the MOE. And they use paper ballots exclusively there. What does THAT tell you?

So...do your homework before you post.





http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=288009&mesg_id=288868

[Nederland]

23. Do your own homework

There was no exit polling in Oregon. Oregon votes entirely by mail. As a result, the Oregon "exit poll" is done like a traditional pre-election poll where a sample of voters is polled by telephone. A poll conducted by phone is, by definition, not an "exit poll".

http://duckhenge.uoregon.edu/hparchive/display.php?q=20041022-exitpolls.html

[IndyPriest]

26. If the question was "Who DID you vote for?" and

not "Who WILL you vote for?", that certainly seems like an "exit poll."

[smartvoter]

31. Me too. nt

[Nederland]

32. Missing the point

Since the poll is conducted by telephone rather than by query people as they leave the polling place, you immediately face the question of "who" to call. Do you call 40% men and 60% women? Do you call 12% African Americans? How many Democrats? How many Republicans?

In the end a poll of this type has all the problems of a pre-election poll and therefore is not considered an exit poll.

[IndyPriest]

37. Are you serious? Those are exactly the same questions you would

face if you were out on the sidewalk on election day. You would determine how many women and men and blacks etc to ask BASED on PRIOR understandings of how the population sifts out in your area. A phone "exit poll" might seem severly crippled in this regard, until you realize: Oh, I can ask those who answer their phones "Are you male or female, Afro., Hispan., etc." When the answers to those questions are known, and you calculate if you need that person's response, you ask "For whom did you vote?"

[Nederland]

38. Not true

It would be helpful if you understood how the 2004 exit polls were conducted before venturing into the discussion. To enlighten yourself, I suggest you read the NEP methodology linked below. As you can see once you've bother to read it, you are wrong. Exit pollsters do not predetermine what demographic makeup they are seeking and then actively ask voters that match the desired profile. They ask every Nth voter to respond regardless of demographics. Adjustment, if any, are made after the data is collected.

Common sense should tell you that this is the only way it could possible work. After all, how could you know what percentage of the voters in a precinct were black unless you took a random sample of voters exiting the polling place and calculated the percentage from the resulting data? Your statement that a determination would be made based upon prior understandings of how the population sifts out in an area is therefore fundamentally flawed. The reason is that you cannot assume that simply because in the last presidential election the percentage of black voters was X, that in this election the percentage will stay the same--especially when voter turnout increases as much as it did in 2004.

This is why post voting polling in Oregon is fundamentally different exit polls in other states. When you select every Nth voter leaving a voting place, you are getting a random sample of people that voted at that location. When you call people on the telephone, you are getting a random sample of people that happen to answer their phone. There is absolutely no guarantee that those samples will match.


http://www.exit-poll.net/election-night/MethodsStatementNationalFinal.pdf

[IndyPriest]

47. Hmmmm -- we must read differently

Or we're talking about two different things. Every exit poll is statistically useful because general knowledge of the subgroups in a population is known IN ADVANCE. From what data? Historical, census, past voting registrations, etc. That general knowledge has to be in place BEFORE you start a poll in order to know the value of the responses you're getting once the poll begins. Isn't that the meaning of "stratification" which the Methods paper refers to? Because you have that knowledge PRIOR to the poll, you can THEN properly (or more properly) WEIGHT the actual responses you receive from every nth voter. Your reference to what "common sense" would dictate would certainly apply IF you were trying to get a truly RANDOM sample of voters and their preferences. But isn't it the point of exit polling that it is NOT random? And isn't it precisely that NONrandomness that gives a (typical) exit poll its SMALLER MoE?

I don't know if Oregon's phone "exit poll" used that methodology or not. But my only point is that if it did, it would qualify for what most mean by an "exit poll."

[KaliTracy]

39. i'm confused. I'm reading this as a contridiction to your post

# 38

<snip> Exit pollsters do not predetermine what demographic makeup they are seeking and then actively ask voters that match the desired profile. They ask every Nth voter to respond regardless of demographics. Adjustment, if any, are made after the data is collected.... <snip>


When they call random people they do not "look for" 40% this or 60% that.


And... according to the House Judiciary Committee Report on Ohio --

"...In addition, this year's poll numbers were designed to account for absentee votes after a large number of absentee votes contributed to the inaccurate projection of the Florida race in 2000. This year Mitofsky and Edison began telephone surveys in Key States before the election to screen for absentee voters and create an accurate estimate of their votes" (footnoted) Page 73

http://www.pdamerica.org/field/final%20status%20report.pdf

Sounds like an exit poll to me. But I'm not a statistician, just a curious soul.

[Nederland]

43. Ohio was exit polled

Oregon was not and could not be.

In the end this is not really an important distinction, more of a technicality. The main problem I and others have with TIA's methodology has to do with completely different issues and an inane tendency to ignore criticism and merely repeat the same thing over and over again. In this case, the repetition is "refute the math if you can..." This, off course, ignores the criticism entirely. Nobody is saying that TIA made a mistake in multiplication. The mistake is assuming without any proof whatsoever that exit poll sample were representative of the voting population as a whole. Until we see raw precinct level data, nothing should be assumed about anything.

[KaliTracy]

44. absentee phone polls in several Key States

"...In addition, this year's poll numbers were designed to account for absentee votes after a large number of absentee votes contributed to the inaccurate projection of the Florida race in 2000. This year Mitofsky and Edison began telephone surveys in Key States before the election to screen for absentee voters and create an accurate estimate of their votes" (footnoted) Page 73

I know it's not Oregon -- but they did use phone polls which in turn went into Exit Poll data. that was my point. sorry if I wasn't clear.

I agree we need to see raw precinct data. But we have been saying that all along.

[Nederland]

16. Answers

We know Mitofsky adjusts the Exit poll to the Actual Vote.
But what if the vote counts are wrong?

If the vote counts are wrong, then using them to adjust exit poll data would yield incorrect results.

What if votes are corrupted by BBV and forever lost in the Ether of Cyberspace?
Wouldn't we want to know?

Yes I would.

We know there is no audited paper backup.
How can we assume that fraud did not occur?
Especially when out of 88 touchscreen incidents, 86 Kerry votes registered for Bush?

We cannot assume fraud did not occur.

Would it not be prudent to assume that fraud may very well have occurred?
Especially knowing that Republicans control the machines and count the votes.

No. You should assume anything. You should let evidence speak for itself and make no assumptions either way.

Could that be why Oregon uses paper ballots and the exit poll perfectly matches the vote?

As discussed elsewhere, there were no exit polls in Oregon.

Why would anyone want to mix corrupted votes with a pristine exit poll?
Does that make any sense?

If the vote was corrupted, you wouldn't want to.

Wouldn't it make more sense to check the votes with exit polls like they do in other countries?
Rather than to adjust the exit polls to legitimize stolen votes?

Yes it would, and I would support designing future US exit polls to do exactly that, unlike its done today.


Now that I've answered all you questions, can we dispense with the idea that those that disagree with you never answer your questions? Can we also agree that since I've answered seven of your questions, it would only be fair that you answer seven questions of mine. How about it TIA--do you have the guts to answer a seven questions? I'll wait for a response, though I doubt I'll get anything from you other than evasions.

[TruthIsAll]

20. Nederland, there is nothing more to say. Oh, except OR was exit polled.

We have been at it long enough.

You are a naysayer from way back.
Fine. You have made your feelings known time and time again.

There is no point in sparring any further.

I have given you enough of my time.
And you keep coming back for more.

Give it up, already.
You can't win.

[Nederland]

22. As expected

I answer all your questions and you run away and hide from mine.

Typical.

A bit of advice, if you want to get your ideas accepted by anyone outside of those on DU that are predisposed to agree with you, you are going to have to engage in logical, rational debate. That means listening to people that disagree with you and responding to questions.

Failing that, the only place you will "win" is on DU. Perhaps the only thing you got the guts for is preaching to the choir. :shrug:

[TruthIsAll]

29. Rational debate is fruitless with you. I've tried it for two years.

I have given you enough of my time.

That last sentence of yours said it all.
Don't edit it out. Leave it in.
We have a permanent record of your incivility.

Questioning my guts.

Quite a statement, especially since you never had the guts to present any analysis of your own.

[Nederland]

34. LOL

We have a permanent record of your incivility.

Now I will be the first to admit that I am not be the nicest person around here, but that comment is just too rich coming from a person who has had countless posts deleted and was in fact banned from DU for a short period of time just weeks ago. In a comparison of who is more "civil", there is no comparison. In my three years here I have had a grand total of four posts deleted, I have never had a moderator give me a warning, and I certainly have never been banned.

Hello pot? This is kettle. You're black.

To get back on topic, I'm confused about why you try to change the subject and refuse to answer my questions. Care to explain?

[Name removed]

51. Deleted message

Message removed by moderator. Click here to review the message board rules.

[euler]

50. A gut feeling is...

... telling me that someone in this forum is not an adult. I'm not kidding. I think he's a teenager.

[smartvoter]

28. Whoaaahhh!!! Hold on here!

Republicans only dislike talking to pollsters in swing states. Get with the program!!!


:-)

[Mistwell]

36. It isn't just swing states...

...in fact, the disparity is greater in some NON swing states. It's just that we always focus out attention on swing states.

[Time for change]

56. But the discrepancy is much greater in the swing states n/t

[Mistwell]

58. No really, it isn't

In fact, I seem to recall New York had one of the biggest deviations.

[Time for change]

60. Look at the Mitofsky report

The discrepancy is outside of the margin of error in 5 of the 11 primary swing states (OH, FL, PA, MN, NH), and only 8 of the other 39 states (Yes, including NY).

[euler]

40. If....

....the point of your post is to ask this question....

Wouldn't it make more sense to check the votes with exit polls like they do in other countries?

....then the answer is 'Yes.'

I assume I am one of the 'naysayers' you refer to. The point to my 'naysaying' was to convey to the forum that we our wasting our valuable time analyzing exit polls. In this post, you're not analyzing exit polls. You're talking about a step that must be taken in order to make sure that there is no doubt about the integrity of future elections. Why would anyone disagree with that? I don't.

You say your points "are so basic but you never address them." Do you have any posts that DO NOT contain some sort of mathematical calculation of exit poll data thats attempt to prove this or that? I've never heard you make the statements you make in this post. If you did, send me a link. In any event, had you made this post anytime in the last 3 months, I would have agreed with you then as well.

Lest you think I've gone soft, let me say: I don't know if Mitofsky's claim about non-response bias from republicans are accurate or not yet. But, I can say that there is an enormous amount of evidence (papers, articles) available that document this phenomenon. It's real. It can't simply be laughed away as if it is some kind of idiotic joke by republicans. REMEMBER, I haven't yet read the report so I don't know if I think Mitofsky's claim is true or not yet. I'm just saying that it does us no good to say the concept itself, is laughable. (If we did say it, we would be wrong. Wrong is bad. It makes the MSM ignore what you're saying even if there is other evidence that is not related to the wrong statements.)

[FreepFryer]

45. The strategic risk/value of exit poll analysis aside...

...it's plainly NOT a waste of time to understand what is being offered as 'evidence' of a clean, fair election.

:hi: :kick: :hi:

[FreepFryer]

53. Conyers on exit polls

"The stakes for our democracy are simply too high for us to allow this matter to pass without a serious and substantive review of the exit poll data."

Rock on, TIA!

:hi: :kick: :hi:

[Name removed]

63. Deleted message

Message removed by moderator. Click here to review the message board rules.

[New Earth]

64. then why do you waste so much time on him?

:shrug:

[TruthIsAll]

66. Nothing serious and substantive? Where is YOUR analysis?

And what makes YOU qualified to critique my analysis?

Although I don't claim to be an expert, the experts seem to be in agreement with my "non-expert" analysis.

While "non-experts", naysayers like yourself, are not.

Case Closed.

[In Truth We Trust]

46. Great post as usual TIA

George W FraWd

[FogerRox]

49. A Lot of good points all around. Would like to add

Many times the actual polls---votes--- are not that accurate.
Spoilage
Chads
You didnt register.
Absentee spoilage
Provisional "
etc.
TIA has done SOOOOOO much here-- that is over my head-- I went to a friend who understands this better than I do--helping me wade thru this-

" A sufficiently randomized sample that's large
enough should have p at .05 or less (plus or minus five points).
Exit polling has become more important because it is a way to measure
and monitor the actual polling. In 2000, the exit polling was, in fact, more accurate than the actual polls. I believe that's what we are seeing again in 2004."

The same friend looking at some of TIAs recent work:

"I don't have much time to digest this. His math's okay, and it looks pretty impressive, but I suspect that there's some state-by-state source of error that might pare these numbers down a bit. For example, states that show exit poll/actual poll disparities favoring Bush could be perfectly legitimate, plus or minus ten points. There's also a possibility, however slight, that you have some polling error that would make the vote swing maybe a point or two more. Maybe he's mis-macro'ed his Excell spreadsheet. There could also be reportage errors in the data, and some might be
explainable through the normal voter spoilage, full moon, bad karma,
etc.p equal to or less than .01 is pretty much close to certainty as we get. p equal to or less than .001, is virtually a done deal. What this author is proposing, however, is that p is approximately .0000029. Forgive me if
that figure is too low to believe right off the bat.
You might want to check the exit polling state by state, look for and
come up with some possible sources of error in his work, and then plug the numbers into a sd equation (you can find them easily on the net, or in any calculus textbook at the library), and figure it out for yourself. If your results lead you to believe that the chances of this actually occurring are one in 100, then you've pretty much found the smoking gun."

I think TIA is going in the right direction. My father would say something about Academic Review, as a retired Math Prof. SUNY.

I'll willing to bet that TIAs work will stand up to some serious review. There are some chinks in the armor TIA but you are still standing.

SOOOOO if TIA did something wrong then lets go with that. MAke the corrections and see what we get---fair enuf--?

Oh yeah TIA said recently 23 states were 2% off in the Exit Poll to results. Mitofsky also said 26 states skewed to Kerry---4 to Bush.
That leaves 22---DID TIA call it at 23? Or get close?
Sorry If I'm reaching at straws again.

[FogerRox]

52. Gotta kick

[TruthIsAll]

65. State Exit Poll/Vote Statistical Update

This is an update/correction of the 2-party exit poll/vote
analysis.

Note the correction:
Fifteen (15) states deviated to Bush beyond the MOE.
I originally calculated 16 (using Nov. 3 data), then using
current vote data, I calculated it incorrectly last week as
20. 

Total	2-party votes: 121,056
(in thousands of votes)

42 of 50 states deviated to Bush
The probability is 0.0000005818
The odds are 1 in 1.719 million

15 states deviated beyond the MOE to Bush
The probability is 0.0000000000009166
The odds are 1 in 1.091 trillion

4 States flipped from Kerry to Bush (FL, OH, IA, NM)	

Kerry (2-party): 
Actual Vote: 59,085 (48.81%) 
Exit Poll:   60,997 (50.39%)	

Bush (2-party): 
Actual Vote: 61,971 (51.19%) 
Exit Poll: 60,059 (49.61%) 

Average MOE: 2.85%	
Average Net Deviation to Bush: -1.73%
				
……	Votes	Poll	Poll	Poll	Vote	Vote	Poll	Dev.	Dev.	Dev	>2.0%Flip	Dev.	Vote	Poll	Vote	Poll
……	(000)	Size	MOE	Kerry	Kerry	Bush	Dev.	/MOE	>MOE	Bush	Bush	Bush	(000)	Kerry	Kerry	Bush	Bush
AK	302	910	3.31%	40.50%	36.77%	63.23%	-3.73%	112%	YES	YES	YES		-11	111	122	191	180
AL	1,870	730	3.70%	41.00%	37.10%	62.90%	-3.90%	105%	YES	YES	YES		-73	694	767	1,176	1,103
AR	1,998	1402	2.67%	46.60%	45.07%	54.93%	-1.53%	57%		YES			-31	900	931	1,097	1,067
AZ	1,043	1859	2.32%	47.00%	44.72%	55.28%	-2.28%	98%		YES	YES		-24	467	490	577	553
CA	12,255	1919	2.28%	54.00%	55.04%	44.96%	1.04%	46%					128	6,745	6,618	5,510	5,637

CO	2,103	2515	1.99%	49.10%	47.63%	52.37%	-1.47%	74%		YES			-31	1,002	1,033	1,101	1,070
CT	1,551	872	3.39%	58.50%	55.27%	44.73%	-3.23%	95%		YES	YES		-50	857	908	694	644
DC	372	795	3.55%	91.00%	90.52%	9.48%	-0.48%	14%		YES			-2	337	338	35	33
DE	224	770	3.60%	58.50%	53.83%	46.17%	-4.67%	130%	YES	YES	YES		-10	121	131	104	93
FL	7,548	2846	1.87%	50.50%	47.48%	52.52%	-3.02%	161%	YES	YES	YES	YES	-228	3,584	3,812	3,965	3,736

GA	3,280	1536	2.55%	43.00%	41.65%	58.35%	-1.35%	53%		YES			-44	1,366	1,411	1,914	1,870
HI	426	499	4.48%	53.30%	54.40%	45.60%	1.10%	25%					5	232	227	194	199
IA	1,494	2502	2.00%	50.60%	49.66%	50.34%	-0.94%	47%		YES		YES	-14	742	756	752	738
ID	590	559	4.23%	33.50%	30.68%	69.32%	-2.82%	67%		YES	YES		-17	181	198	409	393
IL	5,239	1392	2.68%	57.00%	55.21%	44.79%	-1.79%	67%		YES			-94	2,892	2,986	2,347	2,253

IN	2,448	926	3.29%	41.00%	39.58%	60.42%	-1.42%	43%		YES			-35	969	1,004	1,479	1,445
KS	1,171	654	3.91%	35.00%	37.13%	62.87%	2.13%	55%					25	435	410	736	761
KY	1,782	1034	3.11%	41.00%	39.99%	60.01%	-1.01%	32%		YES			-18	713	731	1,069	1,051
LA	1,922	1669	2.45%	44.50%	42.67%	57.33%	-1.83%	75%		YES			-35	820	855	1,102	1,067
MA	2,875	889	3.35%	66.00%	62.74%	37.26%	-3.26%	97%		YES	YES		-94	1,804	1,897	1,071	977

MD	2,359	1000	3.16%	57.00%	56.57%	43.43%	-0.43%	14%		YES			-10	1,334	1,345	1,025	1,014
ME	727	1968	2.25%	54.70%	54.58%	45.42%	-0.12%	5%		YES			-1	397	398	330	329
MI	4,793	2452	2.02%	52.50%	51.73%	48.27%	-0.77%	38%		YES			-37	2,479	2,516	2,314	2,277
MN	2,792	2178	2.14%	54.50%	51.76%	48.24%	-2.74%	128%	YES	YES	YES		-76	1,445	1,521	1,347	1,270
MO	2,715	2158	2.15%	47.50%	46.38%	53.62%	-1.12%	52%		YES			-30	1,259	1,290	1,456	1,425

MS	1,130	798	3.54%	43.30%	40.49%	59.51%	-2.81%	79%		YES	YES		-32	458	489	673	641
MT	440	640	3.95%	39.80%	39.50%	60.50%	-0.30%	8%		YES			-1	174	175	266	265
NC	3,487	2167	2.15%	48.00%	43.76%	56.24%	-4.24%	197%	YES	YES	YES		-148	1,526	1,674	1,961	1,813
ND	308	649	3.93%	34.00%	36.09%	63.91%	2.09%	53%					6	111	105	197	203
NE	767	785	3.57%	36.80%	33.15%	66.85%	-3.65%	102%	YES	YES	YES		-28	254	282	513	485

NH	672	1849	2.33%	55.40%	50.69%	49.31%	-4.71%	203%	YES	YES	YES		-32	341	372	331	300
NJ	3,581	1520	2.56%	55.00%	53.37%	46.63%	-1.63%	64%		YES			-58	1,911	1,970	1,670	1,612
NM	748	1951	2.26%	51.30%	49.60%	50.40%	-1.70%	75%		YES		YES	-13	371	384	377	364
NV	816	2116	2.17%	49.40%	48.68%	51.32%	-0.72%	33%		YES			-6	397	403	419	413
NY	7,277	1452	2.62%	63.00%	59.29%	40.71%	-3.71%	141%	YES	YES	YES		-270	4,314	4,584	2,963	2,692

OH	5,599	1963	2.26%	52.10%	48.94%	51.06%	-3.16%	140%	YES	YES	YES	YES	-177	2,740	2,917	2,859	2,682
OK	1,464	1539	2.55%	35.00%	34.43%	65.57%	-0.57%	22%		YES			-8	504	512	960	951
OR	1,810	1064	3.07%	51.20%	52.11%	47.89%	0.91%	30%					16	943	927	867	883
PA	5,732	1930	2.28%	54.30%	51.26%	48.74%	-3.04%	134%	YES	YES	YES		-174	2,938	3,112	2,794	2,619
RI	429	809	3.52%	64.00%	60.58%	39.42%	-3.42%	97%		YES	YES		-15	260	274	169	154

SC	1,600	1735	2.40%	46.00%	41.36%	58.64%	-4.64%	193%	YES	YES	YES		-74	662	736	938	864
SD	382	1495	2.59%	37.80%	39.09%	60.91%	1.29%	50%					5	149	144	233	237
TN	2,421	1774	2.37%	41.50%	42.81%	57.19%	1.31%	55%					32	1,036	1,005	1,384	1,416
TX	7,360	1671	2.45%	37.00%	38.49%	61.51%	1.49%	61%					110	2,833	2,723	4,527	4,637
UT	905	798	3.54%	30.50%	26.65%	73.35%	-3.85%	109%	YES	YES	YES		-35	241	276	664	629

VA	3,172	1431	2.64%	48.00%	45.87%	54.13%	-2.13%	81%		YES	YES		-68	1,455	1,522	1,717	1,649
VT	305	685	3.82%	65.00%	60.30%	39.70%	-4.70%	123%	YES	YES	YES		-14	184	198	121	107
WA	2,815	2123	2.17%	54.90%	53.65%	46.35%	-1.25%	58%		YES			-35	1,510	1,545	1,305	1,270
WI	2,968	2223	2.12%	52.50%	50.19%	49.81%	-2.31%	109%	YES	YES	YES		-68	1,490	1,558	1,478	1,410
WV	750	1722	2.41%	45.30%	43.52%	56.48%	-1.78%	74%		YES			-13	327	340	424	410
WY	238	684	3.82%	30.90%	29.69%	70.31%	-1.21%	32%		YES			-3	71	74	168	165
																	
Avg	2,374	1443	2.85%	48.81%	50.39%	51.19%	-1.73%	79%									
Med	1,782	1495	2.59%	49.10%	47.48%	52.52%	-1.70%	67%									
Total	121,056	73,607							15	43	22	4	-1,912	59,085	60,997	61,971	60,059
Diff													-1.58%	48.81%	50.39%	51.19%	49.61%

[Bouvet_Island]

62. Hi Truthisall,

I have a question for you and please take this as that, I am not out to attack the exit polls of which I currently hold a neutral opinion.

It is an idea that I got from one article off verified voting I take it you heard about, http://www.verifiedvoting.org/article.php?id=5487 , "Did Networks Fake Exit Polls, While AP 'Accessed' 2,995 Mainframe Computers?"

I am skeptical about that article, as it seems it is not very well sourced or built up IMO. But it holds an interesting point, I believe if more people starts to believe "the exit polls were right" that the first counterattack will be to say it is fraudulent, a claim that probably can be "proven" even if the data are released and very effectively so before. I mean in a Fox sense of proven.

The question is if it is possible to illustrate and "prove" the large proportions and the reality of the poll by any photographic evidence, any pile or warehouse with the documents, pictures or illustration of the computer equipment used? Any interviews or article on pollsters going to work, at work, or reports of them being present? I am not asking a rethorical question, I am sincerely interested in how well it can be quantified from public internet and official record as fraud was the first thing to be brought up on Fox, by the fellah Freeman quotes.

I think this data should be useful no matter what alignment one has to the accurateness of the exit polls, if it is neutral or not.

Also, do anyone know if it is possible to FOIA parts of the mitofsky accounting that are relevant, and crosscheck? I mean, all those pollsters would have to get paid and it would be records of it with the IRS, wouldn´t it? Since we are not getting any data from Mitofsky, it would be nice if we could get some of his data or methods by ourselves.

[TruthIsAll]

67. Sorry, can't help you with that. I calculate the numbers. Let others

investigate or spin as they wish. It's not my purview.

[Bouvet_Island]

68. Well,

that is a wishful approach if I ever saw one.

[berniew1]

69. Documented fraud & dirty tricks acount for Exit Poll/official vote differ

differance

Widespread systematic vote machine fraud, and dirty tricks and suppression of minority registration and voting in at least 15 states in 2004 Election: summary of some of the documentation

http://www.flcv.com/ussumall.html