Mythbusting the Exit Polls

[Mistwell]

I'll say it up front: I think this election involved fraud by Republicans. We have evidence of it, including the lines, the missing machines, the tossing of otherwise valid provisional ballots, and all the other stuff we have talked about here. I do not know if the fraud was enough to change the result of the election, though I support a full investigation to find that out. And, even if it turns out that it wasn't enough to change the result, it still should be corrected to return to people the sense that their vote counts.

That said, I am very tired of all the myths surrounding exit polls that I see here every single day. The deeper I have looked into the exit poll data, the LESS convincing it is. In fact, I think no rationale person who is objectively looking at the exit poll issue can come away with any conclusion other than "Exit polls don't prove anything at all about fraud in this election".

It would be okay if people were just wasting time on a useless thing, if that is all that was happening. However, I think all the intense focus on the exit poll claims is damaging to the anti-fraud movement. Good and smart people, who would otherwise be spending their time looking into evidence of ACTUAL fraud, are instead spinning their wheels on this theoretical and shaky claim about the exit polls. In addition, because the exit poll claims are so easy to refute, it taints the entire anti-fraud effort when those claims are portrayed to the public as one of the stronger reasons to believe there was fraud. It's like crying wolf - do it often enough, and nobody will believe you when you present real proof of fraud.

So, even though I know I will get flack for this, and even though I know I will be accused of being a traitor, a freeper, a Republican, a pessimist, and a jerk, and lots of other stuff, I will once again present at least some of the reasons people should consider before they continue to present the exit poll claims to the public as being an important part of the anti-fraud effort. I do this not because I am trying to drag people down, but because I care about the anti-fraud effort enough to want to help people focus on the things that will get us results - and that is the evidence of actual fraud, and correcting it.

1) US Exit Polls are historically extremely accurate, even in a close race.

FALSE. In fact, raw exit poll data, for the prior four Presidential elections, have always overstated the Democratic vote, sometimes by more than they overstated for this election.

http://www.washingtonmonthly.com/archives/individual/2004_11/005178.php

Year / Exit Poll / Results / Dem Lead / Dem Actual
1988 / Dukakis: 50.3% Bush: 49.7% / +0.6% / -7.7%
1992 / Clinton: 46% Bush: 33.2% / +12.8% / +5.6%
1996 / Clinton: 52.2% Dole: 37.5% / +14.7% / +8.5%
2000 / Gore: 48.5% Bush: 46.2% / +2.3% / +0.5%

"As you can see, the raw exit poll results always overstate the Democratic vote, sometimes by as much as eight percentage points. So the fact that the raw results this year overstated Kerry's actual vote tally is hardly cause for alarm."

2) Internationally, Exit Polls are very accurate – just look at Ukraine.

FALSE. Ukraine’s second election was one of the most observed elections in the history of the world. International observers from all over the globe, including the UN, paid special attention to the last election, and counted and recounted every vote (which were paper votes) to make sure no fraud took place. Despite this, in not one, not two, but ALL THREE EXIT POLLS, the polls deviated from actual election results by a great deal more than the margin of error.

http://www.myrtlebeachonline.com/mld/myrtlebeachonline/10499250.htm

Exit Poll 1: 58.1% to 38.4%, MOE=2
Exit Poll 2: 56.5% to 41.3%, no MOE given
Exit Poll 3: 56% to 41%, MOE=2

Official results: http://tinyurl.com/3ufxp

51.99% to 44.2%

As can be seen, the final results varied from all the exit polls by much more than the margin of error.

3) Experts agree Exit Polls like this one are very accurate.

The leading expert on the subject is Warren Mitofsky, the father of exit polls. Mitofsky is also a lifelong liberal and "apparently holds no brief for Bush." He's the guy who ran this exit poll as well. He's the one that would be called to the stand to testify about this issue.

http://www.russbaker.com/TomPaine_com%20-%20Election%202004%20Stolen%20Or%20Lost.htm

Mitofsky does not believe the early raw exit poll data (the stuff showing a Kerry win - from his own exit poll) indicates fraud. Indeed, he thinks the early raw data is inaccurate, and not useful until it is weighted with his formula. Once the formula was applied after all data came in, the election returned to being within the margin of error.

http://mayflowerhill.blogspot.com/2004/11/mayflower-hill-exclusive-warren.html

He has never, EVER claimed that the margin of error in his early exit polls (the ones that TIA likes to quote constantly) are set for detecting election fraud or even set for the vote itself. The MOE is set ONLY for the Demographic data. That disclaimer is included with every single early release of his exit poll data (though not always repeated by newspapers and television programs that use his data). It is a total myth that the margin or error in the pre-final exit polls are accurate, or ever meant to be accurate, for the vote itself.

4) All the State-by-State deviations between the exit polls came out to favor Bush in the actual counted vote.

FALSE: Of the 50 states, 40 deviated to Bush, and 10 deviated to Kerry. Some of the 10 deviating to Kerry were considered swing states prior to the election, and no explanation has been offered for why exit polls would deviate in this way. It’s usually just ignored by those claiming exit polls prove fraud.

5) Studies all show the exit polls work to detect election fraud in the US.

FALSE. In fact, almost all studies on this exit poll show it is seriously flawed:

http://www.exit-poll.net/election-night/EvaluationJan192005.pdf and see also

http://www.scoop.co.nz/mason/stories/WO0411/S00140.htm

6) There is no explanation offered for why exit polls would favor Democrats.

FALSE: The subject was studied, and a very reasonable explanation offered even prior to this election:

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

7) All the (non-exit) polls prior to the election showed Kerry winning.

FALSE: In the 3 months prior to the election, MOST polls showed Bush winning the popular vote. Here are a few. Now, this is by far not comprehensive, but it does show enough to at least lead any logical person to question this myth about the non-exit pre-election polls all favoring Kerry.

GW/Battleground (1000 LV) 10/31 - 11/1 50% 46% Bush +4
Rasmussen (3,000 LV) 10/30 - 11/1 50.2% 48.5% Bush +1.7
TIPP (936 LV) 10/30 - 11/1 46.9% 44.3% Bush +2.6
FOX News (1200 LV) 10/30 - 10/31 46% 48% Kerry +2
CNN/USAT/Gallup (1573 LV) 10/29 - 10/31 49% 47% Bush +2
CBS/NY Times (643 LV) 10/28 - 10/30 50% 47% Bush +3
ARG (1258 LV) 10/28 - 10/30 48% 49% Kerry +1
Newsweek (882 LV) 10/27 - 10/29 51% 45% Bush +6
Battleground (1000 LV) 10/25 - 10/28 51% 46% Bush +5
CNN/USAT/Gallup (1195 LV) 10/22 - 10/24 52% 46% Bush +6
Los Angeles Times (881 LV) 10/21 - 10/24 49% 48% Bush +1
Newsweek (880 LV) 10/21 - 10/22 48% 47% Bush +1
Time (803 LV) 10/19 - 10/21 52% 47% Bush +5
GW/Battleground (1000 LV) 10/18 - 10/21 49% 45% Bush +4
Rasmussen (3,000 LV) 10/17 - 10/19 48% 47% Bush +1
FOX News (1000 LV) 10/17 - 10/18 48% 43% Bush +5
CBS News (678 LV) 10/14 - 10/17 47% 46% Bush +1
CNN/USAT/Gallup (788 LV) 10/14 - 10/16 52% 44% Bush +8
Time (865 LV w/leaners) 10/14 - 10/15 48% 48% TIE
Newsweek (LV) 10/14 - 10/15 50% 45% Bush +5
GW/Battleground (1000 LV) 10/11 - 10/14 49% 46% Bush +3
CBS News (760 LV) 10/9 - 10/11 47% 46% Bush +1
ICR (763 LV) 10/9 - 10/11 49% 46% Bush +3
CNN/USAT/Gallup (793 LV) 10/9 - 10/10 48% 50% Kerry +2
Rasmussen (3,000 LV) 10/7 - 10/9 50% 46% Bush +4
Time (886 LV w/leaners) 10/6 - 10/7 47% 46% Bush +1
GW/Battleground (1250 LV) 10/3 - 10/7 49% 46% Bush +3
Fox News (1000 LV) 10/3 - 10/4 48% 45% Bush +3
ICR (762 LV)** 10/1 - 10/5 51% 46% Bush +5
ARG (800 LV) 10/2 - 10/4 46% 47% Kerry +1
CBS/NYT (561 LV) 10/1 - 10/3 48% 47% Bush +1
Zogby (1036 LV) 10/1 - 10/3 46% 45% Bush +1
CNN/USAT/Gallup (772 LV) 10/1 - 10/3 49% 49% TIE
Newsweek (1013 RV) 9/30 - 10/2 46% 49% Kerry +3
Battleground (1000 LV) 9/27 - 9/30 51% 44% Bush +7
LA Times (1100 LV) 9/25 - 9/28 51% 46% Bush +5
CNN/USAT/Gallup (758 LV) 9/24 - 9/26 52% 44% Bush +8
IBD/TIPP (649 LV) 9/22 - 9/27 45% 46% Kerry +1
Time (877 LV) 9/21 - 9/23 49% 43% Bush +6
FOX News (1000 LV) 9/21 - 9/22 45% 43% Bush +2
Battleground (1000 LV) 9/20 - 9/23 50% 45% Bush +5
CBS News (931 LV) 9/20 - 9/22 50% 41% Bush +9
Zogby (1066 LV) 9/17 - 9/19 47% 44% Bush +3
IBD/TIPP (650 LV) 9/14 - 9/18 46% 43% Bush +3
CNN/USAT/Gallup (767 LV) 9/13 - 9/15 55% 42% Bush +13
CBS News (1088 RV) 9/12 - 9/16 50% 42% Bush +8
Battleground (1000 LV) 9/12 - 9/15 49% 45% Bush +4
IBD/TIPP (674 LV) 9/7 - 9/12 47% 47% TIE
Newsweek (1003 RV) 9/9 - 9/10 50% 45% Bush +5
Zogby (1018 LV) 9/8 - 9/9 47% 45% Bush +2
Time (857 LV) 9/7 - 9/9 54% 42% Bush +12

This is not the entire case against the "Exit polls prove fraud" claim, but it is a solid start. I am sure there are flaws with some of these claims, and I am happy to discuss them. However, I think any honest person, who is not just playing the "lets argue this because it's in our best interest for this to be wrong" game will at least have some doubts about the "Exit polls prove fraud" claim. I challenge people to do the right thing, the honest thing, and REALLY look into the issue with a more objective eye. I think anyone doing that will inevitably conclude that time is far better spent working on the evidence of actual fraud than this theoretical game of shuffling numbers until they come out the way we want them.

[RaulVB]

1. Rest and quit posting this

Not good and already rejected.

Take a break.

[Mistwell]

2. I've never posted most of this

How can something be "not good and already rejected" when I've never posted most of it?

Why don't you actually read what I wrote, and the links, instead of giving a knee-jerk response. What are you afraid of?

[RaulVB]

3. It didn't work, dude

Keep posting it if that makes you happy.

[Mistwell]

5. What are you afraid of?

Why are you so afraid to consider even the possibility that the exit polls claim might be flawed? What are you so afraid of, that you have totally closed your mind to any deviation from the party line?

Is it that you fear that you might have wasted a lot of time on an issue when your time could have been better spent focusing on actual evidence of fraud? That your investiment in the issue didn't pay off? If that is the case, it's time to swallow your pride. There is still plenty of time to refocus on actual election fraud. You should want to spend your efforts where they will do the most good, even if that means your ego takes a minor hit for a day or two.

[RaulVB]

6. I can SEE the fear in your posting. Don't blame ME FOR IT (n/t)

[Nederland]

25. Help me out

Can you give me a link to where this was debunked?

[Aussie_expat]

4. Well said Mistwell

I for one could not agree with you more.

But hey I'm a freeper too ..... just look at my post count.

[New Earth]

7. *looks at Aussie's post count*

yup, must be a freeper :D :hi:


hehehehehe

[RaulVB]

8. Ha, ha, Faye! (n/t)

[FogerRox]

9. Ok I'll work this guy

Of the 50 states, 40 deviated to Bush, and 10 deviated to Kerry. Some of the 10 deviating to Kerry were considered swing states prior to the election, and no explanation has been offered for why exit polls would deviate in this way. It’s usually just ignored by those claiming exit polls prove fraud.

WOW--deviation like that-- WTF/? Kidding--is 40 to 10 going to get a touch of perspective? is that historical?

And any of those polls that occur before the last debate are meaningless--to me. I forget when that was ---
Most of those polls seem fron the part of the MSM thats owned by 6 guys.
Gallup deserves mention --with new onwer--

[Aussie_expat]

12. Your case would be more credible .......

Your case would be more credible if there had not been considerable deviations since at least 1988.

A change in local demographics when combined with stagnant locations chosen for exit polling would easily add up to faulty data.

[skids]

96. Umm, no.

I'll grant that the history of exit poll performance is something that has been balefully neglected in the posts going around here.

But that assertion is just plain wrong. In NEP's report, they took their weighting formulas for the precincts, fed in the actual vote counts and found that the precincts they chose were biased half a percentage point in Bush's favor.

It would be nice if people on both sides of this argument would stop making statements that they cannot back up, and would expand their research a little. Those questioning the polls need to put their failure in context of past polls and both sides need to know the difference between the methods used in past polls, and those used in 2004, which were supposed to fix a lot of problems with the previous polls.

That is, if you want to take this discussion anywhere instead of still running around in circles.

Crack open the NEP report and read it, all of you, please.

[Aussie_expat]

100. Sources please ....... n/t

.

[Mistwell]

111. Response

"I'll grant that the history of exit poll performance is something that has been balefully neglected in the posts going around here.

But that assertion is just plain wrong. In NEP's report, they took their weighting formulas for the precincts, fed in the actual vote counts and found that the precincts they chose were biased half a percentage point in Bush's favor."

Ahem...you mean, they found a discrepancy WITHIN THE MARGIN OF ERROR. Which is the point.

[nodictators]

119. In the NEP report Mitofsky says it was "Democratic Overstatement"

Mitofsky also used that term on the PBS NewsHour on Nov. 5, 2004.

He used residual errors, Within Precinct Errors (WEP), for his conclusion that assertive Dems and reticent Repubs caused the errors in Ohio, Florida, and Pennsylvania.

I think he's full of BS! I believe it was fraud. I think it can be shown that it was. Many of his projections were correct in other states. Why were they so wrong in the big three above? Fraud!

[Yupster]

104. The fact that half the people approached by the pollsters

refused to answer the exit poll is enough to tell me they're not going to be accurate within a % or two.

[althecat]

10. To respond to two of your points...

Under 1. you produce raw exit poll data

The exit poll data used this time round was not raw... it was weighted data that was wrong.

The data you point to emerged fairly early on in this debate - Nov 17th - and was found to be incomparable as I recall. And as the NEP study confirms with the exception of the 1992 election the 2004 exit poll results were a lot worse than earlier ones.

Under 5. you cite

http://www.scoop.co.nz/mason/stories/WO0411/S00140.htm

This study was based on the adjusted exit poll numbers and has been very thoroughly discredited.

The second study is the NEP poll which was explicitly commissioned and published because the poll was so wrong this time around.

[Aussie_expat]

14. Please explain this statement.....

"The exit poll data used this time round was not raw... it was weighted data that was wrong."

and how was it weighted?

[althecat]

17. The data which was wrong was weighted data...

Have a look for yourself.. at the top of every page it says "weighted"

http://www.scoop.co.nz/stories/pdfs/Mitofsky4zonedata/

[Aussie_expat]

20. How was it weighted? n/t

.

[Mistwell]

30. Response

"The exit poll data used this time round was not raw... it was weighted data that was wrong."

FALSE. It was "partially-weighted" for lack of a better term. The only time the data was "fully-weighted" was with the final poll result. The early ones did not have all weighting factors applied...indeed, not all of the data was even in yet.

While I agree we do not know what was meant by "raw" in the prior years, it's a much closer state to "partially weighted" than "fully-weighted", given that you can look at the "fully-weighted" poll results from those years and see a very similar deviation to what we saw this year between the partial and full weighted results.

"The data you point to emerged fairly early on in this debate - Nov 17th - and was found to be incomparable as I recall."

You recall incorrectly. I've been in every single exit poll thread on this board since the election. Nobody has EVER shown the data to be "incomparable". The best arguement is the one you presented...that the 2004 early data is partially weighted, while the word used for the early polls from prior years is "raw". That's it, that is the sum total of that "incomparable" claim, and it just doesn't hold up. When you look at the deviations for this year between the partial vs full, and compare them to the raw vs full from prior years, it looks very similar...showing that they are, indeed, quite comparable. I never claimed it was a perfect comparison, just a hell of a lot closer to perfect than the claim that the partially-weighted is not at all comparable to the "raw" from prior years.

"And as the NEP study confirms with the exception of the 1992 election the 2004 exit poll results were a lot worse than earlier ones."

I disagree. Could you please cite language in the NEP study that "confirms" this?

"This study was based on the adjusted exit poll numbers and has been very thoroughly discredited."

FALSE. Read the study from starting from the section about "Blue Lemur". It does indeed speak to the non-adjusted exit poll numbers, and it has not been "througly discredited". Things are not discredited by fiat. If you have an arguement to make about it, I am listening. If you have a link, I'd read it. But as far as I can tell, that study has at least some persuasive analysis in it.

[althecat]

11. To reply to no 7.. - National Polls vs The Result

YOU STATE... "In the 3 months prior to the election, MOST polls showed Bush winning the popular vote. "

This is a rather selective use of data. As anyone could see Kerry was gaining on Bush towards the end of the campaign. Plus there were the problems with oversampling of republicans in many of the polls during this period.

For some more useful data on this question see..

http://www.electoral-vote.com/info/polls-2004.html

National polls - Oct 28th to Nov 1st

22 polls in all

8 show Kerry ahead
10 show bush ahead
4 show a tie

This is far from "Most" showing Bush winning. It was in fact as everybody knew on the night a dead heat going into the final days.

More useful would be an analysis of the state polls in the battleground states leading up to the election.

[Aussie_expat]

13. Even your your data shows ....

That is exactly "Most" showing Bush winning.

[althecat]

18. I would call it a tie myself

10 is not most of 22 it is in fact less than half...

[Aussie_expat]

21. Or.....

Who had the most between the Kerry and Bush?

[Name removed]

39. Deleted message

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

[Name removed]

112. Deleted message

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

[Mistwell]

35. Response

""In the 3 months prior to the election, MOST polls showed Bush winning the popular vote. "

This is a rather selective use of data. As anyone could see Kerry was gaining on Bush towards the end of the campaign. Plus there were the problems with oversampling of republicans in many of the polls during this period."

No, what is rather selective is what you chose to exclude from the quote you make from my initial post, as the very next sentence is "Now, this is by far not comprehensive, but it does show enough to at least lead any logical person to question this myth about the non-exit pre-election polls all favoring Kerry."

If you have a problem with the pre-election polls, fair enough. I was ONLY addressing the myth that the pre-election polls showed a Kerry victory. At best, they showed a Kerry trend but still a Bush victory, and that does not in any way add to the anti-fraud fight. I was not trying to claim that the pre-election polls were useful, or showed that Bush DID win the election...just that they don't support our anti-fraud claims, though I have seen the muth repeated more than once that they DID support the anti-fraud effort.

[TruthIsAll]

15. Number 4 is a LIE. Who said ALL? Cite that, Mistwell.

YOU:
4) All the State-by-State deviations between the exit polls came out to favor Bush in the actual counted vote.

ME:
No one ever said ALL states deviated to Bush.
43 out of 51, including DC, did.
THAT IS ENOUGH!

YOU:
Of the 50 states, 40 deviated to Bush, and 10 deviated to Kerry. Some of the 10 deviating to Kerry were considered swing states prior to the election, and no explanation has been offered for why exit polls would deviate in this way. It’s usually just ignored by those claiming exit polls prove fraud.


ME:
First of all EIGHT states deviated to Kerry.
Here are the states: HI,SD,ND,OR,KS,TN,TX,CA.
Only OREGON was a swing state.

OREGON VOTES WITH 100% PAPER MAIL-IN BALLOTS.
WHAT DOES THAT TELL YOU?
THAT KERRY STOLE IT?
IT TELLS ME THAT BUSH COULDN'T.

And the odds of 43 out of 51 states deviating to Bush is
****** 1 in 2.9 million ****

[TruthIsAll]

16. MYSTERY POLLSTER DISAGREES WITH YOU ON THE LAST 15 POLLS

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x302159#304256

[TruthIsAll]

19. KERRY LED IN 11 OF THE FINAL 18 POLLS. YOU SHOW ONE RV POLL OUT OF 50.

YOUR BUDDIES LOST THIS ARGUMENT YESTERDAY.
AND YOU ARE COMING BACK FOR MORE PUNISHMENT.

YOUR LIST WAS SHOWN TO BE PURE BS.
1. IT INCLUDED ONLY LV POLLS.
2. THERE WERE MANY DUPES BY THE SAME POLLSTER.
3. THEY DID NOT INCLUDE THE FINAL POLLS ONLY.
4. THEY ARE REDUNDANT, MISLEADING AND FALSE.

Your buddies presented the list and...
1. They don't match my 18 FINAL polls.
They conveniently forgot Fox and AP
2. Nine (9) polls don't agree with mine.
All but one in favor of Bush.
My numbers are right.
Who did the point shaving?

3. At least 6 of my 18 are LV polls
How many of yours are RV polls?


My 18 Polls
Poll Kerry Bush Kerry Bush Agree?
1 NA AP - - - 1020 49 46
2 NA TIPP - - - 1031 44 45
3 NA FOX - - - 1031 48 45
4 NA ABC - - - 1030 48 47
5 1031 NBC LV 47 48 1031 47 48
6 1030 ARG LV 49 48 1030 49 48
7 1031 GallupRV 48 46 1031 48 46
8 1031 Op 47 45 1026 44 46 NO
9 1024 LAT LV 47 47 1024 48 47 NO
10 1031 Marist 49 48 1031 49 48
11 1029 Nwk RV 44 48 1029 45 48 NO
12 1030 Pew 45 45 1030 46 45 NO
13 1021 Time 43 50 1021 46 51 NO
14 1028 Greenberg 49 46 1031 48 47 NO?
15 1031 CBS LV 46 47 1031 46 47
16 1031 Harris 45 49 1025 48 47 NO
17 1013 Zogby LV 45 46 1030 47 48 NO
18 1101 YouGovLV 50 47 1029 49 45 NO
46.69 47.08 47.17 46.89

[Mistwell]

107. You failed to rea what your arguing about

Since, if you HAD read what I posted, you would not be making this lame response to something I never said.

I never EVER claimed that those polls were representative, or indicative of a Bush victory.

Here is what I wrote: "All the (non-exit) polls prior to the election showed Kerry winning.

FALSE: In the 3 months prior to the election, MOST polls showed Bush winning the popular vote. Here are a few. Now, this is by far not comprehensive, but it does show enough to at least lead any logical person to question this myth about the non-exit pre-election polls all favoring Kerry."

Now, how does that in any way have to do with that wild response you just wrote?

I don't have any buddies here, and I am not claiming the same thing with these polls that you seem to think others have claimed. Perhaps you should actually read what your responding to rather than going off half-cocked and looking like a fool.

[Nederland]

24. Fair enough

What about the other 6 points? Are you just going to ignore them?

[Mistwell]

43. Response

"YOU:
4) All the State-by-State deviations between the exit polls came out to favor Bush in the actual counted vote.

ME:
No one ever said ALL states deviAtated to Bush."

If you think you've never seen that claim, then fair enough - then it should be a moot point in your book. I have seen the claim, several times. While it would make you all uptight to see me talking about a claim that you think is moot is beyond me.

"43 out of 51, including DC, did."

DC does not vount in the election of the President, and so should not be part of this discussion. If you want to debate statehood for DC, be my guest. It's a non issue for me.

"THAT IS ENOUGH!"

Enough for WHAT?

"ME:
First of all EIGHT states deviated to Kerry.
Here are the states: HI,SD,ND,OR,KS,TN,TX,CA.
Only OREGON was a swing state."

My understanding was it was ten states, but for the sake of the argument lets leave it to 8. Of those, South Dakota (Dashelle's state, where he BARELY lost), Oregon and Hawaii (where Cheney and Clinton visited on behalf of the candidates JUST before the election) were all swing states, and Tennesse wasn't far behind in being a swing state. If there is a conspiracy that CAN BE SHOWN BY THE EXIT POLL NUMBERS, then all of those states I just mentioned would have ranked WAY ahead of some of the other states you claim "prove" fraud through exit poll data. Explain why meaningless states would deviate while meaningful states would not if the exit polls can reasonably shown to "prove" a conspiracy of fraud?

"OREGON VOTES WITH 100% PAPER MAIL-IN BALLOTS.
WHAT DOES THAT TELL YOU?
THAT KERRY STOLE IT?
IT TELLS ME THAT BUSH COULDN'T."

Mail-in ballots are both 1) easier to defraud, because ANYONE can fill out the ballot and signatures are NOT checked, and 2) Still counted by machines. So, it doesn't tellme any such thing. I never claimed Kerry stole ANYTHING. I claimed that exit polls do not help in proving that Bush stole the election. Focusing on ACTUAL fraud is what can do that.


"And the odds of 43 out of 51 states deviating to Bush is
****** 1 in 2.9 million ****"

No, my single-minded friend, it does not. Not when the same exact thing happens every Presidential election. What that means is the odds are in FAVOR of it happening in this election...since empirically that is how the exit polls seem to work. Some day, I hope, you will actually respond to that issue rather than hiding from it. I know...the exit poll issue, in your mind, is the only thing that gets you the attention and adoration of your fellow DU'ers, and you are afraid that a focus (by you) would result in a decrease in the attention paid to you. That's an ego issue...one you will some day have to get over. Might as well make it now...while you can still make a difference.

[TruthIsAll]

22. YOU JUST MADE THE CASE FOR ENDEMIC REPUKE FRAUD.

The exit polls always favor the Dems.

But the Repukes always get more votes than they should.
Sometimes enough to steal the election.

Ballot spoilage and fraud are not necessarily mutually exclusive.


Year / Exit Poll / Results / Dem Lead / Dem Actual
1988 / Dukakis: 50.3% Bush: 49.7% / +0.6% / -7.7%
1992 / Clinton: 46% Bush: 33.2% / +12.8% / +5.6%
1996 / Clinton: 52.2% Dole: 37.5% / +14.7% / +8.5%
2000 / Gore: 48.5% Bush: 46.2% / +2.3% / +0.5%

[Name removed]

48. Deleted message

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

[TruthIsAll]

23. MITOFSKY A LIBERAL? MYSTERY POLLSTER A DEMOCRAT?

YOU:
The leading expert on the subject is Warren Mitofsky, the father of exit polls. Mitofsky is also a lifelong liberal and "apparently holds no brief for Bush." He's the guy who ran this exit poll as well. He's the one that would be called to the stand to testify about this issue.

ME:
PURE BS.
Where was Mitofsky when Conyers wanted him to testify?

[Aussie_expat]

26. Mitofsky

You use his figures to support your data when it suits you then you turn around and trash him in the next post.

Sloppy.

[TruthIsAll]

31. MITOFSKY said 1.0% MOE for the National Exit Poll. Case Closed.

.

[qwghlmian]

32. So - when Mitofsky says something,

the case is closed? Does that hold for everything Mitofsky says or are you cherry-picking again?

[Name removed]

34. Deleted message

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

[qwghlmian]

36. Nice evasion Faye

does not answer my question, though, does it?

:hi:

[New Earth]

38. nope

but it answered mine :hi:

[Nederland]

41. They ignore questions they can't answer

Its a long held pattern.

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=292938

[New Earth]

46. who's they?

i don't know anything about exit polls, i'm just goofing off, sorry :D

[Nederland]

49. People that think exit polls are indicative of fraud (nt)

[New Earth]

52. there are plenty of other things

in this election that are indicative of fraud. proving or debunking the exit polls honestly doesn't mean shit.

UK doesn't even use them, nor do many other counties.

[Mistwell]

108. FAYE MUST BE A FREEPER!

Faye, how dare you waiver from the party line and express doubt about the value of the exit polls! You must be a Freeper, a Republican, a liar and a traitor!

Oh yeah folks, while people like TIA were shifting numbers around in an excel spread sheet to make square shapes fit into round holes and claiming he was helping the anti-fraud cause from his warm office chair, Faye actually WENT to Washington DC, in the snow and rain and cold, to protest fraud, and generally posts about ACTUAL fraud events!

I think it's obvious who the real anti-fraud folks are, to anyone who cares to think about it.

[Mistwell]

53. It took me a while too...

But eventually you figure out when Faye is being passionate and serious, and when Faye is poking you for being overly serious.

I think this was one of the later examples.

[Name removed]

56. Deleted message

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

[Melissa G]

58. Love the graphic Faye ROFLMAO! These guys need a life...n/t

[TruthIsAll]

42. MP (Kerry led 11 of 18 polls) and Mitofsky (1% MOE) disagree with you.

http://www.mysterypollster.com/main/2004/11/lessons_likely_.html

Why don't you go argue with them?

[Nederland]

45. Question

Do you trust the opinions of Mystery Pollster, or is this another one of those case where you are selective about when you trust people?

[qwghlmian]

47. You evaded my question -

why would that be :shrug:

[Mistwell]

54. MOE was for demographics only

If you have ANY evidence that the MOE for those polls was supposed to be accurate for the vote count, and not just the demographic data, then post it. Warren Mitofsky says it was just for demographics, and the ACTUAL early poll data always had that disclaimer (though it was removed often by newspapers and TV shows that quoted the data).

[Mistwell]

51. Why spread more myths?

He said that MOE FOR THE DEMOGRAPHICS DATA ONLY in the pre-final exit polls. Why do you pretend he ever said the MOE was for the vote count? You know it's false, and yet it's a basic (flawed) assumption in all of your statistical posts.

If you think differently, then you cite to me one single time where Warren Mitofsky ever said the earlier exit poll MOE was accurate FOR THE VOTE COUNT.

[TruthIsAll]

55. Do you know how to derive the vote count from the demographics?

If the demographic (category) percentages are correct to within 1%, and if you could use this data across a full set of categories to derive the vote percentages, wouldn't the vote percentages also be accurate to within 1.0%?

YES
or
NO?

[Mistwell]

109. Nobody can

It is IMPOSSIBLE to derive how people voted from dempgraphic data alone. You know that. I cannot, you cannot, and nobody can do it. You would need other accurate data to use with it.

If I tell you 10 people voted at a location: 4 were women while 6 were men; 2 were african american and 1 was hispanic and 1 was asian and 6 were caucasian; 3 appeared to be in their 20's, 3 in their 30s, 3 in their 40s, and 1 in their 50s - now, can you tell me how people voted? No, of course not. The answer is a resounding NO, you CAN NOT derive the vote percentages from the demographic category percentages, and the margin of error does not in any way carry over to the vote percentages since THEY ARE NO MEASURING THAT.

[Nederland]

27. Response

As I have repeated pointed out, you seem to be very selective about Mitofsky. When it comes to trusting Mitosfky to produce an accurate exit poll for the election, he is a god that can do no wrong. When it comes to trusting Mitosfky to explain why you can't use his exit poll data in the manner that have, he is an idiot.

So which is it TIA? Is Mitofsky a god or an idiot?

[anaxarchos]

44. You have just succinctly summed up the industry's view of Mitofsky...

Like most of us, he is both. As the inventor of exit polls and the designer of very efficient EP methodology, Mitofsky's talent is widely accepted.

As an apologist for results he doesn't like, as a panderer, and as a defender of his business, he is widely panned -- well before this election too(you can google it if you like).

This is not a contradiction.

[Nederland]

57. Thank you

For a polite response.

The thing I don't get is why, if people believe he is a talented pollster, would they reject his calculations of the MOE when such calculations are done before polls are conducted and therefore cannot fall under the category of "results he doesn't like"?

[anaxarchos]

87. I haven't really followed this thread...

I'll go back and read it later. But from what little I saw, what's being said is not possible. The MOE refers to the randomness of the sample ONLY. Think about it. How can you have one MOE for the poll and another for the survey questions? They are entirely codependent. I might be missing the point above and if so, I'll repost.

What I can tell you about Mitofsky's exit polls is that they are sprinkled with a million disclaimers and CYAs that are meant for commercial consumption only. Much is made of the statements that they are "not intended" for calling elections but if that were really true, none would ever buy his service. Mitofsky is widely dismissed for trying to have it both ways at his convenience... ironically as most people praise his accuracy.

Think about MOEs too. Most pre-election polls claim an MOE of around 3-4% (sometimes less). Exit polls with a much larger sample and on the spot, have to be MUCH more accurate (they are considered to be a huge breakthrough). 1% is widely accepted. Otherwise, the election result/exit poll variations would not have caused half the controversy that they have.

[Yupster]

106. The problem anaxarchos

is the non-response rate for the people asked to fill out an exit poll survey is so astronomicly high. Somewhere around 50 %.

Yes you can ask 240 white men over age 65 to get your 100 white men over age 65 you're looking for to respond, but in the end was it a random sample of white men over 65, or did they self select themselves?

What if liberal white men over 65 are slightly less rude than conservative white men over 65? Then too much of the conservative sample has thrown itself away and the liberal sample would be therefore over-represented.

Is there evidence that this happened? Yes. The studies that have been done long before this election show that liberals are slightly more willing to be polled than conservatives by exit pollsters.

That would explain why Democratic candidates seem to always do well in exit polls.

In my opinion, if the sample is allowed to self eselect itself and half choose not to take part, then a 1 % margin of error can't be possible.

[anaxarchos]

121. Actually that's not true...

Most of the reasons for non-response have nothing to do with self-selection. They are mostly technical or innocent. The most significant factor is probably the length of the survey (which is also the largest reason that response has been going down, though very slowly, over the last 10 years). Other reasons have to do with the logistics of U.S. elections and a host of other factors which have nothing to do with a political pre-disposition. U.S. exit polls have always had a significant non-response differential versus European countries (sometimes 20 points or more).

This figures into the calculation of MOE which is slightly more complicated than has been presented in this thread. It is a multi-tier approach which first attempts to assure randomness and then calculates mathematical variability.

The operative word in your post is "slightly". A "slight" variation which changes "slightly" over time is precisely what the exit polls are very good at correcting for. And, yes, MOEs of 1% (roughly) are possible.

In order to have the impact you describe, the change would have to be "all at once", in a way and at a magnitude that is not anticipated by the exit polls. This is precisely what is claimed for the one "genuine case" of false-response that is most often cited in the literature (I have my doubts about that one too). There is no empirical evidence of any such "dramatic change" in the 2004 election.

I do not agree yet that "Democrats always do well in exit polls" in this context or that, if that is true, it is related to this issue. I just haven't had time to look at it but on the face of it, it does not seem possible. I understand the CLAIM but that's very different.

This reasoned discussion is a pleasant change ;-)

[Mistwell]

50. he DISAGREED with Conyers

You know...you do not have to agree with other liberals to be considered a liberal. People are allowed to form their own opinions, rather than marching in lock step with each other like the good fascists you seem to want all liberals to be.

Warren Mitofsky is his own man. He's always been a liberal, and he doesn't like Bush. However, he has a job to do, and he's trying to do without his personal bias getting in the way...something you cannot say about yourself. The Democratic party is the party of science. It's the party of reason. It's the party that does not hide from difficult truths, like the threat of global warming and the promise of stem cell research. It is entirely consistent with that theme that Warren Mitofsky supports TRUE statistical analysis, which is the tradition of a TRUE liberal, even if his conclusions result in bad news for the political causes he supports. Can you honestly say the same thing?

[TruthIsAll]

60. Oh, stop already. He refused to testify when Conyers asked him to.

Why?

Will he testify now and talk about Reluctant Republicans who would not speak to the exit pollsters?

[Mistwell]

110. I would refuse too!

Mitofsky DISAGREES with the entire premise of the hearing, the hearing was voluntary and would cost Mitofsky money. I would tell him to pund sand as well!

Hey TIA...I'm having a hearing here in two days in Los Angeles...it's about statistics. We are trying to prove a subject you disagree with, but think your testimony will be helpful. Would you mind flying out for the day, at your own expense, to do us this favor which you disagree with?

Riiiight.

[TruthIsAll]

28. THE NATIONAL EXIT POLL MOE IS 1.0%. MITOFSKY SAID SO HIMSELF.

YOU:
"He has never, EVER claimed that the margin of error in his early exit polls (the ones that TIA likes to quote constantly) are set for detecting election fraud or even set for the vote itself. The MOE is set ONLY for the Demographic data. That disclaimer is included with every single early release of his exit poll data (though not always repeated by newspapers and television programs that use his data). It is a total myth that the margin or error in the pre-final exit polls are accurate, or ever meant to be accurate, for the vote itself".

ME:
What a straw man argument. Who cares what he claims as to the purpose of the exit polls. WE can draw our OWN conclusions based on analysis of HIS own numbers.

Why did Edison/Mitofsky claim a 1.0% MOE in the National Exit poll as displayed on the Washington post site in the notes accompanying the results of the poll?

[Aussie_expat]

29. Your answer is in ....

your own post ... "The MOE is set ONLY for the Demographic data."

[TruthIsAll]

33. You don't have a clue.

[Aussie_expat]

37. Well then that makes two of us .......

but I'm not really claiming to know it all am I?

[TruthIsAll]

40. Who cares whether you claim or don't claim to know it all?

All we know is what you say.
And that tells us enough.

[TruthIsAll]

70. Do you know how to derive the vote percentage from the demographic data?

And if the demographics are correct to within 1%, if you could derive the vote percentage from the demographic data, would the vote percentage be accurate to within 1% also?

YES
or
NO?

[Mistwell]

113. No - nobody cab

I've answered you three times now. The answer is no. It is impossible to derive the vote percentage from the demographic data. They are not linked.

[Mistwell]

59. Response

"What a straw man argument. Who cares what he claims as to the purpose of the exit polls. WE can draw our OWN conclusions based on analysis of HIS own numbers."


Yeah, see, that is a misrepresentation of what happens at exit polls.

They have two people who count...one person counts the approximate age, ethnicity, gender, and other data of the people exiting the polls. The other asks specific people specific questions about how and why they voted. The MOE was set to the demographic data, and NOT THE VOTING DATA. You cannot "draw your own conclusions" if your basic assumption about what that MOE is tied to is seriously flawed.

"Why did Edison/Mitofsky claim a 1.0% MOE in the National Exit poll as displayed on the Washington post site in the notes accompanying the results of the poll?"

Because the Washington Post site did NOT list the entire disclaimer, as Mitofsky said. It was NOT supposed to be released, and the entire disclaimer was NOT included, and Mitofsky said that, repeatedly and loudly, right away when it happened. You just choose to ignore that and pretend the MOE meant something it did not.

[Melissa G]

63. Because someone puts in a liability disclaimer for their data

does not mean the data is not worth using. My accountant does it all the time on my books because his monthly work is not as exhaustive as it could be. It does not mean the data is not useful. You are the cherry picker and a sloppy one at that.

[Mistwell]

71. i didnt cherry pick anything

Mitofsky has different margins of error for different types of data. He says so. He said so before any of those polls were released that he uses a drastically difference confidence level to compute the margin of error for the actual vote than he does for demographics data. It's not just a disclaimer, it's a basic statement of the assumptions of the analysis you are looking at. And it's something that people like TIA never talk about because it is a flaw in his basic assumptions of his own statistical analysis.

[Karenca]

61. Why did you post this?

Why is it bothering you so much that:
"I am very tired of all the myths surrounding exit polls "


We've said it , oh, maybe 10,000 times; This forum is for those that want to discuss the fact that this election was stolen, AND THE EXIT POLLS WERE DOCTORED IN FRONT OF OUR EYES, AND NO, PEOPLE DO NOT, LET ME REPEAT, DO NOT LIE TO THE EXIT POLLSTERS BECAUSE THEY ARE ASHAMED, AND YES, EXIT POLLS WERE ALWAYS USED AS A WAY TO CHECK RESULTS, UNTIL THE DICTATOR TOOK OVER OUR COUNTRY!!!!!

Oh. yeah, those September polls are useless, you know that.

[qwghlmian]

62. You are aware of the fact

that if you post a statement in all caps, that alone does not make it true, right?

[Melissa G]

66. about the nineteenth time one of us has to reiterate the

SAME REFUTING POINTS. it does tend to send folks to caps...

[qwghlmian]

68. Just because you repeat something

nineteen times does not make it true either...

Example: Karenca claims that exit polls were "always used as a way to check results" (I removed the caps). I am sure you also agree with that statement. Can you give any citation to this "fact" - that in the US exit polls are "always used as a way to check results". Just one such citation will do.

[Karenca]

80. You are aware that

I understand the cryptic message in your screenname,

[qwghlmian]

81. ROTFL -

yes, being a Neal Stephenson fan is incredibly "cryptic".

[Karenca]

90. A late-night contribution by the NEP to the rewriting of history......

NEPs data transmitted by CNN on the nite of 11/2 & early AM of 11/3—suggest very strongly results of exit polls were fiddled late on 11/2 to make their #'s conform w/tabulated vote tallies.
Its important to remember how big the discrepancy was between early vote tallies & early exit poll figs. By the time polls were closing in the east the vote-count figs. pblshd. by CNN showed * leading Kerry by a huge 11%. At 8:50PM, * was credited w/6,590,476 votes, & Kerry w/5,239,414. This margin gradually shrank. By 9:00PM, * had 8,284,599 votes, & Kerry 6,703,874; by 9:06, * had 9,257,135, & Kerry had 7,652,510, giving * a 9% lead, w/54% of the vote to Kerrys 45%.

At the same time, embarrassingly, NEP figs. reported by CNN showed Kerry holding a decisive lead over *. At 9:06, exit polls indicated womens votes (54% of the tot.) were going 54% to Kerry, 45 to *, & 1 to Nader; men (46% of the tot.) were 51% *, 47% Kerry, 1 Nader. Kerry was leading * by nearly 3%.

Early exit polls appear to have caused some concern to the people at NEP: a gap of 12-14% btwn tallied results & exit polls can't inspire confidence in the legitimacy of an election.

One can surmise that instructions of 2 sorts were issued. The election-massagers working for Diebold, ES&S & the other suppliers of blackbox voting machines may have been told to go easy on their manipulations of back-door ‘Democrat-Delete’ software: mere victory was what the * campaign wanted, not an implausible landslide. & the # crunchers at NEP may have been asked to fix up those awkward exit polls.

Fix them they did. When the polls were last updated, at 1:36AM, mens votes(still 46% of the tot.)had gone 54% to *, 45 to Kerry, 1 to Nader; womens votes (54% of the tot.) had gone 47 to *, 52% to Kerry, and 1 percent to Nader.

How do we know the fix was in? 'Cause the exit poll data also included the tot. # of respondents. At 9:00, this # was well over 13,000; by 1:36 it had risen by less than 3%, to a final total of 13, 531 respondents—but with a corresponding swing of 5% from Kerry to * in voters’ reports of their choices. Given the increase in respondents, a swing of this size is a mathematical impossibility.

The same pattern is evident in the exit polls of OH & FL.

7:32PM, CNN was reporting the following exit poll data for OH. Women voters (53% of the tot.) favored Kerry over * by 53% to 47%; males (47% of the tot.) preferred Kerry over * by 51% to 49%. Kerry was leading * by a little more than 4%. But by 1:41AM, when the exit poll was last updated, a dramatic shift had occurred: women voters had split 50-50 in their preferences for Kerry & *, while men had swung to supporting * over Kerry by 52% to 47%. The final exit polls showed * leading in OH by 2.5%.
At 7:32, there were 1963 respondents; at 1:41AM there was a final total of 2020 respondents. 57 additional respondents must all have voted very powerfully for *— while representing only a 2.8% increase in the # of respondents they managed to produce a swing from Kerry to * of 6.5%.

FLA: exit polls have been tampered with in a similar manner. At 8:40, CNN was reporting exit polls that showed Kerry & * in a near dead heat. Women(54% of the tot.)preferred Kerry over * by 52% to 48%, men(46% of the tot.)preferred * over Kerry 52% to 47%, poll, made at 1:01AM showed a diff. pattern: women now narrowly preferred * over Kerry, by 50 to 49%, while men preferred * by 53 to 46%. These #s gave * a 4% lead.
The # of exit poll respondents in FLA had risen only from 2,846 to 2,862. Again, a powerful numerical magic was at work. A mere 16 respondents—0.55% of the tot. number—produced a 4% swing to *.

What we're witnessing, the evidence suggests, is a late-night contribution by NEP TO THE REWRITING OF HISTORY.
It'S possible at some future moment questions about electoral fraud in the 2004 election might become insistent enough to be embarrassing. The pundits at that point will be able to point to the NEPs final exit poll figs. in the decisive swing states of FLA & OH-& marvel how closely they reflect NEPs vote tallies.

[qwghlmian]

92. This post shows a complete ignorance

of the way exit polls are done, how the data is collected, and how it is weighted. At 9:00 it was 13,000 respondents. At 1:36 it has risen to 13,531 respondens. But guess what - they were not the same respondents. Some were thrown out. Some were added. You really should educate yourself on how exit polling works before posting misinformation.

[Karenca]

93. Ha you're so funny...You call me ignorant, and yet that article came from

The Centre
for Research
on Globilization"!!!

I am laughing so hard right now!! :spank:


http://64.233.161.104/search?q=cache:EbcmMNLf5DsJ:www.globalresearch.ca/articles/KEE411A.html+why+have+exit+polls&hl=en&ie=UTF-8

KERRY!!
KERRY!!!
KERRY!!!!

:dem: :dem: :dem: :dem: :dem: :dem:

[qwghlmian]

94. Wherever it came from,

it shows complete ignorance of exit polling basics.

[Karenca]

95. ha ha ha ha ha ha

can't help it...
too funny :silly: :eyes: :silly:

[Mistwell]

114. I guess you never bothered to follow up on this

As NEP will tell you straight out what was done, and told you that before they even did it. They ALWAYS adjust the final numbers, in the final weighting, to include some of the actual vote. It's no secret conspiracy to "fiddle" with the numbers. It's right there in the formula they always use! That is, my friend, what we've all been debating. NEP will tell you their numbers ALWAYS skew Democratic until the final adjustment. There are studies that explain why, and empirical data shows it does in fact happen that way every single election since the invention of the exit poll.

But you pretending like this is some big, secret attempt to re-write history...that just so happens to be PUBLISHED ON THEIR WEBSITE AND IN THEIR METHODOLOGY DIRECTLY FOR EVERYONE TO READ, EVEN BEFORE THE EXIT POLL IS CONDUCTED, is silly.

[Mistwell]

73. So...you didn't read my post

Nobody who actually read my post would say what you just said.

I gave the reasoning.

You know it is a flat out misrepresentation to claim that exit polls were always used as a way to check results in the US. They have NEVER been used for that purpose...and are no where NEAR the sampling size to get that kind of data. You show me one single expert in the field who EVER said they were ever used for that purpose in the US.

Nobody here claimed anyone lied to exit pollers. See...you really didn't read the post...you're just marching in lock step with the fasist party line that tells you not to think for yourself.

[Karenca]

85. I find it amusing that you're calling Du'ers, "the fasist party".

<<<<See...you really didn't read the post...you're just marching in lock step with the fasist party line that tells you not to think for yourself.>>>>

BTW, the correct spelling is: 'fascist', not "fasist".

[Mistwell]

115. Gee thanks

I don't call DU'ers that by the way...as I *AM* a DU'er (heart and soul). I call people who march in lock step with TIA on this issue "fascists".

As for my spelling, hey, I know it is crap. If that is my worst flaw, I can live with that.

[truckin]

117. Mistwell, have you looked Kathy Dopp's analysis of Florida?

Specifically, the correlation she asserts between the optical scan machines and a disproportional vote to Bush based on registion numbers. I was wondering what your opinion was of her analysis or if you have even reviewed it. Thanks.

[qwghlmian]

118. Correlation does not equal causation

there are lots of northern Florida counties where the fact that someone is a "registered Democrat" does not mean that he/she does NOT consistently, election after election, vote for Republican candidates for President. It has been shown that in those counties the results do not differ drastically in 2004 from what they were in 2000 or 1996. So - unless you claim that there was widespread fraud in 2000 as well (and keep in mind, Florida was meticulously hand-recounted in 2001 by the MSM consortium, so such fraud would have been found) there is nothing to that analysis.

[truckin]

120. Never made any claim, just asked for Mistwell's opinion

[anaxarchos]

88. I'm a big fan of your posts on this subject...

What surprises me is not that you went CAPS but that you have not yet permanently glued down your CAPS LOCK key..... ;-)

[Name removed]

64. Deleted message

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

[New Earth]

65. :O!!!!

you can't call him a freeper for not believing the exit polls! that's so intolerant of you!!!:crazy:

[Melissa G]

69. would he call him one for posting the same garbage day after

day after day?

[Name removed]

72. Deleted message

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

[Melissa G]

79. Oh I am sure we could have a poll here at D U on the likelihood

of TIA being a freeper.... It would rank up there with the likelihood of **** having won the election without FraWd. OH yeah! TIA's point!

[Name removed]

74. Deleted message

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

[Name removed]

78. Deleted message

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

[Name removed]

75. Deleted message

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

[Mistwell]

67. More on Margin of Error

You can read more on Margins of Error here:

http://www.stanford.edu/group/gradethenews/dreamhost%20files/pagesfolder/Pollstory1.htm

In summary: "The margin of error, however, may not mean what most people think. It’s not the total deviation one might expect from the true public opinion."

And yet, that is the basic assumption in ALL of TIA's statsitical analysis...that margin of error is a measure of the total deviation one might expect from the true public opinion.

Also, when Mitofsky is trying to analyze the actual vote, and not the demographics, he uses a different confidence level (which is one of the major factors used to come up with a margin of error) than the one used for the early exit polls (which were set to demographics analysis) that TIA constantly posts.

http://www.mysterypollster.com/main/2004/12/what_is_the_sam.html

"NEP requires a much higher level of confidence to project winners on Election Night. The reasoning is simple: At a 95% confidence level, one poll in twenty will produce a result outside the margin of error. Since they do exit polls in 50 states plus DC, they could miss a call in 2-3 states by chance alone. To reduce that possibility, NEP uses a 99.5% confidence level in determining the statistical significance of its projections (something I was able to confirm with Joe Lenski of Edison Research, who co-directed the NEP exit polls with Warren Mitofsky)."

[TruthIsAll]

77. Do you know how to derive the vote percentage from the demographic data?

Do you know how to derive the vote percentage from the demographic data?

And if the demographics are correct to within 1%, if you could derive the vote percentage from the demographic data, would the vote percentage be accurate to within 1% also?

YES
or
NO?

[Mistwell]

86. You cannot

You CAN NOT derive how people voted based on the demographic data alone. You MUST have additional accuarte data to combine it with. And you know that quite well.

[TruthIsAll]

91. Well, you are wrong. I will show you very quickly why.

54% of women voted vs. 46% men
54% of women voted for Kerry. Only 45% for Bush.
52% of men voted for Bush. Only 47% for Kerry.

The gender demographic (or characteristic, as Mitofsky calls it) was stated to be accurate to within 1.0%. Therefore, the final Kerry winning percentage is also correct to within 1.0%, based on this simple calculation:

Kerry % = .54 *.54 + .46 *.47
Bush % = .54 *.45 + .46 *.52
You can do the multiplications yourself.

Of course, the 54/46 split changed when the poll sample hit 13,660.
It magically became 52% women/48% men. And of course their voting percentages changed in Bush's direction, as well.

The same argument can be applied to the Party ID breakdown.
Originally, in the 13,047 poll, it was 38%Dem/35% Repub/27% Ind.

Of course, the mix also changed when they reached 13,660 respondents:
It magically became 37% Dem/37% Repub/26% Ind.

And the percentages of Dems,Repubs and Independents voting for Kerry or Bush changed as well.

[Mistwell]

116. Wrong!

"52% of men voted for Bush. Only 47% for Kerry."

That is NOT demographic data. Just because it has as an element demographic data, does not in fact make the entire thing demographic data. The MOE was linked to the first two...the percentage of people who voted based on gender...but NOT linked to HOW they voted.

If you think differently, why don't you write Mitofsky and ask him yourself.

[TruthIsAll]

89. Straw-man. We are NOT talking about projecting winners.

YOU:
"And yet, that is the basic assumption in ALL of TIA's statsitical analysis...that margin of error is a measure of the total deviation one might expect from the true public opinion".


ME:
You obviously have no comprehension what the MOE is. You are either very sloppy in your wording, or very ignorant of statistics.

The MOE is NOT a measure of the total deviation one might expect from the true public opinion.

The MOE is the INTERVAL around the SAMPLE MEAN in which the TRUE POPULATION MEAN can be expected 95% of the time, or within 1.96 standard deviations of the SAMPLE MEAN.

Do you know what the standard deviation is?
Do you know how to calculate the MOE given the standard deviation?
Do you know how to calculate the MOE based on the sample size?

Your ignorance is pathetic.
Go back to school and study Stat 101.


YOU:

"NEP requires a much higher level of confidence to project winners on Election Night. The reasoning is simple: At a 95% confidence level, one poll in twenty will produce a result outside the margin of error. Since they do exit polls in 50 states plus DC, they could miss a call in 2-3 states by chance alone. To reduce that possibility, NEP uses a 99.5% confidence level in determining the statistical significance of its projections (something I was able to confirm with Joe Lenski of Edison Research, who co-directed the NEP exit polls with Warren Mitofsky)."

ME:
WRONG. WE ARE TALKING ABOUT EXCEEDING THE EXIT POLLS BY THE MARGIN OF ERROR IN 15 STATES.

THE MYSTERY POLLSTER HAD IT ALL WRONG. THIS IS NOT ABOUT THE 99.5% CONFIDENCE LEVEL USED IN CALLING THE WINNER OF A STATE ON ELECTION DAY. HIS STATEMENT WAS INEXCUSABLE. EITHER IT WAS A DELIBERATE STRAWMAN DESIGNED TO MISLEAD MILLIONS WHO ARE IGNORANT OF THE BASIC STATISTICS OF POLLING, OR HE WAS INADVERTANTLY DUPED INTO MAKING THE STATEMENT. FOR ONE OF HIS SOPHISTICATION, THAT WAS A REMARKABLE ERROR TO MAKE.

THE POLLING MARGIN OF ERROR IS ALWAYS PREDICATED ON A 95% CONFIDENCE INTERVAL. THIS HAS NOTHING WHATSOEVER TO DO WITH CALLING THE WINNER OF A STATE.

IT SO HAPPENED THAT FOUR STATES FLIPPED FROM KERRY TO BUSH OUT OF THE TOTAL OF 15 WHICH DEVIATED BEYOND THE MOE. IT JUST SO HAPPENS THEY WERE THE CRITICAL STATES: OH, FL, NM, IA.

THE DEVIATION OF 15 STATES TO BUSH, ALL BEYOND THE MOE, IS THE ISSUE. THE PROBABILITY OF THIS OCCURRENCE IS 1 OUT OF 1 TRILLION.
I WILL REPEAT THIS MATHEMATICAL FACT UNTIL HELL FREEZES OVER.

AND, OF COURSE, 43 OUT OF 51 STATES DEVIATED TO BUSH.
THE ODDS ARE 1 IN 2.9 MILLION AGAINST THIS OCCURRING BY CHANCE ALONE.
THE MARGIN OF ERROR IS NOT AN ISSUE HERE. WE ARE TALKING ABOUT 43 DEVIATIONS TO BUSH VS. 8 TO KERRY. PERIOD.

YOU GUYS HAVE NOTHING LEFT TO STAND ON.
YOUR ARGUMENTS HAVE ABSOLUTELY NO MERIT.

[Name removed]

76. Deleted message

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

[Mistwell]

83. Freeman was shown to be wrong actually

First, go here:

http://www.mysterypollster.com/main/2004/12/what_is_the_sam.html

"Freeman has recently updated his paper with calculations that rely on this 30% estimate. However, following my post I received an email from Nick Panagakis informing me that the 30% estimate was out of date (Panagakis is the president of a Market Shares Corporation, a polling firm that has conducted exit polls in Wisconsin and other Midwestern states). Panagakis had checked with Warren Mitofsky, director of the NEP exit poll, and learned that the updated design effect used in 2004 assumed a 50% to 80% increase in error over simple random sampling (with the range depending on the number of precincts sampled in a given state). Blogger Rick Brady (Stones Cry Out) has subsequently confirmed that information in an email exchange with Mitofsky that he posted on his website.

Thus, the calculations in Freeman’s revised paper continue to understate the sampling error for the 2004 exit polls (more on this in a post to follow)."

Next, screw you for calling me not "a real DU'er". I am as real as you. I am as anti-fraud as you (probably more so). I just want people to work on the actual fraud and not this BS exit poll crap that is so easy to disprove is harms all the other DU'ers trying to end election fraud.

"Furthermor, your Ukraine argument completely contradicts the entire point you are trying to make."

Oh really? Did you care to articulate your contention, or are you just going to fiat a "I know you are but what am I" and leave it at that?

"Also, the Roper Center posted an actual nationwide exit poll from the Dukakis election that projected a Bush victory by anywhere from 3 to 5 points, so that exit poll was reasonably close to the actual result."

Are you even following this debate? That was the final exit poll...the accurate one. The exit polls are ALWAYS accurate WHEN THEY ARE FINALIZED, it's when they are pre-final (like the one we are debating) that they always mess up. In fact, what you just posted was even IN MY OWN POST.

"Additionally, it might just be that massive electronic fraud has been occuring for a lot longer than anyone realizes."

Oh really...so before there was electronic voting, electronic voting was still to blame? And, all that fraud failed to actually change a single election result? Come on man...you know that's an incredibly silly claim. Stop picking the only possible explanation that matches your world view, and at least consider for a moment that your world view might be wrong this one time.

[truckin]

102. You make points that should at least be considered and discussed

I do not know enough about the exit polls to know how much of an indicator of election fraud they are but Mistwell makes some valid points that should be discussed and considered. I am new to this forum but I have quickly noticed that 2 frequent thread starters are very thin skinned and if you question any of their assumptions they get very defensive. It is too bad that they cannot consider any viewpoints that do not totally support their conclusions, it detracts from their credibility.

That being said, we all must work towards getting rid of the electronic touch screen machines and demand voter verified paper ballots for all elections. Our time could be spent more constructively doing this instead arguing over exit polls.

[livvy]

82. I can't speak for anyone else...however

there has to be some extent of accuracy for the numbers in the polls. Otherwise, why the hell would you pay to have them done? Networks rely on the data to predict who the winner of an election will be. The data may be used for other things, but I don't care what anyone says, I think the main purpose is to predict the winners and losers.
Besides that... exit polls, smedgset scrolls. I don't need any of them to know that Kerry was the legitimate winner. I knew 8 Kerry voters to 1 * voter.
All these exit poll arguments drive me nuts!

[Bill Bored]

84. You know what's worse than spending too much time on exit polls?

Spending too much time ARGUING about them!

[platinumman]

97. Well done, mistwell

I admire you for sticking to your guns. I wish I had the time and patience that you have. I think that all these people who are making these absurd assumptions from the exit polls are just diverting attention from the possibilities of real proof. I expect some of them are doing it on purpose.
I would disagree with you, though when you say that exit polls are generally not so accurate. I think you'll find that they are very accurate measurements of what they are measuring. That is, they are accurate measurements of the polling places that they have set out to measure, and I would take issue with the paper you quote which discusses non-response. The problem is, and always has been, that the sample of polling places are not necessarily an accurate sample of the state as a whole. Nor are they meant to be. They are chosen because swings in those areas tend to reflect swings in the state. Also, the economics of exit polling dictate that you have to choose a few better populated places in urban areas rather than a lot of empty palces in rural areas. Urban areas tend to be Democratic. These are basic commercial polling procedures. I'm sure you understand this, but it bears repeating. When the raw figures are 'weighted' (not a good term), all we are doing is trying to account for the deviation of these particular measured polling places from the actual state-wide vote. This deviation would have already been estimated beforehand, based on previous research.

At the risk of boring you, and pissing off the shouters, I'll just repeat the analogy I always give my students. Most of my hundred American friends are Democrats. In 1996, only ten of them voted Republican. This rose to fifteen in the close race of 2000. Now, if twenty of them voted Republican in 2004, then that would be a pretty strong indication that the Republicans were going to win. I know the voting patterns of my friends, and I know how they relate to the overall pattern. I know what weighting to apply. That is exactly what Mitofsky et al do: they know the voting patterns of the places they are polling.

And then TIA etc get hold of my raw data, showing that Democrats were leading Republicans 80-20 and shout fraud. And they don't stop shouting fraud no matter how much logic they are fed. As I said, I admire you for sticking to your message.

[skids]

98. Hrm...

You do know that precinct selection was verified not to be a source of deviation towards Kerry, and that the deviation was blamed entirely on "within precinct error," right? So I don't think saying that exit polls are "accurate measurements of the polling places they have set out to measure" is really going to calm anyone down.

I will agree to a side-point of that statement: Mitofski has data that is precise enough to identify at least a handful of precincts that deserve major scrutiny. Why he won't turn it over so that individual cases can be investigated before evidence is destroyed is something that, at least in my mind, raises serious questions about him.

[Aussie_expat]

99. Please provide sources for the above statements ..n/t

.

[anaxarchos]

122. You're kidding now, right?

http://www.exit-poll.net/election-night/EvaluationJan192005.pdf

[qwghlmian]

103. He promised to release the data

within 3 months of the elections or so (which is a lot faster than the usual release schedule, which has been about 12 months in previous elections) through the Roper center.

The paper ballots are not destroyed for quite a while (AFAIR, the law mandates a couple years holding time in most places).

[fasttense]

101. A common man's view of the exit polls.

I'm no expert on polling and my ability to do statistical analysis is limited but as an onlooker I see why the exit polls are a controversy.

1. If B*sh had been shown to be in the lead in all the polls before the midnight adjustments, would we be even looking at the exit polls? The answer is no. The exit polls are just another in a long laundry list of strange and questionable results from the 2004 presidential election.

2. If Mitofsky had said my data is good, the election should be reviewed for possible problems, how could that hurt him? For a man, who was paid millions of dollars to do a job, to turn around and say the job he did was wrong is kind of strange. This makes people wonder about him, about his motives, about the true results of the polls. This makes his company look bad, his work look bad and makes me think someone is paying him to keep from saying "my work was good it was the election you need to look at."

Exit polls are not unto themselves absolute proof of a rigged election. It's just this on top of everything else that makes the people think twice about the election.

[chorti]

105. refuting myths with myths

Mistwell - I think some of what you say is true but other pieces are simply myths replacing myths. Your whole argument is based on the notion that the exit poll was a crock, but that is simply not the case. It was a very well constructed poll. Very tight. Have you read the 70-page evaluation report?

It leaves many, many clues for you and I to find the vote fraud.

The evaluation report says only that the exit poll does not prove fraud. However, the report is clear that there is not a convinving explanation on why the exit poll was off. Did you read that in about 30% of the precincts where they exit-polled, the Kerry percentage was more than 15 percentiles greater than what the actual vote tally was? Of course some of that will happen through random error, but not 30%. They suggest this was because of differential response rates. There is a little evidence of that in the report but it is not at all convincing.

It is clear that the sample construction was very sound. Therefore, there can only be four possible reasons for why the exit polls were so far off in many states. First, some pollsters somehow picked out the Democratic-looking voters and interviewed them at a higher rate than Republican-looking voters. (There is very little evidence for that in the report.) Two, Bush voters (especially those who voted for Democratic candidates in other races) refused to respond to exit pollsters. Three, Bush voters (especially those who voted for Democratic candidates in other races) lied to exit pollsters and said that they voted for Kerry. Four, there was vote fraud. Personally, I think all of the latter three explanations happened.

But I think a careful study of the exit polls CAN come very close to proving vote fraud. For example, why would the Bush voters either avoid the exit pollsters or lie to exit pollsters in one state (i.e. Nebraska) but not in the neighboring states (i.e. South Dakota and Kansas)? I would love to see Duke University explain that one. And there are many more examples like that. And actually, your point #4 makes this case exactly. Since there are some states where the exit polls were accurate, how do you explain the huge inaccuracy in so many other (similar) states? It wasn't a sample problem.

[anaxarchos]

123. What you say here sums it up...

On your points 2 and 3, they also would have had to occur somewhat inconsistently (non-random events occuring randomly). Take a look at the Alabama senate and presidential races (by county).

[darthdemocrat]

124. I'm so sick of even wondering...

I think the democratic party (or perhaps just activists) should do some auditing of the election results by calling registered democrats to ask if they voted and who for. What better use could there be for all that left over campaign cash???

That alone would tell us if and how so many democrats could have voted for shrub, since that's the picture they're trying to paint. Somehow the democratic turnout and registration favored him? Shrub picked up heavily in the urban vote?

[Moloch]

125. Locking...

This has turned into a counter-productive flamefest of monolithic proportions.

Some of you should be reminded of the rules:

PERSONAL ATTACKS, CIVILITY, AND RESPECT

The administrators of Democratic Underground are working to provide a place where progressives can share ideas and debate in an atmosphere of mutual respect. Despite our best efforts, many of our members often stray from this ideal and cheapen the quality of discourse for everyone else. Unfortunately, it is simply impossible to write a comprehensive set of rules forbidding every type of antisocial behavior. The fact that the rules don't forbid a certain type of post does not automatically make an uncivil post appropriate, nor does it imply that the administrators approve of disrespectful behavior. Every member of this community has a responsibility to participate in a respectful manner, and to help foster an atmosphere of thoughtful discussion. In this regard, we strongly advise that our members exercise a little common decency, rather than trying to parse the message board rules to figure out what type of antisocial behavior is not forbidden.

Do not post personal attacks or engage in name-calling against other members of this discussion board.

If you are going to disagree with someone, please stick to the message rather than the messenger. For example, if someone posts factually incorrect information, it is appropriate to say, "your facts are wrong," but it is not appropriate to say "you are a liar."

Do not publicly accuse another member of this message board of being a disruptor, troll, conservative, Republican, or FReeper. Do not try to come up with cute ways of skirting around the spirit of this rule. If you think someone is a disruptor, click the "Alert" link below their post so the moderators can deal with it. Unfortunately, it has become all too common for members of this message board to label anyone with a slightly different point of view as a disruptor. We disapprove of this behavior because its intent is to stifle discussion, enforce a particular "party line," and pre-emptively label a particular point of view as inappropriate or unwelcome. This makes thoughtful and open debate virtually impossible.

Democratic Underground is a "big tent" message board which welcomes a broad range of progressive opinions. As such, you are likely to disagree strongly with many of the comments you see expressed here. Please do not take these differences of opinion personally. The simple fact that someone disagrees with you does not give you the right to lash out and break the rules of this message board. A thick skin is usually required to participate on this or any message board.

Please note that, strictly speaking, sweeping statements about entire groups of fellow progressives are not considered personal attacks. However, they are often inflammatory and counterproductive and the moderators have broad discretion to remove such posts in the interests of keeping the peace on the message board.

There is a difference between forceful advocacy for a particular issue (which is allowed), and personally attacking people (which is forbidden). If you can't tell the difference, you are likely to get into trouble here.

Do not "stalk" another member from one discussion thread to another. Do not follow someone into another thread to try to continue a disagreement you had elsewhere. Do not talk negatively about an individual in a thread where they are not participating. Do not start a new discussion thread with the purpose of "calling out" another member or picking a fight with another member. Do not use your signature line to draw negative attention to another member of the board.

If you just don't like someone, please be aware that you have the option of putting that person on your ignore list. Just click on the appropriate icon on one of that person's posts.

We do not typically delete threads which many members may consider to be "flamebait." However, the administrators will occasionally remove threads which we arbitrarily consider too rhetorically hot or too inflammatory. Please use good judgment when starting threads; inflammatory rhetoric does not normally lead to productive discussion.

If you are the type of person who just can't get along with other people, and if you seem to repeatedly cause trouble, eventually we will decide that your presence is a disruption and we will ban you. It doesn't matter if you are a progressive or a long-term member of this board.

There are no exceptions to these civility rules. You cannot attack someone because they attacked you first, or because that person "deserved it," or because you think someone is a disruptor. We consider it a personal attack to call a liar a liar, to call a moron a moron, or to call a jerk a jerk.