EXIT POLL NAYSAYERS: HERE'S A 3RD GRADE ARITHMETIC TEST FOR YOU.

[TruthIsAll]

CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS OF THE 122.3 MILLION WHO VOTED IN 2004.

HINT:
BUSH GOT 50.456 MM VOTES IN 2000.
APPROXIMATELY 48.7 MM ARE STILL ALIVE.
NOW CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS.

AND WHILE YOU ARE AT IT, DO THE SAME FOR GORE 2000 VOTERS.
GORE GOT 50.999 MM VOTES
ABOUT 49.21MM ARE STILL ALIVE.


THAT IS THE TOTAL MATH EXAM.
WILL YOU PASS OR FAIL THIS 2 QUESTION TEST?

GOOD LUCK

FOR REFERENCE ONLY:
THE FINAL 1:25 PM NOV3 NATIONAL EXIT POLL (13660 RESPONDENTS) SAID THAT 43% OF BUSH 2000 VOTERS CAME OUT TO VOTE - AND JUST 37% OF GORE VOTERS.

AT THE NOV 3, 12:22 AM TIMELINE, THE NATIONAL EXIT POLL (13047 RESPONDENTS) SAID THAT 41% OF BUSH 2000 VOTERS CAME OUT TO VOTE - AND 39% OF GORE VOTERS.

[eallen]

1. Speaking as someone who has taught college math...

Please stop pretending. Whatever calculation you're doing isn't telling you what you think it is.

:mad:

[TruthIsAll]

2. No generalities, please. Solve the problem. n/t

.

[eallen]

3. The problem statement is broken. Pretending there's an answer is stupid.

[TruthIsAll]

4. And you taught math? You say there is no answer?

You have just failed a 3rd grade math test.

Show us where the problem statement is broken.
So far, you have not shown us a thing - except that you do not have a clue.

[eallen]

5. It's hard to say where a stupid thing is stupid. I suspect...

I suspect you're trying to hint at some calculation that makes a variety of assumptions about how voters stayed consistently for or against Bush between 2000 and 2004, about what those who voted in 2004 but not 2000 did, or something else. The basic fact remains this: from the assumptions you state, any percentage of votes for Bush in 2004 are possible, up to and including 100%.

That's contrary to reality, of course. But it's not mathematically contrary to your problem statement.

[TruthIsAll]

6. You totally miss the point. I am using actual 2000 and 2004 numbers.

You say:
"I suspect you're trying to hint at some calculation that makes a variety of assumptions about how voters stayed consistently for or against Bush between 2000 and 2004, about what those who voted in 2004 but not 2000 did, or something else. The basic fact remains this: from the assumptions you state, any percentage of votes for Bush in 2004 are possible, up to and including 100%"

That's contrary to reality, of course. But it's not mathematically contrary to your problem statement.

I say:
Please do the math!
You are totally off-base.
I am assuming nothing.

I asked you to do a simple calculation.
Are you familiar with the national exit poll?
I think not.
This is a very simple calculation.

THIS IS NOT A TRICK QUESTION!
Please do the calculation!
HINT: Its simple division.

I will try to be patient with you.

[autorank]

10. It would appear you are you starting out by "baiting"

I just assumed that from the little mad face and the "Please stop pretending." If that was a sincere request, then it should stand as it is. If you were being sarcastic with out intending to bait then you use a emoticon that looks like this :sarcasm:

Just a few of quick questions:

-Do you think that there was election fraud in 2004?
-By your best estimate, how was it committed?
-Do you believe like some of the other academics that visit our faire forum that it's necessary to prove election fraud before we can justify a comprehensive investigation with the goal of assessing whether or not election fraud took place?
-You don't have to answer this but since you say you're an academic, you may be part of the OTOH faction here. OTOH revealed his college affiliation, Febble revealed her affiliation with NEP pollster Warren Mitofsky. Do you plan to do the same?

The first three are the most relevant. The last is entirely optional. It just arose out of curiosity due to a familiarity in style.

Thanks.

[OnTheOtherHand]

12. gee, I only read this post by accident, but

could you please document the assertion that I or any other academic (or any other non-academic, for that matter) has argued that "it's necessary to prove election fraud before we can justify a comprehensive investigation"?

Might depend on what you mean by "comprehensive." Who knows what you mean? Not I.

[autorank]

14. Are you following me around? I'm flattered & I'm not a "coincidence"

theorist. Why don't you deal with TIA on the statistical issues. I'm looking for a "balanced" dialog. You and he seemed to be having one.

Don't bother answering because you've got work you said you were here to do, help progressives and Democrats refine their position on election fraud related to the NEPs. I think thats important work. TIA is our guy for that.

cya on the other thread.

[OnTheOtherHand]

15. let the record reflect

that autorank made no effort to justify the premise of his third question, so that one can safely be stricken.

[TruthIsAll]

20. OTOH, let the record reflect that you have been disengenuous.

You have never properly explained the FINAL NEP 43% Bush/37%
Gore "How voted in 2000" weightings.

It is mathematically impossible. 
And you won't admit it.


IMPOSSIBLE SCENARIO I:
13047 respondents: 12:22am
Kerry wins easy - even with an IMPOSSIBLE 41/39% split

VOTE2000	
	Mix   Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	39%	8%	91%	1%	3.1%	35.5%	0.4%
Bush	41%	90%	10%	0%	36.9%	4.1%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				47.38%	51.23%	1.21%

Probability of Bush going from 47.38% (poll) to 50.73%
(vote): 
 1 in	38,498,885,514


IMPOSSIBLE SCENARIO II:
Final Exit Poll (1:25pm)- 13660 respondents:
BUSH WINS: PURE FICTION - 43% IS EVEN MORE IMPOSSIBLE

VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	37%	8%	91%	1%	3.0%	33.7%	0.4%
Bush	43%	90%	10%	0%	38.7%	4.3%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				49.02%	49.61%	1.19%



PLAUSIBLE SCENARIO III:
Using the calculated maximum possible weightings, Kerry wins
in a
landslide: 

Kerry	63.89mm (52.24% 
Bush	56.77 (46.42%)
Nader/Other 1.49 (1.22%)


VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17.00%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	40.24%	8%	91%	1%	3.2%	36.6%	0.4%
Bush	39.82%	90%	10%	0%	35.8%	4.0%	0.0%
Other	3.00%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				46.42%	52.24%	1.22%
							

[Internut]

22. TIA, here is a question for you -

in the 2002 American National Election Study (that can be found here: http://www.pop.psu.edu/data-archive/codebooks/nes/NES2002.pdf ) there was a question:

V203111: Who did vote for in the 2000 presidential election?

With the following results:

1. Al Gore             431  44%
3. George W. Bush      502  51%
5. Ralph Nader          32   
7. Other {SPECIFY}       8   
8. Don't know            1   
9. Refused              10   
0. NA                    1   

Note: these are raw results, not "weighted", not tampered with in any way.

So, now for the question. Knowing what you know about 2000 election, do you find these results of the American National Election Study "mathematically impossible"?

[TruthIsAll]

26. STRAWMAN. DO YOU THINK YOU CAN GET AWAY WITH THAT?

Internut, I am interested in the facts, not a survey. Got it?

By the way, how does Bush get 43%-37%?
What does that say about rBr?

Just who is being reluctant here with the facts?

YOU LOSE: BOTH WAYS!

[Internut]

27. Your failure to answer the question for the second time is noted.

You love to ask questions of people but seem pretty reluctant to answer them yourself.

Oh, and by the way, for someone who is "interested in facts, not surveys", you seem to spend an inordinate amount of time discussing surveys (especially one in particular, the 2004 exit poll). I think that throws in doubt your professed disinterest in surveys.

[anaxarchos]

33. That's a really good question.... each time you asked it.

I answered your question, Internut. Is that OK?

Actually, I answered OTOH who asked EXACTLY the same question... Is that alright?

Here it is:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=378368&mesg_id=379518

Considering the fact that you asked it in every single thread I looked in (including the gun thread?), you must think it is a really, REALLY good one....

Now, do I really have to track down every single place you asked it or is once enough?

Could you answer TIA now?

[Internut]

34. I answered you in that thread. Your link is irrelevant. n/t

[anaxarchos]

35. I have reviewed the posts on this subject....

I don't believe relevance is all that relevant.

[anaxarchos]

43. Second Try...

"Lemme try this again... awake this time."

"The answer is obvious... 2002 is a Mid Term election in which the Republicans out pointed the Democrats by 6 to 8 points (depending on how you calculate it). The sample is different. That's kinda what I was alluding to above when I pointed to the partisanship spikes for mid-terms...

Take a look at the 1996 NES. The opposite effect is discernible (Perot doubles his vote!) for precisely the same reasons.

BTW, NES post-election is days-to-weeks, not hours-to-days... Thus the "bandwagon" discussion is pertinent.

You just weren't arguing it..."


Copied from the other thread as penance for earlier screw-up.

[OnTheOtherHand]

45. also copied from the other thread: not so obvious

What do you mean when you say "the sample is different"? The NES is a national survey, not an exit poll, and the 2000 vote question was asked in the pre-election wave. There seems to be no difference between reported 2002 voters and non-voters in their response.*

Similar result in the 2004 NES, again in the pre-election wave, although the gap is only 4 points. (These are both unweighted results, by the way, just in case weighting is of the devil. The weighted gaps are wider.)

* Just to clarify what we are talking about, this is people in 2002 -- some of whom will go on to vote in the 2002 election, some of whom won't -- being asked who they voted for in 2000. And then, ditto, replacing "2002" with "2004." Certainly different samples than the 2004 exit poll, but I don't see why they aren't pertinent to the interpretation of TIA's 43/37 result.

[FogerRox]

81. I thought Gore won the popular vote in 2000

ya know that consortium of newspapers, the panel on the election--

[OnTheOtherHand]

85. yes, Gore won the popular vote in 2000

So why is it that in the 2002 National Election Study (which surveys both voters and non-voters -- in fact, this particular question was asked before the election), substantially more people said they had voted for Bush in 2000...

...and in the 2004 National Election Study, the same thing happened again...

...and in the final weighted version of the 2004 exit polls, the same thing happened again...

... and somehow we are supposed to consider the last result more "impossible" than the ones before it?

[autorank]

32. O, we wanted Mano e Mano with you and TIA. You just blinked.

Why don't you answer his question, if you don't mind. It's the central part of tonight's debate. The figures are right there. What more could you ask?

I'm seeing a pattern in the TIA threads that's interesting. He makes a point, people show up and change the topic, the thread gets hijacked (IMHO) with irrelevant dialog (some times by the same people, sometimes by different people saying the same thing), and side issues arise.

Now I know you want to have a coherent debate on this issues. Maybe if those get addressed, TIA will actually get to post a thread and not be the only user on DU to have people show up all the time and give him grief, I repeat, the only user on DU that I know of to consistently have people show up and give him and his work grief.

Is it answer time, dialog time, or more rhetorical meta-analysis?

[autorank]

37. And again!

Note, this is much easier than "refer to post#..."

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=378368&mesg_id=379524

[TruthIsAll]

41. THE WORDING NEEDS TO BE MADE CLEAR.

I APOLOGIZE FOR ANY CONFUSION.

THIS WAS NOT STATED CLEARLY:
FOR REFERENCE ONLY
THE FINAL 1:25PM NOV3 NATIONAL EXIT POLL (13660 RESPONDENTS) SAID THAT 43% OF BUSH 2000 VOTERS CAME OUT TO VOTE - AND JUST 37% OF GORE VOTERS.

AT THE NOV 3, 12:22AM TIMELINE, THE NATIONAL EXIT POLL (13047 RESPONDENTS) SAID THAT 41% OF BUSH 2000 VOTERS CAME OUT TO VOTE - AND 39% OF GORE VOTERS.

***************************************

THIS IS WHAT I MEANT TO SAY:
FOR REFERENCE ONLY:

BASED ON THE FINAL NATIONAL EXIT POLL AT THE NOV 3, 1:25PM TIMELINE (13660 RESPONDENTS) OF THE TOTAL 122.3MM WHO VOTED IN 2004, 43% VOTED FOR BUSH IN 2000 - AND 37% FOR GORE.

BASED ON THE NATIONAL EXIT POLL AT THE NOV 3, 12:22AM TIMELINE (13047 RESPONDENTS) OF THE TOTAL 122.3MM WHO VOTED IN 2004, 41% VOTED FOR BUSH IN 2000 - AND 39% FOR GORE.

[TruthIsAll]

42. HERE ARE THE 13047 AND 13660 HOW VOTED IN 2000 WEIGHTS AND PERCENTAGES

The maximum possible percentage of Bush 2000 voters who could
have voted in 2004 is 39.82%.

Bush 2000 voters: 50.456 million
Bush 2000 voters still alive = 48.69mm

Total 2004 voters: 122.3 mm.

Maximum Bush 2000 voters as a percentage of Total 2004
voters:
Bush 2000/Total 2004 = 48.69/122.3 = 39.82%

The same calculation for Gore voters = 40.25%

Now lets look at the National Exit Poll:


IMPOSSIBLE SCENARIO I:
13047 respondents: 12:22am
Kerry wins easy - even with an IMPOSSIBLE 41% Bush/39% Gore
split

	Mix   Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	39%	8%	91%	1%	3.1%	35.5%	0.4%
Bush	41%	90%	10%	0%	36.9%	4.1%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				47.38%	51.23%	1.21%

Probability of Bush going from 47.38% (poll) to 50.73%
(vote):  
1 in 38,498,885,514


IMPOSSIBLE SCENARIO II:
Final Exit Poll (1:25pm)- 13660 respondents:
BUSH WINS: PURE FICTION - 43% IS EVEN MORE IMPOSSIBLE

MIX	Bush	Kerry	Nader	Bush	Kerry	Nader
NO    17%	45%	54%	1%	7.7%	9.2%	0.2%
Gore  37%	10%	90%	1%	3.7%	33.3%	0.4%
Bush  43%	91%	9%	0%	39.1%	3.9%	0.0%
Other  3%	21%	71%	8%	0.6%	2.1%	0.2%
100%				51.11%	48.48%	0.78%



PLAUSIBLE SCENARIO III:
Using the calculated maximum possible weightings, Kerry wins
in a  landslide: 

Kerry	63.89mm (52.24% 
Bush	56.77 (46.42%)
Nader/Other 1.49 (1.22%)


VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17.00%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	40.24%	8%	91%	1%	3.2%	36.6%	0.4%
Bush	39.82%	90%	10%	0%	35.8%	4.0%	0.0%
Other	3.00%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				46.42%	52.24%	1.22%
							

[WI Independent]

7. 39.8%

of the total 2004 votes could have come from Bush 2000 voters. It could be slightly higher or lower depending on actual death rate. 40.2% maximum if nobody died.

What's your point?

Edit:
Missed part 2:
40.2% (41.7 if nobody died)

[TruthIsAll]

8. WI independent... draw your own conclusions.

What does the 43/37 split tell you about the final (13660) exit poll which matched to the vote and declared Bush 51-48% the winner?

What does the 41/39 split tell you about the first 13047 exit poll which matched to the vote and calculates that Kerry was a 51-48% winner?

Now what if we changed the weightings to 39.82/40.24 and ran it through the same calculation, without making any other changes to the 13047 percentages? What would Kerry's winning margin now be?

Come on, WI.
It's not too hard, you can do it.

[WI Independent]

9. You know my answer...

Either there was widespread fraud (as you believe),
OR
the polling numbers (43/37) are bad (my position).

I might switch to your side when; leading Democrats start delivering the "fraud at a level that changed the outcome" message,
OR
all details of the polling methodology, weightings, etc. are released for public verification.

[autorank]

11. Good position. Leading Democrats should do this and the data..

should be released. I hope it's not a "might" but a "would."

There's actually hope of the first of your criteria being met soon. I'll have a major post on General Discussion later tonight or tomorrow on this. Things are looking much better.

[WI Independent]

18. I "would"...

start getting very suspicious of the election results if leading Democrats started making concrete accusations. I would become convinced (as would the majority of the population) if they started producing hard evidence.

I doubt the exit poll methodology will ever be released. Polls are very accurate IF the sampling is totally random and participation is uniform. Neither was remotely true in this case. IMO, if they released the workings of the whole process we would find there is at least as much art (guessing) as there is science involved in correcting the raw numbers.

So what if hard evidence proves the polls were right? Without an explanation of HOW the pollsters did it, my opinion would be that even a blind squirrel finds an acorn from time to time. :D

[autorank]

23. Well, the election is a public event, the poll material must be released.

I'm actually have a different take on the validity of the poll based on past measures but that's not the real issue. It's our damn election. They took a picture of it. We're the image and there may very well (I believe WAS) a crime committed. What would the police do about a security camera? Get a court order to bring it in to examine it if that's what was required.

The poll data should be released in a way that is PUBLIC (i.e., you, me, everybody). If the networks pull their "it's proprietary" excuse, haul them into court...and remind them that they had no problem releasing intimate details of Clinton's sex life, something we really didn't need the details on.

After such a process, we could sit down and :toast: who ever was right (and you'd be buying haha).

Good point on the need to release the information.

[WI Independent]

29. Not sure about the legal issues...

It was done as a private contract... not sure if the release can be forced unless it was subpoenaed as part of a trial.

See post #28 below... After a bit more reading on how they do the poll, I'm not sure the release of it would prove or disprove anything on a national level since the actual vote totals are included in the weighting.

However, a one-to-one comparison with suspected individual precincts could be quite valuable.

A beer sounds good! Hell, I'll even buy one right or wrong. I have a lot of opinions, but (unlike the current oval office resident) I am wrong sometimes. :)

[Internut]

31. The one-to-one comparison of individual precincts

has been done.

See

This is a scatterplot of WPE (Within-Precinct-Error) on Y axis vs. % of Bush vote in the precinct on the X axis. The WPE means the difference between the exit poll raw data results for the precinct and the official precinct results.

This scatterplot tells me three things:

1. The variance (measure of how "dirty" the data is) is huge.
2. There is really almost as much "Kerry Overestimate" as "Bush Overestimate"
3. Anyone who says that overestimation on one side of the X axis (Bush or Kerry) is the results of fraud has to explain why just as much overestimation exists on the other side of the axis.

[Donkeyboy75]

38. Internut, don't bother doing an actual statistical analysis.

You'll never penetrate the thick skulls of those who toss numbers around without knowing what they mean.

Nice piece of data, BTW. :thumbsup:

[Peace Patriot]

13. Ahem...

"I might switch to your side when; leading Democrats start delivering the 'fraud at a level that changed the outcome' message...". --WI Independent

Oh, you mean those guys who failed to notice that Wally O'Dell (major Bush donor, Bush-Cheney campaign chair) of Diebold and the founders of ES&S (far right funding, major Bush backers) had gained control of our election system with secret, proprietary programming code, and no paper trail? Uh, those same guys who promised, promised, PROMISED throughout the election campaign that "EVERY VOTE WILL BE COUNTED"? The guys who took our money for their legal fund on election night and then walked away from Ohio?

Those guys?

You're waiting for THEM?

[WI Independent]

17. Yes...

That's my barometer. Proof of a stolen election would be total vindication for the strategies used by Kerry. Kerry would be a shoe in for 08. As it stands the only place Kerry has to go is to wait for Kennedy to die/retire so he can be the senior senator from MA. If it really happened, what possible motivation would he have for staying silent on it?

If fraud happened to an extent that would change the election.... there HAS to be proof (not just exit polls). This proof should come out under a full investigation. Why isn't every major Democrat screaming for a Congressional review, a special prosecutor, ANY sort of real investigation?

Possibilities:
- It didn't happen

- It happened, but they don't think they can prove it. In that case, why not at least get some political benefit by talking of evidence that wouldn't hold up in court?

- It happened, but the Dems cheated too, so they can't call for a full investigation. I think this is a real possibility... they very likely didn't cheat to the same extent, but if it comes out both sides cheated to any degree it will become a wash in public opinion.

- The "tin foil hat, skull and bones, it's all for show" conspiracy is correct. I don't really buy this... besides if it's true we might as well all drink our kool-aid and try to be as happy as possible.

- They aren't smart enough to figure it out. I don't believe this... even if I did, why should I get all excited about being lead by a bunch of dumb-asses that aren't even smart enough to defeat the current dumb-asses in power?

- They're to comfy to risk rocking the boat. Again, if this is true why should I give a rip if my corrupt politician has an R or a D after his/her name?

- Others???

[ginnyinWI]

149. I subscribe to the second reason

They know there was fraud, and they can't prove it (yet, anyway).
Why not talk about it? Because of what happened to Gore, and thinking it would damage their credibility more than it would do any good. But with more evidence, they will. Especially if/when a disillusioned population is going be listening.

I agree about the exit polls; they are evidence but not proof.

[TruthIsAll]

19. Your position is typical: you confuse the poll with the facts.

IMPOSSIBLE SCENARIO I:
13047 respondents: 12:22am
Kerry wins easy - even with an IMPOSSIBLE 41/39% split

VOTE2000	
	Mix   Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	39%	8%	91%	1%	3.1%	35.5%	0.4%
Bush	41%	90%	10%	0%	36.9%	4.1%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				47.38%	51.23%	1.21%

Probability of Bush going from 47.38% (poll) to 50.73% (vote):

 1 in	38,498,885,514


IMPOSSIBLE SCENARIO II:
Final Exit Poll (1:25pm)- 13660 respondents:
BUSH WINS: PURE FICTION - 43% IS EVEN MORE IMPOSSIBLE

VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	37%	8%	91%	1%	3.0%	33.7%	0.4%
Bush	43%	90%	10%	0%	38.7%	4.3%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				49.02%	49.61%	1.19%



PLAUSIBLE SCENARIO III:
Using the maximum calculated weightings, Kerry wins in a
landslide

VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17.00%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	40.24%	8%	91%	1%	3.2%	36.6%	0.4%
Bush	39.82%	90%	10%	0%	35.8%	4.0%	0.0%
Other	3.00%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				46.42%	52.24%	1.22%
							

[WI Independent]

21. IMO, you just proved the poll was bad.

The sampling and/or the weighting HAD to be bad, even in the 13047 sampling. No amount of election fraud should affect this number, it's a simple question "who did you vote for in the last election?". The first column should be accurate (if the sampling and weighting were accurate) regardless of anything that happened in this election.

If the poll couldn't even provide an accurate accounting of who voters voted for in the last election, why should we assume it can provide accurate data on any other aspect of the election?

[TruthIsAll]

24. Then you agree the FINAL EXIT POLL which had Bush winning, was also wrong

And do you agree that the first 13047 respondent timeline which Kerry won was OK?

Note: The 41/39 How voted split was still within the 1% MoE.

Thank you for seeing the light.

[Internut]

25. See post #22 -

was the poll that is cited there "bad" or "fraudulent"? After all, for the same question it had the same "impossible" results....

[WI Independent]

28. The percentages from the last election...

were obviously wrong. The question still remains why were they wrong... because of bad sampling or rampant fraud?

The basic problem with all this is you're trying to use the exit polls as independent verification of the election totals. This simply will not work because actual vote counts are an integral part of the Edison/Mitofsky methodology. The raw exit poll interviews are not designed to give meaningful information without weighting. This weighting is based at least partially on actual vote totals. The poll itself is not complete until a significant amount of ACTUAL vote totals are plugged into the weighting formulas.

From the :
"The local figures become more complete as more precincts report vote returns. The county or township vote is put into statistical models, and Edison/Mitofsky make estimates and projections using those models."

I suspect the Bush 2000 voter percentage is caused when actual vote totals didn't match interviews and the weighting formulas forced the percentage impossibly high.

In summary, this poll is not independent verification of the accuracy of the election. By the very nature of it's design, raw poll interview results do not mean anything... and the end result will always match the election results as soon as enough actual results are plugged into the weighting.

[autorank]

30. Well now I'm confused about your position.

What are you doing to get the networks to release the data? I've written them, called, spoken to representatives and others. I follow up and remind them of my concern...and I encourage and help others do the same. You seemed to like that idea earlier but now there is a full scale attack on either the NEPs and, I presume, the state exit polls. What do you think about on-the-ground, in machine or network fraue, actual or suspected? Should we re-install the old premier of the Ukraine since exit polls are meaningless? Have you read the previous bodiy of work on this site, TIA's and others, which address these questions? What do you want to accomplish on DU?

[WI Independent]

39. My position...

- Fraud did not occur to an extent that changed the election.
- The exit polls were not designed to be accurate until weighted with actual vote totals. Therefore they are incapable of detecting fraud.
- I had thought the polls might be helpful in verifying particular precincts, but after seeing the WPE scatterplot Internut posted above, that doesn't look promising.

The Ukraine exit polls were designed to detect fraud, this poll was not.

I've probably missed a few of TIAs posts but I have read most of them. I simply don't agree with the conclusions. For the most part, I think the calculations are accurate, but the input data is bad... "garbage in, garbage out".

I'm not an activist... for me politics is more of a "spectator sport". I respect your commitment, but doubt I will ever get involved to that level. I don't have a goal on DU... I like to read others viewpoints. I occassionly throw in my .02... my read to write ratio is extremely high.

[TruthIsAll]

46. I missed this. GIGO? No siree. I will back up EVERY input data item

you cite.

WI, calling my analyses garbage-in garbage out shows an utter lack of understanding on your part.

I HEREBY ISSUE A CHALLENGE TO YOU:

GO THROUGH ANY AND ALL OF MY POSTS. IF YOU CAN FIND ONE INVALID INPUT DATA ITEM, POINT IT OUT.

THEN WE CAN DISCUSS PARTICULARS.

YOU ARE QUITE ADEPT IN YOUR RECITATION OF THOSE ALL TOO FAMILIAR BOGUS TALKING POINTS:

You:: Fraud did not occur to an extent that changed the election.
Me:
What is the basis for that statement?
Tell us what was the extent of the fraud?

You:
- The exit polls were not designed to be accurate until weighted with actual vote totals. Therefore they are incapable of detecting fraud.
Me:
That is total hogwash I have heard repeated over and over again.
That's like saying kitchen knives are not designed to kill anyone, so they are not capable of it.
Think about how ridiculous that statement is.
It is a classic straw man argument.
WI, they (MSM) have taught you oh so well.
You have been thoroughly brainwashed by RW talking points.
Exit Polls are designed to tell you how people voted.
If they are telling us that people voted for Kerry, does that mean they are wrong, since they do not agree with the vote?
You are implying that the recorded vote was correct.
Such inverted logic.

How very sad to see you backpedal into a world of make-believe.
Black is white.
White is black.
Pure Alice in Wonderland.

You:
- I had thought the polls might be helpful in verifying particular precincts, but after seeing the WPE scatterplot Internut posted above, that doesn't look promising.

Me:
You can't be serious.
You are drawing conclusions from that work of punk art.
YOU SHOULD BE DEMANDING TO SEE THE RAW PRECINCT DATA.
INSTEAD, YOU LET INTERNUT BRAINWASH YOU WITH A SCATTERPLOT.

ARE YOU INCAPABLE OF ANYTHING OTHER THAN RECITING THE GARBAGE THAT OTHERS FEED YOU?

I thought you were making progress.
But it appears that instead of moving forward, you have reverted to that riverboat down denial.

The Ukraine exit polls were designed to detect fraud, this poll was not.



NOW BACK IT UP

[WI Independent]

47. Your source is incomplete...

GIGO: Don't take "garbage" so personally, it's just an expression. "Incomplete data in, Inccurate conclusion out" just doesn't have the same ring to it.

"Fraud did not occur to an extent that changed the election." Basis: John Kerry, the guy with the most to gain if it did. (Not to mention virtually every other prominent Democrat.

Exit Poll Design: see Post #28. The poll was designed to include actual poll results in order to predict election results. If the interviews alone were an accurate predicter, why go through this exercise? The poll (by design) was incomplete until the final numbers were released.

Precinct verification: "doesn't look promising" is hardly a conclusion.

BTW... You might want to have your keyboard checked out, I think your Caps Lock is inadvertantly getting turned on. ;)

[TruthIsAll]

54. Still waiting for that data item....

So far, nothing..

Take a day off and go thtrough all my threads

[WI Independent]

60. Basically any data that came from...

early exit poll interview results, which is the basis for most of your arguments. You're extracting data from the middle of an incomplete process.

So tell me, if the raw exit poll interviews were enough information to do reliable analysis, why were the polls designed from the beginning to mix that data with actual vote tallies?

[tommcintyre]

48. My, my, That's a very "convenient" statement.

"The Ukraine exit polls were designed to detect fraud, this poll was not."

Do you really mean... were not "intended" to detect fraud?

Think about it for a minute. If an exit poll is competently designed, by it's very nature, it will (and historically has been) the most accurate indicator of election fraud. PERIOD - this is NOT a debatable issue.

This "non"-issue has been discussed (and discarded) in the past - probably many times. What we really need is a FAQ for "questions" like this, so we can re-answer such things more efficiently.

Here's a nice little germaine cartoon. Now I gotta get back to breaking some NEW ground regarding election fraud.

[WI Independent]

49. No, I mean not designed.

The design calls for actual vote totals to be used in combination with the exit polls. This indicates the exit polls alone are not considered accurate enough to predict the outcome. Most of the exit poll analysis is based on the fact that early exit poll raw data didn't match actual vote counts. Of course it didn't... it wasn't designed to without the vote tallies included in the formula. There's a lot of questioning and accusation based on how the final released poll "mysteriously" swung to match the tallied results. It's no mystery, when enough actual voting results are entered the poll results will eventually match the counted vote totals. This would be true whether there was fraud in the actual vote counts or not.

This is a competent design for it's intended use... to give the talking heads at the networks something to fill air time. It's not a competent design to detect fraud. An exit poll to detect fraud would need to be totally independent of the actual vote counts. I think it would be a great idea to have such a poll; unfortunately in 2004 we didn't.

[tommcintyre]

51. Here ya go....

http://en.wikipedia.org/wiki/Exit_poll
"Exit polls have historically and throughout the world been used as "parallel vote tabulation", as a check against and rough indicator of the degree of fraud in an election. Some examples of this include the Venezuelan recall referendum, 2004, the Ukrainian presidential election, 2004, and the 2004 U.S. presidential election controversy."
snip
"Leaks of exit poll figures for the 2004 presidential election, mainly via the Internet, appeared to indicate a victory for John Kerry. The discrepancies between the exit poll data and the vote count that where outside of the margin of error, coupled with irregularities in the election which seem to explain the discrepancies and what many perceive as evasive tactics by the polling companies, have shed doubt on the legitimacy of that election among political activists and government officials. (See 2004_U.S._presidential_election_controversy, exit polls for more detail.)"
-----------------------
Here, this is what may be causing your confusion:

http://en.wikipedia.org/wiki/2004_U.S._presidential_election_controversy%2C_exit_polls
"Because final published exit polls in America are matched to vote counts, they cannot be used to determine election fraud. However, in the 2004 election, pre-matched exit polls were leaked onto the internet."
snip
""Exit poll data - asking voters which way they voted as they leave the polls - are used around the world as excellent predictors of actual vote counts, usually accurate within a fraction of a point. Exit polls in this election seemed to match the vote tallies, as usual, except in those areas using touchscreen voting machines (like the Diebold Accuvote) or other software or modem-mediated electronic systems (like those from ES&S) with no paper trail - used by approximately one third of voters, many in swing states."
snip
"<in Germany>... people fill in hand-marked ballots, which are hand-counted by civil servants, watched over by volunteer representatives of the political parties. ... It's totally clean, and easily audited. And even though it takes a week or more to count the vote ... the German people know the election results the night the polls close because the news media’s exit polls, for two generations, have never been more than a tenth of a percent off."
------------------------

There are tons more sources like this; so like I said in the previous post, this issue is a non-starter. It's one of the weakest ways to argue against the high probability that this election was stolen. The argument is purely designed in an attempt to create doubt; and it does it very poorly.

Over the last seven months I have found that two groups of people attempt this argument: those who are poorly informed, or those who appear to be trying to mislead - and frankly, aren't very good at it.

[OnTheOtherHand]

52. kneel to the authority of wikipedia!

Or, not.

Tom, if you understand what wikipedia is, you know that you can't actually win this argument by citing it.

[TruthIsAll]

55. 43/37 ........ n/t

.

[OnTheOtherHand]

56. and if you understand the arguments, this isn't a winner either. n/t

[TruthIsAll]

57. IF YOU UNDERSTOOD THAT DEAD PEOPLE CAN'T VOTE, YOU WOULD SEE THAT IT IS..

/

[mgr]

58. Except in Cook County, Illinois, under the esteemed Richard Daley

You are aware that only one in four questionnaires employed in the national election survey has the question on it--who did you vote for in the last election.

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

So how do you get 13,000 people to respond when only 4,000 may have done so? Apparently the MoE for the question is about 3%.

(NEP FAQs: "What is the Margin of Error for an exit poll?
Every number estimated from a sample may depart from the official vote count. The difference between a sample result and the number one would get if everyone who cast a vote was interviewed in exactly the same way is called the sampling error. That does not mean the sample result is wrong. Instead, it refers to the potential error due to sampling. The margin of error for a 95% confidence interval is about +/- 3% for a typical characteristic from the national exit poll and +/-4% for a typical state exit poll. Characteristics that are more concentrated in a few polling places, such as race, have larger sampling errors. Other nonsampling factors may increase the total error." http://www.exit-poll.net/faq.html)

That may make the bandwagon issue vanish, but it makes your argument do likewise.

Mike

[TruthIsAll]

62. mgr, do your homework:> MOE = 1/sqrt(4000) =1.58%

You are analytically moribund.

[Name removed]

63. Deleted message

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

[TruthIsAll]

65. Then you believe in rBr or eDp; that is your choice. Fine.

I don't.

You are a slave to your own contradictions, mgr.

I will once again prove to all of DU that you are logically moribund:

1)If you believe the election was stolen, as you have stated on this forum, then you believe Kerry won.

2)If you believe Kerry won, then you must also believe the exit polls were right.

3) But yet you argue otherwise

Q.E.D.

Which is it mgr?

DO you see how ridiculous your protestations are?

If you manage the way you express yourself at DU, your employees must be very confused.

I am very consistent.
I do not believe in rBr.
I believe the exit polls were correct - -except for the final NEP.

You are all over the place.
You are a walking mass of confusion.

[OnTheOtherHand]

68. gee, at least insult the right person!

You were replying to Internut.

Ooh, ooh, can I do this one?

1) Untrue or misleading. If the election was stolen through vote suppression, then one would not say that "Kerry won."

2) Untrue. For instance, it could be that Kerry won Ohio but lost the popular vote (so the exit polls could still be wrong).

[mgr]

99. Did not catch this was for me

Interesting syllogism:

If F then K (F=fraud; K=Kerry won)

If F then ~E (E= exit poll correct)(hidden assumption)

Therefore K and ~E.

I affirm the first proposition, but I don't the second. How does this make me inconsistent? Remember fraud is a class of actions that does not necessitate fraud manifest in the national exit poll. If I beleive that Ohio is stolen, how would that be manifest in the national exit poll, Ohio is one state, and enough to give Kerry the electoral college.

Do I believe that Fraud stole the popular vote from Kerry. It could be possible, but only by suppression, and that cannot be measured by an exit poll.

Mike

[Name removed]

74. Deleted message

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

[OnTheOtherHand]

64. dunno the right answer, but that isn't it

Nothing wrong with the math, just the assumptions. This is not a Simple Random Sample.

Anyway, I think it would be nice for you to acknowledge the false assumption that people are bound to have reported their 2000 votes accurately, before we even start to fuss over the false SRS assumption.

[TruthIsAll]

67. OTOH, would you just once entertain the possibility that the

final National Exit Poll weights were set to match the recorded vote?

Just once.
Would you concede the possibility?

Or are you saying that was not the case?

If you refuse to answer the question, your students will take note.

[OnTheOtherHand]

69. hello? hello?

I don't just entertain the possibility, I embrace it. Everyone knows that the final weights were set to match the recorded vote. We are not debating this.

We are debating your specious, at this point seemingly willfully ignorant claim that the 43/37 result is mathematically impossible.

[TruthIsAll]

70. Hello, Hello? You have just capitulated. Thank you.

/

[Name removed]

71. Deleted message

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

[OnTheOtherHand]

73. huh?

It isn't a secret that the final results are weighted to the official returns. That is the method. That is what they say they will do, and they do it.

But it does not prove that the official returns are wrong.

In order for your 43/37 argument to work, you would need to assume (inter alia) that everyone reports their past votes accurately. And this assumption is demonstrably false.

So your calculation goes out the window, no matter how perfect the arithmetic.

[TruthIsAll]

101. I AM ASKING YOU FOR A ONE WORD ANSWER...

YOU:
In order for your 43/37 argument to work, you would need to assume (inter alia) that everyone reports their past votes accurately. And this assumption is demonstrably false.

So your calculation goes out the window, no matter how perfect the arithmetic.

ME:
WHICH OF THESE NUMBERS ARE YOU DISPUTING:
50.456?
50.999?
48.7?
49.2?
122.27?

WHICH ONE(S)?

FOR THE UPTEENTH TIME, I AM NOT TALKING ABOUT WHAT THE RESPONDENTS SAID OR WHAT MITOFSKY SAID WHEN HE RE-WEIGHTED THE POLL.

ARE YOU REALLY AS DENSE AS YOU APPEAR TO BE?

ANSWER THE QUESTION.
RIGHT HERE.
RIGHT NOW.

IS IT POSSIBLE THAT 43% OF THOSE WHO VOTED IN 2004 WERE BUSH 2000 VOTERS?

YES

or

NO


I AM ASKING FOR A ONE WORD ANSWER.
CAN YOU GIVE IT?
WILL YOU GIVE IT?
DARE YOU GIVE IT?

Once I have the answer, we can move on to the next stage, a more interesting discussion, something more worthy of your academic standing..

[OnTheOtherHand]

104. my ONE WORD ANSWER: irrelevant

If you are not willing to make actual arguments yourself, nor to engage my actual arguments, then why are you here?

[TruthIsAll]

105. Irrelevant? You can't even answer a simple YES/NO question?

And yet you have infinite capacity to interminably drone on and on when it comes to debunking the work of USCV and myself.

You cannot answer my simple 3rd grad arithmetic question. And yet you have the disengenuous GALL to engage us in debate. All you have been doing is WASTING our time and energy here. What a tremendous disappointment you have been. Your credibility has been thoroughly shot.

You have not stipulated to one damn fact in all the time you have been posting here which in any way indicated that fraud is a viable explanation for the exit poll discrepancies. All you have tried to do is spin and divert.

OTOH, you are very slick, but not slick enough. You see, 1+ 1 still equals 2. As a professor of Poly Sci, you must know that.

Who would have thought that you would succumb to a simple problem in division? No, OTOH, you have not succeeded in dividing us; in fact, knowing your true agenda, you have made us stronger than ever.

I hope you enjoy the rest of your stay at DU.

[OnTheOtherHand]

107. what is my true agenda? what is yours?

TIA, your simple division problem will not win the fraud argument. And more importantly, it will not win electoral reform.

If you insist on publicly making arguments that public opinion professionals regard as obviously ignorant, that might benefit the cause of electoral reform, but the smart money is not that way.

Maybe once I am gone, you will be able to set aside your ego, think this one through, and get the answer right.

[TruthIsAll]

116. OTOH, IS IT YES OR IS IT NO? DUERS ARE STILL WAITING FOR A RESPONSE

ONE WORD WILL SUFFICE.

RIGHT HERE.
RIGHT NOW.

ONE SIMPLE DIVISION.

YOU HAVE FAILED 3RD GRADE MATH.

I AM GIVING YOU A CHANCE TO RETAKE THE EXAM.

IS IT POSSIBLE THAT 43% OF THOSE WHO VOTED IN 2004 WERE BUSH 2000 VOTERS?



IS IT
YES

OR

IS IT
NO?

[OnTheOtherHand]

137. ok, let's try it here

(Edit: remove inadvertent HTML)

I prefer the multiplication problem, although even that is problematic, but if there were 122.3 million voters in 2004, 43% of that number would be 52.6 million, which is more votes than Bush even got in 2000.

So, I guess we can go with "No."

But I have several questions. I will not chase you around the board haranguing you to answer them. It's a pure gut check for you, a chance to live up to your screen name.

Why do you insist that 43% and 37% (the reported percentages of Bush 2000 and Gore 2000 voters) were used as "weighting percentages" or "weighting factors" or "weighting component(s)" even though (as WI Independent pointed out, and Internut has also tried to explain) the exit poll is designed to be reweighted to official returns as the returns arrive? Do you actually intend to insist that the exit poll was not weighted to match the official returns?

You say in #124 that "AN IMPOSSIBLE RESULT (THE POLL) WAS MATCHED...TO ANOTHER IMPOSSIBLE RESULT (THE VOTE)."

When you refer to the exit poll as an "impossible result," are you referring to the reported percentages of Bush 2000 and Gore 2000 voters?

--If no, then why in heck are we talking about the 2000 figures?

--If yes, since the 2002 National Election Study yields a similar imbalance of reported Bush 2000 and Gore 2000 voters, should we conclude that it, too, was matched to the result of the 2004 election?

TIA, this stuff may work for some audiences, but it will not convince people who actually know about public opinion research. If you are content with that, that is your choice. I think it is a bad choice for the movement.

[TruthIsAll]

143. Thanks for finally answering the question

Now you agree that 43% is impossible.
But you still don't (or won't) get it.
I will be brief.
I have a new thread up which we should move to.

You:
Why do you insist that 43% and 37% (the reported percentages of Bush 2000 and Gore 2000 voters) were used as "weighting percentages" or "weighting factors" or "weighting component(s)" even though (as WI Independent pointed out, and Internut has also tried to explain) the exit poll is designed to be re-weighted to official returns as the returns arrive? Do you actually intend to insist that the exit poll was not weighted to match the official returns?

You say in #124 that "AN IMPOSSIBLE RESULT (THE POLL) WAS MATCHED...TO ANOTHER IMPOSSIBLE RESULT (THE VOTE)."

When you refer to the exit poll as an "impossible result," are you referring to the reported percentages of Bush 2000 and Gore 2000 voters?

Me:
Yes.
Let me spell it out for you and ALL readers as best I can.
As I'm sure you know, but pretend not to, Kerry won every exit poll time line, up to and including the 13047 respondents at 12:22am, which had the 41/39% how voted split.

Obviously, this could not stand.
Twelve hours later, the Final was released at 1:25pm.
Just 613 respondents were added to bring the total to 13660.
Either weights or percentages or both had to be adjusted from the 13047 time line to MATCH THE VOTE.

And so it was done.
The 41% jumped to 43%.
The 57% Kerry share of NEW voters dropped to 54%.
Mission accomplished.
The 51-48 Kerry win was reversed to a 51-48 Bush win.
Simple as that.

Now, let's move on to the Hobson's choice thread, in which we will discuss the impossible 43/37 split and rBr.

You:
--If yes, since the 2002 National Election Study yields a similar imbalance of reported Bush 2000 and Gore 2000 voters, should we conclude that it, too, was matched to the result of the 2004 election?

Me:
That is very interesting.
No sale.
We are talking about the purported 2004 exit poll respondents, aren't we? Anything else is irrelevant to the discussion. The 2004 exit poll took place in 2004, not 2002.

See the Hobson's choice thread.
There is this little thing called rBr.

Let's now advance the discussion in that pristine, newly born thread.

[OnTheOtherHand]

144. you can flee this thread if you want

and you have already demonstrated your ability to spawn new ones. But they will not save your argument.

"As I'm sure you know, but pretend not to...."

I think at least three of us have been very explicit about the NEP's preannounced practice of reweighting its projections to official returns as the returns become available. (It obviously is not their practice to reweight the projections to some target "2000 vote" percentages, since by your own account all their projections through 12:22 am violated that constraint -- if it had been a constraint.)

So, do we all agree that the weighting is done to the 2004 official returns, not to the 2000 reported votes?

Me: "If yes, since the 2002 National Election Study yields a similar imbalance of reported Bush 2000 and Gore 2000 voters, should we conclude that it, too, was matched to the result of the 2004 election?"

You: "That is very interesting. No sale. We are talking about the purported 2004 exit poll respondents, aren't we? Anything else is irrelevant to the discussion. The 2004 exit poll took place in 2004, not 2002."

OK, so if anything else is irrelevant to the discussion, then that includes the 2000 vote totals, right? Or is there some rule here that evidence only counts if it comes from a year that is a multiple of 4?

No, wait, the 2004 NES (a pre-election survey) has a similar imbalance, also. So maybe the rule is....

Hey, good luck with that. Obviously if you get to decide what evidence we are allowed to consider, you will win every time. But in the real world, you don't get to decide that. And so you will lose. Not just for yourself, but for all of us. That is a shame.

[autorank]

120. The pollsters are like the media, held in very low regard.

They don't seem to put a discrete version of their trade in those polls that run every so often of "respected" professions. Well, they are part of the media because they're the media's creature, subservient subcontractor. In this fully informed judgment, based on daily review, the public is correct, Vox populi, vox Dei. So much for casting aspersions on people who correct the creatures.

Before you leave, will you ever answer the question? In fact, if you answer the question, you are free to leave!

And when you leave, will your crew depart with you? Just asking.

[mgr]

77. God I wish I could make supple arguments like this

just think of what I could do, pigs could fly.

Without preamble or syllogism, portions of the NEP methodology are more acceptble than others because they better fit your pet theory. Ah, yes, the leap of faith that forever separated Continental Rationalism from English Empiricism is alive and well in TIA's mathematics.

How do you know that you are thinking, so that you can then know that you exist? But if you know that you exist, how do you know that you are thinking?

Your argument is illogical in the extreme. If you agree that NEP follows their methodology, then you captitulate (lose your head?). You are missing a few steps in that argument. What does the final weighting have to do with what the appropriate error term is to be?

When a gaussian curve is skewed, what happens to the measures of central tendency? When it is kurtotic or leptokurtotic (sp.)? Laugh it off, but to a real statistician this is where they earn their meal.

Mike

[TruthIsAll]

89. OTOH, show DUers how Bush 2000 voters could equal 43% of 122.3 mm in 2004.

Show us how it is possible.

Right here.
Right now.

Do you enjoy this?

I do.

[OnTheOtherHand]

90. dude, they DON'T HAVE TO

TIA, you can declare victory until you are blue in the face, but you still will never convince folks who know about surveys. That is, in a nutshell, why your arithmetic exercise has not taken AAPOR by storm.

The question is not whether 6% more of 2004 voters actually voted for Bush rather than Gore in 2000. It is whether it is plausible that 6% more of 2004 voters would have reported voting for Bush rather than Gore.

If you are able to indicate your comprehension of those two sentences, we can continue. Otherwise, you have a brave -- but UTTERLY IRRELEVANT -- demonstration that "it can't be true that 43% of 2004 voters were Bush 2000 voters."

[TruthIsAll]

92. Answer the question, please.

Is it possible that 43% of 2004 voters were Bush 2000 voters?

YES

or

NO

[Internut]

93. Answer:

No, it is not possible that 43% of 2004 voters were Bush 2000 voters.

But in a poll, the result of "43% of those who were asked said they voted for Bush in 2000" is plausible, for reasons that were explained to you a dozen times at least and that you keep ignoring.

For some reason you keep insisting that if the poll says that 43% of the 2004 voters voted for Bush in 2000, that implies that this actually happened. As was pointed out to you, that could quite easily be explained by people misreporting their past voting behaviors, like they regularly do in other polls.

Another thing you keep misstating is that the "43%/37% in 2000" was the weight that was applied. That is an invention of yours with no basis in reality. No pollster would ever use that statistic as a weight to adjust other variables, because that statistic cannot be relied upon.

Now, since you got the answer to the question that you asked, it is only fair that you answer a question as well:


In the The 2002 American National Election Study there was this question:

V203111: Who did R (respondent) vote for in the 2000 presidential election?

With the following results:

1. Al Gore             431  44%
3. George W. Bush      502  51%
5. Ralph Nader          32   
7. Other {SPECIFY}       8   
8. Don't know            1   
9. Refused              10   
0. NA                    1   

Note: these are raw results, not "weighted", not tampered with in any way.

Do you consider this 2002 American National Election Study a "FRAUD", judging by those results?

[autorank]

112. "No, it is not possible that 43% of 2004 voters were Bush 2000 voters."

Internut gets the prize!!!!

Vanna, open curtain #3!

[Internut]

113. Yes, and 2+2=4. That's about as relevant

as the possibility that 43% of 2004 voters were Bush 2000 voters. The "43%" numbers was not "number of 2004 voters who were Bush voters in 2000", no matter how many times you and TIA repeat that it was. The "43%" number was "number of 2004 voters who SAID they were Bush voters in 2000", and as such, that number is quite possible.

[Internut]

91. No, they could not

equal 43% of 122.3m in 2004. But in a poll, the result of "43% of those who were asked said they voted for Bush in 2000" is plausible, for reasons that were explained to you a dozen times at least.

For some reason you keep insisting that if the poll says that 43% of the 2004 voters voted for Bush in 2000, that implies that this actually happened. As was pointed out to you, that could quite easily be explained by people misreporting their past voting behaviors, like they regularly do in other polls.

Another thing you keep misstating is that the "43%/37% in 2000" was the weight that was applied. That is an invention of yours with no basis in reality. No pollster would ever use that statistic as a weight to adjust other variables, because that statistic cannot be relied upon.

[TruthIsAll]

95. internut, thank you for stipulating that, but you misstate as usual

you say:

No, it could not equal 43% of 122.3m in 2004. But in a poll, the result of "43% of those who were asked said they voted for Bush in 2000" is plausible, for reasons that were explained to you a dozen times at least.

You say:
For some reason you keep insisting that if the poll says that 43% of the 2004 voters voted for Bush in 2000, that implies that this actually happened.

I say:
PURE HOGWASH!
IS YOUR BRAIN FUNCTIONING?

I SAID THE 43% WAS IMPOSSIBLE.
AND YOU AGREE.

THAT IS WHAT I SAID.

You say:
As was pointed out to you, that could quite easily be explained by people misreporting their past voting behaviors, like they regularly do in other polls.

I say:
That is totally irrelevant to the discussion.
The point is simply this, and you have answered correctly:
IT IS IMPOSSIBLE THAT 43% OF TOTAL 2004 VOTERS COULD HAVE BEEN BUSH 2000 VOTERS.

AND YOU AGREE.

You say:
Another thing you keep misstating is that the "43%/37% in 2000" was the weight that was applied. That is an invention of yours with no basis in reality. No pollster would ever use that statistic as a weight to adjust other variables, because that statistic cannot be relied upon.

I say:
YOU ARE DELUSIONAL.
WE ARE TALKING ABOUT THE "HOW VOTED IN 2000 DEMOGRAPHIC" IN THE FINAL NATIONAL EXIT POLL.

TAKE A LOOK AT THE FINAL NATIONAL EXIT POLL.
ITS RIGHT THERE FOR ALL THE WORLD TO SEE.
43% IS A WEIGHTING, NUT.
YOU ARE SADLY LACKING IN EVEN RUDIMENTARY KNOWLEDGE OF EXIT POLLING.
DO YOU REALLY EXPECT TO CONVINCE ANYONE THAT YOU KNOW WHAT YOU ARE TALKING ABOUT, WHEN YOU DON'T KNOW WHAT A WEIGHTING FACTOR IS?
THAT IS SO BASIC.

BUT I DIGRESS.

NUT, THEY DID JUST THAT. THEY CHANGED THE WEIGHTINGS FROM FROM 41% AT THE 13047 EXIT POLL TIME LINE TO 43% IN THE FINAL 13660.
ARE YOU AWARE OF THAT?

NEXT QUESTION:
YOU HAVE AGREED 43% IS IMPOSSIBLE.
GOOD (WHEW!)

IF THAT IS THE CASE, THEN IS IT VALID TO USE THE 43/37 WEIGHTINGS AND CORRESPONDING VOTING PERCENTAGES TO DETERMINE THE FINAL BUSH /KERRY NATIONAL PERCENTAGE?

I ASSUME THAT YOU KNOW HOW TO DO THE CALCULATION.

DO YOU?

IF NOT, I WILL GLADLY HELP.

THE BALL IS IN YOUR COURT.
IS THE HEAT GETTING TO YOU YET?

ARE YOU SURE YOU WANT TO CONTINUE?

[Internut]

102. You seem to have no idea how polls are weighted. Let me try to explain.

When they get raw data, it may be misbalanced. Too many males, or too many young people, or too many whites, or Latinos, etc etc etc. In order to make sure that the sample is close to being "representative", the sample is reweighted. What that means is that some variables (such as Gender, or Ethnicity or Age) are adjusted to confirm to known, correctly representative weights. Those are the weights that we are talking about.

When you do that, other, secondary variables, that were NOT used as weights and that are NOT known (such as "whom did you vote for") are weighted and changed.

What YOU are claiming is that the "43/37 whom did you vote for in 2000" variable was used as a weight. That is absurd, since it is an unreliable result. People do not lie about their gender, so that can be used as a weight, if you know what the representative sample should contain. People regularly lie about whom they voted for in the past.

[TruthIsAll]

118. DID YOU SEE POST # 42? I'LL REPRINT IT HERE. OTOH, THIS IS FOR YOU, TOO.

ARE YOU STILL BREATHING?

TAKE A LONG, HARD LOOK AT SCENARIO II
YOU KNOW WHAT THAT IS?
THAT'S THE FINAL NATIONAL EXIT POLL HOW VOTED IN 2000
DEMOGRAPHIC.

NOW, LOOK AT THE LEFT COLUMN.
DO YOU KNOW WHAT THAT IS?
THOSE ARE THE FAMOUS WEIGHTS.

DO YOU KNOW WHAT THEY ARE USED FOR?
TO CALCULATE FINAL VOTE PERCENTAGES, BASED ON POLLING RESULTS
FOR KERRY, BUSH, AND NADER/OTHER.

DO YOU KNOW WHO WON THE POLL?
BUSH.
BY 51-48.

DO YOU REALIZE THAT 43% WAS USED AS A WEIGHTING COMPONENT OF
BUSH'S VOTE?
DO YOU REALIZE THAT 37% WAS USED AS A WEIGHTING COMPONENT OF
KERRY'S VOTE?

CAN YOU REPRODUCE THE CALCULATIONS?
OR MUST I SPELL THEM OUT FOR YOU.

NOW DO YOU SEE WHY THIS IS VERY RELEVANT?
ARE YOU LISTENING, OTOH?



HERE ARE THE 13047 AND 13660 HOW VOTED IN 2000 WEIGHTS AND
PERCENTAGES 
 
The maximum possible percentage of Bush 2000 voters who could
have voted in 2004 is 39.82%.

Bush 2000 voters: 50.456 million
Bush 2000 voters still alive = 48.69mm

Total 2004 voters: 122.3 mm.

Maximum Bush 2000 voters as a percentage of Total 2004
voters:
Bush 2000/Total 2004 = 48.69/122.3 = 39.82%

The same calculation for Gore voters = 40.25%

Now lets look at the National Exit Poll:


IMPOSSIBLE SCENARIO I:
13047 respondents: 12:22am
Kerry wins easy - even with an IMPOSSIBLE 41% Bush/39% Gore
split

	Mix   Bush	Kerry	Nader	Bush	Kerry	Nader
No	17%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	39%	8%	91%	1%	3.1%	35.5%	0.4%
Bush	41%	90%	10%	0%	36.9%	4.1%	0.0%
Other	3%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				47.38%	51.23%	1.21%

Probability of Bush going from 47.38% (poll) to 50.73%
(vote):  
1 in 38,498,885,514


IMPOSSIBLE SCENARIO II:
Final Exit Poll (1:25pm)- 13660 respondents:
BUSH WINS: PURE FICTION - 43% IS EVEN MORE IMPOSSIBLE

MIX	Bush	Kerry	Nader	Bush	Kerry	Nader
NO    17%	45%	54%	1%	7.7%	9.2%	0.2%
Gore  37%	10%	90%	1%	3.7%	33.3%	0.4%
Bush  43%	91%	9%	0%	39.1%	3.9%	0.0%
Other  3%	21%	71%	8%	0.6%	2.1%	0.2%
100%				51.11%	48.48%	0.78%



PLAUSIBLE SCENARIO III:
Using the calculated maximum possible weightings, Kerry wins
in a  landslide: 

Kerry	63.89mm (52.24% 
Bush	56.77 (46.42%)
Nader/Other 1.49 (1.22%)


VOTE2000	
      Mix	Bush	Kerry	Nader	Bush	Kerry	Nader
No	17.00%	41%	57%	2%	7.0%	9.7%	0.3%
Gore	40.24%	8%	91%	1%	3.2%	36.6%	0.4%
Bush	39.82%	90%	10%	0%	35.8%	4.0%	0.0%
Other	3.00%	13%	65%	16%	0.4%	2.0%	0.5%
	100%				46.42%	52.24%	1.22%
							



 

[Internut]

121. Let's see -

"DO YOU REALIZE THAT 43% WAS USED AS A WEIGHTING COMPONENT OF
BUSH'S VOTE?
DO YOU REALIZE THAT 37% WAS USED AS A WEIGHTING COMPONENT OF
KERRY'S VOTE?"

No they were not, and it shows an amazing ignorance of the polling process to think that they were.

[TruthIsAll]

124. Oh, to reveal your ignorance so graphically....

in this manner, knowing all of DU is watching...

I'm talking about the GODDAMN FINAL EXIT POLL, NUT.

NUT, THE EXIT POLL WAS MATCHED TO THE VOTE.
AND THE 43%/37% WEIGHTS WERE WHAT DID IT.

NOW, YOU HAVE ALREADY STIPULATED THAT THE 43% WAS IMPOSSIBLE.

SO, WE HAVE ARRIVED AT THIS.
TAKE MY HAND AS I LEAD YOU TO THE LIGHT.
TO THE LIGHT OF TRUTH.

AN IMPOSSIBLE RESULT (THE POLL) WAS MATCHED...TO ANOTHER IMPOSSIBLE RESULT (THE VOTE).

CAN YOU WRAP YOUR CEREBELLUM AROUND THAT ONE?

[Internut]

125. Once again -

you have no idea what weights were used to weigh the final exit poll. None. Your contention that it was the "whom did you vote for in 2000" question that was used as weight is laughable.

[Name removed]

128. Deleted message

Message removed by moderator. 
[link:www.democraticunderground.com/forums/rules.html|Click
here] to review the message board rules.

[Internut]

130. You refuse to reply YES/NO to the question -

but you demand that others reply YES/NO to yours.

And I did read your posts. All I can do is tell you again - you have no idea what weights were used to weigh the final exit poll. None whatsoever. Your contention that it was the "whom did you vote for in 2000" question that was used as weight is laughable on its face. You can quote numbers all you want - but as long as they are based on the imagined "fact" that the Mitofsky used the 43/37 numbers as weights to adjust the other variables they don't mean much.

[LightningFlash]

152. And you failed the argument with your own respot...

You say "This includes gender, ethnicity, age, race, color and so on"

You say explicitly that this portion is re-weighted.

Everyone who knows precisely how polls work know this. And if you are intent on following your own logic than you must actually agree right here.

http://www.census.gov/population/www/socdemo/voting/cps2004.html

From the OFFICIAL census results and from 2002 onward, The GENDER, RACE, AGE mainstays have been reported and accurately portrayed.

So then for this fact, do you dispute that over 10% of the population at least mis-reported who they vote for in 2004 even though it has only happened one other time in the final weightings?

And do you dispute your own logic that 4% of the gender side in this election weighting, was in fact subtracted from the final data leaving less exit poll respondents who voted for Kerry than who voted for Bush in the final batch?

DO you dispute your own logic? Answer the question. If you don't, then the argument is finished. You just refuted yourself. If you do, then explain for everyone to see why you dispute it.

They're census statistical facts. Not petty words. :smoke:

[OnTheOtherHand]

153. am I confused, or did you just change the subject? or both?

Did you just totally drop the argument about recalled 2000 vote, and switch over to a different argument about discrepancies between the 2004 exit poll and the 2004 Current Population Survey?

And in the process, did you just agree that the 2004 exit poll is not weighted by recalled 2000 vote?

(Your insistence on calling the CPS "census results" and "census statistical facts" remains perplexing. Why not call things by their names?)

[LightningFlash]

154. On the contrary....

I just pointed back to how it is weighted and was used in that very purpose. Apparently you are the one who is confused.

I was not even addressing your argument which has been rebutted elsewhere, but instead Internut's argument on Gender and Population.

[OnTheOtherHand]

155. so, you have evidence that

the 2004 exit poll was weighted, after the fact, using 2000 recalled vote as a basis to calculate the weighting factor? You think you can use CPS data to determine post facto which variables E/M used for their weights, regardless of the variables they tell us they used for their weights? and you can prove that 2000 recalled vote was one of them? or perhaps the only one?

And my argument was rebutted elsewhere by TIA's trenchant intellectual critique: "I don't care what your poll said"?

Good luck with that.

[LightningFlash]

156. No....

It was rebutted ages ago when you argued against the census results. It was rebutted when you claimed "Voters News Service" had nothing to do with the final 2002 exit results and averages, when they in fact did. It was rebutted when it was proven DRE discrepancies were found all through-out the 2002 Gubernational race.


In short, you rebutted yourself ages ago.

[OnTheOtherHand]

157. high marks for creativity. n/t

[WI Independent]

75. TruthIsAll, They weren't "set" to match the result...

The design of the system is such that they will match in the end.


"How are projections made?
Projections are based on models that use votes from three (3) different sources -- exit poll interviews with voters, vote returns as reported by election officials from the sample precincts, and tabulations of votes by county. The models make estimates from all these vote reports. The models also indicate the likely error in the estimates. The best model estimate may be used to make a projection if it passes a series of tests."

As more of the second and third sources are plugged into the model the model will naturally more closely match tabulated results.

Most of your work is done with a portion of one of the three inputs... incomplete data.

Why did they use 3 sources if they only needed one?

I think this about the third time I've asked this question with no reply addressing it... I'm beginning to feel like I'm "persona non grata". :(

[autorank]

110. Hello there!

"As more of the second and third sources are plugged into the model the model will naturally more closely match tabulated results."

What does this mean.

This sounds like the Republican strategy of saying "fixed" really didn't mean "fixed."

At 12:22 it was Kerry-Bush, 51-48 with 13000 or so respondents; at 1:00, it was just the reverse with 13600 or so. Give me a break.

You have a naive faith in this process that must comfort you and allow you to sleep well at night knowing your country is well run by honest people. Congratulations.

[WI Independent]

140. 13000 and 13600 are ONE of the THREE inputs...

That's not the only thing that changed from 12:22 to 1:00. Read the FAQ, NEP designed it from the beginning for three inputs. As they plugged in more of the actual tallies from the second and third source the results were subject to change even if there were NO more exit poll interviews included.

It's not whether or not I have faith in the process, it's whether or not this data is capable of verifying that process. IMO, it's not. I'm quite aware of the human capacity for dishonesty.

[TruthIsAll]

111. No, I use the published data, after the fact.

I am not running a poll, I am reporting the results.

And these results all said Kerry won, by 51-48:

3:59 pm 8349 respondents
7:33 pm 11027 respondents
12:22am 13047 respondents.

But wait, over the next 12 hours, they added 613 more:
1:25pm 13660 respondents
Bush 51-48

[WI Independent]

141. TruthIsAll, I'm repeating myself...

You are reporting ONE of THREE sources the poll was designed to have as inputs. The affect of the other two sources is what caused the shift, not the 613 additional responses. The other 2 inputs are the weightings. This was the design from the beginning, not some 11th hour cover-up.

If the first source was capable of being accurate alone, why would they go through all of the extra work to include the others?

I know it's hard to accept that much of your efforts are based on incomplete data and doesn't prove anything, but if you do maybe you can refocus your efforts on something more meaningful.




"How are projections made?
Projections are based on models that use votes from three (3) different sources -- exit poll interviews with voters, vote returns as reported by election officials from the sample precincts, and tabulations of votes by county. The models make estimates from all these vote reports. The models also indicate the likely error in the estimates. The best model estimate may be used to make a projection if it passes a series of tests."

[mgr]

72. What happens to your equation when the gaussian curve is

skewed strongly in one direction, such that say the mode is one half a standard deviation from the mean? What happens to the 95% CI, is it skewed as well? What is it that you have to do to make it parametric again? And what does that do to the arithmetic measures you obtain to set your ranges?

Mike

[tommcintyre]

94. Actually, YOU don't understand "what wikipedia is"

Or you do, and you are trying to spin or mislead?

As RonB said here, and I think this statement epitomizes (and sums up) an overall problem:

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x374482#378637

We had our moment when you finally came up with a specific point - but now its back to the misunderstanding and spin I guess.

From another post I made over here, regarding your claims against wikipedia:
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=378368&mesg_id=379967

YOU REALLY SHOULD DO YOUR HOMEWORK BEFORE YOU CRITICIZE

C'mon oto, YOU can do better than that!

Did you bother to check the guidelines to making edits before spouting off? Did you read how contributors have to negotiate a neutral point of view (NPOV)? About the restrictions on citing sources? Did you bother to read the edit or history pages of the individual pages? <I use wikipedia here because of the quality checks/sourcing; and the fact that it saves time hunting down ALL this info that I have CONFIRMED elsewhere.>

In fact, DID YOU BOTHER TO DO ANYTHING OTHER THAN JUMP TO A CONVENIENT CONCLUSION?

oto, I honestly think you are really "grasping at straws" lately. (Pulled from all those "straw-men" you have created around here, maybe? ;) )

I think your arguments are honestly getting weaker and weaker. I don't think you can find better arguments, so, maybe, is it time to seriously consider changing your positions?

There is no shame in admitting it when you've made mistakes; so I strongly suggest you consider it, and move on. I think if you did, you could make some good contributions here.

One final point: No, I did NOT write any of the information at wikipedia yet. I will soon (about Mitofsky, rBr and the C/B Election Hearing - "weak spots" that need attention). That's WHY I have started to do MY homework and already know about the things I mentioned above (NPOV, etc.). When I do, you will know, because I will be sure to put my name on it.
----------------------------------

A link to get you started:

http://en.wikipedia.org/wiki/Neutral_point_of_view
Wikipedia:Neutral point of view
"Wikipedia policy is that all articles should be written from a neutral point of view: without bias, representing all views fairly. According to Wikipedia founder Jimbo Wales, NPOV is "absolute and non-negotiable". <1> (http://mail.wikipedia.org/pipermail/wikien-l/2003-November/008096.html)

For guidance on making an article conform to the neutral point of view (NPOV), please see the Wikipedia:NPOV tutorial."
----------------------------------

P.S. I don’t see any point in answering any more of your direct questions until you get a handle on this “bias” (cognitive dissonance?) problem. Until you do, it is CLEARLY and excercise in futility. <Although, I certainly will comment if (IMO) I see you attempting to spin or miss-lead.>

Good luck! I'm really pulling for ya. :D

[autorank]

96. I second all of the above. Nice analysis!!!

[OnTheOtherHand]

98. oh, wow, better jump off that train... n/t

[autorank]

100. You learned nothing from the "nice" thread? No milk, straight to bed!

:rofl:

[OnTheOtherHand]

97. gee, you think it'll go better for you over here?

Tom, the fact that Wikipedia has a policy of "neutral point of view" does not demonstrate that all its articles are factually infallible (or neutral, for that matter).

As I pointed out on the other thread, British exit polls in 1992, 1997, and 2001 all generated lead errors of 2.6 points or more. (I guess those could be construed as "fractions of a percentage point," if the fractions are allowed to be improper.)

http://pollingreport.co.uk/record.html

[tommcintyre]

106. Oh... I think things go pretty well "over here" too. ;)

You see, both wiki AND DU are pretty darn good examples of true democracy in action.

Given enough time (and usually pretty quickly), the "truth" is "negotiated" and, consequently, revealed.

So, the "truth-seekers (and tellers) are also revealed rather quickly, aren't they? ;)

And, in contrast, those who "have mostly ducked, bobbed, weaved, misled, misdirected, "misunderstood", obfuscated... and, oh yes spun like a top" are also quickly revealed, aren't they? ;) <Similar to what is generally found on the MSM (main stream media), for example.>

So, those of us who are LEGITIMATELY trying to "save democracy" are very fortunate to have DU, wikipedia - and the internet in general. You see, the internet IS very democratic (and I'm not referring to the party affiliation here). So the truth can still be found here.

Unlike, the MSM for example. In contrast, there, the words are controlled by a very few, and misleading "ten second sound-bites" are quickly expelled from obfuscaters (spinners?) mouths, with little or no hard-copy track record to be remembered (and even re-read, like DU?).

So, after attempting to denigrate wikipedia, you asked me the following question: "gee, you think it'll go better for you over here?" <wikipedia>

So, now I will ask YOU a similar question:

"gee, you think it would go better for you over at the MSM?"

[OnTheOtherHand]

108. strangely, "British exit polls" appear nowhere in your answer. n/t

[tommcintyre]

109. And they won't "appear" 'til ya fix your problem

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=378368&mesg_id=379989
snip
"So (IMO), it is long overdue that I no longer take your questions seriously until you straighten out this problem that I perceive.

C'mon oto, be honest. I've answered much tougher "diversions" (IMO) than you have thrown up in that last post, haven't I?

<Caution: If you bother to answer this last question at all; your answer (one way or another) will be VERY revealing, won't it?>

GOOD LUCK! :D "

Actually, it's a pretty simple answer (and, IMO a pathetic and desperate question); but, as I said, I'm tired of these one-sided "debates" where you don't keep up your side of the "good faith" bargain.

But, let me know if you REALLY need an answer to this "burning" question (I don't want you losing sleep over it, or anything ;) ); I will see if I can get someone else to answer it for you (like I said, it's NOT that complicated). But, you have to promise you won't keep pestering them after they post it, OK? ;)

EDIT: for a superfluous "n/t".

[WI Independent]

53. I don't agree with the arguments...

"Because final published exit polls in America are matched to vote counts, they cannot be used to determine election fraud. However, in the 2004 election, pre-matched exit polls were leaked onto the internet."

My position is the raw exit poll data leaked is incomplete and not usable because of the design of THIS poll.

"Leaks of exit poll figures for the 2004 presidential election, mainly via the Internet, appeared to indicate a victory for John Kerry. The discrepancies between the exit poll data and the vote count that where outside of the margin of error, ..."

There is no MOE on a subset of the data collected. The MOE numbers are for the completed poll including weightings by actual vote tallies.

Citing other polls which may not (and in some cases definitely were not) designed the same tells us nothing about THIS poll.

[tommcintyre]

119. Can you back up your "position" with more than your opinion?

"My position is the raw exit poll data leaked is incomplete and not usable because of the design of THIS poll."

Any credible sources for this opinion?

"There is no MOE on a subset of the data collected. The MOE numbers are for the completed poll including weightings by actual vote tallies."

Again, that's a convenient interpretation for your "position"; but not very applicable in this unusual situation.

"Citing other polls which may not (and in some cases definitely were not) designed the same tells us nothing about THIS poll."

WOW! So ya think the Germans, and all those other countries are so much better at designing exit polls? Since it can be done by so many other countries, wouldn't it be incredibly stupid for us to fall short on this? C'mon! Give us a break!

[WI Independent]

139. I have... I'll post the link again...

"How are projections made?
Projections are based on models that use votes from three (3) different sources -- exit poll interviews with voters, vote returns as reported by election officials from the sample precincts, and tabulations of votes by county. The models make estimates from all these vote reports. The models also indicate the likely error in the estimates. The best model estimate may be used to make a projection if it passes a series of tests."

Note: THREE sources... the exit poll interviews are only ONE... incomplete.

The MoE given by NEP is for the completed poll, which says Bush won. Standard Moe calculations are for statistically random samples, which the exit poll interviews were not.

The definition of a good design is dependent on the objective. Other countries aren't better... they were designing to a different objective. The NEP was hired by the networks to give the talking heads air time filler Tuesday night. They seemed to have met that objective fairly well.

[tommcintyre]

142. That's only how the NEP chooses to use the exit poll data

That suits their purpose to give an early report on how the election is going. This doesn't imply the exit polls are faulty; and, therefore, unsuitable to serve the same purpose that exit polls are used for around the world (as a check on the accuracy of the vote count, and to give an early indication of who won). True, once Mitofsky et. al. has had a chance to "weight" ("doctor") the exit polls, they are contaminated. But, the early "leaking" was NOT contaminated.

To claim that the undoctored early exit polls are unreliable, simply because the people who paid for them did not intend for them to be used as a check of election results, is very bad logic.

Like I said earlier, this is among the weakest arguments against the election being stolen; and that's pretty weak, considering just how weak all the other arguments are.

[TruthIsAll]

123. That is incorrect. The sample-sizes and therefore MoEs are available

for category groupings.

For example, over 3000 responded to the How Voted Category.
We can calculate the MoE; its under 1.80%. But that is immaterial here. The relevant question is this:

What could have been the maximum Bush 2000 voter turnout as a percentage of total 2004 voters?

I believed that earlier you correctly told us: 39.82%
(48.7/122.3 = 39.82%)

[WI Independent]

138. That MoE calculations...

are only good for a truly random sample... this was not even close to a random sample in statistical terms. But you already know that.

[TruthIsAll]

150. OH, YES IT WAS, BUT YOU DON'T KNOW THAT; MITOFSKY DOES

The MoE Formula: Mitofsky agrees its 1.0% for over 8000 respondents

Edited on Fri Apr-15-05 05:15 AM
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=357345

[OnTheOtherHand]

151. not quite

As an engineer, you know that 1 is not the same as 1.0.

From your calculated value for 50%, n = 100, of 9.8%, and the MoE table value15%, we can infer that the design effect (at least according to the table) is on the order of 15/9.8 = 1.5 or so.

So, if the nominal MoE for (say) 8000 respondents is (approximately)

1 / sqrt(8000) = 1.1%,

the actual MoE would be closer to 1.7%.

It appears that they did some very funky "rounding" in order to simplify the table a bit. Or maybe you can come up with some other design effect estimate that fits better.

[TruthIsAll]

158. Your "design effect" is designed to obfuscate.What do you need:20%,30%,50%

If you need more "effect" to make your case, I will
take it up to 200%, no further.




MoE = 1/sqrt (N)
.............Design Effect
N	MoE	20%	30%	50%
500	4.47%	5.37%	5.81%	6.71%
1000	3.16%	3.79%	4.11%	4.74%
1500	2.58%	3.10%	3.36%	3.87%
2000	2.24%	2.68%	2.91%	3.35%
2500	2.00%	2.40%	2.60%	3.00%
3000	1.83%	2.19%	2.37%	2.74%

3500	1.69%	2.03%	2.20%	2.54%
4000	1.58%	1.90%	2.06%	2.37%
4500	1.49%	1.79%	1.94%	2.24%
5000	1.41%	1.70%	1.84%	2.12%
5500	1.35%	1.62%	1.75%	2.02%
000	1.29%	1.55%	1.68%	1.94%
6500	1.24%	1.49%	1.61%	1.86%
7000	1.20%	1.43%	1.55%	1.79%
7500	1.15%	1.39%	1.50%	1.73%
8000	1.12%	1.34%	1.45%	1.68%

8500	1.08%	1.30%	1.41%	1.63%
9000	1.05%	1.26%	1.37%	1.58%
9500	1.03%	1.23%	1.33%	1.54%
10000	1.00%	1.20%	1.30%	1.50%
10500	0.98%	1.17%	1.27%	1.46%
11000	0.95%	1.14%	1.24%	1.43%

11500	0.93%	1.12%	1.21%	1.40%
12000	0.91%	1.10%	1.19%	1.37%
12500	0.89%	1.07%	1.16%	1.34%
13000	0.88%	1.05%	1.14%	1.32%

13500	0.86%	1.03%	1.12%	1.29%

14000	0.85%	1.01%	1.10%	1.27%
14500	0.83%	1.00%	1.08%	1.25%
15000	0.82%	0.98%	1.06%	1.22%
15500	0.80%	0.96%	1.04%	1.20%
16000	0.79%	0.95%	1.03%	1.19%
16500	0.78%	0.93%	1.01%	1.17%
17000	0.77%	0.92%	1.00%	1.15%
17500	0.76%	0.91%	0.98%	1.13%
18000	0.75%	0.89%	0.97%	1.12%
18500	0.74%	0.88%	0.96%	1.10%
19000	0.73%	0.87%	0.94%	1.09%
19500	0.72%	0.86%	0.93%	1.07%
20000	0.71%	0.85%	0.92%	1.06%
20500	0.70%	0.84%	0.91%	1.05%
21000	0.69%	0.83%	0.90%	1.04%
21500	0.68%	0.82%	0.89%	1.02%
22000	0.67%	0.81%	0.88%	1.01%
22500	0.67%	0.80%	0.87%	1.00%
23000	0.66%	0.79%	0.86%	0.99%
23500	0.65%	0.78%	0.85%	0.98%
24000	0.65%	0.77%	0.84%	0.97%
24500	0.64%	0.77%	0.83%	0.96%
25000	0.63%	0.76%	0.82%	0.95%
25500	0.63%	0.75%	0.81%	0.94%
26000	0.62%	0.74%	0.81%	0.93%
26500	0.61%	0.74%	0.80%	0.92%
27000	0.61%	0.73%	0.79%	0.91%
27500	0.60%	0.72%	0.78%	0.90%
28000	0.60%	0.72%	0.78%	0.90%
28500	0.59%	0.71%	0.77%	0.89%
29000	0.59%	0.70%	0.76%	0.88%
29500	0.58%	0.70%	0.76%	0.87%
30000	0.58%	0.69%	0.75%	0.87%
30500	0.57%	0.69%	0.74%	0.86%
31000	0.57%	0.68%	0.74%	0.85%
31500	0.56%	0.68%	0.73%	0.85%
32000	0.56%	0.67%	0.73%	0.84%
32500	0.55%	0.67%	0.72%	0.83%
33000	0.55%	0.66%	0.72%	0.83%
33500	0.55%	0.66%	0.71%	0.82%
34000	0.54%	0.65%	0.71%	0.81%
34500	0.54%	0.65%	0.70%	0.81%
35000	0.53%	0.64%	0.69%	0.80%
35500	0.53%	0.64%	0.69%	0.80%
36000	0.53%	0.63%	0.69%	0.79%
36500	0.52%	0.63%	0.68%	0.79%
37000	0.52%	0.62%	0.68%	0.78%
37500	0.52%	0.62%	0.67%	0.77%
38000	0.51%	0.62%	0.67%	0.77%
38500	0.51%	0.61%	0.66%	0.76%
39000	0.51%	0.61%	0.66%	0.76%
39500	0.50%	0.60%	0.65%	0.75%
40000	0.50%	0.60%	0.65%	0.75%
40500	0.50%	0.60%	0.65%	0.75%
41000	0.49%	0.59%	0.64%	0.74%
41500	0.49%	0.59%	0.64%	0.74%
42000	0.49%	0.59%	0.63%	0.73%
42500	0.49%	0.58%	0.63%	0.73%
43000	0.48%	0.58%	0.63%	0.72%
43500	0.48%	0.58%	0.62%	0.72%
44000	0.48%	0.57%	0.62%	0.72%
44500	0.47%	0.57%	0.62%	0.71%
45000	0.47%	0.57%	0.61%	0.71%
45500	0.47%	0.56%	0.61%	0.70%
46000	0.47%	0.56%	0.61%	0.70%
46500	0.46%	0.56%	0.60%	0.70%
47000	0.46%	0.55%	0.60%	0.69%
47500	0.46%	0.55%	0.60%	0.69%
48000	0.46%	0.55%	0.59%	0.68%
48500	0.45%	0.54%	0.59%	0.68%
49000	0.45%	0.54%	0.59%	0.68%
49500	0.45%	0.54%	0.58%	0.67%
50000	0.45%	0.54%	0.58%	0.67%
50500	0.44%	0.53%	0.58%	0.67%
51000	0.44%	0.53%	0.58%	0.66%
51500	0.44%	0.53%	0.57%	0.66%
52000	0.44%	0.53%	0.57%	0.66%
52500	0.44%	0.52%	0.57%	0.65%
53000	0.43%	0.52%	0.56%	0.65%
53500	0.43%	0.52%	0.56%	0.65%
54000	0.43%	0.52%	0.56%	0.65%
54500	0.43%	0.51%	0.56%	0.64%
55000	0.43%	0.51%	0.55%	0.64%
55500	0.42%	0.51%	0.55%	0.64%
56000	0.42%	0.51%	0.55%	0.63%
56500	0.42%	0.50%	0.55%	0.63%
57000	0.42%	0.50%	0.54%	0.63%
57500	0.42%	0.50%	0.54%	0.63%
58000	0.42%	0.50%	0.54%	0.62%
58500	0.41%	0.50%	0.54%	0.62%
59000	0.41%	0.49%	0.54%	0.62%
59500	0.41%	0.49%	0.53%	0.61%
60000	0.41%	0.49%	0.53%	0.61%
60500	0.41%	0.49%	0.53%	0.61%
61000	0.40%	0.49%	0.53%	0.61%
61500	0.40%	0.48%	0.52%	0.60%
62000	0.40%	0.48%	0.52%	0.60%
62500	0.40%	0.48%	0.52%	0.60%
63000	0.40%	0.48%	0.52%	0.60%
63500	0.40%	0.48%	0.52%	0.60%
64000	0.40%	0.47%	0.51%	0.59%
64500	0.39%	0.47%	0.51%	0.59%
65000	0.39%	0.47%	0.51%	0.59%
65500	0.39%	0.47%	0.51%	0.59%
66000	0.39%	0.47%	0.51%	0.58%
66500	0.39%	0.47%	0.50%	0.58%
67000	0.39%	0.46%	0.50%	0.58%
67500	0.38%	0.46%	0.50%	0.58%
68000	0.38%	0.46%	0.50%	0.58%
68500	0.38%	0.46%	0.50%	0.57%
69000	0.38%	0.46%	0.49%	0.57%
69500	0.38%	0.46%	0.49%	0.57%
70000	0.38%	0.45%	0.49%	0.57%
70500	0.38%	0.45%	0.49%	0.56%
71000	0.38%	0.45%	0.49%	0.56%
71500	0.37%	0.45%	0.49%	0.56%
72000	0.37%	0.45%	0.48%	0.56%
72500	0.37%	0.45%	0.48%	0.56%
73000	0.37%	0.44%	0.48%	0.56%
73500	0.37%	0.44%	0.48%	0.55%
74000	0.37%	0.44%	0.48%	0.55%

[OnTheOtherHand]

159. in other words, "I don't care about your design effects"?

Are you aware that design effects actually exist?

By the way, I've been sloppy with my terminology. The MoE multiplier in a cluster sample is the square root of the design effect. The calculated design effect is a ratio of variances.

[TruthIsAll]

36. THE 13047 WAS CLOSE TO THE TRUTH. THERE ARE TWO POSSIBILITIES.

Bush won.
In that case, 13660 was correct (or nearly so).

Kerry won.
In that case, 13047 was correct (or nearly so).

We have shown that 13660 was totally bogus.
43/37 is at least 3.2% too high for the Bush 2000 voter, and 3.2% too low for Gore.

We have shown that 13047 was closer to the truth, since 41/39 is at the edge of the 1% MoE, even though 41/39 is also mathematically impossible.

Where does that leave us?
So who won?

Q.E.D.

[WI Independent]

40. See post #28... n/t

[mgr]

59. MoE= +/- 3%.

Assuming the characteristic is typical, but it may not appear on all questionaires, so the MOE may be higher.

http://www.exit-poll.net/faq.html

Bye-Bye argument hello http://www.democrats.org/vri/ohioreport/index.html

Mike

[TruthIsAll]

61. For How Voted in 2000, MAX for Bush is 39.82%; for Gore: 40.25%

What is the MoE for the Gore vote (50.999 mm)?
What is the MoE for the Bush vote (50.456 mm)?

What is the MoE for the total 2004 vote (122.3 mm)?

What is the MoE for the ratios?
48.7/122.3 = 39.82% ?
49.2/122.3 = 40.25% ?

[mgr]

66. Since you are dealing with what you imagine to be third grade math:

What you are saying is innumerate as far as statistics go. The actual population does not have a MoE, the MoE is applied to the sample taken to characterize the population. The exit poll is the sample.

(BTW you forgot Nader in all this discussion, the fact that he disappears altogether in the 2002 sampling, suggests the band wagon effect is viable)

This is what it means for anyone who has had seventh grade math:

36.8-42.8% of those who voted for Gore participated in the 2004 election;
37.25%-43.25% of those who voted for Bush participated in the 2004 election.

Apply your simplistic demography, and the numbers fall in those ranges; and they cease to be implausible. So much for the math tutorial. Thank you.

Now, did you try that equation at all?

Did you ever look at the questionaires--only about ~4,000 respondants were possible to that question. That's sufficient to get wrong without having to infer the band wagon effect; and the band wagon effect can apply in these circumstances as well (it does not need to be enlisted wholesale).

Mike

[TruthIsAll]

76. You have it all wrong. Learn how to read.

You say:
36.8-42.8% of those who voted for Gore participated in the 2004 election;

37.25%-43.25% of those who voted for Bush participated in the 2004 election.

I say:
You miss the point:

Regardless of what the poll says, I KNOW HOW TO CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS WHO COULD HAVE VOTED IN 2004.

DO YOU?
APPARENTLY NOT.

YOU DON'T KNOW HOW TO DO SIMPLE ARITHMETIC:
48.7/122.3 = 39.82%

I LOVE THIS.
YOU AND YOUR TWO BUDDIES HAVE JUST FAILED 3RD GRADE MATH.

AND YOU CALL YOURSELVES ANALYSTS?

[OnTheOtherHand]

78. second verse...

I think Mike may be highballing the MoE a little, but he is getting it straight from the source. Consider also

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

The MoE is higher than the nominal MoE because the sample is not a simple random sample. What is your best estimate of the design effect, TIA?

But more importantly, why are you assuming that people answer the "2000 vote" question accurately? or if you aren't assuming that, what in heck is your argument, anyway?

[autorank]

80. Isn't he talking about the actual vote? What matter is a question?

I'm jumping in here but that seems to be what's being said.

On edit: This should help. (from above)

BUSH GOT 50.456 MM VOTES IN 2000.
APPROXIMATELY 48.7 MM ARE STILL ALIVE.
NOW CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS.

AND WHILE YOU ARE AT IT, DO THE SAME FOR GORE 2000 VOTERS.
GORE GOT 50.999 MM VOTES
ABOUT 49.21MM ARE STILL ALIVE.

These are the actual votes, no questions necessary to obtain this informaiton.

[OnTheOtherHand]

87. OK, I think we covered that back around message #7... n/t

[mgr]

88. Problem was I made an error that did not change <> my conclusions.

What I did was bracket the MoE around the actual percent of expected Gore and Bush 2000 voters, not around the estimate. Since they still fell within the range of the 3% MOE cite<d> by NEP, the argument is still valid. The thing is that regardless of how I applied the %, in any situation, the same might occur, within the range or outside the range.

I think what TIA is trying to explain is that the optimal estimate of returning republican voters is very close to the lower bounds of the error term, and if we were to apply a more accurate estimate, the percentage would surpass the bounds.

The problem with this assumption is that one presumes equal optimal mortality rates for members of both parties; and one hundred percent fidelity between 2000 and 2004 as to who you voted for. Quite frankly, the republican party is primarily younger and whiter than the democratic party (African American and Hispanic minorities have lower life expectancies), so the mortality for members of the democratic party would be greater than the average over any period of time than republicans; and the republicans would be lesser. This differential could also explain the discrepancy (lets' call it MLDD--the more likely dead democrat; or OWDSTYFWM--Old women die sooner that younger fat white males).

Mike
<edited to delete 'the' between 'change' and 'my', corrected tense>

[TruthIsAll]

114. Here is my best estimate...

Say hello to Mitofsky:
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=357345

You:
But more importantly, why are you assuming that people answer the "2000 vote" question accurately? or if you aren't assuming that, what in heck is your argument, anyway?

Me:
Once more, GREAT OBFUSCATOR, I HAVE NEVER STATED OR ASSUMED THAT.

BUT I DO ASSUME THAT
BUSH 2000 VOTERS STILL ALIVE/TOTAL 2004 votes
= 48.7/122.3 = 39.82%

I assume that because that's what I learned in 3rd grade.
Therefore 39.8% is correct. It is a MAXIMUM.
Therefore 43% is incorrect. IT IS BOGUS.

So the Final 13660 National Exit Poll is therefore incorrect, since it rests on a bogus foundation.

Therefore Bush did not win ONE GOD DAMN EXIT POLL, NOT EVEN THE FINAL, WHICH THE MSM AND YOU BELIEVE TO BE TRUE.

That is my argument, (O)h (T)imeless (O)bfuscating (H)andyman

[Internut]

115. You're fond of Yes/No questions - here is one

"If a poll result shows that a lot more people voted in 2000 for Bush than for Gore, does that mean the poll is "incorrect", and "rests on a foundation of quicksand"? Note that the question is about the poll. Not about that particular question on the poll, but about the whole poll.".

Will TIA answer this YES or NO (like he demands of others) I wonder?

[autorank]

117. If pigs can fly, will you take one to Europe?

Do you have a particular poll in mind, perhaps the Retrpspective Self-Serving Republican Poor Excuse for a Poll Poll?

All this thread was about was a math problem. You got it right.

Move On

[Internut]

122. I asked TIA, not you,

I gave up on getting anything remotely relevant to the discussion from you a while ago.

So, TIA - what's good for the goose should be good for the gander, no? You got your YES/NO answer - now will you give one yourself?

[autorank]

133. I'm so sorry to hear that.

Today's lesson

[TruthIsAll]

126. YOU ARE CONFUSING APPLES AND ORANGES.

"If a poll result shows that a lot more people voted in 2000 for Bush than for Gore, does that mean the poll is "incorrect", and "rests on a foundation of quicksand"? Note that the question is about the poll. Not about that particular question on the poll, but about the whole poll.".

Will TIA answer this YES or NO (like he demands of others) I wonder?

YOU ARE CONFUSING APPLES AND ORANGES.

If I take a private poll of 10 people and ask who they voted for in 2000 and 6 say Bush, 4 say Gore, what does that prove?
Not a damn thing. Just that 6 of ten said they voted for Bush

But on the other hand, since I KNOW that Gore won a majority of the vote in 2000, would I be correct in applying a 60/40 weight to determine what the vote was for Bush and Kerry in 2004?

In other words, would I be correct in EXTRAPOLATING AN OBVIOUSLY INCORRECT RESULT IN A CALCULATION TO DETERMINE HOW THEY VOTED IN 2004? WHY NOT USE THE BEST DATA AVAILABLE - THE ACTUAL GORE/BUSH WEIGHTINGS WHICH WE KNOW TO BE FACTUAL?

AFTER ALL, HOW ARE WEIGHTINGS DETERMINED IN THE FIRST PLACE?
FROM HISTORICAL DATA TRENDS. WHAT CAN BE MORE HISTORIC AND VERIFIABLE THEN THE ACTUAL NUMBERS THEMSELVES.

THIS IS NOT ABOUT WHO THEY VOTED FOR IN 2004. ITS ABOUT WHO THEY VOTED FOR IN 2000.

SINCE WHEN IS IT PREFERABLE TO USE BAD NUMBERS WHEN THE REAL NUMBERS ARE RIGHT IN FRONT OF YOU.

AND IF 43% SAID THEY WERE BUSH VOTERS AND ONLY 37% WERE GORE VOTERS, WHAT DOES THAT TELL YOU ABOUT RBR?

IN ANY CASE, HOW DO YOU JUSTIFY USING AN IMPOSSIBLE WEIGHTING TO CALCULATE THE NATIONAL VOTE PERCENTAGE?

OK, NOW TELL ME THAT IT WAS DUE TO GORE VOTERS THINKING THEY REALLY VOTED FOR BUSH.

DO YOU SEE HOW RIDICULOUS THE ARGUMENT GETS?

GORE VOTERS WITH LAPSED MEMORIES....
EXUBERANT BUSH VOTERS..BUT BUT WHAT DOES THAT DO TO RBR?
MATHEMATICAL IMPOSSIBILITY - ONLY 50.45 MM VOTED FOR BUSH IN 2000,

TO BELIEVE THE FINAL EXIT POLL, YOU MUST BELIEVE THAT
52.59MM (43% OF 122.3) BUSH 2000 VOTERS CAME OUT TO VOTE....
AND NONE OF THEM DIED..WHEN THE ALMANAC TELLS YOU THAT BUSH GOT ONLY 50.456 MM.

AND ON...
AND ON...
AND ON...

THE MERRY-GO-ROUND OF DISBELIEF

[Name removed]

127. Deleted message

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

[Name removed]

129. Deleted message

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

[Internut]

131. Don't know why the message got deleted -

but I'd just like to point out that while you demand that others answer YES/NO to your questions, you refuse to do so yourself.

[autorank]

132. It's his thread. You'd be a terrible guest for bridge. Hey, bad rule!

I quit. Let's play Texas hold 'em instead. I hate Bridge:rofl:

[autorank]

134. Nominated for LINE OF THE YEAR. "Destroying a Fraud in One Line"

And the winner is:

AND IF 43% SAID THEY WERE BUSH VOTERS AND ONLY 37% WERE GORE VOTERS, WHAT DOES THAT TELL YOU ABOUT RBR?

I am stunned. Anyone who says Reluctant Bush Responder again just needs to get the tee-shirt with this line on it, available soon at Cafe Presse.

This is just wonderful to see.

You can lead a horse to water but you can't make him smart enough to drink it.

[mgr]

145. Let's step away for a moment:

There are two distinct arguments being made here, and swapping arguments back and forth does not help any third party to come to understand what is being addressed. They reflect the different experience of the proponents--my statistical work always involves direct measurement of the phenomena to be addressed; OTOH's ( and Internut?) experience involves indirect measurement, and, should be, when assessing expertise, are the arguments that count when discussing exit polling.

Estimating the demographic pattern from 2000 to 2004 is a back of the envelope calculation (BEC). What I am arguing is that for a BEC, the 43% holds up.

All a BEC is is a consistency check. Its rarely used for refuting an argument, more informed investigation would be necessary. Using the published MoE from NEP, we find that the 43% cames close to, or falls within the error range of the 40%. In the nature of a BEC, one usually uses approximations (although OTOH provided an more precise MoE of 2.6% from a table). This is without taking into account the election results for the final weighting for D and R support.

With the 13660, the election results are accounted for, the D and R support weightings are changed (although they are within the range of the older weightings), and the value still remains within the range. This suggests that the weightings are apt.

If you consider this from an empirical standpoint, what else could NEP do? The precinct level sampling has no indication of movement toward either candidate due to noise, and when the election results come in, they are at or just beyond the upper ends of the range using the old weights, so the weights are adjusted. One has to remember that all the exit poll is a better estimate than a BEC. In the next election, it may be necessary to adjust the weights downward.

Now if one considers the information potential of the who did you vote for in 2000, is it likely or not that adjustments to the weightings are to include adjustments for this question? I think not, since each election is an independent event, and this question is contingent on the past election.

What OTOH and Internuts seem to be arguing is that whether the BEC would or would not be consistent, there is no reason to place so much weight on the who voted for who question because the respondents are unreliable. That is a question where I have no expertise.

Part of the problem I think is how one visualizes the population being estimated by the exit poll (the election). TIA appears to see it as a textbook Gaussian curve, with little variability between the three measures of central tendency (mean, median, mode); and also views the exit poll sampling similarly. I visualize the election population as being distorted, with greater numbers to wards the tails, and the center of the distribution flattened with slightly greater spread between the measures of central tendency; I see the exit polling sample population as having measures of central tendency almost exaggerated in their separation, producing a highly skewed distribution, producing lopsided MoEs (e.g. the error range is better characterized by a log transformation), with measures of central tendency offset from the mean of the population (the election results) by 6.5% (The WPE).

In TIA's visualization, a small MoE of 1% would be valid. It does not match reality. My visualization would produce a far greater MoE.

If TIA, as he argues, does not acknowledge that he has visualized the problem, but is only applying the math, then he probably has nothing to say about the problem--since he is running calculations without context.

Mike

[mgr]

83. Slight correction, same outcome

37<39.8<40; 40<40.25<46

Sorry, does not wash. Too busy here celebrating the report.

Mike

[kster]

16. kick.nt

[byronius]

44. Autorank, TruthIsAll -- I want you to know something.

I read most of these threads. I read, but do not understand fully, most of TIA's math. Autorank's voice has become second nature to me -- I can almost see who you both are; perhaps the shape of your soul more than your features. Why do I trust the two of you to know what you're talking about? Why do I consider it a certainty that election fraud was committed, and that it was enough to change the outcome, even if I do not fully understand every detail of these threads?

A lifetime of living with 'them', and around 'them'. A lifetime of reading current and past history, especially political history, and slowly coming to understand what it is they do, and how they do it, and why. It's not at all required to contrive a tangled theory that weaves its way across a contrary landscape to support TIA's contentions -- it flows down reality river pretty smoothly. Massive election fraud is a mild expression of the acts they would commit if they had the courage and organizational focus of some of the proto-versions of their belief system. I think our opposing forces, and these basic dramas, go back to the dawn of human society, and simply-drawn pencil lines will continue to smoothly forecast much of what we're seeing.

I went to the raucous meeting at the CA Secretary of State's office recently. Here is my post describing it:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=378552&mesg_id=378635

Until then, I had not seen it face-to-face. What I saw was both heartening, and confirmed my worst fears. The demeanor of the Diebold representatives, who fled from public comment, and the continuing attempts by the CA GOP to shut down any and all activities (especially discussion) that do not serve their purposes, and to demonize anyone who disagrees with them, just as Karl Rove does on a national level -- all this was telling. Even more telling was the fact that, after Shelley was hounded from office by an extremely well-coordinated and well-financed 'movement' (that taped large full-color signs to every single light pole on Broadway saying 'Shelley Must Go', all at heights requiring a ladder to remove), the very highest priority that his new Republican replacement has is to immediately (<30 Days) attempt to certify both Diebold and ES&S machines for use all throughout California. These machines have a 'paper trail' of sorts, but it's printed on thermal paper that begins to degrade immediately, and becomes unreadable at temperatures over 90 degrees. These machines also contain wireless devices intended to be used for both software upload and central tabulation, in addition to rewritable SmartCards that are physically transported to the central tabulators.

These machines could not have been better designed for the purpose of committing wholesale electronic election fraud. The Assistant Secretary of State, William Wood, made it clear in his opening statements that he considered the immediate, pre-HABA deadline installation of these machines to be top priority. After the hearing proved to be a wonderful display of common sense from all walks of the public that flew directly in the face of Wood's smug intentions, the panel chose not to recommend on the issue -- a first-time-in-history event in California. However, I have no doubt at all that they will simply regroup, and try to find a way to avoid public scrutiny of their further actions. It Is What They Do.

I consider both of you to be heroes. I have written a new serial comedy that references election fraud in a more serious moment where characters discuss current US culture, and references both of your DU user names. I talk about these threads constantly. I am about to embark on a period of increased activism and writing about these issues -- Peace Patriot has provided a list of all CA Representatives, and I'm going to start working my way through them, first with letters and calls, and then with personal meetings if possible. I intend to become much more deeply involved with the CA Democratic Party and its processes.

Autorank -- I would frankly love a short Primer of your opinions on these matters. Where should I put my energy? Can you sketch a quick outline of your ideas for action? I've seen bits and pieces from you on this, but given more detail of where I am and what I want to do -- what should I do?

Honor to both of you. Thank you for your imperturbability, and your willingness to get into the nitty-gritty details of the arguments presented by others. The dialogue always reveals the truth, frankly.
Let's Go!

:bounce:

[autorank]

50. KICK for the TRUTH

:kick:

[David Dunham]

79. You forget that bunches of Gore 2000 voters went for Bush in 2004

You miss the fact that because of the national security issue, and because of the denunciation of Kerry by the Catholic clergy, bunches of Gore 2000 voters went for Bush in 2004.

[OnTheOtherHand]

82. actually, that is in there

The exit poll gives estimates of Gore voters for Bush, and Bush '00 voters for Kerry -- quite a few of each.

TIA's argument would be pretty good, if not for the facts that (1) other surveys show that people don't answer the past-vote question accurately, and (2) he seems to be underestimating the margin of error. (Forgive me for not explaining his argument; that would be painful for me right now. Sniff.)

[autorank]

84. You're making it too hard. It's a simple question.

From the initial post:

CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS OF THE 122.3 MILLION WHO VOTED IN 2004.

HINT:
BUSH GOT 50.456 MM VOTES IN 2000.
APPROXIMATELY 48.7 MM ARE STILL ALIVE.
NOW CALCULATE THE MAXIMUM PERCENTAGE OF BUSH 2000 VOTERS.

AND WHILE YOU ARE AT IT, DO THE SAME FOR GORE 2000 VOTERS.
GORE GOT 50.999 MM VOTES
ABOUT 49.21MM ARE STILL ALIVE.


THAT IS THE TOTAL MATH EXAM.

It's not about who defected or even trying to figure that out. It's aobut Bush 2000 voters who survived and were alive in 2004, thus the "maximum percentage of Bush 2000 voters" who could have voted in 2004. The next step might interest you.

[FogerRox]

86. yeah right. ---- Pfffft

Republicans for Kerry .com

[autorank]

103. Summary: A bunch of "nay-sayer" bullshit because they won't answer the ?

Amazing. I mean if you don't like the question or you can't answer it, fine but why all the blablabla over and over and over as a means of avoiding the question. It was a simple math exam.

Your grade: F

Your dismissed! :hi:

[Bouvet_Island]

135. Whatever the premise,

I notice and I mean I read your threads since before the election, I noticed that you only calculate STANDARD MARGIN OF ERROR.

Which, as I believe your stack of statistics books will tell yah, it only regards the kinda error you get under straight flush perfect conditions.

Like, it does only say something about the error you get because the poll wasn´t large enough. Like if you draw a map of the world, you make it smaller than the earth itself, your plastic globus perhaps isn´t precise enough to plan your sightseeing inside New York and your tourist map of New York might not have that tiny side street on it? Please clarify, this is the problem you are talking about?

It actually doesn´t say anything about any other possible human or machine error, no matter the cause or intent? So, to give a direct answer to your question, if the world would adhere to really really simple rules like in Microsoft Minesweeper tm or Ludo, your logic would be a full 100% correct. I mean given that noone used the official cheat to Minesweeper tm?

It just seems to me that anything beyond that simply aren´t analysized in your analisysus or calculacated in your calculacations? So, clearly, Deep blue would simply have to agree, given that the pawn have no mind of it own to tell it to move sideways facing the queen, it´s check mate?

I mean if there are or were no sources of error in any science, at any time through history, beyond sampling error, I mean then your logic is as hard as the diamond and to be frank as flat as flat earth.

And if else you are simply not that sure about your point as you say, I mean since if there were other things to consider, I mean then you would say so? That the margin of error could just as well be 20% as .3% for such a complex material as the US vote count, depending on HUMAN FACTORS. That if say Mr MITOFSKY is a raving loon or as previously discussed here a Bush Croonie, that your point here could go both ways? That if the leaked polls were planned, that it would mean that nothing of what you say would be correct? How on earth can you be sure the Leaked polls wasn´t a trick, that it was intended as distraction? What´s your material?

Now to "change the subject" or "spin" or whatever you might like to call it. I have a question, I read a fair amount of your stuff and in particular I just read this whole thread, I believe I have the right.

1. What on earth gives You the right to act errh RUDE (or substitute) with anyone of different opinion on the particular question of Exit Polling Methodology and Results, given that they still count themselves at our side, or at least lending us their ear? I mean about every single thread you post here is like watching the Hindenburg land, to quote your own title ""EXIT POLL NAYSAYERS: HERE'S A 3RD GRADE ARITHMETIC TEST FOR YOU."

In grown up terms you are calling me, being of a different opinion than you on a question so *mhmfffh* marginal given what we really are talking about, US election fraud since the 19th century and in particular the past few elections, and in many cases I mean we are *talking* about it, we are having civilized and open discussions about it, considering both sides of the question, trying to give our own arguments the same critical treatment we give our opponents´. And before I even get to read your *mfmmfm mfm point, You are calling me a Naysayer, sorry, A NAYSAYER, NEGATIVE a description as is.

To illustrate my point I am calling you to your face a YES-SAYER, I sincerely believe you are saying to much "yes" for your points to be scientificly valid. Notice how you sound Positive and I, I the NAY-SAYHYENA, I sound NEGATIVE.

And frankly I suspect the reality is the direct opposite, I just got a job as a cook on an Island in the Mexican Gulf! :] Wohooo! How about that !?!

[TruthIsAll]

136. I'm sorry, I cannot respond. You are incoherent. n/t

/

[Bouvet_Island]

146. You mean you would not like to,

that´s OK, I didn´t really expect Civilized Discussion here after reading your title.

I believe you are not sorry to not give a response here, rather You should be sorry you treated so many of your debate opponents here at DU so disrespectfully, it makes the content of whatever you have to say seem very weak and in need of compensation, it´s not an effective way of building support or weight to a theory.

If you had a scent of scientific spirit or training you would realise that criticism and good, civilized discussion is in fact a good thing, and that you don´t win a science debate or much of anything by SHOUTING YOUR POINTS AT PEOPLE, calling them names or avoiding the real critical questions.

It´d just be great if you rather would direct the Negative energy outwards, in the direction of the enemy, where it can be productive.

[bleever]

147. "Nay"

Nay is a curious word,
From those who could have demurred.
But instead of dismiss you,
They hang out and dis you,
For fear all this talk might be heard.

[kster]

148. kick.nt

[tommcintyre]

160. kick - In tribute to his tremendous contributions - WHY??? :( n/t

[bleever]

161. Kicking for truth, justice, and TIA's invaluable work.