15% OF GORE VOTERS (7.65 MM) SUFFERED FROM ALZHEIMER'S IN 2004

[TruthIsAll]

Concerning the Final National Exit Poll, in which 43% of 2004 voters said they voted for Bush. We have shown 43% is impossible, since 43% of 122.3 = 52.57 million, and Bush only got 50.456mm votes in 2000.

So how do the naysayers now explain this?
Simple.
Gore voters forgot that they voted for Gore when they said they voted for Bush.

Ok, now let's restate the 43/37 problem slightly.

Assume that 40% Bush/40% Gore was the actual response, but 3 out of 40 Gore voters forgot they voted for Gore and said they voted for Bush. That's how 40/40 became 43/37.

But let's look closely at that calculation:

If a Gore voter forgot who he/she voted for, couldn't he/she have been equally likely to say that they voted for Gore as for Bush?

By the law of averages (or the Law of Large Numbers) 50% of forgetful Gore voters would say they voted for Gore and 50% would say they voted for Bush. Just flip a coin.

Therefore, 6 out of the 40 Gore voters must have forgotten who they voted for, with 3 saying they voted for Bush and 3 saying they voted for Gore.

Now 6/40 = 15%. That means that 15% (7.65 MILLION) of 51 million Gore voters forgot who they voted for. Let's assume for the purpose of this discussion, that there were no liars among them, just those who forgot.

Are we really expected to believe that not only did 7.65 MILLION Gore voters suffer from Alzheimer's or Amnesia, but AT THE SAME TIME ALSO BELIEVE THAT NOT A SINGLE BUSH VOTER SUFFERED FROM EITHER MALADY?

[truth_is_extreme]

1. ask Kerry

he's the one who conceded.

[BamaLefty]

30. C'mon now... n/t

[uppityperson]

2. I voted for Kerry, I think. I don't remember voting for Gore. Or Bush.

maybe I did and I too have alzheimers or amnesia? Very funny.

[sweetheart]

3. Obviously, its clinton's fault

Clinton did it... that is why. No further explanations are needed.

Its a great excuse. I've started using it to explain all phenomena
in the universe without apparent purpose... clinton phenomena. All
events in the universe are related to his dick.

[Peace Patriot]

4. Great subject line, TIA! Har-har!

Don't you know, Bush voters get injections of babies' brain cells in Paris when they hop the pond to load up on French wine, cheap prescriptions, good organic food and almost free medical care?

Alzheimer's is for the unwashed.

[autorank]

5. I have the answer...but it's gonna cost you.

OK, now let me compose myself here...this is important. All right, here goes:

Some Gore voters acquired a special neurological condition called "Diebold's Syndrome." Prior to Election 2004, Management beamed a coded signal that was picked up by only Gore voters. The signal inserts a circuit to perform a highly specific default function in only one instance; in the case they're subject to exit poll questions after the Kerry-Bush 2004 Presidential Election. When they're asked, a sufficient proportion (according the algorithm called Mitofsky's fancy function) default automatically to a memory of voting for Bush.

TIA, this explains it.

The larger issue is the residue of Diebold Syndrome in 7.5 million Americans, good Democrats. They can only be relieved of their agony by answering the question, "who did you vote for in 2000?"

Please ask all your friends this question. It may be their only hope.

For those of you who question my logic on this, please see my source. He's impeccable.

[Chi]

6. It's not like the election went into quadruple overtime, or anything.

It was so unmemorable.
And it's not like the electorate was polarized.
There must of been what....20-25% undecideds right up till the election.

/Sarcasm

[Melissa G]

7. it's just the pesky 11 million come back to haunt us another way...

That's where we started. Well, Thanks to TIA, we have learned part of the answer is Zombies.. The risen dead who came back from their graves to vote for the shrub... Another part must be the living dead( was it a full moon?) because Shrub had to retain every single living and dead 2000 voter before he added all that incomprehensible new influx of people. So many of the Texas 'pubs I know plus lots of the fair weather Dems who thought it would be good for Texas to vote for Blivet in 2000 told me unequivocally that even though they had been fooled once no way in hell they were doing it twice. Well, obviously they turned into ballot vampires or werewolves or some other living dead creature who had no real will to choose when the see they voting booth..Quick, go get the garlic!!!

Maybe the the new influx of Blivet voters were a science fiction import from the planet DB-ESS...Naturalized aliens used by Neocons to install the puppet! When that leaks it could inflame the far right.... Naturalizing aliens always upsets them!

[St. Jarvitude]

8. Explain something to me...

I've been gone a while.

Concerning the Final National Exit Poll, in which 43% of 2004 voters said they voted for Bush. We have shown 43% is impossible, since 43% of 122.3 = 52.57 million, and Bush only got 50.456mm votes in 2000.

Why is this so ridiculous? I believe the total national voter turnout in 2004 was considerably higher than in 2000.

Just checked - yes, 105,406,100 in 2000 vs. 122,293,332 in 2004- a 16% increase .

[TruthIsAll]

9. LET'S MAKE IT ABSOLUTELY CLEAR

YOU
Concerning the Final National Exit Poll, in which 43% of 2004 voters said they voted for Bush. We have shown 43% is impossible, since 43% of 122.3 = 52.57 million, and Bush only got 50.456mm votes in 2000.

Why is this so ridiculous? I believe the total national voter turnout in 2004 was considerably higher than in 2000.

ME
You are missing the point.
Its very simple.
But maybe it's my fault for not being absolutely clear.

So let's make it so:
change it to read "43% of 2004 voters said they voted for Bush IN 2000.

"... 43% of 2004 voters said they voted for Bush IN 2000. We have shown 43% is impossible, since 43% of 122.3 = 52.57 million, and Bush only got 50.456mm votes in 2000".

[Hippo_Tron]

15. 43% is a sample, not an exact figure

Usually these samples have like a 4% margin of error.

[TruthIsAll]

17. Bush 2000/2004: 50.456 -1.77 = 48.7/122.3 = 39.8% has ZERO MoE

You miss the essence of the post.

The Sample size was 3168 for this question.
The MoE is 1.77%
MoE= 1/sqrt(3168)= 1.77%, not 4%.

But the MoE and sample size are irrelevant for this demographic.
We KNOW the proper weight for Bush 2000 can be no more than 39.8%.

Just focus on the facts:
Given that 48.7 mm former Bush 2000 were alive in Nov. 2004, what is the maximum percentage Bush 2000 voters could be of the total 122.3 million who voted in 2004?

That's it.

[mgr]

18. I believe that On The Other Hand pointed out your mistake

with addressing MoE when the variance is greater, and/or for clustering/treatment effects. Ignoring when you have been corrected by someone who may constitute an expert, without a good argument why you don't think you should follow the same direction that Simon accepted, is, how shall we say, problematic.


I believe that the MoE he determined for response to all ~13,000 questions, which I think is what the FAQs refer to having the 3.0 MoE, was 2.6. Since this was only a question responded to by ~3,000 respondants, the MoE may in fact be higher, to where the range indicated by the MoE can include plausible estimates of voter fidelity.

The 4.0 MoE referred to in the previous post is associated with the State level exit polls.

www.exit-poll.net I recommend that all review this web site thoroughly before continuing this discussion, because I believe all are arguing from willful ignorance or what is commonly referred to as faith.

It also took a great deal of mathematical work to demonstrate Bishop Usher's assertion that the world was created at 10 am 2,000BC. The math was right, its just the assumptions that went into the calculation were wrong.

Mike

[TruthIsAll]

19. In every one of your myriad posts, you naysay as if on cue..

You have never ONCE accepted a single fact or analysis of mine as being even remotely correct.

Why should DU expect anything different from you?

Well, mgr, the numbers speak for themselves.
Who do you speak for?

You are nothing more than a broken record.
Which never stops spinning.

Keep it coming.
We all enjoy a good laugh around here.

[mgr]

22. I did not accept Bishop Usher's explanation either.

So you cannot speak to the point I make. Speak to the issue, convince me. Explain why you reject an expert's opinion? And don't start a new thread in hopes some will forget.

As to on cue, let's see this thread started on Monday, I responded Wednesday, yep right on cue. You seem to forget that this thread follows an unresolved argument from a previous thread, that followed the same unresolved argument in a previous thread, from an argument you lost in a previous thread. I seem to have been in on that original argument, so it would stand to reason that I would persist.

Mike

[TruthIsAll]

24. Take the mgr challenge which autorank gave you..

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=381311

[mgr]

25. See response 20 below n/t

[mgr]

26. Poor deflection.

Hi Bishop Ussher:

In other words, you cannot defend that decision? People, myself included, have been trying to hammer home to you that your MoE is incorrect, and present<s> to<o> narrow a range. Several method for appropriate calculation have been provided, but you reject these without argument.

Why is it that Simon, the original professor of the red shift argument accepts this modification to MoE, but you do not? I mean all that you did was take Simon's work with the battleground states, and apply it wholesale to all the states. So what makes it that you differ with him as to how MoE is to be applied? This question has been put to you since December, and you have yet to answer it in any manner other than ad hominen arguments.

Mike

<edited to correct spelling and tense,--lazy fingers>

[TruthIsAll]

27. MoE = 1/sqrt(N)

MoE = 1/sqrt(N)
N MoE
500 4.47%
1000 3.16%
1500 2.58%
2000 2.24%
2500 2.00%
3000 1.83%
3500 1.69%
4000 1.58%
4500 1.49%
5000 1.41%
5500 1.35%
6000 1.29%
6500 1.24%
7000 1.20%
7500 1.15%
8000 1.12%
8500 1.08%
9000 1.05%
9500 1.03%
10000 1.00%
10500 0.98%
11000 0.95%
11500 0.93%
12000 0.91%
12500 0.89%
13000 0.88%
13500 0.86%
14000 0.85%
14500 0.83%
15000 0.82%
15500 0.80%
16000 0.79%
16500 0.78%
17000 0.77%
17500 0.76%
18000 0.75%
18500 0.74%
19000 0.73%
19500 0.72%
20000 0.71%
20500 0.70%
21000 0.69%
21500 0.68%
22000 0.67%
22500 0.67%
23000 0.66%
23500 0.65%
24000 0.65%
24500 0.64%
25000 0.63%
25500 0.63%
26000 0.62%
26500 0.61%
27000 0.61%
27500 0.60%
28000 0.60%
28500 0.59%
29000 0.59%
29500 0.58%
30000 0.58%

[mgr]

28. So you disagree with Simon.

So clustering does not affect the measure of the range in variability? Lets throw out all those text books on indirect sampling and polling, Bishop Ussher, the software engineer and amateur statistician, has shown that they all have it wrong. That is is where your argument leads. So much for USCV, so much for Fischer, so much for Gauss.

Mike

[TruthIsAll]

31. Two facts and a challenge for you, mgr

Fact 1.
Exit Polls have historically been more accurate than pre-election polls.

Fact 2.
Pre-election polls by national and state pollsters use the formula
MoE = 1/sqrt(n) when they provide the margin of error for their polls.

I challenge YOU, mgr, to find a SINGLE POLL OR SURVEY which deviates from the formula provided for MoE by more than 0.1%.

Conclusion:
Due to Facts 1 and 2, the MoE formula is therefore a CONSERVATIVE measure for measuring Exit Polling accuracy.

Q.E.D.

Your unrelenting design to disrupt has had no effect other than that of reducing the credibility of your arguments to the vanishing point.

Did you ever come up with that analysis you promised (threatened) us with months ago in which you would attempt to reconstruct the basic foundations of statistics; namely, the Law of Large Numbers?

Have you EVER, EVER contributed a DAMN thing to DU other than to revel in your role as Chief Naysayer and TIA disruptor?

Have you ever shown us ONE piece of statistical analysis using available pre-election and exit poll data, state or national to advance or dispute the likelihood of fraud?

Have you ever?

If you have, lets see a thread of links to YOUR analysis. In lieu of this, since none exist, would you create a thread of your replies to my posts?

Let's look closely at your historical record of diversion, straw men, and nay-saying. If you won't do it, I will. Let DUers see your track record of manufactured fog.

You cannot hide from it.

[LightningFlash]

33. MGR, we're waiting for your answer and your evidence.

Do not move away from the argument and run away.

You said you'd take the challenge, and provide the evidence.

Provide it now which would prove your theory correct. We're still waiting for you to provide that.

What is going on, MGR? It's been several hours.

Oh what's that? Can't prove the 2004 census results didn't happen.....Which show more africanamerican/hispanics voting than who were counted....

[autorank]

10. KICK -- for some attention, this is hot!!!

[mgr]

11. Once again you misrepresent the argument

The issue is not switching from Gore 2000 to Bush 2004; but those that may have (not) participated in 2000 claiming that they did. What you do not address is that the fidelity of Gore 2000 to Kerry 2004 is also very high. So where did these additional voters come from?

Or is it that the number when you consider the MoE is not so impossible when you reduce 43% to 40.3%?

Mike

edited to include negative.

[TruthIsAll]

12. Of course, mgr, I knew you would show up sooner or later.

"The issue is not switching from Gore 2000 to Bush 2004; but those that may have (not) participated in 2000 claiming that they did. What you do not address is that the fidelity of Gore 2000 to Kerry 2004 is also very high. So where did these additional voters come from?"

Are you saying that New voters lied?

What about the fact that New voters went to Kerry by 54-45 at minimum (bogus final exit poll 13660 respondents) to 57-41% (13047) to 59-39% (11027)

According to you, NEW voters were liars, Gore voters suffer from Alzheimer's or were liars, and Bush 2000 voters were absolutely truthful.

Ok, I see your point. Makes a lot of sense.

If you live in a Bizarro world.

[mgr]

13. You are confused but its a nice straw man

First off, new voters were not measured by this question, it is from a different questionaire.

Second, I've already pointed out that there are three pools of possible voters to consider--those that voted in 2000, the newly registered (new), and those that are registered, but did not vote in 2000 (the casual voter).

So which of these is likely to be the least honest? Why would a new voter lie, since it is their first election? Not very likely. Those that voted in 2000? Well if they did, it was a very small percentage, since so many already indicated what their 2004 choices were. So that would leave the casual voter.

The thing that you are dancing around is my point about the MoE which is my own argument, not OTOH's or Internut's or Foo's. I have not come out in support of the explanatory power of the band wagon effect, because it does not need to apply here. If you had a very low G2000 to K2004; and the numbers for B2000 to B2004 were very high, then I would say it is suspicious, and examining the band wagon effect might be in order. But as you are aware, the numbers to K2004 and B2004 are too high, and suggest fidelity.

If you can show that 40.3% of those voting for Bush in 2000 going to Bush in 2004 is impossible, using actuary data, then I will likely agree. Just remember, that the demographics of the republican party membership suggests that their overall death rate to be <delete higher, insert greater> than the national average, so you might want to adjust it for your estimate.

I agree, it is bizarro to consider only two categories when three exist. It also is bizarro to misrepresent an argument that you know is a misrepresentation (sneering optional) in a forum where honesty and truth is prized.

Mike

<edited, words changed for greater clarity>

[Hippo_Tron]

21. Does "new voter" mean didn't vote in 2000, or never voted at all?

[mgr]

23. New voter should mean never voted at all. n/t

[TruthIsAll]

35. 17% of 2004 voters (21mm) did not vote in 2000; 11% were New.

17% of 2004 voters (21mm) did not vote in 2000.

Of the 17%, Kerry won 54-59%, depending on the exit poll
timeline:
11% were first time voters. Kerry won 53-55% of them.
6% voted prior to 2000. Kerry won about 60% of them.



Voted in 2000 for:
	
     Mix	Bush	Kerry	Nader

13047	respondents 12:22am		
NONE	17%	41%	57%	2%
GORE	39%	8%	91%	1%
BUSH	41%	90%	10%	0%
Nader	3%	13%	65%	12%
				
13660	respondents 1:25pm		
NONE	17%	45%	54%	1%
GORE	37%	10%	90%	0%
BUSH	43%	91%	9%	0%
Nader 3%	21%	71%	3%
				
				
FIRST TIME VOTED in 2004				
13047	respondents 12:22am		
Yes	11%	43%	55%	1%
No	89%	48%	50%	1%
				
13660	respondents 1:25pm		
Yes	11%	46%	53%	1%
No	89%	51%	48%	1%

[autorank]

16. "The mgr challenge..."

Why not put up here...or, well, you know...

[mgr]

20. Who are you to set deadlines?

Why is it now that we need to close ranks, and agree; and if we don't we should shut up.

The position presented by TIA is disputed by the DNC. The DNC report is closer to what I understand to have happened in this election than wholesale theft of the popular vote. Voter suppression cannot be measured by an exit poll, since suppressed voters never make it to where they show up at the voting booth. I think it may have been orchestrated, and has been expanded to address both African Americans and Jews, from just African Americans.

I have reviewed both exit polls' estimates of the Jewish vote, compared it to congressional districts I am familiar with in Los Angeles with large jewish populations to see how democrats fared, and suspect that the estimate of 80% participation for Kerry the correct estimate given in the Los Angeles Times exit poll; and the lower ~70 percent estimate given by NEP (with only three respondants) incorrect.

Palm Beach and Volusia Counties in Florida have large Jewish populations, and if you review the long lines/machine problems record in the database, you should note a large number of problems associated with the Jewish voting places. When you look at the voting places, and look for synagogues, in each that I have examined, synagogues are close by, if not the voting place. And if you are ignorant of jewish religious practice, the admonition of no work on the Sabbath, requires one to travel on foot to the place of worship (with the exception of reformed jews, who may follow this practice but are not compelled to). Therefore one would expect jews to be far closer to their place of worship than gentiles.


I also suppose you are unfamiliar with Bishop Usher who calculated a specific date and time for the biblical creation. I am sure that he, like TIA, found his math compelling. Refuting him would not require being familiar with the family trees in Numbers would it? So just what am I supposed to put up if I cannot argue that TIA's presumptions are wrong, and do not conform to the standard practices performed by those familiar with experimental design and statistics?

Or are you familiar with the mathematician who sent Bert(r)and Russell an elegant mathematical proof of God's existence. It is said Russell refuted him on a post card. It had something to do with an assumption confusing the value 0 and the null set. I have shown that TIA's assumptions are wanting in the necessary independence to allow determinations regarding MoE; and that when the appropriate MoE is considered, the margin of victory in this election is within the range of the NEP estimate.

Mike

<edited to correct spelling of Lord Russell's name>

[mgr]

29. Back of envelop calculation to show significance

The estimated population of Jews in the US is six million. Lets assume a threeway split: 1/3 Los Angeles, 1/3 New York, 1/3 Florida.

Lets assume the 1/3 Florida is in Palm Beach and Volusia Counties: so we have 2 million jews residing in the voting population.

So, according to LA Times, 1.6 million votes for democrats can potentially come from the Jewish vote. The turnout in Los Angeles was in the 75 to 80%, so 0.8* 1,600,000= 1,200,000 to Kerry; 320,000 to Bush. Say machine problems/longlines limits the turnout to 65%, that means a loss of 15% to both candidates: so Kerry's margin is reduced by 180,000 votes.

Mike

[LightningFlash]

34. That isn't evidence, MGR.

That's strawman theories.

This is direct undisputed evidence:

http://www.census.gov/population/www/socdemo/voting/cps2004.html
http://www.votersunite.org/info/content/newmessup-17.asp

Provide yours, now.

[autorank]

32. So we'll take that as a no.

[autorank]

36. I'm familiar with Joey Bishop and the rap star Usher.

Are you casting yourself as a latter day Bertrand Russell, lordy.

Relax...

[TruthIsAll]

37. Here is your MOE and a probability calc

GENDER						
13047 respondents, 12:22am	

     MIX 	Bush	Kerry	Nader		Bush	Kerry Nader	
Male 	46%	52%	47%	1%		23.9%	21.6%	0.5%	
Female54%	45%	54%	1%		24.3%	29.2%	0.5%	
	100%					48.22% 50.78% 1.00%	

The probability of Bush gaining from 48.22% in the exit poll
to 50.73% in the vote (using a 1% MOE):

Prob = 4.3424E-07	

=NORMDIST(0.4822,0.5073,0.01/1.96,TRUE)	

= 1 in 2,302,876


My reply to this thread (post #38) shows you that Mitofsky
calculates 1% as the MoE for  8000+ respondents...

Once again, mgr, your fog is cleared by the facts.

[TruthIsAll]

38. And here is what Mitofsky has to say...about MoE for 8000 respondents

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

[kster]

14. kick.nt

[tommcintyre]

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

[bleever]

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