FOUR SCENARIOS: WILL THE MSM REPORT THIS?

[TruthIsAll]

Don't hold your breath.

Kerry wins the only plausible scenario - and even two of the
three implausible ones which all favor Bush.

IMPLAUSIBLE SCENARIO 1:
Use the Final Exit Poll (13660) with PLAUSIBLE weights
(assuming only those Bush 2000 voters who were still alive
turned out to vote in 2004).

Also assume that an IMPLAUSIBLE 16% of former Gore 2000 voters
crossed over to vote for Bush. That's 8.16 million Gore
democrats voting for the guy who stole the election in 2000.
Not likely.

Bush wins by: 2.28 million votes.

Voted	
2000		         Mix	Bush	Kerry	Nader
22.386	None		18.31%	45%	54%	1%
48.598	Gore		39.75%	16%	84%	0%
48.073	Bush		39.32%	90%	9%	1%
3.203	Other		2.62%	21%	71%	8%

	TOTAL		100.00%	50.54%	48.68%	0.79%
122.26	Votes		122.26	61.79	59.51	0.96

......................

PLAUSIBLE SCENARIO 2. 
Assume that 10% of Gore voters cross over to Bush, plausible,
but not too likely.

Kerry	wins by 3.56 million.			

Voted	
2000		         Mix	Bush	Kerry	Nader
22.386	None		18.31%	45%	54%	1%
48.598	Gore		39.75%	10%	90%	0%
48.073	Bush		39.32%	90%	9%	1%
3.203	Other		2.62%	21%	71%	8%

	TOTAL		100.00%	48.15%	51.06%	0.79%
122.26	Votes		122.26	58.87	62.43	0.96

.......................................

IMPLAUSIBLE SCENARIO 3: 
Same as above, but now assume an IMPLAUSIBLE 13% of Gore
voters voted for Bush (that's 6.63 million Democrats crossing
over to vote for the guy who stole the 2000 election from
Gore).

Kerry	wins by 640,000 votes.			

Voted	2000		Mix	Bush	Kerry	Nader
22.386	None		18.31%	45%	54%	1%
48.598	Gore		39.75%	13%	87%	0%
48.073	Bush		39.32%	90%	9%	1%
3.203	Other		2.62%	21%	71%	8%

	TOTAL		100.00%	49.35%	49.87%	0.79%
122.26	Votes		122.26	60.33	60.97	0.96
.....................................................

IMPLAUSIBLE SCENARIO 4:
Let's make it an IMPLAUSIBLE 14% of Gore voters for Bush, but
now assume that Kerry wins new voters by 57-41%, as per the
Preliminary Exit Poll (13047):

Even with 7.14 million former Gore voters crossing over to
Bush, Kerry still wins by 1.24 million votes.

Voted	2000		Mix	Bush	Kerry	Nader
22.386	None		18.31%	41%	57%	2%
48.598	Gore		39.75%	14%	86%	0%
48.073	Bush		39.32%	90%	9%	1%
3.203	Other		2.62%	21%	71%	8%
	TOTAL		100.00%	49.01%	50.02%	0.97%

122.26	Votes		122.26	59.92	61.16	1.18
			



 
 

[kiwi_expat]

1. Please run this scenario:

11.5% Gore voters and 94% Bush voters voting for the "War President".

Thanks.


(I posted this on your "Seatbelts" thread.)

[TruthIsAll]

2. Done

Kerry loses by 180,000 votes...
IF WE ASSUME HE WINS 57-41% OF NEW VOTERS 
(PRELIMINARY, PRISTINE 13047) EXIT POLL

Voted	
2000		         Mix	Bush	Kerry	Nader
22.386	None		18.31%	41%	57%	2%
48.598	Gore		39.75%	11.5%	88.5%	0%
48.073	Bush		39.32%	94%	5%	1%
3.203	Other		2.62%	21%	71%	8%

	TOTAL		100.00%	49.59%	49.44%	0.97%
122.26	Votes		122.26	60.63	60.45	1.18


Kerry loses by 1.75 MILLION...
IF WE ASSUME HE WINS 54-45% OF NEW VOTERS 
(FINAL, SOILED 13660 EXIT POLL)

Voted	2000		Mix	Bush	Kerry	Nader
22.386	None		18.31%	45%	54%	1%
48.598	Gore		39.75%	11.5%	88.5%	0%
48.073	Bush		39.32%	94%	5%	1%
3.203	Other		2.62%	21%	71%	8%
	TOTAL		100.00%	50.32%	48.89%	0.79%
122.26	Votes		122.26	61.52	59.78	0.96
Kerry	Margin		-1.75			

[kiwi_expat]

3. Thanks

[TruthIsAll]

4. Breakeven is 15.3% of Gore voters for Bush, or 1 of every 6.5 dems)

Oh, yeah, I can believe that.
Sure. Like I believe pigs can fly.


Voted	2000		    Mix	Bush	Kerry	Nader
22.386	None		18.31%	41%	57%	2%
48.598	Gore		39.75%	15.3%	84.7%	0%
48.073	Bush		39.32%	90%	9%	1%
3.203	Other		2.62%	21%	71%	8%
	TOTAL		100.0%	49.53%	49.50%	0.97%

122.26	Votes		122.26	60.55	60.52	1.18
Kerry	Margin		-0.03			
		

[kiwi_expat]

7. That is assuming that only 90% of Bush-2000 voters voted for him in 2004

[TruthIsAll]

9. Well, kiwi, how long will you hold out on us? You still

have not responded.

Are you struggling to come up with a plausible implausibility to foist on us?

Or are you now convinced that Kerry won, since he wins all the plausible scenarios and even some which are implausible.

And Bush only wins scenarios which are categorically IMPOSSIBLE or so beyond the realm of our experience and common sense, that they must be rejected.

Will the foreman please read the verdict?

[kiwi_expat]

11. The verdict is not mine to make.

"Are you struggling to come up with a plausible implausibility to foist on us?" -TIA

What, my "11.5% Gore and 94% Bush voters for the 'War Prez'" scenario isn't good enough for you??

Something to keep in mind when labeling a scenario "implausible" because it does not correspond to the narrow 90-92% party switches in the polling-data demographics, is that Mitofsky, himself, said that the demographic data has a high MOE (I can't remember if it is 4% or 6%).

Surely the "11.5 G, 94 B" scenario is within the "realm of our experience and common sense".

So, if I were the judge, I would say the case has not been proven.


But, hey, I've always thought that Kerry probably won both the pop& EC vote. The people you need to convince are folks like liam_laddie's liberal friends (see "Seatbelts" thread). Maybe they are not as strict in their definition of "proof" as I.

[TruthIsAll]

14. Oh, no, not the MOE again. Come on Kiwi. It was 1.0%

How can you be so far off?
Did you read the Edison-Mitofksy notes at the bottom of the WP 13047 Exit poll screen shot?

I am paraphrasing:
"The exit poll is a randomly-selected sample with a 1.0% MOE.
Percentages may change slightly".

Kiwi, I thought you were up on the basics.

The MOE is 1.0% for demographic characteristics near 50%.
And the "How Voted in 2000" demographic percentages are near 50%.

[kiwi_expat]

17. I remember someone who had read the NEP pdf file saying that...

the demographic characteristics had a high MOE. (I have never been able to read the NEP pdf file, myself. I am dependent upon the mercies of others.)

Do you have a non-pdf link for the statement "The MOE is 1.0% for demographic characteristics near 50%"?

Also, do you happen to know if the "How Voted in 2000" question was on the first or the second page of the questionnaire? And what percent of respondents did not complete the second page?

Thanks.

[TruthIsAll]

19. I must say, I'm really surprised at your 4-6% MOE..

I have posted on the 1% MOE hundreds of times. And you you rely on the "mercies of others"? Why don't you just rely on the verifiable facts? Try Google once in a while. What did you say about trying to be more "rigorous"?

I can't tell you if it was the 1st or second page or what the percentage was. I just compared Edison-Mitofky's statement below and it to the little formula:
MOE = 1/sqrt(N)=0.875%, where N =13047.
They apparently rounded UP to 1%.

Here is the gif of the 13047 poll posted on the NEP/WP site 11/03/04.

Apparently Kerry did slightly better than I reported.

10% of Bush voters voted for Kerry.
8% of Gore voters voted for Bush.

I got the missing 3% "Other" from Final 13660 Poll.

Kerry won by 7.06 million votes after the weights were adjusted from an impossible 39%Kerry/41%Bush to a "plausible" 39.75/39.32...

Assumes 99% turnout of Gore and Bush voters
Voted 2000 Mix Bush Kerry Nader
22.386 None 18.31% 41% 57% 2%
48.598 Gore 39.75% 8% 91% 1%
48.073 Bush 39.32% 90% 10% 0%
3.203 Other 2.62% 21% 71% 8%
TOTAL 100.00% 46.63% 52.40% 0.97%
122.26 Votes 122.26 57.00 64.07 1.19
Kerry Margin 7.06

Please read the notes at the bottom.

http:
//media.washingtonpost.com/wp-srv/politics/elections/2004/graphics/exitpolls_us_110204.gif

[kiwi_expat]

22. I have googled my head off and here is the best I can find:

The NEP table showing the MOE (at 95% confidence interval):

You are correct, if all of the respondents have answered a question, the MOE is 1%. However there are four different questionnaires, and each questionnaire has an optional "backside". It is possible that a question that appears on the backside of only one of the four questionnaires would have a MOE of 4% or greater.

You might expect that the question "who did you vote for in 2000?" would be on all four questionnaires - but I'm not sure that it was. I have a detailed spreadsheet of the Ohio NEP data (prepared by Blue22) and the "Vote in 2000" question is no-where to be found. (It does have a question about party and one asking if a first time voter.)

The answers to all my questions are in the 4 NEP pdf files. This will spur me to buy a laptop and replace my steam-powered PC.

[TruthIsAll]

24. Excellent work! Great find! Now compare to the formula: MOE= 1/sqrt(n)

You have just driven another nail into the high MOE strawman.
That is a great find.

It MEANS that 1.0% is the universal MOE, regardless of characteristic weighting, if the sample size is large.

The MOE is 1% REGARDLESS OF THE CHARACTERISTIC WEIGHT WHEN YOU GET TO 8,000 SAMPLE-SIZE.

We have 11027-13,047 sampled in the PRELIMINARY POLLS, the plausible ones..

Let's not even consider the FINAL 13660, implausible, matched-to-the-vote sham. That poll has no credibility left.

MoE = 1/sqrt(N)

n 1/sqrt(n)
100 10.00%
200 7.07%
500 4.47%
950 3.24%
2350 2.06%
5250 1.38%
8000 1.12%

11027 0.95% Prelim 7:38pm Kerry wins 51-48
13047 0.88% Prelim 12:22am Kerry wins 51-48
13660 0.86% Final 2:05pm Bush wins 51-48

15000 0.82%

[kiwi_expat]

27. Minvis has verified that the "Vote in 2000" ques. was not asked in Ohio.

Ohio had about 2000 respondents. We don't know if the question was on the questionnaires in the other heavily-polled battle-ground states.

[kiwi_expat]

28. "Vote 2000" appears on only 1 of 4 questionnaires. It is on the backside.

I can finally read the NEP pdf reports!

The table I posted in #22, above, is from the following NEP National Methods Statement, and needs to be read in context:
http://www.exit-poll.net/election-night/MethodsStatementNationalFinal.pdf

The question "Who did you vote for in 2000?" appears on only 1 of the 4 National NEP questionnaires: the Version 3 questionnaire. The question is on the backside of the questionnaire. See:
http://www.exit-poll.net/election-night/Nat_Final.pdf

[dzika]

30. There were 102 Vote2000 responders for Ohio in the NEP Raw data

The raw data file released by NEP has 102 records for Ohio with responses to the Vote 2000 question. Some of the weights seem very high in comparison to others.


Weighted Vote2000 responders in Ohio

		Bush	Kerry	NoVote	Total
		
	Total	67	70	1	138
	
Bush 2000	50	12	1	63
Gore 2000	3	44	0	47
No Vote 2000	14	11	0	25
Other 2000	0	3	0	3

Raw Vote2000 responders in Ohio:

		Bush	Kerry	NoVote	Total
		
	Total	39	62	1	102
	
Bush 2000	30	9	1	40
Gore 2000	1	38	0	39
No Vote 2000	8	13	0	21
Other 2000	0	2	0	2

[kiwi_expat]

31. Thanks dzika !

How interesting that NEP showed an Ohio KERRY win for the Vote2000 respondents in the final weighted data for the national survey.

(The data I was originally looking at was for the Ohio state-questionnaire, which had no Vote2000 question. The Vote2000 question was on one-fourth of the national-questionnaires - in all states.)


Cheers

[dzika]

32. The weights

I find the weights to be interesting too. With the exception of nader 2000 voters, the Bush weights are significantly higher and Kerry still wins this sample.

		Bush	Kerry	NoVote
	
Bush 2000	1.65	1.48	1.44
Gore 2000	3.09	2.13	0.00
No Vote 2000	1.76	1.30	0.00
Other 2000	0.00	1.28	0.00

[TruthIsAll]

16. What experience are you referring to?

Surely the "11.5 G, 94 B" scenario is within the "realm of our experience and common sense".

Point out just one election in recent history where there was anything even approaching those numbers...

[kiwi_expat]

18. The realm of our experience: mid-term polling.

Common sense is a personal thing. My common sense (based upon years of observing the fickleness of voters) has lead me to believe it is plausible that large numbers of otherwise not-too-stupid voters might panic and choose the "War President", at the last second.

[TruthIsAll]

5. Can I ask how you came up with that particular scenario?

Inquiring minds would like to know.

[kiwi_expat]

8. I tried to find a combo least likely to be labeled "totally implausible"..

...(based on your previous reactions), that might produce a Bush win. Then I did some crude calculations with my little calculator.

Was that cheating?

[TruthIsAll]

10. Not at all. I assumed that's what you did. And I give you credit

for trying to keep the possibility alive that Bush won fair and square. Bill Clinton, that liberal champion of democracy, would be very happy to be proven he was right when he said that Bush won it fair and square.

[kiwi_expat]

12. IMO neither Kerry nor Bush have yet been "proven" to have won.

And even if Bush could be proven to have won (which I consider to be extremely unlikely), he certainly didn't win it "fair and square". Surely you don't think I would think he did.

Are you suggesting that, by trying to be rigorous, I am giving aid and comfort to the enemy? I thought this discussion was just among friends.

[kiwi_expat]

13. Maybe we should think of this as a civil - not a criminal - court case

Perhaps the concept of "proof" is inappropriate here.

How about settling for making a decision based on the "preponderance" of evidence. If I were a judge and I had to decide between a Kerry popular vote win and a Bush popular vote win, I would definitely decide for Kerry.

[TruthIsAll]

15. Ok, I take that as a mild concession. Welcome aboard. n/t

.

[vince3]

6. You Rock, TIA

I have been reading your posts since the last stolen election. You have done a great service over these last few months.

[TruthIsAll]

20. Thanks, vince. The naysayers always pop up whenever I come

up with new analyses, as I'm sure you have seen...

[kiwi_expat]

23. I must be paranoid....

I get the impression you are talking about me.

You are calling me a "naysayer" right after you have welcomed me "aboard"? That is not very nice. :-(

[TruthIsAll]

25. Relax. It was not meant for you. Your motivation is to get at the truth.

On the contrary, I commend your work.
You get down to brass tacks.
No BS.
You crunch numbers.

We are on the same team.

[kiwi_expat]

26. Thank you.

[TruthIsAll]

21. Come to think of it, even 3% of Gore voters crossing over is IMPLAUSIBLE.

This scenario seems VERY PLAUSIBLE to me.

99% Turnout
98% of Gore voters for Kerry.
 4% of Bush voters for Kerry.
 3% of Bush voters for Nader/Bednarik et al.

Kerry wins by over 9 million votes.

Voted	2000
		         Mix	Bush	Kerry	Nader
22.386	None		18.31%	41%	57%	2%
48.598	Gore		39.75%	2%	98%	0%
48.073	Bush		39.32%	93%	4%	3%
3.203	Other		2.62%	10%	70%	20%
	TOTAL		100.0%	45.13%	52.80%	2.07%

122.26	Votes		122.26	55.18	64.55	2.53
Kerry	Margin		9.37			

[LatePeriduct]

33. Implausible is right!

Unless you happen to think independants always get the votes. Tada.

[LatePeriduct]

29. I think I saw the dominionists reel back!

One, two, three, four.

Four scenarios you can't argue with. Totally amazing.

[LatePeriduct]

34. kick