100% ABSOLUTE PROOF: FINAL NEP OVERSTATED BUSH VOTE BY AT LEAST 2 MIL.
Edited on Mon Mar-21-05 02:01 PM by TruthIsAll
Here is incontrovertible proof that the FINAL National Exit Poll (13660) overstated The Bush vote by a MINIMUM OF 2 MILLION VOTES.
After proving it was at least 2 million, a little common sense will reveal that it was probably overstated by more than 3 million.
In the FINAL NATIONAL EXIT POLL of 13660 RESPONDENTS, POSTED ON CNN at 2:05pm on 11/3, 43% of voters claimed to have voted for Bush in 2000 while only 37% said they voted for Gore.
The PRELIMINARY EXIT POLL (13047 RESPONDENTS, 12:22am on 11/3) had this demographic at 41% Bush/38% Gore. So the numbers were adjusted from 41% in the Preliminary poll to 43% in the FINAL, OFFICIAL EXIT POLL.
BUT IT IS IMPOSSIBLE FOR 43% OF 122.26 million 2004 VOTERS TO HAVE VOTED FOR BUSH IN 2000 BECAUSE 43% of 122.26 = 52.57 MILLION VOTERS. AND THE BUSH 2000 VOTE WAS JUST 50.5 MILLION.
THAT'S AN OFFICIAL ERROR, NOT A THEORETICAL ONE. IT'S A 4% ERROR IN FAVOR OF BUSH. BUT THE EXIT POLL MARGIN OF ERROR, ACCORDING TO MITOFSKY EDISON, WAS +/-1.0%
THE ODDS ARE NEAR ZERO THAT THE 1.0% MOE WOULD BE EXCEED BY 3.0%. THEY ARE EXTREMELY LOW EVEN IF THE MOE WERE 2.0% (IT WOULD BE OFF BY ONE MILLION VOTES).
WHERE DID THESE 2 MILLION VOTERS COME FROM?
We KNOW that at least SOME fraction of Bush 2000 voters DIED. Some were INCAPACITATED and did not vote. Some decided to STAY HOME and did not vote. And some, believe it or not, actually VOTED FOR KERRY.
SO COMMON SENSE TELLS US THAT FEWER BUSH 2000 VOTERS HAD TO HAVE VOTED FOR HIM IN 2004, NOT MORE.
ASSUMING THAT 1-2 MILLION FEWER 2000 BUSH VOTERS ACTUALLY VOTED FOR HIM THIS TIME, THEN HIS 2004 VOTE WAS OVERSTATED BY 3-4 MILLION VOTES. THAT'S A 6-8% DIFFERENCE.
One final thought. The 2000 vote is an historical fact. In the ONLY demographic for which the final exit poll is matched to an actual vote count (in 2000), it is found to be absolutely bogus.
SO HOW IS ANYONE TO BELIEVE ANY OF THE OTHER RESULTS IN THE FINAL EXIT POLL?
75. Another way of proving the NEP is bogus and Kerry won....
Edited on Fri Mar-25-05 10:49 AM by TruthIsAll
According to the FINAL Exit poll, Bush won by 3.22 million
votes (51.11%-48.48%), based on how respondents voted in 2000.
The Final Exit Poll is MATCHED very close to the recorded
Mix Bush Kerry Nader
No 17% 45% 54% 1%
Gore 37% 10% 90% 1%
Bush 43% 91% 9% 0%
Other 3% 21% 71% 3%
100% 51.11% 48.48% 0.63%
Votes 62.49 59.27 3.22
But we know this result is IMPOSSIBLE since the MAXIMUM Bush
assuming NOT A SINGLE BUSH 2000 VOTER HAS DIED, was 41.26% of
122.26, or 50.45 million.
Let's determine what is both possible AND likely:
Assume the following:
1) 3.5% of Election 2000 voters have since died (multiply the
national annual death rate of 8.7 per thousand by 4)
2) ALL those living voted in 2004.
Then the absolute maximum percentage of returning voters was
40.2% for Gore and 39.8% for Bush.
Using these percentages and keeping all the other FEP weights
constant, then Kerry won by over 3 million votes.
VOTED IN 2000
Mix Bush Kerry Nader
No 17.0% 45% 54% 1%
Gore 40.2% 10% 90% 1%
Bush 39.8% 91% 9% 0%
Other 3.0% 21% 71% 3%
100.0% 48.52% 51.07% 0.66%
Tot. Votes 59.32 62.44 -3.12
an absolute squeaker of an election. There's a slim chance Fuckwit actually won by a hudredth of a percent. There's a bigger chance that Kerry won by as much as a percent. The process has been so completely corrupted by the right that we'll never really know for sure. Machines have been cleared and the data lost forever.
What I do know about NM is that to believe they won this state, you'd have to believe 17,000 people left their homes and TVs to stand in line to vote for a judge, but not vote in the national election. That was the discrepancy here. Apparently our ES&S machines simply deleted Democratic votes across the board, varying percentages in every county that used those damned machines, instead of flipping them the way the Diebold machines appeared to.
Do you have any data on the performance of those two companies? Did the ES&S machines delete votes nationwide, or were they clever enough to fiddle them differently in each state?
Thanks for these posts. They've been very enlightening, to say the least.
12. BASED ON THE US. ANNUAL DEATH RATE OF 8.7 PER 1000...
Edited on Mon Mar-21-05 07:49 PM by TruthIsAll
that is about 35 per 1000 over a four year period or 3.5%.
That means that approximately 3.5 million voters in 2000 died.
Assuming half voted for Bush in 2000, at least 1.75 million voters could not have voted this time. We know that 50.45 million voted for Bush in 2000.
The maximum number who could have voted for him in 2004 was therefore 50.45 - 1.75 = 48.7. And so 48.7/122.2 = 39.8% is the maximum percentage of 2004 voters who could have voted for Bush in 2000. That assumes that nor a single Bush 2000 voter: 1) sat out the election, 2) voted for Kerry, or 3) was incapacitated and unable to get to the polls.
If you believe the final exit poll 43%, then you must also believe that 52.57 million voted for Bush. We have just shown that this is TOO HIGH BY 3.8 MILLION AT MINIMUM, WHEN THE DEATH RATE IS CONSIDERED AND PROBABLY 2-3 MILLION MORE WHEN YOU CONSIDER THE OTHER THREE POSSIBILITIES.
If you believe the final exit poll 43% demographic, then you
must also believe that 52.57 million voted for Bush. We have
just shown that this is TOO HIGH BY 3.9 MILLION = 52.6 - 48.7
AT MINIMUM - AND PROBABLY 2-3 MILLION MORE WHEN YOU CONSIDER
THE OTHER THREE POSSIBILITIES.
THEREFORE, IT MAY VERY WELL BE THAT BUSH'S VOTE WAS 6 MILLION
TOO HIGH. REMEMBER, KERRY WON THE 13047 EXIT POLL BY 51-48%
AND EVEN IN THIS PRELIMINARY POLL BUSH GOT 41%, WHICH WE HAVE
JUST SHOWN TO BE TOO HIGH AS WELL.
USING THE PRELIMINARY EXIT POLL WEIGHTS, IF 39% OF 2004 VOTERS
VOTED FOR BUSH IN 2000 AND 40% VOTED FOR GORE,
THEN KERRY WON BY 52.2% TO 45.8%,
OR 63.4-56 MILLION.
USING THE IMPOSSIBLE FINAL EXIT POLL WEIGHTS, WITH 43% OF 2004
VOTERS VOTED FOR BUSH IN 2000 AND 37% FOR GORE, THEN BUSH WON
OR 62.5-59 MILLION.
THE ACTUAL RECORDED VOTE WAS BUSH:
OR 62.03-59.03 MILLION..
SO THE BOGUS VOTE WAS EXACTLY EQUAL TO THE FINAL BOGUS EXIT
HERE ARE THE CALCULATIONS:
VOTED IN 2000
PRELIMINARY 13047.........................FINAL 13660
Mix Bush Kerry Nader Mix Bush Kerry Nader
No 17% 41% 57% 1% 17% 45% 54% 1%
Gore 40% 8% 91% 1% 37% 10% 90% 0%
Bush 39% 90% 9% 0% 43% 91% 9% 0%
Other 4% 13% 65% 16% 3% 21% 71% 3%
100% 45.8% 52.2% 1.2% 100% 51.1% 48.58% 0.3%
I've only just reached this forum. I read one of your massive battle-dialogues recently -- to be frank, I find this stuff addictive because it seems to me to be the central issue of our times. The 2000 election was a terrible blow to democracy, and the 2004 election has us on critical life-support. Jeez. Never been this bad, not even 1980.
Having no idea who you are makes it more difficult to fully trust, of course, so I need to ask probing questions. No offense intended. At all.
These statements -- where what how?:
'We KNOW that at least SOME fraction of Bush 2000 voters DIED. Some were INCAPACITATED and did not vote. Some decided to STAY HOME and did not vote. And some, believe it or not, actually VOTED FOR KERRY.'
And where do you get most of your numbers? Is there a central research source you use?
Perhaps you just have a link to a compendium of your work on this matter. Or a compendium of links.
P.S. When's your book coming out? I call firsties.
20. No verification is necessary. All the information is readily available.
Edited on Mon Mar-21-05 10:47 PM by TruthIsAll
You can look it all up. Check the web sites for the national exit poll:
CNN 11027 respondents at 7:38pm, 11/2 WP 13047 at 12:22am, 11/3 CNN 13660 at 2:05pm
Check the state exit poll data downloaded by Simon at 12:22am Check the analysis done by Freeman and uscountvotes.org Check the Baiman/Simon analysis. Check out the Mitofsky report. Check out the articles of various Math/Stat PhDs. You want verification?
And Zogby was not stumped. He knows he got it exactly right. Just like he did in 2000 and 2002.
The stolen election should be at the forefront everyday here on DU, but incredible as it seems, some still don't believe in it while others have moved on. I think you should write a book of your journey proving how the election was stolen. It might just end up on the best seller lists and in turn boot that asshole out of office! :bounce:
27. But Gore won the popular vote in 2000 by 500,000 votes
So how in the hell could the 2000 exit poll show him getting just 37% to Bush's 43%? The "actual" result was Gore 50,999,897 (48.38% of 105,405,100 total votes cast for all candidates), vs. Bush's popular vote count of 50,546,002 (47.87%). That's a 15% MOE in the Gore exit poll number. Your numbers can't be right--could you post a URL so we can look at the source?
When you say: "BUT IT IS IMPOSSIBLE FOR 43% OF 122.26 million 2004 VOTERS TO HAVE VOTED FOR BUSH IN 2000 BECAUSE 43% of 122.26 = 52.57 MILLION VOTERS. AND THE BUSH 2000 VOTE WAS JUST 50.5 MILLION."
You assume that everyone told the truth about who they voted for in 2000. It's not altogether unlikely, IMO, that a lot of folks who leaned toward Bush in 2000 but didn't get out to vote would claim they voted for him in an exit poll four years later. I think your case is fairly weak here, and that the best arguments for massive vote fraud remain statistical (swing-state "red shift" is very convincing) and anecdotal--all those voting machines that registered Bush votes when the Kerry button was pushed.
40. PERHAPS THIS WILL CLEAR IT UP. PLEASE RESPOND EITHER WAY.
Edited on Tue Mar-22-05 04:30 PM by TruthIsAll
YOU: "I agree with all you have said - but perhaps I do not follow exactly the argument being made"
The argument being made is this: THE FINAL EXIT POLL DEMOGRAPHIC CATEGORY OF "HOW THEY VOTED IN 2000" WAS WEIGHTED USING AN IMPOSSIBLE NUMBERS (43%-37%) TO MAKE IT APPEAR THAT BUSH WON, EVEN IF IT MEANT ONE HAD TO IGNORE THE FACTS OF HISTORY BY OVERSTATING THE BUSH 2000 VOTE BY OVER 2 MILLION (4%) IN ORDER TO FORCE THE FINAL 2004 EXIT POLL TO MATCH THE RECORDED VOTE.
IF THIS WAS DONE WITH A VERIFIABLE DATA SOURCE (THE 2000 VOTE), THEN WE HAVE TO ASSUME THAT OTHER ARBITRARY ADJUSTMENTS HAD TO HAVE BEEN DONE TO ALL THE OTHER UNVERIFIABLE DEMOGRAPHIC WEIGHTS, AS WELL.
YOU: "I thought it was that the FINAL 2004 percentages as per 2004 questionnaire and Mitofsky weighting in 2004 do not reflect 2000 election history".
THE WEIGHTINGS DO NOT REFLECT 2000 HISTORY, BUT THEY SHOULD HAVE. SINCE MITOFSKY ADJUSTED THE WEIGHTINGS TO MATCH THE 2004 RECORDED VOTE WITHIN 14 HOURS AFTER THE POLLS CLOSED, THEN HE ALSO HAD TO MATCH TO THE 2000 VOTES FOR THIS CATEGORY.
THIS WAS AN OPPORTUNITY TO QUESTION THE NUMBER, BUT HE DID NOT DO SO, AND IN FACT RAISED THE NUMBER FROM 41% TO 43% (WHICH WAS IMPOSSIBLE) WHEN HE SHOULD HAVE LOWERED IT.
THERE IS NO EXCUSE FOR RAISING THE 41% NUMBER, WHICH IS ALSO OBVIOUSLY TOO HIGH TO BEGIN WITH, SINCE APPROXIMATELY 1.75 MILLION BUSH 2000 VOTERS DIED, BASED ON ANNUAL DEATH RATES - THAT MEANS THAT IT HAD TO BE UNDER 40%. THAT IS THE ABSOLUTE MAXIMUM NUMBER OF BUSH 2000 VOTERS WHO COULD HAVE VOTED IN 2004.
BUT 40% MUST ALSO BE TOO HIGH, BECAUSE IT ASSUMES THAT NOT A SINGLE LIVING BUSH 2000 VOTER STAYED HOME IN 2004 AND DID NOT VOTE.
"AND IF THIS WAS DONE WITH A VERIFIABLE DATA SOURCE (THE 2000 VOTE), THEN WE HAVE TO ASSUME THAT OTHER ARBITRARY ADJUSTMENTS HAD TO HAVE BEEN DONE TO ALL THE OTHER UNVERIFIABLE DEMOGRAPHIC WEIGHTS, AS WELL."
We are on the same page here, except I do not see it as a given. However,the above was my conclusion also after my smell test of the adjustments. There is a real "say you voted for the winner" effect - so the percentage of folks claiming in 04 that they voted for Bush in 00 can be expected to be higher than the actual percentage of the vote that Bush got in 00. I'd expect the "error" - just not as large as the FINAL exit poll says it was
"1.75 MILLION BUSH 2000 VOTERS DIED" - the 50 million Bush voters, using 1994 GAR Proj. 2002 mortality, would if average age 50 have a .99122 chance of living 4 years, meaning about 500,000 Bush voters in 2000 would have died by 2004. It does not change your logic - but I'd revise the statement to "450,000 to 500,000 Bush voters were expected to die". No need to go beyond the data you have - which I agree implies that the race was won by Kerry - and that the exit poll adjustments are out of the range of reasonableness.
YOU SAY: "1.75 MILLION BUSH 2000 VOTERS DIED" - the 50 million Bush voters, using 1994 GAR Proj. 2002 mortality, would if average age 50 have a .99122 chance of living 4 years, meaning about 500,000 Bush voters in 2000 would have died by 2004. It does not change your logic - but I'd revise the statement to "450,000 to 500,000 Bush voters were expected to die". No need to go beyond the data you have - which I agree implies that the race was won by Kerry - and that the exit poll adjustments are out of the range of reasonableness".
Why do you quote a mortality stat for an average age of 50+? Why not apply the 8.5 per 1000 annual death rate across the board (3.5% over a four year period)?
What about voters who died from 20 on up? There were 105 million voters in 2000. Why would the annual death rate not apply here?
Why do you quote a mortality stat for an average age of 50+?
The death rate increases with age. Indeed if the average Bush voter was about 73 years old, a 3% death rate would be correct per year
Why not apply the 8.5 per 1000 annual death rate across the board (3.5% over a four year period)?
I do not know where the 3.5% death rate comes from that - it should not be 4 times the death rate in my post because that rate was for a 4 year period - the 8.5 per 1000 is the death rate for 4 years - about 2 per year per 1000.
What about voters who died from 20 on up?
- near zero death rates when you are young do not affect the mix that much - but I assume an average age of 50 - so a 20 year old and a 80 year old are 2 aver age 50 persons. They are all in the stat. To the extend older folks bias the death rate up - and vote more - we should use a weighted by number voting at a given age death rate rather than an average age. At age 80 220 out of a 1000 will die in 4 years, so an average would be 110 out of a 1000, at least - and not 8 per 1000 that the age 50 implies. We die fast when we get old! So I am not saying your numbers are unreasonable - just that they imply a lot of old folks voted for Bush.
There were 105 million voters in 2000. Why would the annual death rate not apply here?
It does. ============================================== I am coming to the conclusion your 1.75 mil died is not so unreasonable - just that it assumes few young (under age 50) voters
I'm not arguing that--the degree of swing state "red shift" is so improbable that vote fraud is the only reasonable explanation. I'm saying that your assertion--that too many voters claimed to have voted for Bush in 2000, therefore vote fraud must have occurred--is weak in comparison. But whatever. It's obviously important to you, so have at it.
39. You agree Kerry won, so you should also agree the exit poll was rigged.
"I'm not arguing that--the degree of swing state "red shift" is so improbable that vote fraud is the only reasonable explanation. I'm saying that your assertion--that too many voters claimed to have voted for Bush in 2000, therefore vote fraud must have occurred--is weak in comparison. But whatever. It's obviously important to you, so have at it".
Fine. You agree on the conclusion. But you disagree on the demonstrable proof that the exit poll was manipulated.
How can you call my argument "weak" when it is based on a mathematical fact?
How can you call my argument "weak" when it is the ONLY demographic which references ACTUAL historic vote data?
Sure there is much factual evidence of voting "glitches" which favored Bush 99.9% of the time. The exit poll discrepancies confirm that evidence.
44. It is still just a smell test - the word rigged for the election is solid
but I would not use the word rigged for the exit poll.
The rules of the game required a final fit be made between the "real" votes and the exit poll.
It was the "real" votes that were rigged. The FINAL exit poll was messed up by that fraud.
The only exit poll rigging is the fact that our pollesters do not scream that the final fit violates the smell test - and that it therefore implies the election "real vote" numbers are a fraud.
In other countries we call for new elections IF - and only if - there are reports of observed fraud in the election. For some reason the 40,000 reports in the US Election are deemed not enough to trigger that response - so we are left with "fixing our exit poll procedure" -
48. No, papau, they changed the preliminary exit poll to match the vote.
Why did they do this, knowing that at least one demographic was demonstrably false?
1-Why not leave well-enough alone and just say that the 13047 exit poll, which Kerry won by 51-48 with a 1.0% MOE, did not agree with the recorded vote?
2-Why not just say that there was no statistical explanation for it?
3-Why propose the Reluctant Bush Responder theory, while at the same time expecting us to believe the complete opposite (and impossible)43% Bush/37% Gore spread of those repondents who voted in the 2000 election?
4-Why raise the virtually impossible 41% Bush 2000 voter weight to the absolutely impossible 43%?
49. I agree the changes stink - but making final changes is SOP for exit
So "changes" that produce exit polls closer to real vote is standard operating procedure.
But they should have just noted that the 13047 preliminary exit poll did not agree with the recorded vote?
Since they - for a reason I do not know but which I suspect has to do with GOP lawyers and media threats - could not reference the 40,000 events indicating massive fraud, they could not say anything other than the model was way wrong - Leading to the Reluctant Bush Responder theory - while as you note they do not describe or discuss all the other implied "facts" of the exit model that are at odds with logic (such as percent of Bush 2000 voters voting and the hispanic split (other hispanic exit polls showed no change from 2000 - but this official one has a huge increase in GOP Hispanics).
We agree -The smell of rotting fish overwhelms the senses.
YOU "I'm not arguing that--the degree of swing state "red shift" is so improbable that vote fraud is the only reasonable explanation. I'm saying that your assertion--that too many voters claimed to have voted for Bush in 2000, therefore vote fraud must have occurred--is weak in comparison. But whatever. It's obviously important to you, so have at it".
Not quite. I am saying that the final exit poll was bogus, because it replaced a very improbable weighting (41%) to an absolutely impossible one (43%) in order to match the recorded vote count. It is a mathematically impossibilty, because it added phantom voters to Bush's 2000 vote total, rather than subtracting out the voters who actually passed on. This could not have been due to pure chance.
The Final Exit Poll was supposed to produce a more "accurate" result by matching to the recorded vote count by re-weighting the Preliminary Exit Poll of 13047. But in order to achieve this result, it had to create 2 million phantom Bush 2000 voters. This produced a 4% discrepancy (2/50) in a poll which has a 1.0% margin of error. In actuality the error was 6-8% if you consider that based on mortality tables, 1.5 to 2 million Bush 2000 voters passed on before the election.
Now, since the 43% poll weighting is impossible for this demographic, so must the vote count be impossible - because use of the bogus 43% weight produced the same final Bush vote margin (51-48%) as do all the other category weightings.
THE VARIOUS CATEGORY WEIGHTINGS SHOULD ALL PRODUCE BUSH MARGINS WITHIN 1% OF EACH OTHER. THIS CATEGORY WEIGHTING IS IN ERROR BY 6-8%. IF THIS CATEGORY IS THAT FAR OFF, THEN WE MUST ASSUME ALL THE OTHERS ARE AS WELL.
53. TIA, I would like to thank you for your work on this as well.
Here's what I am seeing from your analysis. The problem ISN'T whether there was a "bandwagon" effect to those polled in 2004 aabout whom they voted for in 2000, that is irrelevant. The PROBLEM is the data gleaaned from that question was very easily varified by the hard data of number of votes as recorded in the 2000 election.
To show my point let's pretend that EVERYONE who answered the question in 2004 said they voted for bush* in 2000. That would mean that 100 percent of 2004 voters would have voted for bush* in 2000. We can clearly see this is BAD DATA and in no way, shape, or form should we use this # to adjust our final exit polling data. That would mean we would have to say John Kerry got ZERO votes.
So what is a valid number? Easy, once the final number of votes are in simply divide the number of bush* votes in 2000 by the TOTAL number of votes recorded in 2004. So forgetting any of the factors of people changing from bush* votes in 2000 to nonbush* votes in 2004 the MAX percent the question should garner with the actual numbers you come up with a MAX of 42 percent. It doesn't matter then if 70 percent of the people answered due to the "bandwagon" effect because an upper limit is easily figured and any number above that should be thrown out as garbage.
The evidence of shenanigans is this number being above what it could possibly be. It appears the data was massaged by reverse engineering. That is they had the total number of votes received and in order for the result to be what it ended up being they had to fall back into the exit poll data and were stuck with an impossible result.
So how can any final poll be adjusted using a number that is IMPOSSIBLE to achieve? In my far fetched example of 100 percent we can see it leads to a crazy result, but in reality only a point or two is needed to produce the desired results, but the deception is the same.
60. Thanks for the clarification, I didn't have the exact number at hand
when posting and used 120 million as the 2004 total so my 42% is rough. That's the number then that should have triggered a "WTF?" when any final number juggling resulted in that percentage being surpassed, unless fo course you're rigging the vote and can't cover it up any other way.
persistance is still inspiring. Seeing you still hard at work over 4 and 1/2 months after the evil deeds is quite reasuring that something has got to get throught to these numbskulls. People seem to be pushing legal battles all over the country in their prospective states about the nov 2004 election fraud and yet others are trying the legislative angle.
Please keep it coming. There is no election reform 2004 without you.
Allow me to play Devil's Advocate here: Weren't most of the exit polls conducted in heavily contested areas. Areas which would have had a greater "Get out the Vote" campaigns by both sides over the last 2 Federal Elections; and therefore a higher than average turnout?
This would mean a greater number of voters compared to polling stations that did not have as good a turnout. So yes, the numbers don't add up, but the error could easily be the pollsters fault for not taking into account 'slower'& 'average' polling stations and creating an average only from the busiest stations.
Not saying you're wrong, just something that has to be looked into before this goes further. It definitely is evidence that something is wrong, I just think it's more likely to be a pollsters error.
59. You divert from the point of the post: 41% to 43% was IMPOSSIBLE
Edited on Wed Mar-23-05 10:01 AM by TruthIsAll
YOU: Allow me to play Devil's Advocate here: Weren't most of the exit polls conducted in heavily contested areas. Areas which would have had a greater "Get out the Vote" campaigns by both sides over the last 2 Federal Elections; and therefore a higher than average turnout?
ME: THERE IS ABSOLUTELY NO EVIDENCE OF THAT
This would mean a greater number of voters compared to polling stations that did not have as good a turnout.
ME: SEE ABOVE
So yes, the numbers don't add up, but the error could easily be the pollsters fault for not taking into account 'slower'& 'average' polling stations and creating an average only from the busiest stations.
ME: ARE YOU SUGGESTING MITOFSKY/EDISON, AFTER 25 YEARS IN THE BUSINESS, DON'T KNOW THE BUSINESS?
Not saying you're wrong, just something that has to be looked into before this goes further. It definitely is evidence that something is wrong, I just think it's more likely to be a pollsters error
ME: I JUST THINK IT'S MORE LIKELY THAT YOU ARE THROWING OUT TALKING POINTS BECAUSE YOU DON'T OR WON'T ACCEPT THE TRUTH..
In order for the percentages to go from 41 to 43 over the whole poll numbers, it seems that an extraordinary number of votes would have had to been listed in the bush late additions in order to make a 2% shift in the 13047 number.
IOW, there would have had to of been something like Gore 20% bush 80%, in the final 600+ poll.
I made a stab at the calcs, rattled me brain. But I think there is something there, eh?
There is no way the 41% is possible, much less the 43%.
The only way they could get from 41 to 43 in 613 is that of the 613 the b poll would have to be @80%, or 490 votes, and 123 for Kerry, and that is just another piece in the big pan of fudge Mitkosfsky cooked up.
And they thought we'd just eat it and go our merry way? Fools!
TIA, I am impressed with your efforts. You've hung in there and fought off the waves of BS thrown at you by the naysayers. Numbers don't lie, only people do, eh?
I was convinced by this argument for about 2 hours yesterday. I hate to say it, but having thought over it and worked on the figures I am now less convinced for 2 reasons.
1 /the calculations are based on the fixed (present CNN exit polls that have been fiddled. If you take the 12.20 exit polls which were "less fixed", the figures are much less convincing. (I agree that that serves them right for fiddling with all the figures)
2/The argument relies on the fact that respondents remember accurately how they voted 4 years ago. I remember reading somewhere else (a discussion on exit polls in the UK) that the question about how you voted in the last election (several years before) is the one to which the replies are the most unreliable. It is possible that quite a few (5% or more?) either think they voted differently or don't really remember.
OK you say that on average those errors would tend to cancel each other out (same number wrongly stating they voted for Gore as those wrongly stating they voted for Bush). But who knows maybe (some) people are more likely to incorrectly think they voted for the winner than the loser last time. You can be sure that the answer from NEP would go something like that.
As you know I have spent a lot of time like you on the exit polls, and would love the smoking gun to be found. I think this argument is interesting but I doubt it's the smoking gun. It needs more data like indications from other years about the accuracy of the "HOW I VOTED LAST TIME" replies.
But we should not give up, the smoking gun may be in all those figures, this could be it, or it could be something else. We may still be missing the obvious.
What is clear is that NEP either 1/ lied again in their report about what the exit polls were really stating. or 2/ deliberately sent out vote predictions that were not backed up by any of their data, neither their "best GEO survey indicator" nor their so called "composite estimate" which is based partly on the exit polls and partly on the opinion polls of the previous week!
This became clear to for Colorado - simply because I have an original document with the Vote estimate on it. The vote estimates were confidential and in general not published, neither are they in the NEP report! You can check out the data on www.exitpollz.org - (then Select Colorado from the CNN 02/11 interface.
I am still trying to figure out whether what they did in Colorado was typical or whether whether there is a particular reason why they would pretend that their Colorado Presidential exit poll was spot on. (You may remember that I pointed out way back in January that it was strange that the Senatorial exit poll was spot on but not the presidential - especially as it was in reality the same respondents!
I'll try and spend more time on checking that out. Keep up the good work.
i wanted to highlight an issue in case it affects some of your analysis. it doesnt affect your primary point about the maximum number of bush votes, but it might be significant wrt some of your subsequent thinking/analysis.
the NEP didnt conduct 13660 interviews (or 13047, or any of those numbers) - despite what cnn (and others) are *still* reporting on their site. they actually only conducted 11719 interviews on election, and 500 pre-election day interviews, for a total of 12,219.
mysterypollster explains it best, so go visit there, but basically the 500 early (phone) interviews were counted 4 times, before being reweighted down by a factor of 0.25, but they forgot to change the total number - therefore the 13660 is too high by approx 1500.
note also that all 3 sets of exit poll data have this same error - basically, before election day started, there were already 2000 (or 500) entries in the system.
it isnt obvious whether MER has told cnn et al about the error, and if not, why not. similarly, its still not obvious why there is a discrepancy between the 13660 and the 13047.
Ever have one of those situations where 2 equations come to different conclusions, but both look right. I'm having one right now.
I must be simplifying too much....but to me, 3.27% of 13660 is 447.
I was looking for how many respondents would need to be oversampled (or give a wrong answer) in order to make the poll 4 million off. I wasn't looking for a comparison to the 13047 number (but I certainly understand what you are saying about the final 613).
I'm not questioning your math, I'm trying to find my error.
Feel free to say 'I'm not here to teach you basic math, ya dumbass' 8)
"I must be simplifying too much....but to me, 3.27% of 13660 is 447".
Fist consider the popular vote percentage, then apply it to the 13660.
Bush ended up with 50.73%. To match this, he needed 667 of the final 613 (WTF?). He already had 48% (6263) of the first 13047. IOW, to gain 2.73% in the popular vote, he needed 109% (667) of the remaining respondents (613).
Powered by DCForum+ Version 1.1 Copyright 1997-2002 DCScripts.com
Software has been extensively modified by the DU administrators
Important Notices: By participating on this discussion
board, visitors agree to abide by the rules outlined on our Rules
page. Messages posted on the Democratic Underground Discussion Forums are the
opinions of the individuals who post them, and do not necessarily represent
the opinions of Democratic Underground, LLC.