Final demonstration based on public data that Bush DID NOT WIN.

RESIGN NOW, NOT LATER. TAKE CHENEY WITH YOU (AND SCHMIDT FOR THAT MATTER)
(Reprinted with the permission of the author.)

PLEASE SHARE THIS POST WITH A FRIEND. Happy Thanks Giving!

All pre-election and election day polls showed Bush 47.8% to 48.7%
1) Bush's 11-poll average election day job approval was 48.5% (1.0% MoE).
2) His pre-election national 18 poll weighted share was 48.7% (0.7% MoE).
3) His pre-election 50 state poll weighted share was 48.5% (0.6% MoE).
4) His National Exit Poll (12:22am timeline) vote share (gender demographic)
was 47.8% (1.2% MoE, assuming a 40% cluster effect).
5) His State Exit Poll weighted national vote share was 48.3% (0.50% MoE,
assuming a 40% cluster effect).

Here is the Pre-election 50 state poll share calculation:
Bush's weighted poll share was 47.0%, as compared to Kerry's 47.5%.
That accounts for 94.5% of the total.

Add 1.0% for third parties, for 95.5% of the total.
That leaves 4.5% undecided.
Of the 4.5%, add 1.50% (1/3) to the Bush share.
Therefore, Bush's pre-election state poll share: 48.5%

And the Pre-election National 18 poll share calculation:
Bush's weighted share was 47.30%, as compared to Kerry's 47.55%.
That accounts for 94.85% of the total.
Add 1.0% for third parties, for 95.85% of the total.
That leaves 4.15% undecided.
Of the 4.15%, add 1.40% (1/3) to the Bush share.
Therefore, Bush's pre-election National 18 poll share: 48.7%

Consider the Law of Large Numbers.
The mean of the the FOUR independent pre-and post election poll
group means {48.7, 48.5, 47.8, 48.3} is 48.33%.
That's within 0.17% of Bush's 48.5% PRE-ELECTION JOB APPROVAL!

The probability is 97.5% that Bush got LESS THAN 48.7% of the vote.
It's virtually 100% that he got LESS THAN 49.0%.

Want more of this?
Bush's current 37% job approval is confirmed by TWO INDEPENDENT poll sets:
1) the weighted average of 50 state polls (0.6% MoE).
2) the unweighted average of 12 national polls (1.0% MoE).

These results confirm prior election studies.
An incumbent's TRUE vote is directly correlated to job approval.
They EXACTLY matched in 2004.
It's also additional confirmation that the 12:22am exit polls were correct.

So naysayers, will you now claim that
1) 50 pre-election state polls were wrong?
2) 18 pre-election national polls were wrong?
3) 11 pre-election Bush approval polls were wrong?
4) 50 post-election state exit polls were wrong?
5) the National Exit poll (12:22am, 13047 respondents) was wrong?
6) 12 post-election national approval polls are wrong?

At the same time, will you claim that the Final National Exit Poll,
which was the ONLY poll matched to the recorded vote, was correct?
Even though it is a fact that impossible Voted 2000 demographic
weightings are necessary for Bush to have won it?

Naysayers,
You were wrong a year ago.
You were wrong 6 months ago.
And you are wrong now.

If the election were held today,
Bush would lose in a landslide of epic proportions.
Even Diebold couldn't steal it for him.

Kerry won.
He really did.
He got 12 million more votes (63mm) than Al Gore (51mm).
Maybe this analysis will convince you.

But it's a moot point.
Al Gore is still President.

Salon?
Mother Jones?
What ever happened to investigative journalism?
Time to get with the program.

Prove it to yourself.

Download the Excel Interactive Election Model.
Find the link at TruthIsAll.Net (link below)

TruthIsAll.Net--Comprehensive Democrratioins of Election Fraud Here including the Excel Interactive Election Model!

In this chart, note the PERFECT correlation.
BUSH APPROVAL RATING vs. EXIT POLL. Survey USA 11/13/05
Bush exit poll and CURRENT approval rating trend lines have identical slope.


BUSH STATE APPROVAL DEVIATIONS FROM EXIT POLL.Survey USA 11/13/05
This related chart shows the deviations between the state exit polls and current approval ratings


State and National Pre-Election/Exit Poll Simulations and National Exit Poll Timelines.
|

:kick:

I miss TIA!:cry:

Peace.

post-election!

:cry:

is

http://www.democraticunderground.com/discuss/duboard.php?az=user_profiles&u_id=106042

and

http://progressiveindependent.com/dc/dcboard.php?az=show_topics&forum=120&page=

:kick:

:dem:

What a great resource!

Thanks!:hi:

:hi:

:hi:

-- for Bush -- were certified in January 2001? And again in January 2005?

Whether Al Gore or John Kerry won the popular vote is irrelevant. Whether Diebold or the BFEE or whatever funky boogieman you want to blame stole the election is irrelevant, too.

The Electoral College voted, and it voted for Bush, twice. That is the only vote that matters, Constitutionally speaking, in determining a president. Investigation, impeachment, resignation, flat-out confession of theft, whatever does NOT make Gore president. We don't get a reset button. Sorry. (If anything -- assuming Bush and Cheney both resign/are removed -- it would make Dennis Hastert president.)

That and there's no guarantee that if Gore had won in 2000, he'd have been reelected in 2004.

Until we change the Constitution, whoever wins the Electoral College - wins - no matter how much it sucks. We need to focus on fair, future elections.

(And I think in 20-20 hindsight, Gore probably would have lost the 2004 election - contrary to that they say about Bush and 911 - the GOP would have gone full bore in blaming Gore (and Clinton) for 911 and would probably have tried to impeach).)

if Gore was the president at the time, you mean to tell me that you still believe 19 Arabs who took a few hours of flight training beat these guys to there target in a commercial airplane. Puleeeease

This



Beat this, come on!

:eyes:

An F-16 does like mach 2.5 without a full load

at that time, the fighter jets would have been scrambled. They would have broke a whole lot of windows along the way (literally) but they would have stopped them planes from hitting the trade center.

I don't think it would have. Bush ignored a boatload of intelligence indicating that Al Quaeda was going to attack us.

Gore wouldn't have.

so no, Gore would not have allowed 9/11 to happen because he would have done his job instead of taking the entire month of August off.

There were a lot of attacks prevented during the 90's. Terrorism worldwide is up dramatically since then. And people are scared shitless. The only people in danger during the 90's were slutty whitehouse interns. :evilgrin:

How did 2000 work, the Supreme Court was easy breezy with the law,
there is nothing in the constitution that provides for the Supereme
Court to interfere in the elections process, disputed elections are to be handled by the states or by congress at the national level.
So how is it legit? I still say, we were robbed.

That's who won it for Bush in 2000 and it happened because of bully-boy tactics and the inability to expose the "felon-purge" eliminating black voters in Florida who would have voted had their names not been wrongly removed from the registration database.

Every vote matters, including votes that are denied.

Black Americans are sick to death of "spoiled ballots," "registration problems," etc. It's been going on for decades. All this is known and available to the public.

Enough of the election corruption, of all times.

EVERY VOTE MATTERS (except a bunch of right wing judges pushing states rights who suddenly assert the Federal prerogative when it suits their ultimate political purposes)

I'm not even arguing whether Bush won or lost, or whether there was cheating or not. This is just not evidence. It's a gross misrepresentation of how polling is done.

A poll figures out how the 100 million people will vote by questioning 500 or 1000 of them. We all know this. They use formulas to determine what groups of people will vote, and how they will vote as a group. So they find a couple of token white rich guys, and few token middle class women, and a few people from every other demographic group they think will matter, and they get their opinions, and they calculate totals from those numbers, using their formulas.

Their formulas are based on guesses. These guesses take into account how people voted in the last few elections, and a bunch of other details. But they are just guesses. If the formulas are wrong, then the results are wrong. The same is true of the exit polls. They take random samples and extrapolate outwards. Exit polls are even less reliable, because until the final voting is tallied, they don't know how to use the data they have. They have to factor in regional turnout, precinct turnout, total turnout, to understand what groups are voting and what groups aren't.

One minor change in turnout throws all polls off. They guess that equal numbers of Repubs and Dems will vote, for instance, but for some reason Dems stay home in greater numbers than expected, or they guess that all of the new voters registered will vote for Kerry but instead a larger than expected number vote for Bush. How could they know that, since they are new voters? Every pollster makes similar assumptions, and if those assumptions are off, they affect all the polls.

Polls are just educated guesses. That's why they come with such large Margins of Error. They aren't evidence unless you can tie them to something else. These types of formulas mean nothing, and that't the reason no investigative journalist is paying attention to such numbers.

Again, I'm not arguing that there wasn't cheating. But to prove it you need evidence, not number games.

I'll accept that, But here is a "red flag" for you, They are still pushing these vote count and tabulating machines that count the vote in secret. There are no checks and balances for them, thats a FACT.

Now if we put two and two together "numbers way off" in 2004, and the FACT that they are parading these election theft machines across the country, if this doesn't throw up a red flag to you I'm SHOCKED.

I specifically said that I wasn't even questioning whether it was rigged. I believe there is life on other planets, but I don't believe goofy videos of alien autopsies in Roswell prove it.

:sarcasm:
Bush Won
the Election
:rofl:

But to prove it you need evidence, not number games.

The numbers you dismiss so readily have no party or platform, no political ideology whatsoever.

TRUTH

has relatively little to do with the accuracy of polls. It is normally computed on the assumption of random sampling, which means that the larger the sample the smaller the MoE. Exit polls therefore have a very small MoE; pre-election polls rather larger.

However, the MoE does nothing to protect your "educated guess" from any violation of randomness in your sampling, and it is extraordinarily difficult to get a truly random sample - which is where the educated guesses actually start. Pollsters often aim at a "representative" rather than a "random" sample, by having demographic quotas, by stratifying the sample, or by weighting for the visible demographic characteristics of non-responders.

It is a scandal and a tragedy that we cannot rely on your vote-counting system (or even rely on your entire electoral system to ensure that everyone entitled a vote can get as far as casting a vote). But that simply means that both the vote count and the polls may be inaccurate; it doesn't not magically confer accuracy on the polls.

With the size of the national sample and the accumulated size of all the state samples, the discrepancy between the exit polls and the pre-election polling and the outcome is ridiculously unlikely.

Yes. You are absolutely right. It was not due to chance. No-one, least of all Mitofsky disputes that.

The question is: what was it due to? Something was biased. It was either the count or the poll.

The evidence IMO to date strongly favours the poll, although I would agree that the data leave room for enough fraud to swing Ohio (with a following wind and, of course, voter suppression).

The MoE only tells you the margin of error due to random sampling. It tells you absolutely nothing about non-sampling error. Sources of non-sampling in surveys include non-response bias and sampling bias. They are challenging methodological problems that are remarkably hard to eliminate, and you can rarely be sure you have done so, certainly not if your response rate is only around 50%.

"...unlikely to be due to chance.

The question is: what was it due to? Something was biased. It was either the count or the poll."
-------------------------------------------------------------------------------

Right.
Either the count or the poll was biased.
Must have been the poll.
There is no way the count was biased.
There is no evidence to suspect foul play in the count.

Not in Ohio.
Not in Florida.
Not in Pennsylvania.
Not in New Mexico.
Not in Nevada.
Not in Missouri.
Not in North Carolina.
Not in Minnesota.
Not in New Hampshire.
Not in Texas.
Not in Washington.

No vote count bias whatsoever.
Right.
Must be the poll.
All 130 of them:
11 Pre-election Bush approvals
50 Pre-election states
18 Pre-election nationals
50 Post-election state exits
12:22am Post-election national exit (13047 respondents)

Even Mitofky's exit polls.
Except for the ONLY one which was matched to the vote:
The FINAL.
1:25pm Nov. 3 National Exit (13660).
In this case, BOTH poll and count had to be unbiased.
Even if the 43%Bush/37% Gore demographic weighting is IMPOSSIBLE.

But it can't be.
BushCo would never cheat.
They may lie about WMD, but never about the vote.
So let's just disregard the FACT that 43/37 is IMPOSSIBLE.

Thanks.
This make a lot of sense.
The pollsters were not wrong.
"It's the BIAS, stupid."

You convinced us:
Exit poll respondents lied.
Gore voters forgot.
Bush voters were reluctant.
Even if the Final Exit Poll (43/37) says otherwise.

But, but....
You agree that 43/37 is impossible.
And that 43/37 implies

1) Bush 2000 voters were oversampled; in fact, they were anxious.

but, but...

2) Bush 2000 voters could not have comprised 43% of the 2004 vote.
That would mean Bush got 52.57mm votes.
We know he only got 50.45.
But 1.75mm of them died.
So at most 48.7mm Bush 2000 voters (39.8%) could have turned out in 2004.

Sorry, but...
There goes your rBr (reluctant Bush responder).
There goes your Gore Voter False Recall.

You realize that equal turnout makes Kerry a landslide winner
of the Final Exit Poll, don't you?
Hell, Kerry won by 51-48 (4mm votes) when it was 41Bush/39 Gore at 12:22am.
And that means Kerry won EVERY time line.

Bias?
Or just BS?

Naysayers, come out of the dark.
And into the light.

It's time to view the forest.

As Deep Throat told Woodward (he was a journalist then):
"You are missing the overall".
-------------------------------------------------------
-------------------------------------------------------

To quote Febble (my responses in UPPER case):
"If you read both pieces you will see that there are a number of ways in which the plot could be compatible with fraud, even with widespread fraud..."
THANK YOU

it is, however, very difficult to reconcile the plot with a fraud mechanism that could have contributed substantially to the exit poll discrepancy.
DEFINE SUBSTANTIALLY.
BUT WAS IT ENOUGH TO STEAL THE ELECTION?

It would therefore seem more likely than not that the exit poll discrepancy was caused largely by polling bias rather than fraud.
WHAT ABOUT THE PRE-ELECTION POLLING DISCREPANCIES?

However BECAUSE the analysis is perfectly compatible with the kinds of electoral injustice that we KNOW happened in 2004, just as they did in 2000, then the analysis does nothing to rule out fraud.
THANK YOU

The analysis simply says that the magnitude is not indexed by the exit poll discrepancy. It could (probably) be less. It could (conceivably, I suppose) be greater.
COULD BE MORE, COULD BE LESS.
FOR HEAVEN'S SAKE, FEBBLE, TELL US SOMETHING WE DON'T KNOW.

And the take-home message is: stop regarding the exit polls as prima facie evidence of a stolen election. There is plenty of other work to be done, and frankly, the exit polls are getting in the way.
NO, FEBBLE, THE EXIT POLLS WERE THE FIRST MECHANISM TO RAISE SERIOUS FLAGS.
SINCE THEN, WITH THE ANALYSIS OF THE NATIONAL EXIT POLL TIME LINE
(THE ONLY ONE MATCHED TO THE PURLOINED VOTE)
AND WITH PROOF OF THE IMPOSSIBILITY OF THE FINAL'S
BUSH/GORE 2000 WEIGHTS, THE EXIT POLLS HAVE BEEN CERTIFIED AS
THE GIFTS THAT ALWAYS KEEP ON GIVING.

OF COURSE, YOU WOULD LIKE ALL EXIT POLL DISCUSSION TO CEASE.
YOU WANT US TO "MOVE ON".
SO WOULD MITOFSKY, I BET.
DO YOU WANT TO TALK ABOUT THE PRE-ELECTION POLLS?
THAT'S FINE WITH ME.

And while I'm attuned to the other world, let me channel another of the dead:
YUP, THAT'S ME.
RUMORS OF MY DEATH HAVE BEEN GREATLY EXAGGERATED.
THOUGH BANNED FROM DU, I'M STILL KICKING, THANK YOU.
KICKING ENOUGH TO MAKE ALL YOU NAYSAYS COME OUT IN FORCE.



I believe the citation is from this document:

http://electionarchive.org/ucvAnalysis/US/exit-polls/ES...

In which the hypothesis I referred to is incorrectly stated.

See the second of the links above for clarification.

However, the channeled spirit correctly spots one of the fraud mechanisms that could account for the exit poll discrepancy and not produce a swing-shift correlation:
THANK YOU

If fraud in 2004 was precisely calibrated to the fraud that occurred in 2000 (and I mean precisely), then yes, that might do it.
NO THANKS
2004 FRAUD MUST BE BE PRECISELY CALIBRATED TO 2000 FRAUD?
SURELY YOU JEST.
THERE ARE INFINITE SCENARIO COMBINATIONS WHICH WILL YIELD
RESULTS SIMILAR TO THE MODEL WHICH YOU JUST REFEREED TO.

If every extra 1 percent in Bush's vote share due to fraud in 2000 was mirrored by X times 1% fraud in 2004, yup, maybe you could pull it off.
UMM, THANK YOU...
I GUESS.

An alternative would be uniform fraud in all precincts.
POSSIBLE.
NOT LIKELY.
WHY LEAVE A TRAIL?
BETTER TO TARGET INDIVIDUALLY.

Another alternative is fraud confined to precincts in which Bush was anticipated to do badly relative to 2000 (although you would have to be careful to cover a substantial majority of precincts and be sure to avoid any where he was doing well).
AGAIN, POSSIBLE,
BUT NOT LIKELY

THERE ARE MANY WAYS TO SKIN A CAT.
TOUCH SCREENS.
OPTI-SCANS.
PUNCH CARDS.
CENTRAL TABULATORS.
AND THE VOTERS WOULD NEVER KNOW IT.

BUT YOU NEED COOPERATION AT THE STATE AND/OR LOCAL LEVEL.
LIKE IN FLORIDA AND OHIO.

If anyone can suggest an practical algorithm that could achieve this, bearing in mind that it in all NEP precincts where the vote counts for are collected at the precinct, that the fraud must be perpetrated at precinct level, not at tabulator level (about 60% of precincts) while the remainder has to be done at tabulator level, then I will concede, yes, perhaps the exit poll discrepancy was due to fraud.

NOW WE MUST COME UP WITH THE THEFT ALGORITHM FOR YOU?
YOU ASSUME THERE WAS JUST ONE.
THE SMOKING ALOGORITHM...
SURELY, FEBBLE, YOU JEST.
FRAUD CAME IN MANY SHAPES AND SIZES.
IT HAD TO BE FLUID.
1. OHIO AND FLORIDA -ABSOLUTELY NECESSARY
2. PICK UP ENOUGH VOTES ELSEWHERE TO WIN BY AT LEAST 3MM VOTES
IN ORDER TO AVOID ANOTHER 2000 AND GET THAT "MANDATE"
3. SET THE TONE EARLY: THE EAST WAS THE BEAST.
MAKE SURE THE EARLY TOTALS SHOW BUSH AHEAD.
4. AND, MOST IMPORTANT, MAKE SURE THE INITIAL EXIT POLL TIMELINESS
ARE NOT RELEASED. ONLY SHOW THE FINAL AFTER MATCHING TO THE VOTE...

But remember - there is very little room for any deviation from the fraud perpetrated in 2000. It has to be directly proportional.
MUST BE DIRECTLY PROPORTIONAL?
SURELY, FEBBLE, YOU JEST.
WHY DO YOU EXPECT THE PERPS WOULD USE A SIMPLE FRAUD FORMULA?

Any variance and it will show up in a swing-shift correlation. I challenge my channeled spirit to produce a model of 1250 precincts (not 10) with a realistic distribution of WPEs, vote shares, counting methods and swing values in which the WPE (or any other measure of bias) is not correlated with a measure of swing, and yet fraud is responsible for both (I'm happy with 70% shared variance, or even less). Oh, and you can't end up with a large mean WPE in 2000 because there wasn't one.

LET'S BACK UP HERE.
I CHALLENGED THE NAYSAYERS TO COME UP WITH A SINGLE
PLAUSIBLE BUSH WIN SCENARIO.

THE ORIGINAL SCENARIO YOU CAME UP WITH IN THE DU "GAME" THREAD
IN WHICH I WAS BANNED (HOW DID THAT HAPPEN?) WAS ESSENTIALLY THIS:
1- 15% OF GORE VOTERS (3 MILLION) VOTED FOR BUSH
2- KERRY WINS JUST 52% OF NEW VOTERS.

NOT VERY PLAUSIBLE.

THE FINAL EXIT POLL HAD KERRY WINNING 54% OF NEW VOTERS (TOO LOW).
AND 10% OF GORE VOTERS VOTING FOR BUSH (TOO HIGH).
IN THE 1:25PM FINAL, THE WEIGHTS/PERCENTAGES WERE CHANGED
TO MATCH THE VOTE.

SO, OF COURSE, BUSH WON THE FINAL EXIT POLL. HE HAD TO.
BECAUSE THEY FORCED THE NUMBERS TO MAKE IT SO...

43/37, 43/37, 43/37, 43/37...
IMPOSSIBLE, IMPOSSIBLE, IMPOSSIBLE, IMPOSSIBLE

AND THAT WAS THE ONLY TIME LINE THAT HE WON.

AT THE 12:22AM TIME LINE, KERRY WON 91% OF GORE VOTERS.
AND 57% OF NEW VOTERS.

AND THE FINAL'S 200O VOTER TURNOUT
43% OF 2004 VOTERS FOR BUSH AND 37% FOR GORE?
- CLEARLY IMPOSSIBLE.
OOPS, I REPEAT MY SELF.

SO I ASKED THE NAYSAYERS TO COME UP WITH ANOTHER SCENARIO.
YOU ARE WELCOME TO DOWNLOAD MY EXCEL INTERACTIVE ELECTION MODEL.
IT'S FREE.

BUT NO ONE TOOK ME UP ON THE OFFER.
AND NOW YOU CHALLENGE ME TO DEVELOP A RIDICULOUS, IMPOSSIBLE ALGORITHM?
I THINK I'LL STICK WITH 43/37.
SINCE YOU ARE ALREADY STUCK WITH IT.

FROM HERE TO ETERNITY.

SINATRA AND LANCASTER.
TWO OF MY ALL-TIME FAVORITES, BTW.
NO BS FROM EITHER OF THEM.
HONEST. HARD WORKING. TALENTED. PROFESSIONAL.


THOSE WERE THE DAYS.

Even if the Final Exit Poll (43/37) says otherwise.

But, but....
You agree that 43/37 is impossible.
And that 43/37 implies

1) Bush 2000 voters were oversampled; in fact, they were anxious.

but, but...

2) Bush 2000 voters could not have comprised 43% of the 2004 vote.
That would mean Bush got 52.57mm votes.
We know he only got 50.45.
But 1.75mm of them died.
So at most 48.7mm Bush 2000 voters (39.8%) could have turned out in 2004.

Sorry, but...
There goes your rBr (reluctant Bush responder).
There goes your Gore Voter False Recall.

This makes no sense. It is perfectly possible for Kerry voters to have been sampled at a higher rate than Bush voters, and that hypothesis is simply not incompatible with the Gore-Bush recall data, given what we know about recall. People may choose to disbelieve either but there is nothing inherently impossible about either.


To quote Febble (my responses in UPPER case):
"If you read both pieces you will see that there are a number of ways in which the plot could be compatible with fraud, even with widespread fraud..."
THANK YOU

it is, however, very difficult to reconcile the plot with a fraud mechanism that could have contributed substantially to the exit poll discrepancy.
DEFINE SUBSTANTIALLY.
BUT WAS IT ENOUGH TO STEAL THE ELECTION?

If fraud was confined entirely to precincts in which Bush was doing badly, and scrupulously avoided in precincts where Bush was doing well, then I suppose, theoretically, about half. But that is theoretical. To do this would require a sophisticated algorithm which would have to applied at county tabulation level in about 40% of precincts and at machine level in about 60%. And it would have to devised in such a way as to undo any vote-flipping if a precinct turned out to be going too far Bush's way - or done on the final totals. No-one to my knowledge has yet proposed a workable algorithm to do this, but I would be interested to know if anyone has done.

It would therefore seem more likely than not that the exit poll discrepancy was caused largely by polling bias rather than fraud.
WHAT ABOUT THE PRE-ELECTION POLLING DISCREPANCIES?

I was watching every pre-election poll, and I was not expecting Kerry to win without a lot of assumptions going his way. You made those assumptions apparently, and predicted a Kerry win. I hoped for a Kerry win but predicted a Kerry loss, on the basis of the pre-election polls.

And the take-home message is: stop regarding the exit polls as prima facie evidence of a stolen election. There is plenty of other work to be done, and frankly, the exit polls are getting in the way.
NO, FEBBLE, THE EXIT POLLS WERE THE FIRST MECHANISM TO RAISE SERIOUS FLAGS.
SINCE THEN, WITH THE ANALYSIS OF THE NATIONAL EXIT POLL TIME LINE
(THE ONLY ONE MATCHED TO THE PURLOINED VOTE)
AND WITH PROOF OF THE IMPOSSIBILITY OF THE FINAL'S
BUSH/GORE 2000 WEIGHTS, THE EXIT POLLS HAVE BEEN CERTIFIED AS
THE GIFTS THAT ALWAYS KEEP ON GIVING.

Yes, the exit polls raised serious flags (though the first flag to me were the long lines in Ohio). That's why I got interested in them.

OF COURSE, YOU WOULD LIKE ALL EXIT POLL DISCUSSION TO CEASE.
YOU WANT US TO "MOVE ON".

Only because I don't think they support the fraud case. I think other evidence supports it far better. I actually think that what we now know about the exit poll data indicates that the discrepancy is unlikely to have been due to fraud. Fraud may nonetheless have occurred, particularly voter-suppression and vote rejection.

.

However, the channeled spirit correctly spots one of the fraud mechanisms that could account for the exit poll discrepancy and not produce a swing-shift correlation:
THANK YOU

On further reflection, I'm not sure that this is correct.

If fraud in 2004 was precisely calibrated to the fraud that occurred in 2000 (and I mean precisely), then yes, that might do it.
NO THANKS
2004 FRAUD MUST BE BE PRECISELY CALIBRATED TO 2000 FRAUD?
SURELY YOU JEST.
THERE ARE INFINITE SCENARIO COMBINATIONS WHICH WILL YIELD
RESULTS SIMILAR TO THE MODEL WHICH YOU JUST REFEREED TO.

Yes there are infinite combinations. What we need is the probability distribution of those combinations. If you can compute the probability distribution, I will listen.

If every extra 1 percent in Bush's vote share due to fraud in 2000 was mirrored by X times 1% fraud in 2004, yup, maybe you could pull it off.
UMM, THANK YOU...
I GUESS.

An alternative would be uniform fraud in all precincts.
POSSIBLE.
NOT LIKELY.
WHY LEAVE A TRAIL?
BETTER TO TARGET INDIVIDUALLY.

I agree. On reflection, I don't think the calibrate-to-2000 will work. My best shot so far is link vote-flipping to Bush's 2000 performance regardless of fraud in 2000, and only target precincts where Bush is doing badly. This is not as easy as it may sound.

Another alternative is fraud confined to precincts in which Bush was anticipated to do badly relative to 2000 (although you would have to be careful to cover a substantial majority of precincts and be sure to avoid any where he was doing well).
AGAIN, POSSIBLE,
BUT NOT LIKELY

No, I agree, not likely, which is why I think the non-correlation argues against widespread vote-flipping fraud. But YMMV.

THERE ARE MANY WAYS TO SKIN A CAT.
TOUCH SCREENS.
OPTI-SCANS.
PUNCH CARDS.
CENTRAL TABULATORS.
AND THE VOTERS WOULD NEVER KNOW IT.

No, but a statistician might if they found a swing-shift correlation.

BUT YOU NEED COOPERATION AT THE STATE AND/OR LOCAL LEVEL.
LIKE IN FLORIDA AND OHIO.

Yes.

If anyone can suggest an practical algorithm that could achieve this, bearing in mind that it in all NEP precincts where the vote counts for are collected at the precinct, that the fraud must be perpetrated at precinct level, not at tabulator level (about 60% of precincts) while the remainder has to be done at tabulator level, then I will concede, yes, perhaps the exit poll discrepancy was due to fraud.

NOW WE MUST COME UP WITH THE THEFT ALGORITHM FOR YOU?
YOU ASSUME THERE WAS JUST ONE.
THE SMOKING ALOGORITHM...
SURELY, FEBBLE, YOU JEST.
FRAUD CAME IN MANY SHAPES AND SIZES.
IT HAD TO BE FLUID.

I absolutely agree. The problem is, if there is variance in the methodology, and the methodology involves vote-flipping, then it will produce a swing-shift correlation. As long as it involves voter suppression, or vote suppression in precincts where there has always been vote suppression of the Democratic vote (eg provisional ballot rejection, undervotes, overvotes etc) then it won't show up as a swing-shift correlation. Nor will blanket spoilage in Dem precincts. And this may well have happened - there is lots of evidence that it did. However, most of this kind of fraud won't show up in the exit poll (differential spoilage/suppression of Dem ballots will, but not in a swing-shift correlation).

I agree - it had to be fluid. And I would argue that vote-flipping had to be a minor component, given the lack of swing-shift correlation. The form of fraud that would produce red-shift but not correlated swing-shift is differential spoilage. Other forms could swing the election, but not the exit poll.


But remember - there is very little room for any deviation from the fraud perpetrated in 2000. It has to be directly proportional.
MUST BE DIRECTLY PROPORTIONAL?
SURELY, FEBBLE, YOU JEST.
WHY DO YOU EXPECT THE PERPS WOULD USE A SIMPLE FRAUD FORMULA?

No. This is my point. Only if they used a simple formula could they avoid the swing-shift formula - or if the methodology was mainly voter/vote suppression.


LET'S BACK UP HERE.
I CHALLENGED THE NAYSAYERS TO COME UP WITH A SINGLE
PLAUSIBLE BUSH WIN SCENARIO.


SO I ASKED THE NAYSAYERS TO COME UP WITH ANOTHER SCENARIO.
YOU ARE WELCOME TO DOWNLOAD MY EXCEL INTERACTIVE ELECTION MODEL.
IT'S FREE.

BUT NO ONE TOOK ME UP ON THE OFFER.

No, because, I don't accept the assumptions underlying the model.

AND NOW YOU CHALLENGE ME TO DEVELOP A RIDICULOUS, IMPOSSIBLE ALGORITHM?

Take it or leave it.

This was left out of the response above. Sensei also said:

---------------------------------------------------------------

THE ORIGINAL SCENARIO YOU CAME UP WITH IN THE DU "GAME" THREAD
IN WHICH I WAS BANNED (HOW DID THAT HAPPEN?) WAS ESSENTIALLY THIS:
1- 15% OF GORE VOTERS (3 MILLION) VOTED FOR BUSH
2- KERRY WINS JUST 52% OF NEW VOTERS.

NOT VERY PLAUSIBLE.

THE FINAL EXIT POLL HAD KERRY WINNING 54% OF NEW VOTERS (TOO LOW).
AND 10% OF GORE VOTERS VOTING FOR BUSH (TOO HIGH).
IN THE 1:25PM FINAL, THE WEIGHTS/PERCENTAGES WERE CHANGED
TO MATCH THE VOTE.

SO, OF COURSE, BUSH WON THE FINAL EXIT POLL. HE HAD TO.
BECAUSE THEY FORCED THE NUMBERS TO MAKE IT SO...

43/37, 43/37, 43/37, 43/37...
IMPOSSIBLE, IMPOSSIBLE, IMPOSSIBLE, IMPOSSIBLE

AND THAT WAS THE ONLY TIME LINE THAT HE WON.

AT THE 12:22AM TIME LINE, KERRY WON 91% OF GORE VOTERS.
AND 57% OF NEW VOTERS.

AND THE FINAL'S 200O VOTER TURNOUT
43% OF 2004 VOTERS FOR BUSH AND 37% FOR GORE?
- CLEARLY IMPOSSIBLE.
OOPS, I REPEAT MY SELF.

but thought that pasting it in might be skating on thin ice, rule-wise.

It is perfectly true that the Gore-Bush2000 proportions are "impossible" when extrapolated to the population. I've already said this elsewhere. The question is: what are the possible explanations? One, of course is fraud. But using actual known "false report" percentages from longitudinal studies of 2000 voters, OTOH was able to reproduce the data in terms of false recall.

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=389901&mesg_id=390176
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=389901&mesg_id=390176

You, and others, may decide that you don't believe the "false report" explanation, but unless you deny the evidence that 9% of people who voted for Gore in 2000 later "recalled" voting for Bush, then the "impossible" proportions are perfectly explicable in terms of known patterns of recalled vote.

I am not even saying that this is what happened. But I would certainly dispute that the "43/37" mantra is anything like a slam dunk for fraud, no matter how many times anyone cares to type IMPOSSIBLE in all caps, and given other evidence (lack of swing-shift correlation for one) I find the false recall explanation at least plausible.

WHICH IS NOT TO SAY THAT THERE WAS NO FRAUD (Caps seem to be catching).

But the harder I stare at the exit poll evidence the less it appears to me that more than a fraction of the discrepancy can be attributed to vote switching fraud, and IMO puts strong constraints on the amount of vote-switching fraud that could have plausibly happened. But the exit poll evidence does NOT rule out (IMO):

Voter suppression, including rationing of voting machines in Democratic precincts
Differential spoilage of Democratic votes in precincts where it has always happened (also rejected provisional Democratic ballots)
Spoilage of votes in largely Democratic precincts
A limited amount of vote flipping (but it has to be limited to avoid producing the swing-shift correlation).

And frankly that's enough to make me pretty mad.

(Love the photo - Delphi was the first place I ever travelled out of England, and I was bowled over - I can still recall the taste of the cold spring water in the sunshine, under the olives. I was thirteen).

Line above which reads:
"1- 15% OF GORE VOTERS (3 MILLION) VOTED FOR BUSH"
should have read
"I= 15% OF GORE VOTERS (7.5 MILLION) VOTED FOR BUSH"
The Oracle apologizes to Sensei for his mistake. Will do 50 push ups.

I was there at 21 and got a message: "In 2004 you will be stunned and sickened; in 2005 you will awaken and see the light." What a a coincidence.

Let me add to your list:
"Voter suppression, including rationing of voting machines in Democratic precincts
Differential spoilage of Democratic votes in precincts where it has always happened (also rejected provisional Democratic ballots)
Spoilage of votes in largely Democratic precincts
A limited amount of vote flipping (but it has to be limited to avoid producing the swing-shift correlation)."
Padded Republican voter registration in select precincts (e.g. SW Ohio).
Padded Republican totals during tabulation in Ohio.
Padded Republican totals during tabulation in select localities (e.g. south Florida and heavily heavily Democratic states (e.g., CA & NY).

Tabulation in 2004 should not be left out as a huge potential source of fraud. Unlike voting machines in 2004, Diebold had deep penetration in the tabulation market. Who is Jeff Dean is not a rhetorical question.

Here is a point that cries out for simple reason.

The Columbus Dispatch runs polls which are highly respected and have been very accurate. The ran one just a few days prior to the special issues election in 11/05. Their predictions for the vote were so far off it's stunning. For issues 2 & 3, there's a completely unbelievable swing of 40-50% from the poll compared to the vote totals. This isn't a 3-5 point spread which is worthy enough of debate to involve a lot of bright people in the task. It's a clear case of either total incompetence on the part of the Dispatch OR election fraud. I think it was a Blackwell "throw down" to see if the Democrats would take it as a prelude to fixing the 2006 elections there and elsewhere. Of note 44 Ohio counties installed Diebold voting machines prior to this vote.

What do you say about that? Some apologize for this which I find amazing. What's more amazing is the total cave in. What the heck is going on here.

Here are the #'s:


Issue 2: Vote by mail.
Poll YES 59% | NO 33%
Rslt YES 36% | NO 63%

Issue 3: Limit campaign contributions.
Poll YES 61% | NO 25%
Rslt YES 33% | NO 67%


Boss Tweed and Secretary Blackwell: Fond admiration

"Padded Republican voter registration in select precincts (e.g. SW Ohio).
Padded Republican totals during tabulation in Ohio." -a.rank

Hi Auto,

Do you have any sources on this Republican padding? Conyers or Phillips, for example?

Would you like to take a stab at quantifying the padding? All fraud citations and attempts at quantification (to the nearest ten thousand, say) would be welcomed on the Election Fraud Analysis Evaluation thread:
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=399216&mesg_id=399216

My guess is that tabulator frauders are more likely to use vote-SWITCHING than padding, because vote-switching is less likely to be detected. Padding could be easily detected by checking the ballot totals (and signatures) in the poll books. Vote-switching could only be detected with a manual recount.

Cheers!

is more likely to show up in a swing-shift correlation - and there wasn't a significant swing-shift correlation in Ohio. However, there is probably still room for a bit of it in the statistical noise (which is why I've included a bit in my list).

Regarding the referendum in Ohio: I agree it it needs investigation. I've read a number of posts on DU that have argued that it is explicable in terms of the campaign, the timeline of the polls and the wording of the questions on the ballot (which apparently differed from the the question in the polls), and MysteryPollster has an extensive piece here:

http://www.mysterypollster.com/main/2005/11/columbus_dispat.html

However, given the culture of corruption (and laxity) in Ohio, I wouldn't want to rule out fraud either.

My bottom line is: if you have to rely on polls as the only check on the integrity of your electoral system then your democracy is in Deep Deep Trouble.

http://www.elftrance.com

As for MP, he's started New Years early. You're right to hedge, although your caution is not necessary: "However, given the culture of corruption (and laxity) in Ohio, I wouldn't want to rule out fraud either."

One of the most respected newspaper polls in America...NO WAY IT'S MAKES THAT TYPE OF HUGE ERROR.

Here is a little experiment for MP those blithely bold enough to defend this lunacy: Is it easier to hack several thousand pieces of mail randomly delivered to respondents and returned to the Dispatch or, well, you guessed it. Diebold-Blackwell Uber Allis.

This is an extremely serious event. I do not appreciate MP or anyone else that tries to put lipstick on a pig, or better yet, a smiley face on the monster that has become the Ohio electoral system. This is their warm up, just like Poland was a warm up. They're out to take it all again in 2006 and they don't care how ridiculous it looks.

The arbitrary ban on the election fraud story, in any of its iterations carries a price measured on a global scale.

when he was eight, wrote a poem with the superb title:

"Don't let your caution waver" - and I don't. It's part of being a scientist. Pollsters are scientists too.

So I will defend Mark Blumenthal, whom I count as a friend, against the charge of putting lipstick on pig. What scientists have to do is to check all livestock for authenticity, regardless of the cosmetics used. I would agree that the discrepancy between poll and count in Ohio 2005 needs investigating. However, investigating any discrepancy between two findings involves the investigation of both findings. Blumenthal is a pollster. He's done a good analysis on the half of the problem he knows about, and he makes good points, as do the posters in this thread:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=401315&mesg_id=401509

But just because it is perfectly possible to propose plausible hypotheses as to why the polls might have been wrong doesn't mean it is not equally possible to hypothesise why the count might have been, as you have done. Both hypotheses need to be tested.

All I would argue is that it can be counter-productive to regard a poll-count discrepancy as prima-facie evidence of fraud when there are so many ways in which polls can give you wrong answers. All social scientists know this, which is why failing to take into account perfectly valid arguments as to why a poll might have given a discrepant result risks alienating those who might (as I do) argue that the fraud charge should also be seriously investigated.

As well, of course, as arguing the case that no electoral system in a democracy should allow any room for doubt about the result (ours doesn't, which is why we know that polls are often wrong).

Which brings me back to my bottom line (above, if a bottom can be above....)

I'll respond to this later after some thought.

(although on checking, he must have been nine)



Kinda applicable, don't you think?

I think it's unique unto itself:)

by email today. You might like to check them out if you haven't already:

http://www.freep.com/apps/pbcs.dll/article?AID=/20051124/NEWS01/511240355/1003/NEWS

http://toledoblade.com/apps/pbcs.dll/article?AID=/20051120/NEWS09/511200359

Click here if you would like to know identity of my correspondent.

I now know of two counties where it is quite unlikely that there was tabulator fraud: Hamilton (punch card) and Lake (DRE). In the Greens' 2004 3% manual recounts, both counties appear to have had transparent recount procedures and acceptable precinct-selection processes. (Hamilton had a virtually-random selection process. In Lake county, officials let the recount-observers select the precincts.) Of course both counties probably had other types of fraud, such as voter suppression, and perhaps even card-punching/DRE fraud, as well.

But, even though Ohio state-wide tabulator fraud was unlikely, it is quite possible that SOME Ohio counties had tabulator fraud. If so, I would imagine that the perpetrator(s) would try to choose a method (e.g., vote switching) that would be least likely to be detected - and use it in counties where manual recounts could be expected to be a shambles (e.g., Cuyahoga).


Regarding the lack of a swing-shift correlation in Ohio, let's just hypothesize that the vote-switching was not done for NEP precincts.

...like Frank Sinatra's "retirement tour" (there were several, each a success!). I respond on the Ohio phenomenon at the post specifically addressing that.

btw, I, personally, would very much like to see your response to the election model opportunity. I'm tempted to use the Cadrinals and Galileo analogy but even I admit, that would not be fair (although despite the ambivalent attitute of The Church on science, it has recently "walked tall" in rejecting "intelligent (sic) design";)

-----------------------
TIA:
SO I ASKED THE NAYSAYERS TO COME UP WITH ANOTHER SCENARIO.
YOU ARE WELCOME TO DOWNLOAD MY EXCEL INTERACTIVE ELECTION MODEL.
IT'S FREE.
BUT NO ONE TOOK ME UP ON THE OFFER.

Febble:
No, because, I don't accept the assumptions underlying the model.


-----------------------------------------------------------------------

You didn't think I'd let you off so quickly with THAT one, did you?
I ask you: what assumptions?
Be specific, please.

IS IT THE 18 PRE-ELECTION NATIONAL POLLS?
I make only ONE assumption: the undecided vote.
Do you have a problem with that?

Check out the National Exit Poll's "When Voted" demographic.
It has Kerry winning 60% of the late undecided.
Go ahead. Run the model.
Try whatever assumptions you wish.

IS IT THE 50 PRE-ELECTION STATE POLLS?
I make only ONE assumption: the undecided vote.
Do you have a problem with that?

IS IT THE 50 STATE EXIT POLLS?
There is just one input assumption: the "cluster" effect.
This only effects the probability of exceeding the MoE.
Do you take issue with that?
Then go ahead. Run the model.
Try whatever assumptions you wish.

IS IT THE NATIONAL EXIT POLL?
There is just one input assumption: the "cluster" effect.
This only effects the probability of exceeding the MoE.
Do you take issue with that?
Then go ahead. Run the model.
Try whatever assumptions you wish.

I use the 12:22am timeline of state and national exit poll data.
Do you have a problem with that?
Or would you prefer the 1:25pm Final National,
which was matched to the vote count?
Fine but there is ONE restriction which you MUST adhere to:

For Bush/Gore weights in the "How Voted in 2000" demographic:
THE ABSOLUTE MAXIMUM BUSH SHARE OF 2004 VOTERS IS 39.8%.
That's not my restriction; it's the Almighty's.
Bush 2000 voters who died could not have voted in 2004.
Would you stipulate to that?

Ok,
Ready.
Set.
Go.

Assume the pose.
Run the model.
Play what-if with any of the demographic categories.

THIS IS THE CHALLENGE:
COME UP WITH ONE PLAUSIBLE SCEANARIO OF A BUSH WIN

Consider real numbers, not non-response.
Consider feasibility, not wishful thinking.
Consider actual scenarios, not abstract generalities.
Consider probabilities, not intelligent design.

Is that fair enough?

------------------------

as per #96:

random sampling.

But to elaborate, we are asked to:

"Consider real numbers, not non-response."

If you assume random sampling (i.e. no non-response bias) then I am sure the model will do what its author says it will do. The only non-fraud explanation for the discrepancy between polls and count is some kind of bias in the poll. This is sometimes called "non-response bias" although that needs also to include "selection bias" - i.e. bias in the selection of respondents. "Non-response bias" strictly means that those who refused were different in some way from those who agreed to participate. "Selection bias" can occur if there is bias in the selection process itself. Both can happen all too easily in any survey.

If we assume that there was neither non-response bias nor sampling bias, then the model will, I am sure, reproducably give a Kerry win. I would certainly expect it to. But that is the assumption I do not share. Social scientists never do, as selection bias and non-response bias are a pretty intractable methodological problems, and, unlike "sampling error" (on which the MoE is calculated) do not reduce as sample size increases. They are also not addressed by increasing your variance estimate to allow for clustered sampling (e.g. using the DESR - Design Effect Square Root. They have nothing to do with clustered sampling, which will introduce greater variance, but should not introduce bias. The sad fact is that we cannot mathematically compute selection bias, nor non-response bias.

The only way of investigating whether what caused a poll-count discrepancy was bias in the poll or bias in the count is some kind of correlational analysis: is the bias greater in circumstances where, a priori*, you would expect it, if fraud was the factor? Or the bias greater in circumstances where, a priori, you would expect polling bias to be greater?

But we can no more investigate polling bias by ruling it out, than we can investigate fraud by ruling it out, and it is the former that our undead friend asks us to do.



*see, I can do Latin when I try!

Here's the reference (one of many) on Ohio precincts. It carries weight with me, particularly in light of the ongoing problems in Ohio (special election, 2nd District, 8/05 producing, as if by magic, Rep. Jean Schmidt, R). http://web.northnet.org/minstrel/southwest.htm

With regard to tabulator concerns, my response was to febble's statement about exit polls, " But the exit poll evidence does NOT rule out (IMO):" Tabulator manipulation is one that you may have ruled out, which is fine with me, but is appropriate to the list febble generated.

Now here's my question to you, given:

a) the Dispatch poll versus the results just on issues 2-3;
b) the esteemed nature of the Dispatch poll and it's outstanding track record in 30+ Ohio races see first reference in my post above;
c) the placement of new Diebolds in 44 counties just before the election (and God knows how many precincts); and
d) the highly questionable track record of the Ohio state elections authorities

how do you explain the huge discrepancy between the Dispatch poll results and the vote totals reported?

This is the best thing you can ad to that board:

Issue 2: Vote by mail.
Poll YES 59% | NO 33%
Rslt YES 36% | NO 63%

Issue 3: Limit campaign contributions.
Poll YES 61% | NO 25%
Rslt YES 33% | NO 67%

Somebody needs to explain that.

Take care.

See his SW Ohio (Butler, Warren and Clermont) estimate at http://www.freepress.org/images/departments/Vote_Count_Ohio.pdf

As for the Dispatch poll versus the results on issues 2-3, DUers who currently live in Ohio would have more insight than I. I have found the comments by Ms. Toad persuasive:

"It was OH voters, but not those who lean Republican - voters who lean in a lot of different ways killed issues 2-5. More accurately, the amendments committed suicide. Read here: http://www.reformohionow.org/downloads/ron_amendments.p... . Pages 1-8 were actually on the ballot as summaries of the amendments, which can be found on pages 9-19. I voted against 2 of them, and more than half of my liberal friends voted against all 4 (Most switching from leaning Yes to voting No upon reading the actual text of the amendments or a summary for the first time in the 24 hours preceding the election." - Ms. Toad
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=401315&mesg_id=401486


However, I do agree that the new Diebolds, in 44 of the 88 Ohio counties, are a very big worry. Do you, happen to know whether or not the new machines have voter-verified paper receipts?

Cheers.

Don't know about the receipts or whether or not they're retained.

With regard to how the issues lost, I'd say this. First, people go into vote on issues, amendments, bond issues with an opinion in mind when they issues mean something. Issue 2 "No" grew 30%. Issue 3 "No" grew 42% just 9-10 days before the election. Those are stunning shifts.

Lets say there is an election and candidate "Hell Yes" is running against candidate, "No Way." Hell Yes is leading No Way by 42%. All of a sudden, three days before the election it's revealed that Hell Yes is married to his first cousin and owns a topless bar. Then I could imagine a 42% shift, no problem. Saying that the campaign was less than it should have been or that people actually read an 8 page summary on the ballot (I need a link to believe that) is simply not enough. Many in Ohio are furious about 2004, you will agree. Taft is below 20% (or was) and the rest are faring poorly. The initiatives were a clear choice.

I can't believe this at all. I do accept that there are instant post-election recriminations by supports when a cause seemingly loses. That normal, par for the course reaction does not explain the loss any more than my opinion that the Raiders where cheated by the refs every time they played Pittsburgh explains those loses.

The poll is excellent evidence. It has a paper trail that can be reviewed.

It will be interesting to see how the Dispatch responds and also how the academics who praised the Dispatch's accuracy in over 30 races, most accurate polling around they say, respond.

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=401315

But I would nonetheless be highly suspicious of the count. Those machines are going to make errors whether deliberate or not, and if they make errors anyway, it is going to be highly tempting to make them deliberate - because you have built-in plausible deniability.

Voting on machines is SO stupid, it leaves me speechless.

There are opportunities to cheat with paper but they become extremely difficult with observers and a motivated public. With regard to errors, the CA Scty of State low balled the Diebold test results as I recall at about 10% when they were around 19%. 1 out of 10 or 1 out of 5 errors, what a debate.

When a reporter asked the famous bank robber Willie Sutton why he robbed banks, his response was, "Because that's where the money is." Imagine how many banks he would have robbed if the security was 5-10 years behind the time and easy toi beat. It is crazy.

I agree you can cheat with paper too. You need transparency and scrutiny at all points from the casting of ballots to the count and recount.

The weak point of our system is, ironically, the postal ballot. We had our first hint of serious corruption this year, in which, for the first time, offered postal voting on demand.

http://news.bbc.co.uk/1/hi/england/west_midlands/4406575.stm

But so far that's the only real vulnerability, although occasionally sealed ballot boxes go missing between the polling station and the count. Each constituency has a number of polling stations and a central counting place (usually a school or town hall). The votes are cast at the polling system (paper ballots posted into black metal ballot boxes, which are sealed at close of polls) then the numbered ballot boxes are transported to the count. From that moment onwards everything is under close bipartisan scrutiny, and the count is open to the public who can view from a slight distance (usually behind a rope barrier). There are also TV cameras at every count.

But getting the ballots on to paper is a good first step!




Waiting for the ballot boxes.



Ballot boxes being emptied at the count.



Counting the ballots



Counted ballots stacked in piles



Result is announced (notice candidates for each party are present).

And if there is any dispute, the result is not announced, and the whole lot are counted over again.


http://images.google.co.uk/imgres?imgurl=&imgrefurl=http://www.bbc.co.uk/wales/mid/sites/in_pictures/pages/election2005_ceredigion.shtml&h=255&w=340&sz=14&tbnid=82WFgU7Ez68J:&tbnh=86&tbnw=115&hl=en&start=7&prev=/images%3Fq%3Dcount%2Bconstituency%26svnum%3D10%26hl%3Den%26lr%3D%26sa%3DG

(Edited to make sense!)

random sampling.

If you question 1000 people randomly chosen, you can predict the actions of millions within 1/2 of a percent.

TIA isn't the only statistician that questioned the last election. A number of univeristy professors have made essentially the same arguments.

"randomly".

There is no guarantee that the people chosen in a poll are random unless you get a 100% response rate and your selection process is designed to ensure absolutely uniform coverage and random selection. Neither of these apply to either the pre-election polls or the exit polls, and there is strong evidence to suggest that the exit polls at least were subject to systematic non-participation bias.

there has been sufficient debate throughout the last year that well demonstrated TIA's ignorance of statistics. I would point out that the primary problem is that TIA confounds the heuristic model of statistical inference with the actual patterns encountered by the experts, thus that his models lack sufficient error range to prevent acceptance of hypotheses when data fails to support the assertion.

What Febble describes is what occurs when the sampling methodology can go south, or when an unchecked bias enters into the sampling--it is the art of statistics that is never discussed in that intro to stats taken in many of our sophomore years; and is really never touched upon until grad. school--coursework not part of TIA's CV.

Mike

"A poll figures out how the 100 million people will vote by questioning 500 or 1000 of them."

Except for "The" Exit poll TIA refers to, polled um-- ah --- I think- it -- was


about 13 thousand people

"They guess that equal numbers of Repubs and Dems will vote"

Wrong check something called weighting


I think its time to get a search warrant - - - - - - -

I think what is being referred to are the pre election polls, not the exit polls. A pre-election prediction (kerry landslide) performed by TIA combined several pre election polls (pre-selected to include only those favorable to TIA's POV), and I suspect this is what's being referred to. The equal D/R proportions is an issue taken with the Gallup Poll, there were several times where R's were oversampled.

Mike

So, what do you think Diebold's and ES&S's 'TRADE SECRET,' PROPRIETARY programming code is based on? Hm-m?

Lay it out for for me. Show me how they tabulate our votes. Explain the code. Give me a url where I can review it, or have a programmer review it. EXPLAIN TO ME HOW THEY COUNT ARE VOTES!

I'm sorry, but you DON'T NEED EVIDENCE when a system is INHERENTLY FRAUDULENT and is designed TO HIDE THE EVIDENCE.

We had a fraudulent election system going in. And...

AND!

...all external evidence points to a Bush loss. And...

AND!

...numerous internal anomalies--such as nearly a hundred reports of touchscreens changing Kerry votes to Bush votes, with almost none going the other way; or, studies of paper vs. electronic voting showing tens of thousands of phantom votes for Bush in electronic voting--also pointing to a Bush loss. And...

AND!

...among all the external evidence pointing to a Bush loss, all polls--not just some polls--ALL polls point to a Bush loss.

Consider the scientific principle of cumulative effect. Cut down one tree. The birds and fish can still make it. The streams are still clear. Cut down ten trees. A few birds and fish fall by the wayside. The streams run brown for a short period. Cut down a hundred trees. Some birds and fish stop reproducing. The streams get browner for longer period. Clear-cut the hillside, and spray it with garlon. Those local fish, if they are coho salmon (and therefore are genetically tied to only one stream and can spawn nowhere else) disappear from the face of the earth; sensitive birds become stressed, can't find habitat, and some die without laying eggs, or their young die for lack of dispersal habitat. Further, frogs start being born with three legs or no eyes; they can't reproduce. Their predators starve. The forest itself, now a near desert, pours mud into the stream with every rainstorm. Young trees fry in the hot sun. Invasive species of plants take over some of the ground. A forest that took a thousand years to become tall and green and wet and full of life never grows back. Its scrawny, sparsely placed trees never again can adequately protect nesting birds. And the combination of sun, heat, pesticides, muddy waters, and loss of species, known and unknown, destroys that once beautiful ecosystem forever.

It happened in Greece. It happened in North Africa. It happened in China. Sometimes cumulative effects are definitive and final, and the original conditions cannot be restored.

Industry foresters will deny, deny, deny that cumulative effects are real. They will say, "Show me one dead fish." They will ignore the evidence of their own eyes. They will say one tree, a hundred trees, what does it matter? It'll all grow back. They will call the old fisherman a liar who says that, when he was a boy, the stream was teeming with thousands of fish, so thick you could walk across the stream on their backs. They will say that's just anecdotal. They will say the hot sun on your face--where cool forest should be--and the nausea you feel from pesticides, and the feeling of death all around you, are anecdotal. They will require you to raise $100,000 for a scientific study; then they will get their lawyers to deny that evidence in court.

One poll. Two polls. Three polls. Four polls. Five polls and more. All saying the same thing. New registration figures saying the same thing. Polls of new voters, independent voters, and former Nader voters, all saying the same thing. Pre-polls. Post-polls. Polls leading up to the election. Poll today. All pointing to a Bush loss. The issue polls all pointing to huge disapproval of every major Bush policy, foreign and domestic, way up in the 60% to 70% range, over the last two years. Anecdotes of enthusiasm among voters to get rid of Bush. One, two, three, four, five--many anecdotes, all saying the same thing: My brother who never votes; my Republican grandfather who voted for Bush in '00; my co-worker who let her registration lapse--all inspired to vote THIS TIME, because "this is the most important election in our history", for Kerry, because "I can't stand Bush any more"--or dragged to the polls by Gore voters who are still angry about '00. The spontaneous enthusiasm at Kerry-Edwards rallies. The highly vetted, bused-in dullness of Bush-Cheney events.

The war that 58% of the American people opposed before it ever started. The torture photos revealed--such horror!--six months before the election, with 63% of the American people opposing torture "under any circumstances." The failure to find WMDs. The outing of a CIA agent and her entire WMD counter-proliferation project--an act of treason--for political revenge. The secrecy--on everything from budgets to Halliburton's no bid contracts, to 9/11. The massive loss of jobs. Bush's utterly dismal performance in the debates--an idiot, an embarrassment. Continued carnage in Iraq.

Where does the cumulative effect begin, and get acknowledged? Does this not ADD UP to an overwhelming picture, some of it detailed and some of sketched in, but, cumulatively compelling, of a wrong outcome in the 2004 election?

And when you combine it with the secret, proprietary programming code, owned and controlled by Bushite corporations, and the lack of a paper trail, and the Republicans fighting a paper trail in Congress, and the voting machine companies lobbying against it everywhere--egregious and deliberate non-transparency--it seems to me deliberate know-nothingism, deliberate blindness, of the kind that the industry forester exhibits, to NOT SEE THIS REALITY.

"Show me one dead fish." Well, what I can show you is that no fish come up here any more, so, I can't show a dead example of something that no longer exists.

"To prove it you need evidence." Ask Diebold and ES&S.

--------

People who demand proof and evidence, in this instance, fail to grasp that evidence is the very thing that the Bushite Republicans and their electronic voting companies, with their secret, proprietary programming code, sought to deny us; and that the denial of that evidence IS evidence. AND, everything else says Bush lost.

The Diebold and ES&S vote tabulation result, which the war profiteering corporate news monopolies "matched" their exit polls to--thus disguising and hiding some of the evidence of election fraud--is the ONLY evidence that Bush won. The secret vote count of his buds at these companies is the only "proof" THEY have. Everything else--including the real exit poll results--says he lost.

Why should anyone presume that Bushite corporations, counting the votes with secret formulae, would produce a correct result? Their very secrecy cries out against them. And why make--or retain--this presumption, when all other evidence says the opposite?

Trends MEAN SOMETHING. The cumulative sweep of an aggregate of facts MEANS SOMETHING. All polls, and all other evidence, trends to Kerry. AND, they counted the votes behind a veil of secrecy.

The scientific mode of thought tends to take things apart, and dissect them down to their smallest constituent parts. We can do that to a frog, for instance. But we can't MAKE a frog, nor understand it as a whole creature, nor understand it IN ITS NATURAL ENVIRONMENT, without some fuzzy, intuitive thinking.

You have got to get out of the "proof and evidence" mode, for a moment, and ask yourself: What the hell are you looking at? Are you looking at a tongue? Are you looking at an eyeball? Or are you looking at a whole thing, a living creature, with a face, with being, and with intricate relationship to other living things, and to air, and water, and rocks?

Now try to "see" that frog, and determine what it is, while it hides under a black veil. Come on, you can do it. Something's hopping around under there. Something's croaking.

It's not a cat. It's not a tarantula. It's not parrot. It's not a rock. It's a....

....stolen election!

Who is "us"?

-------------------

"...and that't the reason no investigative journalist is paying attention to such numbers."--jobycom

You determined this scientifically, did you? How many journalists did you ask, as to why they aren't paying attention to such numbers? And, how many journalists are there in the world?

You're just guessing. I could put forward an equally--and, actually, much better--hypothesis, and that is that journalists are paid good money by war profiteering corporate news monopolies to ignore all evidence that Bush lost, and especially to ignore the evidence commissioned by the war profiteering corporate news monopolies, their own exit polls, which they fiddled on election night to "fit" the results of Diebold's and ES&S's secret formulae.

Anyway, considering what Judith Miller, Bob Novak and Bob Woodward have done to this country, with their "journalism," I wouldn't be inclined to give a crap what corporate journalists did or did not pay attention to. Is "journalist" some sort of guarantee of the integrity of information, or the importance of a subject for investigation, in your view? Who, among corporate news journalists, would you name, as particularly reliable? I can't think of a single one of them that I trust any more. Tens of thousands of innocent people were slaughtered because of them. Our democracy lay in ruins because of them. They have failed us, catastrophically. And you are putting THEM up here as some sort of measure of the accuracy and aptness of information?

As Patrick Fitzgerald said at his press conference, when asked about the Republican "talking point" that his indictment of Scooter Libby was a trivial matter: "That talking point won't fly!"

:kick:

:kick::kick:

:kick::kick::kick:

:kick::kick::kick::kick:

:kick::kick::kick::kick::kick:

:kick::kick::kick::kick::kick::kick:

:dem: back at you!:dem:

between the exit poll estimate and the final count is in the region of 0.95, then you will get a similar result if you correlate the final count with Bush's current approval ratings.

This one really is a doozy.

Polling data sources:
National Pre-election polls
http://www.pollingreport.com/wh04gen.htm
http://www.economist.com/media/pdf/YouGovS.pdf

Ohio exit poll
http://www.exitpollz.org/cnn2004epolls/Pres_epolls/OH_P.html

National Exit Poll (pdf)
11/2/04, 3:59pm 8349 respondents: Kerry 51-Bush 48
http://www.exitpollz.org/mitof4zone/US2004G_3737_PRES04_NONE_H_Data.pdf

11/2/04, 7:33pm 11027 respondents: Kerry 51-Bush 48
http://www.exitpollz.org/mitof4zone/US2004G_3798_PRES04_NONE_H_Data.pdf

11/3/04, 12:22am 13047 respondents: Kerry 51-Bush 48
‘
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=265121

11/3/04, 1:25pm 13660 respondents: Kerry 48-Bush 51
http://www.exitpollz.org/mitof4zone/US2004G_3970_PRES04_NONE_H_Data.pdf

POLL SAMPLE-SIZE AND MARGIN OF ERROR
The Law of Large Numbers is the basis for statistical sampling.
All things being equal, polling accuracy is directly related to sample size. The larger the sample, the smaller the margin of error (MoE).

The MoE for a random sample of size N is .98/sqrt(N)

In an unbiased random sample at the 95% confidence level,
there is a 95% probability that the true population mean
falls within the MoE (or within 1.96 standard deviations) of the sample mean.

In pre-election state polls, about 600 were sampled in each state.
That gives a 4.0% MoE for any given state. That may seem high, but...
the combined national sample size is 30,000 (50 states * 600).
So the combined MoE is equal to .98/sqrt(30000) or 0.56%.

In 18 pre-election national polls, the sample-size ranged from 800 (3.5% MoE) to 3500 (1.7%).
The total 27,229 sample size reduces the combined MoE to 0.59%.

The post-election state exit polls sampled 73,607 nation-wide,
with respondents ranging from 600 (4% MoE) to 2800 (1.8%).
The total 73,607 sample-size gives an MoE of just 0.37%.
Assuming (for the naysayers) a very generous 40% cluster effect,
the combined MoE is 1.4*(0.37%) or 0.52%

In the National Exit Poll of 13047 respondents, the MoE is 0.88%.
Once again, assuming a 40% cluster effect, the MoE is 1.4*.88 or 1.2%.
Kerry won the exit poll (gender demographic) by 50.8% - 47.8%.

Therefore, there is a 95% probability that Kerry won BETWEEN 49.6% and 52.0%
of the true vote. And 97.5% that his vote EXCEEDED 49.6%.

According to the 1:25pm Final Exit poll (13660 respondents),
WHICH WAS MATCHED TO THE VOTE, Bush won by 51.1 - 47.9%.

that TIA found out what the "cluster effect" really is, as he seems to regard it as some kind of peanut butter you spread on an MoE to make it a bit bigger.

It has to be calculated quite precisely, depending on the kind of "clustering" is being compensated for, but it is, in essence, a way of compensating for the underestimate of variance you will get if, instead of giving all voters an equal chance of participating in a poll, you first of all select a number of precincts, and then select voters from within those precincts. Once you have selected your precincts, some voters will have no chance of being selected for the poll, whereas voters on the selected precincts may have a 1 in 10 chance, or higher. Moreover, two voters from the same precinct will tend to have more in common than two voters from different precincts. Thus "clustered" sampling will tend to result in less variance than a true random sample. You therefore need to increase your variance estimate to compensate, by a formula known as the DESR (Design Effect Square Root) which is the square root of the ratio of the variance of the estimate to the variance obtained by assuming the sample is a simple random sample.

More information here:

http://www.spss.com/complex_samples/data_analysis.htm

It's not peanut butter.

And it is also high time that TIA learned that polls have two broad category of error: Sampling error and non-sampling error. All his statistical inferences completely ignore non-sampling error.

Non-sampling error is a huge problem in any survey, precisely because it cannot be easily calculated. For TIA to continue to propagate analyses that blithely state that:

"All things being equal, polling accuracy is directly related to sample size. The larger the sample, the smaller the margin of error" (MoE)."

is, frankly, no longer excusable. All things are NOT equal, by any stretch of the imagination, and TIA knows this (or if he doesn't he needs to learn it). Ignoring all error except for sampling error may work in casinos, but that's about it. It doesn't even work in engineering. It certainly doesn't work in surveys.

in 1992, for example - we thought we'd got rid of the Tories, and we hadn't.

http://politics.guardian.co.uk/electionspast/story/0,15867,1450621,00.html

But the difference is that we can pretty well trust our count, so we know it's the polls that are wrong.

And so we know polls can be wrong.

it should always be the polls that are off, but in this country they don't know if it was the polls or the count , because the count is done in secret by corporate America. So Mitofsky is the fall guy even though he more than likely had the numbers right, he has to discredit himself due to the corporate vote counting machines.

Can you imagine being in Mitofsky's shoes and someone says to you, Hey Mitofsky your doing a great job, then you have to turn and say "no I'm not" ??

If Mitofsky wants to do exit polls in this country after all the election theft machines are in place he is gonna need a

Surprised about the ones regarding education. Usually the more education voters have, the more likely they are to vote progressive.

As one TIA's earliest 'litmus' naysayers still on this board, I've been basking in the smug satisfaction that what I suggested long ago as to how the election fraudently fell to Bush is the only viable hypothesis still standing (sorry, landshark).

The weakness here is that TIA confuses his fruit when he needs to use apples. Each estimate is its own, has its own methodology and problems, and stands unique in comparison to another estimate. You cannot combine and weight them to obtain an inference that Bush stole the election, but at best are descriptive, and require further evaluation and analysis, something TIA is loathe to do, since once one undertakes this step, the ignorance of what is good or bad sampling design, of what may be unintended bias can make the case that yes, these polls have sufficient error to where one cannot discriminate whether Kerry won or not.

Secondly, these are not all the polls taken, but those selected by TIA for their utility in supporting his preconcieved notion of who would win. As a pre-election poll junkie, my favorite site was http://www.electoral-vote.com/; take a look at the map (and commentaries!) just prior to the election:http://www.electoral-vote.com/2004/nov/nov02.html. That should refresh some memories, and hopefully not re-open any old wounds.

Mike

:hi:

I hope there is justice and this fraud is exposed and the perpetrators are brought to justice.

From the bogus notices in Baltimore that the election day had been changed (unforgivable!) to the Diebold nonsense to the other attempts to limit voting, it will all be addressed.

It's our time. 35% and dropping.

HOW LONG CAN THEY KEEP THE SECRET?

since Bush is so fond of bald guys

It's long past the time with the elections coming up soon.

Md's Governor pushes for ESS & Diebold! Anyone wondering why? Of course not. The Sunpaper won't give him a pass, hopefully a second time 'round.

I admire his courage as well as his statistical abilities. It is all to easy for some people to take pot shots at someone who is not here to defend himself.
Edited for dyslexia.

and Happy Thanksgiving to all!

for Kerry. If they know they're wrong, why don't they DO something about it?

(Psychic hotline activated, message from the Truth Is All I get)

The Law of Large Numbers.
The Central Limit theorem.

Final Pre-election polls.
50 state
18 national

no rBr
no faulty recall
no poll workers
no long lines
no Mitofsky
no cluster

but ...
there are many pollsters
all scientific
some corporate (CBS, Gallup, FOX, ABC, NBC, AP, etc.)
some independent (Zogby, Harris, Economist, etc.)

Who are you to say their efforts are for naught?
After all, Mitofsky's exit polls confirm the pre-election pollsters.
And Mitofsky knows how to poll.
Doesn't he?
He's been doing it for 30 years.
All over the world.
Makes a good living, too.
He can afford to pay lots of consultants.

Why don't you contact these pollsters and tell them that,
based on your massive evidence of non-response bias,
they're no longer relevant?

That their services are no longer necessary.
Just fire them all.
And that includes Mitofsky.
That's what the Republicans want, don't they?

No more exit polls.
They just mislead voters into believing the elections are fixed.
We can't have that, can we?

So tell E-M and all the rest to exit the business.
What good are they?

No more damn exit polls.
Because what they're doing is just an exercise in futility.
They're polling the wrong people.
They should be polling the non-responders.
And even then, naysayers like you would disagree.

Whose idea was it to do a poll first, anyway?

Go ahead.
Set them straight.
You're the expert.
You have many years of experience in the field.
You know all about U.S. elections from overseas.
You know a lot more about polling than
Harris or Zogby or ARG or Gallup or SUSA or...

They should all just listen to you from now on.
And get out of the business.

Apparently, the Law of Large Numbers no longer applies.
It's the pollsters' bedrock.
Without it, they're out of business.
And that applies to all other statistical researchers.

Why don't we listen to Bush?
Let's replace Probability and Statistics with...
Intelligent Design 101.


FINAL NATIONAL PRE-ELECTION POLLS

PRESS F9 TO SIMULATE 200 ELECTION TRIALS

Undecided vote
Total Poll Total Weighted Average 67% 33%
Sample Sample MoE KERRY BUSH KERRY BUSH
Date 26961 Group 0.60% 47.55 47.30 50.43 48.57

1-Nov Marist 1166 LV 2.87% 50 49 48.97 50.03
1-Nov Econom 2903 RV 1.82% 50 47 50.77 48.23
1-Nov TIPP 1284 LV 2.73% 44 47 50.66 48.34
1-Nov CBS 1125 RV 2.92% 47 48 52.18 46.82
1-Nov Harris 1509 LV 2.52% 48 49 49.17 49.83

31-Oct Zogby 1200 LV 2.83% 47 48 48.48 50.52
31-Oct FOX 1400 RV 2.62% 48 45 52.44 46.56
31-Oct DemCorp 1018 LV 3.07% 48 47 49.65 49.35
31-Oct Gallup 1866 RV 2.27% 48 46 51.05 47.95
31-Oct NBC 1014 LV 3.08% 47 48 53.34 45.66

31-Oct ABC 3511 RV 1.65% 47 48 49.27 49.73
30-Oct ARG 1258 LV 2.76% 49 48 50.85 48.15
30-Oct Pew 2408 RV 2.00% 46 45 51.56 47.44
29-Oct News 1005 RV 3.09% 44 48 47.27 51.73
26-Oct ICR 817 RV 3.43% 48 48 50.87 48.13

24-Oct LAT 1698 RV 2.38% 48 47 50.76 48.24
21-Oct Time 803 LV 3.46% 46 51 46.27 52.73
20-Oct AP 976 LV 3.14% 49 46 53.48 45.52


18 Poll Summary:
Kerry won 9, Bush 8, 1 tie
Kerry won 5 of 9 Registered Voter (RV) Polls
and 4 of 9 Likely Voter (LV) Polls

Probability of Vote Deviation from poll:
Kerry:50.43% to 48.31%: 1 in 502,718mm
Bush: 48.57% to 50.77%: 1 in 3,904,291mm



Polling Data Source:
http://www.pollingreport.com/wh04gen.htm
http://www.economist.com/media/pdf/YouGovS.pdf


BUSH KERRY

Zogby Poll
1 LV 48 47 10/4-31/04 REUTERS/ZOGBY TRACKING POLL: 3-day rolling sample of approx. 1,200 likely voters nationwide. MoE ± 2.9.
Bush Kerry Nader Other
10/29-31/04 48 47 1 4

2 LV 49 50 Marist College Poll. Nov. 1, 2004. N=1,166 registered voters nationwide (MoE ± 3); 1,026 likely voters (MoE ± 3).
Bush Kerry Unsure
11/1/2004 49 50 1


3 RV 47 50 Economist YouGov 2903 total; MoE +/-2%
10/30-11/01
Bush Kerry
45 49

4 LV 47 44 TIPP tracking poll conducted by TechnoMetrica Market Intelligence. Oct. 30-Nov. 1, 2004. N=1,284 likely voters nationwide. MoE ± 2.8.
Bush Kerry
10/30 - 11/1/04 47 44
5 RV 48 47 CBS News Poll. Oct. 29-Nov. 1, 2004. N=1,125 likely voters nationwide. MoE ± 3.
Bush/ Kerry/
Cheney Edwards
10/29 - 11/1/04 48 47

6 LV 49 48 The Harris Poll. Oct. 29-Nov. 1, 2004: N=1,509 likely voters nationwide who express a preference. MoE ± 2.5.
Bush Kerry Nader Other (vol.)
10/29 - 11/1/04 49 48 2 1

7 RV 45 48 FOX News/Opinion Dynamics Poll. Oct. 30-31, 2004. N=1,400 registered voters nationwide (MoE ± 3); 1,200 likely voters (MoE ± 3).
George John Other Wouldn't
W. Bush Kerry Not Sure Vote (vol.)
10/30-31/04 45 48 7 -

8 LV 47 48 Democracy Corps Poll conducted by Greenberg Quinlan Rosner Research (D). Oct. 29-31, 2004. N=1,018 likely voters nationwide. MoE ± 3.1.
George John Ralph Other Unsure
Bush Kerry Nader (vol.)
10/29-31/04 47 48 1 1 3

9 RV 46 48 CNN/USA Today/Gallup Poll. Oct. 29-31, 2004. N=1,866 registered voters nationwide (MoE ± 3); 1,573 likely voters (MoE ± 3).
Bush/ Kerry/ Nader/ Other None/
Cheney Edwards Camejo (vol.) Unsure
10/29-31/04 46 48 1 1 4

10 LV 48 47 NBC News/Wall Street Journal Poll conducted by the polling organizations of Peter Hart (D) and Bill McInturff (R). Oct. 29-31, 2004. N=1,014 likely voters nationwide. MoE ± 3.1.
Bush/ Kerry/ Nader/ None/ Unsure
Cheney Edwards Camejo Other (vol.)
10/29-31/04 48 47 1 2 2

11 RV 48 47 ABC News Tracking Poll and Washington Post Tracking Poll. Rolling sample. Fieldwork by TNS. ABC News and The Washington Post share data collection for this tracking poll, but calculate and report the results independently. WASHINGTON POST: Oct. 28-31, 200
Bush/ Kerry/ Nader/ None/ No
Cheney Edwards Camejo Wouldn't Opinion
ABC News Tracking Poll
10/28-31/04 48 47 1 2 2


12 LV 48 49 American Research Group Poll. Oct. 28-30, 2004. N=1,500 registered voters nationwide (MoE ± 2.5); 1,258 likely voters (MoE ± 2.8).
Bush/ Kerry/ Other/
Cheney Edwards Unsure
48 49 3
13 RV 45 46 Pew Research Center for the People & the Press survey conducted by Princeton Survey Research Associates International. Oct. 27-30, 2004. N=2,408 registered voters nationwide (MoE ± 2.5); 1,925 likely voters (MoE ± 2.5).
Bush/ Kerry/ Nader/ Other/
Cheney Edwards Camejo Unsure
10/27-30/04 45 46 1 8

14 RV 48 44 Newsweek Poll conducted by Princeton Survey Research Associates International. Oct. 27-29, 2004. N=1,005 registered voters nationwide (MoE ± 4); 882 likely voters (MoE ± 4).
Bush/ Kerry/ Nader/ Other (vol.)/
Cheney Edwards Camejo Undecided
10/27-29/04 48 44 1 7
15 RV 48 48 ICR/International Communications Research poll. Oct. 22-26, 2004. N=817 registered voters nationwide (MoE ± 3.4); 741 likely voters (MoE ± 3.6).
Bush/ Kerry/ Other Neither Unsure
Cheney Edwards (vol.) (vol.)
10/22-26/04 48 48 - 1 4

16 RV 47 48 Los Angeles Times Poll. Oct. 21-24, 2004. N=1,698 registered voters nationwide (MoE ± 3); 881 likely voters (MoE ± 3).
Bush/ Kerry/ Unsure
Cheney Edwards
10/21-24/04 47 48 5

17 LV 51 46 Time Poll conducted by Schulman, Ronca & Bucuvalas (SRBI) Public Affairs. Oct. 19-21, 2004. N=1,059 registered voters nationwide (MoE ± 3); 803 likely voters (MoE ± 4).
Bush Kerry Nader Unsure
10/19-21/04 51 46 2 1
.
18 LV 46 49 Associated Press-Ipsos poll conducted by Ipsos-Public Affairs. Oct. 18-20, 2004. N=1,330 registered voters nationwide (MoE ± 2.5); 976 likely voters (MoE ± 3).
Bush/ Kerry/ Nader/ Other/
Cheney Edwards Camejo None (vol.)/
10/18-20/04 46 49 2 3

And it is completely irrelevant to the point pollsters and others have been making about why the polls might have been in error.

The Central Limit Theorem tells you nothing about non-sampling sources of error in polls.

Probability figures with gazillion zeros may impress headline writers but they are not going to impress a statistician unless TIA shows at least some rudimentary understanding of the nature of non-sampling error in surveys, and the assumptions underlying inferential statistics.

He might also like to comment on the complete lack of correlation between swing and shift in the exit polls and see if he reconcile that with his claim that the early exit poll projections were "right".

If they were "right", how come Bush didn't gain greater advantage in precincts where they were "right" than where they were "wrong"?

The most parsimonious answer is that the error in the polls had little or nothing to do with fraud, and a lot to do with non-sampling error, either selection bias or non-response bias.

Neither of which have anything to do with the Central Limit Theorem.

And yes, polls probably aren't a lot of use for distinguishing between 49% support and 51% support. However, they are pretty good at distinguishing between 80% support and 30% support so we can be pretty confident that Bush is, indeed, toast. Unless some other catastrophic event bumps him up to 80% again (the law of 4th degree polynomials notwithstanding).


(edited for typo, and for clarity)

http://www-groups.dcs.st-and.ac.uk/~history/PictDisplay/Thales.html

(edit - although he was Phoenician...)

I am happy to debate the old one, but I have to correct you: this is not "a new thesis" whatever that means. It is simply a testable hypothesis.

You know this, because we have discussed it before, but for others who might be interested, details are given here:

http://inside.bard.edu/~lindeman/slides.html

Further discussion is given here:

http://inside.bard.edu/~lindeman/doppresponse.pdf

Full disclosure: both Mark Lindeman and myself (Elizabeth Liddle) were credited by Mitofsky as contributing to these analyses.

If you read both pieces you will see that there are a number of ways in which the plot could be compatible with fraud, even with widespread fraud; it is, however, very difficult to reconcile the plot with a fraud mechanism that could have contributed substantially to the exit poll discrepancy. It would therefore seem more likely than not that the exit poll discrepancy was caused largely by polling bias rather than fraud.

However BECAUSE the analysis is perfectly compatible with the kinds of electoral injustice that we KNOW happened in 2004, just as they did in 2000, then the analysis does nothing to rule out fraud. The analysis simply says that the magnitude is not indexed by the exit poll discrepancy. It could (probably) be less. It could (conceivably, I suppose) be greater.

And the take-home message is: stop regarding the exit polls as prima facie evidence of a stolen election. There is plenty of other work to be done, and frankly, the exit polls are getting in the way.

And while I'm attuned to the other world, let me channel another of the dead:



I believe the citation is from this document:

http://electionarchive.org/ucvAnalysis/US/exit-polls/ESI/ESI-hypothesis-illogical.pdf

In which the hypothesis I referred to is incorrectly stated.

See the second of the links above for clarification.

However, the channelled spirit correctly spots one of the fraud mechanisms that could account for the exit poll discrepancy and not produce a swing-shift correlation: If fraud in 2004 was precisely calibrated to the fraud that occurred in 2000 (and I mean precisely), then yes, that might do it. If every extra 1 percent in Bush's vote share due to fraud in 2000 was mirrored by X times 1% fraud in 2004, yup, maybe you could pull it off. An alternative would be uniform fraud in all precincts. Another alternative is fraud confined to precincts in which Bush was anticipated to do badly relative to 2000 (although you would have to be careful to cover a substantial majority of precincts and be sure to avoid any where he was doing well).

If anyone can suggest an practical algorithm that could achieve this, bearing in mind that it in all NEP precincts where the vote counts for are collected at the precinct, that the fraud must be perpetrated at precinct level, not at tabulator level (about 60% of precincts) while the remainder has to be done at tabulator level, then I will concede, yes, perhaps the exit poll discrepancy was due to fraud.

But remember - there is very little room for any deviation from the fraud perpetrated in 2000. It has to be directly proportional. Any variance and it will show up in a swing-shift correlation. I challenge my channelled spirit to produce a model of 1250 precincts (not 10) with a realistic distribution of WPEs, vote shares, counting methods and swing values in which the WPE (or any other measure of bias) is not correlated with a measure of swing, and yet fraud is responsible for both (I'm happy with 70% shared variance, or even less). Oh, and you can't end up with a large mean WPE in 2000 because there wasn't one.

Estimates of realistic variance in bias and vote share can be gleaned from the plots linked in this DKos diary:

http://www.dailykos.com/story/2005/5/24/213011/565



(edited to replace accidentally deleted word)

I have to commend your effort, I thought this was self evident, and would not have modeled it.

Where to now? I think we need to tighten parameters on potential voter suppression in Ohio (I know that you are backing away from the values for Hamilton), Florida, and New Mexico. Although not a complete explanation, will address maybe 50-75% of the margin. I'm trying to get to it, but today is the first slow day at work in a month and half, and I don't see it easing up.

Mike

"I think we need to tighten parameters on potential voter suppression in Ohio (I know that you are backing away from the values for Hamilton), Florida, and New Mexico. Although not a complete explanation, will address maybe 50-75% of the margin." -mgr


Which Hamilton county (Ohio) values are being backed away from? Thanks.

And what "margin" are we talking about? The Bush-count Kerry-count difference? Or the difference between the exit polls and the vote count? (Voter suppression would not affect the latter "margin", of course.)


Cheers!

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x399216#399456

For the record, I now reject outright (I earlier entertained the possibility that this was not the case, but I have not had the pleasure of a credible trained statistician confirm it for me, in fact the opposite has been the case) the possibility that the 2004 NEP exit poll can inform us about anything in the presidential election, and in regards to sub voter preference, the margins for error are too wide to ascribe anything in too close an election. What I was referring to was the margin between K & B in OH specifically, and in general FL and NM.

Mike

Maybe it made it to buzz flash. Good, reject it and we won't be seeing your commentary on it any more since it's a moot point.

As they say..."thanks for sharing." It was highly meaningful.

...you might:

1) Mug me by myself... I'm not sure what to do with being mugged along with TIA... and on two seperate locations in cyberspace at that. It probably doesn't matter because your argument is essentially with yourself. I don't concede "So redshift, and "swing" (Bush's performance relative to 2000) ought to be positively correlated"... Let me repeat: Oh yeah? Prove it.

2) If you are going to mug me on THIS thread, you might at least have the decency to do it in Latin.

Fas est et ab hoste doceri.

I'm afraid I last composed a sentence in Latin 40 years ago. Something about Caesar.

And what little Greek I have is not ancient.

But if you are going to mug me elsewhere in cyberspace, you will have to bear with responses in English here. So will TIA.

As for "proving" that redshift and swing will be correlated if there is fraud - this is a prediction not a proof. It's a testable hypothesis. If one factor causes two effects, you would expect those two effects to be correlated. If exposure to sun causes both tanning and skin cancer, you'd expect incidence of tan and skin cancer rates to be correlated. If fraud causes both swing and redshift, you'd expect swing and redshift to be correlated. If they were, it would be fairly strong evidence that the same factor (fraud) caused both. Unfortunately they are not.

What this means is that if fraud caused both, its relationship with each was not independent of the other, and it is remarkably difficult to devise any plausible fraud mechanism that would not result in a swing-shift correlation. I have tried. I have encouraged others to try.

No mugging intended. I am merely trying to teach, and learn from, my enemies. Except that I wish we weren't enemies. It seems to me that we are all concerned with the same issue - fixing America's broken democracy. I just happen to think that the emphasis on an aspect of the breakage for which there is least (and possibly counter-) evidence is not helpful, and potentially harmful.

It's information we need to have and an understanding that a lot more of us should learn in school. I appreciate your expertise on the topic.

Bush stole the election of 00 and 01 imho. But the polls don't "prove" it. And we've got to nail him on what he's doing today, not how he got there 5 years ago.

Though I do believe we have no need of black boxes in our voting booths...they are simply unnecessary and unduly tempting and risky for those who wish to subvert democracy.

We know he didn't get the nukmbers right in 2004. and we all know he ain't gonna get them right in 06 ,as long as the election theft machines are counting the votes, so whats next for him?

Is from William of Malmsbury's English histories from the 12th century--literally translated the voice of the people is the voice of God. A root idea of English protestantism.

What the Greeks have to do <with> this, or Thales (who has no extant writings attributed to him, and why he was writing in latin though we was an Ionian existing far earlier than the Romans) I am not sure. Possibly it derives from Cicero or Cato during the golden or silver age. Dulce et decorum est pro patria mori (Horace)?

Mike

Like a former jesuit with a predilection to quoting Tellurian and Boethius in latin, I think you confuse erudition with reason. The oblique allusion was to Wilfred Owen, and the current issue relating to the use of white phosphorus on combatants in Fellujah.

The thing is that the concept 'vox populi, vox dei' has more in common with medieval scholasticism than Greco-Roman rationalism. We owe as much to religious dissent for our world view as we may the classical Greeks and Romans. Just because it is Latin, does not mean it's from the period of great classical antiquity; and we need to stop mythmaking that denies this heritage.

This same perversion of history to make the world fit with one's dualistic reductionism is what allows one to ignore what underlies why the exit polls are an inaccurate and an imprecise measure of voter choice in 2004. All statistical scientific measures are apt without any consideration of the inherent variability of the phenomena to be measured (just try telling this to an evolutionary biologist, let alone a pollster)--just as all concepts of political equality and egalitarianism originated with the Greeks and Romans, such that all christianity gave us was fear and superstition.

Mike

It'd take some kind of almost-magic event but look at what Katrina did.

Any kind of whistleblower seems unlikely as does a police investigation 'stumbling' upon evidence.

It would have to be something incontrovertible to even joe 'red state'.

Perhaps some kind of key that puts each different state's 'scam' into perspective. Some document perhaps which simply has a name and a 'code' of some sort. Unlocking that code would bring the picture into absolute focus when investigators look for CERTAIN clues, rather than probing around in the darkness of who knows how many levels of corruption hiding the acutal culprits.

Surely each state really did have totally different MOs for pulling off their vote scams. Some things might be fairly common, but hidden in different ways in different states. Like the 'classic', using database collection and searching of voters who were known that they wouldn't actually vote and so their votes could be happily, ahem, 're-used' by the wholesale vote-stuffers. A time honored vote scam! Surely at each state's database company there would be some things left accidentally un-shredded if the authorities went in knowing exactly what to look for.

Something that you would hear or see in the news and just blurt out "holy toledo!". Especially if the crime was laid out bare while the criminals were still left to be detected. The only upside to the media syndicate's happy compliance with hyping anything that boosts viewer eyeball-hours, including a fake war like Iraq, is something like that would BLOW their ratings sky-high and to their own ultimate evil-emperor's, their capital-holders, they could say "look at all this free money!!" and let this ** administration maybe actually feel some real heat from the fire -- and maybe, just maybe, tap into some kind of red+blue combo backflip reaction where ALL the american people demand the obvious THEFT be corrected.

You've probably seen the pic this week of ** trying to escape the press in China only to find the doors locked? Perhaps that would be a foreshadowing to how such a fantasy would end. The sad picture of the newly un-declared 43rd president of the US realizing he was used and abused with a better-late-than-never feeling of failure and having been the biggest living lie since good old Mr. Germany in 1945.

When all is said and done, 5 or 6 years of such a roller coaster certainly would leave the country in some kind of weird headspace.

Can't blame me for dreaming!

{B^)

I think your dream makes a lot of sense (and it's always a pleasure to see you show up and make sense!).

It represents a necessary function, wish fulfillment. After Katrina, still resonating, people are ready for just about anything as you say.

Bush hits 35% consensus approval and 60% disapproval. How long can people tolerate the fact that this idiot is leading the country and they elected him. People need an excuse, a way to escape their guilt...here it is...the election fraud story. Let it roll down the runway and take off on a national tour.

There will be multiple discoveries of the "varieties of fraudulent experience" in election domains. Someone who talks will actually get covered. The press will actually investigate (I mean crib off of citizen activists, which is A OK!).

Let the expiation begin!

This scenario has the convenient quality of also being true, the merger of wish and reality.

If wishes were horses, we'd all take a ride. In this case it will soon be time to hop on.

Nice seeing you.

Appreciate the link to Progressive Independent too!

:)

It dawned on me this week that the gift of ALL of the Bu$hCO scams is that the sheer revolting number of them makes it more likely that people will clearly see THIS one, the most important one.

So, now every time one of those monsters goes before the media and lies their @ss off, I mentally thank them for their help. :)

The blather and stupidity with which they advance their weak reasoned and pathetically documented case to for Bush legitimacy is truly amazing. It's as though they're some sort of holographic bot in search of a friendly crowd. The crowd is getting smaller and smaller, the lies are getting more and more obvious, and you're right, re-framing idiocy into a message identifying idiocy is a higher level response, and probably more correct than wasting psychic energy (after all I need that for the Truth and my next feat...TBA).

:hi:

That's what I like to see Autorank! A bit of logical follow-through...! The kind of elementary common sense we routinely use all the time in daily life.

Public to * : "It's time, there's the door, you've done badly George and you shouldn't have been here in the first place. And while your'e on your way out take Cheney, Scalia, Roberts, Thomas, and all of your Federal judge appointees with you."

It's only fair, isn't it.

My very best regards to you.