ONE OUT OF 4.5 BILLION!

in the real world his odds are excellent.

Philosophy works fine when you don't have the math behind you

There it is, in black and white!

But we'll just conveniently forget that throughout the history of exit polls there has NEVER been an incident like this. We'll just go ahead and tell the world that some magic thing just happened to go wrong this time, and the sheeple will buy it. Who has time for all that - not quite even college level for Christ's sake!! - stupid math?

Between this and my disintegrating personal life I'm ready to have myself a nice long nervous breakdown.

complicated, sophisticated and supposedly civilized world, everything is buried under a ton of facts an numbers. Since I am an optimist, I want to reassure you by the fact that NOTHING can be hidden forever.

People not well verse in stats may think that they cannot help in exposing the truth about the last election. Wrong! Maths people can and will do their part inspired by emotional people.

And , somewhere in time we will all put our voices together and win the battle against bushco.

Lise

your northern neighbor

I will be unable to respond to queries, criticisms, trolls, etc. until after I return from a much needed weekend trip on Sunday.

I suggest that those who are so inclined to please check the following before posting replies:
a) exit poll data (4pm numbers)
b) definitions of the probability functions

Let's make sure we are all on the same page as far as the basic stattistical methods used. If you have suggestions, be specific.
If you prefer to use alternative methods, state them - but be specific. In fact, calculate the probabilities using your methods. There are many ways to skin a cat.

One out of four billion odds seems totally unrealistic at first blush. We are all familiar with Dr. Freeman's original 250 million to one odds based on exit poll deivations in FL, OH, and PA. He subsequently dropped the odds down to one out of a million or so.
My calculation is based on 16 states exceeding the MOE and all going Bush's way.

So there is a major difference in the calculation. The odds reflect it.

Cheers,

tia

I am no math wiz but it sure looks good to me.

If the shoe was on the other foot, the Republicans would have Kerry in leg irons for stealing the vote.

Kick this folks!

Thank you so much for this.

You mean that Kerry's Drop from the value shown in the exit polls exceeded the MOE in 16 states.

tia

in the voting tally at each precinct so that even if one was discovered, they can say this was just a single eroor and that would not have materially affected the outcome.Your global results are a better way to deal with this issue of voter fraud.

in 2000, but we were too busy paying attention to Florida.

toward them a little bit at a time and not in some massive number at one location.This makes it all but impossible for us to detect the fraud unless we look at it from a global perspective. Even if one precinct, as in Warren, Ohio is caught, they can deny it occurred at all locations.A very nice strategy if you ask me.

It's pretty clear the attack was multi-modal, very sophisticated and yet could only be pulled off with a simplistic elegance of command. Kinda like, well, organized crime syndicates, Hmmmmm.

<Waves hands:SFX:doodle-oodle> "Party on Wayne"

Garth: Can we, like, amass enough individual incidents to file a RICCO suit?

Like, going back to 1996?
</silly>

A few examples:

We have these divergences, Red shift, Does anyone have correlation numbers with Red controlled BOEs? (I was a sub sailor in the Regan era cold war. Red was bad then too )

From 2003, we have HAVA amendments: H.R.2239 and S.1980 (essentially, paper trail requirements) stuck since May 2003 and Dec 2003 respectively, in Rep controlled committees.

The Great Lone Star Gerrymandering.

Eloriel's collection which includes many examples of voter obstruction.

There's the NC analysis that correlates tabulators to Red shift. I'd love to see that for *ALL* states.

This stuff all needs to be collected and organized. The wikipedia article is a good start, but not inclusive of other periods.

The documentation needed would be massive, are we up to it? Maybe DU needs a wiki.

-Hoot

....several groups/individuals are amassing incident reports, by State, and others are doing studies/analyses, like TIA's, but, without coordination we will fail the challenge -- we must force a "Halt and Audit" of THIS National Election.

Reform of practice can come later, but if we don't force a "Halt and Audit" the effort at reform will face even more obstruction than it has since 2000. In other words, we've done that experiment - (default to a selected Pres in 2000 and have folk with an agenda that is anything but supportive of our Constitutional rights obstruct legislation, expand and enhance their skills at 'systemic disenfranchisement).

Thus, those individuals reading this thread who have access to the executives at the ACLU, the AFL-CIO, the Rainbow Coalition, Common Cause, etc -- we need a group of those executives to form a National Task Force to Save the Franchise and they need to bring all their resources to support a centralized repository of incident reports but, even more importantly, coordinate demands for investigations of every 'central tabulating center' and EVEN MORE importantly, broadcast the call for "Prove My Vote Counts, Now," many, many, many times every day, through all media -- full page ads in major papers, 30 sec slots on all major cable and radio/tv, etc.

All I can do is think and type and send dozens of letters per day and I'm doing that, but without our fellow citizens calling "Halt and Audit" we will see an even more concerted effort to render our Constitutional franchise extinct in the next 2-4 years.

Peace.

Excellent post, understandinglife.

Question one - your statement of the problem and recommendation for the solution is so simple and direct - why are people resistant to this do you think? Why would not all members of DU support what you just said, for example?

Question two - you mention the ACLU, the AFL-CIO, the Rainbow Coalition, Common Cause - all groups associated with Democrats. Would there not be great value in bringing in organizations associated with the opposition? Would that be possible do you think?

...to explain what it is and why it matters that we demand the test be done, now:

http://www.dailykos.com/comments/2004/11/8/114232/967/2...

I agree that "Florida" in 2000 was very likely a 'gambit' -- and IT WORKED.

Once in the 'white house' the available gambits for Bush and Rove to use only expanded, but the one thing they underestimated is that a bunch of smart folk happen care deeply about the franchise -- that has been demonstrated by the types of analysis TruthIsAll, Freeman and others have done. And, corroborating these studies, in the most ironic and unpredictable way imaginable, is the Ukrainian National Election!!!

What we need to do is not be distracted by whether one State or another supports a 'recount' -- what we must do, at a National level, is create a massive, and simple demand "Prove My Vote Counts, Now" and if you don't, We The People refuse to validate the election.

We have the leverage, now we must apply it on a scale never seen in this Nation. I think, just about now, way more than the 55+ Americans who voted for Kerry are ready to call 'TIME OUT.'

All they need is a solid reason to motivate them and we've got it -- use the Ukrainian situation to call on the deep emotion all those Americans have for their franchise, and make that call by simply showing them that the exit polls indicate that what they feel in their guts is correct and that the fact that no one can show them If or How their vote was counted.

The only barrier between where we are now and having that happen is getting the message to The People -- please bring maximum resource focus and allocation to delivering the message in the next 72 hours and don't stop until we have millions of Americans behaving like the Ukrainians. No one can stop us from saving our franchise as long as we don't stop ourselves.

Peace.

Do you trust Warren Mitofsky and Edison Media Research?

I'm going to keep on asking this question until I get an answer.

You want to make it personal, raising strawmen.
Are you reduced to this?

I focus on the numbers.
You ignore and avoid the analysis.

Is that because you just don't understand it?
Or is it because you cannot refute it mathematically?
Or both?

As for Mitofsky, we all wait for his official exit poll data.
Till hell freezes over, if need be.

How exactly am I making this personal? I ask you a simple question and you refuse to answer it. Please give us all an answer:

Do you trust Warren Mitofsky and Edison Media Research?

No, I trust Warren O'Dell of Diebold, who contributed over $100,000 to the Bush campaign, a Bush "Pioneer" who promised to "deliver" Ohio to Bush in '04, and who owns and holds as proprietary information the secret source code by which all our votes were "tabulated," and billionaire H. Ahmanson of ES&S, who funds rightwing religious causes, and who, along with O'Dell's Diebold, manufactured many of the electronic machines on which we voted--both of whom insisted that their machines require no paper trail. "Trust us," they said. And who are WE to place non-partisan exit polls, used worldwide to verify elections, above these Dark Masters of the Art of Keeping George W. Bush in the White House?!

Finally someone who answers the question.

So you don't trust Warren Mitofsky and Edison Media Research. Fair enough. Obviously, therefore, you don't believe that the 4pm exit polls they produced are trustworthy either. True?

I knew that at least one statistician would pull it all together like that. I was just waiting patiently . i would also like to hear the other two legitimate reportsSteven Freeman and Berkeley) pull it all together like you did.

And yeah for those of you responding with your mindless matter of fact answers, well you just don't understand math. That is what it all comes down, you just don't get the math.

Reguardless of the fact that the adjusted final numbers teaked the results to match the final vote, just eyeballing it, there still is probably a statistical story to be had by analysing those numers.

You won't get dramatic impossible odds numbers out of it, but if even if you assume a pro-kerry systematic bias, even in the final numbers, by calculating the average error you can account for this. Then you can calculate the odds that all the battleground states would be on the Bush side (weighting by how much they are on the Bush side). Yes, the massaged numbers are all within margin of error, but there is still a probability curve even inside that margin.

My bet is you'll find that there is still a signifigant improbability even in the numbers that have been massaged. But I don't want to presume to do such an analysis lacking the statistical expertise... won't someone with such expertise tackle these numbers, to give us an absolute lower limit on the improbability?

to data manipulation and forgery on a massive scale.

That is what they have to do.

TIA

TIA -

I agree that your analysis reflects what others' have shown and that we KNOW, i.e. that statisically, something ain't stirring the kool-aid. I was curious what the underlying assumptions, the "givens" are for this analysis - assuming that the exit polls correctly show the voters' intent. I think, based on the history of exit polling, this is probably the case, but I just want to play devil's advocate and explore the strengths and weaknesses of these types of analyses. I'm no math whiz, but I'm trying to understand these things - this is too important not to bend some mind-power to it.

Thanks for your work - keep it up!!
KJ

on edit: Furthermore, in light of the post preceding mine (by skids), it occurs to me that comparing the pre-adjusted and adjusted numbers is meritorious in itself because some bright individual could come across some evidence of an algorithm used to shift the numbers uniformly to *. (Some wicked kind of reverse-engineering.) If some kind of formulaic shift was found to be shared across the states from pre- to -post adjustment, that would, de facto, show fraud in my mind.

to get this pattern into the hands of skilled algorithm developers, and I'm sure many of them exist who care deeply for our franchise and would be motivated to jump on this issue, big time.

Put the word out wherever you think it might elicit support.

Peace.

a common denominator, formula or some other pattern that makes one set of numbers not jive with the others. I am no math genius but if relativity can be winnowed down to it's simple equation, surely this can too.

Thanks TIA , I am sure the trollers will be hanging around somewhere in the vicinity Sunday.

to the overall election results.

You have to start with analysis of the basic state polling data Look for suspicious patterns which jump from the stats. Then use the state stats to calculate a robust, global probability estimate of mistabulation or F-R-A-U-D. Or both.

Circumstantial evidence? Sure. But VERY convincing.

Something smells.

Now, will Olbermann be the only one to report thse facts?

That anyone that bakes a cake or builds a house has this certain pattern they follow. It is built in like the lines on the back of hand. You cannot make things with out it. It may not be obvious by even comparing the numbers, but when the formulation is discovered it becomes obvious how you can get the results.

Taking it apart piece by peace from the way it was built would yield answers. A ubiquitous randomization wouldn't work because of the specifics involved (an elimination exercise perhaps?)

....and those who construct algorithms typically use tools to achieve one or more representations of how to generate a pattern.

What I'm suggesting is a focus on how the algorithm might most efficiently be deployed. My hypothesis is that access to several 'central tabulating systems' seems much more efficient than hacking individual optical scanners or e-vote machines.

Now, if I wanted to hide my efficiency, one of the several things I would do is choose certain precincts that are heavily populated by voters registered for my opposition and I would make even their task to vote very, very hard -- call it the 'long line gambit in disadvantaged precincts.'

Another thing I might do is have several of the Evote computer behave in a flaky way: for instance, as the voter attempts to select my opponent, the vote appears to switch for me -- call it the 'Evote head-fake gambit.'

In other words, I'd create several forms of true disenfranchisement but when folk tried to build a case, all the pundits, 'reasonable media anchors and editors' and, even many within the 'leadership' of my opponents party would go -- yeah sure thing dude, we know it's bad but no way is it going to add up to '4 million votes' so forget-about-it-anyway; if your so concerned just go work on reform for the next time.

All the while, the actual mechanism used was a simple algorithm that displaced a certain number of votes at strategically selected central tabulating centers (strategically selected to include not only 'battleground states' but 'sure win states' for BOTH me and my opponent).

Give it some thought; this ain't conspiracy theory folk, this is an attempt to build testable working hypotheses and testable algorithms to determine how an all-time, historical event happened sometime between ~ 6pm EST 2 Nov - ~ 1am EST 3 Nov 2004.

"It's about America" -- currently on life support and in need of immediate, super-smart, heroic treatment.

But what would you say if they figured they would get most of their numbers that way, but just cheat a little also to make sure. Then people found out about it but said it's okay because it was only done a little (this is scenario of what happened currently). I say to you, 30 years ago with Nixon this impropriety would not of passed muster with the rest of the population. I guess we will have to get stuck in deeper dog po before anything changes.

The scum may have risen it to the top, but the undercurrents that accept it might also have something to do with it. Even excepting premise that the vote was stole leaving 48-49% of the "OTHERS" Should be a problem in anyones book. There is policy of a house divided here, make no mistake. Should we make a play of jue-jet-sue or be the shrinking loyal "other"

On Edit: But just don't forget 30-40% of eligible voters don't, children can't and legal and illegal aliens cannot. Many or more are also very Un or mis-informed. It's the information not the people mostly as I see it.

....I think we need to leverage what Prof Ian Solomon has just written in the Baltimore Sun and build a massive "Halt And Audit" demand from as many millions of Americans as we can.

Some relevant comments are posted here:

http://www.dailykos.com/story/2004/11/26/21262/464

And, I'm sorry about the confusion regarding the 30%; I was merely suggesting that if 30 % of the responders to the poll were eligible voters. You are certainly correct that my assumption is perhaps not accurate.

Peace.

http://www.onlinejournal.com/Special_Reports/112504Mads...

Saudis, Enron money helped pay for US rigged election

By Wayne Madsen
Online Journal Contributing Writer


"November 25, 2004—According to informed sources in Washington and Houston, the Bush campaign spent some $29 million to pay polling place operatives around the country to rig the election for Bush. The operatives were posing as Homeland Security and FBI agents but were actually technicians familiar with Diebold, Sequoia, ES&S, Triad, Unilect, and Danaher Controls voting machines. These technicians reportedly hacked the systems to skew the results in favor of Bush."

(there's more at the link)

Firstly, thank you for your considerable and persistent effort to build a framework within which others can begin to grapple with what happened to the franchise of our democracy on '2 Nov 2004.'

You, Professor Freeman, Alastair Thompson with Ed Shalom and others have grounded all our concerns in a basis of analysis. A few others, like myself, have attempted to hypothesize about the mechanisms that might produce the results, the actual results of votes tallied that diverge significantly from the predicted outcome.

Much has already been written about the irony of SoS Powell, Senator Lugar's and others pronouncements about the validity of 'exit polls' in the current Ukrainian election and the lack of any voice to that same issue in the US National election .

Recently, I've attempted to contribute by requesting that everyone ask their neighbors, friends, family, media, etc., three simple questions:

1. Did you vote;
2. Do you remember for whom you voted;
3. Do you Know If and How your vote was Counted.

Those questions are a method, a device to attempt to bring a massive outcry for "Prove My Vote Counts, Now".

It has another tightly linked utility -- I, and others, want exhaustive investigation, BEFORE the election is declared valid, of not only the central tabulating systems , but of all telecom logs, and other access logs, pertinent to whomever had access to those systems. Discussion can be found here:

http://www.dailykos.com/story/2004/11/24/115048/86

The point is that to declare the election invalid what the stakeholders in the franchise need to realize is that the core of the tabulating process, in Every County in Every State, is suspect at best, and could well have been accessed inappropriately .

So, I urge all of you to leverage the outstanding efforts of TruthIsAll, Prof Freeman and others and focus intense and unrelenting pressure on your Congressional representatives, the media, your respect State SoS and AG, and others on the fact that the most fundamental method by which We The People conduct our Constitutional business is not just broken but is the very reason why all our other Constitutional rights are threatened with extinction.

Thank you.

"It's about America" -- currently on life support and in need of massive, focused, heroic and immediate treatment

Is that the most highly improbable swings occurred in only a small subset of states, and thus there is no reason to think the whole exit poll methodology was off.

Good job, TIA!

Of course, it would be nice to know where you got these numbers from!

that when our exit polls are wrong, it was just a malfunction, but when the exit polls in the Ukraine are wrong, the election is not valid.

Welcome to DU!

What amazes me is the blatant boastful hypocrisy of it all. It is like we are stuck in some sort of tragic Shakespearian comedy...

that it was the exit pollsters who were wrong. I understand that you are arguing that the probability of them accidentally being wrong is infinitesimal. Is the freeper answer to this that Zogby et al were PURPOSEFULLY falsifying the exit poll data to make it look like Kerry was winning, either to discourage Republican voters from bothering to turn out on the West Coast, or so that Dems would be able to claim FRAUD once the actual results came in?

To me, these are the only two explanations for the discrepancies. I have heard a lot of talk about how they could not have been wrong accidentally, but how do we prove that it is the voting machines that are lying, rather than the exit pollsters?

Again, I believe the exit polls were correct. But I think among the folks who aren't desperately trying to pretend this situation doesn't even exist, one theory is as paranoid, or as credible, as the other. I think we should be putting energy also into debunking the lying-pollster theory, even as we gather evidence to support theories of election fraud.

Well done TIA!

However, you are missing one other possibility. TIA's and others' analysis prove beyond a shadow of a doubt that random error (chance) could not account for the discrepancy (between exit polls and official election results).

However, that or purposeful falsifying of the exit polls or a fraudulant election are not the only possibilities. The other possibility is exit poll bias. That does not mean that they were purposely falsified, but rather that for some unknown reason they were slanted towards Kerry. It is a methodology issue, or an issue inherant to exit polls.

I believe that a fraudulent election is more likely the case (see Dr. Freeman's excellent discussion of the accuracy of exit polls). However, we cannot and should not summarily rule out biased exit polls as a possibility, because when we do that we lose our credibility.

My understanding from reading these boards and others is that exit polls were MORE wrong in places with electronic voting (and no paper trail). Is this true? It seems like if the methodology was wrong on the exit polls, these discrepancies would have been uniform (or more uniform) throughout the system.

This point keeps getting obscured. People ask were the exit polls accurate or weren't they? My understanding is that they were quite accurate in some places, and quite inaccurate in other places. Where they were inaccurate just happens to be the states that determined the election, and the inaccuracies all favored one candidate. So the question is this - why have exit polls consistently been accurate everywhere and every time for 40+ years within a certain MOE excepting a handful of locations in 2000, 2002, and 2004 that happened to favor one party and that happened to be the swing areas and happened to determine the outcome of the elections?

.....here in the good ol' USoA sometime between ~ 6pm EST on 2 Nov and ~ 1am on 3 Nov 2004:

http://www.dailykos.com/comments/2004/11/26/94921/014/8...

Peace.

......America.

That credibility thingie is coming 'home to roost'....you betcha:

http://www.bradblog.com/archives/00000991.htm

Pravda online:

"Where was this condemnation during the appalling electoral fraud committed in the USA on November 2nd?
...
Mention of "electoral fraud and abuse" from an American observer was risible, after the two fiascos in the USA which saw the most flagrant examples of vote-rigging and electoral fixing in modern history.
...
When the Republican Party deploys electronic voting machines bought from Republican Party fundraisers who promised before the election to help the President to win, the OSCE observers describe it as localised and insignificant incidents. However, when the incompent stooge Yushchenko fails to win in the Ukraine, it is fraud."
====

As I've been indicating:

"It's about America" -- currently on life support and needing massive, immediate heroic treatment"

Suggest starting an IV running 100%, 24/7 infusion of "Prove My Vote Counts, Now"

Peace.

This information corroborates the theory that many states were targeted for fraud.

It was important for Bush to get the popular vote this time as well, so that his electoral victory would not be held up as a sham again, as in 2000.

Could you please explain where your numbers came from?

This analysis and others prove beyond a shadow of a doubt that random error (chance) could not account for the discrepancy (between exit polls and official election results).

However, I believe that it is important that we not summarily dismiss the possibility of exit poll bias (I'm not talking about purposeful falsification of the exit polls, but rather a bias that for some unknown reason, other than chance, worked in Kerry's favor in the good majority of the states.)

I believe that a fraudulent election is more likely the case, perhaps much more likely -- (see Dr. Freeman's excellent discussion of the accuracy of exit polls). However, we cannot and should not summarily rule out biased exit polls as a possibility, because when we do that we lose our credibility.

Fellow bloggers, this is my first post, but there is a first time for everything, right?

I,too, believe the election in selected states was stolen.
However, the above estimate of probability hangs on a pesky little assumption, which is that the exit polls were a representative sample of the vote population. There is a number of reasons why it might not have been, such as Bush voters declining to answer who they voted for. Even if the sampling were done right, the polling ended at 4 PM, which means that they did not sample homogeneously all voters. I am sorry but in light of these things the numerical estimate (1 in 4,5 billion) is meaningless.

Our hope for redress rests squarely on the GAO investigation of specific instances of voter suppression and on the Ohio recount.

but the GAO is the last place i lay my hopes.

other than an early unpaid retirement. Looks like it will be good time to catch up on that reading

FOREWORD

If you strip Stoicism of its paradoxes and its wilful misuse of
language, what is left is simply the moral philosophy of Socrates,
Plato and Aristotle, dashed with the physics of Heraclitus. Stoicism
was not so much a new doctrine as the form under which the old Greek
philosophy finally presented itself to the world at large. It owed
its popularity in some measure to its extravagance. A great deal
might be said about Stoicism as a religion and about the part it
played in the formation of Christianity but these subjects were
excluded by the plan of this volume which was to present a sketch of
the Stoic doctrine based on the original authorities.

ST GEORGE STOCK M A
_Pemb. Coll. Oxford_



A GUIDE TO STOICISM.

ST GEORGE STOCK


PHILOSOPHY AMONG THE GREEKS AND ROMANS.

Among the Greeks and Romans of the classical age philosophy occupied
the place taken by religion among ourselves. Their appeal was to
reason not to revelation. To what, asks Cicero in his Offices, are we
to look for training in virtue, if not to philosophy? Now, if truth
is believed to rest upon authority it is natural that it should be
impressed upon the mind from the earliest age, since the essential
thing is that it should be believed, but a truth which makes its
appeal to reason must be content to wait till reason is developed. We
are born into the Eastern, Western or Anglican communion or some
other denomination, but it was of his own free choice that the
serious minded young Greek or Roman embraced the tenets of one of the
great sects which divided the world of philosophy. The motive which
led him to do so in the first instance may have been merely the
influence of a friend or a discourse from some eloquent speaker, but
the choice once made was his own choice, and he adhered to it as
such. Conversions from one sect to another were of quite rare
occurrence. A certain Dionysius of Heraclea, who went over from the
Stoics to the Cyrenaics, was ever afterward known as "the deserter."
It was as difficult to be independent in philosophy as it is with us
to be independent in politics. When a young man joined a school, he
committed himself to all its opinions, not only as to the end of
life, which was the main point of division, but as to all questions
on all subjects. The Stoic did not differ merely in his ethics from
the Epicurean; he differed also in his theology and his physics and
his metaphysics. Aristotle, as Shakespeare knew, thought young men
"unfit to hear moral philosophy". And yet it was a question--or
rather the question--of moral philosophy, the answer to which decided
the young man's opinions on all other points. The language which
Cicero sometimes uses about the seriousness of the choice made in
early life and how a young man gets entrammelled by a school before
he is really able to judge, reminds us of what we hear said nowadays
about the danger of a young man's taking orders before his opinions
are formed. To this it was replied that a young man only exercised
the right of private judgment in selecting the authority whom he
should follow, and, having once done that, trusted to him for all the
rest. With the analogue of this contention also we are familiar in
modern times. Cicero allows that there would be something in it, if
the selection of the true philosopher did not above all things
require the philosophic mind. But in those days it was probably the
case, as it is now, that, if a man did not form speculative opinions
in youth, the pressure of affairs would not leave him leisure to do
so later.
(snip)
http://www.btinternet.com/~k.h.s/stoic-foundation.htm

Your scenario, although theoretically possible, has no evidence to indicate that it is plausible. Exit polls have a long history of being very accurate. Why would you think that Bush voters would decline to answer, but not Kerry voters?

You are correct, however, that we will need an investigation, or several investigations to make the case. Unfortunately, if the crooks have managed to get rid of the evidence by now, it will be a lost cause -- but we need to keep on pressing.

...they left out all those working poor and minority voters who have to vote at the end of the day because they CAN'T AFFORD to lose work (paid by the hour), and have lousy jobs or lousy bosses, or they have to get the kids off to school in the morning, then go to work, then vote; or have to pick up the kids after school and feed them and see about babysitters after working the night shift, and then have to go vote instead of sleep; or the shits who run their state elections shorted them on voting machines or the number of precincts and made them stand in line for 10 hours, or drive 30 miles to maybe find their polling place (unless election officials gave them the wrong information), or ran to the polling place late in the day after they found out that Republican-distributed flyers were WRONG that they had two days to vote....

...that's who votes late in the day. And THEY didn't get "exit polled."

WITHOUT paper trails??????

if the problem is with the poll methodology or the responders' mindset, why isn't the discrepancy everywhere, instead of only where the actual vote cannot be examined???????????

Had all of the exit polls been consistently outside of the MOE, that would be one thing, and we could start questioning why the exit polls were so strangely off this time. However, they performed poorly selectively. That is the cause for suspicion. Why did they fail only where they coincidentally needed to fail for a certain outcome?

As a statistician (and a hardcore Kerry supporter) I really have to step in and provide some analysis of these numbers and hopefully teach everyone a little about statistics along the way. But first, I want to make an important point:

STATISTICS DO NOT PROVE ANYTHING - they can only be used as a guide or a tool to inspire further investigation. Statistics are not "evidence" or "proof" of vote fraud (no matter how much we all want it to be true) but rather a method of uncovering irregularities that should be investigated further.

Now on to these numbers TIA has posted.

His math is essentially correct BUT HIS CONCLUSIONS ARE NOT. The odds of that many exit polls being that far outside their MOE are very small indeed, but there are several problems with his analysis.

1. As noted in a post above, TIA assumes a normal, random distribution in the exit polls and that is not the case. The problem with statistics is that they rely extremely heavily upon the notion of "randomness" of the samples taken - if the samples are not truly random, the results are skewed and are basically useless.

2. The MOE and Standard Deviation calculations don't seem to be accurate because, as I said before, they assumed a simple random sample of voters. The exit pollsters were not placed randomly around the country but rather strategically grouped so as to estimate "clusters" or "groups" of similar precincts and counties together. (See www.mysterypollster.com for more information on this)

3. Another wrench in the problem is the actual randomness of the voters being polled at each precinct. The pollsters are not professional pollsters but mostly college students looking for extra money and they were trained over the phone and given an instruction packet on how to do the exit polling. They also had to take breaks from polling throughout the day to call in their data every few hours. There is also the theory of oversampling Kerry voters which, while not a prevaling theory, probably does have at lest a slight effect given that Republicans are likely to ignore a pollster than a Democrat (since the pollsters are decked out in ABC-CBS-NBC-AP-etc. apparrell and over 50% of repubs don't trust the media, compared to just 38% of Dems).

All these factors create some more uncertainty as to the "randomness" of the sampling and will create a significantly larger MOE than the one TIA has calculated. The actual calculations are much more complex (Freeman does an excellent job with that aspect in his newest version of his paper: http://www.appliedresearch.us/sf/epdiscrep.htm ).

I do want to commend TIA on his efforts because I know it took him a long time to put it all together. I wish I could agree with your conclusions but sadly I cannot.

I know we all want to do our part to help uncover the fraud in the election, but I urge you not to jump to any conclusions about any of the statistical analyses that come out. So far, every paper has been disproven or cast into doubt (including the papers refuting the pro-fraud papers) so just remember that statistics are fun because they can be easily manipulated and your inherent bias will almost always show up in your analysis unless you make a concerted effort to overcome it.

So, to conclude, don't believe everything you see and hear when it comes to statistics. Especially the ones that are being rushed out in the wake of the election because they aren't submitted to rigorous expert analysis and criticism before they're released and most likely have errors in them. TAKE EVERYTHING WITH A GRAIN OF SALT and just hope that somehow, in some way, the truth will come out.

In the meantime, start changing the terms of political debate by using progressive frames to argue our values and our beliefs.

See we already mostly made the bet, you might of missed it, but that's okay too. See we, us many here, been watching the man behind the curtain. I say whats up with you? You want us to play your game, but you are unwilling to play ours. We are where we belong but are you sure about yourself in this place you are at.

In a world that don't make sense you come to search out the nonsense. We are not basing our corner stones of any logic of perceptions but on history we have already seen. You can crunch them numbers till doomsday and still be calculating them when it gets here. That's where perceptions fit in, hang around for while, maybe you can see too.

Peace out man

and I've been following the vote fraud cases just as close as everyone else here. But do you honestly believe that this election will be overturned? Do you really think they'll be able to uncover indisputable evidence of fraud by January 12th? It's been 4 years since 2000 and we still don't have definitive proof of what exactly happened.

Personally, I plan on hoping for the best but planning for the worst. We have to prepare for the next election, help push for election reform, and be better prepared to fight next time. The worst thing we can do now is throw all our eggs in one basket and claim vote fraud and be wrong.

I wish John Kerry came out on Nov. 3 and said "FUCK YOU DUBYA! John Kerry is coming to take over your office!" and I wish there was indisputable evidence that 3.5 million votes were illegally given to Bush instead of Kerry. It might be true, but I'd rather not throw around ridiculous numbers and claim them as proof of vote fraud.

I am not putting any eggs in a basket. Any election 2 or 4 years out is an excuse for some kind of reasoning to accept was just handed out. I don't but can understand the need to correct a wrong. Understanding that we not dealing with things we will never be able to control until we include everybody.

Having the last Mid term elections 39% participated telling me that people are just way too comfortable and don't feel much trouble is ahead. If you cannot motivate them to act in their own interest, you must wait till something does.

however, if we do not cast massive doubt on this election, we will never get open sourced, transparent, fully paper validated elections ever again. Democracy, or our crappy ass barely Representative Republic version thereof is dying in black boxes made by Diebold, EE&S and Sequoia and unless we scream loud and long, RIGHT NOW!!!!, it is over, my friend. We can pack up our toys and take it on home and it doesn't matter how we reframe the discussion, if these voting issues are not brought out into the bright light, right now, Jesus H. Christ himself could run with Saint Christopher as his running mate and as Democrats, they wouldn't win. Get it? It's rigged.

I gave up on Kerry on November 3rd. This isn't about Kerry!!!

I am certainly no mathematician or statistician so cannot offer a comment one way or the other as to the merits of TIA's analysis. I do strongly disagree with c-macdonald's comment about hoping for the truth, however. We cannot just sit back and watch, and hope that the truth will come out, and in the meantime work on changing the political debate. Time is of the essence here! If the truth doesn't come out NOW, the Dems will never win again. Yes, the Dems. definitely need to work on framing the issues better, etc. But, remember, the repugs OWN the voting machines. If they get away with stealing this election, they're certainly not going to insist on verifiable paper trails for the voting machines, ever. Remember, it was only the few repug leaders in Congress that prevented the verifiable paper trail bills from being voted on. If their plan works this time, nothing's going to stop them in the future. Being a bunch of Pollyannas isn't going to help us do anything, especially to take our country back. The Ukrainians seem to be more realistic about their situation. Perhaps we passive Pollyannas could learn something from them.

but I do think it's important to start planning on another 4 years of Cowboy Dubya in office. I think it's extremely valuable to hear all the stories of voting irregularities and vote fraud and we need to continue to push these stories so that somebody hears us and starts pushing for election reform. But really, the Dems will win again as long as we fight back and not only change the system so that they can't possibly cheat but convince enough people to vote Democratic so that it becomes impossible for fraud to occur. Remember, the best way to prevent fraud is to have a candidate win in a landslide...

Nope. I don't buy it. Even the worse news networks had better exit polling then your own. I mean, hello... CNN'ers had a look of "aww.. da..." when their own big$$ polling boards said Kerry, but reports coming into their offices said Bushit. So, bushit.

You don't have to like what I say, but to dismiss me as a "troll" is a bit childish, isn't it? Get a life.

Of course, of course. But what does framing issues, getting out the vote in 2008, reforming the election process etc. have to do with this thread? Why is it either/or? I am not seeing the connection you are trying to make here.

I am no statistician, but I do have a Ph.D. in a physical science and use statistics quite often. I took the liberty to graph out TIA's numbers. There was a definite "red shift" occurring from the exit poll to the "final" poll of the vote. Does this not look statistically significant to your statistician's eyes? It looks like election fraud to these eyes! I'm working on compiling the 2000 election exit poll versus vote poll numbers for a comparison. I think Rove and Co. only rigged Florida in 2000.



PTV

the numbers TIA presents do look like a comprehensive cause for invesigation of fraud. But my problem is the accuracy of the data, not with his analysis. These numbers look very suspicious to me and I have some questions before I'll accept his work as accurate:

1) Where did you get the exit poll data from?
2) How did you calculate the MOE and Standard Deviation?
3) What considerations did you give to the un-randomness of the data?

Answering these questions will go a long way towards legitimizing his work.

I agree that we do need to verify where the exit poll data came from. Mitofsky/Edison have been refusing to make their complete data set public and we are going on "leaked" numbers for the most part. I am very skeptical of some reports of "final" exit poll data since they have been tainted by applying some kind of factor to more closely match "actual" vote data. We will have to wait until Sunday when TIA returns so that we can have that discussion.

As far as the MOE and SD, TIA gave an example calculation in the original post. However, look at the chart above. I do not rely on MOE or SD in the chart - it is simply the difference between the exit poll and the "final vote". In a perfect exit poll, all differences would be zero. In a more real-life case, the differences should be small and clustered somewhat evenly around zero on both positive (favor Kerry) and negative (favor Bush) sides. What we see is a whopping disparity between how many shifts to favor Bush versus how many shifts to favor Kerry (41 to 9). Also, we see that Bush has many more states shifting to him in more dramatic fashion (let's say more than 2 points).

As for the randomness of the initial polling data, again, we will have to try to get information from Mitofsky/Edison. However, I do not buy any argument that there is enough inherent bias in the exit polling techniques to cause this "red shift".

As a statistician, can you suggest other tests that would be appropriate to apply in this case? I've looked at skew and kurtosis, but they didn't seem quite right for this case.

Again, your graph assumes that the exit-poll-to-real-vote comparison follows a normal distribution. If the samples were truly random then this would be the case. But the samples are not random (the precincts are strategically chosen) and people can and do refuse to participate in the polling. Reports have shown that the past few elections have shown Democratically skewed exit polls (both 1996 and 2000, as far as I have read) so really these results may not be as much of an anomaly as we think.

Regardless of what my statistical instincts are telling me, though, I do find it VERY ODD that so many results ended up favoring Bush by so much. I just wish we had that data from Edison/Mitofsky so we could all make more sense of it.

As for other tests, I really don't think doing tests on this data provides any type of real analysis because this data is sketchy and unreliable at best.

Come on, you said you were a statistician. You know that exit polling (or any opinion poll) is not conducted randomly. For example, if the woman/man ratio increases beyond the state ratio, then pollster will not poll women until the ratio gets back in line.

Don't you think that enough of a sample was taken that effects such as lying or not wanting to be polled are taken care of? Isn't that all part of the margin of error of the poll itself?

Still digging up data from 2000...

PTV

you would know the answer to #2 by looking at the data.

Either you are a statistician who did NOT look at the data, or you are exaggerating your credentials.

BTW, I'm NOT a statistician. But I have three degrees in mathematics, for whatever that's worth.

To wit:
The Margin of Error for a polling sample size n in simply 1/sqrt(n).

For a sample size of 900, the MOE = 1/30 = 3.33%
For a sample size of 1600, the MOE = 1/40 = 2.50%
For a sample size of 2500, the MOE = 1/50 = 2.00%

The standard deviation statistic used to calculate the 95% confidence interval = MOE/1.96

So that is how I derived the MOE and Std dev, based on the exit poll sample size for each state.

You refer to the "unrandomness" of exit poll data. That begs the issue. The question should be "HOW ACCURATE HAVE EXIT POLLS BEEN HISTORICALLY"? Who cares if they are random or not if they are virtually always accurate to with a fraction of ONE percent? This is in fact the case. Let me add (ad nauseam) that exit polls have been extremely accurate wherever they have been used in the world. In fact, the last three German elections have been accurate to within ONE-TENTH OF ONE PERCENT.

As far as OUR exit polls are concerned, they have ALSO been extremely accurate historically. After all, that's how Mitofsky earns a living, isn't it? And he's the INVENTOR of exit polling, isn't he? And doing them for over 25 years, right?

Why was there no question regarding the accuracy of exit polls prior to 2000? Hmmmm....

Remember FL 2000? Initially, they called it for Gore. But then a computer "glich" in Volusia county spread 16,022 of Gore's votes to third party candidates (I believe the Socialist got 10,000) and Voila! - they called FL for Bush.

It turned out, though, that the exit polls were right since 175,000 votes were spoiled (double and triple-punched cards), the great majority of them in minority precincts who voted overwhelmingly for Gore. The minorities left the polling booths thinking their votes were counted. Of course, little did they know...

Yes, the Exit Polls were right after all. Gore won FL by at least 50,000 votes.

PAST is PROLOGUE.

And, thank you again for all your enlightening efforts.

"Halt, Audit & Prove My Vote Counts, Now"

N= 1000 sample size

Sqrt (N)= 31.62278
MOE = 1/sqrt(N)= 1/31.62278 = 3.1623%

You get 3.1%.
I get 3.1623%.

Where's the beef?

I am a firm believer in the KISS philosophy: Keep the calculations in the model as simple as possible.

The simpler the model, the more robust.

Introducing extraneous calculations serves no useful purpose - in fact, it will obscure understanding the basics of the system we are attempting to model.

We should use the minimal number of calculations necessary to derive results which model the real-world problem at hand.

e=mc^2
f=ma

MOE= 1/sqrt(N)

But I digress...

You could have at least kept them there for reference sake.

First of all, to assume the data is random and follows a normal distribution is a stretch and that assumption really wouldn't hold up after rigorous analysis. But let's say the data is random and we want to get a margin of error.

A sample size of 1000 yields a MOE of +/- 3.1% using the formula:

MOE = +/- 1.96* sqrt< (1-p)*p / n >

where p is our estimated probability of an event ocurring and n is our sample size.

Assuming n=1000, p=.5, we get 1.96* sqrt(.25/1000) = 0.0309 = +/- 3.1%

To get an accurate MOE for your data, you have to calculate each state using the sample size (n) and the percentage of voters for Kerry (p) and that would give you a margin of error at a 95% confidence level.

You can only use MOE = 1/sqrt(n) if the percentage you're measuring is unknown (i.e. you haven't gotten the results yet and want a ballpark figure). Since we have actual percentages for our samples, you have to use the formula I gave above.

With that said, it doesn't change the numbers too much but I'm not sure where you got the Std. Dev. = MOE/1.96 formula from. The MOE takes the Std. Dev. into consideration and, in the formula I gave, already factors in the Std. Dev.

So let's take Ohio. The MOE would be:

MOE = +/- 1.96*sqrt< (.521)(.479) /1963 > = +/- 2.21%

So that means his vote total should be between <49.89, 54.31>

but you can't put "simplicity" over "accuracy" when it comes to calculating these numbers.

Using the formula I gave for MOE yields much different results (even though our values don't differ greatly)

My analysis yields 9 states that are outside the MOE, not 16. That changes the results already.

But, like I said before, this analysis is useless because it assumes a normal distribution of the exit poll data and it's very clear that the exit poll data does not follow a normal distribution.

He's already suggested he missed the odds by a factor of trillions. What's a little more inaccuracy? At least it makes racy headlines.

tia

than what does your MOE represent?

What is the distribution if not Normal?

The MOE is related to the standard deviation which is input to the normal distribution function to compute probabilities.

Your argument that the exit polls are not normally distributed HAS NO BASIS IN FACT. You are creating pure mathematical fiction.
What do you KNOW about exit polls? Even Mitofsky talks about the MOE for his exit polls.

You keep throwing out strawmen, and I will keep knocking them down.

You ask, where do I get the MOE from? As a statistician, you should know that the MOE is the interval which coincides with 95% confidence limits. The 95% confidence interval is exactly 1.96 standard deviations on either side of the sample mean.
So the MOE = Stdev * 1.96 and the Stdev = MOE / 1.96

Along with your MOE's, how about some probability calculations? Again, show ALL the data for ALL the states. Just like I did. In fact, why don't you reproduce my analysis with your MOE's. Be sure to include the probabilities of deviaitions (within AND beyond the MOE) for each state, as well.

Finally, assuming your contention that "only" 9 states fell beyond the MOE (which I don't accept) the odds that this occurrence is due to chance alone is:

0.00004766% OR 1 out of 2,098,096.

Because the MOE for exit polls is much smaller than the MOE in standard polls. The corresponding variance (or standard deviation) around the sample mean is also much smaller. Exit polls have been historically much more accurate, all things being equal, than pre-election polls. MUCH lower than ONE percent. Read up on the last three German exit polls.

So all this talk about MOE is basically moot. But not for the reasons you suggest. The MOE's are actually MUCH SMALLER than those I have used. That means an even smaller probability that the deviations could have been due to chance alone.

for 2002? 2002 still smells, after all this time. I do not recall a discussion of exit polls vs actuals in states such as Georgia, at the time. If I did, hope someone fills me in.

Comparisons of 2000, 2002 & 2004 might be helpful. Be nice if the data were available.

Nice graph, PTV.

<snip>There is also the theory of oversampling Kerry voters which, while not a prevaling theory, probably does have at lest a slight effect given that Republicans are likely to ignore a pollster than a Democrat (since the pollsters are decked out in ABC-CBS-NBC-AP-etc. apparrell and over 50% of repubs don't trust the media, compared to just 38% of Dems). <snip>

Yet... the overwhelming 'reason' given by MSM was that "Moral Values"
was important to voters -- which shocked a lot of correspondents. Is it possible, that the Republicans gave info about the things they were most worried about, but didn't tell who they voted for? Anyone have the Numbers for how many picked Moral Values comment And told who they voted for, -vs just giving that reason without telling who they voted for?

just wondering.

thanks.
tracy

That Republicans were less likely to talk to pollsters than Dems. Has anyone ever heard that theory before this election's exit polls contradicted the election results? I don't think so. And I certainly don't think that it's plausible to say that Republicans trust the news media less than Democrats.

When the polls were so skewed, that is what the correspondents said -- right after the election, and I just heard a commentator say that Friday Morning, too.

My question has to do with the data we have now -- OF the republicans that stated "moral values" as something very important to them, how many told they voted for Bush?

Or perhaps, of the Voters that stated "moral values" as something very important to them, how many told who they voted for?

My question has to do with the excuses the MSM was giving at the time -- I don't think it can actually be answered. In one sentence they would say that it was the "republican men" who voted after 4 that turned things around, in another, they say that Republican shy away from media-types and don't give their opinion, and in another they say that Moral Values was the deciding factor in this race....

just wondering if there is any way to crunch those numbers, that's all...

tracy

What the MSM wasn't telling us is that "moral values" actually was named by less people in 2004 than in 2000 and 1996 (35% and 40%, respectively).

As for your question, the data for that analysis just isn't available and we really won't have any answers until the actual data is released.

1) why not do the adjustments you recommend and post them here? yes, it would require more time from you, but if the MOE is corrected and found to be MUCH smaller, don't you think it would make your argument more compelling (more compelling than simply more posts)? On the other hand, wouldn't it be informative if the corrected MOE is not much larger than TIA's estimate?

2) do you have any empirical evidence that Republicans really are less likely to respond to exit pollsters than are Democrats? Even if they are, that seems to be something that could be estimated and corrected as well, by weighting the resulting samples.

In other words, rather than have the two positions seem extremely different and black-and-white (exit polls are worthless vs. exit polls are absolutely accurate), why not resolve some of the differences between TIA's and your positions by doing a little more work and then people can judge the situation more easily?

And most people will probably be somewhat intimidated by it.

As I understand it, the exit poll variance from actual results is highly selective, so speculation on the relative accuracy of exit polls in general, or the exit polls in this particular election are not relevant that I can see.

Looking at 40 years worth of exit polls, why did they fail so dramatically only and always where and when they needed to for a particular outcome to occur?

To demonstrate your point, would one not have to account for the relative accuracy of the exit polls within this election from district to district? Why did they fail in the swing states, and yet perform per expectation elsewhere, and as far as that goes, else when?

Thanks for the warning to not believe everything we see, and for the advice to think about framing issues.

Thanks c-macdonald, I can only add that estimations of probability are only as good as the underlying assumptions.

To get back to why the exit polls could be skewed, one reason could be that if the Bush voters have the "F yourself" attitude toward some college kid asking them how they voted, then there will be a Kerry bias in the poll data. Another could be that they were unwilling to publicly say they voted for Bush because of shame. Another reason might be that democrats were so eager to get Bush out that they rushed to vote as soon as possible to get him out of their system, whereas the bushiites waited for rain to stop (remember, it was cold and rainy in Ohio), then went to their nice precincts equipped with ample numbers of voting machines and voted late.

Therefore, exit polls actually do mean nothing, sad to admit. A couple of percent this way or that could well be explained by various biases.

Like I said, we need to focus on the actual votes. I don't believe that
papertrail has been completely erased. I do not believe the machine cartridges (flash memory, what else could they be?) have been erased. We need to make sure that there is still a right to vote in this country for us and our children. This is not about Kerry anymore - it's about the future.

I'll take any help I can right now, be it from GAO or GLibs.

To thereismore:

You write, "To get back to why the exit polls could be skewed, one reason could be that....," then you give several 'could-be-thats' (Republican shame at voting for Bush, etc.), and finally you say, "Therefore, exit polls actually do mean nothing...".

You provide no proof--not even a hint at proof--of any of your "could be" and "might-be" reasons for the exit polls being skewed, then you make the grand conclusion that the exit polls "mean nothing."

Exit polls use well-tested methods and sampling techniques and are greatly respected and utilized throughout the world to verify elections and to detect election fraud. There is simply no evidence pointing to any error in the 2004 exit poll numbers. And you are merely speculating as to why they differ from the Republican-owned and controlled (and top secret) vote tabulation source code "results."

And there is PLENTY of reason to doubt those Diebold (CEO Warren O'Dell a Bush "Pioneer," i.e. donated over $100,000 to the Bush campaign) "results."

Non-partisan, highly respected exit polls used worldwide
vs
Electronic vote "tabulation" completely controlled by partisan Republicans.

Are you "in denial" or what?

But you are correct that "this is not about Kerry." This is about the people who stood out in the rain for 10 hours to vote for him, and gave up their jobs and all their time and energy for the voter registration, "get out the vote" and poll monitoring campaigns--the people who gave Kerry and the Democrats an overwhelming victory, and have been cheated of democracy's finest hour--the overwhelming repudiation of George W. Bush.

It's about US! Yes. And what we're going to do next. Bury our heads in the sand, and watch our democracy DIE? Or fight back, by: 1) supporting the recounts and lawsuits and investigations that are on-going, and do everything we can to overturn this election; and 2) work state by state to achieve, a) a paper trail for every vote, and b) open source code for vote tabulation.

See http://www.blackboxvoting.org, http://www.votersunite.org, http://www.votecobb.org and http://hungerfordemocracy.com

Ever since it became clear that there was a very good chance that this election had been stolen, I've felt most outraged about the fact that all the people who worked so hard and endured so much might have been cheated out of the right to have their voices heard and validated. While I also feel strongly about wrongly denying Kerry a fair win, this is really about the people, not the candidate.

I'm not saying fraud didn't happen. I'm just saying these numbers don't prove it.

Okay, here is why I said that "exit polls actually do mean nothing":

Assume for the sake of argument that the election was Kerry 50: Bush 50. If only one out of every 50 republicans had lied at the exit poll and said "I voted for Kerry", the exit poll would come out to be Kerry 51: Bush 49. That's how little is needed to throw off the exit poll by 2%. Is that how it happened? I don't know. But you bet your boots that if this ever came to court and you built your case upon the exit poll, the repug defense would come up with 100 repugs who will say "I lied at the exit poll", and your case will go down the tubes. That's why exit polls actually mean nothing unless you can back it up by hard data. Those are in the flash memory sticks
and paper ballots, as well as in the sworn testimonies of disenfranchised and intimidated voters. Forget the exit polls, they prove nothing. They just tickle our emotions. We need to beat them with facts. For all those who waited in the freezing rain, and for those who might again someday dare to think that their vote counts.

You're making claims based on your speculation.

In your hypothetical, if people lie to exit pollsters, why would only Republicans lie? What is anyone's motivation for lying to a pollster (rather than tell them they prefer not to participate)? And if your Republican witnesses would lie to pollsters how credible would they be in court? Besides, what might happen in a lawsuit is not the primary concern at this point. We're talking about finding the truth about what happened in this election and what is needed to make corrections going forward.

There are always errors or weaknesses in data--including errors in your preferred sworn testimonies and paper ballots. But to presume that some error renders the exit data useless and that the electronically recorded voting results are free from error (even after considering the information others have offered here and elsewhere as to why to be concerned about those results), is not good empiricism.

invalid.

We need actual evidence of biases in these data before concluding "exit polls actually do mean nothing."

Please see the Freeman report for some interesting evidence concerning exit poll data.

Your points as to why the exit polls were biased to Kerry:

1. Bush voters were too ashamed to say they voted for Bush?

2. Bush voters have the "F yourself" attitude toward some college kid asking them how they voted?

3. Democrats so eager to get Bush out that they rushed to vote as soon as possible to get him out of their system?

4. Bush voters waited for rain to stop (remember, it was cold and rainy in Ohio), then went to their nice precincts equipped with ample numbers of voting machines and voted late?

5. Because of all these reasons, exit polls mean nothing?

6. A couple of percent this way or that could well be explained by various biases?

Sorry. Your bias is showing.


You guys have no shame, do you?

Analyses of the Votes show the same pattern compared to 2000 vote and number of newly registered voters by county show the same unusual patterns of more votes for Bush in official results than would be expected.

Some examples:

Florida www.flcv.com/EAS.html

www.computerworld.com/governmenttopics/government/polic...

Ohio http://www.opednews.com/shurberg_112004_election.htm
http://web.northnet.org/minstrel/summary.htm

North Carolina see the absentee vote study on this site

....as it seems not to have made it into your comment.

Peace.

See these posts:

http://www.democraticunderground.com/discuss/duboard.ph...

http://www.dailykos.com/story/2004/11/12/233831/06

http://www.dailykos.com/story/2004/11/23/707/32691

http://www.democraticunderground.com/discuss/duboard.ph...

Actually, exit polls don't count anywhere. What counts in Ukraine are hundreds of thousands of people in the streets. The politicians, judges, etc. only worry about the power of the people. The people care about the exit polls, the politicians don't.

collected up to 5pm EST at the latest. Trying to draw conclusions from this is like predicting the winner of a baseball game to be the team that has a one run lead after only 5 innings.

...effectively than most on this little Mitowsky-chestnut you've asked all of us to chew. Here's the link:

http://www.bradblog.com/archives/00000992.htm

In case you don't want to read Brad's commentary then here's a Cannon-chestnut for you:

Joseph Cannon writes -- "No, Mr. Mitowsky. They should be compared to the score at half time at fifty football games. Better analogy: Fifty tosses of a coin. Error should skew in both directions; if the coin keeps coming up heads, something is wrong with that coin."


Enjoy.

Peace.

in turn had this link http://www.metrotimes.com/editorial/story.asp?id=7028 which said, in part:

First we found Simon Jackman, who urged us to stay off the street mob idea.

Jackman is president of the Society for Political Methodology, as well as director of graduate studies for the political science and statistics departments at Stanford University. He contends that Freeman’s math is fine, but that his analysis may be based on a faulty premise. Freeman’s numbers only hold value if the exit polls are accurate, and Jackman says there is reason to suspect that the polls are significantly flawed.

“At this stage, we have a puzzling discrepancy but, alas, more than one plausible hypothesis. We need more data and analysis to sort out the possible explanations, and that will take time. But almost every serious scholar and political operative I have spoken to on this issue believes that we’ll discover problems with exit polls, not large and widespread electoral fraud.”

....who are his 'political operatives and serious scholars' and what basis in fact do they basis their 'beliefs.' Freeman and others worked with the available numbers. And, the M&E folk are not rushing to provide what they have. So, someone hiding something regarding what voters had to say as they left the polls -- about themselves and their votes.

All of which is entirely secondary to the simple fact that, with perhaps a very few exceptions, if those who voted were to request confirmation if and how their vote was counted they would likely not be satisfied with the response -- since a precise accounting of what happened to their vote is either exceptionally difficult to find or doesn't exist.

"Halt and Audit" -- let's, in_deed, stop speculating and gets some facts on the table; that's the least all of those who have died for our franchise deserve; don't you agree?

Peace.

"Error should skew in both directions; if the coin keeps coming up heads, something is wrong with that coin."

Well statisticaly (and realistically) speaking, there isn't something wrong with the coin, there's something wrong with the toss. Coins cannot be biased so as to land heads (or tails) more or less than 50% of the time. The bias comes from the toss.

Cannon's reasoning is absolutely correct but only assuming the exit poll data is a random sample - which I can assure you it certainly was not random.

A better analogy is to have 50 people flip a coin 50 times and then only asking 10 of them to report the results of their first 25 tosses. Basically, it's meaningless. The only exit poll data that IS useful is the final uncalibrated data that Edison-Mitofsky claim will be released in 3 months.

While I cannot claim to speak for all non-Americans, I am personally grateful to people such as yourselves who have taken such a passionate interest in protecting the spirit of the American system. This is especially true since the entire world will be affected by any decisions made by the US yet the entire world is powerless to change the direction of American Politics. I have followed this election closely and I have read virtually everything available regarding the allegations of election “irregularities”; however I respectfully suggest that the real issue is not being addressed by the nation as a whole. The election process in the United States is obviously not trustworthy and therefore the recent election can never be fully respected, even in the absence of damning evidence.

Let’s briefly review the facts:

1. It is possible to hack into at least some of the central tabulation machines used in the recent election and alter results without leaving any evidence.

2. There have been fundamental violations of at least common sense regarding the selection of people with massive influence over the vote counting. The widely publicized comments of Diebold’s Walden O'Dell prove that at least some people of great influence within the system are not opposed to manipulating the votes and disenfranchising American voters.

3. There are no primary source records of large numbers of American votes due to touch screen machines without paper trails.

4. The only barrier to Republican power in the United States has been silent on these and other issues throughout the election process. Democrats continue to participate in the system that COULD so easily manipulated and validate the results of the election by doing so.

The fact of the matter is that Americans have all of the evidence they need to doubt the US electoral system. Virtually everyone agrees that the elections could have been manipulated with no evidence. Now as the information is being to become available everyone seems to have moved on. CNN opinion polls show that Bush’s approval rate has risen since the election. Even people in this forum have resigned themselves to the prospect of another Republican term. Meanwhile people in a country of practically no global significance are in the streets because of a fraudulent election.

The American Election was so poorly designed that short of Bev raiding Whitehouse dumpsters and finding “Please Rig the Election” notes from Mr. Bush, there will never be a CNN newsflash stating “THE ELECTION WAS STOLEN”. The American system requires voters to prove that their vote was not counted in a structure that does not preserve individual votes. “49%” of your voting citizens support a party which refuses to acknowledge the concerns of the voters who “almost” put them in power. Unfortunately, unless the Democratic Party addresses these issues in a unified protest you have already lost your democracy. If the other party doesn’t stand up for you in times like these, why have another party at all?

But the only way to uncover fraud and PROVE it sucks is to actually investigate the machines, the registrations, and all the other aspects of voting fraud that can be investigated. Exit polls can't provide proof of vote fraud any more than Rumsfeld saying Iraq has WMDs provides proof of WMD's in Iraq (well, OK, maybe a little more).

We do need to fight for a transparent, impossible-to-fraud election system and I will support that cause and do everything in my power to advance that cause when it comes up. But in the meantime, I'm getting prepare to fight this administration for the next 4 years with everything I've got.

I second everything you said. Couldn't say it better myself.

“…the only way to uncover fraud…”

“…exit polls can't provide proof of vote fraud…”

In financial prosecutions where there is no "paper trail" (say, in a case in which the suspect shredded the paper evidence), statistical methods are admissible in court to prove a case of fraud. I would imagine proof may be more difficult in such a case (I’m not a lawyer). The point is--and I've heard this now from several sources--the kinds of statistical studies that Freeman and the Berkeley group and TIA have done ARE admissible in such a case.

Another reason to pursue them is that they can be helpful in identifying “means, motive, and opportunity” (the classics of detection), and in locating harder evidence and identifying actual suspects (which "black boxes" to look at? who had access to them? etc. --also, as someone upthread mentioned, discovering by inference the formulae that were used to change the election results).

The Berkeley study wasn’t about exit polls--it was about electronic voting vs. traditional voting methods. They found over 200,000 “excess” votes for Bush in FLA on this data alone (votes that were either manufactured or stolen from Kerry).

ignatzmouse ‘s study of NC was about the 1/3 absentee vote vs. 2/3 electronic vote, with a big discrepancy from the exit polls limited to electronic voting.

Both of these studies don’t just point to fraud, they go a long way toward establishing the “means” (i.e., the murder weapon)--electronic voting.

Continuing efforts to prove the fraud are important for many reasons.

1. To overturn the election, if possible--an effort that would obviously be beneficial to all, but is a long shot, contingent not just on evidence, but also on politics and public perception (--and has a specifically limited time frame in which to be achieved).

(But note also: Watergate happened AFTER Nixon was re-elected in 1972. True, we don't have good conditions for impeachment, but we are in a far better position than anyone of that era to figure out what happened and to get it widely known and understood--through the internet. This is a pivotal scandal--much more serious than Watergate--with, at this point, unknowable consequences for Bush Inc.)

2. To uphold the principles of democracy--that every vote should count; that Republican partisans shouldn't own and keep secret the source code by which all our votes are counted; that an election should be re-countable and verifiable; and that, without these things, we really don't have a democracy.

3. For our own safety --to let the world know that American voters did NOT endorse Bush and his murderous war against Muslims and Arabs (even short of a slam-dunk case, we should at least be letting the world know that there are huge red pointers to election fraud).

4. To spur an effort to save our democracy--which I think requires grass roots citizen effort, state by state, county by county, to achieve, a) a paper trail, and b) open source code. (Congress is NOT going to help, and could make things worse.)

5. To hearten Kerry voters and campaign volunteers across the country, let them know that the election was stolen (not that the country went “red”), and focus them on achieving #4. (#4 is doable, but we may have a limited time window to get it done.)

6. To keep all those newly registered voters active politically, and voting. Right now, what reason do they have to ever vote again? (Even solid, long-time voters are asking that question. But when I tell them what I've learned about all this, their eyes light up, I'll tell you that. Suddenly they UNDERSTAND what their gut has been telling them--that something was very wrong about this election--and they want to do something about it.)

7. To hold the Democratic leadership’s feet to the fire for their catastrophic failure to protect our right to vote--permitting Republican partisans to control the vote tabulation software, with no paper trail. At the least, they should have WARNED voters about this, loudly; talked about it in the campaign. What the hell is the matter with them?!

8. To avoid wasting our time on the delusion that we can influence this fascist coup on the many different issues that they are important to us. It’s scattering our energies; it’s useless--they are NOT into cooperative government; they are on a jihad to destroy everything we hold dear. Stick to fundamentals: unless we recover the right to vote, our democracy is OVER.

Everything so far points to a fraudulent election result. In addition to the growing body of statistical evidence (and, really, it’s gotten to be overwhelming), there is a great volume of anecdotal evidence of severe vote suppression by Republican election officials in certain places--the places the election SEEMED TO come down to (Ohio, Florida), while the main part of the fraud likely occurred way back upstream in less noticeable states, and by the less noticeable method of altering the source code in the central vote tabulation machines in many places.

The very blatant vote suppression by Republican election officials in Ohio and FLA is not the whole story (in my working hypothesis of this election); it appears to me to be part of the larger fraud plan, and something of red herring. (I don’t mean we shouldn’t pursue every “provisional” ballot--what I mean is we shouldn’t feel DEFEATED if those ballots don’t overturn the election. The whole thing seems to me to have been deliberately aimed at making it APPEAR to turn on a relatively few “provisional” ballots in places where highly partisan Republican election officials were in charge, had done a lot of prep in vote suppression, to keep it close, and have official control over which ballots are counted.)

There is also a lot of anecdotal evidence of extraordinary effort by pro-Kerry workers to register voters and to get out the vote, and the perception of many, many people that those efforts were highly successful. Peoples’ rightwing uncles voting against Bush. Like that. Many, many such stories.

Finally, there is solid evidence that this election was a fraud GOING IN: Bush partisans (major donors and supporters) owning the secret tabulation source code that was used to count all our votes; the same people insisting on no paper trail (and achieving it in a third of the country). Made to order for fraud.

One of the most important keys to understanding what happened on Nov. 2 is the TV networks’ handling of the exit polls. People need to be informed about this: that the exit polls showed a Kerry win, that TV networks then started mixing exit poll data with Republican-controlled electronic machine data, making it APPEAR that Bush was winning late in the day, and COVERING UP the big Kerry numbers in the exit polls. The TV networks were key to creating the PERCEPTION of a Bush win late in the day. They FAILED TO DISCLOSE that they had polluted the exit poll data! (A Bush late-in-the-day win made little sense, intuitively--that’s when the working poor and minorities vote, by necessity. But they didn’t mention this either.)

The more we uncover about this fraudulent election, the more it helps fight cynicism, and depression and feelings of powerlessness. We need to FOCUS on this FUNDAMENTAL PROBLEM: the ever-growing evidence that the election was stolen, and what to do about it. This phony division between the “red” states and the “blue” states is causing a lot of distraction. The fact is, WE DON’T KNOW how the “red” states voted. They could have repudiated Bush, for all we know (and I strongly believe some of them in fact did).

You are probably right Peace Patriot about admissibility of statistical "evidence". I am a statistician and we call it statistical inference. This election stinks to high heaven. There is enough evidence to pursue, like what you mention, the machine bias in FL, etc. That's why I've been saying that we need to beat them with facts. Vote counts are facts - that's how they were recorded,
whether fraudulently or not. We need to go after the paper trail, the original flash memory from the Diabol machines, and start from there. The exit poll discrepancy is a great motivator, but we do have more and better evidence, which is the vote counts and testimonies of disenfranchised and intimidated voters.

I just recalled - do you remember when they came up with some astronomically small numbers for probability that the blood on the murder scene wasn't OJ Simpson's? Do you know what happened next? The defense came up with statisticians who questioned the assumptions of that calculation, such as the mean frequency of some traits in the black population. That is exactly our problem. Our equivalent of that mean frequency is the exit polls. Undermine that and we fold. No, we must not be distracted. We must go after the facts: vote counts stratified by technology/county/%party registration, plus testimony of election fraud, such as vote suppression. It might not overturn this election, but it might help us get a decent election reform bill passed, which is even more important in the long run.

Check out this article from Oklahoma. A report in a major newspaper based on 70% of the actual vote COUNT in Oklahoma, showed Kerry with MORE VOTES THAN HE RECEIVED IN THE FINAL COUNT. The machines counted backward for Kerry over a certain point. This is an extreme claim, but it obviously needs to be investigated, because, if true, it blows the whole cover off the Black Boxes. (All OK votes were counted on ES&S Opti-scan machines.)

Link http://okimc.org/newswire.php?story_id=344