Your scenario would be valid if everyone participated in the Exit polls that voted. But since it is only a TINY sample it will not be comparing in the same manner as your scenario.

Could you specify what you mean by "my scenario"? Are you referring to the traffic ticket analogy?

Erasmussimo, as a point of statistics (which I am also familiar with, as an engineer) which data set is a more reliable predictor of outcomes the larger set or the smaller set?

Second point: when measuring people opinions, which data set is more reliable: the anonymous data set or the data set where you must reveal your identity?

Allow me to answer: larger data sets are universally more accurate. Also, when measuring opinion, people are almost universally more honest when they believe they are anonymous.

So we have two sets here. The first set (the Edison/Mitofsky exit poll) was a data set of 11,719 voters. NATIONAL ELECTION POOL EXIT POLLS There is no way (that I am aware of) to determine which of these data points may have misinformed the exit poller. Certainly there would be no surprise had some of them done so.

Conversely, the actual election votes totaled over 121 million (121,000,000) votes. CNN Election results That is four orders of magnitude larger. I will leave it to the reader to determine which data set is more reliable.

When voting, your decision is completely confidential. At least it was the last time around. Some organizations (such as US Count Votes) want to loosen that anonymity in favor of tracking who voted for whom. Such a thing will lead to (among other things) cases of "buyer's remorse" where voters want to change their vote the next day. Anonymity makes things less prone to peer pressure.

On the other hand, exit polling is not anonymous. You may not need to reveal your name and the form may have a large "CONFIDENTIAL" on it link, but that doesn't mean that people feel anonymous when filling it out.

In general, the larger data set with anonymity will be more accurate.

If you look at the Edison/Mitofsky exit poll, in 30 states the actual results differed from the polls prediction by more than one standard deviation. Source



There is something fundamentally flawed in their data collection techniques. The most like (in my mind) is that their data pool was simply too small. They asked, on average, every 10,000th voter who they voted for. That sample size may have been simply too small to accurately predict the results.

Also, keep in mind that statistics can only predict outcomes, they cannot guarantee them.

Back to the statistics textbooks for you! For two data sets, one with N1 data points, with each data point having a sample standard error of s1 and the other with N2 data points, with each data point having a sample standard error of s2, the standard error of the first data set (assuming a normal distribution) is s1/SQRT(N1) while the standard error of the second data set is s2/SQRT(N2). Thus, if N2 > N1, the first data set can still yield more reliable results if s1/SQRT(N1) < s2/SQRT(N2). This is freshman level statistics -- your objection does not even rise to the level of sophomorism.

Your speculations as to the nature of the data gathering problems were discussed, considered, analyzed, examined, argued over, and dealt with many years ago. I don't claim to fully understand the fine points of this study and its critiques, because I have some intellectual humility. You folks are making grand statements about matters about which you know very little. Don't you think it's time for a little intellectual humility, or even realism? We're all too dumb to truly understand what all this really means. Your pitiful attempts to punch holes in a study that is completely over your heads says nothing about the study and everything about your intellectual integrity. So why don't we shine a little sunlight onto this problem, get more people working on it, get some more data, and resolve it with some solid data and analysis rather than blowing ignorant hot air around?

(with hand up) Teacher! Teacher! But isn't the standard deviation of the measurement of the entire population....zero? There is no deviation...it is known with certainty....I imagine that little aspect got all cloudy with all the hot air in your vicinity. Funny how, even though we know oh so very little...we still manage to find holes in your little theory here which your intellectual humility prevents you from addressing. But, then again....this sort of rational thought is probably out of your league.

This will of course probably get me kicked out of class. Fortunately, I don't need this course to graduate.

Oh, by the way, here on planet earth we try to debate such topics civilly. Clearly, this is not the case on your planet from the belittle zone...so just wanted to get you up to speed on some of our customs, as you clearly seemed to have missed that in your acclimation training. Obviously, our intellectual capacity isn't up to snuff, but we can get rather fiesty when repeatedly provoked.

Further, I've yet to hear of the causal mechanism this study points to. As you yourself said (here, let me remind you lest you forgot)....



Again, simply crying out Electoral Fraud! does not a causal mechanism make. But I imagine the authors of this study can join us in our freshmen level classes once they've crossed that bridge. Or, better yet...we'll just get our grand conspiracy machine cranking again, and simply make them and their study disappear. You don't honestly believe that any group with the skill and audacity to steal a Presidential election would have either the slightest compunction against or difficulty with making a few statisticians disappear, do you? You call for further investigation....why wouldnt' the conspirators be able to derail that, as well. What would make an investigation so much harder to 'rig' than an entire national election? But, its clear that such investigations will be called for until one comes up with the desired results. Frankly, this is a good thing, and falls perfectly into our plan. For as long as 'they' are busy studying the past election...they're not devoting time and energy to improving their results in the upcoming elections, are they? Yes, yes...it's all coming together perfectly....soon, they will realize the full force and power of the dark side!

I'm beginning to come around on the investigation aspect on this study....



Far too much emphasis has been placed, of course, on the first alternative suggested by the authors (one which, I might add, clearly shows their bias). The topic of this investigation should be "Why Election Exit Polls Fail to Properly Predict the Actual Outcome." This is naturally not really something that Congress would need to be involved with, as it has to do solely with a particular industry and is essentially just an academic discussion, and has no national impact.




Here's the problem I have with this. This simulation is essentially data massaging. Without some proposed mechanism explaining how such a vote shift could have occurred outside of the simulation, there's nothing to investigate. Especially when a much more plausible explanation exists....there is a problem with exit polling data. This is a known fact, and equally explains the discrepancy. Why waste time investigating the implausible and unknown, when the known and plausible already provides the explanation?



A 6% vote shift? That would be approximately 7 million votes. Are we expected to seriously consider that this is what happened (note that nowhere in the paper do the authors even make this suggestion)? Again, what is more plausible....7 million votes being manipulated, or a 6% poll response error (this error doesn't need to be the one suggested by E/M)?



I could just as easily simulate the effects of shifting voter behavior do to attacks on voting day by alien beings, due to earthquakes or other natural phenomenon, etc. etc. These are all hypothetical also. Only someone who was predisposed to believe voter embezzlement occurred would single out that one hypothetical for investigation. Without going to the extremes above, I could create a simulation showing any result I wanted with the 'People don't respond to exit polls accurately' hypothesis. This hypothesis has a lot more data to back it up (review the poll/actual results listed here in this thread), and could recreate the exact same results as this study (or any desired results, for that matter). Why presume 'vote shifting' when 'poll response shifting' is much more plausible, and has extensive data to support it? This is essentially a hypothesis looking for data to fit it, which violates one of the rules of scientific study, which is that the hypothesis must be proposed before the data is gathered. Further...isn't this something that should be handled through law enforcement, and not Congress? Why not take the argument there? If the hypothesis is true, then clearly having Congress investigate would be like having the Mafia investigate its wrongdoings.



...and we do this....why? Again, why assume 'vote shift' when 'poll response error' would achieve the exact same results. As many here in this thread have stated, this paper is based upon the assumption that poll response data is more accurate than the actual vote. There is nothing in this paper that provides any support for that assumption.



Ahhh...here we have the only real conclusion that that can be drawn from their analysis. How you get from non-uniform poll response rates to massive voter fraud is undiscussed. In fact, definitively ruling out uniform bias is not even done. There is nothing in this paper to support the conclusion they drew. Here's what happened in this paper....they took the 6% response bias proposed by E/M (56%-50%), and changed it to a 6% vote shift. Viola.....the data match! Remarkable? Hardly. Newsworthy? No. Anything the 'man on the street' should even be made aware of? No. This is an academic discussion, analyzing one person's explanation of a data set. Valid conclusions regarding E/M's analysis are drawn, and unfounded assertions are then made.



...yet we really have no data to back this assertion up. We have simply performed a simulation in which a vote shift gets one poll data set to match up with another, and we completely disregard the known issues with poll data. Other, much more plausible explanations for our stated discrepancy are available, but don't suit our purpose, yet we still make these wild accusations.

Now, on to their recommendations....



I can get on board with these. No need for unfounded 'doomsday' scenarios for that. In fact, none of these recommendations really has anything at all to do with their study. I'm not sure what they expect to see in the database...without any way to match up people's poll results with their actual vote, polling data will always be suspect, and no means to reconcile that is available. So, you can construct the biggest database you want of poll results, and no definitive conclusions regarding vote count accuracy could be drawn. Ditto for the need for a thorough investigation of the past election. Without the ability to draw any definitive conclusions, what really would be the point? Now, if that same investigation is process-focused, you might get somewhere....but that is already under way.

So, I've read the paper, and nothing I've seen in there shows me any evidence of any grand conspiracy. The authors don't even make the assertion anywhere other than in their conclusion (which I must say is very convenient for them). This is simply an academic exercise showing various mechanisms for getting data sets and simulated data to match up. Wonderful academic exercise, with little practical application. That said, I don't have an issue with most of their recommendations.

No, there are errors in the counting process. The best demonstration of this fact is that recounts always yield different results.


One thing at a time. To insist that the entire story be made clear before an investigation begins is merely a subterfuge to prevent the story from ever being discovered. All that needs to be demonstrated at this point is that electoral fraud is plausible. That has been accomplished through demonstrations of many security failures in the vote-counting machines.


The use of Monte Carlo methods to examine the behavior of complex stochastic systems is an old and trusted procedure.


You are objecting to a mathematical procedure because you don't like the numerical results of the procedure. Logic requires you to challenge the procedure on its internal merits, not on its numerical results (unless, of course, you have other results that contest it, in which case you must examine the procedures leading to both numerical results.)


No, running hypothetical scenarios through a simulation is a means for examining those hypotheses. If you wish to examine the hypothesis of earthquakes, by all means do so and present your results. Such examination on your part would not indicate that you are predisposed toward or against earthquakes.


We don't assume vote shift, we hypothesize vote shift and examine the hypothesis. Big difference.


For the umpteenth time, I repeat that this study does not attempt to establish any mechanism for electoral fraud. It demonstrates that there is a significant discrepancy between the exit poll results and the actual vote counts, a discrepancy that cannot be explained by any innocent means.


The difference between "definitive ruling out" and "demonstrating" is a matter of degree. They are honest that their study does not constitute a smoking gun. I have already stated this.


I disagree. There's a lot of evidence and analysis there, and I find that the statistical analysis it presents is convincing but not compelling.

This is true. However, I find it very difficult to buy that such errors would be greater than the errors obtained through the use of polling data to predict/explain the actual results...especially since those same counting errors could also be present in the polling data, with their effects magnified through margin of error.



Pointing out that standard scientific procedure was not followed is clearly subterfuge? I'm sorry...you'll have to explain that one. As for the plausibility of the scenario....going from potential security failures in vote-counting machines (which I don't mind investigating and removing, btw) to carrying out 7 million fraudelent exploitations of same is a rather large leap to make without setting forth any process through which this happened. Which is why, of course, the attorney general is not taking this as a call to arms to investigate...there is indeed no evidence it happened.



I don't have a problem with the simulation, or their analysis. Where there is a huge, gaping hole is between their simulation (which only attempts to simulate sample data) and the actual results. Why are you so steadfastly refusing to address the much more plausible explanation that I have put forth...and that does have evidence behind it. Exit polls simply are not accurate predictors of actual results. Until that hypothesis can be ruled out, this study does nothing to warrant investigation of the actual results.


Why, oh why, is it so hard to have this debate without inferring my thoughts into it? I have challenged it on its internal merits...it is based upon potentially flawed data. I have further pointed out that the numerical results are only valid for the purpose to which the analysis was applied...that being to compare simulated results with sample data. If you want to use the simulation to question the sample data...fine, by all means, do so. In fact, that is precisely what I am saying should be done. I do have other results which demonstrate this....all the other exit polls which have been done. So, let me turn the tables on this...given that I have done this, and also demonstrated that all that was done in the study was to substiture one correction factor for another...and even though I have pointed out that the authors themselves don't draw any inferences for their conclusion anywhere in their own paper...and even though I have demonstrated that the magnitude of what their conclusion suggests in essentially beyond reason without any other evidence pointing to it...why do you still insist that I am the one clinging to my predispositions? All the evidence points to the opposite being the case.



Yes, there is a big difference. Note that nowhere in their paper other than in their conclusion do they even make the vote shift hypothesis. They simply show that one data set which assumed a 6% shift can be best approximated by a 6% shift. Hardly earthshattering...not even worthy of publication, actually, at least for the conclusions it draws (and it should be noted that apparently this has not been published)



Ahh, but of course it can, and I have in fact done so. Even the authors don't rule out (or even address, for that matter) such means. Which is why I ask, also for the umpteenth time....why assume grand conspiracy when there are indeed known innocent means which also fully explain the discrepancy?



The statistical analysis is fine...its the conclusions drawn which are being questioned. Does it not concern you at all that the authors conclude vote manipulation on a massive scale, and yet don't even have one single line in their paper even postulating that, much less defending it as plausible. I would think your intellectual integrity meter would start flashing somewhere along the giant leap they take from analysis to conclusion in the absence (which you admit and note several times) of any causal factors.

No, Erasmussimo, I am only left to conclude that I am not the one here being biased. I have examined the paper, and pointed out several areas of concern...none of which have been addressed. I have further pointed out, and not been refuted, more plausible explanations of their results than what they conclude. Most auspiciously, I have pointed out that the authors themselves don't even infer what they conclude anywhere in their paper or their analysis. Do they even state their hypothesis anywhere? No, they do not (at least not the one they apparently conclude). Why didn't they do that? I have a hypothesis of my own...because it would look so preposterous, if actually stated, that their paper would have been discarded out of hand. Further, because they knew that their analysis offered, at best, scant evidence towards that hypothesis. Why publish it then...especially with that conclusion? Because of the obvious political impact, which, I might add, you are very effectively demonstrating. But that doesn't mean that the paper's conclusions arent'open to criticism, when they so clearly deserve it.

One final point....note that E/M is of course not just a polling company, but the one that performed the poll analysis in question. It is not surprising at all that a polling company would seek to put forth some explanation other than 'Exit polls are not reliable', is it? But that, of course, does nothing to increase their merit. Ditto for the folks at UVC. They are composed of statisticians, who need such data to perform their analysis, and other pollsters. So, it is hardly surprising that neither side of this dispute has put this hypothesis forward. But that does not detract from its merit...particularly with such a long history of evidence supporting that very conclusion.

There is of course a very reasonable investigation that could be performed, and one which would cost nothing, and be completely under your own control. E-mail the authors, and ask them how they get from their analysis to their conclusion (since the paper clearly does not provide that information). Their response would either put an end to the need for any further discussion, or provide something further to debate.

Please specify what you mean by "standard scientific procedure" in this context.


We are talking plausibility here, not proof. The fact that it can be done on one machine means that it can be done on many machines.


Your "plausible explanation" consists of the bald statement the exit polls don't work. Your statement is prima facie false; exit polls do "work". The question is, "how reliable are they?" More precisely, "What is the standard deviation of the difference between exit poll results and actual vote counts in elections in which properly conducted exit polls have been executed?" The answer to that question is, "under 1%".


Very well, provide me with your own figure for the standard deviation of the difference between exit poll results and actual vote counts in elections in which properly conducted exit polls have been executed.


You're right. I apologize for making unwarranted assumptions.



It would appear that you are inferring my thoughts...
But to address your points: you have not done "this", you have not demonstrated that all that was done in the study was to substitute one correction factor for another (the study goes into a great many factors), you are incorrect in claiming that the authors don't draw any inferences for their conclusion in their paper, and you have not demonstrated that the magnitude of what their conclusion suggests is beyond reason.

Moreover, you are repeating your earlier mistake of objecting to the numerical result because, in your opinion, it's too big. Focus on the procedure, not the results. If we get the procedure right, then whatever number comes out of the chute is the number we go with.



Here are the conclusions of the paper, on page 10:



Show me where the phrase "vote shift" lies in these conclusions.


Your "innocent explanation" is "Exit polls are wrong." That's not what I consider a compelling logical argument.


Because the study and its predecessors examine each and every innocent hypothesis and reject them. You are basing your entire case on a blind condemnation of exit polls, without offering any calculations of the effects any of the various factors you cite. If you want to prove that exit polls are garbage, then show me your calculations.


Please refer to the conclusions quoted above, and show me where they concluded vote manipulation on a massive scale.


Could you reprise those for me? I thought I responded to every one of them.

Should Congress subpoena the Edison-Mitosfky raw data so that it can be analysed to (possibly) clear up the exit poll question?

Yes.

Should a federal-level investigation be initiated into possible electoral fraud?

I dont think there was a fruad, but an investigation couldnt hurt.

Particularly, how much investigative money does this controversy deserve?

However much it needs, which shouldnt amount to much.

We now have something just short of a smoking gun on the possibility of electoral fraud. It takes the form of the discovery of computer code inside the optical scan vote-counting machines produced by Diebold. This computer code:

1. permits the alteration of actual vote counts by external manipulations;
2. has no apparent function other than to alter actual vote counts;
3. was written in such a way as to disguise its operation;
4. violates Federal Election Commission regulations;
5. permits undetectable alterations of vote counts;
6. has been used in a demonstration to alter vote counts in less than five minutes.

You can find more information on this at Black Box Voting.

The claims you cite are a "black box voting" exclusive story. I resent your posting Bev Harris as a news source. Please cite a real news source for these allegations.

Resent away. If you want to dismiss this all as a lie, feel free. But she is not declaring opinions, she is reporting events that were witnessed by two members of Congress and a number of other people. Do you want to call them liars?

Bev Harris is a crank. I'm disappointed that you are reading this stuff. A number of congressmen are certainly liars, yes. And this "article" isn't even an article, it's a post in a forum much like this one. which is why I suspect you linked the main site, not this silly-looking bbs posting.

Not even our own Bush-hating scandal-driven media will listen to her ravings. And even she notes that her article does not prove fraud, only the possibility for fraud.
So, she found a POTENTIALLY exploitable flaw in one type of voting system, AFTER the election. I'm sure that Jeb Bush will be FedExing those memory cards to the esteemed Bev Harris anytime now.

I was thinking that this was similar to finding an exploitable flaw in Windows. An analogy - I applied for my Illinois vehicle license plate online using the internet. It is possible to hack my computer with a trojan horse virus. Indeed, if you took a congressman to my home you could show them my screen with the window "NORTON VIRUS ALERT" visible. Am I therefore to conclude that Illinois license plates are fraudulent? Actually, even this isn't accurate, because the flaw she found was hard-coded on the memory card, so a county official would have had to place it there centrally, before hand-distributing them to the precincts. Katherine Harris on a geeky day, I presume.

Maybe she is a crank; I won't take anything she says at face value unless it is falsifiable. In other words, if she makes a statement that could not be checked out against another source, then I wouldn't believe it, but I am willing to believe that she's not stupid enough to get caught in a flat lie. The six characterizations of the computer code that I presented in post #1 were all taken from falsifiable statements on her website.

I remind you that you are presenting a classic ad hominem argument: this statement is false because the person making it is a crank. While I'd be willing to argue within the constraint that nonfalsifiable statements from Ms. Harris are not to be trusted, I remind you that we are discussing the facts here, not Ms. Harris. If you have any objections to any of those six characterizations, then by all means let's hear them. Do you think that the code in the Diebold op-scan machines does not permit the alteration of actual vote counts by means of external manipulations? Do you believe that it has some function other than to alter vote counts? Do you believe that it was not written in such a way as to disguise its operation? And so on...


I linked to the main site because it provides the headline and additional information on the organization, not because I am part of some vast conspiracy to deceive the planet. And while it is a post in the Black Box Voting forum, it is a post from the organization itself. They are honest enough to permit open debate of the report, and they even include many critical comments which are of value to any person wishing to objectively assess the value of the report. Open debate -- what a concept! And if you think that the website is silly-looking, you really should take that up with the webmaster, not me.



True, her article proves only that fraud was possible, not that it actually happened. But this raises the question, why would somebody go out of their way to program the voting machines to permit fraud, if not to commit fraud? You're right, it doesn't prove fraud. But then, a smoking gun in the hand of a man standing near a gunshot victim doesn't prove murder, either. Perhaps further investigation is called for, don't you think?


Indeed, we don't need to wring our hands about all this. Just permit a little investigation of the facts. Why don't we take a look at those memory cards? You don't have any objections, do you?

Um, no. We haven't found a smoking gun in the hand of a man standing near a gunshot victim. We've found an FOID card in the hand of a man taking the train 700 miles away.


I do have objections. As we've seen since the 2000 election, cranks like Bev Harris will create noise, move the goalposts, change the sport ad infinitum to continue to generate conspiracy theories and "fire up the base." There is no there there. I'd remind you that the Miami Herald and USA Today found conclusively that under any standard employed to count ballots, Gore lost Florida in 2000. Yet, we had assertions of fraud up to and after the 2004 election. No matter how her theories are debunked, it won't be good enough. There is no point. It's not healthy for our democracy and so yes, I do have objections to it.

Why are we so willing to accept that the government is unable to renew our drivers license efficiently, quickly and without error in less than 2 hours, at an office that is open and staffed with full-time employees five days a week. But somehow when part-time, volunteer staff supplanted by (democrat and republican) county officials create voting mechanisms that are used once every 2 years, if I have to wait in line it's a freaking conspiracy. And oh, yeah, the guy who made the voting machines is a Republican. Gee, the guy that runs the printing company who made the paper ballots is a Democrat. And the county official who designed the "butterfly ballot" is also. Some conspiracy.

Can I get an AMEN????

The funny thing is that Americans have digressed into a "sore loser" mentality that seemingly could last until the next election, and possibly longer! When Bill Clinton won, most likely the first time due to the third party split, no republicans were screaming CONSPIRACY! ! I am appalled at the lack of political understanding being accompanied by an inordinate amount of whining and talk of conspiracy....

Any objective political expert (or even arm chair theorist!) would tell you that there are going to be flaws in any large scale election. The nature of our system is going to lend itself to being open to mistakes, holes, or other defects.

We're a far cry from the corruption often found in other "open" elections, especially those that rely on paper ballots. There is a bi-partisan constituency present during the process, and I believe largely, they keep it relatively honest.

carlitoswhy and aevans176, you have posted arguments that address anything but the evidence. You raise issues not raised in my posts, you complain about the statements of people not referenced here, you attack the character of Bev Harris, you talk about the 1992 election and the 2000 election, you preach about moral issues -- but you don't say anything about the evidence itself.

Does the fact that your responses avoid the evidence not in itself suggest something?

That this evidence isn't worth taking seriously?

I'm not the one to debunk this specific piece of evidence, so if it makes you feel better, go ahead and believe it until it's debunked. My larger point is that there will always be a theory out there waiting to be debunked. Because the losing side in the last 2 presidential elections just can't believe it. Therefore there must be a conspiracy.
Examples
- despite large Democrat registrations in key Florida counties, huge votes went to Bush. Conspiracy says "Diebolt." Facts say "these registrees are dems from 30 years ago who never re-registered."
- Berkeley students perform statistical analysis showing electronic voting machines may have awarded 130,000-260,000 or more excess votes to President George W. Bush in Florida. Conspiracy says "Diebolt." Facts say they used flawed regression analysis only looking at the change in votes from 2000 to 2004.

edited to add one more example:

You post this conspiracy theory in the original topic. Then you update it with a Bev Harris post. So, I asked myself, out of curiosity, what does Bev Harris say about the programmer and the Republican operative? She doesn't believe it's real

She has a 6-point rebuttal of this alleged hacking story which sounds pretty convincing. Question - when one conspiracy theorist is debunked by another conspiracy theorist, which one exactly am I to believe? Or, can I reasonably conclude that this is all misinformation designed to fire up the far-left base, evidence or none.

It goes on and on. Pointless.


?

Look. I generally applaud an objective debate, but there is no substantial evidence in any of your posts. This is the whole point. That's exactly what we're trying to say.

There are valid arguments that Aliens have abducted humans, that the Mafia killed JFK, and that Elvis is alive. However, most Americans believe none of the three. Frankly, we discuss the previous elections in that the same arguments would apply, but the republicans never made fools of themselves and that the same "election stealing" rhetoric has been spewed for far too long. If I were a Democrat, I'd find a new sounding board as this one smells horribly of absurdity.

You are refusing to take the evidence seriously not on its own merits, but on external considerations. While external considerations might have secondary significance, the evidence itself deserves some consideration. Your refusal to address the evidence itself speaks volumes.


Indeed so. We should bring a healthy skepticism to all hypotheses. That skepticism works both ways: we should not assume the answer to the question "Was the 2004 election fraudulent?" We should instead examine the evidence dispassionately. An investigation would further this end. Why are you opposed to a dispassionate investigation of the evidence?



Neither. You decide based on the evidence -- which you refuse to address.


The substantial evidence is the discovery of a hook in the internal code of a vote-counting machine, written in an illegal programming language.

Inasmuch as neither carlitoswhey nor aevans176 have availed themselves of several opportunities to address the evidence itself, let us proceed on the assumption that they do not question the truth of the evidence (by all means, correct me if this assumption is incorrect, gentlemen). Their position then can be expressed as follows:

"The computer code in the Diebold opscan machines:

1. permits the alteration of actual vote counts by external manipulations;
2. has no apparent function other than to alter actual vote counts;
3. was written in such a way as to disguise its operation;
4. violates Federal Election Commission regulations;
5. permits undetectable alterations of vote counts;
6. has been used in a demonstration to alter vote counts in less than five minutes.

The hypothesis that there was electoral fraud is preposterous."

I suggest that this position is, on its face, illogical.

Erasmussimo,

I'm with you, the Diebold code needs independent auditing. It's ridiculous that federal law does not require this of voting machine manufacturers. I mean, tabulation is the base of computer science, the machine that Herman Hollerith made to do a census, then went on to found IBM. The code ought to be very simple, not intentionally convoluted to allow backdoor data manipulations.

Until such is a requirement, none of our elections are trustworthy. In fact, due to the stonewalling that Diebold put up before the election, I'm certain that fraud indeed did happen. But what can be done to avoid similar fraud in future elections?

Trust but verify, now who said that? Oh yes, Ronald Reagan. Make it federal law.

YES! Absolutely!

That's why everyone who votes in the United States should have to provide proof of their identity and citizenship at the polls. And they should be disqualified (via a computer database) from voting more than once.

Why is it democrats are so vehemently against such common sense verification methods?

With respect to Ohio and Diebold, this is nothing but Area 51, Trilateral Commission, Grassy Knoll conspiracy nonsense.

Those machines went through a very extensive qualification process. The exit polls were bogus because the methodology was incompetent.

I want honest elections. When we have them, fewer people (like the pseudo Governor of Washington State) will be elected by felons and dead people.

What can I possibly evaluate? Some chick wrote on her blog an accusation that there could have maybe been a flaw in the hardware that could potentially have been exploited by some nameless operatives, had they the time, motivation, inclination. But even she says that there is no evidence of fraud.


And who exactly is dispassionately investigation here? Bev Harris Bio

She hasn't even released her "technical report" so even you must admit there is no "evidence" to review. So she got a computer programmer to implant a trojan-horse virus in a tabulating computer. Was he there on election night? No! They concluded that, if you were to hack a memory card and swap it with the "real" memory card, you could affect the outcome. Great? Did this happen? Of course not. Why in the world would anyone believe this, other than they pre-suppose that it's not possible an idiot like Bush won?

lordhelmet, your post contains eight sentences, only two of which have any bearing on the topic. Could you please confine your remarks to the topic?


No, she says that there is no proof of electoral fraud. She says that there is proof of improper procedure in the electoral process.



Which is precisely why I am suggesting an investigation by independent researchers, not partisans. Why do you oppose that?


The evidence consists of the existence of a hook in the programming code. Do you deny this evidence?

There's a new development in this story: the Supervisor of Elections for Leon County, Florida, Ion Sancho, has just released a special report confirming the thrust of Ms. Harris' claims. This smashes your case, which has rested on the assertion that Ms. Harris is an untrustworthy source. It would seem that you are now boxed into a rather tight logical jam, from which I can see only five exit doors:

Door A: OK, maybe we should have an investigation into electoral irregularities.
Door B: Mr. Sancho is a crank, a quack, a tin-hatted conspiracy theorist, and a sore loser.
Door C: Slink away without responding.
Door D: Attack me.
Door E: Change the subject.

Of course, I do not mean to minimize your ability to come up with something creative. Perhaps you can identify some additional doors.

Since when were you appointed as a moderator in this forum? I thought moderators were the only ones allowed to makes statements like that. Am I right? Wrong?

I wasn't responding to your post anyway. "Authormusician" made a point with respect to the verification of valid voting (which I HAVE to believe you think is important) and I replied to it. What's it to you????

Frankly, as carlitoswhey pointed out very well, the Diebold article was written by a political crank who also believe Bush knew in advance of 9/11. What's next? Sources from those who are convinced of UFO's being hidden in Area 51?

With an invitation like this, I choose door B.
democracy now
I offer - Ion Sancho, giving felons the benefit of the doubt.

Ion Sancho, defending optical scanning systems: testimony here

Ion Sancho is a member of Center for Policy Alternatives, a progressive activist group.

Etc.

I'll concede - Ms. Harris and Mr. Sancho have proved that "with the right access and tools" someone could have hacked some voting results if you concede that someone "with the right access and tools" could buy a firearm under the name "Carlito Erassmusino" and ship it to Al Qaeda. Then, you'd have your smoking gun.

These quotes you supply establish his integrity, not his bias. You believe that he's a crank, a quack, a tin-hatted conspiracy theorist, and a sore loser, because he went to great lengths to insure that the law was properly executed, and to insure that he got the best possible technology for his county?

Go look at their website. Here are the top ten headlines on their homepage:

2005 Progressive Agenda - Get Your Copy Today
Minnesota Increases Minimum Wage, Hawaii Set to Follow
Mississippi Enacts Sweeping Juvenile Justice Reform
Tennessee Passes Racial Profiling Bill
Hawaii Establishes Pay Equity Commission
Montana and Washington Enact Prescription Drug Laws
Montana Expands Affordable Insurance
Washington Enacts First Environmental Building Standards
New Hampshire Bans MTBE
Washington Improves Assistance for Sex Trafficking Victims

Do you claim that this organization is obviously populated by crank, quack, tin-hatted conspiracy theorists?


Why do you need to tie a concession of truth to a completely irrelevant issue? Let's just address the topic. And "the right access and tools" in this case consist of a personal computer and physical access to memory chips used to transport the data -- memory chips that were seen lying scattered in the open on tables at precincts.

So, I guess you don't want to investigate Al-Qaeda arms shipments. Fine. He's not a crank, he's just been had by a crank, I guess.


I guess that Leon County should investigate who had access to the central tabulator. They should also take better care of memory cards. I suspect that Mr. Sancho had access, along with his staff. I know that you aren't saying that this man of integrity or his staff altered the vote, so why would you want an investigation, other than you don't like the result?

Now you're choosing Door E: Change the subject.


Because the same machines containing the illegal program code were used in many other counties across the country. The potential for fraud on a scale of millions of votes has now been proven.

The code on these cards was illegal?

The potential for fraudulent Illinois license plates and Al-Qaeda arms shipments have also been proven. You can call it changing the subject, but it merely shows how tenuous your "proof" is. The company you keep also affects the seriousness with which your argument is taken. If you want every piece of alleged election fraud investigated "dispassionately," best not to go around crying "wolf" all the time.

Indeed it was; it was written in a language not certified by the National Elections Commission. Federal regulations require all code used in vote-tabulating machinery to be written in a language certified by the National Election Commission.


If the existence of a hook in the code in the opscan machine is tenuous evidence, by all means, show how or why it is tenuous.


You are taking the ad hominem argument to new heights. Not just "don't believe the case because its author is a bad person" but now: "don't believe the case because its author keeps company with bad people."

LH, I agree. States should restrict voter registration to a reasonable time before Election Day. Also, states should get tighter in their voter registration as people move out of state. Unfortunately, this isn't a federal issue, so some states are better than others -- and it depends to a great extent on how much the state in question is willing to spend on the efforts to clean up voter registration lists.

On the other hand, voting machine manufacturing involves interstate commerce, and is thus under the control of feds.

For Colorado, it is a requirement to produce a valid state ID and sign in (compare signatures) when it's time to vote. Still, the lists get out of date because much of Colorado's population is transcient (military and tourist industry).

I am happy to see that you agree with me that Diebold and others need to have their code (whether firmware or software) audited by an independent, non-partisan, technically savvy organization. The auditors should be required to give up the right to vote in order to audit in a non-partisan manner.

By the way, if the memory chips in question are ROM, there's no danger of hacking. But if PROM (Programmable Read Only Memory), it is possible to hack the chips. I think the law ought to require the non-hackable architecture as well.

It looks like everyone has lost all innocence as to the honesty of democratic elections. Good! That's an inherent weakness of democracies, and the more we demand honesty, the more we might get. You know, if anyone's listening.

The code in the machines is ROM. That's why it has to have a hook in order to cheat: if the cheat code were in the ROM, it could be found. The cheat code is on the memory cards, which I believe is standard EEPROM, and is not subject to any scrutiny -- because the memory cards are supposed to hold only data, not code. The fact that the hook enables code from the memory card to execute is itself incriminating, as the memory cards are not supposed to have any code on them.

Here's a non-technical analogy: imagine a bank where you deposit your cash by a complicated process requiring you to go from one window to the next, getting all the proper forms filled out. At each window, the teller tells you where to go next. The fifth teller always tells you to "go down the hallway and follow the signs to the next window." Down the hallway is a sign pointing to the right. But somebody, at the last minute, inserts a paper sign over the real sign, with the paper sign pointing to the left. They've set up a desk in the left hallway where they take your money, give you a receipt, and send you out the door. After they've taken some money, they remove the paper sign and walk out. Undetectable theft.

I'll take it to even greater heights. I was referring to the entire vote conspiracy theory camp, not just the author. To illustrate - people who have said that the election was rigged for Bush:
- Michael Moore
- Air America (best as I can tell)
- Keith Olbermann
- Bev Harris
- Democratic Underground
- Howard Dean
- Al Gore / John Kerry (depending on when you listen)
- Al Sharpton
- George Soros
- etc.

Each of these entities latches onto every single voting conspiracy possible theory, which is then debunked in short order. They then never bother to inform their constituents of the debunking. Each of these entities has knowingly lied about the election results and how they were rigged. Why would I reward this behavior by listening to the "next" theory and "dispassionately" analyzing it? You act as though you making this argument in a vacuum makes it automatically devoid of any other "facts" we've seen before, and we should drop all of our critical discrimination as to whom to believe because this time, it's a real conspiracy. Sorry, potential for fraud. As I said before, it's crying wolf. I don't need to see it. I disbelieve it. Call me closed-minded, assign multiple-choice letters to my opinions, whatever. We may as well debate the protocols of the elders of Zion, or whether the "K" on the snapple bottle is for Klu Klux Klan.

maxime ad multes homines!


Well, then it shouldn't be any problem at all debunking this one, should it? Go ahead -- give it a whirl.


OK, so here's the situation: you are unable to refute the evidence provided by the hook in the code, but since it yields results that you do not like, you simply refuse to believe it. Fair enough.

LH





http://www.seattleweekly.com/features/0523...ws_election.php



HL,evidently the judge doesn't agree with your statement.

It seems Rossi, doesn't believe justice prevailed and took one last cheap shot.

This couldn't be a conspiracy, could it? No, that only happens to those liberal Democrats. Oh well.

Brace yourselves, oh ye defenders of truth against the barbarian hordes of conspiracy theorists! Yet more teeth-grindingly solid evidence supporting the possibility of electoral fraud has emerged. Interestingly enough, it's not new evidence, but old evidence that was previously discredited for what now turn out to be the wrong reasons.

Here's the story: immediately after the election, a lot of people focused on Florida, for three reasons: first, it had been the site of quite a few electoral shenanigans in the 2000 election; second, it had shown continuing problems, such as the screwed-up list of ineligible felons; and third, it had recently installed a number of touch-screen voting systems lacking any proper paper trail.

So people went to work on the electoral data, poring over every detail. There were lots of false starts, and one problem in particular: the "Dixiecrat effect". This occurs all over the South but is particularly evident in the panhandle of Florida. Dixiecrats register as Democrats and can be counted on to vote for Democrats in local races -- but they often vote for Republicans in national races. This produces some very strange results, such as a county producing three times as many votes for Mr. Bush as there are registered Republicans.

The Dixiecrat effect proved to be immensely difficult to correct for. There were lots of false starts and dead ends. There was definitely something peculiar about the election data, but nobody could clearly isolate it. At last somebody came up with an analysis that examined differences between counties using op-scan machines and counties using touch-screen machines. What was confusing was that the analysis demonstrated that the anomaly was greater with op-scan machines than with touch-screen machines. Everybody had been expecting problems with the touch-screen machines, because these were brand-new and they left no paper trail -- an open invitation to fraud. But the op-scan machines had been in use for many years and were felt to be trustworthy; moreover, they left a paper audit trail, which was solid protection against fraud; surely nobody would attempt to fraudulently manipulate vote counts with those paper tapes in place.

This was puzzling. People talked about it, tried a number of other approaches, and eventually concluded that there just wasn't anything there that could be relied upon. The subject died. And by the way, all those who are certain that the people doing this work are rabid conspiracy theorists willing to grasp at any straw should realize that this line of investigation was abandoned by the investigators themselves. They didn't need any outside criticism to conclude that they were barking up a bad tree. In fact, the author of this study, Elizabeth Liddle, is also the author of "Febble's Fancy Function", which suggests that the exit poll data may not demonstrate anomalies in the vote counts. In other words, this lady has already demonstrated great intellectual integrity by attacking one of the major arguments supporting the hypothesis of electoral fraud.

But now it appears that in fact, they were barking up the right tree. The discovery of a hook in the ROM for the op-scan machines changes everything, because it suggests that the op-scan machines were in fact designed to be hacked.

Thus, the old, discredited analysis that you can find at this page has now become the old, quite pertinent evidence that shows that the statistical anomalies were significantly greater with op-scan machines than with touch-screen machines.

One other thing: this analysis is nowhere near as complicated as the exit poll analysis. Any person with a reasonable grasp of statistics should be able to understand the basic arg