How to FIX the exit polls, as in "the FIX is in."

Can you think of ways that the exit polls could be "fixed" or otherwise altered?

Hacking the central computer would be one. How was data entered? Is there a paper trail of the call ins?

Give this your best brainstorming.

Was the data server connected directly to a network that had access to the Internet, dial-in, or any possible outside connections?

Is there a paper trail?

How do they justify their weighting technique?

How many hands touched the data before it was entered into the system?

Were there any redundant steps to ensure accuracy?

The key to any magic trick is to make people think you are doing one thing while you are really doing another...

lp

sure - any reporting process can be "fixed"

but more than one org did exit polls - and they all say Gore and Kerry won more votes than were reported.

The latino groups private exit poll for lantinos gives similiar data to latinoes in the overall poll.

Multiple exit polls make game playing unlikely.

Unfortunately, 80% of the actual vote counting goes through 2 GOP controlled firms that refuse to release the software used - and whose ownship/leadership seem a bit "in the family".

What purpose do they serve? to give the media an early report of the outcome? Why can't we wait a week or 2 to know the result? What difference does it make if we don't know the winner right away?

also.

Indeed if I did not have personal experience with GOP poll workers than had taught me that they cheat, I'd be in favor of no exit polls.

I would rather have a machine do the counting. But the machine needs to be tested for accuracy every time.

And it is universally accepted (except here) that a discrepancy between the Exit Polls and the "official result" probably means that fraud has occurred, and an investigation is warranted, and probably a re-vote.

Exit Polls must therefore absolutely continue in use here. The problem with the Exit Polls on 11/2/04 (which Kerry won by a 3% margin) is that the TV networks CHANGED the Exit Poll results--"altered" the Exit Poll data that everybody was looking at on their TV screens--to fit the official BushCon-controlled electronc results (the "official vote"). Thus, most Americans DIDN'T KNOW that Kerry won the Exit Polls (unlike the folks in the Ukraine, for instance).

One other problem: Edison/Mitofsky (the 2004 Exit Pollster) chose (or the networks chose) to do a demographic poll, in which they not only ask who you voted for but also gathered information on race, religion, etc. This gave the BushCons a ready-made but DECEITFUL talking point, when the real Exit Poll numbers came out, that "it was just a demographic poll." In reality, there is no reason a demographic poll would be inaccurate on the prez numbers (which Mitofsky admitted). And then Edison/Mitofsky, trying to save their sorry asses, threw them another bone--that Republicans were shy of the pollsters on 11/2. This has been entirely refuted, in a Jan. 29 report signed by NINE Ph.D.s from leading universityies, who say that E/M's Exit Poll data actually indicates just the opposite--that the Exit Poll FAVORED BUSH--and that this means that Kerry's margin in the Exit Polls (3%) is probably even larger.

http://uscountvotes.org/ucvAnalysis/US/USCountVotes_Re_Mitofsky-Edison.pdf

Without the Exit Polls--that is, without their REAL results (which were leaked)--we would not have had any data for analyzing the election, except the BushCon-controlled "results."

The only comparison data would have been paper vs. electronic--and also top of the ticket vs lower ticket, number of voters vs. number of votes, and other internal evidence from the election itself.

In the two '04 cases that have been studied of paper vs. electronic--Florida and N. Carolina--fraud (or utterly impossible skew to Bush) has been found in the electronic vote (vs. the paper vote). Impossible skews to Bush have also been found in the other internal evidence.

But without the Exit Polls, we would have been unable to show the OVERALL skew that occurred NATIONALLY, and our case for fraud would be weaker.

Exit Polls are vital to the integrity of the election process, and would be even with paper ballots and hand counts, and no central electronic tabulators owned by BushCons.

How would you know if the election was stolen?

Here is how we know:
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x316275

The number of interviewers is so small that even if a few are compromised, it can have a drastic effect on the final outcome.

See my post on their response to this and other such arguments.

Take a gander at the link in my signature.

The original poster asked, if you wanted to screw up the exit polls, how would you do it. One way, IMO, is to flood the applicant base for the interviewers and get a 10% or so share of the interviewers. I didn't say that happened in 2004, nor do I have any reason to believe it did.

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x318234

Sorry. I didn't mean to insult.

Not some private firm.

Mitsofsky does boot licking for Nazis.

These are hacker attacks that would flood the server(s) with requests, which, in turn, would cause server crashes.

In fact, Mitofsky claimed that the NEP-poll server crashed at about 11pm EST, Nov. 2. In the 2002 general election, the now-defunct Voter News Service (VNS) claimed a computer crash forced them to abandon their exit poll in late-afternoon of Election Day. Mitofsky was with VNS then.

Of course, I getting tired of hearing excuses involving computer crashes, server crashes, and memory card failures in elections and exit polls. Computers and servers are usually highly reliable. These types of excuses are in the same category as "the dog ate my homework" excuse.

You just infiltrate Mitofsky's organization with Repuke operatives. Whether they are pollsters on the ground, or computer techs, network techs, programmers, or whatever. The goal being to release an early version of the poll to the media showing a Kerry lead, so Limbush and Shammity can get on the radio (and Matt Fudge can get on the Internet) and play Get-Out-The-Vote for Shrub, particularly in the West.

If you look at the biggest weighting change in the adjusted poll (the only major change really), it was the Dem/Rep party affiliation ratio and this change was largest in the West! So if you believe any of this, the voters shifted more toward Bush later in the day, AFTER the release of the polls that showed Kerry ahead.

Then you get a Repuke operative such as Dick Morris to come out with the cover story and accuse the Dems of leaking the polls! He said that WE released the polls early to get out OUR vote. But I believe this would get out the Bush vote if it appeared Kerry was ahead, not ours.

And on edit: To avoid embarrassment at having his polls corrupted in this way, Mitofsky then comes up with his "reluctant Bush responder" explanation. And he did say his interviewers basically sucked too! He said they were "young", but what if they were just Young Republicans?

Why do you ask L. Coyote?

You ask, "Why do you ask L. Coyote?"

What good are exit polls if they can be fixed more easily than an election. If you are going to fix an election, you need to fix the exit polls too!! Simple.

we don't really know.

I'd run background checks on all the interviewers, IT people, etc.
Set security policy for the servers and client PCs, routers, modems, etc. Instruct the employees, etc.

We don't know if they do any of this, do we?

And all this should be put in the RFP the media sends out when they select a polling company. I mean for all we know, Joe's Polls might do a better job with IT security than Mitofsky. A few years ago, no one took IT security seriously except mainframe guys/gals.

If you do, all your work has been for naught.

The motivation for fixing the exit polls in the scenario you draw with Morris et al is very weak.

Why should Repukes infiltrate the exit poll staffs when they could just as easily hack the votes? And why would they fix it for Kerry?

Your post makes no sense.

I didn't raise the issue this time. We were asked how the polls could have been hacked (fixed).

If I were going to do this, that's how I'd do it. I'd do it to get out the vote by making it look like my guy was going to lose. And I'd arrange for a cover story that said the other side did it. We know the polls weren't hacked to make them AGREE with the election, because they did not agree. They were "adjusted" to agree. That's not a hack.

And I still think, with all the weakness of the Diebold stuff, and the other stuff, it's still easier to hack a few servers than hundreds or thousands to swing the vote. There are 3000 counties, so even tabulator-level hacks would have to number at least in the hundreds nationwide. DRE hacks, about a thousand times more than that. I know that bad code could be replicated, but it's still harder to do that than to hack one or two servers in some corporate network somewhere, esp. if you can get someone on the inside.

Dem BOE members, while they may be complicit, stupid and unaware in some areas, are still a deterrent to some extent. They can't ALL be dumb and corrupt. But if Edison-Mitofsky, not exactly a Fortune 500 company, happens to have some weak IT security or human resources investigative staff, they could easily have had their data manipulated.

We simply don't know, so we brainstorm. Nothing wrong with that.

Diebold has just come out with a DRE with a VVPB. I heard it on Air America, so it must be true!

Now, we see bills in Congress that will actually require this nationwide, some of which are even Republican-sponsored.

Coincidence?

Let's say the Exit Polls were hacked, but not necessarily the election, except in Ohio which was mostly voter-suppression (after all, Shrub still had to win right?). But because WE are so sure the election was hacked, the machines defaulted to Bush, the tabulators can't be trusted, etc. WE are demanding VVPBs at the grass roots level.

I bet we get them too. Why? Because Diebold is finally ready to make some money off of it.

Now, of course, Op Scans and maybe even hand counting would work too, but no one gets rich off of that stuff, do they?

If Diebold and the rest had perfect systems, no one would buy the new improved versions, would they? So there had to be a few glitches. Just enough to cast doubt on the process and engender the outcry for change.

Ballot access for the disabled is the other business driver (as they say in the business). Multimedia DREs!

You don't have to steal an election to make money off of it.

Two databases. One favoring Kerry by a few percent, the other actual.

You start out running the bogus one, but when you reboot, you default to the actual one and the bogus one is deleted. Both are updated concurrently from the field, but the only way to get the correct data without discovering the fix or being in on it, is to reboot the server. This is the "server glitch" that is supposed to have occurred late on Nov 2, after which the poll results changed.

This would require only one Republican operative in Mitofsky's data center, or even outside if he/she knew the password and used a wide-area network connection (like the Internet or a dial-up).

IMHO, it would be SO SWEET to hack the exit polls to get out the vote, and once the results was achieved, go back to the real data. If Bush wins, you've made it happen. If Bush loses, no harm done. If you get caught, so what? All you've done is hack an exit poll! A slap on the wrist at the most. And this is SO MUCH EASIER than hacking the actual vote.

Granted, there is some effect if California voters think it's over, but everyone knew CA was going to Kerry anyway. Oregon voted by mail, so that was over. Where's the gain, in Utah? No way.

This early release concept is a red herring, detracting from grander concerns. Can the exit polls be trusted in 2008?

How difficult would it be to fix the election polls?

We know they stole OH, and most likely FL, so the election was already over. But Shrub wanted his "mandate" so he needed the popular vote.

By West, I'm going by Mitofsky's definition of West on the Scoop poll reports that showed 4 regions. He didn't name any states. The Dem/Rep ratio went from 37/34 to 34/36 in the Western region between the 7:30 PM (ET) Nov 2 report and the 2 PM Nov 3 final report. This is a huge shift, bigger than any other region. Since votes run 90% according to party affiliation, this means there could have been a shift late in the day in the west which added to the popular vote, even though he lost CA, OR and WA.

As far as the exit polls, well, there are those who trust them now. Are you saying you do or you don't? And which version?

All I'm saying is, it's easier to rig a poll than to rig an election. One server is all it takes.

And, it logically follows, if you can rig an election you can easily rig the exit polls, and would be foolish not to!!

So, what can be done to ensure the security of future exit polls?

and also have bi-partisan employees or auditors at the polling organization and at the polls doing the interviews.

Forget about the demographic crap! The poll should be designed to determine one thing and one thing only: Who won the frickin' election!
It should not be an exercise designed to tell politicians how to sell soap.

I certainly believe that it’s easier to rig an exit poll than an election. But, it would help to have publicly funded exit polls under the direct supervision of state officials. The raw data should be available immediately after the election for everyone to see. The formula to weight the selected precincts should also be subject to public scrutiny. There needs to be better trained interviewers, larger samples, and supervision at the individual precincts. We need to invest a lot more resources. That being said, it’s still easier to rig exit polls than elections.

like Blackwell in Ohio?????????

If we can't trust public officials, who can we trust? We don't have a lot of choice. At least the data would be public and not someone's private property.

Then I dont think even "public officials" could get away with screwing with them.

I agree, exit polls need to be organized by someone else and need to be entirely open. It is certainly necessary as long as there is some form of E-voting or E-tabulation.

Now, how can that be accomplished?

Paper ballots only are an absolute must. In the final analysis, exit polls are less trustworthy than paper voting. And the current exit polls are not being done for the voter; they serve specific corporate purposes only.