Voting Company Reverses Stand: Flawed software WAS used

[BevHarris]

NEWS RELEASE

Voting Company Reverses Stand: Flawed software WAS used in Georgia and other elections
- There are “kinks” in touch screens

Contact: Bev Harris xxx-xxx-xxxx or [e-mail]

According to an Aug. 4 article in Wired.com: Diebold company spokesman Mike Jacobsen “confirmed that the source code Rubin's team examined was last used in November 2002 general elections in Georgia, Maryland and in counties in California and Kansas.” (http://www.wired.com/news/politics/0,1283,59874,00.html)

Actually, the software may have been used in as many as 13 states and 197 counties, according to Diebold documents given to Santa Clara County in Feb. 2003 (http://www.blackboxvoting.org/mfr.pdf) - list of counties at bottom.

Earlier, Diebold had told reporters that the software which contained “stunning security flaws” that made hacking easy, was an older version and never used in any election.

Yet it was used, and provably so. The Diebold software version was easily verifiable:
- The FEC requires certification of voting machine software by version number
- The certified version number matches what was studied by the Johns Hopkins scientists.

STILL “KINKS” IN THE TOUCH SCREENS: John Silvestro, a voting machine representative in Boston who sells Diebold machines, said the touch-screen system would cost the city about six times as much money as optical scan machines which have a paper audit trail, and that companies like his are still working the kinks out of the touch-screen machines. Silvestro told the Boston City Council that Boston was better off with optical scanners. (http://www.boston.com/dailyglobe2/214/metro/Menino_OK_s_new_voting_machines+.shtml)

Georgia, perhaps hardest hit by the growing Diebold scandal, is now facing renewed questions about missing memory cards and other irregularities. On election night during the 2002 general election, 67 memory cards, containing thousands of votes, went missing in Fulton County. Also, according to documents provided to Santa Clara County, Diebold machines experienced “buffer overrun” problems during the election, requiring poll workers to turn them on and off, and if not done properly, this can also cause loss of votes.

Georgia officials, responding to a Freedom of Information Act request by Georgia voters, admitted that they did not have any of the certification documents clearing use of the machines following a series of unexamined program patches put on the machines right before the election. Georgia law requires that any time software is updated, it must be recertified, but the patches were never examined by testing labs. No one really knows what was on the patches; Diebold denied that patches were done. (http://www.blackboxvoting.org/lies.htm)

According to technicians (http://www.blackboxvoting.org/robgeorgia.htm) who administered the patches, they were told to download them directly off the Internet from the Diebold ftp site, rather than getting them from Microsoft (as they would if they were Windows patches) or from the Independent Testing Authority (as the Georgia examiner for voting machines, Dr. Brit Williams -- http://www.blackboxvoting.org/Williams.htm -- describes the required procedure).

In a new controversy, researchers with BlackBoxVoting.org say they have discovered that changes were made in the Windows operating system used with the voting machines.

According to Dr. Doug Jones, a member of the Iowa Board of Examiners for Elections and an expert congressional witness on electronic voting: “The FEC/NASED Voting System Standards require that all software used in voting systems be passed through a source-code audit, but there is an exemption, in both the 1990 and 2002 editions of this standard, for UNMODIFIED third-party 'COTS' software, that is, commercial off-the-shelf software produced by a third party THAT HAS NOT BEEN MODIFIED for use in the voting context. Use of Microsoft Windows and Microsoft Office clearly qualifies for this exemption.” (http://www.cs.uiowa.edu/~jones/voting/dieboldftp.html)

Jones says that Diebold made representations to him that they made no changes whatsoever to the Windows operating system, yet new information from Black Box Voting indicates that significant changes to Windows were made.

“Source code files clearly show that Windows source code was modified.” says Bev Harris, author of ‘Black Box Voting.’ A new Freedom of Information Act is now being filed in Georgia seeking to clarify whether changes in Windows were disclosed and authorized, and Black Box Voting researchers are examining what impact the changes have on the voting program and its security.

OTHER STORY ANGLES:
- Experts point to secrecy, flaws in certification system
- Following the money trail: Lobbying, kickbacks and contributions
- More on original story: http://avirubin.com/vote/response.html

COUNTIES THAT USE DIEBOLD TOUCH SCREENS
Alameda County, California
Kern County, California
Los Angeles County, California
Marin County, California
Modoc County, California
Plumas County, California
San Diego County, California
San Joaquin County, California
Solano County, California
Tulare County, California
El Paso County, Colorado
Saguache County, Colorado
Weld County, Colorado
===========================
GEORGIA: All 159 counties
===========================
Tippecanoe County, Indiana
Johnson County, Kansas
Jefferson County, Kentucky
Allegany County, Maryland
Dorchester County, Maryland
Montgomery County, Maryland
Prince Georges County, Maryland
===============================
MARYLAND: All counties in Maryland have purchased Diebold touch screen machines
===============================
Gaston County, North Carolina
Rowan County, North Carolina
Douglas County, Nebraska
Lancaster County, Nebraska
===============================
OHIO: Most of Ohio is considering purchase of Diebold machines
===============================
Bradley County, Tennessee
Rutherford County, Tennessee
Shelby County, Tennessee
El Paso County, Texas
Franklin County, Texas
Guadalupe County, Texas
City of Norfolk, Virginia

CONTACTS:
xxxredactedxxx

[BevHarris]

1. Update from Hopkins researchers: Diebold's arguments often miss the point

"Diebold's arguments often miss the point, do not address many of our most serious findings, and demonstrate a considerable lack of knowledge of the technical matter, including a misunderstanding of technical terms," they write (http://avirubin.com/vote/response.html)

"Diebold criticized our network-based attacks as being unrealistic since the voting machines will not be networked in practice. The Diebold code we examined contains many different configuration options, including the use of wired or wireless networks and the use of modems. Any communication, whether wired or wireless, whether over the Internet or over private phone lines, is fair game for an analysis of what can be intercepted by an intruder. If there is no such communication, only then would the Diebold system be safe against such attacks."

"The primary example is Diebold's claim, which was widely cited in the press, that we ran the code on a different platform from the one that was intended. While this is true, it in no way reflects on our analysis. In fact, the main reason that we ran the code at all was to test whether or not it worked, and thus to help draw on opinion about whether or not it was production code. Our entire paper and our entire analysis were based on manual inspection of the source code. Thus, the platform on which we ran the code did not play into any of our findings, and Diebold's attempt to focus attention on that is a clear effort to misdirect readers. Unfortunately, many reporters and election officials have latched onto this issue as though it was meaningful."

[TruthIsAll]

15. Bev, just want to commend you once again. Regarding that insane thread

yesterday where obfuscators, naysayers and other disrupting trolls came out of the woodwork, I shall not name names here, but I will say that I am especially dissapointed in one DUer who I once had a great deal of respect for. That person seemed to change overnight into something quite despicable. I don't understand the DUer's agenda, but I will never trust that person again. Jealous? Maybe. Smug and arrogant- definitely.

Bev, I don't care if this may have been timed to hype your book; more power to you. You deserve to raise awareness and prepare the masses for your tome. The groundwork is laid for the next chapter in this saga.

All America owes you one big HUG of gratitude. You are simply amazing, and you have achieved something quite fantastic. The crooks are on the run, thanks to you and DemActivist and the rest.

As a lifelong programmer from the mainframes to the minis to the PCs, and a student of politics and mathematics, you have done it exactly the way I would have. But there is no way I could have.

TIA

[SharonAnn]

16. Yes, Bev, major kudos to you and if this timing helps your book,

then that's even better.

WE WANT PEOPLE TO BUY AND READ YOUR BOOK! Once they do, they're going to have an entirely different perspective on what they can trust for election machines.

I see this as a WIN-WIN situation. You've done something where truth wins, our election systems get reviewed for accuracy and auditability, citizens win because democracy can take place, and you can win because your books sell. The important thing is that the message gets out.

I think that's great! I'm applauding!

[Eloriel]

83. Timing

Bev, I don't care if this may have been timed to hype your book;

This is what I know on the subject of timing of all this and her book: The book was originally supposed to be out at the end of May. I think she was pushing the deadline for that happening pretty hard, but in addition, more things kept happening and she obviously wanted to add as many of those things to the book as she could. THAT was also an ongoing process because it continues still.

Additionally, it's not as if she was doing nothing but writing her book. She was involved in various research herself which has kept her pretty busy. And some of that too I'm sure she wanted to add to the book, or probably alter some parts to accommodate the new information.

I'm sure there are cynics who see artifice in the timing of the book along with some of these revelations, but actually I think it would have been better to have had the book already published by this time. YMMV.

But those who are more fair-minded will think back over the history here at DU and see a little more to the picture.

Eloriel

[arcane1]

2. this is going to be in my reps' fax machines STAT

this is hot stuff! Does this invalidate any of the '02 election results in those areas??

[Eloriel]

10. Invadlidate any 2002 elections?

It darned well should. After all -- how on earth can a SoS certify ELECTION RESULTS with all this?

Think about it. In Georgia:

* There's no actual evidence that there WAS any state-level "certification" -- at least none that went any farther than collecting the vendor's certification fee, as required by the state

* National certification is something of a joke too

* Purportedly, WINDOWS "patches" (uncertified) were installed AFTER these two levels of "certification" -- which modifies the COTS provision of the FEC regulations

* And then there were all those incredible security vulnerabilities, as outlined by the Hopkins Heroes.

How on earth could anyone think that the election results could conceivably be considered valid and thus the election results certifiable?

I'm hoping there will be some lawsuits filed on the matter.

Eloriel

[ozymandius]

22. Good point.

How can suits alleging fraud be avoided when this revelation is vetted by proper attorneys?

[neuvocat]

3. Thanks Bev!

Outstanding work.

[kainah]

4. Great job, Bev! n/t

[T Roosevelt]

5. Wow

this just keeps getting better and better...

[dusty64]

6. I swear these "voting" machine

companies lie and change their stories almost as much as our regime. How is it that virtually all election officials and political parties buy their crap without question.

[SharonAnn]

20. For those of us who have worked for large tech companies

this is actually what we expected. It's just the way they do business (in all areas - not just voting machines). Razzle-dazzle, avoid the question, tap dance around the issue, throw out a bunch of tech buzz words, make statements, retract or restate statements, sow FUD (Fear, Uncertainty, Doubt). Rinse. Repeat.

[ParanoidPat]

77. ROFLMAO!

:evilgrin: :thumbsup:

[dmr]

80. Right SharonAnn, then they throw in the cost factor

as they did in Boston, and most probably elsewhere.

STILL “KINKS” IN THE TOUCH SCREENS: John Silvestro, a voting machine representative in Boston who sells Diebold machines, said the touch-screen system would cost the city about six times as much money as optical scan machines which have a paper audit trail, and that companies like his are still working the kinks out of the touch-screen machines. Silvestro told the Boston City Council that Boston was better off with optical scanners. (http://www.boston.com/dailyglobe2/214/metro/Menino_OK_s_new_voting_machines+.shtml)

So FUD becomes FUCD = Fear, Uncertainty, Cost, Doubt.

And, FUCD we are until each and every one of us can vote without a shadow of a doubt the integrity of the process is not compromised. I won't settle for less.

[RedEagle]

7. Bev, Check Out the Boston Story

I have talked to an insider and this person says that the individual identified by the reporter as Diebold's president, is really a salesman from another company.

Everyone may be misinformed on this one, because a reporter didn't check facts out.

Story needs to be cross checked and confirmed.:shrug:

[DEMActivist]

LHS Associates = Diebold Distributor

http://www.lhsassociates.com

[BevHarris]

12. He is a Diebold distributor. His web site links straight to Diebold

and I believe he is the Diebold rep in the Boston area. Most large manufacturers work with distributors.

However, I will edit that down.

Bev

[Name removed]

18. Deleted message

Message removed by moderator. Click here to review the message board rules.

[BevHarris]

23. Quoting an article in the Boston Globe

Tinfoil, I hit alert. You have one chance to change your post to conform to rules before I leave DU for good.

Bev

Have alerted mods four times re: publishing personal names. So far, someone took it upon themselves to post my married name and four times I have alerted for posting DemActivist's name. This may be my last visit to DU. It's bullshit. And WOULD YOU CARE TO EXPLAIN HOW YOU GOT THAT NAME???? No. On second thought, don't bother.

It's bullshit, just like the post to me today "Think hard. Watch very carefully what you do." A total fucking threat.

[Occulus]

41. I think that may be

a truly desperate gasp to make you stop what you're doing.

But somehow, I just can't see a total pitbull researcher such as yourself backing off :) Still, and I mean this in the sincerest way possible.... be careful. There are those who stand against you who would do anything to silence you.

That's not a threat, just some advice to be careful. If anything unusual happens, well... just be prepared to disappear with all of your data for a while.

You've done a great job. Keep up the good work :)

[RedEagle]

27. Bet the Reporter Mixed up LHS and Diebold

Okay,

Here is the information I was given:

"John Silvestro is not the president of Diebold. He is the president of a local New England Rep that sells accuvote equipment (LHS Associates)."

Demactivist has a link to the site, and looks like the guy sells Diebold stuff.

The reporter mixed up LHS with Diebold, apparently.

It still requires further checking, but for the sake of an honest record, I alerted people here.

TinFoilHat, if you're so hot on people checking this out, why didn't you jump on first and do you're own checking?

The journalist made a mistake, I doubt even the Boston Globe knows it yet. Sloppy, maybe, but news staff is cut to the bone and given no time for research.

This will get corrected as soon as possible, and BBV will probably be the first.

We don't even know of the LHS guy made any attempt to correct the matter, which should have been the first step and his responsibility to do.

How many presidents are there in Diebold? Can you name them all off the top of your head and the departments they lead?

Has Diebold even cried foul on this yet? That's the one that's got me. You'd think they would have been all over this immediately.

For the sake of the record, it needs to be cleared up. But clearing up the Boston Story does not clear Diebold. It's a snafu and that's all. It won't change the big picture one whit.

[BevHarris]

29. He is the Diebold rep in Boston.

He is president of his company, which represents Diebold there.

Splitting hairs, and I represented it correctly the first time (first write-up called him the Diebold representative).

Red herrings and distractions. These people are desperate.

Bev

[Occulus]

40. Bev,

Ka-BLAM!! But it goes, I'll wager, even deeper than you or any of us have yet proven. Although, I must say, the information trail you've uncovered is certainly a very interesting path to tread.

I'll be looking for your book :)

[TinfoilHatProgrammer]

47. no kidding

I never said it changes anything. That's not the point.

Ms. Harris posted earlier today on DU that it was basically unacceptable for a party (in her case Diebold) to make an inaccurate public statement and then correct/amend it at a later time.

I'm simply pointing out that according to her own logic, it's equally unacceptable for Ms. Harris or DEMActivist to make an inaccurate public statement and then correct/amend it at a later time. I submit to you that Ms. Harris knows full well that the person in question is not the president of Diebold. Their original statement (posted multiple times, on multiple threads) was a baldfaced lie that they (or more accurately, she) chose to correct only after it had been posted and repeated several times.

It's pure hypocrisy, forgive me for pointing it out.

JC

[DEMActivist]

48. What was posted was a direct quote from a newspaper article

and it happens on DU every day. Surprise!

Pull up every post on this.....what was posted was word for word what was printed on the newspaper's web site (The Boston Globe).

If you have a problem, I suggest you take it up with the reporter who printed that.

[TinfoilHatProgrammer]

51. no problem with the reporter

The reporter in question hasn't demonstrated any special hypocrisy regarding inaccurate public statements.

JC

[ibegurpard]

54. The "hypocrisy" you refer to was in reference to THAT article.

Which I'm sure you would've realized had you taken enough time away from axe-grinding to actually READ anything.

[DagmarK]

73. Wow, TinFoil.....you have made my day!

Cause if a silly reference to a person's title is the MOST you can slam "Ms. Harris" and DemA with...... I am a happy, happy camper!

Diebold, however, has compromised democracy in the free world.............. Somehow...I think their intentional lies and manipulations carry a bit more weight.

You should meet a former poster that used to come here..... Freida was her name.....she was cool. You two have a lot in common. HMMMMMMM

[TinfoilHatProgrammer]

79. really, let's be accurate

"A silly reference to a person's title." Really.



DIEBOLD PRESIDENT ANNOUNCES "KINKS" IN ELECTRONIC VOTING PLATFORM, RECOMMENDS PURCHASING OPTICAL SCAN SYSTEMS INSTEAD!!!

or

SOME NON-DIEBOLD SALES FLUNKY ALLEGES "KINKS" IN ELECTRONIC VOTING, TRIES TO MAKE SALE OF OPTICAL SCAN SYSTEM INSTEAD


Which headline is newsworthy, and which one isn't? I'll give you a hint... the one that's a complete, utter lie is way more sensationalistic. But yeah... let's just dismiss it as "a silly reference to a person's title." :thumbsup:

And it's not the most I can slam anyone with, you know full well I slammed people at great, ongoing length in a 350+ post thread yesterday. This is nothing more than an innocent jab at the typical kind of hypocrisy practiced by some of the self-proclaimed BBV heroes.

JC

[stickdog]

94. Now, let's ACTUALLY be accurate here.

Here is the accurate headline:

PRESIDENT OF DIEBOLD SYSTEMS DISTRIBUTOR ANNOUNCES "KINKS" IN ELECTRONIC VOTING PLATFORM, RECOMMENDS PURCHASING OPTICAL SCAN SYSTEMS INSTEAD

So, which of the two headlines below is closer to the truth?

DIEBOLD PRESIDENT ANNOUNCES "KINKS" IN ELECTRONIC VOTING PLATFORM, RECOMMENDS PURCHASING OPTICAL SCAN SYSTEMS INSTEAD

or

SOME NON-DIEBOLD SALES FLUNKY ALLEGES "KINKS" IN ELECTRONIC VOTING, TRIES TO MAKE SALE OF OPTICAL SCAN SYSTEM INSTEAD

[Eloriel]

84. There's a huge difference between

purposely telling lies in the first place (Diebold) and then correcting the story versus innocently repeating the error of a newspaper that apparently no longer has fact-checkers.

I'm SURE you know the difference.

No hypocrisy on Bev's part at all.

Eloriel

[stickdog]

93. Speaking of hypocrites,

why don't your tell us you full real name, address, phone number and personal email account, JC?

I have some important information I'd like to deliver to you personally.

[Name removed]

30. Deleted message

Message removed by moderator. Click here to review the message board rules.

[hang a left]

8. Has there been any

data released on which way elections went in these counties where diebold touchscreens were used? In 2002

edited to add year

[Eloriel]

14. Here are some interesting state-wide results for ya

http://www.bartcop.com/111102fraud.htm

****Poll by Atlanta Journal Constitution/WSB-TV of 800 likely voters on Nov. 1 For Georgia Governor

Roy Barnes (D) 51% up 11
Sonny Perdue (R) 40%

** "Official Results" from the 'Diebold Electronic Voting Machines' on Nov. 5

Roy Barnes (D) 46%
Sonny Perdue (R) 51% up 5 - that's a 16-point pro-Bush swing - was it magic?

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

****Poll by Atlanta Journal Constitution Nov. 1 for Georgia Senate

Max Cleland (D) 49% up 5
Saxby Chambliss (R) 44%

**"Official Results" from the 'Diebold Electronic Voting Machines'

Max Cleland (D) 46%
Saxby Chambliss 53% up 7 - that's a 13-point pro-Bush swing - was it magic?

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

****Poll by MSNBC/Zogby on Nov. 3 for Colorado Senate

Tom Strickland (D) 53% up 9
Wayne Allard (R) 44%

** "Official Results"

Tom Strickland (D) 46%
Wayne Allard (R) 51% up 5 - that's a 14-point pro-Bush swing - was it magic?

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

****Minneapolis Star-Tribune Poll on Nov. 3 for Minnesota Senate

Walter Mondale (D) 46% up 5
Norm Coleman (R) 41%

** "Official Results"

Norm Coleman (R) 50%
Walter Mondale (D) 47% up 3 that's an 8-point pro-Bush swing - was it magic?
Did they let this one stay close because they knew MN loved Mondale?

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

****Poll by St. Louis Dispatch/Zogby on Nov. 3 for Illinois Governor

Rod Blagojevich (D) 52% up 7
Jim Ryan (R) 45%

**"Official Results"

Rod Blagojevich (D) 43%
Jim Ryan (R) 44% up 1 that's an 8-point pro-Bush swing - was it magic?

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

****Poll by Concord, NH Monitor on Nov. 3 for New Hampshire Senate

Jeanne Shaheen (D) 47% up 1
John E. Sununu (R) 46%

**"Official Results"

Jeanne Shaheen (D) 47%
John E. Sununu (R) 51% up 4 that's a 5-point pro-Bush swing - was it magic?


Eloriel

[Pobeka]

28. This is another interesting probabilities question

I weighed in a while ago, and determined that the 3 candidates getting the same vote totals in a single county was, while improbable, actually likely to occur given the number of elections and the number of counties over all the years.

But, now. Could we get data from past elections, with pre-election polls, and see how often we get these kinds of swings, and are they related to Diebold et. al.?

[TruthIsAll]

39. The probability of NO fraud: 1 out of 43,040 (conservatively)

The odds that 4 out of 8 elections would fall outside the MOE (all for the Repubs) is 1 out of 43,040.

The odds that 5 out of 10 elections would fall outside the MOE and ALL go for the Repub is 1 out of 502,435.
The following are the results of 8 hotly contested 2002 Senate races...
Note the deviations between the final polls and the actual results.
Four races fell outside the final poll margin of error and swung sharply from the Democrat to the Republican.

FINAL POLL ELECTION
Dem Rep diff Dem Rep Diff Net change Outside MOE
NC 42 48 -6 45 54 -9 -3 0
MN 47 39 8 47 50 -3 -11 5
AK 51 46 5 54 46 8 3 0
GA 49 44 5 46 53 -7 -12 6
TX 48 49 -1 43 55 -12 -11 5
MO 46 46 0 49 50 -1 -1 0
NH 46 40 6 47 51 -4 -10 4
CO 45 47 -2 45 51 -6 -4 0
Average 46.75 44.875 1.875 47 51.25 -4.25 -6.125 2.5

The probability that a given poll will be correct within the MOE is 19 out of 20 =.95, where the MOE = + or - 3 points.
The average switch was a 6.125% move to the Republican, beyond the margin of error…
Out of 8 races, the probability that ALL will fall WITHIN the margin of error is 0.663420431

This means the chances that 1 or more will fall outside the MOE = .33, or 1 out of 3
But 4 out of 8 races fell outside the MOE! The Probability =BINOMDIST(4,8,0.95,TRUE) = 0.000371751


Probability Matrix
Number of elections (trials) Number of elections outside MOE
N.........3........4.......5.......6
8 0.579% 0.037% 0.002% 0.000%
10 1.150% 0.103% 0.006% 0.000%
13 2.451% 0.310% 0.029% 0.002%
16 4.294% 0.700% 0.086% 0.008%
20 7.548% 1.590% 0.257% 0.033%
25 12.711% 3.409% 0.716% 0.121%
30 18.782% 6.077% 1.564% 0.328%
34 24.065% 8.813% 2.592% 0.627%


Prob (all N) = Repub
3..........4.......5.......6
12.50% 6.25% 3.13% 1.56%


Joint Prob (Outside MOE and all Repub)
N.........3........4.......5.......6
8 0.072% 0.002% 0.000% 0.000%
10 0.144% 0.006% 0.000% 0.000%
13 0.306% 0.019% 0.001% 0.000%
16 0.537% 0.044% 0.003% 0.000%
20 0.944% 0.099% 0.008% 0.001%
25 1.589% 0.213% 0.022% 0.002%
30 2.348% 0.380% 0.049% 0.005%
34 3.008% 0.551% 0.081% 0.010%

Odds =1 out of X
N.........3........4.......5.......6
8 1,382 43,040 2,077,263 159,672,547
10 695 15,557 502,435 23,234,010
13 326 5,156 111,666 3,240,553
16 186 2,284 37,326 791,105
20 106 1,006 12,432 194,355
25 63 469 4,466 52,763
30 43 263 2,047 19,497
34 33 182 1,235 10,208


The odds that 4 out of 8 elections would fall outside the MOE (all for the Repubs) is 1 out of 43,040.
The odds that 4 out of 16 elections would fall outside the MOE (all for the Repubs) is 1 out of 2,284.

The odds that 4 out of 34 elections would fall outside the MOE(all for the Repubs) is 1 out of 182.

The Binomial Distribution Excel Function is expressed as
BINOMDIST(number_s,trials,probability_s,cumulative)

- Number_s   is the number of successes in trials.
- s is the number races within MOE
- Trials is the number of independent trials.
Here Trials is the number of CRITICAL (competitive, close) races
- Probability_s  is the probability of success in each trial.

Here the Probability_s = .95 or 19/20 that the result will fall within the MOE of +/-3%.
Then the probability = 1 -19/20 = 1/20 that the result will fall OUTSIDE the MOE.

- Cumulative is a logical value that determines the form of the function.

If Cumulative is TRUE, then BINOMDIST returns the cumulative distribution function, which is the probability that there are at most number_s successes;
if Cumulative is FALSE, it returns the probability mass function, the probability of exactly number_s successes.

[Pobeka]

42. Yeah, but

You have to consider all the elections that have taken place, over all the years, all the candidates. This isn't a one time event, you can't pick the events to do your analysis after the fact (or my stat prof will hunt me down and beat me senseless for not pointing that out).

Do we have any similar events swinging for Dem's. What about all the past elections? Can we really point to this election and say "wow, this one really did fly in the face of the odds, given the election history that preceded it?".

I think we can (with a heck of a lot of work). And I think the result would be very obvious to the average voter out there when framed in the perspective of:

# 10 point swings in non-diebold elections for dems, repubs
# 10 point swings in all other elections for dems, repubs

See what I mean?

[TruthIsAll]

45. Show this to your stat prof. He will agree the analysis is correct.

It is pure probability theory applied to deviations of final polling results assuming 95% is the probability of a correct poll (within the MOE) and all falling for one side.

Go ahead, get past data. Look at the deviations. Better yet, contact Zogby. He will tell you he never saw anything like the discrepancies in the 2002 senate races. We are way beyond the MOE in at least 4 states here.

This is a pure math problem. We have 10 experiments (races). We know the probability of failure for each race is 1 out of 20. What is the probability that at least 4 out of the 10 will fail?

Use the Cumulative Binomial or Poisson function. Take your pick. That's it. That's all there is. Believe.

[Pobeka]

46. No, I don't believe -- here's why

(A simplified version)

You say the odds are 1 in 20 that we'll have the election swing outside the polls.

Now, you get, say 500 races.

Out of those 500 races, how many are likely to be outside the polls?

0.05 * 500 = 25

You *know* ahead of time that aproximately 25 races will result in swings outside the polls.

This is why you can't pick your races with "hindsight" and say those are "special". They're not special. You predicted it ahead of time.

[TruthIsAll]

52. Wrong. I chose the races which were hotly contested and

could go either way, i.e. the ones in which the repubs had incentive to cheat.

I did not include one-sided races where there was no question of outcome. For instance, would the Repubs try to cheat Kerry when he was leading with 80% in the polls? It would be too obvious if they tried to steal Mass. But they would cheat in a state where the race was close and they needed it to win (steal) the senate.

To satisfy disbelievers, I provide probabilities even for the case assuming all 34 elections are critical (close, which is not the case).

This analysis was originally posted last November, long before you came to DU. It has been verified as matematically correct by Professors of Mathematics here at DU.

Have you completed a course in probability theory?

Show it to your prof, then report back. I do not wish to defend this analysis; I have done so too many times in the past. You missed the threads where this was fully vetted. I am not going to get into a blow by blow discussion here. Consider the case closed. Show it to your prof.

[Pobeka]

53. OK, I understand you''ve been through this before, but 1 *burning* ??

Did you pick the races before, or after the election?

(and can you point to the threads where this was discussed last Nov. I don't want to be a pest if there's nothing new I can add.)

I plead ignorance on previous vetting...

[DagmarK]

75. A subjective consideration.......

the 3 counties in texas that had the 18181 votes FOR the republican...

Those were in S. Texas -- predominantly dem territory. Agricultural areas -- as in.....hard labor workers...spanish speaking population.

AND, interestingly, is the Area that the repukes in Texas are trying to break up with the redistricting!

Coincidence? I don't think so.

On of the Austin districts desired by the repukes runs from Austin straight down to S. Texas......like 600 miles long!

The repukes have a voting fraud plan in place:

1. get redistricting to ensure wins;

2. if that fails...special voting machines.....

But they had to ENSURE a majority in federal and state legislatures to get push the redistricting........hence, the machines.

Purely subjective.....but entirely relevant to the statistical questions at hand.

[gristy]

90. I'm mostly with you, Pebeka

but you don't need to consider all elections. You just need to properly select the subset that you are going to study. I don't think this subset was properly selected.

[gristy]

91. this is subjective

could go either way, i.e. the ones in which the repubs had incentive to cheat.

Is it not possible that for the races where you have the feeling that "the repubs had incentive to cheat", you developed that feeling AFTER seeing the numbers?

To satisfy disbelievers, I provide probabilities even for the case assuming all 34 elections are critical (close, which is not the case).

Please provide me a link to this information.

[shance]

68. Where did you drum up such an idea? All elections are separate events.

And totally different situations, and computerized voting hasnt even been around in years past, so unless you just got here, thats the issue at hand.

[ConsAreLiars]

88. Polls versus elections

Pobeka, you ask the right question:
Can we really point to this election and say "wow, this one really did fly in the face of the odds, given the election history that preceded it?".

The first problem is that the margin of error ("results are accurate within + or - x%") means only that if you go through the same polling routine 20 times you can expect the numbers yielded by one sample to fall outside that range and 19 to fall within (usually). It says nothing at all about the utility of the poll to predict any real world action. Poll results only give estimates of the results one would get if everyone in a certain category were asked the same questions, but for anything else, like predicting actual decisions, a lot of assumptions and guesses are needed. Obviously there are numbers of steps that can be taken to improve the value of the poll as a predictor, and the better operations do a lot to try to devise methods for getting answers that are useful for predicting elections, but the margin of error numbers say nothing at all about how close the poll numbers might be to election numbers. The "margin of error" numbers simply tell you how many people were interviewed.

In fact, one source indicated that election results which actually fall within the margin of error for polling samples seem to be rather rare:
http://commonsense2000.tripod.com/poll.htm

The polling industry interest group NCPP, using a number called "candidate error," reports that of 159 polls on senate and governor races in 2002 16% fell outside that range, but the public document does not give actual distributions. Interestingly, "candidate error" numbers are actually one-half the real error. It seems that the justification is that that deviations from polling sample numbers occur when individuals switch sides at the last minute, rather than assuming the actual numbers reflect simply how many partisans from each side showed up and had their votes counted. Even so, using their industry-friendly analysis, the odds that election results will fall outside the expected margins is one in six rather than the assumed one in twenty. Using the less industry-friendly assumption, results outside the margin of error might be just as likely as results within.
http://www.ncpp.org/2002SenGovPoll/2002ElectionPolls.html

So the first question is "how closely correlated are poll-based estimates and election outcomes?" This is an empirical question. To answer it means devising some method for gathering together in an unbiased manner a large enough amount of data from polls and elections. The next question is whether some factors can be identified which have a significant impact on the correlation between polls and vote tallies.

One hypothesis might be that the more Diebold machines in use in an election the lower the correlation between polls and results. Another level of analysis might be to separate out "hotly contested" versus other elections, using some pre-election criteria to define the two groups. Etc.

The difficulties are several - 159 cases is not a lot to work with if this one cycle of statewide races was the sole focus. Getting more would be a lot of work even if the data behind the NCPP report could be obtained. Also there is no reason to suspect that ES&S systems are any less vulnerable that the Diebold machines, given their common parentage. Further, if there was some sort of tampering going on, and if a detailed comparison of polls and votes were done, the leap from "anomolous" to "fraudulent" is still a matter of inference rather than proof.

On the positive side, what could be done, if those 159 bits of data could be obtained, would be simply to graph them out and color-code the D and R sides and/or something about the voting tech used. Trying to put odds on a specific group of results is impossible with the info now available, and collating all the relevant data would be something that could only be done with a large research grant, unless this data has already been compiled at some location I haven't yet found. Such a graph might show a few races standing off in their own space. apart from the bulk of the cases, or the races mentioned here might just blend in with the rest due to all the noise in that chart.

[ConsAreLiars]

49. You should know better

You can NOT pick a few polls out of hundreds and a few senate races out of dozens - pick them precisely because they are odd - and then run the probabilities that your picks are accidental. Of course they are not accidental! You picked them out precisely because they are strange! The odds that you picked them out at random are close to what you calculate (near zero), but the odds that they are out of the normal range of events requires exactly the kind of data (all races and all polls) that Pobeka requested. Without that data your numbers are just the same sort of malarkey that the "intelligent design" morons use to "prove" creationism. I agree that these numbers are suspicious. But proving this depends on looking at ALL the relevant data, not just a few ideologically selected datapoints. It is a real shame and a warning to small "d" democrats of every stripe that this comprehensive election data and polling data is so hard to find, but such calculations cannot be meaningful without looking at ALL the relevant data.

[Pobeka]

50. So what do you think of my post #42?

It's a heck of a lot of work, would require more than just a couple of people, and I'm not even sure if the data exists?

[TruthIsAll]

55. Bumbler, listen up.

I did not cherry-pick. There were 34 senate races, in case you don't know. There were only 8 or 10 which were competitive. Read any pre-election analysis if you don't agree.

And I chose only those states from the original Scoop analysis. Sure, I did it after the fact. Makes no fucking difference. Before or after, the numbers would work out the same. I did it afterwards out of curiosity to see what the odds were for the elections to change overnight from what the polls predicted.

The UNIVERSE was the universe of HOTLY CONTESTED elections.

THIS UNIVERSE IS NOT IN DISPUTE: SEE THE STATES IN THE ANALYSIS, THEY WERE THE ONES THAT ALL THE PUNDITS WERE TALKING ABOUT BEFORE THE ELECTION.

I SUGGEST YOU RE-THINK THE PROBLEM.

IF I AM WORKING WITH A DECK OF CARDS, I BASE MY PROBABILITIES OF DRAWING A GIVEN HAND FROM THE COMPONENTS OF THE DECK- 52 CARDS.

HERE THE DECK CONSISTED OF TEN HOTLY CONTESTED ELECTIONS. IF YOU WANT TO ARGUE YOUR POINT, THEN GO AHEAD, TELL ME THAT ALL 34 RACES WERE HOTLY CONTESTED. AND YOU WOULD THEN IMMEDIATELY REALIZE THAT YOUR ARGUMENT HAS NO BASIS IN FACT.

AS FOR THE POLLS THEMSELVES, I USED THE ONES GIVEN IN THE SCOOP STUDY, WHICH IS WHAT IS REFERENCED IN THIS THREAD.

YOU HAVE HEARD OF ZOGBY, HAVE YOU NOT? DO YOU FEEL THAT HIS POLLS ARE NOT ACCURATE? POLLS DIFFERED, BUT THE FACT IS THAT VIRTUALLY ALL FELL OUTSIDE THE 6% MOE. THE ONES CITED HERE WERE WAY BEYOND 6%, SO IN FACT MY RESULTS WERE CONSERVATIVE.

BUMBLER, HAVE YOU TAKEN PROBABILITY THEORY? IF SO, SHOW IT TO YOUR PROF; IF NOT, TAKE THE COURSE.

[Pobeka]

56. OK TIA, I see it now.

Yes, that make sense to me. Sorry to put you through the discussion again. the 1 to 43,000 (or whatever) odds are correct in that case.

Do you think a similar analysis could be done for previous elections with Diebold equipment? I really have no idea if the data exists, do you?

Thanks for putting up with the ignorant one here :-)

[TruthIsAll]

58. Pobeka, welcome to DU! I'm glad we agree; your post was legit.

The questions you posed have been posed by others at DU. Unlike you, some of these individuals were too stubborn (or lacked the knowledge) to understand the analysis. They accused me of cherry-picking polls, and that I assumed statistical independence (a fair assumption) when they felt that the states were dependent, etc. All sorts of straw men. And this was long before anyone was even considering fraud in Nov. 2002.

We have come a long way since then. Diebold had the means, the opportunity and the motive. Bev Harris and Johns Hopkins have taken us to where we are now. My faith in probability mathematics and plain gut feeling that some strange shit happened in these elections were confirmed by the analysis.

[TruthIsAll]

59. Pobeka, welcome to DU! I'm glad we agree; your post was legit.

The questions you posed have been posed by others at DU. Unlike you, some of these individuals were too stubborn (or lacked the knowledge) to understand the analysis. They accused me of cherry-picking polls, and that I assumed statistical independence (a fair assumption) when they felt that the states were dependent, etc. All sorts of straw men. And this was long before anyone was even considering fraud in Nov. 2002.

We have come a long way since then. Diebold had the means, the opportunity and the motive. Bev Harris and Johns Hopkins have taken us to where we are now. My faith in probability mathematics and plain gut feeling that some strange shit happened in these elections were confirmed by the analysis.

[Pobeka]

60. But now the challenge

Is how to present it to the average american.

That's where I came up with the how many "swings" occurred in previous elections of this magnitude. You know if we had a simple bar chart, and it turned out like this:

# swings    *
            *
            *
            *
            *
            *          *
------------------------
     Diebold           other

See what I mean -- this would get the average american engaged.

[DEMActivist]

62. The Georgia races are online...

the actual numbers, that is.

http://www.sos.state.ga.us/elections/results/2002_1105/default.htm

[Pobeka]

64. How about relavent poll results?

any similar data -- anyone?

[gbwarming]

61. You confuse MOE of the poll with actual voting.

MOE is a measure internal to the poll. It doesn't have anything to do with predicting how people will actually vote. It only describes the reliability of the poll sample to reflect the population on the poll questions.

MOE is NOT a measure of poll to vote correlation.

[Pobeka]

63. Depends on the poll question

If the question is "who are you going to vote for?", then I'd say for all practical purposes we have "alignment" between the poll and the election outcome.

on edit: assuming of course, all poll respondents have the same chance of voting in the election, which would seem to be the null hypothesis

[gbwarming]

65. No actually, that's not correct

The margin of error ONLY describes the range of poll results you would expect to get if you asked different samples of people the same question. The whole point of polling is to predict what voters will do when they vote but without serious work to correlate polling to voting results across a large sample of races you can't assume that polling and voting numbers are closely related.

I'll just repeat this for TIA: Poll MOE is not a measure of voter behavior. You cannot draw any conclusions by noting that election results fall outside the poll moe.

[Pobeka]

67. It's late, and my brain is slowing up :-)

I understand what you say here. I don't know (this late), is there a way to get from the poll MOE to the standard deviation of the sample? Then one could make the direct comparison.

I'll think on this tomorrow if no one else knows tonight.

[punpirate]

70. You're quite right about the MOE being the fudge factor...

... internal to the poll. However, the poll itself is meant to be a predictor of voter behavior, so what is internal to the poll is a legitimate variable in analyzing it against the final vote. The poll results don't stand in absolute isolation from the MOE--they are integral to it. The MOE defines the margins of the poll. In order to even treat the poll as a predictor of the vote, one has to accept the marginal figures as the outside boundaries. A final vote exceeding those boundaries is an anomaly--as long as one accepts that the poll protocols are legitimate.

Cheers.

[gbwarming]

76. MOE isn't a fudge factor.

Margin of error is precisely defined. It's been a long time since I did these calculations so I had to look them up:
http://www.stat.psu.edu/~antoniou/stat250/ch10.pdf

It is, for any given confidence interval, the boundary on the variation in results you would expect to get by taking a different sample of the population. It does not relate polling to voting in any way.

I think what you need to do, in order to pursue this statistically, is establish the correlation between polling and voting on a large sample of elections and then test the sample of interest (contested 2002 senate races) to see if they fall outside. The definition of contested is slippery too.

[punpirate]

82. I don't think you read what I said.

Look at the poll in total, and see the MOE as a boundary of expected results. That _has_ to be seen in conjunction with actual results, since the poll itself is meant to be a predictor of voting behavior, within the boundaries defined by the MOE.

And, you can say that MOE is precisely defined, but to the ordinary person, it's a fudge. Otherwise, the poll would stand without qualification, without an MOE. The MOE exists, basically, for the pollster to say, "our poll is right, but only within these limits." Those limits must be considered when compared against actual results. You can argue from the basis of statistics, but, you know, too, that an MOE is a tolerance--it's wiggle room, based on the size of the sample.

That doesn't change the fact that the MOE has to be considered when comparing poll results against actual results--the original issue. If the poll were accurate enough (i.e., a large enough sample) that the MOE were negligible, then the poll results could be compared to actual results with considerable accuracy. The inaccuracy implicit in a small sample is the reason for the MOE, and the reason why the MOE must be considered when compared to actual results. Without factoring in the effects of the MOE, one has no choice but to treat the poll results as an absolute, which, with an MOE qualifier, they most certainly are not. To ignore the MOE in such comparisons is to ignore an essential data point. If the MOE is precisely calculated, then it must be considered as a data point in all calculations of odds, because to ignore it would skew results. The MOE adds uncertainty to the odds calculations, rather than reduces uncertainty.

Cheers.

[ConsAreLiars]

81. All the SHOUTING

suggests I may have hit a sore spot. There were lots of polls and more than 8 elections. The message you responded to included an number of governorship races as well. If you don't want look at the whole picture you don't have to, but you cannot claim to present the whole picture if you do not do this. If I recall correctly you did an equally bogus analysis of the three 18181 occurring in one county without factoring in the number of counties in which elections occurred. Pobeka did the full analysis, consulting with professional statisticians, and found the threesome not to be improbable.

Are these reversals (from polling to vote "counts") suspicious? YES! But proving this, or even determining the odds, requires looking at ALL the polls and ALL the elections, not cherrypicking a few. When you only look at the odd cases, and pick them out for examination because they are odd, you prove nothing. This is disappointing, because I truly believe, but still lack the evidence to prove, that the cases you cite are genuinely wierd.

As for consulting with my professors or my personal qualifications, the odds are fairly good those profs are no longer any more responsive to questions than Houdini, and I suspect that the university in which I once taught a class on statistical analysis was no less rigorous in its expectations than whatever school allowed you to confuse post hoc with prior hoc.

That being said, and in order to avoid rancor, i want to say that 95% of your DU posts are in the top 10% in terms of truthtelling, but in this particular area your devotion to the cause lets you skip a few essentials. Pobeka's approach is necessary if we are to prove that what you (and I and others) believe - that those election reversals which you cite are extremely improbable. Pobeka's proposal, to gather all the polling/election data together and do the numbers, is essential if we want to get beyond the charge of selective perception. My own searches have failed to discover any sort of comprehensive election summary or associated polling stats. Maybe some of this work has been done by pollingreport.com or by some academics, but I haven't yet found it. Maybe you can help find this data?

[punpirate]

66. Well, yes, no, maybe...

... if I may, since I've asked a few questions of both TIA and Pobeka about analyzing this statistically. Let's try to redefine the terms, and list some of the variables, and that may help.

First, for the races in question, there were not hundreds of polls, but, rather, just a few. The published margin of error is likely slightly different between them, but, the first essential question is: of the polls taken for a specific race in question, do all the polls err from election results beyond the margin of error, and in an opposite direction? That can be quantified, and odds assessed on the basis of all polls taken shortly before the election.

Second, I agree that trends over time are relevant and can be fit into the general matrix of variables to be considered. However, for them to be relevant, there must be similar polling data and election results to compare. Even then, detailed analysis would be required to define anomalous voting behavior (economics at time of vote, changes in votes cast vs. registered voters, universally popular candidates pulling large numbers of swing voters and many crossover votes, time between poll and vote, etc.). These are not easy to quantify, and for that purpose may render statistical analysis over time as suspect.

But, the largest fly in the ointment is the principal one in question. Is there some means of codifying the probability of voting fraud according to the manner in which the vote is cast? With paper ballots, with lever machines, with electronic voting. Remember, in setting up this problem, that touch screen voting machines are new. The history is long with other methods, but not with DREs. To elaborate, if historical data show that paper ballots, for instance, have a much better alignment between pre-election polls and final tallies, would that not be an indication that, at least for the races chosen, of an anomaly in the DRE-counted races? Yes, certainly. Can this be quantified statistically? Maybe not.

It's for those reasons, and more, that I think that it is easier to see these races in isolation. To go further, each race is fairly unique in its characteristics, and it is only in those races where there are repeat candidates on both sides that historical comparison is useful. With each new set of candidates for a particular office, intangible variations from previous races exist. To define those intangibles further complicates historical comparisons, since one has to know, and somehow quantify, all the reasons why people vote in the ways they do--is one candidate handsome, and the other not, etc. All those approximations vary with each set of candidates for a particular office. In this sense, too many unquantifiable intangibles would invalidate historical data.

As for cherry-picking results (34 races, but only nine or ten examined), maybe there's some merit to that argument, but, again, the means of casting the vote varied enormously in each of those races. It might be more interesting to sort all 34 races by type of equipment used, precinct by precinct and look for anomalous results there first (daunting task producing that data, though). It also might be useful, in those races which did go as predicted by polls, to look for clear historical patterns in just those races, if only to verify that those races were statistically predictable over time. Such would further indicate anomalous behavior in the races in question--one should be able to answer the question, from the available data, as to why this anomalous vote occurred. If one can't, then it's not beyond the pale to treat the odd results with additional suspicion.

Whatever the means of reconciling the divergent views here, it should be clear from just these few messages that trying to statistically analyze voting _in toto_ is a horrible task to contemplate, because of the number of variables (each one is required to be analyzed in isolation from all others, for the sake of mathematical simplicity) and the inability of the statistician to put a numerical value to intangibles should be notable considerations.

Voting behavior is simply a lot more complicated than we might have imagined. Fairly thorny problem, to say the least.

Cheers.

[Pobeka]

69. As usual, you have a lot of good thoughts here

And before I go to bed, let me toss this out.

Can't it be as simple as:

How well do polls predict election results, by voting method?:

Paper
Punch
Touch Screen.
...

I realize there is a nasty hornets nest in the details, but to me, it's a question of what level of abstraction do you want?

Thanks all.

BTW Punpirate, I scratched out the exact algorithm needed to compute the odds of 3 candidates getting the same number of votes. Got the same number as my simulation which is re-assuring. If your interested I'll share the algorithm -- it's a loop within a loop.

[punpirate]

72. Nah, the simulation says enough...

... given the very strict limitations imposed by the number of variables considered.

On the business of the details, it's not so much a question of a level of abstraction _I_ want--it's more that when considering voting as an abstraction, the variables are very ill-defined because they are intangibles, but they still do have considerable influence on the process of voting.

Human behavior can be defined, just so long as the number of variables are limited and quantifiable and are tested by other means. As an example, perhaps one can make a calculated guess about the success of a politician running for office on a primary theme of "security." "Security" as an issue burrows into people's notions about their own safety and security, and might sway one's vote by playing on the voter's elemental fears. Can that be quantified? Yes, by psychological testing separate and apart from the process of voting.

Now, take that same politician with a message about "security," and add in the following: this person is truly ugly, utterly repulsive in both appearance and behavior. This guy picks his nose and eats his own snot while giving his spiel on security. Does he now fit the previous psychological data? Not on your life. Voters run from this guy. He's not successful at all in his campaign for reasons entirely different from the psychological (and statistically proven) foundation of his message.

That's why almost every race has to be seen in some degree of isolation, and in conjunction with poll results--the poll results reflect, to some degree, those intangibles which can't be statistically quantified. And for that reason, I've found the 18181 winning results more aberrant than you have by purely statistical means.

It's truly a bugger, as I've said. Add in the tendency of the press to misquote, or selectively quote from a candidate's message, and it gets very squirrelly, indeed. That's why comparisons against poll results may be the better method in establishing variance from the norm.

Cheers.

[DEMActivist]

71. OK, now let me complicate it a bit more for you.

First, within Georgia you have 2 of the fastest growing counties in the United States - Paulding and Forsyth Counties. A great deal of the influx in these traditional southern counties is "northern" companies moving HQ to this area. So you have a group of well trained, well educated, technically trained people moving into the areas.

Now, if we take Max Cleland's race, and compare it to the 1996 Senate race you'll find few differences in the actual candidates (very conservative R vs. moderate D).

Add to the mix that the "metro Atlanta" area is typically heavily Democratic and typically carries the races for the Democrats. The 2 growing counties are part of the Metro area as well.

The "excuse" for the republican win, immediately after the race, was that an influx of "angry white males" went to the polls to vote against Roy Barnes - the Governor who changed the state flag without putting it to the voters.

However, when you look at the actual statistics, no angry white males show up at the polls.

Some other key points from the CFVR comparative analysis:

Between 1998 and 2002, overall active voter registration fell by 142,682 citizens. This reduction in the voter roll, even as Georgia’s population showed strong growth during the period, is the result of list maintenance efforts established by NVRA (the federal “motor voter” law).

White males, white females and black males all showed reductions in voter registration between 1998 and 2002.

Only black females and “other” voters showed increases.

While the size of the voter roll decreased, the number of Georgians who cast ballots increased -- by over 218,000. Hence, “turnout” (number of those against the base of those registered) grew by nearly eight percent.

Voter registration and participation by black males continues to lag far behind that of black females. Indeed, the lower overall voter participation by African-Americans compared to whites is attributable to this factor. Black females continue to register and vote at rates similar to white males and females.

In 2002 as in the past, voter participation increased significantly with age. The highest percentage turnout was among registrants 60 – 64, with nearly 71 % turnout. The turnout percentage of young adults 18 – 24 was nearly 50 points lower.

http://www.sos.state.ga.us/pressrel/022503.htm

Also, let us not forget that the Democrats held control of the Georgia government for 134 years until 2002. In addition, the Democratic Secretary of State, Lt. Governor and Attorney General were all re-elected that night. Only the Governor, the longest sitting State Speaker (Tom Murphy) and the Senate races were lost and the republicans need the senate seat to take back the majority in DC.

There just isn't anything to support the anomolies we saw in this election.

[punpirate]

78. I read, a week or so ago...

... an analysis of voter behavior in this election compared to previous ones (as a means of refuting the charge that an increase in white male voting combined with low black voter turnout caused the end result) and the changes in who voted in the 2002 election were quite predictable--the ratio of votes from each group remained virtually unchanged (and, wasn't there a slight uptick in 2002 in the percentage of registered black women voters who voted?).

And, it's still very hard to justify the results in the gubernatorial race by citing the raw campaigning done by Chambliss in his campaign against Cleland (Cleland's lack of patriotism, etc.). Crossover votes may have accounted for some of that change from traditional voting patterns, but not all. It was a helluva time for VNS to go belly-up.

On the latter, VNS said, consistently, that they couldn't predict results because of database software failures. Still, there were counters in the field, even if they couldn't submit data. Has anyone tried, in such races as Georgia, to collate data after the fact from VNS workers in GA?

Cheers.

[dansolo]

31. Blagojevich won

Bartcop has the Illinois governor race wrong.

[aquart]

86. At any time, did a Diebold machine

Switch to a Democrat when a Republican was favored?

[SoCalDem]

87. At least 4 times that I remember

2 in Georgia.. senate & gov..

[gristy]

89. I am on democracy's side.

But statistically speaking, it is of no value to take a subset of a data set where the criteria for chosing elements in the subset is to prove your point. You have to take a RANDOM sampling of the data set, or better yet, the ENTIRE data set (e.g., every congressional election in 2002), and apply your analysis. Maybe analyze it as one set, then split that set into two subsets: one where Diebold election systems were used, and one where they weren't used.

[Generic Other]

9. "Kinks"

Those kinks just keep getting harder to find...wooo wooo.

Now there's just one question they need to address--Was it a case of LIHOP or MIHOP?

[acmavm]

11. Absolutely stunning piece of work and research. But beyond showing

that the system has serious flaws, this also illustrates that the American people have others working for the good of this country who will use their precious time and energy to investigate these issues and are not afraid to step up and tell the truth. It also illustrates that we are not as ignorant as the "experts" think we are.

Fantastic!!!!!!!!!!!

[Eloriel]

17. Uhhh

I'm sure you mean that as a compliment for Bev&Crew, and boy I share that. They've done fabulous work and we all owe them so much, even if this were the end of the story (which I'm sure it isn't).

But I have to confess that your comment also makes me a little uncomfortable. Why? Because the American people can't and shouldn't rely on the extreme good fortune of having a Bev and a DEMActivist and a few other people who would pursue this with bulldog tenacity.

IOW, what we should have all learned in 2000 and since is that what our Founding Fathers told us was true (and of course the exact quote isn't coming to me at the moment): we can't take our freedoms or democracy or anything about it for granted.

I now understand -- not just because of Bev&Crew but their work does figure in -- that we need an enormous new and sustained emphasis on Citizenship in this country. There are always going to be forces afoot who try to take it from us. Our foes have learned incrementalism and patience, and they've found ways to leave the illusion of democracy while they distract and entertain us in a thousand million different ways and steal our freedoms and our rights and even our vote from us.

If any of us takes ANY of this for granted, from the stolen election of 2000 to the voting machine issue now or any other of possibly thousands of issues, thinking "someone else will take care of this for us," then we probably don't deserve to keep the democracy after all.

That sounds harsh, but it's realistic. The battle is far from over -- there is MUCH work to do re the voting machines issue and not a lot of time. Primaries start in January (New Hampshire). My hope is that all the states will make an abrupt halt (de-certify) these and other machines. But ThEN what? We're going to have to vote somehow.

And even if we get this voting machine problem solved, there are legions of other things that need to be addressed before we can really and truly take back our country. And even then we will have to remain eternally vigilant. As I said: I now understand that these anti-democratic forces are always with us, always present.

My heartfelt thanks to Bev and DEMActivist and the others. Now what are the REST of us going to do? It's up to us, not them.

Eloriel

[xcentrik]

13. The software came from....!

Was I the only one who just choked when it was revealed that these things ran on off-the-shelf, hackable, security-hole-ridden Microsoft Access?
:wtf:

[Eloriel]

19. You may be one of few who didn't know that before

Even so, it has its ardent admirers here at DU, if you can believe that.

ROTFL.

Eloriel

[Name removed]

57. Deleted message

Message removed by moderator. Click here to review the message board rules.

[ozymandius]

21. Is it too late to add this to your book?

This story makes those letters and phone conversations between my state officials and me sound like utter shit. I doubt that anyone will be broaching this subject so willingly.

You know - this just casts a deep dark shadow of illegitimacy over those who now hold office in the wake of the 2002 elections. Is it too early to call them "squatters"?

[DEMActivist]

32. I know the feeling well

Perhaps now you understand my utter disgust with Cathy Cox, Roy Barnes and some of the other Georgia Democrats we discussed so long ago on the phone.

It really hurts to know, deep in one's psyche that the leaders we placed trust in have so completely and utterly betrayed that trust.

I can't discuss the specifics publicly, but have no doubt these folks have not heard the end of this in Georgia.

[Vadem]

24. WhoooHoooo Bev! You've got Diebold on the run, as well

as the other companies, who are, I'm sure, pouring over every press release and study, holding their breaths and wondering when it is their turn next!! Great job!

[Rose Siding]

25. This Kiffmeyer cookie in GA says-

"She pointed out that Georgia used new touch-screen machines in its 2002 elections without incident."

http://www.wired.com/news/politics/0,1283,59874-2,00.html

How does she think she would KNOW? Warts? Peculiar odor? Tingling sensation?

[punpirate]

33. Kiffmeyer is in Minnesota, I believe...

... she was using Georgia's 2002 election as, ahem, an example. She's been made chair of NASS, so I suppose she has to demonstrate, uh, competence in the election arts....

[DEMActivist]

34. ELECTION 2002: ELECTRONIC VOTING: Glitches mar day for new system

ELECTION 2002: ELECTRONIC VOTING: Glitches mar day for new system
BYLINE: DUANE D. STANFORD, STAFF
DATE: 11-06-2002
PUBLICATION: The Atlanta Journal and Constitution
EDITION: Metro
SECTION: News
PAGE: A.1

Voters across Georgia raved about touch-screen voting Tuesday, but the system's statewide debut had plenty of glitches.

Crashed terminals, jammed voter access cards and too few machines in some counties were among the many technical problems reported in scattered precincts statewide. The troubles created delays at some polling places where turnout was already heavy, and long lines were common from metro Atlanta to rural South Georgia.

Other problems were more serious.

Republican poll monitors took complaints from more than 100 voters in at least 20 counties around the state who said the machines highlighted a Democratic candidate even though they selected the Republican.

And a judge in Terrell County, near Albany, held four polling places open past 7 p.m. after a ballot mix-up forced poll managers to close the sites earlier in the day. The error, which allowed some voters to vote in the wrong County Commission races for a time, could affect the outcome of two races, election officials said.

snip...

In five of the six precincts in Terrell County, the ballots on the voting machines listed the wrong County Commission races. The error was discovered by a poll manager before the polls were opened.

The southwest Georgia county's largest voting district was shut down, but in the other precincts, 121 ballots were cast in which voters could have voted in the wrong County Commission race, officials said. Four of the precincts were shut down for two hours and another for one hour while the problem was fixed.

Ralph L. Phillips, a Republican candidate for Terrell County commissioner, was one of those affected.

"My wife and I went to the polls this morning at 8:15 a.m.," said Phillips, an attorney and farmer. "I wasn't even on the ballot."

All the Terrell County polls were reopened by late morning. A judge later ordered two precincts to remain open until 9 p.m. and one precinct to stay open until 8 p.m.

In other counties, some voters complained the touch-screen units were changing their votes from Republican to Democrat.

State Republican Party attorney Randy Evans said his election monitoring team received reports from at least 100 voters in more than 20 counties around the state who said the machines highlighted the wrong candidate box after they made a selection. The problems appeared to be limited to the races for governor and U.S. Senate, said Evans, who was collecting sworn statements from voters.

"We want to get the information while it's fresh in their minds," he said. "These reports are virtually identical."

At Henderson Mill Elementary School in DeKalb County, a voter approached poll manager Elizabeth Perkins to complain that her vote was jumping on the screen, dancing from Republican to Democrat.

"She had pushed Saxby, and it went up to Cleland," said Perkins. "I said, 'Try it for the governor,' and it did it again. We cleared it again and started again, and it came out all right."

Two other voters on different machines reported the same problem. Technicians arrived in about 15 minutes and rebooted three machines, the poll manager said. The technician gave the poll manager no reason for the problem, and the problem did not recur.

---

How soon they forget, huh?

[DEMActivist]

35. And some more "without a hitch" stories

BYLINE: TY TAGAMI, STAFF
DATE: 11-13-2002
PUBLICATION: The Atlanta Journal and Constitution
EDITION: Home
SECTION: METRO
PAGE: B.6

Fulton County election officials said Tuesday that they discovered more errors while tallying last week's votes, but they said the final results didn't change the winner of any race.

The officials submitted results to the county Board of Registration and Elections for certification Tuesday. The board approved the count without objection, although the department's final tally was 3,621 votes more than a preliminary total released the day after the election.

Officials attributed the difference in part to about 3,100 votes recorded on 67 of the new computer voting machines whose contents went uncounted until Friday. The machines' memory cards had been misplaced by poll workers.

An additional 533 votes turned up after officials discovered they had miscounted absentee and provisional votes recorded on paper ballots, said John Sullivan, chief of voter registration. He said the errors occurred when election workers ran paper ballots through optical scanning devices.

Ballots from a half-dozen precincts were run through the scanners twice, while those from at least a dozen precincts were not counted at all, Sullivan said.

Cynthia Welch, the county elections chief, said some poll managers failed to deliver some or all of their computer memory cards to election headquarters, and no one noticed until later. "It was totally chaos down there, and some people did slip through the cracks," she said. "We definitely have a lot of training we need to do."

[DEMActivist]

36. And another....

Touch and no-go in Fulton? Unreadiness for new voting system raises Florida effect fears

BYLINE: DUANE D. STANFORD, STAFF
DATE: 09-28-2002
PUBLICATION: The Atlanta Journal and Constitution
EDITION: Home
SECTION: METRO
PAGE: E.1

State and local leaders are concerned Georgia's largest county might not be ready for the transition to touch-screen voting in just 38 days. Problems with demonstration machines at Fulton County polling sites during the Aug. 20 primary and a cramped training schedule for poll workers have aggravated the worry.

With this month's botched Florida election still fresh in their minds, state legislators from Fulton gathered Friday for a progress report on the county's preparation for Georgia's November general election.

State Sen. Vincent Fort (D-Atlanta) left angry.

"I am not sure at this point that Fulton County will not be another Broward County," he said, referring to one of two Florida counties where poll worker mistakes with new touch-screen units left a close gubernatorial primary in doubt for a week. "I have heard very little at this point that convinces me this won't be a fiasco."

County elections director Gloria Champion reported that at least 11 percent of the touch-screen machines set up for demonstrations during the primary failed. Part of the blame went to a technical glitch that has since been fixed, state officials said. But poll workers also incorrectly plugged in electrical cables and set some machines on the wrong mode.

Fulton County Commission Chairman Mike Kenn, who was not at the meeting, said he was wary of the computerized voting system. "I'm extremely fearful that we're going to have a repeat performance like Florida," he said.

[scarletwoman]

37. Kiffmeyer is the Repug Secretary of State of MINNESOTA

not Georgia -- btw...

In any case it's true, her remarks are utterly absurd -- but quite typical. We're gearing up here for a significant battle over voting machines -- she is determined to have DRE's installed in every precinct by 2004.

sw

[SnohoDem]

26. Fabulous job, Bev!

n/t

[SimpleMan]

38. A way to give this story some real legs...

This is just a passing thought, but after reading this post and seeing how many machines like this were used in California, wouldn't it be in Gray Davis' interest to make a HUGE deal about these machines?

It would be a great reason to put off this recall until the spring Primary like he wants. Plus, it would get the real truth about this voting machine mess out front and center.

This is a perfect opportunity. We win because Davis has a MUCH better chance in March, AND we win because this voting machine fiasco will be exposed to the nation.

Problem is , I don't have a clue how to get Davis to notice.

[alfredo]

43. A big fat

hairy Kickarooni!!!

good job Bev

[Fredda Weinberg]

44. Whoah ... was there a reversal?

the same day that Ellen Messmer's story in InfoWorld appeared, July 25

http://www.infoworld.com/article/03/07/25/HNvote_1.html

appeared, you had this official statement from Diebold

"The researchers installed and analyzed a prior version of the AccuVote-TS software"

Google link

later, the company posted a pdf version

http://www.diebold.com/technicalresponse.pdf

The claim that the company reversed itself isn't supported here either

http://www2.diebold.com/checksandbalances.pdf

and that link came from Avi Rubin's site.

[stickdog]

95. You can rest easy, Fredda.

Diebold Corporate is still officially lying through their teeth.

[oasis]

74. Encouraging news.Justice is about to turn the corner and run to dayligh

:thumbsup:

[4dog]

85. Giving a kick so I can find later...

[oasis]

92. Good research on a most important issue. (eom)