polls vs diebold

In other words, I find 100% probability that the op-scan tallies were hacked!

Dave

this is great- I get it-

well I know that you know what you're talking about and can explain it to anyone who would challenge you and I can refer them to you- which helps a lot for those of who never took statistics but know that when the polls show Obama winning across the state and indeed he does, except that is where Diebold counts the votes,  we just scanned, oops I mean scammed again.

thanks Dave

 andi

by andi novick (52 articles, 0 quicklinks, 0 diaries, 14 comments) on Thursday, January 31, 2008 at 8:48:27 PM
 

Math Instincts

I'm so glad the qualified confirmed gut instincts. 

The most ridiculous assertion by deniers I've read is that hand-counted precincts in NH are different politically from opti-scam precincts.  That would only work if the official results were 81-19 Clinton, since 81% of the voters voted on optiscam.  Not only did that writer not take any statistics courses, but he seems to have failed basic proportions.

The other assertion that amused me was some DLC whiner accusing us of only attacking NH based on our preference for Obama, since he heard nothing about South Carolina's machines where Obama won.  SC - as has been blogged - conducts illegal, unconstitutional elections whenever the vote count is secret.  

If the DLC whiner were sincere about honest elections, he'd sue SC for violating its own Constitution.

Glad you put your findings out there, David.  Thanks.

by Rady Ananda (97 articles, 246 quicklinks, 19 diaries, 697 comments) on Thursday, January 31, 2008 at 9:32:48 PM
 

Which Set of Numbers Is Correct?

Previusly, EDA posted the following results and percenatges:

Clinton Op-Scan : 91,717  --  52.95% 

Obama Op-Scan : 81,495  --  47.05%

Clinton Hand-Count :  20,889  -- 47.05%

Obama Hand-Count : 23,509  -- 52.95%

 

Now we're being told that the vote totals and percentages are different:

Clinton Op-Scan: 52.73%

Obama Op-Scan: 47.27%

Clinton Hand-Count: 46.75%

Obama Hand-Count: 53.25%  

That's with 81,753 total votes for Op-Scan and 35,864 total votes for Hand-Count, whereas the previous vote totals were (if I can add correctly) 173,212 and 44,398 respectively.

So why have the percentages and vote totals changed from EDA's initial reporting?!? Which set of vote totals and percentages is correct?

by Sam Adams (0 articles, 0 quicklinks, 0 diaries, 81 comments) on Friday, February 1, 2008 at 5:39:18 AM
 

My understanding of the history of this

I don't know the origin of the original numbers, which were so fantastically mirror images as to evoke immediate suspicion.

The second set of numbers came from Bruce O'Dell after he had carefully waded through the official numbers. I am only guessing, but I think that Bruce became suspicious that the earlier numbers were too good to be true and that EDA would be discredited if it tried to make hay of false numbers. Anyway, I am of this view, which is why I quoted only Bruce's numbers in my article

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Friday, February 1, 2008 at 11:28:50 AM
 

Correct NH Numbers Updated at EDA Weeks Ago, Sam.

CORRECTED Statewide New Hampshire 2008 Democratic Primary Analysis
See CORRECTION Note further below (jump link: Read More)

Clinton: statewide optical scan tally
95,843
52.73%

Obama: statewide optical scan tally
85,910
47.27%

Clinton: statewide hand-count tally
16,767
46.75%

Obama: hand count
19,097
53.25%

While the actual difference between Obama and Clinton hand count and optical scan margins are not a mirror image of each other to four decimal places as we had initially believed*, the undeniable fact that Obama appears to have carried the hand-counted tally statewide, while Clinton carried the optical scan statewide tally -- by almost exactly opposite margins -- remains a remarkable result.

* See Correction Note below for full explanation of originally posted information.

---

CORRECTION Note on EDA's Previously Reported Statewide Anomaly

On January 10, 2008, analysts at the Election Defense Alliance (EDA) reported that, based on the official results on the New Hampshire Secretary of state web site, there was a remarkable relationship between Obama and Clinton votes, comparing votes tabulated by op-scan to votes tabulated by hand in a head-to-head contest between the two candidates:

Clinton: statewide optical scan tally
91,717
52.95%

Obama: statewide optical scan tally
81,495
47.05%

Clinton: statewide hand-count tally
20,889
47.05%

Obama: hand count
23,509
52.95%

EDA subsequently learned that the list of New Hampshire hand-count voting districts used in our initial analysis on January 10, 2008 was outdated information. Shortly after that list was downloaded the New Hampshire Secretary of State published a corrected list showing 14 districts previously listed as hand-count, as having in fact been counted by Diebold optical scan.[1]

====================

[1] -- Voting method information was downloaded November 22, 2007 but was subsequently updated November 26, 2007.
The 14 locations that the state of New Hampshire had previously listed as hand-counted but that are actually optical scan, are:

Carroll County -- Moultonborough, Ossipee, Tamworth
Cheshire County -- Fitzwilliam
Grafton County -- Campton, Plymouth
Hillsborough County -- Hillsborough, New Boston
Merrimack County -- Newbury
Rockingham County -- East Kingston
Strafford County -- New Durham
Sullivan County -- Claremont Wards 1, 2 and 3

The net effect was to reduce the hand-count vote and increase the optical scan vote in each county.

[2] -- See Real Clear Politics (below) for a summary of 7 pre-election public tracking polls from 1/5 to 1/7/08, showing Obama in the lead at 38.3%, with Clinton trailing 8 points at 30.0%.

Head to head percentages were calculated as: Clinton = 30/(30+38.3) and Obama= 38.2/(30+38.3)

MORE links to New Hampshire Primary News and Data.

Overview of NH Polling and Vote Count Discrepancies
http://www.electiondefensealliance.org/overview_NH_discrepancies

Table of Machine vs. Hand Count Differentials (Democratic Primary)
http://www.electiondefensealliance.org/machine_vs_paper_count_differenti...

"I Count" Volunteer Signup to Hand Count Paper Ballots
http://www.electiondefensealliance.org/register_i_count_corps_hand_count...

NH Primary Voting Data Spreadsheets -- EDA
2008NHDemPrimComplete-EDA.xls (Democratic) and

2008NHRepPrimComplete-EDA.xls (Republican)

New Hampshire Binomial Statistics
http://www.electiondefensealliance.org/New_Hampshire_Binomial_Statistics

---

 

by Daniel Ashby (0 articles, 0 quicklinks, 0 diaries, 2 comments) on Friday, February 1, 2008 at 1:29:57 PM
 

What's So Remarkable?

Quote: "the undeniable fact that Obama appears to have carried the hand-counted tally statewide, while Clinton carried the optical scan statewide tally -- by almost exactly opposite margins -- remains a remarkable result."

It's not remarkable if the exit polls showed the same discrepancy between rural (hand-count) and urban (op-scan) areas, which I am informed is the case.

The first set of data was remarkable on its face, but alas it has proven to be false, which I suspected all along. The second set of data is far less remarkable and subject to explanations other than fraud, such as patterns in rural versus urban voting behavior. Since this is the more likely explanation, supporters of the fraud hypopthesis still have a long way to go to prove their case. Only a recount could get to the bottom of this, and I understand that isn't going to happen.

by Sam Adams (0 articles, 0 quicklinks, 0 diaries, 81 comments) on Friday, February 1, 2008 at 5:06:50 PM
 

Second-hand anecdotal evidence isn't a counterargument

You say, “The first set of data was remarkable on its face, but alas it has proven to be false.” There are two things wrong with this assertion. First, both sets of data are remarkable for the simple reason that they are slightly different variants of an outcome that I have bothered to do the work to prove is statistically impossible. Second, your information that the exit polls “showed the same discrepancy” isn’t out where I’ve been able to find it. So why don’t you dig up those exit polls and then do the work to find out what they do and don’t prove?

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Saturday, February 2, 2008 at 8:48:42 AM
 

Does Not Compute

No, not your findings, Mr. Griscom. I totally accept them. And congratulations on getting in the Mark Crispin Miller.

Now we just need someone to explain the who and why. Democrats are hacking now? Or Republicans want Hillary to run in order to rally their base against her?

Help me out here.

 

by Russ Wellen (58 articles, 1029 quicklinks, 66 diaries, 335 comments) on Friday, February 1, 2008 at 10:21:11 AM
 

Fascist Republicans want Hillary as their 'opponent'

.
Partly because they expect their criminal ballot-counting frauds can defeat her.

If that fails (by quick-arriving aroused voter oversight?, so she 'wins'), then Plan B is that already she is one of them (powerlusting fascists) anyway (and can be allowed to 'win'), or in Office can be sufficiently bent to their aims by threat, coercions and blackmail.

---
Speaking of 'quick-arriving' ... the rapid momentum of Obama's rise was unanticipated by fascist Republicans, (because their own racism is a blind spot), and so the fascists had to act against him (eclipsing Hillary) in NH, prematurely departing from the fascists' plans. That 'offsides' false start of the fascists' plans, is the irritating 'grain of sand' inception of the later (May'08) disintegration of the plans, and indicative symbol of it (plans failure).

Much irritation and destabilization of the fascist Republicans' plans, can be accomplished by forcing and publicizing the NH anomaly into wider examination, consideration, and recount (although valid recount is already corrupted, as Bev Harris documented, and so the 'second differential' is to take action to force and publicize the corrupted recount -- excoriate NH elections officials and that sleazeball criminal NH Secretary of State ... for instance, call your State's SoS and force that Office's officer to defend or repudiate NH's SoS in the collegium of the 'National Secretaries of State Assn.')

by meremark (1 articles, 3 quicklinks, 23 diaries, 483 comments) on Friday, February 1, 2008 at 1:39:46 PM
 

Good Question!

In order to hack an entire state’s election it would be necessary to have colluding election officials in charge of most of that state’s election computers.  This cannot be done overnight.  Surely it must take a decade or more.  In 2004, New Hampshire was the second most “red-shifted” state in the union (4.9%), right after neighboring Vermont (5.2%).  See table at the end of this document:

http://electionarchive.org/ucvAnalysis/US/Exit_Polls_2004_Mitofsky-Edison.pdf

So, alas, those two freedom-loving New England states seem to have been taken over by folks who stole the 2004 Election for Bush and then stymied Nader’s recount – just as they recently did Kucinich’s.  Bev Harris and others have documented the poor chain of custody of the paper ballots.  So the ability to forge recounts is likely a well-developed part of the election stealers’ system.

This would seem to say that they were Republicans in 2004 and that this crew certainly would not have willingly switched places with Clinton Democrats this time.  But an interesting story seems to be emerging nationally:  It appears accurate to say that the election stealers represent no political party, or even a wing of a political party.  They may be described as “fascists,” as Meremark does, or as representatives of the “military-industrial complex” that Ike warned us against.  Melinda Pillsbury-Foster, who insists that she herself is a right-wing Republican, is ferociously anti-election-theft and identifies the miscreants as “the neocons.”  But the “theocons,” or Christian right, that Mark Crispin Miller profiles so well in his book Fooled Again are definitely involved in the mix (see evidence from Tucson my chapter in MCM’s new book Loser Take All).

In this context, the brightest news comes from Arizona, where Democratic State Representative Ted Downing worked together with Republican State Senator Karen Johnson to draft and pass what may be the best election reform law in the country.  I’ve posted a little history here:

http://impactglassman.blogspot.com/2007/12/concise-history-of-election-integrity.html

Indeed, the most heartwarming cooperation between a Democrat and a Republican to defeat election fraud happened only four weeks ago at a public meeting of the Pima County, AZ, Board of Supervisors:

http://www.soundandfury.tv/pages/brakey.html

What went on in this film clip is so complex that I plan to write an entire OpEdNews column treating it alone.  But in summary, progressive Democrat John Brakey risks ejection from the room by loudly demanding the citizens’ right to be heard.  And the single member of the Board that agreed with this – and the citizen’s demand for the release of election computer data bases from past elections – was Republican Ray Carroll.  Neither Ray nor John knew what the other would say or do, but in their kind of “good cop, bad cop” interplay, they won an enormous victory over the election stealers.  Watch the tremendous grin that Ray flashes toward John at the end...

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Saturday, February 2, 2008 at 10:22:47 AM
 

what is there to disprove?

David, as far as I can tell, if your assumption of random assignment holds, then the optical scan and hand-count jurisdictions should be statistically indistinguishable in every respect, not only Obama/Clinton vote shares. Did you check that?

If your assumption fails, then your statistical result is irrelevant.

Rather than gnash your teeth at the "stolen-election denier community," perhaps you should compare notes with someone who has previously analyzed political and/or spatial data. I am given to understand that that is a common mode of scientific inquiry.

by Mark Lindeman (0 articles, 0 quicklinks, 0 diaries, 3 comments) on Friday, February 1, 2008 at 12:14:28 PM
 

One thing for sure, you won't disprove my calculation.

... “if your assumption of random assignment holds, then the optical scan and hand-count jurisdictions should be statistically indistinguishable in every respect, not only Obama/Clinton vote shares.”

That’s a non sequitur, Mark. As you well know, Binomial Statistics most famously pertains to coin flips. But while the coin always lands heads or tails, when it is flipped tomorrow at the Super Bowl, the outcome will be determined by any number of statistically distinguishable things. There is the rate of revolution of the coin, the height to which it is thrown, the nature of the sod or Astroturf it falls on, and even the wind velocity. If the coin should come down heads for the next 10 Super bowls in a row, it would not imply that any, much less all, of these factors were indistinguishable in all ten cases.

... “you should compare notes with someone who has previously analyzed political and/or spatial data.”

I agree with you that it would be instructive to study the demographics of the optical-scan and HCPB precincts to get an idea of whether or not there were factors strongly correlated with the way a person might vote in the Obama-Clinton contest. I would be happy to look at any such data systematically taken by qualified pollsters, although due to my preoccupation with many other things I have no “notes” of my own to compare them with.

In the meantime, my calculation convinces me that there is 0% probability that demographic effects strong enough to flip the Obama-Clinton vote 13% will be discovered to exist in perfect correlation with op-scan/HCPB boundaries.

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Saturday, February 2, 2008 at 11:29:23 AM
 

goodness gracious

Your calculation rebuts the assumption of no difference between op-scan and hand count jurisdictions. That's all it does. And therefore it does pretty much nothing. Perhaps you have spent too much of your career doing experimental research to understand just how hollow the result is.

Your example of flipping a single coin ten times has no bearing on a contrast between hand-count and op-scan jurisdictions. You would be somewhat closer if you compared one thousand coin tosses in each of one hundred domed stadiums to one thousand coin tosses in each of one hundred open stadiums. Only each "toss" is of a different coin, and the coins have individual characteristics (sometimes referred to as "personalities"), and social influences upon each other, and the ability to choose which stadium they are in, and whether they land heads or tails. And so on (of course the number of "coins" varies from stadium to stadium). A physicist may feel comfortable abstracting away from all of that, but that is no reason for anyone to follow him.

If I read this correctly, you are saying that you are preoccupied with other matters, but not too preoccupied to declare a 100% probability that the optical scanners were hacked. If it isn't obvious why I find that strange, I despair of explaining it.

You might want to look at Herron, Mebane and Wand's paper, Voting Technology and the 2008 New Hampshire Primary. Forgive my skepticism, but I'm just not convinced that you need to school these people in the use of binomials. But if your goal is to convince yourself, well, by all means carry on.

by Mark Lindeman (0 articles, 0 quicklinks, 0 diaries, 3 comments) on Saturday, February 2, 2008 at 12:27:10 PM
 

What is it about ZERO that you don't understand?

Your words, Mark: “Your calculation rebuts the assumption of no difference between op-scan and hand count jurisdictions. That's all it does. And therefore it does pretty much nothing.”

Wrong, wrong, wrong!  The power of the statistics that I worked out is that it gives the probability of the outcome assuming that all causal effects (like the effect of the wind on a flipped coin) are disinterested in the actual outcome.  So my calculation assumes that the boundaries between HCPB and op-scan precincts were, in effect, determined by a lot of coin flips (using un-weighted coins, of course) – and so were the Election-Day variations in demography across the state.

I found a probability of zero that the official results of New Hampshire’s 2008 Obama-Clinton contest could have resulted from a “perfect storm” of certain randomly decided (but well known) precinct boundaries accidentally exactly matching certain separately randomly decided (but so far unproven) demographic factors.  ZERO!

While zero is “nothing” (to use your deprecating tone), it is truly “something,” given that no one – not even yourself – disputes this result!

So let me put the implications of this result in different terms:  It is not a “conspiracy theory” (using these words in their contemporary deprecatory sense) for someone to look at what I’ve found here and say: Hey, this means that there is a 100% probability that the official outcome was contrived by conspirators in control of the election computers!

Corollary: In this case – consequent to my calculation – it’s the “I-can’t-believe-it’s-a-conspiracy theorists” who (rightfully only in this case) merit ridicule.

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Sunday, February 3, 2008 at 10:29:09 AM
 

Exit Polls Not Easy to Come By

Quote: "So why don’t you dig up those exit polls and then do the work to find out what they do and don’t prove?"

Those exit polls are proprietary and they (Edison/Mitofsky) don't hand them out to just anybody. But you can buy them here. I'm going by published reports that there is a distinct pattern of rural voters voting disproportionately for underdog candidates versus urban dwellers in the past several presidential elections as demonstrated by the exit polls.  

The only thing that a purely statistical analysis of vote tallies could prove is that the discrepancy between op-scan and hand-count ballots is not due to chance alone. But op-scan verus hand-count precincts are not randomly assigned. They are self-selected with the op-scans being in higher populated areas where there are more votes to count hence the desire for automation.

Therefore, the investigation of possible vote fraud by purley statistical means of vote tallies is impossible since it fails to take into account disparate voting behavior between urban and rural voters. This behavior is easily observable in the exit polls, but I refuse to buy them just to prove a point. I'll let you buy them instead since you are the one asserting vote fraud has taken place. Personally, I trust the published reports of the exit poll data.

In 2004, we had a case where the exit poll data conflicted with the vote tallies in multiple swing states. Dr. Steve Freeman (MIT) wrote a great book investigating possible vote fraud in those national elections. I believe he made a strong case. But that doesn't mean I buy every assertion of vote fraud whenever it pops up. Great care must be used lest one cries "Wolf!" too often.  

 

 

by Sam Adams (0 articles, 0 quicklinks, 0 diaries, 81 comments) on Saturday, February 2, 2008 at 7:11:44 PM
 

You make some thoughtful points, Sam, but see above...

O.K. Neither you nor I can afford to buy those exit polls ...and neither can any of the dirt-poor election integrity organizations that I know of.  It’s truly a shame that exit polls – already paid for by the news media – should be concealed from the public.  Obviously, most of us would be a lot more convinced of whether or not election theft is going on, if we could only get a hold of the exit poll results.  Then experts like Steve Freeman could do statistics on whatever discrepancies may exist between these polls and the official results.  (Published accounts simply don’t contain enough demographic information to do the statistics.) 

So I strongly encourage anyone reading this that might be wealthy enough to buy the poll to do so and then make it available to Steve – and to any and all EI outfits interested in analyzing it.  (Alternatively a tax-deductible donation earmarked for this purpose could be made to the Election Defense Alliance, which would both do their own analyses and disseminate the poll results to any and all requestors).

B.T.W.  Anyone interested in reading my review of: Was the 2004 Presidential Election Stolen? – Exit Polls, Election Fraud, and the Official Count by Steven F. Freeman and Joel Bleifuss can find it here:

http://www.drivehq.com/file/df.aspx/publish/dlgriscom/ElectionFraudResearch

You won’t find it on Amazon, for whom I originally wrote it, since they tacitly rejected it.  (Wonder why?)

by David Griscom (4 articles, 0 quicklinks, 1 diaries, 28 comments) on Sunday, February 3, 2008 at 11:50:29 AM