When is a perfect oval evidence of vote fraud?

Having sworn off professional sporting events like many of you, I have made what feels like the moribund turn to other pursuits, such as watching a Dominion Voting System Presentation from 2017.  I did this so you don’t have to; unless your Lawrence Welk Betamax of Cissy and Bobby’s greatest hits is fuzzing up, then by all means, take a look. 

In the video we have a team of sales and demo experts (Dr. Eric Coomer, Waldeep Singh, David Marino, CEO John Poulos and other support staff) “…to outline new voting equipment options in Chicago on April 13, 2017.”  This was the 4th of four election equipment videos and had a few items of note that I haven’t heard or read anywhere before and I wanted to share those with you. (video is almost an hour long)


Dr. Eric Coomer YouTube screengrab (cropped)

First, let’s talk about the elephant in the room, Dr. Eric Coomer.  Recognize that name?  In this video, Coomer comes across as a consummate professional, knowing hundreds of thousands of dollars are on the line.  In my professional experience, it was a very good demo highlighting the capabilities and features of the hardware and software that Dominion Voting Systems sells.  As he walks through his measured and informational demo, you would never know the alleged radical and hateful conservative contempt kicking around in the back of his head.  So rather than dismiss Mr. Coomer outright, let me present some information that has been hiding in plain sight in what is otherwise an unremarkable video.  Given the propensity for technocrats to scuttle damaging info, I am surprised Google hasn’t memory holed this item, nor has Dominion has requested a takedown.

The presentation takes an interesting turn when Dr. Coomer presents the Imagecast Evolution Ballot Marking Device (BMD) at the 15:00 mark.  It functions as an all- in-one (votes, tabulates, and prints the ballot) and Americans With Disabilities Act (ADA) solution as well. ADA voters get a blank ballot to insert and the voter can use any ADA compliant input device attached to the unit or use the device itself.  An internal printer marks the ballot for the voter and directly deposits into the ballot collection box.  At the 18:30 mark, David Marino states the following (emphasis added)

Something important, uh, on these marks that are created on the ballot is we have a huge library of handmade marks so it’s not a perfect oval that you are going to be able to identify that that was a mark by a machine. But it’s, it’s ah, it’s a library of different random hand marks that looks like somebody else used a Sharpie to vote the ballot. So you are never going to be able to say this is ah, a ballot voted by the accessible uh voter, this is a ballot voted by a person with a Sharpie for example, for, with the mark.

Dr. Coomer follows immediately with:

Yeah, again it’s all about preserving voter anonymity, um you know if, if you only have one or two disabled voters in a given precinct an if you’re using standard marking techniques where they’re, uh uh an exact perfect fill of that oval um you would be able to uh distinguish that ballot from somebody that just hand marked it. So this is one of those further steps that we do um to preserve that anonymity.

A library of different random marks to avoid a ballot appearing differently from hand marked ballots becomes more critical considering the next segment of his presentation. In other words, their system prints electronically cast ballots that cannot be differentiated from a hand marked ballot.

Let’s move on to the next item of note, the image audit mark. Dr. Coomer continues (emphasis added):

The other key thing here is what I haven’t talked about, so obviously we take an image of every ballot cast, front and back, but we also have what’s called the audit mark. That’s part of the image and it is a text record of how the scanner actually counted that ballot when it was scanned. So it shows here (points to the number on the image), and there’s, there’s a variety of information that’s included on that. It shows which tabulator it was scanned on, which batch it was created in, and for the ICC (ImageCast Central), because it’s central, it actually gives you the number of the ballot within, within the stack. We do the same uh audit mark creation on the precinct uh devices, but we don’t, we don’t index the paper, again for voter anonymity.

Dr. Coomer shows an audit mark contiguous to the image of a scanned ballot that indicates a blank vote because the voter circled the candidate names.  He makes digital checkmarks of voter intent in the correct spot and adjudicates the ballot on screen, thus updating the audit mark but the original audit mark is not erased.  So an adjudicator, on-screen, could divine whatever intent, update the ballot which adds a new audit mark, then move on to the next ballot.  All of this happens with a screen and mouse editing an image record but alters the image of the ballot and the audit trail.  A digital version of divining the hanging chad, if you will.  The remainder is somewhat interesting as he continues with a summary of audit features for recounts, candidate views, and risk-limiting statistical sampling.  However, the products he presents are all tied to one election system and did describe their overseas ballot portal for military voters as well.

As I mentioned, let’s not rush to dismiss Dr. Coomer’s remarks outright, let’s learn from the publically available information he has shared and compare that against BMD auditability.  As recently as September of this year, a fascinating article was published in the Election Law Journal Volume 19, Number 3 (note: paywall) regarding BMD’s such as the one mentioned above in Dominion’s presentation. The abstract states:

Voters can make mistakes in expressing their intent in either technology, but only BMDs are also subject to hacking, bugs, and misconfiguration of the software that prints the marked ballots. Most voters do not review BMD-printed ballots, and those who do often fail to notice when the printed vote is not what they expressed on the touchscreen. Furthermore, there is no action a voter can take to demonstrate to election officials that a BMD altered their expressed votes, nor is there a corrective action that election officials can take if notified by voters—there is no way to deter, contain, or correct computer hacking in BMDs. These are the essential security flaws of BMDs.

Risk-limiting audits can ensure that the votes recorded on paper ballots are tabulated correctly, but no audit can ensure that the votes on paper are the ones expressed by the voter on a touchscreen: Elections conducted on current BMDs cannot be confirmed by audits. We identify two properties of voting systems, contestability and defensibility, necessary for audits to confirm election outcomes. No available BMD certified by the Election Assistance Commission is contestable or defensible.

There is a double-edged sword to technology that mimics hand marking for the sake of anonymity.  I am terribly curious to hear if any ballots in the statistically impossible contested ballot dumps have perfect standard oval marks where Dominion’s library of hand marks was used? Could the thieves have outsmarted themselves? Were the adjudicators relying solely on Dominion’s adjudication ballot recognition software that fails to detect anomalies while rushing through ballot images too fast to tell anything was amiss? Can Dominion’s software correctly distinguish fraudulent ballots that are too perfect? All in the name of voter anonymity. Are we certain the fraud is not anonymous as well? Time will tell.