Scanners like ours: Optical scanner counts differ for same ballots

Brad Friedman reported the story last week <read>

A close race on election night. Rescanned to check but the other candidate won. Then they did a hand count and confirmed the original result. UT like Connecticut is fortunate to have chosen optical scanners with voter completed paper ballots. But we need to verify the accuracy of scanners with audits, recanvasses, and recounts.

The first “recount” of Provo’s Municipal Council District 1 ballots — carried out on the same op-scan systems that tallied them in the first place — was held yesterday, only to be abruptly called off when the results were found to be “extremely in favor of the opposite candidate.”…

“The numbers were varying too much,” Utah County Chief Deputy Clerk/Auditor Scott Hogensen tells the Deseret News about the District 1 race. “It became obvious the machines weren’t counting things correctly.”

But whether the Diebold op-scanners tallied the ballots inaccurately on Election Day or during the so-called “recount” remains unknown at the moment.

According to Deseret News, “Morrow said she asked for the recount to be done by hand in the first place but the request was denied.”…

two hand-counts on Wednesday have now confirmed the accuracy of the original optical-scan count giving the election victory to Gary Winterton after all. The “recount” on the same op-scan systems seem to have been inaccurate, while the original count was accurate. We still don’t know why, of course.

It was not a small, trivial difference, we are talking over 700 votes!

No word yet on why the second scanner might have miscounted. There should be an investigation, however, we suggest that determining the cause is not a complete cure. No matter the cause:

  • It could happen again in Utah or Connecticut
  • Another time it might be the original scanner, not the second one, and/or election day officials making the error
  • It might be far enough off that there is no automatic recount or recanvass
  • It might not be the machine, it might be procedures, yet exonerating the machine does not provide comfort, whatever the cause it can happen again in Utah or Connecticut
  • Perhaps it has happened before – maybe last year in Connecticut one or more of the differences between hand counts and machine counts might not have been human errors as assumed by the Secretary of the State’s office. <read>

We leave with this further item from Brad illustrating the tendency for officials to leap to unfounded, yet assuring conclusions based on assumptions:

Amusingly, and for reasons unknown, [Utah County Chief Deputy Clerk/Auditor Scott] Hogensen told Deseret News that, according to the paper, he “does not believe machine malfunctions affect the outcome of any other races in the county.”

This has happened a couple of times before with other scanners. <one example><another>

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.