Why is this audit tool crucial? Elections present a complex landscape, overseen by numerous individuals who are not immersed in electoral processes daily. Regrettably, election officials often face a tough decision between prioritizing speed or accuracy, as achieving both simultaneously proves challenging. Why? Because the traditional precinct-based voting method has evolved into a multifaceted system encompassing voting centers, mail-in ballots, and drop boxes, each meticulously tracked through sophisticated systems reminiscent of those used by Amazon in how they track packages. These systems ensure that each ballot envelope is linked securely to its rightful voter. Once a ballot's envelope signature is verified, the ballot is removed from its identifying envelope, preserving the anonymity of the voted ballot.
Despite the evolution away from precincts, they still retain significance in a virtual sense, serving as foundational units within the electoral framework. However, many election officials lack comprehensive understanding of these systems, often relying heavily on vendors for support.
Contrary to common belief, modern electoral systems leverage digital technologies instead of traditional optical scan methods, which leave no trace of their operation.
These digital systems do not simply tally paper ballots; instead, they analyze digital images of the ballots, offering notable advantages such as advanced hyperlinking, sorting capabilities, and the provision of "ballot images" as proof of their functioning.
At AUDIT USA, we've leveraged two key public records—the "cast vote record database" and "the ballot images"—to develop "A Better Elections System for canvassing the results, called “ABE". This initiative aims to demystify electoral processes by rendering them transparent, trackable, and publicly verifiable by using public records.
Why so much focus on canvassing and precincts?
The adage of "think globally, act locally" rings true here. Precincts serve as vital units in the electoral process, forming the cornerstone of election integrity and democratic practices. They encapsulate crucial demographic information and historical voting patterns, providing a common-sense approach to understanding voters’ behavior not individually but as a unit.
Where do we currently stand with ABE?
Thanks to Ellen Theisen and the advancements made in version 3.1, our system now boasts the capability to hyperlink Dade County, Wisconsin's extensive collection of ballot images and single-cast vote records. This impressive dataset encompasses approximately 303k ballot images and 250k single-cast vote records. Remarkably, all this data is seamlessly consolidated into a single spreadsheet made of 303k rows, achieved in under a minute, clocking in at just 55 seconds. This remarkable improvement stands in stark contrast to the previous process, which required a lengthy 15 minutes to complete before macro-optimization.
Since March 1st, our primary focus has been on scalability and the development of other essential features, all seamlessly integrated with hyperlinks to the corresponding ballot images.
Among these features included by races in separate excel files organized by:
1. Overvotes by races
2. Write-ins by race.
3. Convert the Cast Vote Record (CVR) into precincts, an Individual CVRs, a critical aspect of “A Better Election” tool for common sense auditing by precinct, 1st checking to make sure that if 600 people voted do you have 600 ballot images? A common practice in doing a precinct canvas report.
Additionally, we've introduced a crucial existing tool called Microsoft Pivot, which enables users to create queries in Excel, retrieving data from the database and exporting it to Excel. Once imported, PivotTables can be used to analyze and summarize the data effectively.
This integration allows you to leverage the strengths of both tools: Access for managing and querying large datasets, and Excel for data analysis and visualization using features like PivotTables.
We are going to soon release Somerset County, which is going to be the beta test to use because it's easier to teach people to climb a hill than a mountain. And it can be done with two gigabytes of data that we can send by “WeTransfer” with three separate macros that do different jobs.
If I was to send you Dane County, WI it would be 44 gigabytes. Somerset County with 6,966 voters and 24 precincts with all features being done at one time takes 21 seconds.
Looking ahead, we anticipate handling data for a county with approximately 700,000 voters. Our goal is to scale up to process close to 900,000 ballot images using Microsoft Excel. If bigger than a million it would have to be in Microsoft Access and then run in two parts in Microsoft Excel.
We aim to integrate ballot Images and Cast Vote Records from other digital vendors' systems.
Информация по комментариям в разработке