As many as 200 voters casting ballots at an in-person polling location in Oakland were disenfranchised because volunteers were poorly trained by the Alameda County Registrar of Voter’s office, the ACLU of Northern California, other groups, said on Friday.

In a letter to the Alameda County Board of Supervisors, the ACLU and other groups laid out a litany of allegations that also include a lack of training for poll workers, in addition, to slow responses by the county registrar’s office after potential pre-elections concerns were brought to its attention.

The confusion at the Mills College polling location in Oakland, however, appears to be the most alarming allegation.

The ALCU of Northern California, along with the Asian Americans Advancing Justice – Asian Law Caucus, and Oakland Rising, allege 100-200 eligible voters using electronic ballot marking devices typically used by the disabled or seniors, did not have their votes counted. The area around the Mills College polling location is also disproportionally Black.

The Mills College location was open for early voting on Oct. 31 through the Nov. 3 election day. The group alleges the registered voters successfully accessed the touch-screen device and completed their ballots. The device does not electronically send the voter’s ballot to a centralized system. Instead, a printed paper ballot is created and added to the batch of ballots to be later counted by the county registrar’s office.

“However, due to inadequate training and inaccurate information they repeatedly received from your office, Mills College poll workers incorrectly told voters who cast their ballots using [Ballot Marking Devices] (BMDs) between Saturday and about mid-day Tuesday that the printouts from the machine were ‘receipts’ that the voters should take with them, rather than official ballots that they should deposit in the ballot box,” the group wrote.

The group alleges the confusion over printed ballots created by touch-screen devices extended from poll workers to “captains” who oversee operations of the polling location.

Tim Dupuis, the Alameda County registrar, said his office identified 140 instances at the Mills College polling location in which voters mistakenly received their printed ballots as receipts. An effort was made to contact those voters and procure the ballots, which have a unique watermark, Dupuis said. At this time, 22 of those ballots have been retrieved, he added.

“All poll workers should have been aware that the touchscreen machines printed out a ballot for deposit onsite, not as a receipts,” Dupuis said. “The [Registrar of Voters] cannot account for how this occurred given the training and resources available to poll workers. The problem is not systemic. No information indicates this anomaly arose at any other vote center.”

In addition, the county registrar’s omission of ranked-choice tabulations on its official website for Oakland, Berkeley, and San Leandro city council races is also a concern, the group said.

The county registrar did not include a detailed breakdown of how each ranked-choice votes were eventually allocated. The system asks voters to select their top three choices for mayoral and city council races in the three cities. On the occasion the first-place vote-getter does not exceed 50 percent of the vote, second-place, and if needed, third-place votes are allocated until one candidate reaches a simple majority.

None of the ranked-choice voting races in Alameda County last week were not particularly close. Several produced winners without the additional allocation of votes. But the lack of transparency is troubling, the group wrote. “In effect, the public does not have complete and accurate information about how Alameda County voters voted in RCV jurisdictions.”

In past elections, the county registrar’s office has included links on its website that display a matrix of how ranked-choice votes were allocated to each candidate. This election, the results for ranked-choice voting races directed the public to contact the registrar’s office for the complete results.