Voting System Technical Oversight Program (VSTOP)
Ball State University

Voter Verifiable Paper Audit Trail (VVPAP)

Provided by Indiana Secretary  of State – Mar 17, 2023, via Margaret Menge, Crossroads Report

• • Nov 18, 2021. VSTOP P-E RLA White Paper 11-18-21
  • Oct 1, 2021. 2023 VSTOP P-E RLA 2020 General Election audits report 10-1-2120
  • Sep 23, 2020. VSTOP P-E RLA 2020 Primary Pilot Audits report 9-23-20 

Received – Brown County Recorder: Jan 19, 2023.

• • April 6, 2020.  BC Recorder Final Risk Limiting Audit Program Proposal 6.5.2020
  • Sep 23, 2020. BC Recorder Indiana RLA Pilots Primary 2020 Final Report: 
    Brown County Results starting on Page 11

(1) On August 7, 2020, the VSTOP team successfully completed an RLA pilot in Brown County, Indiana. This activity was carried out by VSTOP in collaboration with the Election Assistance Commission, the Brown County Clerk’s Office, and Brown County Commissioner Diana Biddle.

Also participating were two members of the VSTOP CEATS cohort: Karen Wheeler and Beth Sheller.

The VSTOP Team members who participated are: Jay Bagga and Bryan Byers, VSTOP Co-Directors, Molly Owens, Project Specialist, Mani Kilaru, IT Specialist, Chad Kinsella, VSTOP Faculty Fellow, and Sajal Sheel, Graduate Assistant.

The following information was gathered as part of the pre-work and planning phases of the Brown County RLA pilot. This information was considered and entered in the Stark Tool, which assisted the VSTOP team and county election officials in deciding which race should be audited during the RLA pilot. Brown County uses the Unisyn 2.0A Ballot Card Voting System. According to the VSTOP inventory, there are 15 OVO (OpScan) units and 15 FVT (Ballot Marking Device) units. There are 12 precincts in Brown County and the county stores ballots by precincts. These ballots were not separated by Party.

The RLA Team audited the following races:

• Republican races (R): President of the U.S. and County Treasurer, and Jackson 2 Precinct Committeeman

• Democrat races (D): President of the U.S. and US Representative in Congress District 9

Each of the above contests was audited with a risk limit of 10%.

There were a total of 3,143 ballots of which 1,258 were Democrat ballots and 1,885 were Republican ballots.

President of the U.S. (R): The RLA tool instructed the RLA Team to audit a sample of 6 ballots for ballot polling with a diluted margin of 85.52%. The RLA team oversampled the ballots for ballot polling and reviewed 104 ballots. The Stark method functioned as expected and confirmed the “Donald J. Trump” outcome in the Brown County precincts with high levels of statistical assurance (100% for the Ballot Polling RLA).

County Treasurer (R): The RLA tool instructed the RLA Team to audit a sample of 119 ballots for ballot polling with a diluted margin of 18.30%. After reviewing the initially sampled ballots, the risk limit was not met. RLA team sampled a total of 270 ballots to attain the risk limit. The Stark method functioned as expected and confirmed the “Andy Vasquez Bond” outcome in the Brown County precincts with high levels of statistical assurance (96% for the Ballot Polling RLA).

Jackson 2 Precinct Committeeman (R): All 110 ballots were sampled for the Jackson 2 Precinct Committeeman, a contest that was close. Results were verified but the winner received two additional votes in our count. After closely examining the ballots, the RLA team confirmed that the additional ballots might have been counted by the voting machine as undervotes or overvotes.

President of the U.S. (D): The RLA tool instructed the RLA Team to audit a sample of 12 ballots for ballot polling with a diluted margin of 61.21%. The RLA team oversampled the ballots for ballot polling and reviewed 73 ballots. The Stark method functioned as expected and confirmed the “Joseph R. Biden” outcome in the Brown County precincts with high levels of statistical assurance (100% for both the Ballot Polling).

US Representative in Congress District 9 (D): The RLA tool instructed the RLA Team to audit a sample of 43 ballots for ballot polling with a diluted margin of 24.74%. The RLA team oversampled the ballots for ballot polling and reviewed 73 ballots. The Stark method functioned as expected and confirmed the “Andy Ruff” outcome in the Brown County precincts with high levels of statistical assurance (100% for both the Ballot Polling).