Logic and Accuracy Testing

Logic and Accuracy Testing Procedures for the Butler County Board of Elections
Per Directive 2016-22 (Section 1.08)

Under R. C. 3506.14 (B), before every election, each board of elections must publicly test automatic tabulating equipment used in the county to ascertain that it will accurately count the votes cast for all offices and on all questions and issues. 

Refer to the Calendar of Events for the schedule for public testing.  The following preparations take place before logic and accuracy testing for each election:

  1. Several sets of ballots (test decks) have been assembled.  These test desks contain races of contests with multiple candidates, candidate rotations, splits within a precinct, multiple questions or issues within a precinct, and write-in candidates. 
  2. Public Notice has been given as to the date, time and place of the test.
  3. Board staff familiar with the various components of the county’s vote tabulation system will conduct the test, and will be available for questioning before and after the testing of the system.
  4. The Board staff will conduct successful public tests on the vote tabulation system on two occasions:
    • Before tabulating ballots for the unofficial count; and
    • Before tabulating ballots for the official count.

Testing of the High-Speed Central County Optical Scanner

  1. The “Zero Count” is verified that there are no votes currently recorded.
  2. The ballots in the test deck are marked so as to record a predetermined number of valid votes for each candidate and on each question and issue, and include for each office one or more ballots marked with too many votes (over-vote) to test the ability of the equipment to detect and reject those votes. 
  3. Before tabulating the test deck, the votes on the test deck are audited by hand count.  The results from the hand count of the test desk will be recorded and announced before tabulating the test deck on the high-speed optical scanner.
  4. Once the test deck has been voted, the test votes are removed from the scanner and uploaded to the central tabulation server in order to produce a “test count” report.
  5. The “test count” report of votes that were uploaded to the central tabulation server will be printed from the central tabulation server.  The “test count” results from the central tabulation server as they were uploaded from the high-speed optical scanner will then be announced.
  6. The results of the pre-audited group of test deck ballots and the “test count” results from the central tabulation server must then be compared and must match.
  7. If no errors/discrepancies are reported, “closing zero count” reports are generated.

Testing of Direct Recording Electronic (DRE) Voting Machines

  1. An “opening zero count” report will be generated from the central tabulation server by the tester to verify there are no votes currently recorded on the system.An “opening zero count” report will be generated from each of a representative sample of DRE voting machines by the tester to verify there are no votes currently recorded on the individual voting machines.
  2. A set of ballot voting instructions specific to each precinct or precincts for which a DRE voting machine has been programmed, with separate instructions to ensure that every possible ballot position is voted, must be created for each voting machine tested.  Each set of ballot voting instructions must include a predetermined number of votes for each candidate and on each question and issue so that a different number of votes must be assigned to each candidate for an office, including write-in candidates, and for and against each question and issue so that the DRE voting machine tallies votes on each race and issue.
  3. Before voting on the DRE voting machines, a hand count is taken of the votes to be tabulated on the voting machines and announced before tabulating on the DRE machines.
  4. The pre-audited sets of ballot instructions will be voted, summary statements generated by each DRE voting machines, and the results will be announced for each unit.  The results from the summary statements will then be compared to the hand counts, and each must match that count.
  5. The memory cards must be removed from the DRE voting machines and uploaded into the central tabulation system in order to produce a “test count” report for each memory card, and total “test count” report for the server, accumulating the totals for each DRE voting machine.
  6. Those votes must equal the accumulated totals and verified by hand count as can be cumulated from the summary reports that were generated.
  7. A “test count” report of votes uploaded to the central tabulation system shall be printed and the results will then be announced.
  8. The results must match, and if any error/discrepancy is detected, the cause must be determined, corrected, and another test conducted.
  9. If no errors are detected, a “closing zero count” report will be generated from each DRE machine.

After the Board of Elections completes the public testing of the vote tabulation system, all components, including the server, will be secured until Election Day.  The DRE voting machines will be secured before and during use for in person voting and/or for use on Election Day, until transport to the Board of Elections after the election.