[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: GEMS 1.11.2 preference voting errors

Ballot rejection
  • Ballots are returned with a blank ballot return message if the preference race is not marked.  Isn't the preference race optional?
Do you mean blank ballot or blank race?  How are the other races voted?  If the ballot really is blank, then is should be rejected as blank.  I'll try to reproduce this.
  • A candidate and write-in selected in a vote for one race are returned as overvoted, even though the override write-in rule should effect the selection of the write-in only.

Is the write-in override rule broken or does this have to do with preference races?  If the former, submit another report with a minimal example (one race) and I will look into it.  If the latter, let me know.
  • If the first partisan race following the preference race is overvoted the race is returned as overvoted where a political party different from that of the party endorsing the partisan race is selected in the preference race.
This is very difficult to parse.  I think you mean:
Partisan races are rejected as overvoted even if they should not be counted because of party preference.
Right?  If so, this sounds like an AccuVote bug.  Guy, please take a look.
  • A ballot is rejected with the preference race and first partisan candidacy as blank and multi-party votes in all three partisan candidacies where the preference race has been left blank and candidates selected in all three partisan races.  Since all three races should be discarded as blank (due to multi-party overvote, there should either be a blank race return message for all three candidacies or none at all (since the three partisan candidacies are already being returned due to multi-party votes).

Again I can barely parse this.  I think this is the same case as your first point.  Should any votes be counted for the ballot in question or not?  If not, it should be rejected as blank.  If I didn't understand you (this would not suprise me), try explaining it again without the run on sentences.
  • Candidacies marked with a candidate as well as a write-in are being tallied as overvoted even though they should be tallied as write-in using the override write-in rule.

You already said this, right?
  • The Republican party is selected in the preference race, one candidate in the Democrat candidacy and no candidates in either the Republican or Unaffiliated candidacies - the Democrat candidate marked on the ballot is tallied even though the Democrat candidacy should be tallied as blank voted.
I will try to reproduce this.  Do you have a test case?
  • The Democrat party is selected in the preference race, two candidates in each of the vote for one Democrat and Republican candidacies and a candidate and write-in in the Unaffiliated candidacy.  The Democrat candidacy should be logged with an overvote and the Republican and Unaffiliated candidacies with blank votes, yet all three partisan candidacies are returned as blank voted.
I cannot parse this, but I'll try setting up a partisan race voted democrat and a partisan democratic race overvoted and see what happens.  Are you saying the democratic race will be rejected as blank instead of overvoted?
I have to ask.  Is a "candidacy" a race or a candidate?
I understand the preference race to conform to the following logic in a closed primary (all races on a single ballot):  If no selection is made in the preference race and only selections are made in partisan races endorsed by one party, all partisan race choices are considered.  If no selection is made in the preference race and selections are made in races belonging to more than one political party, all partisan race choices are discarded.  If a valid selection is made in the preference race only the results of races endorsed by the political party selected in the preference race are taken.
I am not sure.  For what it is worth, GEMS central count only counts partisan races controlled by a preference race if they match the preference candidate, but now that I think about it that sounds wrong.  It makes more sense if it behaved like a closed primary without a preference race.  That gets really hairy in the face of multiple preference races though.  Guy, how does the AV behave in this respect?
Whoever in ESG that is responsible for accounts that have preference races should follow up with the statutes if they want to make sure this is correct.  I'll make sure CC behaves however you want it.
I understand the override write-in tally rule to cause only write-in selections to be taken in a race in which the combination of candidate and write-in selections cause an overvote.
I look forward to your bug report on the write-in override rules.