Open cwulfman opened 2 years ago
Created a Versasim base for mocking up Hadron Party Primary featuring ranked-choice voting.
I took a look at the wiki page, and everything I see there makes sense to me.
For EDFs, the new VoteVariation
is the perfect way to signal to the client that the contest will need to be presented as RCV (in whatever way is appropriate). Continuing to use the existing VotesAllowed
and ContestSelection
means that no changes will be required to the parsing logic there. Very straightforward!
For CVRs, everything in section 5.8.1 of the NIST-1500-103 specification also looks sensible. The client would generate a one ContestSelection
per voter choice, which will be identical to what we have in plurality contests except that we'll now need to add the Rank
property with whatever the voter chose. Again, very straightforward.
One question about the NIST-1500-103 doc - I found two (here on GitHub and here on NIST's website). Which is correct? Can you link the correct one in your wiki?
I'll post a PR when I have a draft of an actual EDF, so others can review it. As for the NIST-1500-103 specification document: unless we're told otherwise, we should use the published version: https://doi.org/10.6028/NIST.SP.1500-103.
@mattbeattie it turns out the difference between the latest GitHub version and the officially published version is on the last page of each. The published version was created on March 31, 2020 (git: 8b3373ee). The one on the current GitHub version is 3 commits later on Apr. 21, 2020 (git: 6ad813ed), and has these changes:
- Small corrections to the documentation were made.
- For class CVR, definitions for BallotPrePrintedId and BallotSheetId were clarified.
- The basic example in section 5.2 was corrected.
So if you had relied on the GitHub version nothing major changed, though of course the published version is still the one to use.
Added a wiki page about RCV: https://github.com/TrustTheVote-Project/NIST-1500-100-103-examples/wiki/ranked_choice