Results
Thanks for everyone who nominated and voted!!
Turnout was 1,444,708 ENS across 60 voters, reaching the 1M (1% of supply) quorum. Votes were counted with Copeland ranked-choice and Shutter shielded voting.
If you open the Snapshot result, the third seat appears to go to SpikeWatanabe.eth. That is not quite right, and the reason is a part of the counting method that Snapshot never implemented.
What the UI is showing
Copeland ranks candidates by how many head-to-head matchups they win. Abdullah and Spike each won 3 of their 6 matchups, so they finish with an identical Copeland score (Snapshot shows both at 206,386.88). They are tied on victories.
The tiebreaker for exactly this case was defined when we commissioned this voting type. From the RFP: improving Ranked Choice Voting support in Snapshot:
The Average Support for each candidate will also be calculated (…) Average Support is used as a tiebreaker.
Average Support is the voting power a candidate gathered across all of their matchups. On that measure Abdullah leads Spike, 57.5% to 50.0%, so the third seat is Abdullah’s.
Snapshot’s deployed Copeland code does not compute tiebreaks. When candidates tie on victories it leaves their scores identical, and the UI then orders them by the position they were listed in on the ballot. Spike was listed above Abdullah, so the UI shows Spike third. That ordering comes from ballot order, not from the votes.
The fix
I opened a PR on Snapshot’s own repository to add the Average Support tiebreaker to the Copeland implementation: snapshot.js#1210. It resolves ties by Average Support exactly as the RFP specified, and by design, it cannot change any result that is not a genuine tie. I’m in contact with their team to get this merged.
The result
With the tiebreaker applied, the three elected Meta-Governance stewards for Term 7 are:
- Alex Netto / netto.eth
- Sov / sovereignsignal.eth
- Abdullah / abdullahumar.eth
Here is a copeland visualizer for those curious to understand more the votes and result.
Feedback about the vote
This was a pilot using shielded voting. Feedback was positive on the last Metagov call. Please share your thoughts on this thread, about how should we continue to use this primitive.