Next Tuesday, ballots are due in the August Primary Election including a major choice for Seattle voters — picking one person from a field of 15 with hopes your candidate is one of the top two to go through to the November General Election.
There is probably a better way to winnow down the field. Steps to introduce ranked choice voting in King County are on hold but the election format advocates say more accurately captures voter preference while eliminating the need for expensive runoff elections might be a good solution for a summer vote like the mayoral choice Seattle is facing.
To give you a feel for how it could eventually work, please consider casting your ballot in the CHS 2021 Mayoral Primary Ranked Choice Survey. We’ve posted a vote with the 15 Seattle mayoral candidates for you to rank in the top 5 fashion used this summer in New York City’s primary. We’ll leave the poll open until July 31st.
The survey simulates the paper ballots being used by New York, Maine, and Alaska:
- Pick your first-choice candidate and completely fill in the oval next to their name under the 1st column.
- If you have a second-choice candidate, fill in the oval next to their name under the 2nd column.
- You can rank up to 5 candidates. You can still choose to vote for only one candidate if you prefer. Ranking other candidates does not harm your first choice.
It asks you to rank up to five of the 15 candidates. You can rank less than five or only one top candidate if you wish. You cannot rank multiple candidates #1 nor can you rank your favorite candidate more than once.
Here’s how the winners are selected — for the Seattle mayoral primary, our ballot will select two winners:
As the votes are tallied, if a candidate receives more than 50% of the first-choice votes, they win the election. If no candidate earns more than 50% of first-choice votes, then counting will continue in rounds. Each round, the candidate with the fewest votes is eliminated. If your highest rated candidate is eliminated, your vote will move to your next highest rated candidate.
The result? Hopefully, an election where your vote matters even if your top choice does not win.
Need more information about the candidates? You can find answers for Five Capitol Hill and Central District questions for the Seattle mayoral candidates here.
2021 Primary Mayoral Candidates
| Clinton Bliss | M. Lorena González (CHS coverage) | Lance Randall |
| Henry Clay Dennison | Bruce Harrell (CHS coverage) | Don L. Rivers |
| James Donaldson | Andrew Grant Houston (CHS coverage) | Casey Sixkiller (CHS coverage) |
| Colleen Echohawk (CHS coverage) | Arthur K. Langlie (CHS coverage) | Omari Tahir-Garrett |
| Jessyn Farrell (CHS coverage) | Stan Lippmann | Bobby Tucker |
PLEASE HELP KEEP CHS PAYWALL-FREE!
Subscribe to CHS to help us pay writers and photographers to cover the neighborhood. CHS is a pay what you can community news site with no required sign-in or paywall. Become a subscriber to help us cover the neighborhood for as little as $5 a month.
I probably am misunderstanding how this site works but how is it possible for AGH to lose votes between round 1 and round 2 in this simulation? That seems broken unless the vote counts are not actually vote counts and more like an opaque percentage.
So I looked into this. RCV is really complicated when selecting 2 winners.
Basically it’s this:
The second is a vote for a candidate who wins “too much.” Any excess votes beyond the number of votes needed to win doesn’t change the outcome. That’s why those excess votes are also redistributed to their next ranked choice in the following round. You’ll see this in the visualization above as a winner’s green bar reverting back to the dotted “votes needed to win” line.
I’m not sure how it determines which ballots are the excess ones though.
No, only votes from the biggest loser should be redistributed. There’s a bug if any candidate other than the last place has their votes reduced in a round.
No need to speculate. You can watch the round by round distribution on the results page
That is not true.
It’s hard to describe since the results are ever changing. But when I was watching the first time, AGH had 10 more votes needed to move on, so they moved 10 of his votes to Echohawk and Gonzales. It’s very unclear how they chose those 10 votes.
“The second is a vote for a candidate who wins “too much.” Any excess votes beyond the number of votes needed to win doesn’t change the outcome. That’s why those excess votes are also redistributed to their next ranked choice in the following round. You’ll see this in the visualization above as a winner’s green bar reverting back to the dotted “votes needed to win” line.”
Seems unfair if it happens to choose excess votes that go to Gonzales instead of Harrell or something like that.
RCV seems like a poor choice of voting for selecting 2 outcomes, rather approval voting would be much more clear.
Please read my reply. It’s taken directly from the site.
Super interesting. Thanks for sharing, and I hope you publish the final results on here!
It let me vote twice, so I have to assume someone for at least one of the candidates will script a mass-vote run at some point and game the results. In other words, this is going to be exploited. Unless it’s just not counting when you repeat-vote, even though it says it did count.
RankedVote can be set to attempt to remove “suspected duplicate votes” — that setting is activated for this survey
Seems like so-called exploitation of a *demonstration* that attempts to help people understand how RCV works may not be exploitation at all.
I find it extremely suspicious that Andrew Grant Houston is dominating this poll. Something tells me he’s not actually the overwhelming choice of Capitol Hill voters. He has 200 votes so far and I would be surprised if he even manages that in the actual city-wide election. Reminiscent of Ron Paul 2012.
I really wish people would be more specific than “ranked choice voting”. There are plenty of “ranked choice” voting methods, most (if not all) with distinct names. The RCV method being used here seems to be Single Transferable Vote (STV). The vast majority of people who say “Ranked Choice voting” are referring Instant Runoff voting, but rebranded… Other ranked voting methods include Condorcet methods, which are their own family.
Anyway, the voting method used here appears to be STV with a Droop quota (the win threshold is: votes_cast/(seats+1) + 1 ). Explainer video: https://www.youtube.com/watch?v=l8XOZJkozfI