🏅 Top traders

@traders To clarify, I resolved after Harrell conceded and relevant news outlets called the race for Wilson.

Seattle will NEVER elect a mayor for a second term

Assuming normal distribution for simplicity

Mean:

• Scenario A: like last two batches = average Friday and Monday

• Scenario B: trend upwards continue Tuesday

Standard deviation:

• Not so many data points in last elections and among daily batches in Nov 2025

• I’m not so confident my scenarios are the real standard deviation we will see, but at least they should be “not crazy”

If anyone has any different priors for distribution shape, mean or standard deviation, I would love to read your thoughts.

Outcome:

• Wilson will be ahead tonight

• Wilson will not avoid recounting

• If you assume the last batch will be similar to Friday+Monday, then >95% chances of manual recount all votes, 5% automated recount

• If you assume the last batch will continue the upwards trend, then >50% chances of manual recounting all votes, 50% automated recount

so, finally we got 60-40% for the batch

this was quite on the right side of the probability curve, or my toy model was wrong

Wilson YES holders: I can sell you my NO shares

• at 80% before next batch

• 70% in a recount (<0,5% delta)

• 60% in a manual recount (<0,25% delta)
  

@nettika @HillaryClinton @stevjobs

Wilson is up by 91 votes, so I guess 91% is the rigth magic number.

I will vote her down to 91%

I'm not sure how many votes are left to count but it's crazy neck-and-neck still according to today's data from KingCounty.gov

@Stralor it looks like a recount to me. I'm not sure she will get >64,75% in the last batch

@Stralor

I'm not sure how many votes are left to count

You can check the last estimate here: https://results.vote.wa.gov/results/20251104/turnout.html

As of now (Nov 11th), it is 13.000 for King. The press estimates half of them belong to Seattle. This is consistent with previous days' ratio of Seattle vs. King.

A recount is triggered "if the difference between the candidates is less than one half of one percent and also less than 2,000 votes," and a manual recount is done if "the difference between the candidates or responses is less than one quarter of one percent and also less than 1,000 votes.

Based on your experience in previous elections, if we go for a manual recount of all votes, how long would that process take?

It will be an interesting period to negotiate the fairest price for the Limit Orders on Manifold if any of the big holders need some liquidity or wants to reduce their exposure.

@MiguelLM

if we go for a manual recount of all votes, how long would that process take?

highly variable. I cant remember a local election I was paying attention to where that happened. months at the worst, maybe a couple weeks at best for a manual recount of city-sized votes?

@Stralor if we go for a process that takes months, let us see how long they want to keep their mana locked here

@Stralor the Recounts page in King County Elections site is quite clear about the process, but not so much about the expected timeline. I guess it will be quick, as we will go for manual recount.

It is also useful to see the former results. Almost all the recounts changed the end results by 0 votes.
https://kingcounty.gov/en/dept/elections/recounts

It was quite fun
Which Seattle mayoral race is more fun? (see description)

Do you like low liquidity roller-coaster markets like this one?

Your views? It seems early to call it.
Do DDHQ has some info we don't have?

https://election-night.decisiondeskhq.com/date/2025-11-04/states/Washington/race/385189

Then we have this:

KUOW has not called the Seattle mayoral race at this point. King County Elections says there are still approximately 100,000 Seattle ballots left to count.

https://www.kuow.org/stories/live-results-nov-4-seattle-area-general-election

@MiguelLM DDHQ is mistaken, Wilson has a clear path if late ballots lean towards her as much as they have towards other progressives in past elections.

@HenryRodgers This was my first thought. Maybe the path is not completely clear, but realistic enough to be feasible.

But then I tried to inform myself about DDHQ, and apparently their track record is quite good.

1. Why did the call it?

2. If they are wrong, why didn’t they fixed the chart yet? More than 18 hours already passed after KUOW asked them to fix it

@MiguelLM The local politics hobbyists I know here in Seattle all think it's still a squeaker.

@CraigDemel it seems so. It is a funny one:
- data is posted once per day
- significant difference of early votes vs. late votes

I'm not sure how common this is in the US.

@HenryRodgers

Wilson has a clear path if late ballots lean towards her as much as they have towards other progressives in past elections.

If you have a look to previous elections (grey), she will make it.
If you focus on the current progress this week (green), she will not make it.

this chart will be outdated and useless in less than one hour with the new batch of votes, but as I took the time to process the csv files I decided to share it anyhow for your reference

This is the first time I am following a mayoral race in the US. How do you guys manage to track what is going on?

After trying it by myself for a while, I asked ChatGPT 5 Pro for support

Can you get a list of Seattle precincts sorted by number of registered voters? with a column accumulated % of the total, so that I can do a Pareto

2 minutes 9 seconds thinking

It was amazing how many sites ChatGPT had to try for such a simple request, explaining every couple of seconds what it was doing and why it was failing. I finally got the .csv for August primaries, at the 3rd or 4th prompted question.

Give me a table with
5 columns: Precinct name or District name, Registered voters, Ballots returned, Votes counted Wilson, Votes counted Harrell, Primary votes Wilson, Primary votes Harrell

20 rows for the top 20 precincts by number of registered voters

On the first trial ChatGPT directly rejected to even try such a task.
After splitting into several different prompts I kind of managed to get to a long list of steps suggested by ChatGPT.

I would say with good prompting, the best ChatGPT can do is to reduce a ~20 hours process to a 2 hours process, for a rookie like me who knows nothing about US election results sites.

I gave up. I will just invent some random odds, as everyone else is doing.

Is Seattle mayoral race results tracking the new benchmark for Artificial Super Intelligence?

Anyone in the room with AI and elections tracking experience who knows what AI agents are the best for this purpose?

The funniest part where the official sites with the warnings “due to lack of funding this site data is not up to date”

King County elections spokesperson Kendall LeVan Hodson said they expect “at least” another 180,00 ballots countywide, roughly half of which will be in Seattle. (The Seattle Times)

You have to figure out:
1) how many ballots you expect are left
2) with this number, how many more would Wilson need?

For reference, with 90K pending ballots, Wilson would need to have >57% of them to win.

in scenario B of this second model the participation is assumed to follow the same pattern as in 2021: elections have +28% total ballots compared with primaries

It is almost the same estimate as the previous one, but this time after checking the actual data