Two important factors:

1.) I think the Taliban will be smart enough not to attract more ire

2.) I think the US knows how to extract diplomats and staff without incident

I think the community may be underrating how a few scattered incidents could be labeled "civil war" by news outlets, even if it isn't remotely close to "civil war" at the scale of 1861.

Even if 1861-style war has a chance of <1%, I think there is a 10% chance of a "few scattered incidents" and a 30% chance these incidents may get labeled as "civil war".

@Sylvain Thanks. In the future I will try actually reading the question.

What if Russia installs a new gov and the new gov does Minsk? Is that possible?

My forecast: How will the Cuomo situation end?

60% - Cuomo resigns before his term is up

15% - Cuomo is impeached and removed

15% - Cuomo makes it to the end of his term + does not run again

7% - Cuomo survives + runs again, but loses

3% - Cuomo survives + runs + is re-elected

— edited by PeterWildeford

I made a Monte Carlo model that suggests a lower 25% of 2021-10-29, a mean of 2021-11-24, and an upper 75% of 2021-12-24.

To account for model uncertainty, I used two components to add some fat tails.

Davis lost his recall by 10 points, mainly from conservative-leaning districts. But now California is twenty points more blue leaning than it was during the Davis recall and the candidates running against Newsom are not Schwarzenegger. Newsom is very likely going to prevail.

— edited by PeterHurford

I think the community is underrating some ">=$9, <$15" compromise legislation.

ChickenWatch [now tracks](https://chickenwatch.org/progress-tracker/) 1811 cage-free pledges, which means ~58 pledges were added in the past two months. This also means that 3189 more pledges are needed to get to 5000. [Past data](https://docs.google.com/document/d/1knvmDBqwqDGLp-31u7l80i9mN963d2xUYl4ZM7m8sKA/edit) shows this is part of an exponential (or - more likely - sigmoidal) trend: 2005-2013 - +69 commitments 2014 - +55 commitments 2015 - +128 commitments 2016 - +265 commitments 2017 - +278 commitments 2018 - +443 commitments 2019 - +573 c...

I'm going to go out on a limb here and say that recency bias plus Trump being impeached twice is going to make historians look really unfavorably upon him.

@Uncle Jeff @johnnycaffeine @casens Maybe one easy solution would be that predictions from level 1 users are not included in the community median

Seems like there are very strong incentives to avoid this sort of thing.

@Jgalt More than two-thirds of Americans are bad at forecasting