my apologies for the mistaken resolution notification. no points have been awarded, predicting will continue.

@predictors I again apologize for the mistaken resolution email sent today. The fault is my own; and essentially comes down to me having more than one tab open and not seeing which question I was resolving, until it was too late. I know tensions are quite high, and this error caused significant distress. The Metaculus team will be working on safeguards to prevent this kind of error in the future.

Emails are (in a small way) a bit like bombs -- they can't be un-delivered.

@(galaga) the question terms were "will gamestop reach 420.69 by **end of year**", so the question was open for a wide period (initially to the end of november, then revised to end of march). when a question resolves early, usually it is appropriate to not change the closing date, so you're awarded points for only a fraction of the total period (about 3 days out of 60, here). if this rule were not in place, it would create an incentive to always predict 99% early in the question, then gradually revise down to 1% as the question approaches the close dat...

the performance of PaLM indicates this is closer to a matter of "when will somebody bother to run a test that meets the criteria" than "when will AI be capable of meeting the criteria"

@mitrontatidis the opposite happened: the community's prediction moved a lot and then people griped that we should change it to "weakly" general AI, and we agreed.

resolved as Yes according to twitter's amended 13D/A form

@blednotik in questions like this, forecasters are asked to express their truly held beliefs in respect to the monumental importance of having public, quantitative forecasts. forecasters are asked to ignore the expectations of metaculus no longer existing, or the forecaster no longer existing or valuing points, etc.

# casens candidacy my fellow metaculoids, i believe in the dream of metaculus. in so many fields, *reading about* a subject confers much less knowledge than *doing* the subject: you can learn programming by reading about python, but you learn 50 times more by *writing* python. you can learn about the history of the telegraph in a well-written book, but you'll learn more by trying to answer a question that hasn't been answered yet, piecing together the evidence yourself. metaculus is like this: instead of passively taking in information about politic...

resolved as Yes according to this SEC filing.

not that it matters to anyone who isn't me, but kudos to wired for being the first source i read to actually cite the SEC filing, instead of just reporting the story on prestige journalism authority alone.

@rob.louis there was a brief period where metaculus users generally noticed that all binary questions had a 30% chance of resolving positively. as such, there were some attempts to randomly invert the +/- sign on binary questions, which i don't think anyone on the moderation team is currently doing.

@(Primer) the problem with this is that there are very credible claims of fighting on february 25 from [washington post](https://web.archive.org/web/20220306092949/https://www.washingtonpost.com/world/2022/03/06/ukraine-kyiv-russia-zoo/) and [reuters](https://www.reuters.com/world/europe/russian-troops-attack-kyiv-military-base-are-repelled-ukraine-military-2022-02-26/). since they do not give a clear indication of whether there were more or less than 100 soldiers, it would be overconfident to resolve this as yes or no. our options were: 1. wait to get...

i've just re-read the terms of the prediction, which state that unless there is media consensus on the winner of the election within 1 week of election day, the question resolves ambiguous. that would either imply very quick counting of absentee ballots, or a hugely decisive victory for either trump or biden, and under those outcomes, a biden victory is clearly more likely. it's a shame though, there's probably a 40% chance or greater that this question will resolve ambiguously under these terms. a lot of people will probably be unhappy. unless we wa...

@(Rexracer63) we have some thoughts about revising overall user ranking, and when we do make those updates, we'll want to make it a fully-polished system. two objections i can think of to a "points per question" system: - the overall amount of points a user gets in a question goes down over the lifetime of the question. a user incentivised to maximize points per question could try to predict as early as possible on every question; if they come to a question late, the total point award might hurt their metric even if their forecast is better than the c...

@Linch although coronatime can be disorienting, it is not yet 2021-01-21

@jabowery

1. almost half of people have already voted

2. BLM has more favorability than unfavorability, and biden's support actually went up after the george floyd protests

3. there's not that many undecided voters, and it's unlikely that they'd be influenced more by this riot than the george floyd riots

4. i'd say biden still has an 85% chance of winning pennsylvania, but even if he didn't, that's not a guaranteed trump victory. given trump wins pennsylvania, i'd say biden has a 30% chance.

@Joker i actually think the downvotes on your comment specifically are an order of quality much higher than you would typically see on reddit. share your forecasts, reasoning, and sources and you'll be welcomed. calling people stupid and "woke" doesn't contribute to a healthy community where dissenting opinions are treated respectfully.

@jacobef2 i think metaculus is comparable to the pyramids of giza, at least to some degree

@satisficer thanks, good question. a limited ceasefire such as this would not count. i've added this language to the resolution text: "The ceasefire must apply to all military operations in all official and disputed Ukrainian and Russian territory. In other words, a limited ceasefire (such as granting safety to humanitarian corridors or specific regions) is insufficient to resolve the question as Yes"

@predictors please be aware that i've added this clarification to the text.

i understand that people dislike the historical bias of "~80% of binary questions will resolve negative" but this question is a good example of why inverting the question phrasing is more trouble than it's worth. it leads to unneccesary linguistic contrivance, which probably increases people's errors in recording their true predictions. might also discourage people from participating at all.

but if i had a gun to my head, i'd phrase the question as "will trump's first term end before the next supreme court appointment?"

— edited by casens

this currently holds the record for "most points i've earned on a question" as well as "most points earned on a question which i didn't deserve"