[FTX still at 14%](https://ftx.com/trade/TRUMPFEB) Given earlier comments about how supposedly the fees aren't that high, I'm sorely tempted to money against it. Still worried about the "too good to be true" effect. How worried should I be that FTX will scam me, or something else will make this not worthwhile? I find it *really* hard to believe this could actually be a reasonable price. I've been following a bunch of right-wing claims to see if there might be something there, and while I can't refute every one of them, there are enough that are obvious...
Good Judgment's superforecasters have given Democrats implied odds of [98%](https://goodjudgment.io/covid-recovery) (can't direct link, Ctrl +F Congress) to win the House. That seems too high to me. GJ's forecast has Trump at 26% to win, Metaculus has him at 35%, other forecasts are mostly in this ballpark. Conditional upon Trump winning, what is the probability Republicans win the House? I think it would be decently likely, maybe not 50%, but certainly higher than the ~10% implied by Good Judgment. Reasons: - The House of Representatives has a pro-...

This seems quite scary to me. Conditional upon being true, is the world basically back to April 2020? Or is it plausible that a VOHC would fail to become a dominant strain?

I think betting markets are overreacting to Florida and county-level swings elsewhere look decent for Biden in the midwest. Nonetheless, looking very close and uncertain.

New poll

Party support
Labour Party - 48% (down 5%)
National Party - 31% (down 1%)
ACT - 7% (up 2%)
Green Party - 6% (up 1%)
New Zealand First - 2%

— edited by honeypuppy

@honeypuppy

This got me thinking: If you are Tom Brady (or Elon Musk, or anyone else who is incredibly successful), how should that change your credence that the universe is a simulation? Presumably, it's more likely that the creators of simulations would put you in a "starring" role.

Metaculus is always right. All praise Metaculus.

@(SimonM) Predictit had Yang at 10%+ to win the 2020 Dem nomination for significant periods (e.g. [see this screenshot](https://twitter.com/NateSilver538/status/1179778622040657923/photo/1)), in which Yang is 3rd at 11%, ahead of even Bernie Sanders). I also recall other prediction markets (like Intrade) having Ron Paul at ~10% to win at some points in '08/'12. I think there's a good case to be made that prediction markets have a systematic bias towards "internet-popular" candidates (specifically, outsider candidates without widespread popular suppor...

Inspired by the "Will Navalny become President of Russia question, why not simply "When will Vladimir Putin leave office?" (i.e. be neither President nor Prime Minister of Russia).

How do we account for the possibility that the universe is a simulation built around Tom Brady?

I'd tempted to predict 41% (betting lines + 2% for this possibility). But I'm now wondering that betting lines have already priced this in.

@rodeo_flagellum

Interesting, though without proof. I doubt it's made up, but if it is, I don't begrudge you it. Metaculus trying to converge on an "ironic" answer because of a lie is interesting enough. As does people coming up with increasingly implausible stories about what Scott Alexander told them.

Liz Cheney was removed from her leadership position in the US House. According to liberal publications like Vox, [this could be indicative of Republicans refusing to certify a future Democratic win](https://www.vox.com/22420764/liz-cheney-trump-republicans-democracy-2024). I'd to see one or more questions that try to grapple with this prospect. Possibilities: - Will there will be a serious effort to overturn the results of the 2024 election? - Will be there be a *successful* effort to overturn the results of the 2024 election? The wording of what is c...

Slightly tinfoil-hat suggestion: The Economist ranks China as a democracy, not because it really is, but because China has the political clout to influence rankings like this.

@SimonM For the vast majority of questions (really all but these ACX ones), the phenomena in question is not predicting the behaviour of a single person who is aware of the Metaculus questions and wants to influence them, potentially mischievously. Questions about unemployment et al don't have the risk of someone deliberately trying to change the real-world phenomena in order to find a loophole.

I'd say conditional upon a very close election, likely one Republicans narrowly lost there might be a 20% chance of this happening. But I think the probability of the election being close enough for this to be plausibly pulled off is generously 25% at most.

For a few minutes a couple of hours ago I really thought Trump might really have tried to pull off a coup while egging on the protestors, and I mentally updated to ~5%.

Now, it's quickly becoming apparent that these are a handful of crazies with no mainstream support, and there is no support from Republican leadership to try to reinstall Trump.

I'm at 1% because I can't go any lower, but I'm not sure what odds I'd give if I could go into decimals. Any Trump second term would have to be an "outside-the-model" event, in which either there really has been widespread Democratic fraud sufficient to swing the election that will be brought to light, Trump will successfully engineer a literal coup, or something kind of in the middle which gives some plausibly legal way for Trump to stay in power. I think all such events are very unlikely, but are they closer to 1 in 10,000 likely? 1 in 1000 likely? 1 ...
@(SimonM) I think the assumptions of 50% probability and statistical independence are unreasonable. Firstly, Republicans have a significant structural advantage in the Senate (the median Senate seat [votes about 7 points to the right]("https://fivethirtyeight.com/features/the-senates-rural-skew-makes-it-very-hard-for-democrats-to-win-the-supreme-court/") of the country as a whole. Secondly, especially because Senate terms are staggered, a solid Senate majority is more likely to persist. This helped Democrats hold the Senate in 2010 and Republicans i...