Posts
Comments
Note that it says "the degree of uncertainty remaining is insufficient to render the market interesting" AND <5% or >95%.
This seems reasonable to me. Note that degree varies by market - an absolute probability by itself wouldn't be a good rule. If it was already very unlikely (e.g. "Will a nuclear bomb be detonated in NYC this year"), then for the degree of uncertainty to become uninteresting it has to become much more unlikely in a way that rules out most of the previous probability space.
I've thought about this type of case a decent amount, and I think a perfectly reasonable approach is to resove early when the uncertainty is almost entirely gone, but with the commitment to re-open and re-resolve as needed in the event that the "almost entirely" turns out to be relevant.
Classic example: "Who wins election X?" Resolve as soon as projected by mainstream decision desks. But on a large number of election markets, there will surely be some small number of wrong projections, so you undo the resolution, reopen the market, and eventually re-resolve as needed.
Pro: avoids locking up mana unnecessarily. Con: undoing resolutions can cause people to have negative mana.
Philosophically, you can look at this as giving everyone a loan based on the presumptive resolution.
Also, on the Petrov day market, let's suppose the question had been "What % of Petrov Day will elapse before someone uses the big red button to take down Less Wrong's frontpage for the rest of the day?" ("for the rest of the day" is the only change.) I would consider this reasonably unambiguous - if LW decides to bring the page back up because it was a "mistake" then it shouldn't resolve yet. But I suspect that people would have bet on that nearly the same as the actual question, and your hypothetical user who saw the site was down at 10% would also have been burned. It's still better to avoid the ambiguity, I agree, but the problem of traders being burned by details is still there even if you avoid ambiguity in the details.
This sort of thing happens in the financial markets too. I'm thinking of all the games that happen over credit-default-swaps, for example. It would be nice if we could magically reduce complexity to reduce the impact of these sorts of issues, but it's a risk market participants are taking on by trading in the markets, and I think the value the markets provide is still clearly worth it (or else people wouldn't be willing to trade in them)!
I agree that ambiguity is bad, and most Manifold markets are probably too imprecise and ambiguous. My usual style is trying to be fairly precise in the forecasting questions I write, and I definitely second your recommendations!
However, I want to point out that the problem isn't just ambiguity, but really complexity. The more you try to nail down the resolution criteria, the more likely it becomes for there to be a serious mismatch between the top-line summary of the question (the question that you are actually interested in answering) and the detailed decision tree used to resolve it.
Here's a great example with a real-money, CFTC-regulated prediction market "Will student loan debt be forgiven?" on Kalshi. The rules were as unambiguous as possible up-front, specifying what would count as "broad-based debt relief" and things like that - things that people would reasonably think are important details to check. Unfortunately, it turned out that there was a huge detail in the resolution criteria that the market missed - the resolution rules required that the forgiveness be done via Executive Order or law. And in this case, the debt relief was clearly done but it wasn't done by law or executive order. So lots of traders lost lots of real money, and the prediction market clearly failed at its job. Link to the market: https://kalshi.com/events/SDEBT/markets/SDEBT-23JAN01. (Note: the page has been updated after the fact to highlight this issue.)
This is to highlight the point that writing unambiguous resolution criteria that also match the top-line summary question is really really hard! It's not uncommon on Metaculus for there to be a fairly clear subjective answer that differs from the actual resolution because of some issue that most people agree is a mere technicality. So I think that trying to disambiguate is good, but being willing to step back from technicalities is also potentially good. There are advantages to both approaches. I agree with you that the average market on Manifold is way too underspecified and ambiguous, but if you require Metaculus levels of precision then many people wouldn't be willing to spend the effort to ask the questions at all, and that is also a loss.
I've made several updates to my guide to incorporate your input. I'll be trying the 7500 and 2291s next. Thanks again!
Thanks! I didn't realize that Zoro was owned by Grainger, that makes me feel much better about them. I also had a decent experience buying from them, it was still much slower than Amazon but not too bad. I'll add a recommended link to them.
And to add to my last comment, just to explain why I wrote things the way I did, I think the more important thing in terms of recommendations to a general audience is that in my personal experience, most people I've talked to have indeed upgraded to an N95, but extremely few people I've talked to have actually ended up wearing a P100 elastomeric. They have a ton of advantages, but most people simply don't want to use them. So in my mask article I explain the advantages, but in my general advice article I feel like it's more useful to simply say upgrade to a N95 or P100.
I agree with most of your points, especially about poor seal for typical N95s, but in my personal experience, I am often wearing a mask intermittently, not wearing it for the whole day, and in that context I typically prefer to use a N95 which I can fold and put in my pocket - that convenience is very important for me. I definitely do recommend an elastomeric P100 for long-duration wear where the convenience factor is not an issue. And I found it basically impossible to make myself audible in a loud crowded environment, although it's fine in quieter environments.
I have not done a lot of detailed research into different P100 filters, my cursory reading suggested that they were generally "good enough" for my purposes (N95->P100 being a large difference, while different P100 filters having only a relatively small difference, as far as I know), but I appreciate your info and suggestions.
The counterfeiting issues I've heard of mostly apply to N95s/KN95s/etc. I had not previously heard of it in the context of elastomeric respirators and their filters, although it wouldn't surprise me. My experience with a couple other vendors I tried has been unpleasant (issues with long delays and no stock etc), so I ended up just going with Amazon. Do you have recommended vendors?
I mostly agree with you and go into that in detail on https://firstsigma.github.io/masks but I think the "average" reader (especially one who is still wearing cloth/surgical) is better served by a recommendation to upgrade to N95 first. I find that P100s are great for many situations, but they do have their downsides.
I don't think it's the worst idea but I have a couple reasons I am actively avoiding getting Omicron this month:
- I expect a strong chance that a new variant displaces Omicron in the next say year (we had ~3 displacements in the last ~year: Alpha Delta Omicron). So then the question becomes, how much does Omicron protect against the next big variant (probably a good amount). Vs how much does your vaccine already protect you against the next big variant (probably similar). If that's the case, I don't see much advantage to getting Omicron.
- I'd expect Omicron to infect a large fraction of the population, but it's unlikely to infect "almost everyone" (before Omicron is displaced by the next big variant). So I definitely think it's still avoidable.
- Getting Omicron in January is the highest risk time due to strain on treatment capacity. Also Paxlovid is becoming more available. I think the risk of very bad outcomes will be much lower in March.
- If what you're optimizing for is protection from severe disease, vaccines already do really well.
- The severity of Omicron looks less because more of Omicron's cases are breakthrough infections, which are less severe. If you look at the severity in people of a given vaccination status, it's about ~50% less severe by the latest estimates I could find, which is a lot better than Delta but similar to Alpha.
I have a post on some of these topics: https://firstsigma.github.io/omicron