Please read this update before participating in this market.

Let’s do it again!!!

@PatrickDelaney The OWID GitHub page says they are going to integrate the WHO data imminently. I don’t know for sure, but it seems like China just arbitrarily decided they would rather report to WHO (where they have some influence and relationships) than Johns Hopkins (where they don’t). Salem doesn’t have to resolve any market technically until it’s closed.

@BTE My understanding is that China isn't reporting to anyone. It's the China CDC, which puts out count reports. The WHO, according to their website creates their own estimates based upon government reports and their own expertise, it's opaque - so it may be that the WHO put their information after some consultation with whatever health experts from wherever. Contrast this to Johns Hopkins, which reports directly from Government reports, they literally have people manually doing data entry from government reports, and it's all open source on Github.

@PatrickDelaney The epidemiologists at the CDC report cases here in the US. Do you think every single case in US data is based on a positive test logged in a registry somewhere? No, lots of cases reported everywhere are estimated from waste water using statistics. China is not special in the sense you are claiming. The head epidemiologist at China’s CDC personally guesses there have been 900 million cases since mid-December. We changed the way we tallied and reported cases at about the same point in our first wave that China did. The fact that the WHO is reporting anything means they have the permission of the CCP, there should be no doubt about that. I can’t find any government officials denying the WHO numbers too, which is China’s version of confirmation.

@BTE Well, even John's Hopkins instructed users to defer to WHO data for now so that's all that should matter. But again Salem lives in its own special fantasy world so we're not seeing a resolution yet.

@PatrickDelaney This has been fun!

@PatrickDelaney Not to mention the 98 million "confirmed cases" since January 3rd.

@BTE We still have yet to see where this goes, it's not over until it's over. Salem could still decide to change their criteria, or just fail to resolve. Andy Martin could do the same.

@PatrickDelaney The folks in that competition are not going to let them get away with not resolving it, but they could resolve it dishonorably. There is no way they can resolve NO justifiably. I appreciate your work tracking down the numbers!!

@BTE dang, even Xi Jinping liked my comments...can't really get more of an endorsement of official numbers than the big man himself. Big shout out to his Excellency.

@PatrickDelaney Absolutely zero justification for a NO resolution.

@PatrickDelaney Alright, here's my attempt at a version of this market with hopefully unambiguous rules. It will resolve 100% based on that one Our World in Data graph.

@1941159478 Appreciated.

@PatrickDelaney That link is dead. But nice find!

@BTE My guess would be, it's an NGinx error, their server has crashed because of so many visitors. It's an unprecedented report for the past month and a half.

@BTE keep in mind...billion people in China, they all want to know about this. Seems to be a sort of glasnost moment?

@PatrickDelaney But the sentence after that is important too!

[…] if the Salem market resolves in a way that seems in bad faith or unreasonable, I'll resolve this as N/A. Otherwise I'll match their resolution, even if I disagree with or am disappointed by their reasoning or decision.

I’m assuming the Salem Center is just taking their stated resolution criteria very literally, including the part saying ‘Source for settling the market will be Our World in Data at the link below’. And if you look only at Our World in Data, this should be a NO so far. Obviously, that is disagreeable and disappointing to some. But it’s internally consistent.

@1941159478 That is NOT what Salem originally said. You explicitly skip the part where they say if that source isn’t available they will use another source.

And the 60,000 deaths reported yesterday should make it clear that they are going to report cases eventually and 100k is a drop in the bucket. That is like two blocks of Beijing, and the entire city is apparently going to get Covid by the end of the month. It is so much easier for this question to resolve YES than is appreciated by the NOs, but that’s fine with me.

@BTE Yeah I think a lot depends on how to interpret "unavailable" since OWID is still around but not very meaningful in this case.

It’s not like I’ve got any special insight into the Salem Center’s attitudes, but I feel like if they were up for resolving this based on other forms of evidence (like officials making announcements that commonsensically imply YES) they would have already. You’ve provided some articles that plausibly could have sufficed for that already.

But who knows! My tummy is at like 10% right now. My point was just that a NO resolution wouldn’t necessarily be bad faith or unreasonable.

According to the NHC, fever clinic visits – both in cities and rural areas – have been declining since the peak when more than 2.86 million people visited them on December 23, 2022.

Salem is not going to be able to ignore “official” government statistics.

@brp That is incorrect.

The OWID data is indeed an automated feed from Johns Hopkins.

China shifted from the China NIH to the China CDC, as there was an announcement from the China NIH that their reporting would shift to the CDC.

There is a few days lag between the Johns Hopkins raw GitHub feed and what shows up on their own dashboard and then another lag between what shows up on OWID.

I pulled the raw data myself in a Colab notebook, compared everything, compared the China CDC reports, did a checksum.

The data from the Chinese CDC seems to be punched in manually by Johns Hopkins. Johns Hopkins announced a reconciliation three days ago starting from Dec 18th through 2022 and now there seems to be some erroneous / data mismatch. There has been no response from Johns Hopkins on this for three days.

So all of that being said, Johns Hopkins contextualizes their data and publishes, "data notes," which help interpret country raw data feeds, which OWID does not do.

OWID is just an auto-feed, it is meaningless. Johns Hopkins is an official source. Beyond that, as discussed below, we have the actual facts that Government Officials have reported well over the benchmark, with 88.5M cases in Henan reported by a health official alone. If you linearly extrapolate that out to 90 days, that's still in the hunedreds of thousands of cases. You would need to go back to September 6th, 2020 and assume uniform distribution to get 100,000 cases per day...in just Henan.

China be big.

@PatrickDelaney Strong arguments. Thank you for explaining your reasoning. I no longer hold NO.

@BTE PRC has made several ... interesting moves in its reporting of cases. The Henan numbers are estimates by local health officials, while the national numbers are now counts of "symptomatic" positive tests. I for one joined this market in the expectation that the testing infrastructure would either fail or be abandoned in the case of a large wave of cases. Less Than Zero Covid indeed.

@brp Keep in mind, the overarching bet here has to do with Salem Center's source market being reasonable and acting in good faith.

Salem YES / Reasonable -> This market resolves YES.

Salem YES / Unreasonable -> This market resolves N/A.

Salem NO / Reasonable -> This market resolves to NO.

Salem NO / Unreasonable -> This market resolves N/A.

So assuming uniform distribution for all of the possible outcomes, the NO side is actually only at about 25%.

Now, looking at Salem and what they actually said on Dec 17th, 2022:

This market was never meant to estimate the true number of cases in China, but was always to be based on government reports. Thus, any alternative source to Our World in Data will have to rely on official numbers to be used. If at some point China stops reporting data completely, this market will resolve as NO as long as there was never 100,000 cases reported for any 7-day rolling average.

Since the start of this market, there was a lot of discussion on Manifold.Markets about whether Salem would resolve this bet properly, since it was clear that at a city level there were reports of millions of cases, and they were not saying anything, but the market was highly incongruent with the Manifold.Markets bet right here.

After December 17th, 2022 they opened the door to all government reports coming from China, not just Our World in Data.

If they keep raising the bar and saying, "oh well no it needs to be this kind of Government report, oh well it needs to be a report that supports our original hypothesis, oh well it needs to be so that only our friends win this bet" -- that's not reasonable or fair, so it needs to resolve in N/A.

So basically, taking your prior distribution of NO being at 25%, and Salem making increasing accommodations to change the bet to be whatever they need it to be, you're now much lower than 25% as a NO, and a YES bet would be a much better way to make money, because if it's N/A, everyone gets their money back, if it's a YES, you have really cheap YES shares right now.

@BTE I actually physically went in and logged in to Salem Center's Manifold, linked to this market, let them know about my research thread. Again, I don't trust that they are going to resolve things properly, wanted to get my take, but I definitely stand my research below and think that you and any other bettors should rightfully be rewarded for that.

@PatrickDelaney I disagree with your take and would say that it’s ambiguous and both YES and NO would be defensible. I feel you’ve got the YES side covered, and I of course agree that there almost certainly are enough cases right now and that the Chinese government acknowledges that. But here’s the NO side:

The question stated that they would resolve as per Our World in Data and there’s a reasonably argument that resolutions should be very literal based on their description. And if you just look at that graph, it’s a clear NO so far. The graph now also comes with a note on China when you hover reading ‘No reporting since 25 Dec’. Also note that OWID is reporting Hong Kong, Macao and Taiwan separately from (mainland) China, which it just calls China. It’s imprecise, but equating the two is also pretty common, people say things like ‘China does zero-COVID’ all the time. Everyone in the comments has also treated this as being about the mainland so far.

There was an exception noted in the original description in case OWID was ‘unavailable’. But they are up and running and aware of the situation judging from their note.

@1941159478 Look at my below research on the topic. The original wager was, "Will China do X," not, "Will Johns Hopkins report that China does X." If you follow the original Github repo, and show when the link from Johns Hopkins was taken off, and then look at the original source data, they passed the threshold.

@PatrickDelaney I read your comments here, but I’d argue the market actually was "Will OWID [in this case based on Johns Hopkins] report that China does X". And as far as I can tell the OWID/Johns Hopkins data is the same as the one from the link http://www.nhc.gov.cn/xcs/yqtb/list_gzbd.shtml in their source, which makes sense assuming they haven't committed some sort of horrible transcription error. With that data China has not yet passed the threshold with the highest reported 7-day rolling average of new cases only being about 40,000: https://ourworldindata.org/explorers/coronavirus-data-explorer?facet=none&Metric=Confirmed+cases&Interval=7-day+rolling+average&Relative+to+Population=false&Color+by+test+positivity=false&country=~CHN

@1941159478 Look at the individual reports that you linked from NhC.Gov.CN, for example:

http://www.nhc.gov.cn/xcs/yqtb/202212/cb666dbd11864171b6586887c964791c.shtml

Add up: 1. Current cases from 31 provinces. 2. Cases in Hong Kong.

So, you can review all of the days in between yourself, they go up from 105,143 + 76,361 to 97,195 + 108,371 which is clearly well over 100k as a seven day average.

@PatrickDelaney Whoa!! You just slayed Salem. This is awesome.

@PatrickDelaney Even if you reject the notion of Taiwan being a part of China...which would be totally a fair point on Salem's part. You still have Among them, 495,614 were discharged from the Hong Kong Special Administrative Region ... OK, so NO ONE serious thinks that Hong Kong is not part of China. No one. You have almost a half million cases having been discharged in Hong Kong, as reported by China. Regardless of how shady China is being, they reported over 100k cases within mutually Internally accepted, "China."

@PatrickDelaney A possible problem with your analysis is the 7 day average not total cases.

@PatrickDelaney The thing most are missing is China simply reassigned reporting to a different department in the same agency. We did the same thing at the CDC in the US. They will start reporting again at some point.

Above in the report it states regarding the 31 provinces on Mainland China:

There are 97,195 cases, and there are 24 suspected cases.

I mean...come on...3k short of 100k? It's going to be 100k, no problem...and when you take Hong Kong, Macao into account...easily 100k current cases are being reported.

I think the question is, "Does China Report X?" - which, here's the report that Salem linked to, here's what they are reporting...end of story, doesn't matter anything else. They linked us to a third-party data report and expect us to make bets off of that. If they claim that 2+2 =5 then this market is an N/A.

@PatrickDelaney Again though that is not a 7 day rolling average. That is an absolute total. Salem was technically asking about 100k/day.

@BTE open up the page link...put it through a translator. You will see at the top of the report it says:

From 00:00 to 24:00 on December 23

So I'm reading that as a daily report from midnight to midnight. Pretty clear, they even use military time.

@PatrickDelaney I absolutely follow what you are saying and I agree with you 100 percent. But still that is not a 7 day rolling average, that is one day of cases.

@PatrickDelaney OK let's look at Dec 16th-23rd. Starting at Dec 16th:

http://www.nhc.gov.cn/xcs/yqtb/202212/6a7d9897b43e42b3903bc067f8154789.shtml

31 provinces:

76,361 cases

Hong Kong:

482,878 cases discharged

Now compare to Dec 23rd:

http://www.nhc.gov.cn/xcs/yqtb/202212/cb666dbd11864171b6586887c964791c.shtml

31 provinces:

97,195 cases

Hong Kong:

495,614

So, I'm not sure why they repeat the, "cases discharged" number for Hong Kong, but even if you take the lower of the two numbers it's around 100k, so they are well within >100k over a 7 day average, just taking into account the 31 provinces and Hong Kong.

Now, China is almost certainly inflating the Hong Kong numbers and deflating the mainland numbers....but, they still went well over the 100k line for a 7 day average in terms of reporting! Again, case closed.

@PatrickDelaney are you sure those aren't running totals since the beginning of the pandemic to the day of the report? may be missing something, but the ones I've checked seem to be monotonically increasing and I'm not sure the diffs add up to > 100k

@AndyMartin not sure how I can make it more clear? Bailing on this market based on that question.

@AndyMartin Here's my track record in interpreting these reports thus far. https://manifold.markets/ManifoldMarkets/china-officially-abandons-covid-zer -- on this maret, when the score was down around 25%, I pulled up these reports and explained what was really going on in terms of China's Zero COVID policy. A lot of people doubted me. I exited once I got some profit because I didn't feel good about how the market maker, Manifold was going to resolve the bet, and I want to make sure I make my Manna while I can. Of course now it's at 98%...even prior to the source data being placed at 98%, the general market was at around 75% just based upon media reports catching up. Just read through everything I wrote very carefully, I am confident in what I wrote on your market here too. I am not 100% confident that Salem Center will resolve appropriately which would force you to resolve N/A, meaning I would not profit so that's why I'm choosing to step out for now.

@PatrickDelaney I should mention that I have worked and lived in China previously extensively, speak a limited amount of Chinese, though I don't read it, I have always used translators to figure out what's going on with these types of events.

@PatrickDelaney I think @AndyMartin asked a fair question as I had been under similar impression about the Chinese reported numbers. I don’t know how to feel about the fact that you went from being certain of your translation initially to being decidedly uncertain in your last comment….

@BTE I'm very certain and I will be happy to continue engaging on this market and supporting your point. However, I am a conservative investor. Ultimately I don't want my Manna sitting there, I had made $M 100 and I could use that to create two more markets of things that are interesting to me. I'm not as much of a ride or die investor, because ultimately I can't be sure that this is going to be resolved properly. If it's resolved N/A I get nothing. However, I do believe it should be YES based upon what I read and I will continue to comment - same as I did with the Zero COVID market.

@PatrickDelaney If you think about it from the other point, if this market does get resolved properly, I'm going to miss out on a huge potential profit. So there's that way I'm getting punished for not sticking with it.

@PatrickDelaney You're misinterpreting running totals figures.

@AlvarodeMenard Yeah @PatrickDelaney this is the point I was making in the previous comment

If you look at the first and last days listed on http://www.nhc.gov.cn/xcs/yqtb/list_gzbd.shtml , Taiwan was at 8,329,008 on December 1 and 8,663,616 on December 24

With a population of ~24M, it's not possible that they would report 8M new cases each day for many weeks. It seems straightforward that this would be a running total of all cases (accurate on the day of the report), and that the total new infections from Dec 1 to Dec 24 was ~334,608 or ~15k/day

When you look at the diffs in other regions, I don't think it comes close to 100k/day

@AndyMartin and @AlvarodeMenard ... the oringal wager:

7-day rolling average of at least 100,000 covid-19 cases at any point

*7-day rolling average* is the wager, not new cases

@AndyMartin you are changing the bet now to new cases from 7-day rolling average?

You can't just say, "when you look at the diffs I don't think," ...no one cares what you think, e.g. no one cares about your opinion, they care about what the actual results of the threshold upon what the bet is. You are the judge of this market, so have to actually copy and paste the source material, ideally linking back to the source material, as I have. Honestly, you can't be sloppy about it, unfortunately it's a lot of work, but you really can't just say, "Well I think it's X." -- that's just being arbitrary, no offense intended.

@PatrickDelaney Apologies about how harsh that sounds, I don't mean to have a rude tone. Hopefully I'm not being too hard on you here. But, the point still stands, that for any market, not just you personally, the opinion doesn't matter, the source material upon which the wager was based matters, so that should be cited first, then opinion and interpretation second.

@AlvarodeMenard Same point goes to you... "You're misinterpreting X," is meaningless, literally meaningless. You have to demonstrate and argue why it's misinterpreted in a clear way, you can't just, "throw a comment on the thread," this is not Twitter. Look and read my research, read the translations of the webpages. Please show me exactly where the error was.

@AlvarodeMenard and again, I hope I'm not coming off to rude or too hard on you...and if so I apologize. But, at the end of the day, this is about adjudicating based upon a numerical threshold, not just yelling louder.

Hi @PatrickDelaney , just to confirm: you're saying you interpreted the original prompt as being a 7-day average of total (cumulative) cases?

Honestly, you can't be sloppy about it, unfortunately it's a lot of work, but you really can't just say, "Well I think it's X." -- that's just being arbitrary, no offense intended.

I'd be interested to hear if anyone else had a similar interpretation -- that said, I don't think using the 7-day average of new daily case counts is arbitrary at all.

1. If you click on the OWID link in the market description, "7-day rolling average" in the graph options is a non-cumulative rolling average of daily new case counts

2. This is how the previous market for 20k was resolved https://salemcenter.manifold.markets/SalemCenter/china-covid-wave-by-winter (i.e. it resolved yes after the 7d rolling average of new cases crossed 20k, not total cumulative cases)

3. Reading the comments/discussion here shows that this seems to be the most common interpretation (control-F for "per day")

@AndyMartin You are correct about the meaning of 7-day rolling average and how that aspect of this market has been understood by most people.

@PatrickDelaney "Ultimately I don't want my Manna sitting there, I had made $M 100 and I could use that to create two more markets of things that are interesting to me. I'm not as much of a ride or die investor, because ultimately I can't be sure that this is going to be resolved properly. If it's resolved N/A I get nothing."

FYI - your Mana gets "clawed back" when a market is resolved N/A, even if you exited your position entirely. It's as if all your trades never happened. Nobody profits, nobody loses. You'll have a negative balance if you already reinvested those profits.

@Mira I did not know that.

@NicoDelon @AndyMartin

OK, so it sounds like the interpretation of daily cases is based upon some standard figure that I disagree with, that's fine, I understand that's how it goes, and I have to go by popular interpretation, I'm not going to dispute that.

Let's look at the WHO report, what they are showing:

https://covid19.who.int/region/wpro/country/cn

So that's around 26k new daily cases for now, which matches up with one of the figures within the reports I'm linking to. So it'll get there, it's going to get to 100k, still a safe YES bet.

Also, this doesn't take into account the Hong Kong numbers, which, OK let's say my interpretation is wrong, so we have to not count the Hong Kong numbers, you can still Google Hong Kong numbers independently and get 20k cases per day, curiously, cited by Johns Hopkins as well.

So, we're at 40k per day in both sets, based upon reports from the originally cited data sources. So we're up from 20k total two weeks ago...we'll get there by Feb 28th. However, we need to understand that China is indeed reporting this data through the source originally cited by Salem. Even though we all know that many Chinese government officials have already reported many more cases, etc., and that this data underreports, it's the original source data cited for the purpose of the bet.

I am back in market disclaimer be damned.

@AndyMartin You can update the close time to yesterday and let it stay closed but unresolved and in the meantime continue the debate here. @jack has done that before and I think it might make sense here but that is your call. That would also prevent whales like me from dumping my huge position and wiping out all the other participants betting YES with me.

@BTE ah, thanks. I hadn't thought of that option. I think the new title should be good enough to prevent new people from being misled while also allowing trading to continue for people who know what they're getting into.

@AndyMartin I can't decide whether to leave the Mana in and hope for sanity to prevail or take a huge loss just so I don't have to keep it all locked up. This market was a honeypot!!! LOL!

@BTE What's misleading is not the description. What's misleading is the mismatch between the title and the description. Most users interface and place their bets purely based upon title from the main page, either clicking, "left or right," -- however, as soon as you say, "Reported by China as measured by Salem," it throws the probability distribution out the window. The title should be, "Will China report 100,000 daily Covid Cases by February 28, 2023? (Outcome Determined by Salem Center Market)" ... otherwise new betters are going to pile on from the front page as soon as the holiday season is over and they catch wind of frontline news that China is getting slammed.

@PatrickDelaney I think you nailed it.