For the record, I recognize that the linked article contains multiple things that make people skeptical about Scott Alexander (extreme length, and taking weird views seriously), but I had not heard of any of the people before reading this summary and I thought it was a really interesting discussion of what can or cannot be learned from a process like that.

The page seemed to load fine for me, videos and all, but then it slowed & froze.

I ended up using the anti-paywall trick: select-all, copy into text editor, close webpage.

If you disagree with any of his Rootclaim analyses - you think Putin does have cancer, or whatever - he and the Rootclaim team will bet you $100,000 that they're right.

Given how rich Saar is and his stated desire to make it a meaningful challenge / improve the state of debate, it seems assholish of him to make the bet $100,000 and not scale it to challengers' means. Unless he prefers to only debate other rich people.

I mostly skimmed the long debate back-and-forth, but this stood out:

Weibo lists the most popular places for people to check in to their network on their phones, and the wet market was the 1600th most popular place in Wuhan, meaning that if you weight locations by busy-ness, there's a less than 1/1600 chance that the first cases would be in the wet market.

It seems dopey to make this particular inference - ordinal rank is not percentage! It would almost certainly be far, far less than 1/1600. (Which is even better for Peter's anti-lab-leak argument, to be clear.)

Haven't tried loading the page yet, don't want to derail other stuff. Here's a source probably redundant with evidence on that page, especially if people putting the debate server and framework together are careful. It's the 2022 science paper with strong preponderance of evidence for natural origin: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9348750/ .
Having lengthy discussions of evidence for and against propositions like this is valuable, especially with dying or corrupted news organizations. Sponsoring debates with reasonable rules of conduct and evidence is a great idea. Not sure how to get around a high entry fee-- along with reasonable people without savings or a patron, there are a lot of motivated trolls/enthusiastically deluded people. Balancing a policy of openness with the real work of curation and filtering for quality of marginal submissions (those that almost no-one reads, that is, most of them) that needs to be paid for is a real challenge. This is a big issue for unrefereed forums (ArXiv, BioRxiv) and open-refereeing publications like e-Life (both Eisens are interesting figures, so is their backstory)
e-Life controversy from a for-profit competitor.

As late as the closing arguments, Saar was citing examples with no more reliable source than the Daily Mail as if they were established.

This seems like a useful comment from one of the judges:

I am skeptical that the Bayesian decision making/evaluation methods are any more "objective" than [intuitive reasoning]. I think they maximize legibility, not objectivity, and tend to hide the intuitive/heuristic portion in the data inclusion step and values, where it's harder to see...

There's also steps where someone has to assign 1/100 or 1/10,000 or 1/million odds to some component which have huge impacts on the outcomes - and they're of objective but also unknowable factors where human mentation seems pretty inadequate to distinguish between different extreme odds, so it's best seen as proxy for what outcome you want. And it results in gibberish like:

Saar said he "could have" argued that there was only a one-in-a-million chance COVID's furin cleavage site evolved naturally. Instead, he gave a Bayes factor of 1-in-20, because he wanted to leave room for out-of-model error. But (as Saar understands it) the judges' probabilistic analysis took Peter's unadjusted 1-in-10,000 wet market claim seriously and Saar's fully-adjusted 1-in-20 furin cleavage site claim literally, compared them, found Peter's evidence stronger, and gave him the victory. You can see why he's upset.

And of course, Saar didn't change his mind, either about the issue being debated or about the overall viability of his grand idea for this purpose. He just had tweaks to address where he thought the process fell short (for him).

The best recent truth-finding bet on a matter of public import. Don't think I linked it here.

I won't.

Jesus H. Christ, who cares about these nerds' desire to live in a permanent freshman week?

Yes, this is distracting us from the important work of commenting on Arby's strengths and weaknesses.

The best recent truth-finding bet on a matter of public import. Don't think I linked it here.

But, in that case, there really weren't any good arguments on the hyper-inflation side. The linked post left me believing that there were enough arguments for "lab leak" that it was worth trying to respond to them.

Jesus H. Christ, who cares about these nerds' desire to live in a permanent freshman week

There are two possible answers. First is that it can be interesting to watch someone take something way further than would normally be reasonable. The second is that it feels like the sense of shared "public reason" is fraying at the moment, and it is important to try to figure out what sorts of things can help restore it. The linked post is clear on the limitations of public debate, but it's such an obvious possible answer to the question of, "how can we agree on anything" that it's worth thinking about why it does or doesn't work.

But, in that case, there really weren't any good arguments on the hyper-inflation side.

Yes--it wasn't a debate but a bet, a bet whose underlying stupidity (it started life as a joke by the other party) helped illuminate the underlying stupidity of the hyperinflation side.

The big thing the original post tells me is that every billionaire is a public policy mistake. The secondary thing is that debate teams are bad for the world and should be sentenced to doing something worthwhile like watching every episode of Naruto.

I've only read a bit of the link, but here's the original claim that a lab leak is 94% likely.

I ain't a statistician and only vaguely understand the Bayes thing, but isn't this basically what people are describing when they talk about ridiculous interpretations of Bayesian reasoning?

I mean, the whole thing depends on where you're starting out, and he starts out with a 91% lab leak probability.

Minivet cites the key point here:

I am skeptical that the Bayesian decision making/evaluation methods are any more "objective" than [intuitive reasoning].

My heuristic on the lab leak theory is that its major proponents tend to be bullshit artists. That doesn't make the lab leak theory wrong, but it makes my starting point something like 90% against that theory.

At first glance Peter's last remarks (about intermediaries and progenitors) make me think Peter has bothered to learn what's going on and Yuri/Saar are completely full of shit.

That was supposed to say "last remarks in Session 1.5"

The more I read the more of a massacre this is. I like this Peter guy! Really does a good job laying things out.

"In Yuri's model, Wuhan Institute of Virology picked up a discarded grant and decided to do the gain-of-function half allotted to a different university, despite their relative inexperience. They skipped over all the SARS-like viruses they were supposed to work on, and all the standard gain-of-function model backbones, in favor of BANAL-52, a virus which would not be discovered for another two years, but which they somehow had samples of, which they had for some reason decided to keep secret despite its total lack of interestingness. Then they would have had to eschew all usual gain-of-function practices in favor of inserting a weird furin cleavage site that shouldn't have worked according to the theory they had at the time, via a frameshift mutation. Then they would have had to culture it, a technique beyond their limited capabilities. Then it would have had to leak, and magically show up again in front of the raccoon-dog stall at a wet market."

14: But at least, as predicted by the rich guy (ho ho), being trounced in a formal debate has made him rethink his Bayesian priors. And we all lived happily ever after.

I really liked this paragraph, which I thought did a really good job of capturing something I've tried to say before when we've discussed this here, namely that lab-leak was initially a very plausible hypothesis and there's maybe even a small chance it's true, but that all the people who think there's actually evidence for it is a conspiracy theorist.

"Peter's position is that, although the lab leak theory is inherently plausible and didn't start as pseudoscience, it gradually accreted a community around it with bad epistemic norms. Once lab leak became A Thing - after people became obsessed with getting one over on the experts - they developed dozens of further arguments which ranged from flawed to completely false. Peter spent most of the debate debunking these - Mr. Chen's supposed 12/8 COVID case, Connor Reed's supposed 11/25 COVID case, the rumors of WIV researchers falling sick, the 90 early cases supposedly "hidden" in a random paper, etc, etc, etc. Peter compares this to QAnon, where an early "seed" idea created an entire community of people riffing off of it to create more and more bad facts and arguments until they had constructed an entire alternative epistemic edifice."

12: Why do you hate Sally Rooney?

The alternative to billionaires is ? and when? News organizations are dying now, and they are necessary for a living democracy. While this particular billionaire is an idiot about the lab leak, having the debate in public (with all of its flaws and the improvised Bayesian priors BS) is still worthwhile. Yes, it would be more cost-effective to just cut a check to ProPublica or whoever, but this kind of long-winded public nerdery is a way to maybe keep interested people from checking out completely, or have the like of Alex Jones for a guide. But I don't think this guy would support ProPublica, this weirdo venue makes a libertarian impulse somewhat useful. The idea behind Vox or Jim Simons' Quanta, charitable support for longform (compared to twitter or tabloids) writing to provide context for news still seems pretty sound to me. I'd give experiments in that direction that ultimately suppport popular science or news writing a 2% prior probability of benign outcome.

20: Just bigoted against the Irish, I guess.

Yes, this is distracting us from the important work of commenting on Arby's strengths and weaknesses.

Now serving the Plausible Sandwich, containing 0-100% Bayesian Beef™.

it feels like the sense of shared "public reason" is fraying at the moment, and it is important to try to figure out what sorts of things can help restore it. The linked post is clear on the limitations of public debate, but it's such an obvious possible answer to the question of, "how can we agree on anything" that it's worth thinking about why it does or doesn't work.

It didn't work! Saar didn't change his mind and he's now saying that the method was flawed. I'll give Saar credit for putting his money into some kind of truth-hunting effort as opposed to a propaganda-disseminating effort or coke-snorting effort like so many other rich people, but not as much as I'd give him if he had admitted the problems with his starting position.

23: Update your priors at Arby's.

21: I kind of enjoy reading Quanta articles about math topics, where I don't know a whole lot, but a lot of their articles are about things I do know a lot about and they are often bad, and getting worse over time, as far as I can tell. So I don't really trust anything they publish.

25: A Facebook friend shared a photo of the neon-hat Arby's sign with letters burned out so it reads RAT BEAST, so I think I've already done that. But I'm also whatever the opposite of a billionaire is so it might not count.

As for the OP my "never engage with anything on the internet written by anyone named Scott A" rule has served me well so far.

27: Thanks. I've been spending less time on Facebook, but I know that does mean I will miss something important from time to time.

Agree that Quanta's math articles that I've read are generally pretty good. The Fields medal profiles are especially fantastic. It's possible though that I only run across the good ones. It's also possible that it's less about Quanta in general or math as opposed to physics, and more that Erica Klarreich is very good at her job.

28: ALAS.

I guess it was RAT BEEF, sorry. I'm not paying enough attention while trying to convince this poor Californian who's afraid of normal rain that it's fine for her to take a nap in the lull between a storm that was bad but not awful and the coming one that might have hail and winds and tornados, though I think they said that about the last one too.

It didn't work! Saar didn't change his mind and he's now saying that the method was flawed.

That doesn't mean it didn't work! Saar is not the only person who counts, and his stubbornness was probably overdetermined (see the slide in the linked post of odds that "Saar will be a sore loser."). I do appreciate that the post makes an effort to explain why he might be frustrated (which is not to claim that he's being entirely rational).

One of the challenges of a debate like that is determining burden of proof. Does Peter need to prove that Covid spread from a racoon dog or can he just say that is the likely animal host? Does Saar need to prove that WIV was doing gain-of-function research or can he just say that's a likely possibility among a range of scenarios in which the institute would be the source of the viral spread?

In theory the whole framework about doing Baysian estimates should help resolve those questions. In practice, as I read the post, it did not solve that problem, and that one of the take-aways from the experience (and not a surprising one, but you can understand why Saar might say, "no, there should be a better way to do this").

I'm not reading the whole thing because come on, but this isn't a coincidence:
"Either a zoonotic virus crossed over to humans fifteen miles from the biggest coronavirus laboratory in the Eastern Hemisphere."
The virology lab is in Wuhan because of the Chinese research institutes it's the largest in a region with a lot of zoonotic viral reservoirs. It's like saying Wow it's so weird the CDC is in Atlanta when most of the government is in DC without knowing the history of malaria in the US.

I mean, it's a kind of surprising coincidence. SARS1 started in Guangdong, and SARS3 probably won't start in Wuhan either!

34: Also, IPSIHSTBB "fifteen miles from" - the entire urbanized area of the city of Wuhan has an average diameter of 27.4 miles. Fifteen miles is the distance from near the city center to a research institute way out in the suburbs, a radius in which eight million people live.

Arby's: Let our priors comprise your posterior

Right, the "fifteen miles" thing is a good indication you're talking to a crank.

You expect SARS-like epidemics to start in southern China (or maybe Hanoi?) most likely in a large city, so Wuhan shouldn't be terribly surprising, but it's still relatively low-likelihood and does make you think that you should look into whether it came from the lab (and then you look into it and realize it almost certainly didn't). The initial instinct of "huh, that's kinda interesting that it's in Wuhan" on it's own is perfectly fine.

I do wish people talking about it as a coincidence would put it into context with SARS1. It's not "how surprised should we be that this one thing singular thing that's never happened before would happen in Wuhan," it's "how surprised should we be that this thing that has happened *twice* in 20 years happened once in Foshan and once in Wuhan" and I think the answer to that should be "not all that surprising."

Now if it happened a third time and that time was also in Wuhan, then it'd be a little weird. But more likely it'll be somewhere new like Shenzen or Chonqing or whatever.

I should become a "SARS1 was a lab leak" crank.

"how surprised should we be that this thing that has happened *twice* in 20 years happened once in Foshan and once in Wuhan" and I think the answer to that should be "not all that surprising."

I agree with you, but I also note that's begging the question (in the classic sense of assuming your conclusion). If you assume that SARS1 and COVID originated from the same process then looking at them as a sample of 2 does make it seem less surprising. But they would disagree with the premise that SARS1 and COVID are 2 examples in the same category.

(It is possible that I was motivated to comment just so I could use "begging the question" in the technically correct way).

I don't think this guy would support ProPublica

ProPublica's baffling article in favor of the lab leak theory makes me think maybe he would.

I mean, they're obviously two things in the same category! That doesn't mean that they necessarily came about in the same way, but if you're not looking at them as a sample of 2 then you're doing things utterly wrong and I'm not sure how to help you.

42 True; I was being sloppy to equate "two things in the same category" and "same origin." (and I think you are making a good point).

Arby's: You don't want to frequentist. (You might have to say it out loud. Or it might just be a bad pun.)

Like a serious argument for lab leak would involve a lot of "If SARS2 had a similar zoonotic origin like SARS1 did, we would expect to see X happen with both of them, but in fact SARS2 is very different and Y happened instead" and there should be *many* such differences not just one or two.

One of the challenges of a debate like that is determining burden of proof.

Outside of the "who won the debate" question, you also need to be open to the idea that the burden of proof for any theory may not yet have been met. A debate between two "some guy"s isn't going to settle that even if you could say clearly that some guy won the debate.

Regular eyes regularize regular fries.

As a possibly interesting point of comparison. The one example that I have used of a productive online debate was between Sam Harris and Bruce Schneier. In that case, nobody was declared a winner, I doubt either of them changed their minds, but it did provide a platform for each of them to talk through their positions and respond to challenges. I think Bruce Schneier clearly gets the better of it, but I think the e-mail format works fairly well for allowing each of them to make their case.

This debate was conducted entirely by email, without a moderator. While the gloves came off early, Bruce and I permitted one another to modify previous statements and to insert comments into each other's text. This occasionally complicated matters--requiring further work from the freshly injured party--but the resulting exchange is more temperate than it would have otherwise been, as well as more complete. Of course, there is only so much ripping and mending that a linear conversation can accommodate. And, as readers will see, Bruce and I still occasionally talk past one another, grow a little prickly, and leave important issues unresolved. Despite its imperfections, I think the following debate is a good example of how two people with very different perspectives on a controversial topic can engage in a rational conversation.

48: However procedurally productive the debate may have been, it has the fatal flaw that it included Sam Harris.

. . . it has the fatal flaw that it included Sam Harris.

From my perspective he plays the important role of a foil, representing a certain naïve conventional wisdom.

I don't know who Sam Harris is, but I'm still mad at Sam Wang so that's close enough.

So, a prat.

So, a prat.

I feel like 90% of the time anyone suggests a "debate" on some politically controversial issue at least one of the proposed debaters is being a prat. So I feel like the Schneier / Harris debate is a generally successful example of the typical use-case.

I think I heard Sam Harris on a podcast and decided he was a twat.

If he wasn't awful, why would he have a podcast?

You must do your research to settle some issues. Having now looked into it, I am convinced the sign says Rat Beef Sandwich.

I actually conducted a lengthy email debate with an atheist friend over whether Sam Harris is a twat. (Technically, I suppose we were discussing the merits of the New Atheism.)

Did you win the debate? I think you would have the stronger side.

I do wish people talking about it as a coincidence would put it into context with SARS1. It's not "how surprised should we be that this one thing singular thing that's never happened before would happen in Wuhan," it's "how surprised should we be that this thing that has happened *twice* in 20 years happened once in Foshan and once in Wuhan" and I think the answer to that should be "not all that surprising."

Equally, though, given that the first time the Chinese discovered a novel human coronavirus they by their own admission let it leak out of their virology labs multiple times (SARS-1), how surprised should we be by a second leak?

"Either a zoonotic virus crossed over to humans fifteen miles from the biggest coronavirus laboratory in the Eastern Hemisphere."

This also happened before! Ebola Reston was the first outbreak of Ebola in the US and it happened about 30 km from Fort Detrick.

58: I am gratified that you know, without my saying, which side I took in that debate. But alas, my friend and Sam Harris had Objective Truth on their side -- or, you know, their version of it. So I was unable to convince my friend that Harris is a twat.

Harris and the New Atheists like to talk about how, when discussing religion, they are advocating for objective reality -- without acknowledging that their own ethical systems rely on foundations that are inventions, and not the result of the discernment of facts. This error (I contend) leads fairly inevitably to things like Harris's bigotry and Dawkins' sexism.

If we keep working at it, we can develop an AI with Harris's sexism and Dawkins's bigotry.

"Either a zoonotic virus crossed over to humans fifteen miles from the biggest coronavirus laboratory in the Eastern Hemisphere."

This doesn't actually seem like a coincidence. Like, the laboratory was built in that location largely because of its proximity to places considered high-risk for the emergence of a coronavirus.

Was it? Honest question. I thought that southern China was the zoonosis hotspot - Wuhan's in Hubei which is more central. SARS-1 came from Yunnan IIRC, and all the new flu strains seem to start in Guangdong. Wuhan Institute of Virology was founded in the fifties and I don't think that studying coronaviruses was a huge priority then.

I don't think we should be that surprised by either scenario before looking into it. Both are initially plausible. But once you look into it all the evidence points one way.

||NMM to John Barth, who I was surprised to discover was still alive. He wrote a lot of tedious self-indulgent twaddle but some good stuff in between. Also I had a long phase where I was very tedious self-indulgent twaddle tolerant and read a lot of his stuff. And somehow around age 12 I checked out Lost in the Funhouse which added a few aberrant wrinkled to my adolescent brain. |>

In this particular case the admonition in 65 may only be necessary for neb.

Southern China is conventionally south of the Qinling-Huaihe Line, which is well north of Wuhan; and Wuhan, though on the river is also fairly close to the Southern Uplands and their exuberant biology. Yunnan is conventionally SW, Guandong SE. (I've no insight on WIV.)

Weirdly, swiping Bumble last night I was shown a woman mentioned liking Sam Harris's podcasts.

Also relevant to zoonotic transmission is not just the local climate but also (perhaps even more so) the local economy and how much exotic bushmeat it's buying. Wuhan in 2019 was spectacularly richer than in 2003.

19: Once lab leak became A Thing - after people became obsessed with getting one over on the experts - they developed dozens of further arguments which ranged from flawed to completely false. Peter spent most of the debate debunking these

That is all well and good (and necessary!) but I don't think it in itself really changes any assessment of the likely truth of the more serious alternative hypotheses. Does the fact that grievance farmers like Nate Silver have glommed on to the lab leak in fairly stupid ways actually make it less likely to be true? I guess maybe, but am not really seeing it myself.

Am somewhat thinking of something like the JFK assassination where I believe after Ruby there were (and still are) legitimate doubts about some key aspects of the official story. But of course a veritable blizzard of inane and insane bullshit ensued which only served to make constructing plausible alternatives well nigh impossible. Debunking the nuts is a necessary step, but in and of itself does not really shed light on what actually happened. (And for the JFK thing it is now entirely hopeless.)

Right, that's what makes it interesting to figure out how exactly to make this point. Lab leak isn't necessarily *impossible*, but it's unlikely, and all the people advocating for it are being crazy conspiracy theorists. And every time someone tries to make lab-leak more specific than just "huh, I wonder if the lab was somehow involved?" whatever specific version they propose is easily debunked. Analogies are banned, but I think JFK works as an analogy here, yeah you can see why people might not buy the main explanation, but if you look at any specific alternative explanation it's always obviously full of shit.

My absolute favorite moon-landing theory is that the first landing was staged so that they could meet JFK's promise of standing on the moon before the end of the decade. They were very close, though, and the later landings were all real.

I don't have a favorite moon-landing theory, but I do have a favorite JFK theory (which I'm about 40% convinced by); that the final bullet was an accidental (and perhaps unknown) shot from the secret service agent in the back of the car.

Conducting his own investigation, Donahue eventually decided that the bullet that struck Kennedy in the head had in fact been fired by United States Secret Service Special Agent George Warren Hickey Jr. (March 24, 1923 - February 25, 2005) from an AR-15 rifle carried in the car immediately following the President's vehicle. The proposed series of events is as follows: After the first shot (which hit the street) was fired, Hickey turns completely around and looks toward Oswald, who is on the sixth floor of the school book depository building. His turned head is documented in an AP photograph by James Altgens. Hickey reaches for the AR-15 under the seat, releases the safety and begins to lift the gun. The second shot is fired by Oswald, hitting the president and Texas Governor John Connally. The president's car and the follow-up car containing Hickey suddenly sped up. This is attested to by Secret Service agent Clint Hill. Hickey, who is unstable because he is standing on the cushion of the seat, rather than the floor of the car, begins to fall back due to the acceleration of the vehicle, pulling the trigger of the AR-15. The gun is pointed toward Kennedy at that instant, and the bullet strikes him squarely in the back of the head.

In parallel, he believes Oswald's second shot through Kennedy's neck may have already critically wounded the president before the third shot was fired.

Major keeps biting Secret Service agents because Biden told him who shot JFK.

Biden told him who shot JFK

An idle thought: how many US presidents have used their access to classified documents just to satisfy their curiosity? "I've always wondered who shot JFK but it's completely irrelevant to any decision I need to make now."

75: Related Doonesbury comic: https://www.gocomics.com/doonesbury/1991/11/15

73 is appealing but I just can't believe that a Secret Service agent could lift a rifle up, point it at the back of the head of the President of the United States, and pull the trigger, and no one watching noticed it happen, or heard the (extremely loud) noise it would have made. Oswald fired three fairly distant shots (or two, in Hickeys theory) and witnesses heard three fairly distant shots, not two fairly distant shots and one colossal bang right in front of them. An AR-15 going off ten feet away from you is not something you will easily mistake for a rifle being fired from 100 feet away inside a building.

Yeah, the trouble with so many of these theories is that they need supporting evidence.

And once you have the theory whittled down to something like 72 or 73, why try so hard to prove it? It doesn't change how one ought to view the world like far-reaching conspiracy theories do, it's just a curiosity. It's not even going to calm down conspiracy cranks by addressing their minor nitpicks, because those nitpicks are excuses for their believes, not causes for them.

Beliefs.

Grammar is funny. It makes more sense your way.

Living with grammar is being discussed in the multigenerational thread.

And once you have the theory whittled down to something like 72 or 73, why try so hard to prove it? It doesn't change how one ought to view the world like far-reaching conspiracy theories do, it's just a curiosity.

Presumably, to stop people believing in conspiracy theories?

Plus they found three empty casings where Oswald shot from, which would imply three shots - except that Hickey argues the third one was already empty, and Oswald just used to go round with an empty casing in the breech of his rifle out of habit, which is a really weird idea.

As is well known, committed Marxists habitually travel with a spent round chambered, symbolizing the exhausted power of bourgeois false consciousness; which, upon making ready for action, the soldier-worker ejects by dialectic cycling of the firing mechanism, thus replacing the old and exhausted with the new and the potent.

Checking your weapon is unloaded by checking the breech is empty is right-deviationism. Just point it at the back of the head of a passing ethnic minority and pull the trigger.

Trotskyite! Back of the neck!

It's a great theory, as we all know the Secret Service can and does cover up their crimes and gets away with it, but as ajay says it's pretty clearly wrong in this instance.

I like the Zoolander theory.

Weibo lists the most popular places for people to check in to their network on their phones, and the wet market was the 1600th most popular place in Wuhan, meaning that if you weight locations by busy-ness, there's a less than 1/1600 chance that the first cases would be in the wet market

This is really strange reasoning. How "busy" somewhere is in terms of generic footfall (or in fact aspirational Weibo posting) might be informative about subsequent spread, but it has nothing to do with origins. Guano caves in remote villages in Yunnan and mysterious Dr Evil labs have in common that neither of them attract massive human traffic or social media posting.

Also there's a clue about the Huanan South China Wholesale Market. It's in the name. In fact there are two clues and they're BOTH in the name. First of all, the point of its existence was to serve *South China*. That's a big area and a big deal. Second, it's the *wholesale* market; you buy there if you're a shopkeeper or you have a restaurant.

[Late to this but I am banned from posting from a mobile web browser, apparently irrespective of network. I get a 403 Forbidden.]