[skissane]

> Then again, too many idiots think proof means "The CCP notoriously likes to hide stuff, therefore they're hiding the lab leak, therefore the lab leak theory is real, QED". That sort of disinformation does deserve a ban.

"The CCP notoriously likes to hide stuff, which makes it more likely that they're hiding the lab leak, which makes it more likely that the lab leak theory is true" isn't that far off but seems much more reasonable.

Should we ban people for making strawman arguments, or should we steelman their argument in response?

[bellyfullofbac]

> which makes it more likely that they're hiding the lab leak, which makes it more likely that the lab leak theory is true.

Geezus Christ, just as I'm talking about scientific thinking and lack of it. No and no! It's possible, but they're not more likely!

Why they don't want Western investigators free reign investigating the lab? Imagine if the WHO asked the US to let in Chinese investigators in their virology labs. Maybe they're hiding bio-weapons research, or cancer research that they want bragging rights for, but it doesn't automatically mean there's the "Covid smoking gun" being hidden there. It's possible, but them being cagey doesn't make it more possible.

[fighterpilot]

It sounds like you've misunderstood the statement.

They're not saying the lab-leak is more likely than not given the CCP likes to hide things. They're saying the lab-leak becomes relatively more likely after conditioning on the information that the CCP likes to hide things.

That's completely reasonable if we think the tendency to hide info is going to asymmetrically impact the investigation into a leak more than an investigation into zoonotic origins. It's also reasonable even if we think they're going to hide things whether or not it really was a leak, since the impact can still be asymmetric.

[skissane]

If someone has a track record of hiding stuff, that fact makes it more likely [0] they are actually hiding something. It doesn’t definitively prove they are hiding anything in particular; something being more likely is compatible with it being false. And saying “more likely” is saying nothing about how much more likely; it could just be slightly more likely.

I don’t see how saying that is a “lack of scientific thinking”. I think that criticism is just false.

[0] “more likely” is talking about epistemic probability, Bayesian probability

[bellyfullofbac]

> If someone has a track record of hiding stuff, that fact makes it more likely [0] they are actually hiding something.

That's a tautology. So what, what if someone came with a theory that the virus comes from meteorite that people from this lab had been researching? Does their cageyness mean more likely that the meteorite theory is true? Geez...

[Closi]

> So what, what if someone came with a theory that the virus comes from meteorite that people from this lab had been researching?

The two scenarios are not the same, and denying a plausible scenario isn’t the same as denying an unplausible scenario, particularly when there are negative incentives.

Is your argument really comparing the theory that a coronavirus that hopped species might have come from the lab next door which was doing research on how coronaviruses hopped species with it coming in on a meteorite?

A better example would be if a giant smouldering crater suddenly appeared, and the Chinese government started saying it wasn’t a meteor and holes sometimes just appear because of tectonic activity, while at the same time blocking other nations scientists from seeing the crater.

[fighterpilot]

You're attempting a reductio ad absurdum but it's not valid because the meteorite has a tiny prior probability of being true, whereas the lab leak doesn't have such a tiny prior due to historical precedent of it happening a number of times, even if it is much less likely to be true as the zoonotic origin explanation in the case of COVID.

[skissane]

Thank you. I think this discussion makes more sense if one has a basic understanding of Bayesianism – one starts with a prior probability, one then has various pieces of evidence and associated conditional probabilities than one uses to update that probability, and then one arrives at a posterior probability after updating based on those pieces of evidence. I was talking about the conditional probability for one particular piece of evidence; I was never saying anything about the prior or posterior probability. Obviously you, and others, get this, but it seems bellyfullofbac doesn't.