Did you catch it ?

[PeterNSteinmetz]

At most he said he was a scientist: "I'm rooting for my fellow scientists..."
I didn't see where any of his points depended on his authority as a scientist.

And since arguing from authority is a fallacy, his credentials really don't matter much since his "authority" is irrelevant anyway. I'd just let it go.

Yes, I am always more interested in the data and the arguments.

There are several points in this thread where he has made bold assertions and refused to provide the claimed analyses or any published studies as well as claiming that those who disagree with him are engaged in magical thinking. So I do not agree that there have been no attempts at an argument from authority as I think the essential thrust of such statements are “trust me, I’m an authority”

However, I do agree that such personal characteristics are uninteresting as a point of discussion. The main problem is the lack of data, claimed analyses, and other fallacies. Would you like me to quote those problems with the arguments that have been made? Sort of beating a dead horse but if you really want to understand the problems with the science, I can do so.

 

[PeterNSteinmetz]

Reasoning: IgG takes on average 10 days to develop since infection and be measurable. Medium time to death since infection is 20 days. So there's about 10 days of measurable IgG cases that are still at risk of death.

I don’t believe this is the correct way to compute this. If the population is in equilibrium, one should just divide the current numbers and that will make the units work out.

Of note, the Santa Clara study estimated 0.12-0.3. There are a range of estimates presently, but the 0.57 from NY is definitely on the high end.

The different estimates may well reflect differences in the age distribution as there is a strong age dependence in the IFR for SARS Cov-2. Might also reflect differences in the type of care being received in different populations.

More dangerous than the seasonal flu, definitely, but a far cry from the 8% everyone was fearing.

 

[PeterNSteinmetz]

Which is a long winded way of saying if you mean to discredit the input of someone

That is not my point. My main point is that if you are going to try and make an argument from authority and claim that others are engaged in magical thinking (sort of an ad hominem fallacy), that is fallacious reasoning.

And if you are going to engage in that type of fallacious reasoning, then it is fair for people to ask what on earth the basis of this alleged authority is?

But hey we can go back over the fallacious lines of reasoning and arguments and lack of data and ability to cite if you like. That is what really shows the level of credibility that should be assigned to the speaker. But I think that point has already been made pretty clearly.

 

[PeterNSteinmetz]

The Santa Clara study is flawed. The test they used was not sensitive/specific enough for such a low prevelence population. If you ran that exact test 10 years ago it would show a similar number of positive cases.

That has been the argument. But have you read the study and the controls the authors and the manufacturer ran?

If you do, I think you will see that this argument which has been made by a number of people is incorrect.

The manufacturer of the test used had about 300 pre-Covid-19 controls which tested negative.

While a 95% confidence interval for the false positive rate is broader, the best estimate of the sensitivity and specificity of the test is as stated by the authors.

And the best estimate of the IFR for that population is the 0.12-0.3% they stated.

But arguing about this is relatively meaningless unless one has read and understood the study and the statistics used. It is not a simple “that study is flawed” sort of issue.

 

[PeterNSteinmetz]

If the answer to that question would put your livelihood in jeopardy then I don't think you'd be too keen to answer but maybe that's just me.

I agree that would make me reluctant to do such a thing. OTOH, I am also reluctant to engage in that type of fallacious reasoning and making statements based upon same because it would reflect poorly upon me.

Much easier to engage in that kind of behavior when writing under a pseudonym - don’t you think? A side note - I don’t imagine your wife, as a well accomplished and established scientist does that sort of thing in fields related to her area of expertise.

 

[PeterNSteinmetz]

Reduction ad absurdum: Infect everybody at the same time. 10 days later you will measure 100% IgG but measure 0% deaths. You'd have to wait another few days to see actual deaths coming in. If you wait too long though you'll get deaths that did not have IgG picked up by the serology test. So using average time to There will be a percentage of people who are IgG positive that will still be dieing. If it was just important to be alive on the day of the test then you can compare it but most people will find it important to be alive 2 weeks later as well.

.

An interesting way to think about it, however, such a population is not in equilibrium. To compute in a non-equilibrium situation you will have to use different types of averaging over the solution to a differential equation and estimate parameters based on that because the variables are changing.

Easier in the equilibrium situation where one can see that if you multiply the fraction infected by the size of the population infected you get number infected. If you use that to divide deaths in the population, that will give you deaths per person infected.

I can see one fine point to note here though which might cause some confusion. The test for antibodies is assumed to reflect effectively lifetime exposure, so the number infected computed in this manner is the total number ever infected in the population.

 

[PeterNSteinmetz]

This is an internet board for pilots, not a learned professional society. He didn't even, AFAIK, even make the claim he was a scientist until one of the later posts in this thread, that I quoted, but I didn't bother to look.

I don't think there are problems with the science itself. Perhaps the interpretation of the science, and also there was a lot we didn't know, not very long ago, that we do know now. Tie that in a discussion about how to balance safety and the economy, and there will be differences in opinion.

Well we may have to disagree a bit on this I guess. I think there is a major problem with:

A) Stating you have performed an analysis which is conclusive but being unable to provide the data or analysis or even a summary of same.

B) Stating it is absolutely clear that social distancing has reduced Covid-19 spread but being unable to provide citations to any studies, data or analysis that shows that is true.

C) Ignoring repeated requests for same while claiming you are a scientist.

D) After that, claiming that anyone who disagrees is engaged in “magical thinking”.

I think these behaviors are a problem, scientifically speaking. That is not how one makes scientific progress. Also not the sort of person I think of as a responsible discussant, even on a pilot’s board.

I didn’t even bring it up until the claim and argument to authority was made. In my experience, these are not the sort of behaviors which scientists properly trained in research engage in.

And you are correct that the ability to process and remain objective about things when the data is uncertain, rather than always thinking that some one answer has to always be “absolutely clear” and recognize that data changes and one may say “based on the new data, I changed my mind” are one of the things people trained in research learn how to do.

 

[PeterNSteinmetz]

What?

I get how you'd think that. But you'd be wrong. She very much cares about what the public thinks and she very much pays attention to what the public thinks and she very much wants to engage the public in the conversation about what the public thinks. Both what they think that is correct and what they think that is completely incorrect.

And lots of what they think is incorrect. But her ability to engage and educate them is limited because if she says too much, she'd be fired on the spot no matter how talented she is.

Sorry about any confusion. Maybe to many “that’s” involved.

I will try to be briefer and clearer. I think it is much easier for people who are writing under a pseudonym to engage in fallacious attacks and reasoning. I do not engage in such fallacies personally mostly because I think they are the wrong way to make progress in science and policy, are disrespectful to the people I am discussing with, and are not persuasive. Consequently I would not do so even if I were writing anonymously.

I imagine your wife would not do so either for the same reasons, even as she attempts to help educate the public. I am sorry that she is restricted by her employer but I know that larger organizations can be very conservative in this regard.

 

[PeterNSteinmetz]

I imagine most employers, large and small, would not be happy about employees discussing things related to their business on a public forum such as this. As far as members using their "real name" instead of a screen name, anyone can sign up with a name that sounds real. They can claim all the credentials they want. That doesn't mean any of it will be true. This is a message board, not a scientific journal.

Most more academically oriented employers are more flexible on this point. It is historically one of the reasons professors have tenure - so they can speak their minds. But about commercial employers I agree.

It is usually possible to verify people’s academic credentials by name, if one really wants to. And one could verify identity of a poster in a number of ways, again, if one really wanted to.

My point, as I noted above, is that if you are going to engage in all sorts of fallacious reasoning and attacks, that reflects poorly on the quality of the poster, whatever their credentials. But if you then claim to be a scientist why doing so, it is quite reasonable for people to ask “oh, so where were you trained?”.

I think the faulty reasoning and attacks in this case make it fairly clear the level of training or lack thereof. And that is reinforced by the inability to provide any credentials to back up the authority which was claimed.

But I am glad to hear that people primarily base their judgements based on the quality of reasoning and facts adduced, rather than an argument from authority.

You are correct it is a message board and not a scientific journal. Nonetheless, I continue to expect that responsible people I discuss scientific issues with can engage in polite discussion lacking fallacies and based on actual facts and arguments. Otherwise, I find it quite uninteresting I’m afraid.

 

[PeterNSteinmetz]

Actually I think your attacks on the other poster by asking that he or she identify themselves rather than simply analyzing their argument reflects more poorly on you than the other way around. People who are still reading this thread can make their own judgments.

Well please go back and read the whole thread. I think you will see that I repeatedly attempted to do just that and never brought the matter up until the other poster starting making an argument from authority and indirect ad hominem attacks on anyone disagreeing with him. Then I simply noted (post #249) that if he was going to make such arguments, anonymously, that perhaps he could please identify himself and his credentials.

Others then began becoming very concerned about this request, which I maintain was a reasonable one under the circumstances.

If you focus on arguments and data and don’t try and make arguments from authority, then I agree that individual credentials should not be relevant here at all (as I noted in my post containing the question.)

OTOH, I don’t think one should simply let repeated fallacies while trying to cloak oneself under some likely false authority go unnoted. I do find such behavior a bit exasperating and will call it out from time to time. A much more outrageous example we previously saw was the user mryan75, though he didn’t claim to be a scientist.

 

[PeterNSteinmetz]

So if I'm reading this correctly, you're saying I have a point that validly counters your point, but because, and only because, it is technically possible for someone to post something malicious anonymously, nothing I've said is therefore valid.

I must confess I do not understand how you get there.

If it is important to you, perhaps expand on your reasoning a bit more? I do not understand which point you think you have made somehow counters some point I have made. Truly puzzled actually. So maybe spell out those points out explicitly? And I certainly would not claim none of your points were valid - I thought you brought some interesting points up.

 

[PeterNSteinmetz]

The two of you obviously disagree. So what? The people reading can make their own judgment. I have been reading the thread, and you repeat yourself over and over about how there isn't any proof (that you accept, anyway) that coercive distancing has worked. See, I can even quote you from memory. But this is not a controlled experiment, so there can be no empirical proof, only anecdotal proof.

Well here is an actual scientific point to discuss. I disagree on this last statement. There are plenty of actual empirical data which could show that social distancing works in the US for Covid-19. This is not a non-falsifiable belief.

For example, one could correlate measures of social distancing such as cell phone records with decreases in the growth of actual death rates.

Or alternately one could show that coercive social distancing policies (which I tend to be more concerned with) have had a measurable effect of the growth rate of deaths by comparing states or counties.

I think one can imagine lots of other potential studies which could bear on this question.

So my general point has been where is such data and analysis? And without it, why should we conclude that it is absolutely clear that social distancing works here in the US for Covid-19?

There actually was a preprint which reported a study that analyzed to the end of March which purported to show that such increased isolation measured from cell phone records correlated with decreased confirmed cases. So one could cite and discuss that (which unfortunately didn’t happen). That of course is not a correlation with deaths, and was a quite modest effect (about 20%), and was so poorly reported that it makes one wonder about it’s validity.

On the other hand we now see the report by Reilly that there is no effect of having a coercive social distancing order on deaths when adjusted for population size and density. I have personally examined that analysis and confirmed that result on a similar dataset. Obviously that is a limited way to make the comparison, but it is accurate so far as it goes.

I have been searching and the problem of empirically determining whether an intervention has an effect on the growth rate of epidemics has been much more lightly covered than I would have expected. But I am continuing my research into this hoping there has been more and better studies on past epidemics.

There are a lot of model studies on Covid-19 which show that the model predicted more deaths and there have been fewer. But that is not particularly strong evidence that the interventions which have been applied were causative of fewer deaths, as many have argued.

Do I think it is reasonable to argue that the data are not terribly clear at this point, and to debate the relative strength of the evidence one way or another - absolutely - that is how science advances. But I do not think it reasonable to assert that there is no empirical data which can bear on this question, or that the answer is crystal clear, or consistently de-emphasize data because it doesn’t agree with one’s pre-conceived notions, or engage in fallacious arguments to try and support one’s ideas.

 

[PeterNSteinmetz]

It's clear that social distancing "works". What's less clear is how well it works, and whether it works well enough to be worth the cost. We'll never know how well it works or worked, since we haven't tested the alternative (I.e. no social distancing) in the same conditions. Nor do we have a good handle on the cost, indeed different people will assess the cost differently. So it's a judgement call, based on imperfect data and assumptions, and we'll never truly know whether it was worth the cost.

Largely I agree. From a policy perspective the important question is the cost benefit analysis, which as noted is poorly understood at this point.

Also important I think to distinguish between coercive lockdowns and voluntary social distancing because a lot of collateral damage is caused by coercive policies, which tend to be broader and less finely tuned to circumstances than what people do voluntarily.

I do not agree that there is no data or will not be data to address the effectiveness of social distancing or coercive lockdowns. The data and analyses on coercive lockdowns that is presently available suggest they have not worked in the US to reduce deaths due to SARS Cov-2. See the Reilly study as well as the analysis out of Israel which is more international in nature. What is unclear in that is whether that is because the effect of social distancing is too weak to reduce deaths or people don’t distance enough.

Data is available which can be used to analyze social distancing generally. It perhaps weakly suggests there is an effect on measured confirmed cases though a small one. There is also data to suggest that social distancing increased in the US as this epidemic progressed in late March early April. A key question in terms of policy is whether that was voluntary or caused by coercive measures.

I think that as the dust settles here a bit there will be more data and analyses bearing on this question, though the social sciences are never so clear as harder sciences in their conclusions.

 

[PeterNSteinmetz]

The thing is, you only know how many cases/fatalities there are in your location (I guess you are talking about Syracuse) with social distancing implemented. You don't know what the count would have been if it had not been implemented.

There are 8 states which did not have coercive lockdowns. The comparison between those and the states with coercive lockdowns, as performed by Reilly, showed that if you control for population size and density, there is no independent effect of coercive lockdowns on deaths due to Covid-19 when looking at the most recent day in his dataset.

I have actually looked at the development of deaths over time since the state crossed a threshold of 0.3 deaths per million and there is no obvious separation of coercive lockdown states versus non-lockdown in that either (proper statistical testing is still being worked on).

Quite correct that the population even in states without a coercive lockdown may have been engaging in some level of social distancing voluntarily. One can get some handle on this with indirect measurements such as cell phone mobility, though these measurements obviously have their own limitations. As far as I am aware there is no study yet looking at correlating these cellphone records with deaths rates. Would be an interesting one.

 

[PeterNSteinmetz]

The states you mention may not have statewide orders, but various local agencies have mandated distancing requirements. In addition, even the states that do have statewide orders leave many exceptions. Therefore there is no way to conduct any kind of accurate study. Garbage in, garbage out.

Well maybe best to approach this from a different angle. Some questions:

Is the any evidence, at least possibly obtainable, which would persuade you that the statement “the coercive lockdowns have decreased Covid-19 deaths or confirmed cases” is false?

And do you assert that statement is true?

 

[PeterNSteinmetz]

A harrowing perspective on the worldwide consequences of shutdowns based on the recent WHO warning. 250 million may starve to death.

https://medium.com/@davidgoldhill/treat-the-patient-not-the-virus-f8dcd87231e7

 

[PeterNSteinmetz]

Population size is one thing, but how do you control for population density? Is the effect linear? Polynomial? Exponential? The difference between NYC and, say, Omaha, aren't just a matter of density, but demographics, income level, education level, mobility (consider the NYC subway system and the intense commuter culture), the relationship certainly won't be linear.

Reilly’s analysis used a linear factor on population in a state and it’s average population density. I used those plus their interaction. All 3 were significant, so that is including a population squared term, effectively. The analysis also included several of the other items you mentioned. Perhaps read his article?

If you look at the graphs of deaths on a log scale, normalized for the maximum values on a common day as a function of time since reaching the threshold of 0.3 deaths per million, there is no obvious distinction between lockdown and non-lockdown states in the shape of the curves. (I have not figured out the best statistical test yet.)

 

[PeterNSteinmetz]

Well maybe best to approach this from a different angle. Some questions:

Is the any evidence, at least possibly obtainable, which would persuade you that the statement “the coercive lockdowns have decreased Covid-19 deaths or confirmed cases” is false?

And do you assert that statement is true?

Also for [USEr]@Cap’n Jack [/User]and @Dana. How would you answer these questions?

 

[PeterNSteinmetz]

Since Covid-19 is spread by contact with other people, if there is 100% no contact, it wouldn't spread. That is impossible to achieve though, so there have been varying degrees of limiting contact. This is just logic.

So it doesn’t sound like you can think of any evidence that would ever persuade you of the falsity of the statement.

If there is none, that means you have a non-falsifiable belief that either coercive lockdowns or social distancing has worked in the US with respect to Covid-19. There really is no point then discussing it further pretending evidence matters, because it does not in terms of your belief.

Thinking about it from this other angle often clarifies the issue quite quickly.

 

[PeterNSteinmetz]

As @Everskyward said, another question is whether the shutdowns were worth it.

At this point, it may be possible to say the lock-downs weren't needed only with 20/20 hind-sight. What happens if the next one is more deadly?

Well that’s a candid evaluation of knowledge, cool.

I think it will turn out to have been an enormous error, disastrous really. And my own view is that politicians should have known better. Ionniades at Stanford warned in early to mid-March that we did not have to type of information needed to be taking these kind of drastic steps.

I also agree we foolishly followed the Chinese lead on this. They are a despotic regime which censors the media - their government cannot be trusted. We still don’t know the full story on how this virus came to infect people in Wuhan. If it did come out of the lab by accident their incompetence in handling this and the panicy response of other world governments may well end up killing more people than Mao did with his cultural revolution.

It is quite sobering and appalling to consider the magnitude of this set of mistakes.