980 replies to this topic

[Daniel Cooper]

 

What is the argument now.. that if less than 8,000 Swedes die before October, herd immunity was a success?

 

My argument is that so far the models have been very bad. Which that 8k number (which probably isn't too far off base) supports.  8k is quite a bit away from 32.5k to 65k. So saying that we'd get an 80% reduction if everyone wore a mask based on a model isn't exactly an established fact (I think it's in the realm of possibility). I have no opinion on Sweden's approach and I personally do wear a mask in public.  I do however have an issue with discussing model outputs as if they were a posteriori truth.

 

These models have really served us poorly.  We are making decisions in the US alone of perhaps 5 ~ 6 trillion dollars or more (economic loss due to shutdown + new federal expenditures) based on models that have been off by factor of perhaps 4x to an order magnitude or even greater.  

 

 

 

 

 

 

[gamesguru]

My argument is that so far the models have been very bad. Which that 8k number (which probably isn't too far off base) supports.  8k is quite a bit away from 32.5k to 65k. So saying that we'd get an 80% reduction if everyone wore a mask based on a model isn't exactly an established fact

 

These models have really served us poorly.

 

The "model".  Problem is there are lots.  The IHME was off, sure, and a lot of early policy changes were based on that.  But now we're seeing more use of the MIT model, which I think it a lot more accurate.

 

Problem is we're bleeding edge.  The IHME was literally based off a linear differential equation you learn in 1st year math.  No one trusts it anymore.  The MIT model is more of a, here's what happened in Italy, so here's how it might apply in UK.  But they don't have good dimensions yet for the social distancing parameters.  You can't just apply the model to Sweden or Korea, yet.  And then what will happen in the Fall?  Everyone's bets are off, no model can promise us.

 

We have to make guesses, decisions based on uncertainty, and Sweden is not seeing the steady decline in new cases which other locked-down countries are, they are instead experiencing steady growth through the Summer months.  It's possible herd immunity is setting it, but it's also possible the seasonal nature of respiratory viruses is setting in.  And I am cautious of a hasty return to normal before a vaccine or more knowledge is in hand.

 

 

As for the idea we can save "trillions" of dollars by putting everyone back to work.  Sure, if the virus didn't exist.  Losing 1 out of 100 people has costs beyond ethics, it spills out into the whole decade of growth: a great poet lost here, a wise doctor there, all for what?  To keep a system pumping and churning that cares about nothing but itself?  Maybe we should be hanging a question mark on a system that impoverishes others when we choose not to work, even when there is no shortage of essential goods?  Maybe something needs to change.

 

[Daniel Cooper]

 

 

As for the idea we can save "trillions" of dollars by putting everyone back to work.  Sure, if the virus didn't exist.  Losing 1 out of 100 people has costs beyond ethics, it spills out into the whole decade of growth: a great poet lost here, a wise doctor there, all for what?  To keep a system pumping and churning that cares about nothing but itself?  Maybe we should be hanging a question mark on a system that impoverishes others when we choose not to work, even when there is no shortage of essential goods?  Maybe something needs to change.

 

 

 

Show me where I said we can save trillions by putting everyone back to work.

 

You can't, because I never said it. I think you are responding to what you imagine I believe rather than what I've said.

 

What I said was that we are making multi trillion dollar decisions based on very poor models.  Do you disagree with that?  Making decisions based on poor assumptions generally leads to less than optimal outcomes. Agree?

 

Will the models get better going forward? Maybe. That remains to be seen.  But I don't think it's disputable that the models to date haven't been very accurate.

[Florin]

So saying that we'd get an 80% reduction if everyone wore a mask based on a model isn't exactly an established fact (I think it's in the realm of possibility).

 

Models, shmadels. It's not all theoretical; there are plenty of universal mask wearing examples. And if cloth masks don't do it, DIY N95s almost certainly will. Have the army manufacture them if necessary. Yeah, I know you probably agree, but I'm just trying to emphasize this or beating a dead horse or something.

[Daniel Cooper]

Models, shmadels. It's not all theoretical; there are plenty of universal mask wearing examples. And if cloth masks don't do it, DIY N95s almost certainly will. Have the army manufacture them if necessary. Yeah, I know you probably agree, but I'm just trying to emphasize this or beating a dead horse or something.

 

Yeah, I do agree.  I've probably let a pet peave (speaking of model numbers as if they are established fact) run amok

 

It's hard to see how wearing a mask is a bad idea. In fact, I was in a thread supporting mask wearing back when the CDC was saying mask wearing was useless (way back in the distant past 4 weeks ago).

 

The Asian countries where mask wearing is pervasive have had a much better time of things compared to most of Europe and the US. 

[Florin]

It's hard to see how wearing a mask is a bad idea. In fact, I was in a thread supporting mask wearing back when the CDC was saying mask wearing was useless (way back in the distant past 4 weeks ago).

 

Well, there were dumb arguments floating around that mask wearing could actually be harmful. Yeah, I was fighting the good fight in the same kind of thread, and even today (in yet another thread), a lot of people prefer lockdowns to mask wearing because "experts."

 

And now for something that's not completely different:

 

[gamesguru]

Yes it's naive to think we're still using the original IHME model to make decisions.  There are new models.

 

There are multivariate analyses, for example in Germany.  They use test results and extrapolate it against other things we know—antibody prevalence, asymptomatic spread, false positives, incomplete test results, positivity rate—to calculate an instantaneous Rt value.  From there they decide if it's enough or too much or what.  I sincerely hope no one is using that antiquated IHME model from the 1700s that a first-year algebra student could improve on at recess.

 

I'm not sure how this is being examined in Sweden, but I have questions certainly.

 

[Daniel Cooper]

I think if I wanted a really good model at this point, I'd take the historical data we've built up so far and use it to train a machine learning algorithm.  I think we probably have a big enough data set now.  You'd probably spend the most time thinking about what are the relevant inputs and how to quantify some of them.  For instance, the degree to which a particular country has implemented various mitigation steps.  I suppose you could break mitigation down into separate categories (mandatory shut down of stores, restaurants, schools, etc., etc.) and make them simple binary inputs, but some sort of numerical scale would be better except that is going to be somewhat subjective.  I suppose if you have fine enough granularity on the categories that a binary scale would probably work just fine.  Obviously demographic information is another input (age distribution being a big one). The incidence of various co-morbid  disorders in the population.  

 

If you do a reasonable job of getting good inputs in to your ML algorithm the niece thing about it is that it is usually more immune from preconceived notions of how things should work that humans can fall victim to.

 

Of course, a ML algorithm will handle a highly non-linear problem, which certainly can be addressed by a good piecewise approach using traditional methods, it's just a lot more work and can fall apart if you don't know where your regions should fall.

[gamesguru]

147 deaths so far today in Sweden

 

The MIT model is an ML.  Problem is still too many variables.. scenario uncertainty vs. scientific uncertainty.  How do you model a town protest?  I guess there is some scientific uncertainty too, in that even if we continued predictable behavior, it is far from clear the model will tell us anything useful about 3-5 years down the line.. it just hasn't been trained for that half of the outbreak.

[Daniel Cooper]

Beware of that daily death data. It's lumpier than reality.

 

If you look at a daily histogram you'll see that it has mysterious dips that correspond to the weekends.  Death doesn't actually take a holiday, it's just that deaths don't get recorded on the weekend much and sometimes those accumulated deaths get counted in one batch a few days later.

 

There's really no reason that Sweden should be taking an uptick.  Unlike the US they of course mainly used voluntary restrictions with minimal mandated (compared to us).  As far as I know, they are under the same mitigation regime that they've been under for awhile now.  I suppose the voluntary mitigation might be eroding but I've seen nothing to indicate that.

 

 

[gamesguru]

Beware of that daily death data. It's lumpier than reality.

I suppose the voluntary mitigation might be eroding but I've seen nothing to indicate that.

 

It could also be retrospective inclusions.  Doesn't matter how you slice it my friend, Sweden is now a Top 10 country for Deaths per capita.

[Florin]

It could also be retrospective inclusions.  Doesn't matter how you slice it my friend, Sweden is now a Top 10 country for Deaths per capita.

 

While Sweden isn't perfect, it's doing a lot better than many of the strict lockdown countries.

 

https://www.euromomo...graphs-and-maps

[Daniel Cooper]

It could also be retrospective inclusions.  Doesn't matter how you slice it my friend, Sweden is now a Top 10 country for Deaths per capita.

 

It is, but considerably below Italy, France, Spain, and the United Kingdom. 

 

Once the virus breaks out, as it has done over most of Western Europe and the United State, all you can do is spread out the cases over time.  Nobody dying of this disease is not an option unless you propose to keep everything almost completely locked down until a vaccine or a highly effective antiviral is widely available.  Maybe September, maybe the first of next year. Maybe later.

 

The question is, is that realistic? 

 

What Sweden has done has a certain logic to it.  They are in slightly better shape than a great deal of Western Europe but they haven't substantially damaged their economy.  I know there is a tendency to say "what's money compared to a human life?", but the fact of the matter is to some extent money and economic activity is life.  Look at what happens when you have a major economic depression - suicides go up, drug related deaths go up, crime related deaths go up, and drug abuse goes up.  We know this will happen.  So, its not that there are deaths on one side of the ledger and none on the other.  It's also true that you can not run an economy off of printed and borrowed money indefinitely.

 

If it is likely that most people are going to end up getting this disease, it is smart to damage your economy as little as possible so long as you don't overwhelm your healthcare system. That did happen to in Italy.  It hasn't happened in the rest of Europe and it hasn't happened in the US (NYC was strained at one point but not overwhelmed. Hospitals in the rest of the country are currently under utilized).

 

My expectation is that at some point I will likely get covid unless a vaccine is generally available sooner than I think likely.  I might delay getting it, but I seriously doubt I can prevent it.  I can take steps to potentially ameliorate the impact by being healthy and maybe taking some selected supplements (zinc, quercetin, maybe a few others).

 

Let me point out another issue.  I think if I polled the membership of Longecity I'd find that we're heavily biased towards middle class and upper middle class white collar professionals and college students.  These two groups are perhaps the least impacted so far. Most professionals still have a job and a great many are working at home.  Students have had schools shut down, but most are still dependents of their parents and are at this point getting an extended summer vacation.

 

This pandemic looks a lot different if you're a blue collar worker and your employer has been shut down. If you work for a restaurant or a retail outlet, you've been stilling at home drawing some sort of unemployment.  But that will not continue forever.  And the fact is, if we get a vaccine tomorrow, many of these employers are closed forever.  In my area we've had both local and chain restaurants and retail say that they will not be reopening.  So it's much easier for people like myself and likely you to say "shut it down indefinitely".  We aren't the ones likely to be taking it on the chin. I think we should be mindful of that and try not to write checks that others will have to cash.

 

There are no easy answers to this pandemic.

 

 

 

 

 

 

 

Edited by Daniel Cooper, 13 May 2020 - 11:10 PM.

[gamesguru]

Nobody dying of this disease is not an option unless you propose to keep everything almost completely locked down until a vaccine or a highly effective antiviral is widely available.

 

South Korea would beg to disagree with its aggressive contact tracing program and isolating all cases at the hospital.  They've closed schools and made a few restrictions, generally encouraged masks, but their story is largely one of success without lockdown.

 

And I'm not exactly clear how other nearby countries, Vietnam and Thailand, also have it under control.  In fact, Vietnam had its first case before the USA but no one has died yet in Vietnam!!  I think part of their success comes not from authoritarian control but civilian caution, in those areas of the world they have much more caution, care for the elders, and it is culturally expected to wear a mask when you are sick even with the cold.  All this I believe greases the wheels of cooperation and sacrifice.  This level of excellence could never be expected on a widespread basis in America.  We're sadly never going to work or think together like that.

[Florin]

The lockdowns may have been necessary at the beginning of the pandemic at least in some places due to the lack of a mask-wearing culture. But instead of using the lockdown to institute mandatory masking, the opportunity was squandered; most of the United States (and probably the EU as well) still doesn't have mandatory mask wearing laws. Even a strict lockdown state like California recommends only voluntary masking. And as the economy starts opening up, the death rate might start going up again and the only solution that will be offered is another lockdown. The beatings will continue until mask wearing improves.

[Daniel Cooper]

South Korea would beg to disagree with its aggressive contact tracing program and isolating all cases at the hospital.  They've closed schools and made a few restrictions, generally encouraged masks, but their story is largely one of success without lockdown.

 

And I'm not exactly clear how other nearby countries, Vietnam and Thailand, also have it under control.  In fact, Vietnam had its first case before the USA but no one has died yet in Vietnam!!  I think part of their success comes not from authoritarian control but civilian caution, in those areas of the world they have much more caution, care for the elders, and it is culturally expected to wear a mask when you are sick even with the cold.  All this I believe greases the wheels of cooperation and sacrifice.  This level of excellence could never be expected on a widespread basis in America.  We're sadly never going to work or think together like that.

 

You don't understand.  Those countries stopped it before the genie got out of the bottle.  They started working on controlling it in early January when the Chinese and the WHO were still saying "no human to human transmission". They stopped it when there were a relatively few number of infected.  South Korea's cumulative total infected was just under 11,000.  Which means they only had low single digit thousands actively infected and contagious at any one time.  You can potentially shut it down at that stage.

 

Most Western European countries are at a quarter million cases and we're at 1.4M.  You can't trace contacts and quarantine effectively when you've got that many infected. Once the genie breaks out of the bottle, you really can't stuff it back in.  The time for shutting down this pandemic in Europe and the US is long past, if it were even ever possible.  If South Korea currently had a quarter million cumulative infected, they wouldn't be shutting it down either.

 

You can play "what if" all day long, but the ship has sailed for shutting it down in Western Europe and the US at this point.  You can control the rate at which it spreads, but a total shutdown isn't within the realm of possibility.

 

I wish this were not so but this is the situation we find ourselves in. Blame whomever you will, but we have to deal with the world as it is, not how we wish it would be.  The virus is here, and it's not going be stopped by closing restaurants, or retail stores, or bars or anything like that.  At this point the only option on the table is to balance the damage to our economy with the need to not overwhelm our hospitals.  And this is what we'll be doing until we either have a vaccine or an really effective antiviral. 

 

I mean, if you've got a master plan to drive the number of new infections to zero I think we'd all like to hear about it.

[Florin]

Once the genie breaks out of the bottle, you really can't stuff it back in.
 
You can play "what if" all day long, but the ship has sailed for shutting it down in Western Europe and the US at this point.

Yeah, others also claim that the genie has left the building and that the ship has left the barn. But once M.A.S.K.s force the horse back into the garage, dogs might be able to sniff out its contacts.

[gamesguru]

I'm all in favor of masks instead of uncalculated shutdowns.  The problem is how our loony-breathed president insists the pandemic is over, and refuses to do anything about the mask shortage or even wear one himself because of image and status concerns.  Rather than spook the stock market now, he'd rather put off the inevitable until it can cause more damage.  In this respect nothing has changed since January.  Masks are one of the ways Germany and Italy have gotten ahead of this thing, while the UK and US have fallen miserably behind.

 

So yes, shutdowns are pretty much required til we get a further game plan together—more masks and more testing—which it feels like the US has absolutely squandered these months I've spent at home.  I literally stayed home to buy my president time to tweet and set up his incompetent son-in-law to supposedly solve the PPE shortage, which also supposedly never existed, along with all the fake news on poor testing capacity.

 

It's possible things will improve enough during the summer months for policymakers to say fuck it.  But what if we get slammed by 200,000 more deaths in the Fall?  I guess then they'll take it seriously—ramp up mask production, temporarily lock down again, probably pass the UBI legislation.  It won't be pretty if he's wrong, but the president is basically banking on the fact the pandemic is over.

[Florin]

I'm all in favor of masks instead of uncalculated shutdowns.  The problem is how our loony-breathed president insists the pandemic is over, and refuses to do anything about the mask shortage or even wear one himself because of image and status concerns.  Rather than spook the stock market now, he'd rather put off the inevitable until it can cause more damage.  In this respect nothing has changed since January.  Masks are one of the ways Germany and Italy have gotten ahead of this thing, while the UK and US have fallen miserably behind.

 
The Orange Fatso is a problem but so are the Demented Governors. As I mentioned before, most of the States still don't required mask wearing. Even New York mandated mask wearing only after the death rate peaked but only when social distancing isn't possible, whatever that means, and didn't give the law any teeth. You can't make this stuff up.
 

So yes, shutdowns are pretty much required til we get a further game plan together—more masks and more testing—which it feels like the US has absolutely squandered these months I've spent at home.  I literally stayed home to buy my president time to tweet and set up his incompetent son-in-law to supposedly solve the PPE shortage, which also supposedly never existed, along with all the fake news on poor testing capacity.
 
It's possible things will improve enough during the summer months for policymakers to say fuck it.  But what if we get slammed by 200,000 more deaths in the Fall?  I guess then they'll take it seriously—ramp up mask production, temporarily lock down again, probably pass the UBI legislation.  It won't be pretty if he's wrong, but the president is basically banking on the fact the pandemic is over.

The time for lockdowns is over; it's time for individuals to take responsibility even if their government won't. There are plenty of cloth masks and face shields out there, and if people work in close quarters, N95s are fairly easy to make. If the death rate spikes again, the only people that should suffer are irresponsible individuals that didn't use PPE (hopefully, health care organizations will have learnt their lesson). So, it makes no sense to punish everyone by doing lockdowns again. But who am I kidding, a lot of (most?) people seem to be dumb about this, and I'm not sure if they'll get smarter any time soon.

Edited by Florin, 14 May 2020 - 08:22 PM.

[gamesguru]

As I mentioned before, most of the States still don't required mask wearing.

The time for lockdowns is over; it's time for individuals to take responsibility even if their government won't.

 

That's because it's become political.  They can't punish for not having masks when there's no supply of masks for people who aren't equipped to sew their own.  They're trying to procure their own masks and testing supplies, but it's quite challenging without the full cooperation and support of the clown in the oval office.  Texas is not even trying to procure masks, California is.  That's like blaming States for not having bedspace for everyone at the peak when the Federal government provided virtually zero surge capacity, and merely used it as a chance to redirect PPE to swing states?  Kushner is a big baddie in this mangled equation.

 

It's pretty optimistic to think people are intelligent enough to act responsibly on their own without any regulation.

[Florin]

They can't punish for not having masks when there's no supply of masks for people who aren't equipped to sew their own.

 

It's pretty optimistic to think people are intelligent enough to act responsibly on their own without any regulation.

 

You can easily buy masks today, make no-sew masks, or you can just wrap a t-shirt over your face. Ditto face shields. Even if some people don't mask up or there are no mandatory mask laws, the easy-to-DIY-no-sew N95s will protect you. There has been plenty of time for people to learn about this stuff. No more excuses.

[gamesguru]

You can easily buy masks today, make no-sew masks, or you can just wrap a t-shirt over your face. Ditto face shields. Even if some people don't mask up or there are no mandatory mask laws, the easy-to-DIY-no-sew N95s will protect you. There has been plenty of time for people to learn about this stuff. No more excuses.

 

I would hardly call that guide easy.  Some people are missing an arm, or fingers.  You literally aren't thinking this "everyone needs a mask" regulation through.  Michigan's Governor is proof if you try to boss a conservative base around too much they'll literally threaten to kill you.

[Daniel Cooper]

I would hardly call that guide easy.  Some people are missing an arm, or fingers.  You literally aren't thinking this "everyone needs a mask" regulation through.  Michigan's Governor is proof if you try to boss a conservative base around too much they'll literally threaten to kill you.

 

Unlike say .... ANTIFA?

 

 

[gamesguru]

Unlike say .... ANTIFA?

 

Hah, at least they represent an intelligible cause.

 

Plus Whitmer's position is fully reasonable and she's willing to discuss how the barber shops ought to remain closed til the hospitals feel relief.  Trump runs off stage like a wimpy little girl at the first sign of real questions, and he won't discuss except with OANN and Fox reporters.  What a dimwitted stooge he is

[Florin]

I would hardly call that guide easy.  Some people are missing an arm, or fingers.  You literally aren't thinking this "everyone needs a mask" regulation through.  Michigan's Governor is proof if you try to boss a conservative base around too much they'll literally threaten to kill you.

 

Are most of these people's family and friends missing an arm or fingers? Yeah, it's a rhetorical question.

 

They're mostly pissed about the lockdown, and I don't entirely blame them. The solution is to lift all of the lockdowns and mandate mass masking. Most of the people that won't wear masks are probably concentrated in less dense areas and will present less of a problem.

[gamesguru]

Are most of these people's family and friends missing an arm or fingers? Yeah, it's a rhetorical question.

 

They're mostly pissed about the lockdown, and I don't entirely blame them. The solution is to lift all of the lockdowns and mandate mass masking. Most of the people that won't wear masks are probably concentrated in less dense areas and will present less of a problem.

 

"just make a mask, get creative"

Sounds like you'd be assassinated on day 1 for lack of sympathy

[Florin]

"just make a mask, get creative"

Sounds like you'd be assassinated on day 1 for lack of sympathy

 

No, I'd make a standard N95 design and have them mass produced and distributed. But I'm not the guy in charge, so the only thing I can do is point to Youtube how-tos. While using non-standard N95s might not be ideal, it's a lot better than lockdowns.

Edited by Florin, 15 May 2020 - 03:07 AM.

[Daniel Cooper]

Hah, at least they represent an intelligible cause.

 

 

So it's not that you disagree with threatening to kill people with whom you disagree on principle? 

 

 

[Daniel Cooper]

Are most of these people's family and friends missing an arm or fingers? Yeah, it's a rhetorical question.

 

They're mostly pissed about the lockdown, and I don't entirely blame them. The solution is to lift all of the lockdowns and mandate mass masking. Most of the people that won't wear masks are probably concentrated in less dense areas and will present less of a problem.

 

The problem is that some are proposing to treat the entire country as if it's NYC, where half our deaths are.

 

I live in a county of about half a million.  We've had 5 deaths so far.  That's about 10 deaths per million.  NYC's rate is about 1,400 per million.  Would a reasonable person insist on a one size fits all approach to these to very disparate situations?

 

 

[Florin]

Would a reasonable person insist on a one size fits all approach to these to very disparate situations?

At first, I'd follow the examples of the successful Asian countries which don't seem to have any exceptions for less populated areas; If it ain't broke, don't fix it. After a while, I might start to reduce mandatory masking in less populated areas. But if we're talking N95s, I might be less concerned about mandating mask wearing in less populated areas. So, the approach would probably depend on the type of mask we're talking about and how fast it can be produced and distributed. I'm sure a reasonable person could pick from a wide variety of sizes including one-size-fits-all to match their situation. ;-)

Edited by Florin, 15 May 2020 - 05:25 AM.