COVID-19

[Sclass]

I'm not arguing for or against ending the lockdown (moot in most places anyway). It's more a morbid fascination with the disease and how hard it is to pin down, even six months in. Makes me long for an old-fashioned influenza pandemic.

You mean like Spanish flu ?

Does this happen to be related to the fact you’re a prepper?

Not judging. I’ve had the desire for an “I told you so” moment for a few friends and family lately. Mostly regarding the financial sh.t storm they’re in right now. I’ve been smart enough to keep these dark thoughts to myself.

FWIW I did do some modeling in grad school. I was in a corner of science where we didn’t have a lot of experimental data and we had to come up with all these phenomenological models for the little peeks at reality our science provided. I really sucked at the modeling. I remember having my adviser sh.t can my paper because I was 2x off predicting my experimental result. I tried really hard on that one...lots of fudge factors and special terms and still I just couldn’t predict what I’d found in the lab. I got the shape of the curve but the amplitude was way off. We used to have a name for our friends who got results that magically matched their (a priori) experimental data...frauds. Some of them just over-conditioned their models with the very data they were trying to model...how about that. That was the curse of having the data before the model.

As I said, I’m shocked people can actually get as close as they do to predicting anything real.

[7Wannabe5]

@SClass:

Yeah, the author of the modeling book I am reading (BTW: Probably too easy for anybody at Jacob’s level. Third chapter is attempt to teach just enough calculus for the topic) mentions quote often attributed to Einstein “models should be as simple as possible and no simpler.”

Also, if scientific models were perfect, the Onion wouldn’t be funny. For instance, I would like to attempt to construct a complex model that uses a mathematical proof related to dominant eigenvalues to predict the effect of Covid deaths/hospitalizations on the November election in a purple state based on assumptions regarding social contacts / effective transmission contact in population compartmentalized into Rrr, Rrd, Rdr, Rdd. This is of some interest to me because I live in a mixed allegiance household and according to Nate Silver half of all households in purple states are mixed allegiance, and my strongly r BF thinks the Covid deniers are nutcases and Gretchen Whitmer is doing a good job “ being a leader.” However, it is somewhat likely that any such model I might construct would be more like unto a semi-obnoxious piece of data performance art than accurately predictive instrument.

[Tyler9000]

Apologies if this has been discussed already, but the latest data from the CDC is interesting:

https://www.cdc.gov/coronavirus/2019-nc ... arios.html

Take a look at the "current best estimate" of the symptomatic case fatality ratio stratified by age.

<50: 0.05%
50-64: 0.2%
65+: 1.3%
Overall: 0.4%

For reference, I believe the fatality rate for the seasonal flu is about 0.1% in the US.

Long story short, I think it's time to stop treating mitigation steps as a one-size-fits-all solution and start seriously discussing practical approaches appropriate for each age group.

[7Wannabe5]

@Tyler9000:

Don’t forget that there should also be different solutions for men and women, since 50% higher fatality rate for men across all age cohorts.

[Peanut]

A new interview from the premier anti-lockdown site:
https://unherd.com/2020/05/oxford-doubl ... interview/

Highlights:
Prof. Sunetra Gupta of Oxford put out a competing model in March a week after Ferguson that suggested 50% population may have already been infected, and IFR could be as low as 0.1%. She states that both models were compatible with the evidence back then. But Imperial College's was a scenario in which the disease had just arrived. (To me, this looks like a bad assumption today.) Hers assumed it had arrived a month earlier.

-Now estimates IFR as between 0.1% and 0.05%
-R rate is not useful guide bc it depends on how many have immunity and we don't have that information
-Resistance to infection could be genetic or from previous exposure to related pathogens
-Long-term lockdown makes us more, not less vulnerable as a population
-Strong possibility full return to BAU would be fine (UK specific)

I don't think I need to become an amateur epidemiologist to arrive at the likelier truth. I need to be able to figure out which expert I should believe. I prioritize experience and credentials, both of which Ferguson has to some extent, but certainly less so than some of his peers.

[7Wannabe5]

Here is a link to the pre-print paper For which Gupta was a contributor:

https://www.medrxiv.org/content/10.1101 ... 20042291v1

I agree that experts can be judged on credentials, but also reading a textbook on the topic doesn’t make one less well able to judge experts. It just renders you somewhat better able to skim their papers for merit.

As I noted in post way upstream somewhere, it is possible that innate immunity exists within general population, I just currently find the “lockdown and/or public cautiousness due to information has been very effective “ argument more plausible. I can’t define the exact boundary that will change my mind on the matter, because I intend to become more knowledgeable on the topic as time goes by and thereby be able to update my opinion more readily, but certainly if no large heterogeneous population has exceeded .2 % CFR by this time next year, absent vaccination or curative therapy, I will be singing a different tune.

The question many of us keep asking of those who believe(d) the CFR to be less than .1% is for some similar commitment to possibility of rethinking the matter.

[George the original one]

Daily cases in Alabama still on the increase, but per my calibrated eyeball looks more linear than exponential.

Like California, Alabama's rates for new cases are in the 25-30 day doubling range, so still exponential. This may or may not be an artifact of increased testing. Use the buttons on "show doubling time" to change the values.
https://aatishb.com/covidtrends/?region ... on=Alabama

[7Wannabe5]

What mechanism for effective contact would justify a linear growth model? Seems to me that it would have to involve exposure to something that was either strictly limited or linear in growth itself. For instance, a particular worm you might catch only by swimming in a particular lake or a popular toy ramping up production that gives kids a rash.

Maybe social distancing behavior damps down the growth for kind of the same reason that territoriality* damps down the population growth of a species. So, maybe towards the quadratic?

*difference between using “60% won’t have second baby unless/until can afford 3 bedroom house” vs. observed birth rate of 1.9 to 2.1 per female.

[jacob]

Rt ~ 1 in a susceptible population => linear growth. My hypothesis is that a self-aware population might adjust their behavior towards Rt->1.

Apparent linear growth might also obtain from a linear growth in testing capacity.

[7Wannabe5]

@jacob:

True, but kind of mathematically trivial.

[steveo73]

I thought I'd leave this a little bit but it's very interesting. I think we have a couple of themes here - how to discuss/debate an issue and then having a good philosophical framework to discuss science.

I'll take the how to discuss/debate an issue firstly and make some points:-

1. If you utilize logical fallacies in your argument it shows to me that you have lost the ability to think and talk about issues rationally and logically. Some examples of this are for instance stating that I know science and you don't without any clear facts to explain this point of view. It's like the appeal to nature fallacy. It is simply a way to avoid discussing the issue logically and rationally. Another example of this is the trust the expert approach because they are the experts.
2. If you have to resort to ad-hominen arguments I think you've also lost the debate/discussion.

I've seen a number of posters utilize these techniques. If you have to use these types of techniques you personally lack the intelligence and/or have cognitive biases and/or do not have the knowledge to discuss the issue rationally and logically and/or are simply choosing to not engage rationally and logically.

If you are doing this it's a personal development issue that you need to work on. These techniques in a scientific debate are really weird and incongruous as they go completely against a robust scientific method.

[steveo73]

The next topic is a framework to understand and discuss science.

I'd make some key points here:-

1. Science is not objective reality. It is a way to deepen our understanding of the world.
2. Science evolves. It changes.
3. Math is a tool.
4. Modern day statistical analysis is a huge competent of modern science.
5. Models are now built to test and understand our understanding of the science. Models are also used for predictive purposes. These two points basically feed on each other when a robust scientific method is utilized. So if a model has low predictive ability we can state the model was inaccurate and try and deepen out understanding of reality by changing our assessment of how the world/universe works.
6. Models are subject to our understanding of reality (you could alternatively call this our lack of understanding of reality) and data.
7. Poor data and poor understanding lead to inaccurate results.

=> To sum this up you would be foolish to look at science and in particular mathematical models as somehow being infallible. Now you are entitled to opinions on the science but they are opinions. That is why so many scientists have different opinions. Science isn't objective reality.

[daylen]

If you believe that science cannot model objective reality, then what is the purpose of science? ..and what is the purpose of understanding if there is no objective reality?

Unless you take a subjectivist stance, of course.

[steveo73]

If you believe that science cannot model objective reality, then what is the purpose of science? ..and what is the purpose of understanding if there is no objective reality?

You made some great points in your post earlier on. I think we are basically on the same page.

There is nothing at all wrong with creating models in an attempt to try and describe objective reality. This works really well if we are trying to understand simplistic processes but when it comes to modelling complexity we have to understand that these models will have heaps of flaws. Typically we don't understand complex processes well. So the current state of the science for complex process will typically not describe objective reality well. This is going to be a matter of degree. I should add that in a system where behavior can adapt and change we need to be especially careful. In these situations the science can be very fluid.

So simplistic processes are easier to get to an objective reality whereas complex processes are harder. Hopefully when it comes to simplistic processes we can get to a position where we do understand the objective reality of that process but who knows.

The purpose of science is to try and work through that objective reality. We may not end up with perfect information but we may end up with better information.

I don't think the problem is with science. The problem is with people who have a poor framework from which to understand science. We can't just state person A doesn't understand the science we need to develop pointers to that. An obvious example is when people use logical fallacies. If you can't even debate logically it's unlikely you understand science at all. If you believe that the science is settled and the science is 100% accurate (or close to it) it's going to be a pointer to a person who has a poor framework from which to understand science. If you are appealing to group think it's a poor framework. If you state you have to read paper A or textbook B again we have a pointer in relation to a poor framework from which to understand science. If someone cannot see the flaws in the science it's a pointer. If people think that because a study is statistically significant it holds true for the entire population it's a pointer.

Scientists often disagree. The scientific process requires rational debate. We want to deepen our understanding of the world and science in the hands of people with a good philosophical framework is a great tool.

[theanimal]

To me, this comes across as condescending. Pretty common around here.

Jacob is being remarkably patient and helpful. If you haven't already, you can read some of the Climate change thread and see that this exact exchange is being reenacted once more. And once more it is falling upon deaf ears. I won't speak for jacob, but I can imagine that would cause some frustration. If you are actually looking to have a discussion, everyone here is generally receptive. Not so much if buzzwords are just thrown around and listening is thrown out the window.

Who would have thought we'd ever have a thread that is essentially the sum of the Trump and climate change threads?

[daylen]

It sounds like you decide what science is worthwhile (i.e. "simple") based on whether the experts appear to agree. Correct?

If so, then you are using group-think as a framework from which to understand science. This implies that you are not using a scientific framework, because a scientific framework would decide based on models.

Some scientific fields have obvious moral/political implications (e.g. covid-19, climate). As you say, scientists are humans with emotions and values so it should be expected that a sub-set would speak out in a polarizing way. If so, then it would appear that your framework is flawed in such cases. This would imply that there are more decisive models lurking deeper in these fields. If so, then how can you justify your dismal of these fields when you do not understand the more complex models not being advertised(*) to you?

(*) Although, Jacob has tried.

[steveo73]

@daylen - I was going to put group think into the section on logical fallacies. It's a terrible terrible way to discuss issues. I think it has nothing at all to do with science. I also don't judge like that. I try and listen to a broad purview of experts and I try to get into the detail. I definitely haven't dismissed science.

I'm also completely cool with not knowing. I may have a lot of information but hopefully I can see the flaws in that information. I haven't come up with in hindsight silly predictions or said anything stupid so I think personally my understanding has been pretty good although that is a subjective opinion. Put it this way people aren't quoting me and stating this is incorrect or inaccurate. Your post was a response to nothing at all that I actually stated.

Your comments on group think are what I think is going on here. I think what I'm stating makes sense but it doesn't conform to the group think on this forum/thread and that is why there are a bunch of logical fallacies being utilized and illogical arguments (responding to stuff that I haven't even stated).

[ZAFCorrection]

I'm a big believer in asking myself about the edge cases of a situation. What if this pandemic (we'll still keep that terminology) kills exactly one person if left unchecked? What if this pandemic kills everyone if unchecked? Somewhere in the middle of those two extremes is an evolution from "zero fucks given" to "all possible measures."

Unless I'm missing something, the models being used say something about the number, but nothing about going from ZFG to OMFG. That's a whole extra stew of non-epidemiology science and value judgements. The models are cool and meaningful and ... *genuflects towards the nearest statue of Einstein* ... but they don't really get you all the way to policy. In this case there is a certain amount of scienciness in the invocation of science, even good science.

Edit:

That's also why I don't agree with comparisons to the climate change issue. Climate skeptics are usually disagreeing with the conclusion that the earth is going to change due to increased thermal loading. Most of the people that I know of who are skeptical of the lockdown don't care if 1-2% of the population dies. They think it is a lesser evil than maintaining shutdowns. The models don't even attempt to answer that question, so people are arguing over different things.

Though, I think a lot of people are also using some unspecified and fiddly moral calculus to get from lives lost and hospitalizations to cost-benefit weighted policy proposals.

Personally, I can't figure out if the lockdowns are/were a good idea, but I'm not down with the "42 people died so tomorrow I'm going to wear a red hat" logic. Just because someone won a Nobel prize confirming the 42, it doesn't mean the red hat necessarily makes any sense.

[IlliniDave]

Like California, Alabama's rates for new cases are in the 25-30 day doubling range, so still exponential. This may or may not be an artifact of increased testing. Use the buttons on "show doubling time" to change the values.
https://aatishb.com/covidtrends/?region ... on=Alabama

Okay, this may be a case of the inadequacy of my grasp of the jargon showing. I was looking at a histogram vs time (daily) of the number of new cases confirmed and doing an eyeball curve fit, but qualitatively I wasn't far off from a 14-day moving average calculation (see the purple curve on tab 4 of the bottom of the page at the link below).

Of course, if the number of new cases per day are increasing linearly (proportional to x), then the cumulative case count is increasing proportional to x^2.

https://covid19.alabama.gov/

To jacob's point, I don't have much insight into the testing statistics, although from the same link, about 40% of the total tests administered to date have been administered in the last 14 days, which would imply increased testing per unit time over time. The positive test rate after an initial "spike" in April has fallen to something under 10%, while the new cases per unit time has increased, which I believe also implies expansion of testing (tab 9).

Apologies for my potentially misleading statements. The point is that the apparent trend when taken superficially is not good given it reversed right on the eve of "opening back up".

Although places are opening, many are doing so reasonably. I had to take my old dog to the vet yesterday. They were taking only one appointment at a time, kept the front door locked, had people wait in their cars with their animals, and after calling when I arrived, they came out and got me when they were ready for us. I do worry about the reopening of businesses like restaurants and such, even at a fraction of seating capacity. Overall though, the large majority are not behaving like stereotype red state politically-motivated rubes with a death wish.

[jennypenny]

You mean like Spanish flu ?

Does this happen to be related to the fact you’re a prepper?

Nah, I didn’t mean it like that. I meant that an influenza pandemic wouldn’t be so filled with unknowns. The uncertainty and unpredictability up to this point is what has bothered me.

I worked on a project for a while that dealt with this stuff, so I’ve read many of the books recommended upthread. Most of them focused on influenza because we’re familiar with it and it has (had?) the most potential for blooming into a pandemic. In those books, experts usually focused on influenza pandemics — where they start, how they spread, how fast people react. They rarely focused on the disease itself because we’re familiar with influenza.

At first I thought COVID was odd in that it didn’t seem to spread as fast as one would expect with a pandemic-level disease (geographically, not R rate), even with the lockdowns. The slow roll was different than any scenario I’d read about. Now that we know that many people are asymptomatic, maybe it wasn’t such a slow roll after all?(I wish the powers that be hadn’t F’d up the testing.) We’ll see on that one, I guess.

COVID also seems different to me because we’re still identifying symptoms six months out. Symptoms are also different depending on age. Most frightening to me is that some damage seems to occur silently in asymptomatic people. On a personal/unqualified-to-judge medical note ... it feels strange how many asymptomatic people there are considering how sick some people get, and some of the symptoms are bizarre.

I’m certainly not saying that medical professionals or modelers have been wrong. They are working with the information they have and the information keeps changing — still, at six months out, which is most of my point — so then models and recommendations change too.


As far as ‘I told you so’, I’ve avoided it mostly because I’ve tried to help people out so they’ve been pretty appreciative in return. I pre-staged preps at everyone’s house in my sphere, and I continue to make masks for all of them now that everyone is returning to work/life. (I’ll admit it’s partly selfish — if I want people to wear a mask around me, maybe I’ll guilt them into wearing one if I make it for them, like when you get an ugly sweater for christmas and only wear it around the person who gave it to you.)


tl;dr I wasn’t criticizing anyone. My point was that I would have been more comfortable/confident dealing with an influenza pandemic than this one. I don’t like unknowns and I don’t like the incongruous aspects of COVID.