OT: Brady caught in closed Tampa park

[Ross12]

If you "re-open" when there are 5,000 active infections, it's a lot different than "re-opening" when there are 100 active infections. (Just pulling numbers here to illustrate.) Exponential growth happens a lot faster from the former case, and so you end up with overwhelmed hospitals and the need to "re-close" for another lengthy period of time.

The point of the stay-home isn't just to avoid overloading the hospitals.

It's also taking into factor that we don't have vaccine, and we still don't have a proven treatment for this virus.

In an ideal situation (or country), the stay-home restrictions would be relaxed if we had one or both of the above.

My friend's doctor is a well known specialist in infectious diseases and is on the MA coronavirus task force for the Boston area hospitals and she told him that there's still so much we don't know about this virus other than that it's mutating. She told him that the idea that this virus is being contained or managed, is total ********.

Sorry guys - when I say "re-open" I do mean relaxing restrictions, not a "switch-flipping" situation.

What interests me is the anti-body tests - the ones I have seen so far seem to show that a much larger percent of the population, at least in more densely populated areas, appear to have had this virus but never knew. I think the data from Stanford yesterday was something like 28 to 55 times higher.

Not sure if this is getting "off-topic" or not since it relates to the article at hand...

 

[oldrover]

The point of the stay-home isn't just to avoid overloading the hospitals.

Actually, yes, it is. It's called flattening the curve... we did it... and now it's time to get back out there.

Also, that virus has infected way more people than we thought... 28 to 55 times more, as @Ross12 said, and guess what... they're not presenting with symptoms... so the mortality rate is way lower than we thought too. Win win.

______

Just saying my piece...




...before this thread inevitably implodes.

 

[primetime]

Sorry guys - when I say "re-open" I do mean relaxing restrictions, not a "switch-flipping" situation.

What interests me is the anti-body tests - the ones I have seen so far seem to show that a much larger percent of the population, at least in more densely populated areas, appear to have had this virus but never knew. I think the data from Stanford yesterday was something like 28 to 55 times higher.

Not sure if this is getting "off-topic" or not since it relates to the article at hand...

There's much more we don't know than what we do. We don't know the true rate of infection and we don't even know if antibodies provide immunity, or for how long if they do. For seasonal coronavirus it's less than a year, for SARS it was three to four years.

 

[sb1]

I think the absolute lockdown will be relaxed soon but distancing will be with us for a while. I think it can be managed in restaurants and stores but concerts and sports will be tricky.

 

[Ross12]

There's much more we don't know than what we do. We don't know the true rate of infection and we don't even know if antibodies provide immunity, or for how long if they do. For seasonal coronavirus it's less than a year, for SARS it was three to four years.

Hopefully that can be tested. If the true number is somewhere in between (say, 30x more infected than reported) and if the antibodies provide even 6 months of immunity, that makes your mortality rate comparable to influenza as I understand it and it wouldn't seem reasonable to me to keep things completely shut down.

 

[primetime]

Actually, yes, it is. It's called flattening the curve... we did it... and now it's time to get back out there.

Also, that virus has infected way more people than we thought... 28 to 55 times more, as @Ross12 said, and guess what... they're not presenting with symptoms... so the mortality rate is way lower than we thought too. Win win.

______

Just saying my piece...




...before this thread inevitably implodes.

That study hasn't been peer-reviewed. I'd be careful interpreting the results, because they don't make a ton of sense and there were some issues with weighting the sample and the statistical methods involved. In raw numbers, only 50 out of the 3,000 people in the Stanford study had positive antibody tests (1.6% of the sample). On top of that, the antibody testing kits they used are also brand new and were rolled out quickly from China and were not approved by China's FDA equivalent, so they may not even work. Other serological testing has been all over the place, from 1% of a Scottish sample to 15% of a single German town where they tested everyone and, again, all using kits that haven't been thoroughly vetted.

 

[oldrover]

That study hasn't been peer-reviewed.

So, how do we know social distancing has slowed the virus from spreading? That hasn't been peer-reviewed either, nor does it have a control group.

Answer: we DON'T know it's slowed the virus. The virus could have spread just as quickly if we'd all kept going to work. We have no idea.

 

[IllegalContact]

Who the f*ck is Tom Brady?

f*ck that dude, he sucks

 

[Ross12]

So, how do we know social distancing has slowed the virus from spreading? That hasn't been peer-reviewed either, nor does it have a control group.

Answer: we DON'T know it's slowed the virus. The virus could have spread just as quickly if we'd all kept going to work. We have no idea.

Read an interesting article last night about Sweden who seemed to do just that.

 

[primetime]

So, how do we know social distancing has slowed the virus from spreading? That hasn't been peer-reviewed either, nor does it have a control group.

Answer: we DON'T know it's slowed the virus. The virus could have spread just as quickly if we'd all kept going to work. We have no idea.

How do we know how, like, anything works, maaaan?

Because isolation and quarantine are a proven method of reducing contagion? I mean, the virus doesn't just magically appear in the air. Coronaviruses aren't a new thing, we generally understand how they work.

Andrew Gelman's one of the great statisticians of the world and he's been keeping tabs on coronavirus research on his blog, where he often picks apart studies looking for methodological question marks. I recommend reading it, he has some thoughts on the antibody papers:

Statistical Modeling, Causal Inference, and Social Science

 

[FreeTedWilliams]

The one thing this whole thing has showed us is the speed in which the government will take your civil liberties away! Imagine the insanity of someone walking up to a person in a park (which is open space) and telling them they have to leave because of some arbitrary unconstitutional order.

Now think about how stupid it is to do that to Tom Brady! I could see if he had gathered a crowd or had a bunch of guys with him, but he didn't he was most likley just out for a run.

INSANITY!! This can't end fast enough

 

[Tony2046]

That study hasn't been peer-reviewed. I'd be careful interpreting the results, because they don't make a ton of sense and there were some issues with weighting the sample and the statistical methods involved. In raw numbers, only 50 out of the 3,000 people in the Stanford study had positive antibody tests (1.6% of the sample). On top of that, the antibody testing kits they used are also brand new and were rolled out quickly from China and were not approved by China's FDA equivalent, so they may not even work. Other serological testing has been all over the place, from 1% of a Scottish sample to 15% of a single German town where they tested everyone and, again, all using kits that haven't been thoroughly vetted.

One of the criticisms I heard was that the study asked for volunteers rather than picking random people.

Aside from that, I'm for slowly opening things up sooner rather than later but it does seem we're at least two weeks away from a more appropriate time to do so.

 

[oldrover]

How do we know how, like, anything works, maaaan?

Because isolation and quarantine are a proven method of reducing contagion? I mean, the virus doesn't just magically appear in the air. Coronaviruses aren't a new thing, we generally understand how they work.

Show me the double-blind peer review studies that prove it. Not statistical modeling. You can argue both sides of a story with statistics. And look how often those have been wrong in the last month alone.

 

[primetime]

One of the criticisms I heard was that the study asked for volunteers rather than picking random people.

Aside from that, I'm for slowly opening things up sooner rather than later but it does seem we're at least two weeks away from a more appropriate time to do so.

Gelman's post here is useful: Concerns with that Stanford study of coronavirus prevalence « Statistical Modeling, Causal Inference, and Social Science

There's a lot of potential issues - false positives in tests, selection bias, and the fact that Santa Clara County where they conducted the study doesn't have all that many positives to begin with. So even if we accept those numbers as plausible, it's reasonable to think that many more people have it in Santa Clara County, but less plausible in a place like New York where the numbers they're talking about would mean 3/4 of the city was already infected.

@Ross12 mentioned Sweden, good Gelman post on Sweden here too: Coronavirus in Sweden, what’s the story? « Statistical Modeling, Causal Inference, and Social Science

 

[PatsBoy12]

Once a cheater, always a cheater.



The Pats will probably lose their 2021 first rounder for this though.

 

[primetime]

Show me the double-blind peer review studies that prove it. Not statistical modeling. You can argue both sides of a story with statistics. And look how often those have been wrong in the last month alone.

How would you even do a "double-blind study" on social distancing? You're just regurgitating words you don't know or understand right now.

That Stanford study you're quoting is... a statistical model. They try to account for test specificity using best guesses and then they weight based on demographics. All those are statistical choices. There's no blinding or experimental design there. It's just a cross-sectional study with statistical techniques used to attempt to correct for sampling and measurement error.

 

[Simpelton]

And while you're busy arguing both sides of the story you have a choice to make. Are you going to choose to be reckless, taking the chance that it'll be OK and risking the massive consequences of being wrong, or are you going to moderately inconvenience people a few weeks longer until you're sure.

I know which option I prefer

 

[Ross12]

One of the criticisms I heard was that the study asked for volunteers rather than picking random people.

Aside from that, I'm for slowly opening things up sooner rather than later but it does seem we're at least two weeks away from a more appropriate time to do so.

The Stanford study reached out to individuals based on a statistical model, as I understand it.

 

[Ross12]

And while you're busy arguing both sides of the story you have a choice to make. Are you going to choose to be reckless, taking the chance that it'll be OK and risking the massive consequences of being wrong, or are you going to moderately inconvenience people a few weeks longer until you're sure.

I know which option I prefer

But will we be "sure" in 2 or 3 weeks? That's the question.

I don't think many of us would argue against a May 1st/May 8th target date. But I don't know what is necessarily going to change much between now and then. My hope is more reliable antibody studies.

 

[Teej]

The Diva probably hoped he'd be filmed and praised like he was up here during his suspension but it just shows he's tone deaf and thinks the rules don't apply to him.