Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Mithaldu
Sep 25, 2007

Let's cuddle. :3:
While the Nazis are gathering clout on Twitter by playing the clown ...


Here's two german studies on covid:

https://jamanetwork.com/journals/jamacardiology/fullarticle/2768916

https://jamanetwork.com/journals/jamacardiology/fullarticle/2768914

quote:

The first study, published Monday in the Journal of the American Medical Association Cardiology, looked at the cardiac MRIs of 100 relatively young patients who had recovered from COVID-19 and compared them to MRIs of 100 similar people who had not contracted the disease. Two-thirds of the patients recovered at home.

Two months following their recovery, 78 infected patients were found to have structural changes to their hearts. A biomarker indicating myocardial injury similar to that occurring in heart attacks was found in 76 patients. Sixty patients suffered inflammation of the heart.

The average age of the infected patients was 49. None had previous heart issues or other pre-existing conditions; in fact, many were skiers returning from vacation.

The second study, also published in JAMA Cardiology on Monday, examined autopsy reports from 39 people, 78 to 89 years old, who died in April. Analysis of cardiac tissue revealed the virus had infiltrated the hearts of 24 of the patients.

The studies’ results appeared to validate concerns expressed earlier this month by John Swartzberg, clinical professor emeritus of infectious diseases and vaccinology in the the UC Berkeley-UCSF Joint Medical Program.

“There is evidence now that the virus can directly attack heart muscle cells, and there’s also evidence that the cytokine storm that the virus triggers in the body not only damages the lungs, but can damage the heart,” Swartzberg told Berkeley News. “We don’t know what the long-term effects of that may be, but it could be that we will have a population of people who survive COVID-19 only to go on and have chronic cardiac problems.”

Among those problems are chronic heart failure, a progressive illness in which the heart slowly becomes less able to pump blood throughout the body.

Adbot
ADBOT LOVES YOU

Tarquinn
Jul 3, 2007


I know I’ve made some very poor decisions recently, but I can give you
my complete assurance that my work will be back to normal.
Hell Gem

AndreTheGiantBoned posted:

He mistook Fahr-Spur-Ende for Farhrspurende, as in Flüchtende or Lehrende

I... I couldn’t believe that was true.

skipThings
May 21, 2007

Tell me more about this
"Wireless fun-adaptor" you were speaking of.
I can't wait for the slow collapse of society and people's hearts just giving out randomly, just like they did metaphorically already, really poetic

no, I am not trying to be funny

Zwille
Aug 18, 2006

* For the Ghost Who Walks Funny
Yeah a lot of the long term Covid stuff sounds creepy as hell, like old wounds/scars/broken bones flaring back up and poo poo.

GABA ghoul
Oct 29, 2011

All this new information makes the US' decision to go full Plague Nation slightly questionable

DTurtle
Apr 10, 2011


MonikaTSarn posted:

I find this study questionable. How did they not mention at all the substandard living conditions the workers were in, or the cantina video ? Maybe workers on the same shift together were eating together as well, were being transported in to small vans to their homes together ?
They did look at the car pools and the living situation. The strongest correlation by far for early infection (by looking at mutations it was possible to somewhat figure out the timeline of infections) was proximity of the working station. The only significant correlation with regards to carpool and living situation also happened to be the same people who worked in close proximity together.

Researchers posted:

Positive rates were statistically significant only for a single shared apartment and associated carpool (a1 and c3), and a shared bedroom (r5). The fact that 5 of 7 members in a1/c3, and 2 out of 3 members in r5 have fixed work stations within the 8m area (Supplementary Table S3), however, suggests that high infection rates in these units primarily reflect the number of group members who work in close proximity of the index case, rather than resulting from independent infection chains within the units themselves.
In addition, they did mention the living conditions and the transportation:

Researchers posted:

Description of housing conditions, work area conditions, and working conditions
Housing conditions: Many of the workers share apartments and those usually commute together to their workplace in vans organized by the company. The company provided us with anonymized information about the housing situation of the workers regarding information on shared apartments, bedrooms and carpools. The largest housing unit encompassed seven workers for the initial outbreak in month 1 in the beef processing plant.
...
During breaks, the workers from a shift visit the canteen. Workers do not have fixed seats in the canteen. Supervisory staff does not spend the break times together with the production line workers. The supervisors do not share housing or transport facilities with the production line workers.

DTurtle fucked around with this message at 20:05 on Jul 28, 2020

Smirr
Jun 28, 2012

I don't want to spend too much time on this post, since I don't really want to spend time on statistics even when I'm getting paid for it, but: the methodology of that study is insanely noisy. I'm not the biggest fan of p-values to begin with, but I hate it when people report "significant" results when n = 3. Like, look at this table:



That's the split by apartment / carpool / etc. group. Look at r5, their one "significant" result. That's a sample size of three. Except there are repeated measurements (which makes everything a lot worse, although with a sample size of three it doesn't matter - it's all noise anyway). And you don't even know what the baseline for each group is, because:

quote:

Similarly, for each shared unit in Fig. 3C and Supplementary Table
S3 we calculated p-values for infection frequencies being equal or higher than observed among all
individuals who share one or more unit (22 out of 65 = 33.8%; please note that due to data protection
regulations we cannot reveal which worker IDs belong to a shared unit).

With sample sizes that low, and overlapping multiple comparisons (that they appear to not even have tried to adjust for), p-values are completely, absolutely meaningless. They always are, but in this specific example, they are straight garbage. My advice, take the raw data (which is mostly meaningless because of the data protection issues - you don't know which positive tests occurs twice or more times in which specific groups) and ignore any significance thresholds. The proximity data from Figure 3a is the one thing in this paper I would say is useful (but it's also something we already knew). Figure 3c is :gas: (3b too, but that's such horseshit I didn't even try to parse it)

aphid_licker
Jan 7, 2009


GABA ghoul posted:

All this new information makes the US' decision to go full Plague Nation slightly questionable

Millions of people trying to cope with their chronic damage under the conditions of the US healthcare system is going to drive the next wave of deaths of despair.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

aphid_licker posted:

Millions of people trying to cope with their chronic damage under the conditions of the US healthcare system is going to drive the next wave of deaths of despair.

Well, for the moment this is the wall the USA is driving straight at.

Not having a house.

Lord Stimperor
Jun 13, 2018

I'm a lovable meme.

Mithaldu posted:

Well, for the moment this is the wall the USA is driving straight at.

Not having a house.



Holy moly I see pitchforks in the future

niethan
Nov 22, 2005

Don't be scared, homie!
Can't vote without address, right?

Wipfmetz
Oct 12, 2007

Sitzen ein oder mehrere Wipfe in einer Lore, so kann man sie ueber den Rand der Lore hinausschauen sehen.

Mithaldu posted:

Well, for the moment this is the wall the USA is driving straight at.

Not having a house.
So "37% AK" means "37% of renter households in Alaska are at risk of eviction"?
What's the percentage of renter vs owner households in the US?

DTurtle
Apr 10, 2011


Smirr posted:

That's the split by apartment / carpool / etc. group. Look at r5, their one "significant" result. That's a sample size of three. Except there are repeated measurements (which makes everything a lot worse, although with a sample size of three it doesn't matter - it's all noise anyway). And you don't even know what the baseline for each group is, because:

With sample sizes that low, and overlapping multiple comparisons (that they appear to not even have tried to adjust for), p-values are completely, absolutely meaningless. They always are, but in this specific example, they are straight garbage. My advice, take the raw data (which is mostly meaningless because of the data protection issues - you don't know which positive tests occurs twice or more times in which specific groups) and ignore any significance thresholds. The proximity data from Figure 3a is the one thing in this paper I would say is useful (but it's also something we already knew). Figure 3c is :gas: (3b too, but that's such horseshit I didn't even try to parse it)
I agree that some of the parts of that analysis are silly - especially the the bedrooms (with 2 people in a room, no possible result could be "statistically significant"). As far as I can tell, that part was mainly about quickly checking if the living quarters, car pools, etc were significant.

However, 3B doesn't look that bad? Its a comparison of the actual distribution of infection vs distance in comparison to the expected distribution if distance didn't play a role.

Mithaldu posted:

Well, for the moment this is the wall the USA is driving straight at.

Not having a house.


Hmm, it would be interesting to know the "normal" numbers for that (pre-Corona). I can imagine it was already quite high.

Wipfmetz posted:

So "37% AK" means "37% of renter households in Alaska are at risk of eviction"?
What's the percentage of renter vs owner households in the US?
Roughly 2/3 owner, 1/3 renter.

DTurtle fucked around with this message at 22:18 on Jul 28, 2020

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

Wipfmetz posted:

So "37% AK" means "37% of renter households in Alaska are at risk of eviction"?
What's the percentage of renter vs owner households in the US?

Good question.

26.83% - 46.2% according to https://www.move.org/states-with-highest-lowest-owner-occupied-homes/

With a complete Milchmädchenrechnung this gives me 16% of the US population threatened by eviction.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

DTurtle posted:

Hmm, it would be interesting to know the "normal" numbers for that (pre-Corona). I can imagine it was already quite high.

Roughly 2/3 owner, 1/3 renter.

https://www.cnbc.com/2020/07/27/how-the-eviction-crisis-will-impact-each-state.html

quote:

In 2016, there were 2.3 million evictions, Pollock said. “There could be that many evictions in August,” he said.

Which, if you guesstimate 4 people per household makes 9.2 million people evicted, which would be 2.5% of the population.

Things definitely look worse in that picture than in real numbers i think.

DTurtle
Apr 10, 2011


Hmm, looking here:
https://www.nytimes.com/interactive/2018/04/07/upshot/millions-of-eviction-records-a-sweeping-new-look-at-housing-in-america.html



The highest cities they had there had a eviction filing right of more than 30% and an actual eviction judgment rate of a bit more than 10%.

GABA ghoul
Oct 29, 2011


Wanna hold every smug idiot's face from 2016 who yammered about how it doesn't matter who is president or who controls Congress(~they are all the same~) into that huge pile of poo poo.

Probably still dreaming about that sweet gay space communism that they'll get through accelerationism when the much more likely outcome is a decade of semi-apocalypse and then a pivot to open mega-authoritarianism, ultra-corruption and hyper-capitalism like Russia in the 90s. Or even just straight up decent into some hosed up nationalist totalitarianism à la Weimar.

oliwan
Jul 20, 2005

by Nyc_Tattoo

GABA ghoul posted:

Wanna hold every smug idiot's face from 2016 who yammered about how it doesn't matter who is president or who controls Congress(~they are all the same~) into that huge pile of poo poo.

Probably still dreaming about that sweet gay space communism that they'll get through accelerationism when the much more likely outcome is a decade of semi-apocalypse and then a pivot to open mega-authoritarianism, ultra-corruption and hyper-capitalism like Russia in the 90s. Or even just straight up decent into some hosed up nationalist totalitarianism à la Weimar.

lots of posters in c spam who refuse to vote for biden

Healbot
Jul 7, 2006

very very very fucjable
very vywr very


GABA ghoul posted:

Wanna hold every smug idiot's face from 2016 who yammered about how it doesn't matter who is president or who controls Congress(~they are all the same~) into that huge pile of poo poo.

Probably still dreaming about that sweet gay space communism that they'll get through accelerationism when the much more likely outcome is a decade of semi-apocalypse and then a pivot to open mega-authoritarianism, ultra-corruption and hyper-capitalism like Russia in the 90s. Or even just straight up decent into some hosed up nationalist totalitarianism á la Weimar.

Take a good look at what the dems are doing right now and say that again lol. The functional difference would've been an earlier moratorium reporting on the protests.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

GABA ghoul posted:

Wanna hold every smug idiot's face from 2016 who yammered about how it doesn't matter who is president or who controls Congress(~they are all the same~) into that huge pile of poo poo.

Probably still dreaming about that sweet gay space communism that they'll get through accelerationism when the much more likely outcome is a decade of semi-apocalypse and then a pivot to open mega-authoritarianism, ultra-corruption and hyper-capitalism like Russia in the 90s. Or even just straight up decent into some hosed up nationalist totalitarianism à la Weimar.


oliwan posted:

lots of posters in c spam who refuse to vote for biden

Meanwhile some lefties online:

https://twitter.com/HappyShiiba/status/1288060070115176449

Mr. Lobe
Feb 23, 2007

... Dry bones...


Name an evil thing the US does. I guarantee you it will continue under Biden, only with more crocodile tears.

AndreTheGiantBoned
Oct 28, 2010
I compared with Germany, in which there are 40 million Haushalte, 47% of which are rented, and read a piece of news where it says that 1.6 million are facing this issue. So about 8%. Lol usa

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

Mr. Lobe posted:

Name an evil thing the US does. I guarantee you it will continue under Biden, only with more crocodile tears.

The USA tried to gently caress around with the korean peninsula in massively inept ways. Wouldn't have happened under Biden.

The USA worsened relationships with Cuba. Wouldn't have happened under Biden.

The USA betrayed the Kurds and singlehandedly kickstarted a genocide. That absolutely wouldn't have happened under Biden.

(And i sure hope you wouldn't be the type of person to claim that what the USA did to the Kurds wasn't a massive crime.)

If you're curious about such things, hit up @DylanBurns1776.

Mithaldu fucked around with this message at 00:36 on Jul 29, 2020

elbkaida
Jan 13, 2008
Look!
Note that it would've been Clinton, not Biden, in your examples. Also I think you shouldn't underestimate the ability of any US president to be completely terrible when it comes to foreign policy.

Healbot
Jul 7, 2006

very very very fucjable
very vywr very


If history had started in 2016 for you this might’ve been surprising, but strategic allies stop being allies when the strategy changes.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

elbkaida posted:

Note that it would've been Clinton, not Biden, in your examples. Also I think you shouldn't underestimate the ability of any US president to be completely terrible when it comes to foreign policy.
We're comparing Trump and Biden here.

Yes, and theoretically i could win the lottery too. Similar probabilities regarding these specific cases.

Healbot posted:

If history had started in 2016 for you this might’ve been surprising, but strategic allies stop being allies when the strategy changes.
Is this your opinion on the Rojava situation?

Torrannor
Apr 27, 2013

---FAGNER---
TEAM-MATE

Mithaldu posted:

Is this your opinion on the Rojava situation?

Isn't it accurate? The previous strategy was to contain Assad/Russia by keeping the Kurds semi-independent. Trump loves Russia so the strategy changed, and the Kurds ceased to be strategic allies, so they were left to fend for themselves. I don't think that would have happened under Clinton.

This nicely sidesteps the issue whether Clinton would have been fine with genocide in general.

GABA ghoul
Oct 29, 2011

Mr. Lobe posted:

Name an evil thing the US does. I guarantee you it will continue under Biden, only with more crocodile tears.

And if I do, will you stop doing that bit? Probably not, so what's the point?

e: eh, what the hell: Taking the second largest CO2 emitter in the world out of the Paris Climate Agreement

Smirr
Jun 28, 2012

DTurtle posted:

However, 3B doesn't look that bad? Its a comparison of the actual distribution of infection vs distance in comparison to the expected distribution if distance didn't play a role.

My side beef is "-log10(p-val)" (also in graph 3c) - as far as I can tell, this serves only to turn large downward outliers into even larger upward outliers optically. This might be common practice in their fied, but I find it extremely :nallears:

My main beef is their use of p-values and/or an extremely uninteresting null hypothesis (we know that having someone with covid cough within 1 m of you is more dangerous than within 100 m of you - yes, at some point you do need to show this, but the null hypothesis that distance does not play a role is refutable by eyeball by looking at graph 3a, and by common sense). Specifically, they create the dangerous (and false) at-a-glance takeaway that, since the results for 2-4 m distance are not significant, that those distances are not dangerous. They are, though, it's just that n = 1 for a distance of 2 m. I don't have an issue with the upper part of graph 3b - it shows that covid rates are higher than expected at closer distances. That's interesting information. The lower part of graph 3b only adds the additional information for which distances they have large enough sample sizes for those results to be "significant". But people who are not that used to handling p-values (journalists, for example, or scientists) might see the "significant" arrows and think "most dangerous part".

And yes, I'm aware that many researchers jump through p-value hoops because editors ask for them, but this is a pre-print and the authors appear pretty convinced by what they're doing.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:

Torrannor posted:

Isn't it accurate? The previous strategy was to contain Assad/Russia by keeping the Kurds semi-independent. Trump loves Russia so the strategy changed, and the Kurds ceased to be strategic allies, so they were left to fend for themselves. I don't think that would have happened under Clinton.

This nicely sidesteps the issue whether Clinton would have been fine with genocide in general.
I'm not talking about Clinton. Already said so. And Biden specifically has a pro-kurd track record.

And as for Healbot's post, they're vague-posting, so it's really really hard to tell what they actually mean, but to respond to your interpretation:

I seriously doubt calling what trump was doing a strategy is merited. Also uh, Turkey and Assad were the ones being contained. Russia actually somewhat stepped in to dampen the worst after the Trump decided keeping 50 soldiers in key locations was too much effort.

GABA ghoul posted:

And if I do, will you stop doing that bit? Probably not, so what's the point?

e: eh, what the hell: Taking the second largest CO2 emitter in the world out of the Paris Climate Agreement
Also true.

Antigravitas
Dec 8, 2019

Die Rettung fuer die Landwirte:
I'm sure a democrat would've turned 180°, colluded with Iranian hard liners, and scuppered the JCPOA they just negotiated. 100% sure.

genericnick
Dec 26, 2012

Mithaldu posted:

The USA tried to gently caress around with the korean peninsula in massively inept ways. Wouldn't have happened under Biden.

The USA worsened relationships with Cuba. Wouldn't have happened under Biden.

The USA betrayed the Kurds and singlehandedly kickstarted a genocide. That absolutely wouldn't have happened under Biden.

(And i sure hope you wouldn't be the type of person to claim that what the USA did to the Kurds wasn't a massive crime.)

If you're curious about such things, hit up @DylanBurns1776.

Not sure why we are doing this in the Germany thread but loving around in inept ways in the Korean peninsula and betraying the Kurds is basically in the job description. Those are some of the worst examples you could pick, because not doing those things would be a departure from longstanding US policy. And what was the supposed problem in Korea? The South tried to get a peace initiative going and Trump got a few photo ops in. Doesn't seem to have gone anywhere, but no one is really worse off for it than before the start of history (2016) and treating it as in anyway equal to the Kurdish betrayal is weird to me. You could have brought up Iran though. That escalation probably wouldn't have happened under Clinton or Biden. Probably, because hostility towards Iran is extremely ingrained in the US security apparatus. Biden also probably wouldn't have advised to drink bleach.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:
Gotta admit i have the feeling you straightup just didn't grok the flow of the convo. That first sentence alone is just ... nonsensical on its own AND in context. And pretending i treated it as equal is just ... feels like you had a conclusion about my post you wanted to be true really hard.

--

In any case, people who think all of the bad stuff trump did will also be done by biden and worse ... they're big 'ol dummies.

genericnick
Dec 26, 2012

Mithaldu posted:

And pretending i treated it as equal is just ... feels like you had a conclusion about my post you wanted to be true really hard.

Yeah, sorry that was uncalled for and I blame the lack of coffee. But the Korea situation really shouldn't be brought up in a line with the reescalation against Cuba and Iran, and the betrayal of the Kurds. Or the genocide in Yemen that is somehow never brought up. The discourse on this has been utterly deranged. Trump's initial support of the ROK's is literally the only thing in his whole presidency I would give him some limited credit for. (Credit for not getting in the way, but of course as soon as he forgot about it the US went back into sabotage mode.) Yet by the way the Trump-Kim photo op is talked about it's the Molotov-Ribbentrop pact.

Mithaldu
Sep 25, 2007

Let's cuddle. :3:
Thanks. :)

And in general i agree there, my post mainly was that way because i'm not universally educated in those matters, and sorted some i thought of off the top of my head roughly from dumbest to horriblest.

suck my woke dick
Oct 10, 2012

:siren:I CANNOT EJACULATE WITHOUT SEEING NATIVE AMERICANS BRUTALISED!:siren:

Put this cum-loving slave on ignore immediately!

oliwan posted:

lots of posters in c spam who refuse to vote for biden

counterpoint: biden would be doing the same poo poo domestically (but rachel maddow would twist herself into knots explaining how it's better because he's not controlled by the perfidious russkies), and in terms of foreign policy he'd probably actually be worse because he'd be trying to rebuild us hegemony instead of destroying it as rapidly as possible like trump

Peggotty
May 9, 2014

Counter-counterpoint: When your Americanized "leftism" makes you defend Donald Trump you should probably consume less internet for a while.

Venomous
Nov 7, 2011





Biden is garbage + if he hadn't run and the conservative Democrats hadn't gone Voltron, we'd be looking at President Bernie Sanders in November

Wipfmetz
Oct 12, 2007

Sitzen ein oder mehrere Wipfe in einer Lore, so kann man sie ueber den Rand der Lore hinausschauen sehen.

Mr. Lobe posted:

Name an evil thing the US does. I guarantee you it will continue under Biden, only with more crocodile tears.
Maybe. But a lesson I've taken from the Trump years is, that the masquerate and crocodile tears do matter.
Because they influence what's considered "normal" or "acceptable", and it does induce the cost of maintaining excuses to the "evil things".

Wipfmetz fucked around with this message at 18:10 on Jul 29, 2020

Adbot
ADBOT LOVES YOU

DTurtle
Apr 10, 2011


Sounds like a fair take.

For me the most important part of that paper is simply the evidence that there is nothing magical about 2m (or 1.5m in the UK, I think). Depending on the circumstances, Sars-Cov-2 will travel with the air flow, spread with the air flow, and infect with the air flow - even over larger distances.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply