Can You Spot the Flaw?

[Mick]

Originally Posted by FLEXjs

Same as in Canada. They are not testing everyone. They are telling people if you have symptoms stay home and self-isolate.

So they are only testing the people that require critical care. Death is therefore way artificially over-inflated.

Why are there only a couple of hundred thousand positive tests for seasonal flu in the US every year when we know that millions of people get it? Because we do the same thing. If you are sick as a dog, and go to your doctor, he will tell you you have the flu, and tell you to go home and stay there until you are well. No test.

We only test you if we need to admit you to the hospital for intensive treatment, and the doctors want to be sure it IS the flu they are up against.

This is why the comparison I laid out above, although certainly flawed in its own right, is the best one we have. In the 2018-2019 flu season, 19% of those deemed sick enough to justify a test ultimately died of the flu. So far, of the people with sufficient symptoms to justify a coronavirus test, only 6% of them have died. This number is known to be overstated because the CDC is knowingly adding deaths of people who didn't die of COVID-19, but it is also likely to rise somewhat as we reach the end of flu season so fewer cases get added, but some of those who got sick a few weeks ago will ultimately succumb to the disease.

[Aerovette]

Originally Posted by FLEXjs

Same as in Canada. They are not testing everyone. They are telling people if you have symptoms stay home and self-isolate.

So they are only testing the people that require critical care. Death is therefore way artificially over-inflated.

Somehow they believe they can take the number of DEATHS from the number of CASES...and use that to project impact on the entire population. That's the biggest pile of shit ever.

[dvarapala]

Originally Posted by Mick

Why are there only a couple of hundred thousand positive tests for seasonal flu in the US every year when we know that millions of people get it? Because we do the same thing.

So from a purely mathematical standpoint, how many people would need to be tested in order to make a statistically sound extrapolation of the infection rate to the entire population? And would those tests needs to be random, as opposed to the self-selected subjects who have been tested thus far?

[Aerovette]

Originally Posted by dvarapala

So from a purely mathematical standpoint, how many people would need to be tested in order to make a statistically sound extrapolation of the infection rate to the entire population? And would those tests needs to be random, as opposed to the self-selected subjects who have been tested thus far?

I think 20% of the population coast to coast ought to provide very meaningful data. By percentage of population for each state.

[Mick]

Originally Posted by dvarapala

So from a purely mathematical standpoint, how many people would need to be tested in order to make a statistically sound extrapolation of the infection rate to the entire population? And would those tests needs to be random, as opposed to the self-selected subjects who have been tested thus far?

"From a purely mathematical standpoint" huh?

OK, statistics, when done correctly, IS an exact science. If you sample correctly (meaning making all efforts to remove or correct for sample bias and skew), relatively tiny samples of large populations can provide statistically meaningful answers. The larger your sample, the smaller your confidence intervals get.

[dvarapala]

Originally Posted by Mick

"From a purely mathematical standpoint" huh?

OK, statistics, when done correctly, IS an exact science. If you sample correctly (meaning making all efforts to remove or correct for sample bias and skew), relatively tiny samples of large populations can provide statistically meaningful answers. The larger your sample, the smaller your confidence intervals get.

And have we done those things here in America with our COVID testing?

[Aerovette]

Originally Posted by dvarapala

And have we done those things here in America with our COVID testing?

No. We keep going to the cemetery and checking bodies to see how many people in the nation have a pulse.

Testing sick people to see if they are sick is downright idiotic.

[Mick]

Originally Posted by dvarapala

And have we done those things here in America with our COVID testing?

Is this a serious question?

I have lost confidence that any number I read on the virus has ANY basis in "science" or meets even the loosest definition of "data" I can imagine.

Let's talk about "Total Cases" for a minute. I learned this week that there apparently exists two different types of tests, a "screening" test and an actual test, that have known differences in the percentage of false positives and false negatives. Have we kept track of the Total Cases data by type of test used to determine that case? Apparently not. Have we further mucked up the "data" be knowingly and purposefully including "presumed cases" where some dude had a fever and coughed once so we threw him in there with other "confirmed cases"? Yup.

At the end of the day, mathematically speaking, we don't have any "data" at all. We have a collection of wild ass guesses. AFAIK, there is no mathematical procedure for calculating confidence intervals on a pile of wild ass guesses, so all these politicians saying their "reopening plans" will be "science based", only goes to show how clueless they, and those advising them, really are.

[Aerovette]

Mick,the tactic changed from going down the wrong path (which is too easily picked apart), to going down 30 paths and spitting out information like an Enron shredder, so no one is ever clear on anything.

[dvarapala]

Originally Posted by Mick

What I think the author is trying to point out is exactly what I have been saying all along: that Erickson's 12% infection rate is nothing more than a WILD ASSED GUESS.

I think he's also trying to say that the sampling is biased, as pretty much everyone who has been tested up to this point has either been symptomatic or exposed to a known case.

[Mick]

Originally Posted by dvarapala

I think he's also trying to say that the sampling is biased, as pretty much everyone who has been tested up to this point has either been symptomatic or exposed to a known case.

Of course the sampling is biased, that's kind of the point. The sampling of the testing for seasonal flu is also heavily biased, and in the same way, we don't test you unless you might need serious medical intervention. This is why I have been saying the proper comparator number for mortality isn't the 0.1% the media loves so much, but the 19% of those tested (which includes both positive and negative test results), who ended up dying of seasonal flu in 2018-2019.

[dvarapala]

[Mick]

Originally Posted by dvarapala

Another 43 minute video, where I watched the first 6 minutes, and he still hadn't said anything. Ain't nobody got time fo' dat.

[Aerovette]

HCQ - Used for malaria, good results, miracle recovery - Fauci Verdict: Anecdotal and unproven.

Remdesivir - Used for Ebola, good result, miracle recovery - Fauci verdict: Ethical responsibility to fast track.

There's a head scratcher...but not really.

[dvarapala]

Originally Posted by Mick

Another 43 minute video, where I watched the first 6 minutes, and he still hadn't said anything. Ain't nobody got time fo' dat.

1) Use the 2x speed option.
2) You don't need to watch the entire thing to get the gist. He's basically calling bullshit on their sloppy math and their obvious biases (they own a chain of health clinics and their business is hemorrhaging cash, so of course they want everything to open back up again).

[Aerovette]

I'm going to state the obvious because so what?

On a micro scale, we know that fake screen names get created and are used to get around the system or are used sacrificially to say things that aren't allowed.
In a bigger picture, I have no doubt that the same isn't done on Facebook to pump an agenda, to push false information, to support an unpopular perspective until it becomes popular. The moment those doctors put that video out, there is NO DOUBT that a mad scramble ensued, to pick it apart, pick them apart, find an alternate motivation...because God forbid someone just wanted to provide information without any self-interest. Google, Youtube, Facebook, hell, even Match.com has fakery to obtain a goal.

I agree with the those doctor's regardless of their motivation. Much like I feel about Trump...is he getting the job done? Is he doing the right things? If so, I don't really care WHY?

[Mick]

Originally Posted by dvarapala

1) Use the 2x speed option.
2) You don't need to watch the entire thing to get the gist. He's basically calling bullshit on their sloppy math and their obvious biases (they own a chain of health clinics and their business is hemorrhaging cash, so of course they want everything to open back up again).

Ain't nobody got 22 minutes to hear some other guy's line of bullshit. The docs in the original video explained where they got their estimates. If you don't like them, fine with me.

None of that changes the reality that we did this whole "lockdown" to "flatten the curve" so the healthcare system would not get overwhelmed. Now, the healthcare system is so dramatically underwhelmed, that it is basically going out of business. What are we going to do next winter/early spring when the next wave hits, and all our healthcare businesses are closed?

[Aerovette]

This has been a series of moving the goal posts because the trigger got pulled too early. Trump has time to fix the economy so the Dems have to slow it all down any way possible.