Mailman,
Yes it is, but hardly worth losing sleep over. I find a more satisfactory way of losing sleep is spending hours watching those @SuperfastMatt videos. Rather addictive.

Robbo,

Its all a but of a shit show eh.

Mailman,
I prefer your reasoning to that in the article. Yours is, like mine was back at the time, based on seat-of-the-pants/common sense. The article used IMO bogus stats to give it a sciency feel of authority. Just because I favour its conclusions doesn't mean I should consider it helpful. Touting it would weaken our position. if I can pick holes in it, so can a PCR-test defender.

To put it in terms of your metaphor: it's a rotten apple; bad idea to put it in the barrel with the good ones.

tomo,
I remember Feynman describing his "they just don't get it" exasperation at psychologists' reverence for published scientific papers. Once a result was out there, it was *true*, and the last thing you would want to do is go over the old ground. Feynman's response was (more or less) that physicists *had* *to* repeat old experiments: How can you advance if you don't start by getting to the current position?

30 years since that book was published and the psychologists are winning. Shouldn't be a surprise: Feynmans have always been very thin on the ground, and the academies have set their bars so very low.

Also, does anyone know if the Government has found itself not guilty in its Chinese Covid enquiry yet?

Robbo,

I fear you are making a "Red and Green apples taste different, therefore both cant be apples" kind of argument. Whether its exactly 86.5% or whether there should be some variance or discrepancy or what not noted is neither here nor there. The important factor is these tests were about as useful as c0ckflavoured lollipops and should NEVER have been used for the intended reasons they were used.

Anyone capable of rubbing two or more functioning brain cells together at the time already knew the tests were pointless. Just a pity so many people were more in the 'f8ck em and load em in to the cattle carts for shipping off for education because they did not do as they were told".

Mailman

tomo,
You're only meant to read the headline. It's not cricket to go looking under the covers.

Sadly, that also applies to things we might tend to agree with. This article caught my eye (linked in a comment at Jo Nova's today). It looks into recent disclosures from the German public health people at the Robert Koch Institute, and makes some statements you and I would find very believable. A couple of examples:

The German government’s reactions to the SARS-CoV-2 pandemic had nothing to do with science.
...
The PCR tests were false positive in 86.5% of all those who tested positive and measures imposed on the basis of these tests had no scientific basis.

I think the article makes a fair case for that first statement, but the second looks far too sciency for me to trust it. It's a pity, because I argued the point here at length with TinyCO2 on the need to calibrate PCR to actual infections and the article would be vindication.

You might remember that my point was that a test at the back of the throat is external and can't possibly be conclusive that the virus has gained entry. Secondary argument was that it only tests for material, not whether the organism is active.

This neat 86.5% figure, with no statement on variability doesn't go very well with my seat-of-the-pants reasoning, so I looked at what they'd done to come up with it. Looks like the familiar voodoo statistics of climate science.

It's not clear to me, but I think RKI did followup IgG blood tests on people who had had positive PCR tests. Excellent. Apparently they stopped doing IgG tests half-way through '21. To calibrate PCR to IgG, our researchers have taken cumulative figures of each type of test, then come up with a linear regression to adjust the large number of PCR positives down to the the IgG positives. Result: IgG=PCR×0.135+2. Magic!

That 2 looks a bit odd though, and the explanation in the text — the 2-week delay for IgG to turn positive — is fit for a used-car salesman. The y-axis is percentage, and the 2 gives you a fixed 2% infection base, which makes no sense at all.

Bigger question: why use *cumulative* figures?

For starters, I'm not at all clear what they're accumulating. There's only a percentages label on the Y axis, but it's meaningless to add yesterday's 5% positives of 200 tests to today's 3% positives of 180 tests to make 8% positives... You could add numerators and denominators to get a percentage so-far, but the resulting graph would gravitate towards the average percentage, not ramp up as their graph does.

But why cumulative figures anyway? They've plotted the raw figures for PCR (in the darker blue); why don't we see a similar graph for IgG? If their conclusion is right, it should be a scaled down and (apparently) 2-week delayed picture of the PCR figures. I can only guess that they tried with the day-by-day figures, but found they didn't give a compelling graph.

And if such a graph doesn't leap out at you, I'm going with Rutherford: better "experiment" needed. Can we put out a general plea for more brains and less computers?

$11 billion high-speed rail bridge to nowhere that took 9 years to build