update: before you let someone try to scare you with US case counts, please remember this:

"reporting case counts without reference to testing levels is tantamount to lying."

US cases are not 142% of april peak and rising.

they are 17% of april peak and at the lows.

it can be easier to see as % of april peak

this is such a complete inversion of the popular narrative as to render most of current public discourse here absurd.

mistaking a rise in sample rate for a rise in prevalence is a mistake no first year stats student would make.

it can, at this point, only be willful, especially from arch-karens and political water-carriers like ding and krugman.

they 1000% know better.

but THIS is looking a bit ominous. this looks like some sort of positivity floor.

epidemiology does gompertz, not straight lines.

what DOES do straight lines is probabilities iterated over a large data set, like, oh, say a false positive rate on a test being applied a million times a day.

we know the PCR tests are massively over sensitive.

probably 1000X over sensitive. https://twitter.com/boriquagato/status/1299748995372384257?s=20

they are detecting trace virus in recovered people and in those who never had any clinical infection whatsoever.

at Ct=40, 1 virus becomes a trillion.

and false positive rates go WAY up.

we're currently doing ~1 million tests a day in the US.

every point of false + is 10k cases and 77 deaths a day.

based on this impossibly flat deaths data from IL (that i happened to have handy) this false + rate looks to be around 7% in that locale.

this likely sounds like a crazy number, but it isn't.

consider: people who die of most illnesses and pathologies are highly vulnerable to opportunistic infection.

many wind up in the hospital.

is it such a stretch that covid might move in, even just a little?

you get even a few copies of this in your blood before you die, and PCR at 40 Ct is gonna find it.

this 7-7.5% rate of positivity for all deaths in IL looks pretty low when you consider that and add in picking up traces for 12 weeks in those who recovered.

it would also mean that the false (non-clinical) positive rate in deaths could be much higher than in cases.

that divergence gets interesting.

i'm not clear how far we can extrapolate that, and i doubt the US is that high, but even half that rate, 35k cases a day in the US are false. that's 70-80% (and similar to claims the NYT made) https://www.nytimes.com/2020/08/29/health/coronavirus-testing.html

it also implies that roughly 40% of US deaths daily are false (assuming 3.5% false + rate, all deaths tested).

at 7%, 80% of deaths are false +'s.

this is obviously a very rough approximation, but honestly, i suspect it's not wildly off on state data (see IL above)

what if we are flat out not measuring real disease anymore.

what if it's 80% false positives that were not and possibly never were clinical cases.

honestly, this is starting to look like the odds on bet to me now.

there is a reason the FDA and drug/vaccine companies do not accept a positive PCR tests as conformation of a covid case.

they know this test is ridiculous and over-sensitive.

when you ask real scientists do do real science, they do this: https://twitter.com/boriquagato/status/1308451578869809152?s=20

when you ask government and the covid-industrial complex to do it, they pick a crazily oversensitive test that maximizes political power and revenue under the CARES act and then do more covid tests this year than we'll do flu tests in a century.

it's just a gravy train.

and we need to get off.

this is all driven by the massive flows of federal funds. they are the convection building this hurricane of perverse incentives.

cut it off, and this goes away.

so STOP THROWING MONEY AT THIS!

you're literally paying to keep yourself in jail.