[conspire] Numbers racket

[Rick Moen]

Yesterday, one Bob Cohen of West Menlo Park wrote on NextDoor:


Let's remember that the quarantine was implemented to "flatten the
curve", primarily in order to avoid overflow in hospitals.  That was
done very successfully, at least partially because the predictions for
hospital bed use here in California were way on the high side, as were
the mortality figures.

Now we learn that the virus has likely been around much longer than we
originally thought, and many were were exposed, but showed little or no
symptoms.  The recalculated mortality rate is now similar to the
seasonal flu.

Yet despite  we are still told that we must shelter in place, religious
services are not allowed, and a liquor store is an essential business.  

We will soon be seeing a situation where the conditions imposed by the
government will cause more suffering than the virus due to suicide,
domestic abuse, depression etc.

Will we be ready to debate this then?



I have not yet met the gentleman, though he lives in my immediate area.
Anyway, I just posted this comment in reply:

Bob Cohen:  Annual US deaths from the seasonal flu had an arithmetic
mean ("average") value of 23,607 between the years studied (1976 to
2007) in a CDC study
(https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5933a1.htm).  

As of yesterday, the NY Times tracker showed 40,316 US deaths officially
attributable to COVID-19 so far
(https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html)
since the first known US death on Feb. 26, i.e., over slightly less than
a two-month period.  However, notably this (as mentioned at NY Times)
excludes, among others, approximately 4,800 people in NYC alone who were
reported by the NY Public Health Dept. as highly probable COVID-19
deaths -- so make that a bare minimum of 40,316 + 4,800 = 45,116 US
deaths in less than two months.  But again, it's certain that 45,116 too
is a severe undercount.

Why?  Because the ongoing dire shortage of testing has the consequence
that "known cases" for reporting purposes are limited in most states to
those confirmed by testing kits to be COVID-19 deaths as opposed to
deaths that were almost certainly COVID-19 but are excluded for lack of
testing ability.  How drastic is the resulting undercount?  Nobody is
quite certain, and the degree of undercounting certainly differs between
states (see:
https://www.thedailybeast.com/the-covid-19-death-undercount-is-scarier-than-you-think)
and it'll probably take months to years for an accurate estimate, but
I'll go out on a limb and guesstimate from broad readings that we'll
eventually determine a nationwide average undercount factor of 5x-10x
for the two months of death stats so far -- as misattributed deaths get
reclassified.

Flu season in the US is usually construed to be October to May, numbers
peaking in February in a typical year, with the curve of deaths over
time looking sort of like a Gaussian ("normal") curve with a jagged peak
poking above that around February.  Famously, the COVID-19 death curves,
using the badly-undercounted data so far, are exponential.

As they said on Sesame Street, one of these things is not like the
other.  ;->

You seemed to be seeking a "debate" (your word).  I'm not offering one,
but I'm offering some pretty hard data plus one personal surmise where I
guesstimate an undercount factor.  (Ignore that if you don't trust my
surmise, which is certainly fair.)  The hard data, I hope and expect,
might be a lot more useful than any amount of "debate".