Crap Writing Like This Explains Why Public Is So Confused About Science

[Fred Bloggs]

"The obscure maths theorem that governs the reliability of Covid testing"

https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability

He must be getting paid by the word.

[Rick C]

I think it is well written for the most part. At first I thought he was mucking this up but then I realized it's like the Monty Hall problem where you have to understand the question and the context. So it is correct to say the probability of the chance of a false positive increases when the disease is vary rare. He explains this well and goes on to describe specific scenarios of similar probability issues.

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

[Robert Latest]

I remember exactly this scenario as a "warm up" exercise in my 11th grade math
class. There was heated discussion about the apparent paradox that the accuracy
of the test depended on the prevalence of the infection in the population.

Assuming that the general public doesn't have that much more grasp on math than
a class of 11th grade math majors I believe the article did a fairly good job
of explaining it. Including the catchy headline event though it annoys the
math-savvy among us.

--
robert

[bitrex]

Probability "has disease given a positive test" means the probability
"positive test given has disease", has to be weighted by the prior
probability "has disease", which changes with the number of people
infected, the and the probability the test is positive under _any_
circumstances, which is a joint probability of the positive test results
you get in both cases, when the test subject both does and does not
actually have the disease.

A sanity check is that the probability "has disease given positive test"
must be zero in the limiting case when the probability of having the
disease is zero, i.e. the disease is extinct in the population.

[Rick C]

The devil is in the details and in the wording. The accuracy of the test doesn't change with the rate of infection in the population. It's the probability of any given POSITIVE test being a false positive that changes. The point is that is a different statistic. Until that is understood any other discussion is not meaningful.

Look at it this way. Before you are tested, you have the 1 in 1,000 chance of getting a false positive. Once you get the false positive it puts you in a different group and so you now have a higher chance it was a false positive. If there are very few people in that group with the disease, it skews the odds tremendously.

I'm not trying to explain this to you. I understand it, but I'm still trying to figure out how to explain it, so I'm practicing. ;)

--

Rick C.

+ Get 1,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209

[Bill Sloman]

This explains why Fred is so confused about science. He loses patience with any explanation that covers all the facts he needs to know (and he seems to need to be told rather more of them than most of the rest us ).

--
Bill Sloman, Sydney

[Robert Latest]

Rick C wrote:
> I'm not trying to explain this to you. I understand it, but I'm still
> trying to figure out how to explain it, so I'm practicing. ;)

I find the easiest way is to do what the Guardian did, which is to use actual
numbers. That way you get away without multiplying incidences and
probabilities, and -- more importantly -- you can avoid scary words like
"probability" and "statistics" which can cause people to switch off internally.

Remember back when the "three door" problem was the rage? Easiest explanation
went with a hypothetical 100 games being played.

--
robert

[Clive Arthur]

Monty Hall? Simple.

Say the prize is in A.

You choose A, so Monty opens B or C. It would be bad to switch.
You choose B, so Monty opens C. It would be good to switch.
You choose C, so Monty opens B. It would be good to switch.

Two out of three times it would be good to switch. Nothing more to it.


Now for the real life scenarios...

You don't switch and lose. People pat you on the back and commiserate.
Bad luck old chap.

You do switch and lose. When you get home you see a packed suitcase on
your doorstep and find that the door locks have been changed.

--
Cheers
Clive

[DecadentLinux...@decadence.org]

Clive Arthur <cl...@nowaytoday.co.uk> wrote in
news:s5jq0r$t0h$1...@dont-email.me:

The suitcases should be on your doorstep in any case because you
just spouted exactly ZERO science all the while claiming to be using
science.

Bad luck old chap? No... You're a stupid lying fuck, old hack.
Nothing more to it.

[bitrex]

On 4/19/2021 6:53 AM, Robert Latest wrote:
> Rick C wrote:
>> I'm not trying to explain this to you. I understand it, but I'm still
>> trying to figure out how to explain it, so I'm practicing. ;)
>
> I find the easiest way is to do what the Guardian did, which is to use actual
> numbers. That way you get away without multiplying incidences and
> probabilities, and -- more importantly -- you can avoid scary words like
> "probability" and "statistics" which can cause people to switch off internally.

The easiest way to explain it is with math! But for a general audience
we can't do that.

So the second-best option is to explain it with words. But the reading
comprehension ability of a general audience tends to be only slightly
better on average than their mathematics. So ah, hmm.

> Remember back when the "three door" problem was the rage? Easiest explanation
> went with a hypothetical 100 games being played.

Monty Hall Monte Carlo.

And if that technique were used to analyze the same game as played by
Carlo in Monte Carlo it would be Carlo's Monte Carlo Monty Hall Monte Carlo.

[Martin Brown]

It isn't a bad quick introduction to Bayesian statistics for layman.
(and without any of the scary maths that puts people off)

My only criticism would be that it really needs a diagram to illustrate
the theorem (and a suitably frivolous example like handedness of
kangaroos and their fondness drinking Fosters lager).

It could be much worse! When Laplace first came up with it in an effort
to estimate the weight of Saturn from available observational data he
called it "the principle of insufficient reason". That phrase did not go
down at all well although Bernoulli at one point picked it up.

It is only really in the twentieth century that it really took off. Most
modern data analysis is now Bayesian (as are most spam filters).

In its purest form it is a way of mathematically codifying your prior
knowledge (ignorance) about the system before you have any observations.

--
Regards,
Martin Brown

[DecadentLinux...@decadence.org]

bitrex <us...@example.net> wrote in
news:ecgfI.8517$v3H9...@fx01.iad:

> The easiest way to explain it is with math! But for a general
> audience we can't do that.
>
> So the second-best option is to explain it with words. But the
> reading comprehension ability of a general audience tends to be
> only slightly better on average than their mathematics. So ah,
> hmm.
>

You're a complete idiot to make that presumption.

Or should I say 'general idiot'.

[Clive Arthur]

On 19/04/2021 15:50, Martin Brown wrote:

<snipped>

>
> It isn't a bad quick introduction to Bayesian statistics for layman.
> (and without any of the scary maths that puts people off)
>
> My only criticism would be that it really needs a diagram to illustrate
> the theorem (and a suitably frivolous example like handedness of
> kangaroos and their fondness drinking Fosters lager).

I can't take any science journalism seriously unless they use the proper
units. Olympic swimming pools for volume, blue whales for mass, areas
the size of Wales, and Eiffel Towers for height (or Empire State
buildings for non-SI countries).

--
Cheers
Clive

[bitrex]

<https://www.youtube.com/watch?v=pWdd6_ZxX8c>

[Bill Sloman]

The trouble is that you've only read English language science journalism - though your enthusiasm for popular science units of measurement suggests that you do read "New Scientist", which does a whole lot better than English language newspaper science journalism (which is mostly written by people who want to become foreign correspondents when they grow up). Dutch science journalists frequently at least a first degree in the subject they write about, and I'd been living in the Netherlands for years before I caught them getting something wrong. A great improvement on English newspapers (and I never read any of the Murdoch owned papers, let alone the total crap like the Daily Mail and the Express).

--
Bill Sloman, Sydney

[Clive Arthur]

I used to subscribe to New Scientist, but sadly, it's now owned by DMGT
- Daily Mail and General Trust. I don't know if it's changed, but I
won't give them my money.

--
Cheers
Clive

[Rick C]

I saw something the other day when looking at electrical distribution. "pu" for per-unit. I'm not sure I competely understand it, but it is somehow normalized so that without worrying about voltage or amps comparisons can be made across transformers. I didn't feel like wading through the math so I'm not certain of any of it. I was just looking for the voltage on the input side of the distribution transformer feeding your house in the US. Also, I was trying to find how many houses are typically connected to the 240V output. Seems 7 kV is not an unusual number for the input and 1 to 4 houses is common for connections to the 240V output.

--

Rick C.

-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209

[bitrex]

The football field is the standard non-SI unit of length

[bitrex]

Er, sorry. the American football field, to be clear

[John Robertson]

Soccer or that other game?

John ;-#)#

[John Doe]

"the concepts "male" and "female" are essentially social constructions" (Bill Sloman)

Here is Bozo the Clown...

--
Bill Sloman <bill....@ieee.org> wrote:

> X-Received: by 2002:ac8:5744:: with SMTP id 4mr9784447qtx.10.1618801218153; Sun, 18 Apr 2021 20:00:18 -0700 (PDT)
> X-Received: by 2002:ae9:f205:: with SMTP id m5mr9956646qkg.101.1618801217954; Sun, 18 Apr 2021 20:00:17 -0700 (PDT)
> Path: eternal-september.org!reader02.eternal-september.org!news.mixmin.net!proxad.net!feeder1-2.proxad.net!209.85.160.216.MISMATCH!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
> Newsgroups: sci.electronics.design
> Date: Sun, 18 Apr 2021 20:00:17 -0700 (PDT)
> In-Reply-To: <9492007c-68be-442e...@googlegroups.com>
> Injection-Info: google-groups.googlegroups.com; posting-host=124.168.64.63; posting-account=SJ46pgoAAABuUDuHc5uDiXN30ATE-zi-
> NNTP-Posting-Host: 124.168.64.63
> References: <9492007c-68be-442e...@googlegroups.com>
> User-Agent: G2/1.0
> MIME-Version: 1.0
> Message-ID: <ae4bfe9e-4488-487b...@googlegroups.com>
> Subject: Re: Crap Writing Like This Explains Why Public Is So Confused About Science
> From: Bill Sloman <bill....@ieee.org>
> Injection-Date: Mon, 19 Apr 2021 03:00:18 +0000
> Content-Type: text/plain; charset="UTF-8"
> Xref: reader02.eternal-september.org sci.electronics.design:630348

[Fred Bloggs]

On Sunday, April 18, 2021 at 4:32:40 PM UTC-4, Robert Latest wrote:

He's meandering all over the place in addition to being obnoxious with his opening remarks. Notice the idiot first gives a numerical of actual chance of being positive having tested positive by taking ratio of true positive to total positive tests, then switches to inordinately large false positive rate calculation by looking at ratio of false positive to total positive test results. He's too stupid to understand that NOT false positive test doesn't mean true positive. Not getting a false positive can also mean being true/false negative as well as being a true positive. So there's no reason to expect the proportions to be complementary. His idiotic diatribe what with changing the statistics and subjects frequently and entirely missing the crux of the calculation gets this article a bad rating in my book.


>
> --
> robert

[Fred Bloggs]

LOL- it is? That must be some extension in modern times but it has nothing to do with Bayes original thinking.

I'm not sure why that statement was even elevated to the status of theorem, it is simple calculation of the probability of occurrence of one of a finite number of mutually exclusive and exhaustive causative associations with a separate event under study.

You have to reconfigure so the associated causatives are the state of infected/ uninfected, and the event under study is either a positive or negative test result, not both, for Bayes to make a statement about your infection status based upon a test result.



>
> --
> Regards,
> Martin Brown

[Fred Bloggs]

There were no facts. He didn't begin to explain Bayes calculation, he just name dropped once and then left it never to return, after trying to wow everyone with that simple conditional probability conversion. Most likely you identify with him because he uses the same approach as you do what with insinuation of deeper understanding and profundity that's just not there, an especially laughable characteristic of your posts, being deliberately vague, insulting his readers and telling them they're wrong.


>
> --
> Bill Sloman, Sydney

[Bill Sloman]

That's one explanation. The simpler one is that Fred can't understand anything anything even vaguely complicated and blames the writer's poor exposition for the problems created by his own defective comprehension. I've never been able to get him to understand that "highly conserved" doesn't mean "doesn't mutate" but rather that "when it does mutate most of the mutates get selected out rapidly". He'll probably post yet another explanation of why he thinks I'm wrong.

--
Bill Sloman, Sydney

[Bill Sloman]

On Friday, April 23, 2021 at 7:33:32 AM UTC+10, Fred Bloggs wrote:
> On Monday, April 19, 2021 at 10:50:13 AM UTC-4, Martin Brown wrote:
> > On 18/04/2021 16:34, Fred Bloggs wrote:
> > > "The obscure maths theorem that governs the reliability of Covid testing"
> > >
> > > https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> > >
> > > He must be getting paid by the word.
> > It isn't a bad quick introduction to Bayesian statistics for layman.
> > (and without any of the scary maths that puts people off)
> >
> > My only criticism would be that it really needs a diagram to illustrate
> > the theorem (and a suitably frivolous example like handedness of
> > kangaroos and their fondness drinking Fosters lager).
> >
> > It could be much worse! When Laplace first came up with it in an effort
> > to estimate the weight of Saturn from available observational data he
> > called it "the principle of insufficient reason". That phrase did not go
> > down at all well although Bernoulli at one point picked it up.
> >
> > It is only really in the twentieth century that it really took off. Most
> > modern data analysis is now Bayesian (as are most spam filters).
> >
> > In its purest form it is a way of mathematically codifying your prior
> > knowledge (ignorance) about the system before you have any observations.
> LOL- it is? That must be some extension in modern times but it has nothing to do with Bayes original thinking.
>
> I'm not sure why that statement was even elevated to the status of theorem, it is simple calculation of the probability of occurrence of one of a finite number of mutually exclusive and exhaustive causative associations with a separate event under study.

https://en.wikipedia.org/wiki/Bayes%27_theorem

It probably reflects the need to straighten the thinking of people like Fred who don't seem to get the point.

> You have to reconfigure so the associated causatives are the state of infected/ uninfected, and the event under study is either a positive or negative test result, not both, for Bayes to make a statement about your infection status based upon a test result.

The problem is that there are four test results - true negative, true positive, false negative and false positive. You can't immediately tell the difference between a false positive and true positive, but your reaction to a positive result should be informed by the likelihood that it is true.

--
Bill Sloman, Sydney

[Bill Sloman]

On Friday, April 23, 2021 at 6:58:34 AM UTC+10, Fred Bloggs wrote:
> On Sunday, April 18, 2021 at 4:32:40 PM UTC-4, Robert Latest wrote:
> > Fred Bloggs wrote:
> > > "The obscure maths theorem that governs the reliability of Covid testing"
> > >
> > > https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-
> > theorem-reliability-covid-lateral-flow-tests-probability
> > >
> > > He must be getting paid by the word.
> > I remember exactly this scenario as a "warm up" exercise in my 11th grade math
> > class. There was heated discussion about the apparent paradox that the accuracy
> > of the test depended on the prevalence of the infection in the population.
> >
> > Assuming that the general public doesn't have that much more grasp on math than
> > a class of 11th grade math majors I believe the article did a fairly good job
> > of explaining it. Including the catchy headline event though it annoys the
> > math-savvy among us.
> He's meandering all over the place in addition to being obnoxious with his opening remarks. Notice the idiot first gives a numerical of actual chance of being positive having tested positive by taking ratio of true positive to total positive tests, then switches to inordinately large false positive rate calculation by looking at ratio of false positive to total positive test results. He's too stupid to understand that NOT false positive test doesn't mean true positive.

Since you only know whether the test result is either positive or negative, nobody cares what a NOT false positive test result could mean.

> Not getting a false positive can also mean being true/false negative as well as being a true positive.

It can't mean anything. You can only know that you've had a false positive result on the basis of further evidence - probably a more reliable follow-up test. Not getting symptoms of Covid-19 wouldn't be evidence (since 15% of Covid19 infections are asymptomatic).

> So there's no reason to expect the proportions to be complementary. His idiotic diatribe what with changing the statistics and subjects frequently and entirely missing the crux of the calculation gets this article a bad rating in my book.

The results that you can know have to be complementary. If you vanish down a rabbit-hole of worrying about indistinguishable conditions you can get just as confused as Fred.

--
Bill Sloman, Sydney

[Martin Brown]

His only failing if there is one is that he didn't dumb it down to a low
enough level where Fred Bloggs could actually understand it.

You could perhaps take issue with his use of prior probability to
describe the known (from other experiments) proportion of the population
with Covid. The basic analysis that he presented could just as easily
have been done by a careful frequentallist statistician.

Uninformative prior probability distributions for Bayesian analysis
codify a perfect state of ignorance in the absence of any data at all.

Mathematicians were very squeamish about them since they are typically
un-normalisable with one or more infinities in the range. This flaw
disappears once you have some actual data to infer an answer from.

Jeffries formalised it at the start of the twentieth century but it
wasn't really until the likes of Ed Jaynes that it really began to take
off as a realistic way of getting tiny signals out of lots of noise. It
takes a fair amount of computing power to do it for serious problems.

Most of modern astronomical image processing relies heavily on it.
Notably the detailed image of the black hole accretion disk.

--
Regards,
Martin Brown

[jla...@highlandsniptechnology.com]

On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
<bloggs.fred...@gmail.com> wrote:

>"The obscure maths theorem that governs the reliability of Covid testing"
>
>https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>
>He must be getting paid by the word.

Seems to me that most scientists are confused by science too.

The great thing about science is that there's so much of it to choose
from.



--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.

[Bill Sloman]

On Friday, April 23, 2021 at 9:51:50 PM UTC+10, jla...@highlandsniptechnology.com wrote:
> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
>
> >"The obscure maths theorem that governs the reliability of Covid testing"
> >
> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> >
> >He must be getting paid by the word.
>
> Seems to me that most scientists are confused by science too.

But at a much higher level than you and Fred Bloggs, and within a much narrower range of choices.

> The great thing about science is that there's so much of it to choose from.

It called reviewing the literature. It helps if you can understand what you are reviewing - which sometimes takes quite a lot of work of the kind you don't seem to be willing to put in, even if you had the capacity to do it.

You seems to include crap like climate change denial propaganda within what you think of as "science", which presents you with a whole lot of unnecessary extra nonsense to get even more confused about.

--
Bill Sloman, Sydney

[Fred Bloggs]

As usual you miss the point entirely. I'm not seeing the world's public agencies running to the nearest astronomical observatory asking for advice in analyzing their data. You're idea of what's going is as ridiculous as it gets. The health testing being done is a class of surveillance testing. There is no lack of "signal" , ample confirmation techniques through alternative testing and clinical trials, necessary to completely characterize the performance of the test. The test is chosen based on considerations of very low cost. very rapid turnaround, ease of administration by minimally trained personnel, and wide availability. This has NOTHING to do with ET Jaynes! You give yourself away as a completely out of touch mouthy nerd who doesn't have a clue in hell what he's talking about.



>
> --
> Regards,
> Martin Brown

[Fred Bloggs]

More of your pathetically transparent posturing. As usual you're going to shoot your mouth off about anything and everything except the subject at hand. Then you pretend to have some kind of insight into Bayesian statistics but somehow can't begin to scribble the slightest and most elementary calculation because even that much is way over your dimwit abilities.

The little wiki article says it all.

P(infected | positive test)= P(positive test | infected) x P(infected)/[P(positive test | infected) x P(infected) + P(positive test | uninfected) x P(uninfected)]

In case you can't figure it out, that is the answer to the original question of the article. How likely are you to be infected given a positive test result?

You're too damn dumb to even understand that little tidbit. And that's just the algebra. You miss the point entirely about P(positive test | infected) and P(positive test | uninfected) are both very rigorously characterized by ongoing statistical sampling fully backed by confirmation testing techniques of extreme precision.

You can plug in some numbers of well known captured data. In an advanced country:
P(positive test | infected) is about 0.99
P(positive test | uninfected) is about 0.001 ( according to article example)
P(infected) runs about 5% in most advanced nations ( other places in can be ral bad like 30%)
P(uninfected)= 1- P(infected)=95%

This makes
P(infected | positive test)= 0.99 x 0.05/[0.99 x 0.05 + 0.001 x 0.95]=0.98
which answers the original question.

You don't like that answer, you can always run the various stats through their 90% confidence intervals to smear it through a range with confidence you're happy with. Good luck with that given your capabilities though.

Notice this has nothing to do with legal interpretation of DNA results or prosecution of murders or any other unrelated trivia.

You don't like the idea of thousands of people being falsely told their positive? Awww.... tough shit!




>
> --
> Bill Sloman, Sydney

[Fred Bloggs]

On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
>
> >"The obscure maths theorem that governs the reliability of Covid testing"
> >
> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> >
> >He must be getting paid by the word.
> Seems to me that most scientists are confused by science too.
>
> The great thing about science is that there's so much of it to choose
> from.

LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.

[jla...@highlandsniptechnology.com]

On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs

<bloggs.fred...@gmail.com> wrote:

>On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
>> <bloggs.fred...@gmail.com> wrote:
>>
>> >"The obscure maths theorem that governs the reliability of Covid testing"
>> >
>> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>> >
>> >He must be getting paid by the word.
>> Seems to me that most scientists are confused by science too.
>>
>> The great thing about science is that there's so much of it to choose
>> from.
>
>LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.

How many masks do you wear outdoors?

[Fred Bloggs]

I wear a mask if there are people around. You never know what's going to be upwind.
I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover

[jla...@highlandsniptechnology.com]

On Fri, 23 Apr 2021 08:45:43 -0700 (PDT), Fred Bloggs

<bloggs.fred...@gmail.com> wrote:

>On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>> On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs
>> <bloggs.fred...@gmail.com> wrote:
>>
>> >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>> >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
>> >> <bloggs.fred...@gmail.com> wrote:
>> >>
>> >> >"The obscure maths theorem that governs the reliability of Covid testing"
>> >> >
>> >> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>> >> >
>> >> >He must be getting paid by the word.
>> >> Seems to me that most scientists are confused by science too.
>> >>
>> >> The great thing about science is that there's so much of it to choose
>> >> from.
>> >
>> >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.
>> How many masks do you wear outdoors?
>
>I wear a mask if there are people around. You never know what's going to be upwind.
>I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
>https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover
>

Do you wear that when you drive?

I bet there are people who wear masks to bed.

[jla...@highlandsniptechnology.com]

On Fri, 23 Apr 2021 09:20:14 -0700, jla...@highlandsniptechnology.com

wrote:

>On Fri, 23 Apr 2021 08:45:43 -0700 (PDT), Fred Bloggs
><bloggs.fred...@gmail.com> wrote:
>
>>On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>>> On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs
>>> <bloggs.fred...@gmail.com> wrote:
>>>
>>> >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>>> >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
>>> >> <bloggs.fred...@gmail.com> wrote:
>>> >>
>>> >> >"The obscure maths theorem that governs the reliability of Covid testing"
>>> >> >
>>> >> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>>> >> >
>>> >> >He must be getting paid by the word.
>>> >> Seems to me that most scientists are confused by science too.
>>> >>
>>> >> The great thing about science is that there's so much of it to choose
>>> >> from.
>>> >
>>> >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.
>>> How many masks do you wear outdoors?
>>
>>I wear a mask if there are people around. You never know what's going to be upwind.
>>I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
>>https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover
>>
>
>Do you wear that when you drive?
>
>I bet there are people who wear masks to bed.

Apparently some people do!

Funny: one source suggests that people sleeping in the same bed sleep
head-to-toe. Mo's Toes in my face!

Another suggests that families should mask indoors.

[Fred Bloggs]

No- no need for it. But I do spray the soles of my shoes down with HOCL before getting back in the car :-)

>
> I bet there are people who wear masks to bed.

They should get separate bedrooms if they think they need that extreme.

[Fred Bloggs]

Actually that one makes sense if you have teens that are running around to who knows where.

[Bill Sloman]

On Saturday, April 24, 2021 at 1:45:46 AM UTC+10, Fred Bloggs wrote:
> On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs
> > <bloggs.fred...@gmail.com> wrote:
> >
> > >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
> > >> <bloggs.fred...@gmail.com> wrote:
> > >>
> > >> >"The obscure maths theorem that governs the reliability of Covid testing"
> > >> >
> > >> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> > >> >
> > >> >He must be getting paid by the word.
> > >> Seems to me that most scientists are confused by science too.
> > >>
> > >>
> > >> The great thing about science is that there's so much of it to choose
> > >> from.
> > >
> > >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.

So out of touch that they notice when Fred gets stuff wrong?

> > How many masks do you wear outdoors?
>
> I wear a mask if there are people around. You never know what's going to be upwind.
> I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
> https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover

Trust Fred to be smug about stuff he doesn't really understand. Most mask users aim for a snug fit, but Fred's all about convincing himself that he has got it right.

--
Bill Sloman, Sydney

[Bill Sloman]

On Saturday, April 24, 2021 at 1:20:34 AM UTC+10, Fred Bloggs wrote:
> On Thursday, April 22, 2021 at 11:42:54 PM UTC-4, Bill Sloman wrote:
> > On Friday, April 23, 2021 at 7:42:30 AM UTC+10, Fred Bloggs wrote:
> > > On Sunday, April 18, 2021 at 11:00:20 PM UTC-4, Bill Sloman wrote:
> > > > On Monday, April 19, 2021 at 1:34:40 AM UTC+10, Fred Bloggs wrote:
> > > > > "The obscure maths theorem that governs the reliability of Covid testing"
> > > > >
> > > > > https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> > > > >
> > > > > He must be getting paid by the word.
> > > >
> > > > This explains why Fred is so confused about science. He loses patience with any explanation that covers all the facts he needs to know (and he seems to need to be told rather more of them than most of the rest us ).
> > >
> > > There were no facts. He didn't begin to explain Bayes calculation, he just name dropped once and then left it never to return, after trying to wow everyone with that simple conditional probability conversion. Most likely you identify with him because he uses the same approach as you do what with insinuation of deeper understanding and profundity that's just not there, an especially laughable characteristic of your posts, being deliberately vague, insulting his readers and telling them they're wrong.
>
> > That's one explanation. The simpler one is that Fred can't understand anything anything even vaguely complicated and blames the writer's poor exposition for the problems created by his own defective comprehension. I've never been able to get him to understand that "highly conserved" doesn't mean "doesn't mutate" but rather that "when it does mutate most of the mutates get selected out rapidly". He'll probably post yet another explanation of why he thinks I'm wrong.
>
> More of your pathetically transparent posturing. As usual you're going to shoot your mouth off about anything and everything except the subject at hand. Then you pretend to have some kind of insight into Bayesian statistics but somehow can't begin to scribble the slightest and most elementary calculation because even that much is way over your dimwit abilities.

Not exactly. I've got to keep the abilities of my audience in mind. As Stephen Hawking famously observed, each additional equation in the text halves the likely readership.
It wouldn't put off the readers I don't need to address.

> The little wiki article says it all.
>
> P(infected | positive test)= P(positive test | infected) x P(infected)/[P(positive test | infected) x P(infected) + P(positive test | uninfected) x P(uninfected)]
>
> In case you can't figure it out, that is the answer to the original question of the article. How likely are you to be infected given a positive test result?

Not without actual values for probabilities.

> You're too damn dumb to even understand that little tidbit. And that's just the algebra. You miss the point entirely about P(positive test | infected) and P(positive test | uninfected) are both very rigorously characterized by ongoing statistical sampling fully backed by confirmation testing techniques of extreme precision.

Sometimes. Useful probabilities are data about the world as it is now, and since you can't take a large number of samples in a very short time you can't get extremely precise data that is up-to-date. Epidemics have a way of changing the real world inconveniently rapidly. And new viral strains make life even more complicated.

> You can plug in some numbers of well known captured data. In an advanced country:
> P(positive test | infected) is about 0.99

> P(positive test | uninfected) is about 0.001 ( according to article example) .> P(infected) runs about 5% in most advanced nations ( other places in can be real bad like 30%)
> P(uninfected)= 1- P(infected)=95%

That very much depends where you are doing the testing and in which country. The rule is that you should get less than 5% positive outcomes - but all it means is that if you get more than 5% you should be testing more people. In Queensland they recently tested some 28,000 people and didn't get one positive outcome

> This makes P(infected | positive test)= 0.99 x 0.05/[0.99 x 0.05 + 0.001 x 0.95]=0.98
> which answers the original question.

Incorrectly, in any place that has got the epidemic under control.

> You don't like that answer, you can always run the various stats through their 90% confidence intervals to smear it through a range with confidence you're happy with. Good luck with that given your capabilities though.

I don't like the answer because you've used 5% as the likely level for the proportion of people who really are infected, when correct answer is going to be very context dependent.

> Notice this has nothing to do with legal interpretation of DNA results or prosecution of murders or any other unrelated trivia.
>

> You don't like the idea of thousands of people being falsely told they are positive? Awww.... tough shit!

It doesn't worry me at all. Doing lots of testing is usually a good idea, but it the people tested are told the results of their tests (which isn't a good idea if there are going to be many more false positives than true positives) they do need to be informed about what the test result actually means.

--
Bill Sloman, Sydney

[Fred Bloggs]

On Saturday, April 24, 2021 at 2:34:58 AM UTC-4, Bill Sloman wrote:
> On Saturday, April 24, 2021 at 1:20:34 AM UTC+10, Fred Bloggs wrote:
> > On Thursday, April 22, 2021 at 11:42:54 PM UTC-4, Bill Sloman wrote:
> > > On Friday, April 23, 2021 at 7:42:30 AM UTC+10, Fred Bloggs wrote:
> > > > On Sunday, April 18, 2021 at 11:00:20 PM UTC-4, Bill Sloman wrote:
> > > > > On Monday, April 19, 2021 at 1:34:40 AM UTC+10, Fred Bloggs wrote:
> > > > > > "The obscure maths theorem that governs the reliability of Covid testing"
> > > > > >
> > > > > > https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> > > > > >
> > > > > > He must be getting paid by the word.
> > > > >
> > > > > This explains why Fred is so confused about science. He loses patience with any explanation that covers all the facts he needs to know (and he seems to need to be told rather more of them than most of the rest us ).
> > > >
> > > > There were no facts. He didn't begin to explain Bayes calculation, he just name dropped once and then left it never to return, after trying to wow everyone with that simple conditional probability conversion. Most likely you identify with him because he uses the same approach as you do what with insinuation of deeper understanding and profundity that's just not there, an especially laughable characteristic of your posts, being deliberately vague, insulting his readers and telling them they're wrong.
> >
> > > That's one explanation. The simpler one is that Fred can't understand anything anything even vaguely complicated and blames the writer's poor exposition for the problems created by his own defective comprehension. I've never been able to get him to understand that "highly conserved" doesn't mean "doesn't mutate" but rather that "when it does mutate most of the mutates get selected out rapidly". He'll probably post yet another explanation of why he thinks I'm wrong.
> >
> > More of your pathetically transparent posturing. As usual you're going to shoot your mouth off about anything and everything except the subject at hand. Then you pretend to have some kind of insight into Bayesian statistics but somehow can't begin to scribble the slightest and most elementary calculation because even that much is way over your dimwit abilities.
> Not exactly. I've got to keep the abilities of my audience in mind. As Stephen Hawking famously observed, each additional equation in the text halves the likely readership.
> It wouldn't put off the readers I don't need to address.
> > The little wiki article says it all.
> >
> > P(infected | positive test)= P(positive test | infected) x P(infected)/[P(positive test | infected) x P(infected) + P(positive test | uninfected) x P(uninfected)]
> >
> > In case you can't figure it out, that is the answer to the original question of the article. How likely are you to be infected given a positive test result?
> Not without actual values for probabilities.
> > You're too damn dumb to even understand that little tidbit. And that's just the algebra. You miss the point entirely about P(positive test | infected) and P(positive test | uninfected) are both very rigorously characterized by ongoing statistical sampling fully backed by confirmation testing techniques of extreme precision.
> Sometimes. Useful probabilities are data about the world as it is now, and since you can't take a large number of samples in a very short time you can't get extremely precise data that is up-to-date. Epidemics have a way of changing the real world inconveniently rapidly. And new viral strains make life even more complicated.

The test performance is not one bit a function of time and circumstance, it is very well characterized. If it does deviate, then you're talking about a variant.

> > You can plug in some numbers of well known captured data. In an advanced country:
> > P(positive test | infected) is about 0.99
> > P(positive test | uninfected) is about 0.001 ( according to article example) .> P(infected) runs about 5% in most advanced nations ( other places in can be real bad like 30%)
> > P(uninfected)= 1- P(infected)=95%
>
> That very much depends where you are doing the testing and in which country. The rule is that you should get less than 5% positive outcomes - but all it means is that if you get more than 5% you should be testing more people. In Queensland they recently tested some 28,000 people and didn't get one positive outcome

Didn't I say that in the write-up?! In most of U.S. it is approximately 5%:
"The percentage of COVID-19 RT-PCR tests that are positive (percent positivity) has decreased from the previous week. The 7-day average of percent positivity from tests is now 5.2%. The 7-day average test volume for April 9-April 15, 2021, was 1,189,820, up 1.6% from 1,170,968 for the prior 7 days."
https://www.cdc.gov/coronavirus/2019-ncov/covid-data/covidview/index.html

Right now India is running plus 30%, and that pushes the Bayes estimate to above 99%, effectively 100%. Again making perfect sense because the relative fraction of infected people to the total subpopulation testing positive is overwhelmingly infected. It does show too that a 600% increase in infection rate slows to a 2-point increase in conditional probability, which is a fair desensitization.

> > This makes P(infected | positive test)= 0.99 x 0.05/[0.99 x 0.05 + 0.001 x 0.95]=0.98
> > which answers the original question.
> Incorrectly, in any place that has got the epidemic under control.

Again, you don't understand basic algebra. And you certainly don't know how to interpret
analytic results. You try to fool people by parroting some numbers you've read somewhere, but you can't draw any inference from it because you're too dumb.

For example, if the pandemic is under control to the point of 0% infected, Bayes equation predicts a 0% chance of infection given a positive test result. And that makes perfect sense because no one has it. At the other extreme, if the test has a 0% false alarm rate, Bayes' equation states you have 100% chance of being infected if you get a positive test. And that makes perfect sense since it's impossible for an uninfected person to get a positive result. Your problem with Bayes Rule is you don't understand conditional probability, which is proportionately relative measure, not an absolute one. Another use for Bayes result would be to estimate the infection rate from the P(infected | positive test) data such as it is.

> > You don't like that answer, you can always run the various stats through their 90% confidence intervals to smear it through a range with confidence you're happy with. Good luck with that given your capabilities though.
> I don't like the answer because you've used 5% as the likely level for the proportion of people who really are infected, when correct answer is going to be very context dependent.

Uh-huh, and this without knowing just how the Bayes result is sensitized to infection rate. Brilliant.

> > Notice this has nothing to do with legal interpretation of DNA results or prosecution of murders or any other unrelated trivia.
> >
> > You don't like the idea of thousands of people being falsely told they are positive? Awww.... tough shit!
>
> It doesn't worry me at all. Doing lots of testing is usually a good idea, but it the people tested are told the results of their tests (which isn't a good idea if there are going to be many more false positives than true positives) they do need to be informed about what the test result actually means.

Surveillance testing is not medical diagnostic testing. You're out of touch with all things practical so you don't have anything useful to say about anything. The main reason you're so ignorant is because you think you're too smart to be told anything.

>
> --
> Bill Sloman, Sydney

[Fred Bloggs]

On Saturday, April 24, 2021 at 2:02:34 AM UTC-4, Bill Sloman wrote:
> On Saturday, April 24, 2021 at 1:45:46 AM UTC+10, Fred Bloggs wrote:
> > On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > > On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs
> > > <bloggs.fred...@gmail.com> wrote:
> > >
> > > >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > > >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
> > > >> <bloggs.fred...@gmail.com> wrote:
> > > >>
> > > >> >"The obscure maths theorem that governs the reliability of Covid testing"
> > > >> >
> > > >> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> > > >> >
> > > >> >He must be getting paid by the word.
> > > >> Seems to me that most scientists are confused by science too.
> > > >>
> > > >>
> > > >> The great thing about science is that there's so much of it to choose
> > > >> from.
> > > >
> > > >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.
> So out of touch that they notice when Fred gets stuff wrong?

You think you noticed that but the reality is you're completely wrong. The reason you're completely wrong so often is due to a combination of ignorance, lack of intelligence and mental illness.

> > > How many masks do you wear outdoors?
> >
> > I wear a mask if there are people around. You never know what's going to be upwind.
> > I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
> > https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover
> Trust Fred to be smug about stuff he doesn't really understand. Most mask users aim for a snug fit, but Fred's all about convincing himself that he has got it right.

You can't begin to get a snug fit with those crummy little throwaway jobs. They're little more than a suggestion of a PPD.

>
> --
> Bill Sloman, Sydney

[Bill Sloman]

On Sunday, April 25, 2021 at 12:16:56 AM UTC+10, Fred Bloggs wrote:
> On Saturday, April 24, 2021 at 2:02:34 AM UTC-4, Bill Sloman wrote:
> > On Saturday, April 24, 2021 at 1:45:46 AM UTC+10, Fred Bloggs wrote:
> > > On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > > > On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs <bloggs.fred...@gmail.com> wrote:
> > > > >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
> > > > >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs <bloggs.fred...@gmail.com> wrote:

<snip>

> > > > >> The great thing about science is that there's so much of it to choose
> > > > >> from.
> > > > >
> > > > >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.
> >
> > So out of touch that they notice when Fred gets stuff wrong?
> >
> You think you noticed that but the reality is you're completely wrong. The reason you're completely wrong so often is due to a combination of ignorance, lack of intelligence and mental illness.

Fred does inhabit his own reality. Mental illness can do that.

> > > > How many masks do you wear outdoors?
> > >
> > > I wear a mask if there are people around. You never know what's going to be upwind.
> > > I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
> > > https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover
>
> > Trust Fred to be smug about stuff he doesn't really understand. Most mask users aim for a snug fit, but Fred's all about convincing himself that he has got it right.
>
> You can't begin to get a snug fit with those crummy little throwaway jobs. They're little more than a suggestion of a PPD.

And Fred still hasn't noticed his typo ...

--
Bill Sloman, Sydney

[jla...@highlandsniptechnology.com]

On Sat, 24 Apr 2021 07:16:51 -0700 (PDT), Fred Bloggs

<bloggs.fred...@gmail.com> wrote:

>On Saturday, April 24, 2021 at 2:02:34 AM UTC-4, Bill Sloman wrote:
>> On Saturday, April 24, 2021 at 1:45:46 AM UTC+10, Fred Bloggs wrote:
>> > On Friday, April 23, 2021 at 11:35:30 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>> > > On Fri, 23 Apr 2021 08:26:38 -0700 (PDT), Fred Bloggs
>> > > <bloggs.fred...@gmail.com> wrote:
>> > >
>> > > >On Friday, April 23, 2021 at 7:51:50 AM UTC-4, jla...@highlandsniptechnology.com wrote:
>> > > >> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
>> > > >> <bloggs.fred...@gmail.com> wrote:
>> > > >>
>> > > >> >"The obscure maths theorem that governs the reliability of Covid testing"
>> > > >> >
>> > > >> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>> > > >> >
>> > > >> >He must be getting paid by the word.
>> > > >> Seems to me that most scientists are confused by science too.
>> > > >>
>> > > >>
>> > > >> The great thing about science is that there's so much of it to choose
>> > > >> from.
>> > > >
>> > > >LOL- we have two examples of self-designated scientists in this thread who are hopelessly out of touch.
>> So out of touch that they notice when Fred gets stuff wrong?
>
>You think you noticed that but the reality is you're completely wrong. The reason you're completely wrong so often is due to a combination of ignorance, lack of intelligence and mental illness.
>
>> > > How many masks do you wear outdoors?
>> >
>> > I wear a mask if there are people around. You never know what's going to be upwind.
>> > I use the Honeywell Dual Layer Face Cover with single use filter inserts. I like the fact the ear straps are adjustable so it makes a pretty smug fit.
>> > https://sps.honeywell.com/us/en/campaigns/safety/dual-layer-face-cover
>> Trust Fred to be smug about stuff he doesn't really understand. Most mask users aim for a snug fit, but Fred's all about convincing himself that he has got it right.
>
>You can't begin to get a snug fit with those crummy little throwaway jobs. They're little more than a suggestion of a PPD.
>

My mask is very comfortable in my back pocket with my phone.

[jla...@highlandsniptechnology.com]

On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
<bloggs.fred...@gmail.com> wrote:

>"The obscure maths theorem that governs the reliability of Covid testing"
>
>https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
>
>He must be getting paid by the word.

People would be less confused about science if the science weren't so
confused.

https://justthenews.com/politics-policy/coronavirus/mit-researchers-challenge-indoor-social-distancing-rules-no-safer-60-ft

[whit3rd]

On Friday, April 23, 2021 at 8:35:30 AM UTC-7, jla...@highlandsniptechnology.com wrote:

> How many masks do you wear outdoors?

One is enough. Because a mask (or one's fingers) may become contaminated, it is
prudent to handle masks infrequently and carefully, so... for short errands,
wash, apply mask, go forth. Don't disturb the mask (or touch your face) until
the errand is over, and you return home, and wash again.

Outdoors isn't completely safe, or wild critters wouldn't have picked up the ancestor
of COVID from bats, and we wouldn't have picked it up from the critters.

[Bill Sloman]

On Monday, April 26, 2021 at 1:29:01 AM UTC+10, jla...@highlandsniptechnology.com wrote:
> On Sun, 18 Apr 2021 08:34:35 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
>
> >"The obscure maths theorem that governs the reliability of Covid testing"
> >
> >https://www.theguardian.com/world/2021/apr/18/obscure-maths-bayes-theorem-reliability-covid-lateral-flow-tests-probability
> >
> >He must be getting paid by the word.
>
> People would be less confused about science if the science weren't so confused.
>
> https://justthenews.com/politics-policy/coronavirus/mit-researchers-challenge-indoor-social-distancing-rules-no-safer-60-ft

The confusion is mostly John Larkin's. The argument presented in the link is that because small droplets of saliva that can carry the Covid-9 virus can travel further than six feet, we need to go for bigger "social distance" than six feet. This misses the point that we only emit a finite number of these droplets and the concentration of droplet per inhaled breath drops as the cube of the separation between infected person exhaling, and the person who might inhale the infected droplet.

Six feet seems to work. Sixty feet might work better, but is a trifle impractical.

--
Bill Sloman, Sydney

[Martin Brown]

I'm not convinced that pubs can ever really be safe indoors until the
general public have been vaccinated to the point of herd immunity. There
is no such thing as a socially distanced drunk!

UK suffered very badly in the run up to Xmas where the pubs were still
open - the former hotspots were completely eclipsed by the new super
spreading events in the previously safer zones with pubs still open.
(and loads of students in them with their Covid negative certificates)

Testing negative for Covid doesn't mean you haven't got it - which is a
big problem for how lateral flow tests are being deployed in the UK.

> Six feet seems to work. Sixty feet might work better, but is a trifle
> impractical.

UK uses 2m as its social distance - being a metric country.

Many energy efficient air conditioning systems effectively spread any
fine aerosol that might escape into an indoor environment like a
supermarket through the entire space on a relatively short timescale.

There is therefore an advantage to wearing a mask in such indoors spaces
mainly to prevent infecting others if you have Covid but also with a
marginal gain for the wearer (provided you don't touch your eyes).

It is amazing how many people out and about touch their eyes, lips or
face with potentially contaminated hands - often while yacking on a
mobile phone. People watching in Covid queues is interesting.

--
Regards,
Martin Brown

[jla...@highlandsniptechnology.com]

https://www.nature.com/articles/d41586-021-00251-4

[jla...@highlandsniptechnology.com]

On Mon, 26 Apr 2021 10:23:00 +0100, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:

Pubs probably transmit all sorts of bacterial and viral diseases. STDs
too.

[Martin Brown]

Undoubtedly, but none of them are quite so potentially lethal as Covid -
until a multiple drug resistant version of the plague makes a comeback.

Multiple drug resistant strains of TB have become a problem in the USA
and elsewhere when antibiotics are handed out like Smarties (M&Ms).

--
Regards,
Martin Brown

[jla...@highlandsniptechnology.com]

On Tue, 27 Apr 2021 15:37:23 +0100, Martin Brown

Great book: And The Band Played On by Randy Shilts. He was dying of
AIDS as he wrote it.

I lived 2 blocks from Castro Street just as AIDS hit. I doubt that
many Roman orgies were as wild as the bar scenes there.