Scrutable

Here was today's cabinet meeting.



I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

New numbers in - 1,789 deaths in total now, as of 5pm yesterday. The increase of 381 is more than double yesterday's increase. On March 15th, Italy had 1,809 deaths.

Italy's numbers do look to be slowing down, even if they're still getting horrendous numbers per day. If they've passed the peak then they'll end up probably somewhere around 18k-20k deaths in this phase of the pandemic.

El Pollo Diablo wrote: Tue Mar 31, 2020 3:07 pm Italy's numbers do look to be slowing down, even if they're still getting horrendous numbers per day. If they've passed the peak then they'll end up probably somewhere around 18k-20k deaths in this phase of the pandemic.

Imperial's modelling (linked above by someone) estimates that as of today, Italy's lockdown restrictions have prevented 38,000 deaths.

FlammableFlower wrote: Tue Mar 31, 2020 10:21 am Would be very interested in getting tested (once an accurate antibody test is sufficiently widespread enough for me to warrant being tested):

As I mentioned upthread our Chinese CEO looked at importing Chinese kits. He must have bought a trial pack and sent them to various members of senior management. As I said my line manager (who I was car sharing with) came back from Boston with a persistent cough after meeting someone who was later diagnosed with COVID19 meaning me and him have had to self-isolate since the beginning of March.

He’s better now and used two of the kits today which came back negative on both IgG and IgM which basically leaves us no further forward, either the kits are sh.t or he never had it.

He was a bit deflated as ideally had he come back positive (and we could trust it) he would know it was behind him and get on with his life, as it is he’s stuck in the same boat as the rest of us.

El Pollo Diablo wrote: Tue Mar 31, 2020 3:07 pm Italy's numbers do look to be slowing down, even if they're still getting horrendous numbers per day. If they've passed the peak then they'll end up probably somewhere around 18k-20k deaths in this phase of the pandemic.

Today's number of new total cases is similar (but not the same) as the number of new total cases yesterday, i.e., it didn't obviously go down again.

shpalman wrote: Tue Mar 31, 2020 4:32 pm

El Pollo Diablo wrote: Tue Mar 31, 2020 3:07 pm Italy's numbers do look to be slowing down, even if they're still getting horrendous numbers per day. If they've passed the peak then they'll end up probably somewhere around 18k-20k deaths in this phase of the pandemic.

Today's number of new total cases is similar (but not the same) as the number of new total cases yesterday, i.e., it didn't obviously go down again.

If Italy has flattened the curve and R0 is about 1 then there may be a long plateau at the top before we see the numbers trend down consistently.

This collection of COVID-19-related reports prepared for the U.S. Congress may be of interest:
https://crsreports.congress.gov/Content ... vid19.html

Woodchopper wrote: Tue Mar 31, 2020 5:05 pm

shpalman wrote: Tue Mar 31, 2020 4:32 pm

El Pollo Diablo wrote: Tue Mar 31, 2020 3:07 pm Italy's numbers do look to be slowing down, even if they're still getting horrendous numbers per day. If they've passed the peak then they'll end up probably somewhere around 18k-20k deaths in this phase of the pandemic.

Today's number of new total cases is similar (but not the same) as the number of new total cases yesterday, i.e., it didn't obviously go down again.

If Italy has flattened the curve and R0 is about 1 then there may be a long plateau at the top before we see the numbers trend down consistently.

The lockdown might have a second lag phase as the infections work through households.

lpm wrote: Tue Mar 31, 2020 2:10 pm Here was today's cabinet meeting.



I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

Rees Mogg never not looks like a tw.t.

Anyway - did you give them the benefit of your wisdom? And can you give us the meeting password so we can all join in on the next one? Caring is sharing, after all.

Some good news for Iran - European countries have finally managed to get their trade vehicle working to sell them medicine. https://www.independent.co.uk/news/worl ... 37961.html

On this occasion the supplies aren't related to coronavirus, but any extra equipment has got to help. It seems likely that the US sanctions (and European tardiness with help) will have contributed to a lot of unnecessary deaths in Iran.

lpm wrote: Tue Mar 31, 2020 2:10 pm Here was today's cabinet meeting.



I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

Umm, no you didn't. Top right of the screen shows "Cabinet Room (Host, me)", so its a screen shot from whoever was running the meeting from the Cabinet Room.

On the other hand, I am surprised they are using Zoom.
ProtonMail Blog: Using Zoom? Here are the privacy issues you need to be aware of Posted on March 20, 2020 by Richie Koch

Pucksoppet wrote: Tue Mar 31, 2020 5:47 pm

lpm wrote: Tue Mar 31, 2020 2:10 pm I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

Umm, no you didn't. Top right of the screen shows "Cabinet Room (Host, me)", so its a screen shot from whoever was running the meeting from the Cabinet Room.

You got me. Everyone else but you thought I was on the call.

badger wrote: Tue Mar 31, 2020 5:35 pm And can you give us the meeting password so we can all join in on the next one? Caring is sharing, after all.

Password, ha ha. Like they'd have remembered to use a password.

Pucksoppet wrote: Tue Mar 31, 2020 5:47 pm

lpm wrote: Tue Mar 31, 2020 2:10 pm Here was today's cabinet meeting.

<image excised>

I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

Umm, no you didn't. Top right of the screen shows "Cabinet Room (Host, me)", so its a screen shot from whoever was running the meeting from the Cabinet Room.

On the other hand, I am surprised they are using Zoom.
ProtonMail Blog: Using Zoom? Here are the privacy issues you need to be aware of Posted on March 20, 2020 by Richie Koch

Also

The Intercept: Zoom Meetings Aren’t End-to-End Encrypted, Despite Misleading Marketing (2020-03-31)

AMS wrote: Tue Mar 31, 2020 5:29 pm

Woodchopper wrote: Tue Mar 31, 2020 5:05 pm

shpalman wrote: Tue Mar 31, 2020 4:32 pm
Today's number of new total cases is similar (but not the same) as the number of new total cases yesterday, i.e., it didn't obviously go down again.

If Italy has flattened the curve and R0 is about 1 then there may be a long plateau at the top before we see the numbers trend down consistently.

The lockdown might have a second lag phase as the infections work through households.

I'm more worried about a delayed phase towards the south of Italy, and the infection working through care homes.

Como apparently had twice as many deaths in March this year compared to March last year.

lpm wrote: Tue Mar 31, 2020 6:24 pm

Pucksoppet wrote: Tue Mar 31, 2020 5:47 pm

lpm wrote: Tue Mar 31, 2020 2:10 pm I joined as "iPhone", at the bottom of the list on the right. Was easy to join - meeting number 539-544-323. Now I've got the names of the other joiners, I'll be able to evesdrop even more easily next time.

Umm, no you didn't. Top right of the screen shows "Cabinet Room (Host, me)", so its a screen shot from whoever was running the meeting from the Cabinet Room.

You got me. Everyone else but you thought I was on the call.

Well, given the problem of people guessing Zoom meeting numbers and crashing meetings with unsavoury images, it was not beyond the bounds of possibility. Also, with a bit of careful editing, a plausible fake image is relatively easily produced.

But it would not surprise me if the Cabinet decided to use Zoom without running it past security advisers first. It's the kind of detail they would not normally bother thinking about.

Don't know if we've already had it but this is quite a useful way of conceptualizing the numbers: plotting the number of new cases in the past week vs. the total number of confirmed cases.

shpalman wrote: Tue Mar 31, 2020 7:55 am A preprint from Imperial College is Estimating the number of infections and the impact of nonpharmaceutical interventions on COVID-19 in 11 European countries.

It estimates that about 10% of the Italian population is infected, for example, or about 3% of the UK.

James Annan makes the point that unconstrained R0 was very high and might be quite difficult to get/keep below 0

https://julesandjames.blogspot.com/2020 ... om-ic.html

PeteB wrote: Wed Apr 01, 2020 8:36 am

shpalman wrote: Tue Mar 31, 2020 7:55 am A preprint from Imperial College is Estimating the number of infections and the impact of nonpharmaceutical interventions on COVID-19 in 11 European countries.

It estimates that about 10% of the Italian population is infected, for example, or about 3% of the UK.

James Annan makes the point that unconstrained R0 was very high and might be quite difficult to get/keep below 0

https://julesandjames.blogspot.com/2020 ... om-ic.html

I know... and the original UK policy was to let things run and then step in calmly to flatten the peak in July, or something. As everyone now knows, the peak arrives much sooner than that if you actually look at the data rather than using your "thousands of lines of undocumented C" from 13+ years ago, and it requires all the measures just to get an R0 which might be less than 1 but at least isn't that much more than 1. I was looking at the Italian data a month ago and getting an R0 of around 10 and a peak in mid April and I'm just some c.nt with 70 lines of documented python.

(My R0 of around 10 wasn't valid because of the long recovery time, in which an infected person doesn't actually infect that many other people because they're quarantined once their infection status is known; you would need a Susceptible->Exposed->Infectious->Quarantined->Recovered model in which you fit Q to the official number of cases, and you would probably need separate channels for asymptomatic/mild infectious cases which never get quarantined and never show up in the official numbers.)

PeteB wrote: Wed Apr 01, 2020 8:36 am

shpalman wrote: Tue Mar 31, 2020 7:55 am A preprint from Imperial College is Estimating the number of infections and the impact of nonpharmaceutical interventions on COVID-19 in 11 European countries.

It estimates that about 10% of the Italian population is infected, for example, or about 3% of the UK.

James Annan makes the point that unconstrained R0 was very high and might be quite difficult to get/keep below 0

https://julesandjames.blogspot.com/2020 ... om-ic.html

Less than 0 will be a challenge. So uncontagious you'd go around uninfecting people.

It was a typo obviously for 1 - but is 1 so hard? Social distancing alone knocks a chunk out, then strict isolation for people with symptoms another chunk, and then the game changer - mass testing so that people can be quarantined and anyone they met recently warned to strictly self isolate.

We think of herd immunity as being biological. The virus bumps up against immune people and starts to struggle. But like the xkcd, humans don't need to stick with just biological. We can create pseudo immunity with intelligent behaviours.

To take an extreme imaginary scenario, imagine we all wear bracelets. They flash red the moment we are infectious. They are green when we are immune. Go red and you'd run like a Cummings for home, see someone red and you jump out their way. Greens would approach reds without constraint. You would thus create herd immunity artificially, because the virus would be living in a world where 99% of people it met were green immunes. It would be extinct within a couple of weeks.

In the real world, drive through testing and antibody testing will get us quite a long way towards this imaginary scenario. Test millions per month and you easily get R0 well below 1.

The bottom axis is R0

First graph is before interventions


Second graph is with current interventions



So, even with current interventions, struggling to get below 1 - although take the point that a test to see if you have ever had C-19 would help

The covid makes it hard because of the number of asymptomatic but infectious carrier. It may well be that up to about 80% of cases never even know they have it, while passing it on to other people. And the other 20% have a couple of days of being asymptomatic but infectious before symptoms develop, so by the time they self-isolate they could already have infected other people.

Everyone must behave as if they are simultaneously infectious and everyone else is susceptible, and susceptible while everyone else is infectious.

The virus couldn't give a sh.t about whether you thought you had a good reason to leave the house.

But China got it well below 1. And South Korea.

And Italy looks to be <1 now, eventually, although only a few days into the plateau/decline. Austria too.

It takes time for measure to work - we get better at them.

For the US-ditherer in chief: Graphic of Trump's words vs spread in US

We may get better at doing the measures. Only evidence will help us design more effective (for the amount of disruption) measures. Thats not just evidence to do with transmission of the covids, but evidence of which activities are needed to keep society going, to keep people doing the measures and to stave off complete economic collapse and which aren't.