Scrutable

lpm wrote: ↑

Sat Oct 03, 2020 7:52 pm

No UK stats today?

Just been updated.

I was braced for a high number due to reading something 20 minutes ago about "the total reported over the coming days will include some additional cases from the period between 24 September and 1 October" but

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

This is what the daily (and 7 day average centred on the date) death numbers look like for the UK - for anyone still saying that it's just increased testing and deaths aren't climbing.

Data from ECDC website

https://www.ecdc.europa.eu/en/publicati ... -worldwide

368 patients in ventilator beds (as of yesterday) - was 77 three weeks ago.

2428 in hospital (as of two days ago) - was 726 three weeks ago.

shpalman wrote: ↑

Sat Oct 03, 2020 8:15 pm

368 patients in ventilator beds (as of yesterday) - was 77 three weeks ago.

2428 in hospital (as of two days ago) - was 726 three weeks ago.

Yes and the likes of Ivor Cummins (seemingly previously a fad diet promoter) and @alistairHaimes (who's blocked me) are still denying this.

I am morbidly intrigued how they will explain away the consequences of what's already locked in in the UK, for example.

So we're very roughly in the same place as when the lockdown started (too late) in March? The main differences seem to be lower R (1.4 v 3ish), and a different regional distribution.

jimbob wrote: ↑

Sat Oct 03, 2020 8:03 pm

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

This is what the daily (and 7 day average centred on the date) death numbers look like for the UK - for anyone still saying that it's just increased testing and deaths aren't climbing.

Data from ECDC website

https://www.ecdc.europa.eu/en/publicati ... -worldwide

Screenshot 2020-10-03 210133.png

Nicking that graph.

AMS wrote: ↑

Sat Oct 03, 2020 9:56 pm

So we're very roughly in the same place as when the lockdown started (too late) in March? The main differences seem to be lower R (1.4 v 3ish), and a different regional distribution.

There will be fewer "deaths in 2 weeks per case now" than in March, because to become a case (with a positive test) in March typically you had to be pretty ill. We can't be sure what the multiplier of the number of cases to produce each fatality will be, but there will be a component for a younger age profile of infectees (if only because older people are taking more care) and another for more testing, meaning that many more cases will have only very mild or even no symptoms.

The scary thing is that with complacency, including a lack of scenes of ICU meltdown, the prevalence of the virus may rise to a point where widespread infection of vulnerable groups becomes statistically inevitable.

Grumble wrote: ↑

Sat Oct 03, 2020 10:00 pm

jimbob wrote: ↑

Sat Oct 03, 2020 8:03 pm

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

This is what the daily (and 7 day average centred on the date) death numbers look like for the UK - for anyone still saying that it's just increased testing and deaths aren't climbing.

Data from ECDC website

https://www.ecdc.europa.eu/en/publicati ... -worldwide

Screenshot 2020-10-03 210133.png

Nicking that graph.

Feel free, I have most of those graphs on my pinned tweet @parkinjim

sTeamTraen wrote: ↑

Sat Oct 03, 2020 11:40 pm

AMS wrote: ↑

Sat Oct 03, 2020 9:56 pm

So we're very roughly in the same place as when the lockdown started (too late) in March? The main differences seem to be lower R (1.4 v 3ish), and a different regional distribution.

There will be fewer "deaths in 2 weeks per case now" than in March, because to become a case (with a positive test) in March typically you had to be pretty ill. We can't be sure what the multiplier of the number of cases to produce each fatality will be, but there will be a component for a younger age profile of infectees (if only because older people are taking more care) and another for more testing, meaning that many more cases will have only very mild or even no symptoms.

The scary thing is that with complacency, including a lack of scenes of ICU meltdown, the prevalence of the virus may rise to a point where widespread infection of vulnerable groups becomes statistically inevitable.

Yeah it's difficult to compare because back then the official figures probably missed about half the deaths and ~90% of the cases.

I only have hospitalization figures from 27th March onwards, on which day it was 7043 but doubled in a week. I have "in ventilator beds" numbers starting from the second of April at 1813. A week later the UK started to reach its inflection point, with nearly 20,000 in hospital and 3300 on ventilation. That was in the phase that the number of new cases per day went from about 2000 to 4000. (I won't call it a peak, rather it's the time period in which covid patients were dying faster than new subjects were being infected. Maybe they were also recovering but we don't know because the UK had no protocol for recovery and to be honest it still doesn't.)

The point is that the official figures don't go far back enough to see similar numbers for those metrics as the UK has now.

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

Delayed today too by the look of it.

shpalman wrote: ↑

Sun Oct 04, 2020 6:02 pm

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

Delayed today too by the look of it.

"We are updating the data…

The process takes approximately 15 minutes to complete. Please do not refresh the website until 10:38pm."

That would be 10:38pm my time, so 9:38pm for you lot.

"The cases by publish date for 3 and 4 October include 15,841 additional cases with specimen dates between 25 September and 2 October — they are therefore artificially high for England and the UK."

That's why today's number is 22,961. It's ok, it's only ~2000 extra cases per day for a that week.

WTF?

There were 12,000 cases Friday, they reported only 7,000?

You couldn't be this wrong, even if you tried.

I never thought I'd say this. But this government is proving to be incompetent.

There are currently 11,404 cases with a specimen date of the 30th of September. I doubt that they've even processed all of them yet so it's only going to increase.

For context, during April the number of cases per day was usually between 4000 and 5000.

I think, at a quick look, the doubling time is still about 12 days. Faster than 14 days. But nowhere near doubling every 7 days.

The hopes of a flattening off were illusory.

shpalman wrote: ↑

Sun Oct 04, 2020 8:32 pm

shpalman wrote: ↑

Sun Oct 04, 2020 6:02 pm

lpm wrote: ↑

Sat Oct 03, 2020 7:59 pm

12,872.

Delayed because of a f.ck up - under counted previous days.

Delayed today too by the look of it.

"We are updating the data…

The process takes approximately 15 minutes to complete. Please do not refresh the website until 10:38pm."

That would be 10:38pm my time, so 9:38pm for you lot.

"The cases by publish date for 3 and 4 October include 15,841 additional cases with specimen dates between 25 September and 2 October — they are therefore artificially high for England and the UK."

That's why today's number is 22,961. It's ok, it's only ~2000 extra cases per day for a that week.

Definite spike in the graphs...
https://coronavirus.data.gov.uk/cases

What's harder to do is to compare the "by specimen date" data from previous days to the data now.

I have that data for Lincolnshire though, saved in my spreadsheet.

A couple of days ago there were 20 cases reported for the 30th of September, now there are 45. And 58 for the 1st of October when yesterday's data said 29.

Whilst looking at whatever dross Ivor Cummins has been spouting I see the claim that hospitalisations in Ireland are slowing down their rate of increase and have been since August

https://twitter.com/LGradaigh/status/13 ... 7471759365

Growth rate of hospital confirmed cases has been in decline since beginning of Aug. This is not a reason to panic. Why is NPHET distorting the facts?
@FatEmperor
Quote Tweet

With a nice log-scale regression plot, which I'm not entirely sure is about - partly due to the complete lack of explanation what "x" is
That isn't what the ECDC data says

https://t.co/XnjPbP4uYd?amp=1
Rarely have I seen such a perfect example of exponential growth

FOR THE LOVE OF GOD CAN WE JUST HAVE SOME GOOD NEWS FOR ONCE PLEASE, f.cking HELL




ahem as you were

discovolante wrote: ↑

Sun Oct 04, 2020 9:13 pm

FOR THE LOVE OF GOD CAN WE JUST HAVE SOME GOOD NEWS FOR ONCE PLEASE, f.cking HELL




ahem as you were

Actually there is a smidgen from the covid.joinzoe.com app data suggestingthat the rate of increase is now slowing

jimbob wrote: ↑

Sun Oct 04, 2020 9:43 pm

discovolante wrote: ↑

Sun Oct 04, 2020 9:13 pm

FOR THE LOVE OF GOD CAN WE JUST HAVE SOME GOOD NEWS FOR ONCE PLEASE, f.cking HELL




ahem as you were

Actually there is a smidgen from the covid.joinzoe.com app data suggestingthat the rate of increase is now slowing

Ejg7jAJWsAIQ6nY.jpeg

No offence Jim but that gradient looks like when you've been walking almost vertically up a hill for about two hours and you finally get to a marginally less steep bit only for the wind to change direction and start blowing directly into your face.

discovolante wrote: ↑

Sun Oct 04, 2020 9:57 pm

Actually there is a smidgen from the covid.joinzoe.com app data suggestingthat the rate of increase is now slowing

Ejg7jAJWsAIQ6nY.jpeg

No offence Jim but that gradient looks like when you've been walking almost vertically up a hill for about two hours and you finally get to a marginally less steep bit only for the wind to change direction and start blowing directly into your face.
[/quote]

I always find it difficult to get very excited about arguments along the lines of "rate of increase in the increase is starting to increase less rapidly". For some sufficiently large N, the Nth order derivative is always negative, but it's the zero-order number (i.e., people dying) that concerns most of us.

I agree, it's not much, and 8% a day isn't brilliant either but it's better than nearly 11%

Woodchopper wrote: ↑

Sun Oct 04, 2020 8:53 pm

shpalman wrote: ↑

Sun Oct 04, 2020 8:32 pm

shpalman wrote: ↑

Sun Oct 04, 2020 6:02 pm

Delayed today too by the look of it.

"We are updating the data…

The process takes approximately 15 minutes to complete. Please do not refresh the website until 10:38pm."

That would be 10:38pm my time, so 9:38pm for you lot.

"The cases by publish date for 3 and 4 October include 15,841 additional cases with specimen dates between 25 September and 2 October — they are therefore artificially high for England and the UK."

That's why today's number is 22,961. It's ok, it's only ~2000 extra cases per day for a that week.

Definite spike in the graphs...
https://coronavirus.data.gov.uk/cases

https://twitter.com/Dr_D_Robertson/stat ... 69187?s=20

All outstanding cases were immediately transferred to the contact tracing system by 1am on 3 October and a thorough public health risk assessment was undertaken to ensure outstanding cases were prioritised for contact tracing effectively."
Dr Duncan Robertson
@Dr_D_Robertson
·
1h
The implications are that nearly 16,000 people were not contacted by NHS Test and Trace until this error was identified. Some of these cases go back to 25 September.

This is not good

jimbob wrote: ↑

Sun Oct 04, 2020 10:51 pm

I agree, it's not much, and 8% a day isn't brilliant either but it's better than nearly 11%

Yes, although if it is dependent on people's behaviour then we may expect people to tire of ongoing restrictions over the coming weeks, especially as we get closer to Christmas. I realise I have no expertise in this at all but to be honest it seems marginal.

Soz I know you were trying to help, it's not your fault, it's reality that's the problem.