View previous topic | View next topic

Coronavirus... tidbits

Page 221 of 291
Goto page Previous  1, 2, 3 ... 220, 221, 222 ... 289, 290, 291  Next

barbados
1375388.  Tue Feb 23, 2021 5:32 am Reply with quote

I fear you’re right Brock, looking at the rush to book overseas holidays last night seems to confirm, people are stupid.

 
dr.bob
1375400.  Tue Feb 23, 2021 6:42 am Reply with quote

Brock wrote:
Well, there were an awful lot of "no earlier than" dates


Which is kind of my point.

Brock wrote:
The Government has said that the decision on whether to move to the later stages will be guided by the available data.


And yet they've imposed limits on which date that will happen by. If the data show that things improve much faster than expected, the government have categorically stated that they will ignore this data and not move on to the later stages before an arbitrary date.

If they were truly being guided by data and not dates, yesterday's announcement would've said something like "This stage will be implemented once the number of hospitalisations reaches x and the number of new cases reaches y."

 
Brock
1375404.  Tue Feb 23, 2021 7:00 am Reply with quote

dr.bob wrote:
Brock wrote:
Well, there were an awful lot of "no earlier than" dates


Which is kind of my point.

Brock wrote:
The Government has said that the decision on whether to move to the later stages will be guided by the available data.


And yet they've imposed limits on which date that will happen by.


No, they've imposed no limits on when it will happen by. These are "no earlier than" dates, not "no later than" dates.

Quote:
If the data show that things improve much faster than expected, the government have categorically stated that they will ignore this data and not move on to the later stages before an arbitrary date.


The reason for the minimum five-week gap between stages is that it allows four weeks to collect and assess data, and then a week for people and businesses to prepare for the next step. Remember that there's a time-lag in assessing the effects of any measures introduced.

Quote:
If they were truly being guided by data and not dates, yesterday's announcement would've said something like "This stage will be implemented once the number of hospitalisations reaches x and the number of new cases reaches y."


This is how the Government says it will make the assessment on whether to move on to each stage:

HMG wrote:
"Before taking each step, the Government will review the latest data on the impact of the previous step against four tests. The tests are:
1.The vaccine deployment programme continues successfully.
2.Evidence shows vaccines are sufficiently effective in reducing hospitalisations and deaths in those vaccinated.
3.Infection rates do not risk a surge in hospitalisations which would put unsustainable pressure on the NHS.
4.Our assessment of the risks is not fundamentally changed by new Variants of Concern.


In other words, the Government needs to be satisfied that the previous easing of restrictions has not had an adverse effect on infection rates before moving on to the next one. There don't seem to be any actual numbers involved.


Last edited by Brock on Tue Feb 23, 2021 8:15 am; edited 1 time in total

 
crissdee
1375405.  Tue Feb 23, 2021 7:05 am Reply with quote

Brock wrote:
No, they've imposed no limits on when it will happen by. These are "no earlier than" dates, not "no later than" dates.


Which is still surely a limit, just a retrograde limit. "No earlier than" is just as much of a limit as "No later than".

 
Brock
1375406.  Tue Feb 23, 2021 7:15 am Reply with quote

crissdee wrote:
Brock wrote:
No, they've imposed no limits on when it will happen by. These are "no earlier than" dates, not "no later than" dates.


Which is still surely a limit, just a retrograde limit. "No earlier than" is just as much of a limit as "No later than".


Yes. but that's not a limit on when it will happen by. In the context of time, "by" means "no later than".

If you lend me a tenner, and I say "I'll pay you back by next Wednesday", that means "no later than next Wednesday". I think you'd have a right to be rather disgruntled if I chose to pay you back in three months' time!

 
barbados
1375407.  Tue Feb 23, 2021 7:22 am Reply with quote

It is and it isn’t criss
Imagine you setting similar limits to when you have your dinner, and you will be judged on that limit as to whether you have succeeded or failed.
Your limit is you will sit down and eat no earlier than 4pm, or you will sit down and eat no later than 6pm.
If you have a power cut (which in the terms of the virus would equate to a spike in cases) at say 3:30pm, and you don’t get to sit down and eat until 6:30pm as a result of that - which “target” have you failed to meet?

That is why the two are different- both are targets, yet only one will have people like some here coming back and saying “you failed to deliver on your promises” even though it was totally out of your control and the right thing to do in the circumstances

 
suze
1375408.  Tue Feb 23, 2021 7:55 am Reply with quote

I think the point here is that the government is acutely aware of a whole string of broken promises, and it really doesn't want any more. Especially not immediately before the various elections which happen in May.

Much as the government is saying "X will happen not earlier than date Y", I don't rule out that it might. There are votes in Mr Johnson going on television on a sunny Friday evening and saying "We've decided that the risk associated with doing Z is minimal, so I announce that Z is allowed with immediate effect. Get out there and Z away this weekend!"

What Z might be, well Mr Johnson probably does know. I suspect that he has a couple of potential Zs in his mind, and hopes to be able to use one of them in April or May.

 
dr.bob
1375417.  Tue Feb 23, 2021 9:26 am Reply with quote

Brock wrote:
There don't seem to be any actual numbers involved.


Which rather makes my point. If they were being lead by the data, they'd put hard numbers on each step. If they were being lead by dates, they'd put hard dates on each step.

Making vague statements with no quantifiable targets allows them to change their minds randomly in the future and still claim they're following their original plan. That's a perfectly valid way to run a government, but don't pretend that you're "following the data".

crissdee wrote:
"No earlier than" is just as much of a limit as "No later than".


Quite so.

 
barbados
1375420.  Tue Feb 23, 2021 9:46 am Reply with quote

dr.bob wrote:
Brock wrote:
There don't seem to be any actual numbers involved.


Which rather makes my point. If they were being lead by the data, they'd put hard numbers on each step. If they were being lead by dates, they'd put hard dates on each step.

Making vague statements with no quantifiable targets allows them to change their minds randomly in the future and still claim they're following their original plan. That's a perfectly valid way to run a government, but don't pretend that you're "following the data".

The trouble with that approach is you are relying on a number of data streams.
If you are solely working on data rather than the data over a given period, then how do you manage it - do you open up bits of the economy on a Monday, when things are catching up from the weekend, then close them again for the spike mid week while the data catches up, or do you give yourself a sensible amount of time (arbitrary timescale, but 5 weeks seems long enough to look at a trend)

 
dr.bob
1375422.  Tue Feb 23, 2021 10:08 am Reply with quote

barbados wrote:
If you are solely working on data rather than the data over a given period, then how do you manage it


What an odd thing to say. Nearly all coronavirus data these days is communicated in terms of a seven-day rolling average. This is not a new concept.

barbados wrote:
do you open up bits of the economy on a Monday, when things are catching up from the weekend, then close them again for the spike mid week while the data catches up


That would, of course, be a remarkably silly thing to do. Do you honestly believe anyone is suggesting (or would suggest) doing this?

 
Brock
1375424.  Tue Feb 23, 2021 10:15 am Reply with quote

dr.bob wrote:


crissdee wrote:
"No earlier than" is just as much of a limit as "No later than".


Quite so.


On this purely technical point, please see post 1375423.

 
barbados
1375425.  Tue Feb 23, 2021 10:18 am Reply with quote

dr.bob wrote:
barbados wrote:
If you are solely working on data rather than the data over a given period, then how do you manage it


What an odd thing to say. Nearly all coronavirus data these days is communicated in terms of a seven-day rolling average. This is not a new concept.

barbados wrote:
do you open up bits of the economy on a Monday, when things are catching up from the weekend, then close them again for the spike mid week while the data catches up


That would, of course, be a remarkably silly thing to do. Do you honestly believe anyone is suggesting (or would suggest) doing this?
Of course it would be silly.
If only they had come up with a timescale that gave some kind of overview of the trends. A few weeks at a time for instance - who knows? 5 being about a month might be a good option.

 
dr.bob
1375430.  Tue Feb 23, 2021 10:52 am Reply with quote

barbados wrote:
Of course it would be silly.


Then would you mind explaining to me why you mentioned it?

barbados wrote:
If only they had come up with a timescale that gave some kind of overview of the trends. A few weeks at a time for instance - who knows? 5 being about a month might be a good option.


This is a timescale you described above as "arbitrary".

If the government were lead by the data, they would let the data decide when criteria are being met. If they choose to impose arbitrary limits just to make sure that things are not going wrong, that is (as I said above) perfectly valid. It is not, however, being lead by the data.

That is all.

 
barbados
1375434.  Tue Feb 23, 2021 11:01 am Reply with quote

Yes but you need to know a timeframe for your data - you can't just say we'll keep looking at the data and let that decide - because as you rightly said, that is silly.
The five week review is as good an amount of time as any, you can review it weekly, fortnightly, monthly, every third Tuesday if you wish, but for data to mean anything at all, you need to be consistent in your reviews.
Unless of course you can think of a better way to measure data without having a time frame to measure it against?

 
dr.bob
1375436.  Tue Feb 23, 2021 11:29 am Reply with quote

barbados wrote:
Yes but you need to know a timeframe for your data


No you don't. You certainly need to smooth out your data to make sure it's not too spiky, hence my mention of the bog-standard 7-day rolling average which you seem to have chosen to ignore. That is entirely separate from imposing an arbitrary time frame on which to decide what the data is doing which is, as you have pointed out, arbitrary.

barbados wrote:
you can't just say we'll keep looking at the data and let that decide - because as you rightly said, that is silly.


I said no such thing. Please do not misrepresent my views. I said that changing your mind on a day-to-day basis was silly. That has precisely nothing to do with being data lead.

Please stop trying to make an argument that just isn’t there.

 

Page 221 of 291
Goto page Previous  1, 2, 3 ... 220, 221, 222 ... 289, 290, 291  Next

All times are GMT - 5 Hours


Display posts from previous:   

Search Search Forums

Powered by phpBB © 2001, 2002 phpBB Group