slim_boy_fat
Member
- Messages
- 29,532
- Location
- Scottish Highlands
Remember "There are three kinds of lies - lies, damned lies, and statistics" according to Disraeli. 
- Thread starter Johnnybravo
- Start date
Morrisman
Forum Supporter
- Messages
- 8,931
- Location
- Staffordshire, England
Interesting article on the web, about statistics modelling and covid etc:Remember "There are three kinds of lies - lies, damned lies, and statistics" according to Disraeli.
If you thought that Covid modelling, which predicts the spread of the virus was inaccurate and flaky, think again. It's arguably dishonest. In a majestic piece of journalism, Fraser Nelson from the Spectator magazine had a Twitter exchange with one of the key modellers advising the government – Graham Medley, the chair of the Sage modelling committee. This whole exchange has been written up into an article in the magazine, and it’s well worth a read.
In this exchange Mr Medley seemed to suggest that he has been given a very limited brief, and was asked to churn out worse-case scenarios without being asked to comment on how plausible they are.
“We generally model what we are asked to model” said Mr Medley. “There is a dialogue in which policy teams discuss with the modellers what they need, to inform their policy”.
Fraser nelson asks whether this remit could mean leaving out just-as-plausible, quite important scenarios that would not require lockdown? His reply:
“Decision-makers are generally on only interested in situations where decisions have to be made.......
Mark Dolan: Let’s not cancel Christmas - let’s cancel the doomsday Sage scientists
The Court of King Boris is crumbling, as is the case for any future lockdowns.
mtt.tr
Member
- Messages
- 8,458
- Location
- Essex/Suffolk border
Well it has to be, what's the point in modeling for best case
mtt.tr
Member
- Messages
- 8,458
- Location
- Essex/Suffolk border
Interesting article on the web, about statistics modelling and covid etc:
If you thought that Covid modelling, which predicts the spread of the virus was inaccurate and flaky, think again. It's arguably dishonest. In a majestic piece of journalism, Fraser Nelson from the Spectator magazine had a Twitter exchange with one of the key modellers advising the government – Graham Medley, the chair of the Sage modelling committee. This whole exchange has been written up into an article in the magazine, and it’s well worth a read.
In this exchange Mr Medley seemed to suggest that he has been given a very limited brief, and was asked to churn out worse-case scenarios without being asked to comment on how plausible they are.
“We generally model what we are asked to model” said Mr Medley. “There is a dialogue in which policy teams discuss with the modellers what they need, to inform their policy”.
Fraser nelson asks whether this remit could mean leaving out just-as-plausible, quite important scenarios that would not require lockdown? His reply:
“Decision-makers are generally on only interested in situations where decisions have to be made.......
Mark Dolan: Let’s not cancel Christmas - let’s cancel the doomsday Sage scientists
The Court of King Boris is crumbling, as is the case for any future lockdowns.www.gbnews.uk
Haha from GB news ok then, no agenda.
This is actually a particularly easy thing to model once you've got the base.
I did first time round using icl stats. It was correct with 10% day on day so they do know what they are talking about.
eLuSiVeMiTe
Member
- Messages
- 16,441
- Location
- ......
Often the best way mate
I'd rather tell someone it's gonna take 3 days and be done in two than tell them two and be there for 3 days.
Newlands
Member
- Messages
- 784
- Location
- Sussex, UK
The exact meaning of the terminology is often completely ignored when these statistics are bandied about. 10,000 confirmed cases would mean that there are 10,000 people who have had some sort of medical intervention - a call to a GP, visit to AandE or a hospital admission, for example. 200,000 cases a day refers to the best estimate and would be based on data collected by all sorts of means, including testing for the virus at sewage treatment works, collection of the data returned by those testing at home, and data collected by postal surveys (my other half was recently contacted to do a random test not long ago).
We know that statistical results are never 100% accurate, but we can work out a level of confidence, and they are a lot better than a wild stab in the dark. The problem with these results is that they are misinterpreted, sometimes deliberately, to suit an agenda. A typical example is the notion that the vaccine doesn’t work because of the high number of vaccinated people who contract covid. If you take this to the extreme and have a 100% vaccinated population, then every single covid case would be a vaccinated patient - to say vaccines therefore don’t work is the conclusion reached by those who don’t understand the data or the disease.
We know that statistical results are never 100% accurate, but we can work out a level of confidence, and they are a lot better than a wild stab in the dark. The problem with these results is that they are misinterpreted, sometimes deliberately, to suit an agenda. A typical example is the notion that the vaccine doesn’t work because of the high number of vaccinated people who contract covid. If you take this to the extreme and have a 100% vaccinated population, then every single covid case would be a vaccinated patient - to say vaccines therefore don’t work is the conclusion reached by those who don’t understand the data or the disease.
Seadog
Save the planet. It's the only one with rum!
- Messages
- 12,846
- Location
- NE London - UK
Quite right. "Statistical figures" are usually (always) presented with no reference point. For example, "There was an 80% increase in infections over the last week." It sounds bad, but the piece of information missing from that statement is that there were only six people infected to start with. A figure with no reference is worthless. It's a technique as old as statistics.
- Messages
- 23,220
- Location
- Birmingham
My sister and her partner both take regular LF tests due to working at a school, when he got the symptoms he took a PCR test which came back positive, she felt unwell a day later so also took a PCR which came back positive too, the odd thing is she continued the routine of LF tests which still showed a negative result, it was 4 LF tests after the positive PCR that one gave a positive reading.
It's not making much sense to me.
At the start we were shielding and even left parcels in the garage for two days before opening. Now all the hand washing has gone and we wash hands when dirty etc.
So blaze have we become the only thing we do is wear masks and not go to.pubs etc.
BUT in the last 3 weeks the whole family has had Noro virus and then 2 weeks later ended up with a virus. Head cold, very sore throat, nasty chesty hacking cough that goes for hours and returns.
Strange thing is the symptoms sound like Omicron but my grandson has had a PCR and my son has had 3 tests as they are falling like flies with covid (don't know what type) at his work place.
His brother in law and his girlfriend have tested positive for covid . But anyone who has been doing lateral flow tests has always shown up negative.
So how is not and this other virus getting round so easily but covid seems to be not as bad (yet)
Yes that seems to be the case.
Lateral flow tests are slow to detect .
Chances of getting several negatives but actually being positive makes them 1. next to useless 2. Quite dangerous 3. An expensive white elephant.
You choose.....
Add to that the lying and deceitfulness going back to Barnard Castle and is it any wonder a large majority will ignore any measures they put in place next.Government statistics state that the new strain is very contagious and 200,000 people a day are catching the latest it....
......but in the next breath states there are 10,000 confirmed cases.
And they wonder why people are getting disillusioned with it all.
Morrisman
Forum Supporter
- Messages
- 8,931
- Location
- Staffordshire, England
Doesn't matter the source, as someone will always ridicule it, whether it's Al Jazeera, Daily Wail, or New York Times.
However, it is a copy of what some senior bod has actually said, so makes sense.
Here's a link to an article that was published four days before the UK went into first lockdown:
tomaspueyo.medium.com
There is much in the article that is still relevant today.
Coronavirus: The Hammer and the Dance
What the Next 18 Months Can Look Like, if Leaders Buy Us Time
There is much in the article that is still relevant today.
mcostello
Member
- Messages
- 910
- Location
- Lancaster, (the other one) Ohio, USA
Just been to the Doc today. He said two weeks ago the whole office had the covid. They seem to got it from a worker who was totally up on all covid shots. This was the SECOND time He has had it. He said herd immunity is not working.
But that's the folly of their argument.
You can still spread covid DESPITE being vaccinated. The only thing being vaccinated does (hopefully) is prevent that person being so ill they need hospitalisation.
Burdekin
Chief Bodger
- Messages
- 6,685
- Location
- Aberdeen
There won't be herd immunity as you can still catch it if you've had it before or have the vaccine. You should be less ill though so hopefully it is less deadly/serious as time goes by.
keithski122
Member
- Messages
- 2,023
Were they off due to illness or off because they had to isolate?
MattF
Forum Supporter
- Messages
- 17,626
- Location
- South Yorkshire
I've given up trying to impress that fact on people.
