I know we are tired of Covid19 posts but this is worth a separate one....

You state you are a Data analyst fine. A large part of my job in my last few years working in IT was also of a similar nature- IMHO, you are focussing on assumptions to support your own agenda i.e. testing is not the panacea (Did you read post 25?). but crucially you are making incorrect assumptions to how I interpreted the outcome of the data You also keep referencing South Korea which is irrelevant to the situation the World at large is now in.

To be clear. ALL I got from the report was that 9 out of 10 infected turned out to be asymptomatic - nothing more nothing less. If you project that onto the population at large (and I admit there are many caveats) it therefore follows that far fewer people who catch the virus will be hospitalised and so when much of the population has unknowingly been exposed to it the impact on the Health service (and death rate will not be quite as bad as originally projected simply because they could not factor in a ration of Asymptomatic: Symptomatic victims. (i.e. known unknowns) - The Vo results strongly suggest there is a considerable proportion of population with natural immunity. Again I am NOT misintepreting the results (and it would appear most on here, and elsewhere, agree).

Whilst we can all agree, Testing in certain areas is vital, Mass testing is not and is a distraction and moreover, deflects resources from what IS important i.e Care for those affected, protecting the economy and finances of industry, businesses and individuals for post Pandemic life.

Several week ago I made an embarrassing basic error regarding a simple percentage calculation for which I was rightly ridiculed. I had the sense to accept that and readily acknowledged I was wrong. Whilst this situation is not quite as black and white, claiming superiority and comparing yourself favourably to the "population at large" with an "I am right everyone else is wrong" attitude is frankly, arrogant, and merely digging yourself a deeper hole. My advice is, when you are in a hole -stop digging.

 

You're confusing actual death rate with reported death rate and you're correct. Because we don't test many the reported death rate will appear much higher but the actual death rate won't be and tbh I care more about how many people die than what percentage the UK government erroneously think it is due to inaccurate testing policy.

The UK failing to test won't out half a million bodies in the ground

 

Considering your job it's worrying that you refuse to acknowledge the fact that only 10% have symptoms and that's why our figures are much lower and the death rate LOOKS much higher. Can't really believe that you don't understand why it's important to realise that the many untested and none dying people are important to include in statistics when panicking people and telling them that half a million will die.

 

Agree. In my last post I mentioned 'known unknowns' ANY data analyst knows that without a complete set of relevant data, conclusions regarding outcomes are guesswork. My experience of data gathering e.g. led me to conclude mere data is worthless. Information extrapolated by data analysts into meaningful reports should clarify the muddle. Unfortunately Governments, in particular (though not uniquely) are often guilty of wanting excessive amounts of data. This enables them to cherry pick and support facts to suit their argument (rather like Donny-Tyke appears to be doing here). I could give specific examples but the OS Act prevents me from doing so.

 

Again; dragging down rabbit holes of your own making.
My simple point which you’re trying to turn into something I haven’t said is that trying to ascertain anything from the results of a heavily tested village in Italy and make an assumption for the UK is ridiculous.

if you believe I have posted more than that, I apologise for not being clear.

 

Again the 'heavy' testing is irrelevant other than it has provided data (albeit limited) on a previous unknown i.e. the aforementioned ratio of asymptomatic/symptomatic victims. Wat would be valuable in the UK to carry out a test on a cross section of the UK public as presumably those who were or are infected without symptoms will have COvid19 antibodies present. Then the Vo results can either be confirmed or discounted. I also am concerned that as a data analyst you seem blinkered ....and also please stop the use of annoying corporate speak (it is not big and it is NOT clever) .... 'rabbit holes', 'throwaway points' etc.

What next ...'helicoptering above the situation' to which our response in meetings if anyone ever said it was ' in the hope the downdraft blows the problems away? OT....(We used to play bull*hit bingo on conference calls)

 

Read what you posted yourself above.

I’ll repeat; the ‘death rate’ is exactly what it is depending on the question asked.

So:
The death rate in the UK is ‘as a percentage of those diagnosed’. And the answer is whatever today’s update says.

The death rate in a single village in Italy ‘is a percentage of the whole population, whilst held in isolation and monitored’, and the answer is whatever was the outcome of that.

it’s simply incorrect to:

• Extrapolate the Italian figure across the UK population.
• Suggest that the Italian figure is more accurate - simply because the question is different.

Once more: all statistics answer a specific question; change the question and the answer needs fresh data. It’s a lot less complicated than people are trying to make it.


it’s odd you accuse governments of

 

Only if you take the data out of context.

The government aren’t pretending that x% of the population will die based on the current figures.

And whilst it’s interesting (that word again) the problem with the Italian data is that it shows the outcomes within a heavily monitored small population.

And as everyone keeps pointing out, the UK can’t afford to mass test, so we can’t hope to replicate their outcomes either.

like counting road deaths if all journeys were heavily monitored by the police.

 

Just give it up guys...Donny-Red is right and everyone else is wrong....well in his World at least.

 

You are completely misinterpreting everything everyone is saying and I'm not sure why. You seem intent to keep saying "it depends on the question asked". I'll tell you what the question is. It's how many people are going to die. Not how many people named Doris who have pink hair and have coughed every 12 seconds.

If you insist on ignoring the stats for the whole population then you may as well narrow it down further and say that 99% of those who go into respiratory arrest due to covid-19 will die therefore it is a good idea to shout out to the public that the death rate from covid-19 is 99% because it depends on the question asked. It may depend on that to doctors who are treating people and have to judge their treatment based in their symptoms but to the public bi that doesn't matter. What matters is how many will die out of everybody

 

Some good points on all sides. IIRC the cruise ship tested everyone and found 50% of those had no symptoms. Useful but perhaps not completely reliable, yet they do lend to the theory that as with regular flu many people can have it but show no symptoms.

Testing is important and the stats can be used to show two sides of the same equation.

It can show 'after the event' numbers as they had in this town (what was the population BTW?) in that we know the outcome of a specific place becoming infecting and then finding out the results in terms of deaths, infections, people not showing symptoms etc..

Testing can be used during the event to isolate infected people showing no symptoms and thus reduce further infection.

 

And the answer to that is dependant on all the factors I posted earlier.

whilst you might want it to be a simple % that can be found in some unrelated data from Italy, I’m sorry but it can’t*.

of course you’re free to believe whatever you like, and to do sums based on the data from Italy. Knock yourself out, but you can’t demand others believe it too. And if you post about it, you can expect to be challenged.

*if the data was sound, the experts would be shouting it from the rooftops.

 

One issue of isolation without testing is that you can group vulnerable people together with asymptomatic people, thus increasing infection in households. I don't know enough about Italian society to say beyond basic stereotypes but there may be some social reasons, as above, as to why it's spread so much faster.

It could also be blind bad luck. There will no doubt be many a study to explain it. I hope the isolation kicks in soon and those death rates fall ASAP.

Tragic stuff

 

You'll notice that I said 'if the data is correct'. You just refused to give any credence to it at all saying that it isn't answering the same question as how many will die because to answer that you have to ignore the asymptomatic cases

 

Even in Italy 87% of deaths have been people over the age of 70. It just so happens that sadly they have a lot of people aged over 70.
They also use a system whereby if you have covid-19 and you die then you are listed as dying due to it whether it's actually the cause or not. 78% of death certificates haven't actually show coronavirus as the cause of death but because they tested positive for it it is counted in figures as a death from it which is misleading.
Then there's the fact that on the whole they are only testing symptomatic people and not asymptomatic which makes the mortality rate appear much higher.
Italy also has more people per household than the UK and properties are closer together with more apartments that have communal areas.

 

I never suggested the data was ‘incorrect’.

I thought it was a simple concept, it appears not. Sorry to have confused so many

 

To be perfectly honest I've enjoyed the debate and the differing explanations and thinking and interpretation of the data. Just because I don't really understand your point doesn't mean I don't appreciate you putting it across.

 

In Italy 90 tested positive and 6 were asymptomatic So that means roughly 94% have symptoms.