beorhthelm
A. Virgo, Football Genius
- Jul 21, 2003
- 35,290
optimistic for Spain, seems they discovered resurection.
Got something to say or just want fewer pesky ads? Join us... 😊
Number of Deaths
- Thread starter Tim Over Whelmed
- Start date
optimistic for Spain, seems they discovered resurection.
Not for the first time (I commented on this previously). Some weeks ago there was a mass resurrection in Italy, too. My view is this is proof of the power of prayer, and proof that there is only One True Church.
Not for the first time (I commented on this previously). Some weeks ago there was a mass resurrection in Italy, too. My view is this is proof of the power of prayer, and proof that there is only One True Church.
Maybe we all should start prayer here ?, the figures just announced state 111 new deaths recorded yet the amount of deaths in total are announced as 39,045 this is an increase of 556, from yesterdays total of 38489, yet only 111 are recorded in accordance with the daily figures ? what gives H. ?
too late to edit but apparently they've adjusted the way they count deaths and have increased the overall total to reflect this
?Billy the Fish
Technocrat
HWT, what's the explanation for your numbers not looking any more positive than a month ago when the infections and deaths are dropping rapidly?
Previously missed historic data was added today, giving the 556.
https://www.independent.co.uk/news/uk/home-news/coronavirus-uk-death-toll-cases-nhs-hospital-latest-a9542981.html
Spain did a similar thing a week ago, but deducted 1,700 deaths.
The quotes are drawn from different data sources (apples and oranges) and the fact our government is relaxed with all this in the media is the same reason as that for the governments of Spain and Italy.
Tin pot.
Ah! that's because my numbers are ratios of deaths to cases (call it RC), and difference (D) between RCs on different days. The important thing is whether the D is positive or negative. It is all explained in various tedious posts of mine above.
The Wizard
Well-known member
- Jul 2, 2009
- 18,383
The backdating is getting silly now really, less than 100 of the deaths from today’s 179 NHS total are actually from the last 5 days, how can they be so backlogged?
NHS actual date of death reporting is a far more relevant barometer of where we are at.
The Wizard
Well-known member
- Jul 2, 2009
- 18,383
Absolutely, it seems bizzare to me there is deaths from March 24th being added now 10 weeks after. Perhaps they are going through everyone who passed away samples and checking for COVID?
A friend of mine this morning posted an angry post on Facebook that his mother had died in early April but they’ve notified him they have included COVID on the death certificate in spite of the fact she either caught it at hospital or had it asymptomatically, I wonder if other countries are doing this?
- Jul 10, 2003
- 25,777
I think there are only two consistent bases for data relating to Covid, depending on what you are doing.
If you want to look at trend data for the UK, you really need to look at the ONS data 'by date of death' released 10 days in arrears every Tuesday (and even then the last few days of that is subject to some change). If you want to compare across different countries, then use the 'Total Excess Deaths published' in the FT each day (but updated by different countries on different days).
Individual countries methods of counting (including the UK daily figures) have been subject to interpretation and change depending on what the government thinks at the time and as such, are very difficult to draw conclusions from, particularly in relation to the most recent few days/weeks.
And obviously HWT's stats which are intended to measure a different set of trends entirely.
So if there were only 7 new cases on one day and one of them died, the number would be "7" and the difference "0"?
LamieRobertson
Not awoke
I much prefer your polls
LamieRobertson
Not awoke
You’ve gone down in my estimation Harry .....sigh
LamieRobertson
Not awoke
Do you think that is just soppy bllx as per Harrys thinking?
Yeah no you’re right, as Harry’s bizarre calculations confirm, we are indeed all going to die and **** anyone who tries to stay positive through very tough times mentally.
Something like that.
AmexRuislip
Trainee Spy 🕵️♂️
No. It depends on the total number of cases and the total number of deaths. Get the ration today and compare it with yesterday. Assume that deaths occur some time (maybe 2 weeks) after infection, so if the increase in new cases will be followed some days later by an proportionate increase in the number of deaths. Thus the difference between the ratios over time will fall as deaths catched up with cases. When a peak is reaced the ratio will go negative (or positive - it is purely a convention and I can't recall which I selected - it is indicated in the spreadsheet). When the numbers of new cases steadily falls that chage in the sign will remain. What we are seeing is a much weaker indication of a fall in new cases in some countries. In the US, Sweden, UK and Brazil in particular the failure for the epidemic to truly go away is evident.
But as I said from the start, this is all 'if this, then that'. It sort of hangs together. The key info on what has happened is the excess deaths from ONS, not my silly spreadsheet. It is possible my calculation better predicts the start of new trends, but going forward a surge in new cases will be sufficient for panic.
Why did I bother with all this shit? At the start people refused to 'believe' numbers of new cases or deaths, in the UK because we were not 'testing', in Russia because they don't seem to be attributing deaths to Covid meaning they appear to be resistant to the virus, in China because people thought they were simply lying, in Iran because the numbers look so weird, and in India and Pakistan because vast swathes of poor people are simply not part of any statistics. I wouldn't have bothered faffing about if we 'trusted' the data in the beginning, or had a clear idea what was wrong with some data, what to ignore, etc., and I thought maybe there are trends in global data, signal among the noise, that I could identify.
The one thing I like about my statistic is the similarity in ratios between some countries, and the fact the ratios did fall worldwide over time, indicating that the rubric and assumptions may have had some validity.
You can put holes in it easily of course, but I'm an experimentalist as you know so the name of the game is make some assumptions, generate a hypothesis and invent a means of testing it. Here we have multiple variables that are misrepresented differently by different nations, so we needed a good signal if it was to emerge from the noise. The most clear signal is the one I note above about the ratio falling.
I have got a bit bored with this now as we are not looking at lots of curves of new cases and deaths at the start of the epidemic that all looked frightening and some mad, needing a deeper data mine. My main concerns now are the fate of Brazil, India and Pakistan, and the possibility (still entirely uncertain) whether we get a second surge in cases in the autumn. The data from some nations remains unfathomable. Oh and America, with its increasing cases and deaths in the red neck states, can **** off.
LamieRobertson
Not awoke
Lol ... dare I say it .. but back at work?
..,and Harry’s alrite