Are The Virus Stats Accurate?

[BilSuth]

If Joe Bloggs self-diagnoses (correctly), he does his 14 day isolation without bothering or notifying any medical authority, then returns healthily to normal activity, how can he possibly show up on the official figures?

[ichkeria]

The number of new cases in Italy has fallen fpr two days in a row, and the rate new cases and deaths in the UK has - temporarily at least - not continued to leap.
But I notice on the news they instead compared it with the situation this time last week. ....

[BilSuth]

NickorWan, I prefer to look at the UK/Scottish (similar trends) infection figures. Each daily figure is 10 days out of date (i.e. that's roughly when they actually got infected, not just today) and the consistent graph suggests that the real figure of infections today is perhaps 10 times the figure just released. Even if that is inaccurate it is still exponential stuff and cannot go on.
As I said, hopefully, distancing seems to be happening and will flatten the graph but we won't see that till April ... because each day's published infections have already happened and can't be undone.

[jim360]

It's worth noting that the UK's number of cases reported cases has dipped on Sundays and in particular Mondays. Not sure if that's just a fluke, or something to do with data gathering and processing over the weekends. Either way, it shouldn't surprise me to see another ~20% leap tomorrow.

[jim360]

Also, the UK simply hasn't done enough testing to know the full extent of the disease here. In the US, by contrast, case have exploded in the last few days, at least in part because they've finally been able to roll out extensive testing -- apparently about 90% of those tested have come back negative, meaning that about half a million tests have been carried out. In the UK, it's about 85,000 (again, only 10% or so of those tested were positive).

It's difficult, for me at least, to estimate how many of those not tested have the disease, but if we take the government's own suggested figures from last week then we are probably up to in the region of 100,000 total cases so far nationwide.

[Peter Pedant]

Zhukov who does a uni course on this
posted his lecture a few days ago

[Peter Pedant]

but warning - you need more than 1 + 1 = 2 maff

a few points have come up amongst the dross
10% mortality seems terribly high on the quoted stats - 2% more like it
the conversion rates are sky high - 22 ( low ) to 35% of the population seroconverting per day -
the lead in to death is 20 d so a 20 d lag should be built in to any conclusions about death rates or effects of restrictions

the statisticians are telling the govt that we are following Italy but laggin by 14 d - so in two weeks we will the predict be having 500 deaths a day - 50 000 cases ;active;

[Peter Pedant]

// Also, the UK simply hasn't done enough testing//

yeah I am not sure why everyone wants to test
it tells you where you are ( but we know - in the doggy doo deep) but not really what the future is

and Jims stats are about right 1/2 m done and 10% +ve

I dont know why they dont do the stats on death rates ( it is quite obvious when someone is dead) and then multiply by 50

[Khandro]

As Niels Bohr (may have) said;

"It is difficult to make predictions, especially about the future."

[douglas9401]

I am not a number, I am a FREE MAN!