The analysis of data and conclusions drawn from it, are only as good as the validity of the data being input and propagated.
The issue some people have, myself included, is the quality of data we are being fed, and the narrative attached to it. Such as yesterday’s headline grabbing ‘96 thousand cases per day’.
Now that is way higher than all other case tracking data, such as Zoe and ONS, which are normally fairly similar, with a slight lag between them.
The modelling used would appear to be a little flawed. Some night say is this deliberate? Modelling has been fairly spectacularly wide of the mark throughout this, in SOME aspects, which is a shame, given it’s driving policy (or has been previously.... who knows now?)
Of course, that doesn’t mean the 96 thousand cases per day is correct, or won’t be correct on the future. But if it isn’t and if it is wildly out, then why? And why is it given such far reaching audience?
It's 100% absolutely not unrealistic to be on 96,000 new case per day, we could be there easily.
We have a 3 day moving average of 320 deaths per day on 29th Oct, it takes about 18 days to die (so look back at 11th Oct, we were averaging 12,000 cases per day), so 320 deaths per 12,000 cases, which is 2.7% fatality rate, which we know is not right. So, say it's 0.9% like I used earlier in the thread. This means that 12,000 cases is actually 36,000 cases.
Since then we've gone from 12,000 per day to 24,000 per day positives. So take that 36,000 expected cases and double it, you're now on 72,000 cases.
Now also factor in that when testing is overwhelmed, 24 hours test results are not happening, tracking and tracing gets overwhelmed, and it hits the south more, then all of these are reasons for the cases to go up sharply. It could easily have gone up 33% to to get to 96,000. Some parts of London have a suspected "R" of 3 at the minute.
So, if the fatality rate isn't 0.9%, say it's 0.6%.
2.7%/ 0.6% = 4.5
12,000 cases is actually 54,000 cases (for what you're missing)
12,000 cases on 11 Oct is now 24,000 on 29th Oct (actual positives)
So 54,000 cases is actually 108,000 cases (missed cases doubling, as actual cases have doubled)
But that would assume that we don't get more cases when we get overwhelmed, or that 24 hour tests are nt better than 5 day test results and that there isn't a higher "R" in more densely populated areas.
That's how you get 96,000 cases, it's extremely easy.