Quote:
Originally Posted by Sephiroth
Sure - but what we have from Downquark is an estimated startpoint and a consequential current point. The points in between are as true as all the other points. But what is the method for correction? The unknown cases, the asymptomatic cases and so on.
What factor are they using to include these unknowns and what is the basis for that factor? For example can you project the curve backward because you presumably have the function/equation from current statistics.
|
Excellent questions. You can correct the data with antibody tests and randomised testing. The antibody tests tell you if you have had the virus, because your body is keeping antibodies for it. Random testing then gives you a sample you can extrapolate to the entire population. This is currently being dogged by a lack of reliable tests (every country in the world now wants them and China is selling fakes) and issues about knowing how long you keep immunity to the virus. Some reports are saying you lose immunity fairly quickly.
If you have the R value you can project the curve anyway you want. However as mentioned as you enable or remove precautions that will alter the R value so you will get a kink in the curve as you start or stop a lockdown.
The horny gentleman from imperial college had a computer model for predicting how it changes. It has somewhat been pillared for containing bugs (of the nature I fix in my job), but they insist it is still accurate. But this will be judged by history.
What I would be worrying about is the difference between the lock down R value and the semi-lock down social distancing R value. If there is no difference we need not keep the lock down. If it shoots back up then we may have to lock down again in a few months.
I realise this is all frustrating but viruses are not easy things to understand.