1. The chart below updates the new CV19 case numbers to include today's. As before, the general picture is of flatness, though with a slight upward drift of around 92 cases per day. The drift has got me wondering. What might it show? One possible hypothesis follows. 
2. Up until yesterday 269,598 people had been tested yielding 78,991 cases. That's an average case rate of 29.3%. We know the testing is not random & is biased toward finding a +ve result, relative to a random sample. We know also that the rate of testing has been increasing.
3. As testing is rolled out, the +ve average bias can be expected to fall. Still, to assess today's numbers, take the conservative assumption that the average of +ves from previous testing remained good for today's. Since it was reported today that an extra 12,776 people were...
4. … tested, that would imply an expectation of 3,743 +ves, almost 1,500 less than the reported +ves (5,234). So where do these extra positives come from. The hypothesis is that they reflect an extensive spread of infections among the population. How extensive a spread?
5. Well, if the expectation had been right there would have been about 9,000 discovered -ves. The 1,500 eats into that total: t is about a sixth of it. That is therefore a possible estimate of the population spread. It is in the same ballpark as the preliminary results of ...
6. … the German Heinsburg Ab study. There are also good reasons why it might be higher: (1) it's antigen testing, so only picks up those currently infected (unlike an Ab test), (2) Germany is at an earlier stage of contagion than the UK, (3) the German results are lagged, & ...
7. … the conservative assumption made. Add test processing lags to the mix and the percentage of the population who, as of this moment, have been infected at some time or another could easily be estimated at significantly over 20%. i.e. over 12.5 million.
8. Of course we do not know these things, because no-one advising the govt. appears to have been capable of raising the thought that it might be a good idea to put in place the randomized, rolling testing that would have enabled continuous monitoring of the progress of the enemy.
