I think I've got my head around the apparent discrepancy in case numbers between the PHE regional figures and the ONS. Forgive me if it's been obvious it may be my cognitive problem, but it's been bothering me for a while. The ONS is giving a country-wide average based on random swab samples taken in homes, and so quoting a straight up 1 in 2000 people. It's likely to be undersampled and so doesn't capture the peaks at all well, smoothing out those numbers greatly. They will also include samples from a lot of people who are largely isolated, working at home etc etc, as well as those who live in more rural and secluded locations. Those people just aren't important at this point in time because they are extremely unlikely to be spreaders and unlikely to be infected anyway, so could perhaps be excluded from the 'active population'. However, that's the number usually given by the BBC. As a measure it's pretty useless, but politically very convenient if the aim is to downplay the problem. PHE is capturing postcode information from real infections and averaging over a much smaller population and so will be a much truer representation of the peak numbers. Clearly the virus will spread from the peak locations where people are more congested, i.e. in the northwest at present, just as it did from the London area previously, unless strict measures are in place, so that's the number we need. I've long suspected that Government is using a much finer distribution of postcodes than PHE is actually publishing, but is keeping it to themselves in order not to scare us.