Fourier analysis of UK COVID (reported) deaths shows a periodicity that peaks quite strongly at 6 days. With peaks at half the level at 9 and 11 days. I would expect to see something going on at the 7 day frequency but it is a trough, the 14 day level barely shows.

This is based on the most recent 32 days (a restriction of Excel/FFTs) so it shows only qualitative indications, nothing more. When there are 64 day's worth of stats it may be more relevant but I suspect the 6 day periodicity to be still there. As to why will be a mystery. But if it could be reasoned, it would be a tool to reduce the numbers.

(because I am an engineer, in lock-down and I can do these things, that's why!)