I think the UK is turning around

UK:

I spent some time yesterday trying to see if there was a noticeable pattern between lockdown dates and flattening off in the rate of new cases across various countries. Thanks to a handy analysis published by Imperial College London on lockdown dates, I was able to analyse countries comparable to the UK in terms of size and treatment of coronavirus.

This gave me the following table:

What the data suggests is that on average it takes about 9-10 days from the start of the lockdown for the rate of new cases to flatten, the plateau then last for a further 9 days before the first noticeable signs of declining new cases.

I'm using some smoothing algorithms to look at the data and judging the plateau start and end by visual inspection of the chart (aka I'm eyeballing it!). But the patterns are there is you look carefully.

Admittedly this is a small sample (n=6 and n=3) but the similarity in the number is all too plain to see, so we can at least regard these numbers at least as indicative of what happens and when - there will be other factors at play as well, but these numbers fit well and reflect the sort of incubation period lengths as well.

So, if we apply the same date range to the UK, what do we get?

Well lockdown officially started on March 24th, add nine days and you get April 2nd. Add a further 9 days and you get April 11th for the start of the decline. Now let us look at the daily new cases chart below in more detail.

Apart from the outlier on April 5th, we have been on a plateau (my yellow highlight for emphasis) of new daily cases since April 1st (April 2nd with smoothing), which is very encouraging. 

Now, I'm the first to admit that we have had plateaus in new cases before, but previously these have lasted only 4 or five days, whereas the one we are on at the moment has lasted for 7 days so far. If you take an average over those days you get just under 4300 cases per day, so there is sufficient dips to compensate for the outlier as well, which is another good sign. 

So we have two pieces of evidence suggesting that this might be the plateau - the lag since lockdown and the length of the plateau. This, I believe, is fairly strong evidence to be optimistic, but we await the next few day's figures on new cases with interest.