AngharadHafod @AngharadHafod@toot.wales

I've just been told that there is a story on the BBC that we won't peak for another week and a half.
They should have been reading my analysis 🙂

- when averaged over 3 days to try to eliminate noise (if you like, the second differential is lowering ), for the past 6 days. This is only a very slight reversal of that (12.6% to 12.9%). At this stage in their outbreak, Italy's curve was at a similar stage.

So, whilst we could see a higher figure than 938 deaths in one day, I'm going to stick my neck out here and say I don't think we'll see a much higher daily number, and this could indeed be the highest.

Today's #Covid19UK death figures may have caused some alarm. However, I don't think it is cause for dismay. As has been observed by a number of people (and denied by others - mostly not statisticians?), we have been following Italy's curve for some time.

Italy's highest daily deaths figure of 919 occurred on 27th March, i.e. 12 days ago. We're tracking at Italy+14. The data are noisy (they go up and down); we've had a constant decrease in the % increase of deaths per day, and this is crucial

Today's #Covid19Cymru figures are somewhat less encouraging than of late, with 33 new deaths in the past 24 hours, bringing the total to 245. This is above the projected 236. However, the projection is based on the average increase over the last three days, and the weekend figures were suspiciously low. So although it looks bad, it's not cause for alarm. At least, not yet.

Just a quick word on today's #Covid19UK figures. The total of 6159 (increase 786) was only slightly above the 6142 projected on the average of the past 3 days. However, viewed in the context of the considerable suspicion over yesterday's 441 deaths, this is actually not a bad result. It's probably a correction, after the weekend.
The rate of increase averaged over the past three days continues to fall: 27.6% 26.4% 22.4% 19.1% 14.3% 12.6%

Today's #Covid19Cymru death figures are just slightly less than the 3-day trend predicted (19 new, to 212 - predicted 214).
This is about what we would expect, as cases slow slightly following social distancing measures 2-3 weeks ago. It also seems to confirm my thoughts that yesterday's higher number was a correction after a weekend with two low figures.

Today's #Covid19UK figures:
Deaths today: 441
Total to date: 5373
For my discussion, see yesterday's comments, with increased emphasis.
However, I would note that today's figures are unexpectedly low, and I would not be surprised to see something of a correction tomorrow.

Today's #Covid19Cymru graph shows a turn upwards again - 27 new deaths today, for a total of 193. However, this is probably more of a correction after two lower-than-expected days than a concerning increase. If we were to average out the last three days, we would have 17.33 deaths per day, which is still a reduction over the previous few days.

At roughly 18 days (mean time infection-death) after people started taking measures (lockdown, social distancing, home working) the number will deviate from that curve. But we're only on our way to winning when the curve changes direction. In the UK figures, it's too early to suggest that's happening. 4/4

Today's total #Covid19UK deaths of 4932 is therefore significantly below what yesterday's projection had suggested. The projection assumes the average rate of increase over the past three days will continue, and in earlier stages this is a reasonable assumption. 3/4

We've now seen the %age increase in deaths falling for four consecutive days (31.5%, 24.2%, 23.4%, 19.6%, 14.4%), and so this is looking like a trend. It suggests that we really might be approaching the peak. However, it does not suggest necessarily that we have peaked in deaths per day just yet. 2/4

The #Covid19UK figures for deaths today are again encouraging. As yesterday, they show a reduction of the rate of increase in deaths (d²y/dx²), and indeed today they show a reduction in the actual number of deaths per day (dy/dx), at 619 (from 708 yesterday). The total deaths in the UK is now recorded as 4932. 1/4

CORRECTION: "yesterday's 13 was NOT a blip"

But we're only on our way to winning when the curve changes direction.This appears to be happening now.
3/2 (😀 )

Today's total #Covid19Cymru deaths - 166 - is therefore significantly below what the projection had suggested.
The projection assumes the average rate of increase over the past three days will continue, and in earlier stages this is a reasonable assumption. At roughly 18 days (mean time infection-death) after people started taking measures (lockdown, social distancing, home working) the number will deviate from that curve.

2/2

The #Covid19Cymru figures for deaths today are very encouraging indeed. They show only 12 deaths; the lowest since 31st March, and the lowest one I think we could describe as (even vaguely) reliable since 29th March.
This suggests that yesterday's 13 was a blip, and that we really might be approaching (or even past) the peak. However, 2 days is not sufficient for confidence. For that we will need another day or two of lower figures.
1/2

I know it's more difficult for my servants to go out and get me cat food these days. But that's ok, I'll bring in my own instead.
#Caturday