The Lifeboat News
[ Message Archive | The Lifeboat News ]

    Re: Some country coronavirus trend data Archived Message

    Posted by walter on May 3, 2020, 10:48 am, in reply to "Re: Some country coronavirus trend data "

    Hi Flash.

    Yes - as statisticians ought to be saying, you can claim anything with words

    The peak deaths cannot literally fall, obviously (though even the illogical connotation has a comforting subliminal ring to it), so that must mean the peak rate eg daily rate is past. It is true that the initial acceleration in daily deaths has ended but it hasn't gone into clear deceleration, it's flattened out. This should be a worry!

    Yeah the 6-day acceleration indicator (0.92 on 18 April) has not improved in the previous two weeks (0.95 yesterday). Its rough and maybe yesterday was a small blip but so far it has been around 0.94 within about 6%.

    The deaths coming in to the numbers have been subject to delays in reporting. I did see something where the actual reporting process was looked at, and it was estimated when the actual deaths occurred. I'm not sure if this was accurate or educated guesswork. Looking at it this way it was said that the daily rate of actual deaths (as opposed to the number of deaths reported daily, ie after whatever the delay is) was actually falling. I didn't check out the process to see if it was correct, I kind of doubted it.

    We're 5 weeks into the lockdown, which started when deaths were running at around 60 per day; if the present (reported) death rate continues for another five weeks the further death toll will be of the order of 200,000. Small comfort if that represents a slowdown! The length of time it stays high is also a big factor in the total.

    Good point you raised.

    Cheers (though no cheerfulness intended).

    Message Thread: