Citat:
Ursprungligen postat av
Revej
Länk!
Tack.
We have three days of underreported statistics, Sat/Mon to chart.
The previous week Mon/Fri reported very equal numbers, whereas the middle days of the week reported significantly higher. While this is a sign of under-reporting, it will be included into our calculations.
Our dataset of deaths for the last week: 76, 77, 96, 106, 114
Our median is 96 and our mean is 93,4. Assuming our trend is constant and flat, that would give us 93,4 x 3 (sat/sun/mon) = 280,2 deaths to be reported over the weekend. An interval of 25% uncertainty sounds fitting, and leads to somewhere around 210 – 350 deaths.
Over this period a total of 49 deaths were reported. Adding that to today’s figures of 114 deaths gives us 163.
210 - 163 = 47 deaths uncounted for if deaths sudden decrease exponentially by 25%. I am not arguing for if these cases have died over the weekend, I haven’t checked the restructured statistics where they are matched to dates, but Tegnell said a “stable” rate of 60 a day.
At any rate, there are numbers that haven’t been processed over the weekend and that amounts to anywhere from 40—190 deaths not announced currently and a total of around 200/300 in the log over the weekend. Which days these will fall on and how the weekend will look after all backlogging is complete in a week and a half’s time, I don’t know. But I’ll estimate that Tegnell’s observation of 60 a day holds true and maybe +10 a day on top of that will be backlogged in, and we’ll arrive at around 210 deaths from Sat/Mon.