Covid worldometer Spike 6.10.21

[pinball1970]

TL;DR

A huge spike representing over 23,000 deaths in one day on October 6th.

I check this pretty regularly so I am sure this is a recent adjustment. China made a similar adjustment last year, around May adding 1200. The UK too some numbers off towards the end of last year (around 5,000?)
How does this work?
This spike seems very large.

https://www.worldometers.info/coronavirus/EDIT: It has disappeared!

It was definitely there, deaths 6.10.21 23,000 deaths.

@StatGuy2000 @atyy @Ygggdrasil may have seen it?

 

[jedishrfu]

Maybe someone double counted some deaths and then fixed it.

 

[phinds]

Often, reporting lags and then catches up all at once so the "spike" could be just that.

 

[pinball1970]

Ok, thanks for the input.

 

[jim mcnamara]

It usually relates to how excess deaths are determined and reported. Somewhat like an audit, for an analogy.
For example, the US state of New Mexico often reports a few Covid deaths as much as 30 days after the fact. The number gets lumped with the larger, recent group of deaths. So Worldometer shows a large deaths spike in the new New Mexico graph where several counties all reported death well after the fact - plus, reporting is limited by the fact that local government offices are closed on weekends and holidays.

Go here: https://www.worldometers.info/coronavirus/usa/new-mexico/
Go to the deaths graph, put your cursor on Nov 19, 2021.

Also note the very periodic nature of the data presentation in general. Hundreds of mini-spikes. To get an idea of what is closer to actual death rates, click the 7-day average box below the x-axis. Then move your cursor over the same spot on the graph - again.