HN user
yes, in any system where you are sampling from a population as you increase the bias the variance decreases.
Increase competition, i.e. apply bias to the filter that selected samples(people) for continuation in an academic or professional environment, and the variance of the people who make it past that filter will go down.
There is no way to avoid this.
Looking at ONS and NHS statistics for deaths and using the same criteria to look at deaths within X days of a vaccine, and adjusting the excess deaths in each as compared to the baseline rate of deaths in previous years, adjusted by age group, has been considered anti vaccine propaganda for a lot of this covid debacle.
UNfortunately what your family member is doing is actually the mirror opposite of what the official reports and mainline consensus has been, that all deaths in X days of a positive covid test are caused by covid but that no deaths after vaccination are are caused by vaccines. Also consider that harms from covid have been reported averaged amongst all age groups to give and inflated risk for the young, while risks from vaccination have been averaged amongst all age groups to lower the stated risk to the young. I state this last point just to illustrate how deliberate misinformation has been government policy with regards to covid stats, so the same technique used by anti vaxxers is of no surprise whatsoever.
When the mainstream consensus uses precisely the behaviour nudging abuse of data as a conspiracy theorist, do not be supposed when some people are unable to see what the problem is.
This is horrifying.
Literally looks like a step towards algorithmic driven totalitarian hell.
Imagine....
"worker drone 2349243A, your mask slipped down your nose for 0.34s before you corrected it. This is in violation of your employment contract, and therefore your employment is now terminated.
Email auto generated, do not reply "