[FRIAM] numbers

[uǝlƃ ↙↙↙]

And to make matters worse, we won't be able to trust the Census 2020 data *either* because they're cutting it off early:

https://www.nytimes.com/2020/10/13/us/supreme-court-census.html

All models are always wrong. Applying the population divisor, which is a wrong model, over the top of your wrong model (cases) produces an even more wrong, wrong model. But false models atop false models sure is USEFUL.

Which brings me to a post I thought about making last week: ad hominem *as* critical thinking. I may find the time to make that post if I'm lucky.


On 11/2/20 11:46 AM, Tom Johnson wrote:
> Nick:
> I would suggest that the 2010 population data is pretty thin for analysis.  Off the top of my head, our 87505 ZIP has about 25k people as of 2019.  ZIP 87507 has about 50k.  Hence the crying need for an appropriate denominator when looking at cases by ZIP.

-- 
↙↙↙ uǝlƃ