[quote=zk][quote=ucodegen]
313,000 tests completed, more than 41,000 tests positive.
That means current infection rate within population is about 13%. [/quote]

That would only be a reasonable conclusion if those tests were all given to random people. I didn’t see anything in the article to indicate that that was the case. (Correct me if I missed it.)
[/quote]
Correct, you didn’t miss it.. and that is also why I was saying that this may be an upper limit. It doesn’t handle the asymptomatic.

[quote=zk]
If tests were given to at least some of those people because those people had symptoms, then the 13% number would be high.

That said, if it’s not 13% now, it sounds to me like it could be before too long.[/quote]
I have a feeling that some are because the person actually had the flu(H1N1) fairly bad and wanted to check if it was actually COVID-19. I figure there are things that could make the number larger (new infections) and smaller (extrapolation of existing death rates ignoring the asymptomatic and mild cases that get better). I figure it is an ‘ok’ shot in the dark, certainly better than some of the panic mongering I have been seeing.

I looked at another piece of data (not completely) buried on the CDC site.. It is under the “COVID-19 cases by date of illness onset”. I think this is not ‘presumptive’ data – but proved? I extracted the table and did the sum – it definitely differs from other data values (I don’t like discrepancies)
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html
Note that the section “COVID-19: US at a Glance” – includes confirmed AND presumptive positive cases. I am uncomfortable with the term ‘presumptive’ when it comes to data.