[JustAnotherFan]

The confirmed case total for the US was 113,000 this morning, which is ~0.03% of the population.  This thread makes me thing the actual is A LOT more.

Anyone know people at a polling firm?  I think that with a little statistical knowledge*, you could poll a bunch of people, ask how many they know confirmed and suspected, and get a better estimate of the total than anyone's given us so far.

* you'd have to figure out how many people each person knows on average to get a denominator, but I think if you broke it down into friends, family work acquaintances, etc. there must be data out there on how big the average person's social network is.  You could also use the confirmed numbers as a correction factor, since they have to add up to ~113,000.  The only real guesswork would be what % of the suspected cases are covid-19 vs flu, etc.  My guess is that covid is making up a larger proportion of the flu-like illnesses every day.  For me at least, this would be super interesting (my day job involves analyzing similar messy datasets).

Considering this is the top news item most days and people are scared I would guess your chances of finding this out versus finding out someone has cancer dramatically increase. You're also more likely to register it. So that's one. Two, Stanford grads are disproportionately located in hot spots as are their networks of friends (SF Bay Area, NYC, Seattle, LA) so that would greatly increase our chances of knowing someone over the average American (as if there is such a thing). But I agree there are a lot of cases that are not confirmed and I cannot fathom the infection rate being as low as they claim it might be in my town.

[JustAnotherFan]

The confirmed case total for the US was 113,000 this morning, which is ~0.03% of the population.  This thread makes me thing the actual is A LOT more.

Anyone know people at a polling firm?  I think that with a little statistical knowledge*, you could poll a bunch of people, ask how many they know confirmed and suspected, and get a better estimate of the total than anyone's given us so far.

* you'd have to figure out how many people each person knows on average to get a denominator, but I think if you broke it down into friends, family work acquaintances, etc. there must be data out there on how big the average person's social network is.  You could also use the confirmed numbers as a correction factor, since they have to add up to ~113,000.  The only real guesswork would be what % of the suspected cases are covid-19 vs flu, etc.  My guess is that covid is making up a larger proportion of the flu-like illnesses every day.  For me at least, this would be super interesting (my day job involves analyzing similar messy datasets).

Considering this is the top news item most days and people are scared I would guess your chances of finding this out versus finding out someone has cancer dramatically increase. You're also more likely to register it. So that's one. Two, Stanford grads are disproportionately located in hot spots as are their networks of friends (SF Bay Area, NYC, Seattle, LA) so that would greatly increase our chances of knowing someone over the average American (as if there is such a thing). But I agree there are a lot of cases that are not confirmed and I cannot fathom the infection rate being as low as they claim it might be in my town.