The Covid-19 Info thread (keep politics out please)

[hacker]

Why do you think the negatives are reported later than the positives? Are you seeing that somewhere or are you just assuming they are eventually reported?

Because looking at historical data, usually the day after no tests are reported, twice as many tests are reported. Then the following days it regresses to the mean.

[Johnson85]

Because looking at historical data, usually the day after no tests are reported, twice as many tests are reported. Then the following days it regresses to the mean.

I get that there are days that have dumps of prior tests. But why do you think they include just the negatives from days where positives were already reported? It doesn't seem far fetched to me that that would happen. I could see labs reporting positives as they come up, and then doing a weekly dump of the rest. But I wouldn't just assume that happens. Has MDS said that's what happening? Or do you see something in the data to suggest it?

[hacker]

I get that there are days that have dumps of prior tests. But why do you think they include just the negatives from days where positives were already reported? It doesn't seem far fetched to me that that would happen. I could see labs reporting positives as they come up, and then doing a weekly dump of the rest. But I wouldn't just assume that happens. Has MDS said that's what happening? Or do you see something in the data to suggest it?

It just seems the most plausible explanation given that most weeks look something like this:

Monday: 0 tests reported
Tuesday: 10000 tests reported
Wednesday-Sunday: 5000 tests reported

[Johnson85]

It just seems the most plausible explanation given that most weeks look something like this:

Monday: 0 tests reported
Tuesday: 10000 tests reported
Wednesday-Sunday: 5000 tests reported

But why do you think they are just reporting the negatives? It seems at least as plausible that some clinics report positives but don't report negatives. Some clinics either don't report everyday, or alternatively the MSDH only incporates those labs tests into the data certain days. Unless the data also shows the positive test rate going from 100%, to 5% to 15% or some pattern like that, it seems just as plausible that the negative tests from some clinics just aren't being reported. Not that they are being reported on different days from the positive tests.

[hacker]

But why do you think they are just reporting the negatives? It seems at least as plausible that some clinics report positives but don't report negatives. Some clinics either don't report everyday, or alternatively the MSDH only incporates those labs tests into the data certain days. Unless the data also shows the positive test rate going from 100%, to 5% to 15% or some pattern like that, it seems just as plausible that the negative tests from some clinics just aren't being reported. Not that they are being reported on different days from the positive tests.

MSDH just doesn't update its test numbers some days, usually on weekends.

I don't think it's just as plausible that 100% of our cases reported the past two days came only from clinics that don't report negatives and also today we suddenly have 3x as many tests reported for no reason.

[Dawgology]

MSDH just doesn't update its test numbers some days, usually on weekends.

I don't think it's just as plausible that 100% of our cases reported the past two days came only from clinics that don't report negatives and also today we suddenly have 3x as many tests reported for no reason.

No doubt that shouldn't be happening anymore. Test number lag. Test results lag. Mortality rate lags. We are constantly looking at the picture as it was 3 days - 2 weeks ago.

[Johnson85]

MSDH just doesn't update its test numbers some days, usually on weekends.

I don't think it's just as plausible that 100% of our cases reported the past two days came only from clinics that don't report negatives and also today we suddenly have 3x as many tests reported for no reason.

Where are you looking to see the cases? Have you just kept up the last few days? Or is there a graph or chart for tests per day that I am missing?

[hacker]

Where are you looking to see the cases? Have you just kept up the last few days? Or is there a graph or chart for tests per day that I am missing?

Actually, I'm wrong. They didn't update tests again today. I got confused about Total (PCR + Antibody) and PCR tests (there's about a 16k difference).

Today: https://covidtracking.com/screenshot...804-182114.png
July 31: https://covidtracking.com/screenshot...801-181637.png

To answer your question, you can see historical data and a chart of new tests here: https://covidtracking.com/data/state/mississippi

I agree that 7 day average positivity % will be out of whack until they dump tests, but I expect it to average out by the end of the week.

[hacker]

It just seems the most plausible explanation given that most weeks look something like this:

Monday: 0 tests reported
Tuesday: 10000 tests reported
Wednesday-Sunday: 5000 tests reported

Today's numbers have been reported.

This week:

Sunday: 0 tests reported
Monday: 0 tests reported
Today: 15000 tests reported

We average around 5000 tests a day. 1000 of those positive. 20% positivity rate. Everything checks out.

(liberal use of rounding in calculations to help make my point)