NH? According to this chart, NH seems to have a rather middle-of-the-road amount of testing per million inhabitants, compared to the other states.

I have never taken worldometers as being accurate, but maybe reasonable? Also, it’s not good to change data sources once started. I also know that in NH there is very limited testing. Even so, the active cases continues to climb abut 30,000 per day and serious cases (about 3% of the active cases) also climbing. First indication I would think is serious cases going down. Active cases could continue to climb just due to testing. I was hopeful earlier this week when serious cases starting leveling for 3 days, but then surged again.

Only a sample size of 1, so take it for what it’s worth, but my 82 year old mother living in the Chicago area has been ill for over 3 weeks; a few COVID-esque symptoms. We finally convinced her to go to an urgent care (she is a very tough person who always plays down her own aches and pains), where they tested her for Influenza A and sent her home. They called later and told her the test was negative and told her what to watch for, COVID-wise. So far, so good. My dad is keeping a close eye on her.

The point being, even an 82 year old woman who has some of the symptoms and is not really getting better on her own (she’s also not getting worse) was not tested in Chicago. If my mother’s experience is common in Chicago, then people are mostly only showing as positive cases when they are in serious condition because they are only admitting folks who present with serious conditions, and that’s when they are tested. Different states, even different areas within states appear to have more tests, so testing varies.

That would be incredible, if it turns out to be true. There is so much we do not yet understand.

Constant rate of change is a linear function

I didn’t say it would continue forever, I said it marked the peak. Sigmoid curves are approximately linear in the middle, which means constant rate of change. The question is how long that linear portion continues. The models are pretty simple, https://ihmeuw-msca.github.io/CurveFit/methods/.

My interpretation: We don’t have a national trend, we have localized trends. (Husband: “It’s a non-uniform distribution.”) The next few days will show whether these localized trends continue to improve.

I have to disagree. Constant rate of change is a linear function and will keep growing; it’s definitely not a peak. A linear function will reach infinity, just not as fast as a power or exponential function. We need to see a decrease in new active cases each day to declare getting close to peak, and then finally, zero new cases at the peak.