Kenneth Chang

You are making the same kind of assumption that antivaxxers make when it comes to autism: that apparent increases are associated with actual increases rather than increasing identification.

Look, you're smart. Construct a reasonable model that explains the exponential mathematical model that produces exponential growth in confirmed cases without an underlying exponential growth in the spreading of the virus through the population.

A constant infection level with an exponentially increasing level of testing would yield those results.

I believe a linearly increasing level of infection with a linearly increasing level of testing would also be exponential yield similar results, though it might be parabolic rather than exponential.

Then again, it's hard for the human eye to differentiate between parabolic and exponential growth, and on the log scale, it appears to be sublinear, which would suggest more parabolic growth in actual figures.

A linear infection rate doesn't make biological sense. Testing hasn't been increasing linearly either. And I don't see anything suggesting sublinear in the US data (on the log graf). I've been waiting anxiously for that. You're good w/ numbers. Show me.

The log graph is clearly sublinear, yes? That means the original graph is sub-exponential.

Though again, the human eye isn't great at analyzing graphs objectively. And you're right. An infection doesn't grow linearly. It grows logistically.