You're right that Iowa's test sample is skewed. In general, data from all states in the US should contain a significant time period where tests could only be performed on people who had symptoms. In general, we haven't tested random sweeps of people, at least not in the US. Nearly all the testing data we have is based on people who go to get tested because they suspect they might have the virus. That is the nature of the testing data that we have. So if you say that 80% of people tested positive have no symptoms most people are going to assume that you are referring to the people who have actually been tested, not a hypothetical study group of people tested at random.
Iowa's data says at least 64% have symptoms. The CDC estimates that 65% will have symptoms. That is the data I know of, which strongly contradicts the 80% no symptoms number. If we have data that says today the positive cases are less likely to be symptomatic based on wider testing, then that's great. Or if there has been a study done that attempts to control for the bias in testing data we have. But, so far as I know we don't have that data at least publicly available.
And the point of this related to football - data shows that most people who tested positive do have symptoms. So, we can't just say it's fine for people to play even on a positive test because there's a low chance they'll have symptoms. If someone tests positive, they need to sit out until they're no longer contagious.
The 80% thing comes from people mis-interpreting some data of people coming off of a cruise ship where they tested everyone. It was spreading rapidly on the ship and they tested everyone. 80% of the people that tested positive hadn't shown any symptoms up to that point but they never went back to check and see what percentage of them ended up showing symptoms. But a lot of people have latched on to the 80% number as if it was a wide scientific study or something.
Last edited: