Covid & High School Football

I'm glad there has been a super spreading event.  I am interested to know how many of those schools have in-person teaching and the like, though.  Not that it should invalidate the success teams have had but comparing state-wide infection rates when most people remain virtual, as appears to be the case with Indiana (for example), and claiming success that you haven't had any major spreader events is a bit like comparing apples and oranges.  Also, a quick Google search does show that some HS teams did have breakouts and had to quarantine relatively recently, so it hasn't been completely smooth sailing.

That said, I want to make it clear that I commend HS kids for seemingly taking the necessary steps to keep themselves safe.  It would have been nice if us adults had taken their lead in the prior months.

I guess it's time to start the B1G season! Lets play football!

The lack of traffic in this thread is interesting.

So, basic premise seems to be, that if there is no growth in cases, there is no increase in community spread.   If there is no increase in community spread, high school football can’t be contributing to community spread.

This is a similar concept to one that many used to ask the question about whether protests lead to community spread.   I am one of the people who cited such analysis, so it would be pretty hypocritical of me to disagree with it here.

I do have two key questions?

1. Is it right to use State level data rather than more local level data?    Ideally we’d be looking at more local data (as was done for most of the protest analysis).   I think it unquestionably would, but understand why the author did not.   State level data can easily obscure more local increases, but would obviously take much more effort.

2. More importantly - did the author even do what he purported to do well.   Here I have very significant questions.   The author seems to have just looked at a bunch of bar charts, eyeballed them and said, I see know significant increases.   He did several things in his approach that I would question.    First, he used data from a much longer period than he was looking at.   In many of the states, they had reasonably dramatic humps earlier in the summer, which makes it harder to discern smaller increases by eye.   Second, he appears to have neglected the tool he easily had at his disposal that would have helped rectify this.   He could have clicked the little button at the bottom of the graph that would have provided a rolling average more clearly showing whether there was an increase or decrease.

I have no intention of going back and redoing his graphs, but using something like Covid exit strategy They actually provide 14 day rolling average increases.   Several states where the author suggests that his conclusions about increases may be wrong.    To me, the most obvious is Missouri.   Even just eyeballing the authors charts, it looks like there is an increase.   Looking at the 14 day rolling average in fact confirms an 18% increase.   Other states that the author looked at that have increases include Utah - 11%, Alaska - 19%, Nebraska - 13%, Ohio - 5%, Oklahoma 19% and Arkansas 29%.

While the author is wrong that there are no places that are playing football where at the state level, the metric of cases is increasing, there are certainly some where it is not.   To me this is why all these questions about how much should we open up are way more nuanced than play football or don’t play football.    We are in a significantly better place than we were In March with regards to being able to look at the question of community spread because we have significantly more testing available (and in a better place than we were in July, because that testing is becoming much more timely).   We should be using that data to be making decisions about where community spread is low enough to allow more activity and where it is high, considering limiting activity.

For instance: Arkansa with 243 new cases per million per day (one of the highest in the country), a 14 day increase of 29% in number of new cases and a positivity-rate of 11% snd increasing may not be the poster child for where HS football is safest.   New Mexico, with a relatively low new case count of 49 per million per day that has been decreasing 6% over the last two weeks and which has a low and stable positivity rate of 2.1% seems like a much better place to be allowing activities like HS football.

Driving to the games.