I would rather have questions that can’t be answered than answers that can’t be questioned” –– Richard Feynman, Nobel Physicist

I would like to introduce a new variety of post: “From the Mailbag.” A number of readers have posed questions and comments that may be of interest to all of us. This is the first in an occasional post presenting reader feedback.

We’ll limit this initial mailbag post to some observations and questions (edited for clarity) regarding Post #10 from a “self-described Skeptical Old Curmudgeon” [SOC] in Wisconsin.

SOC: Without disputing the data (or criticality), is it fair to compare raw numbers with Italy knowing that the US population is 5 times larger? Would the curve look different if that is factored in?

     >>>>> This is a very good point. You’re right — if we look at absolute numbers the comparison is not fair: i,t does not take into account the overall population. A population-proportional plot would offer the best method for comparing number of cases. In that case we would see that, relatively speaking, population-adjusted U.S. cases would indeed be five times lower than Italy’s. However, the curves were actually intended to highlight rate of growth. The U.S. curve is growing at a rate noticeably faster than Italy’s. This is what is concerning. What are the implications of such a rapid rise relative, say, to “final” peak? I don‘t think we have enough to conjecture an answer – yet.

SOC: Are we also factoring in age and smoking habits?

     >>>>> Demographic and behavioral factors like this have not been considered in any of our statistical analyses. Our focus right now is to provide a higher-level perspective on the situation. It is not clear that we have enough details on these factors yet to draw statistically significant hypotheses or conclusions. There have been some observations such as a higher infection rate among Chinese men than women that seems to correlate well with smoking habits (many more men smoke than women in China). But even in this case, is there only a fundamental gender factor at play? And we haven’t talked about age & underlying illness factors —  both of which seem to be closely associated with overall fatality rates. We have yet to delve into this kind of “secondary” correlation.

SOC: My casual observations of reported deaths in Wisconsin show advanced age and presence of COPD. All of the deaths in one county have been from a population which may not accurately represent total population. It has also been reported that social distancing has been largely ignored in certain areas. If we were to view this as “natural selection with collateral damage“ does it change the analysis?

     >>>>> One could say that what is happening worldwide is by definition natural selection — further confounded by environmental and cultural factors. For example, one-winged birds will quickly die out, but two-winged birds will also die out if their environment becomes arid or bereft of prey. One could argue that natural selection is at play in both cases. There is always a danger in drawing conclusions from limited data sets in which not all factors are considered. In your example one might wonder what other factors might be at play. Does the population and/or areas you refer to have dietary, environmental, economic, genetic, etc. factors that might affect conclusions that rely on these unique factors differing from the general population? Are there more smokers than in the general population? Is education level or access to media more (or less) limited? Lots of analysis remains to be done. I’m thinking that there is a great deal of fodder for a generation of Ph.D. dissertations in the aftermath of this “plague!”

SOC: What would have happened if the press had sensationalized the 19 million cases and 10,000 US deaths caused by influenzas back in January (from https://www.contagionlive.com/news/us-flu-cases-increased-by-4-million-over-the-last-week)? I am not a statistician. I hated the subject back in college. But I recall some of my six-sigma training (remember when that was all the rage?). We were always cautioned about collecting data to support a conclusion before sorting and weighing data to determine contributing factors.

      >>>>> You are right — the media is always after eyeballs. The January example you refer to would not rate as eyeball-catching news: it happens every year. Covid-19 IS news. Happens once in 100 years. What makes Covid different – what alarmed me when I first noticed it in February — was its similarity to the 1918 Spanish flu: specifically, a contagion rate (Ro) greater than 2, lack of vaccine, and — maybe what really caught my attention — the Chinese 1,000 bed hospital built in 10 days to handle the situation. The Chinese are not stupid. They were seeing something. Something the U.S. should have acted on back then (but that’s another story, beyond the scope of PanDispatch). Having studied the 1918 pandemic I began to think that this would be big. And real. My goal was simply to get the message out in a readable, reliable way. And you’re right about collecting data to support a conclusion. That’s precisely the reason for blind and double-blind testing with control groups and peer review. This alludes to current claims that there is a “game-changer” drug out there — that Tony Fauci continues to emphasize has not been scientifically proven to work (don’t go drinking any fish-tank cleaner!).