The dramatic June resurgence of Covid-19 cases in my current home of Florida strikes me as film noir that has been playing out in four acts. You could write this story in a number of ways, but I see the plot line as a battle among the exciting(?) mathematical concepts of uncertainty, probability, and fecundity, with some ugly politics thrown in.
I am writing this while safely holed up at home on the Gulf Coast, but today it feels more like the classic Humphrey Bogart film Key Largo from 1948. A hurricane is blowing hard outside while these mobsters who do not have a clue as what they are doing are running around with guns drawn. And like that old movie, the hurricane is currently whipping the thugs badly, while Bogart, Bacall and my family hang on, hoping to survive to the end of the storm when the bad guys get carted away.
From the film Key Largo (1948), starring Humphrey Bogart and Lauren Bacall, with Donald Trump Edward G. Robinson making the untoward advance.
This Florida story, as I interpret it, tells us about one month of mathematical uncertainty (March), followed by a month of probability (April), when better medical knowledge begins to triumph over that uncertainty. During May, the voices of doubt, innumeracy and “social distancing fatigue” fight the scientists and doctors to a tie. But now in June, the levees have been breached, and nature’s mighty force of fecundity is on a strong run toward some unknown fate.
Florida confirmed Covid-19 case trend. Chart source: endcoronavirus.org
Fecundity and power laws
I wrote several posts about the emerging coronavirus during March from the perspective of the many exponential growth curves found in nature and economics, also called power laws. [1] A power law is one where a key factor increases at an exponential rate, in other words, X to the third power, fourth power, and so on.
The most familiar non-biology example of these curves is the time value of money. An initial sum invested in a company expanding at, say, 10% per year, grows not in a straight line, but rather in a curve that gets increasingly steeper over time. In this case, an initial investment of $1 million takes seven years to double the first time, however by 30 years out that investment will have grown to over $17 million. In just another 10 years, it balloons to over $45 million. A “payday” loan, where interest can have an effective compounded rate of over 200% per year, blows up in a curve like this in a matter of just months, kind of like this coronavirus.
The unimpeded coronavirus does follow that very same math in its replication in real life. Nature’s viruses, when they find a fertile home like the human respiratory system, or a cold, damp meatpacking plant, can replicate exponentially even faster than payday loan balances. They are fecund, a principle of life on Earth which means to be capable of producing numerous “descendants.” We are all here thanks to the fecundity of our parents and grandparents going back some four billion years on this planet.
Act I – March: Uncertainty
The virus causing this illness called Covid-19 is especially (ahem) virulent. In mid-March, we had only about 1000 cases of Covid-19 confirmed nationwide, but the count was growing exponentially at nearly 40% per day, with total confirmed cases each day being up to seven times higher than the week before it. We have now, at this writing, passed 2.5 million confirmed cases, and over 45,000 new cases per day, of Covid-19 in the U.S. And there are likely millions more unconfirmed cases out there among us.
During March, nobody knew how to stop the spread, but doctors and scientists were throwing a lot of mitigation techniques and medicines at the problem trying to stem the tide and save the sickest. While scientist and economists may toss around percentages trying to quantify the impacts of this disease and ways to mitigate it, those percentages in March represented more uncertainty than probability. These are two quite different concepts, although there is often a fuzzy line between them.
My best analogy contrasting uncertainty with probability is the current pundit odds-making on the upcoming November presidential election. The percentages expressed here are more of a quantitative statement of people’s confidence in their outcome prediction, rather than some probabilistic and testable outcome of, say, the chance of you winning the next blackjack hand at a casino. Elections are never exactly repeated under the same conditions, so they are not “testable.” What about the coronavirus and its mitigation options? Uncertainty or probability?
For the most part, the doctors and scientists have not just been throwing darts in the dark. Instead, the professionals who practice “science-based medicine” often use Bayesian prior probabilities to advocate the best candidates for mitigation strategies, drugs, and other treatment methods. A “Bayesian prior” is a known probability from documented experience with some related challenge, perhaps one caused by a different, earlier form of coronavirus like SARS. When the reality you face is uncertain, you have to pick your shots, and Bayesian decision-making is often the best method for weeding out the unlikely paths from the promising ones.
Not all doctors do it this way. You might instead stake your reputation on a treatment for an unrelated illness using poorly designed tests. Some advocates might even have a major financial stake in the outcome that could possibly make them rich. This appears to be how we went down a dead-end path with Hydroxychloroquine. The smart doctors called this stinker in April.
Politics rears its ugly head in times of uncertainty as well. In recent testimony, Dr. Anthony Fauci confirmed that political fears of mask shortages, rather than medical experience, drove the decision to downplay mask use during March, despite strong “Bayesian prior” evidence of their effectiveness. This political call has had horrendous downstream consequences, as the refusal to wear a mask has now become an indicator of political fealty to Donald Trump, contrary to growing probabilistic evidence of the effectiveness of even homemade cloth masks. Which brings us to…
Act II – April: Probability
Cases and deaths from Covid-19 rose relentlessly at an exponential rate through March, still running at about a 15% daily case growth at the end of the month, and over 20% daily increases in deaths. When exponential growth rates are this high, accurate forecasting becomes impossible because every day that the growth rate is not attenuated can throw off forecasts for 60 days out by 100% or more. This was the lesson that places like New Zealand learned while the United States did not. In this case dawdling and denial are literally fatal strategies.
A country or state needs to hit this virus early to minimize an important factor called R0, or “R-naught,” which is the probabilistic estimate of how many people the average infectious Covid-19 patient will infect in turn. You need to get this measure below 1.0 in order to “reverse the curve.” The U.S. still has not. Even low growth rates, where R0 is slightly over 1.0, are still exponential, just like a low-interest bank account will grow over a longer period of time.
By April, however, some of the March’s Bayesian guesses by scientists and doctors on the front lines began turning into probabilistic data, and the growth rates slowed. Daily new case averages even began to decline in Florida and some other states. But these are “messy probabilities,” with lots of uncertainty still in them. Not every mitigation technique nor every medicine, even the most promising ones, worked every time. Multiple studies currently put the effectiveness of cloth masks, for instance, in the fuzzy 30% to 50% range. The data on social distancing outside now indicates a higher probability of effectiveness than inside distancing, as another example.
An ongoing theme of this blog is that we live daily in a probabilistic “God throws dice” world. My analogy here is that we are trying to move the odds of dying from coronavirus down from a level similar to the chance of dying in a highway accident when driving drunk on a rainy night versus your risk of death while driving on a nice, clear day at safe speeds. The probabilistic differences are huge, by several “orders of magnitude,” or powers of 10, lower, with lots of insurance data to prove it. I’ll take the risk level of “safe driving” in living with the coronavirus in my world, but I do not want to be anywhere near the “drunk drivers.”
My favorite new analogy here is that of Swiss cheese. You can plug a hole in a boat with slices of Swiss cheese if you align the holes right. A combination of multiple 30%-effective mitigation techniques can easily bring your chances of contracting Covid-19 down to “safe driver” levels. It would be even safer if not for the many “new drunks” on the road.
One fan of online role-playing games expressed it this way: If you came across a “magic cape” lying on the ground during your “quest” that gave you a 30% level of immunity from your opponent’s weapons, you would surely pick up that garment and wear it. To toss it aside would be a very stupid thing to do. Well, a face mask in this environment is exactly that “magic cape.” Draw your own conclusion here.
Act III – May: Social distance fatigue and media doubts
By May in Florida, however, the political winds were changing, despite stubbornly-level case counts in many places. People were tired of the isolation of “shelter in place.” Businesses were suffering. The President openly refused to wear a mask because it messed up his makeup or for some other more disturbing reason. Certain broadcast networks began portraying mask use and social distancing as “socialist” intrusions into individual rights or “unmanly.”
Recognize that during this period where case counts were low in some communities, there were still “hot spot” infections in immigrant-dominant tight workplaces and eldercare facilities. My county’s nursing homes were among the hardest hit. People were still dying in these places at an alarming rate, and the virus was waiting for any opportunity to break out.
And that, perhaps, was a key part of the problem. The deaths have been demographically highest in elderly and minority demographics. During May I wrote of a counter-push to what I called “Sucks to be you ethics” increasingly being spoken out loud. “If this disease affects only minorities and old folks, then what do I care?” Add to this numerous YouTube videos of hostile confrontations between the masked and the unmasked. Certain businesses in Florida enforced social distancing rules while others openly flouted them.
Florida Governor Ron DeSantis was bound and determined to open his state and its famous beaches for unfettered business in May, even though most major population centers had failed to meet the “safe opening” guidelines from the Centers for Disease Control and the White House itself. The case numbers in Florida spent the month of May wobbling up and down. But people had forgotten that key basic principle…
Act IV – June: Fecundity returns
As you can tell from that graph at the beginning of this post, Covid-19 case counts have been skyrocketing again in Florida, beginning from about two weeks after widespread openings began (the probabilistic math of incubation time for this virus). As of this writing, confirmed cases are growing at an exponential rate of over 5% per day.
It is four times more dangerous for an at-risk person to be out on the Florida streets now than it was when the state was first locked down in April. Despite massively increased testing, the percentage of Covid-positive tests found has also been going up, a sure sign of faster “community spread” of the coronavirus. A similar surge is happening in over half of the states.
Some of us have been writing since March that this virus really does not give a rip about political intentions. It can, however, be mitigated by smart science, smart medicine, and some simple responses by citizens that can bring R-naught below 1.0, as has now happened in some states and many other countries.
At this point I have no idea what Act V will look like in Florida or the rest of the country. But I do know the power-law math of this virus pretty well. We all need to get our leaders and neighbors to understand this odd word called fecundity. This remains a virus hurricane that will whip your butt before it is through.
Here is the trailer for the film noir classic Key Largo from 1948:
Notes:
- For more on the math of fecundity, see my March post, “Kids, can you say Fecundity?”
Related posts:
- Kids, can you say Fecundity?
- Drunk driving in Coronavirus World
- The ethical theory of “Sucks to be you!”
- Bad habits and the Bayesian brain
For additional posts on probability, volition and ethics, follow the Dice icon back or forward where it appears.
