Quantitative thinking during a pandemic

Today is not right. The weather is fine. My family and friends are healthy and waiting to hear from me, ready for ordinary things like coffee and conversation. Normally, I’d be taking my grandkids to daycare and checking up on grocery and laundry lists. Then, a bit of reading and some writing. But, instead of my usual activity I sit alone in a world of similars. No driving grandkids to school, no personal contact, no “I’ll-pick-it-up-when-next-I’m-at-the-store.” How peculiar this sudden-forced life. We live at the mercy of the coronavirus, specifically the COVID-19 strain. Now, our life is dictated by the statistical models for infection and survival rates, as calculated by mathematicians working with arcane algorithms. Our thoughts are about “flattening the curve” and “opening up,” terminology we have learned to express the theory of having fewer people get sick at once so that needed medical facilities do not become overwhelmed. We accept the restrictions to our daily activities because we believe them to be in our best interests; and, of course, they are enforced to some degree. Then, too, we each know we are not alone. Virtually all countries in the world—140 or so sovereign regions—have experienced coronavirus.

Soon, the statistical modelers tell us, some businesses and public places will be open, following a phased plan that progresses stepwise toward normality. But each day the modelers posit new predictions. We react by revising our expectations. Millions of people have lost jobs and income, savings are wiped out, many businesses have no expectation of restarting much less thriving. The psychological stress has been devastating. Depression, torn relationships, drug use, and alcoholism are rising to unprecedented levels.  But, for other millions, families and partners have grown closer. Couples and individuals spend time with their children—time together that otherwise would not have happened.

Our circumstances are led by the statistical modelers with their continuous flow of information to medical experts who hope that politicians will follow their recommendations. So . . . where are we?  Safer, or have we just endured the most colossal mistake in the history of mankind?  All because of statistical predictions. The statistical models are recalculated to currently reveal a less severe, but not brightening, future. The infection-and-survival models follow the available medical data and standard biometric risk-analyses assumptions. Technically, the information is processed by various types of regressions, odds-ratios, and correlational relationships. One presumes these algorithms are accurate and reliable.

Interestingly, we do not imagine that we are victims at the mercy of these statisticians—in fact, quite the opposite.  We look to them for truth and guidance. We believe their calculations, trusting that they get the math right. Even when their projections do not materialize with exactness, we realize they are working with a paucity of data and that their assumptions are presumed from an unprecedented and unknowable circumstance. We trust science as the most accurate route to truth that we have. We live with a quantitative perspective.  Apart from ages past, we do not perceive ourselves as beings who are at the whim of fate or providence. For us today, quantification is our moment-by-moment reality. With this viewpoint, we live with an internal sense of odds and probability.

There is a truly remarkable story of how these maths (and specifically, probability theory) used by today’s modelers came about. It was not through the slow progression of technical advances across the centuries. Rather, the formulas these statistical models employ were developed within a short period of time, and not too long ago: about 130 years or so, beginning during the last (but most significant) years of the Enlightenment through the first third of the twentieth century, roughly the 1790s through the 1920s. It can be traced that the period’s history drove many of the technical developments.

Even more astounding is the fact that the measurement of uncertainty—statistical modeling—stem from the work of a very few men and women (fewer than about fifty principals), each of almost unimaginably high intellect and most working in about the same time period. Many of the major characters knew one another personally or at least were aware of the others’ corresponding work in prob- ability theory or statistics, and more broadly in mathematics. Yet, aside from knowing one another and their reputation in intellectual circles, most of these individuals worked in relative obscurity and were not popularly known. Even today, most of these individuals are still not well known, except to biographers and scholars. But, as we shall see, their influence on us today is nothing short of astounding. In large part, this is what makes the story of quantification so interesting.

Our quantitative mindset—how we view things today—is what brings us hope and despair. The outcome is simultaneously both.

Image by clay-banks-U0-r0JMypE0 via Unsplash.