When it comes to the coronavirus, be leery of forecasts
TAKING A CLOSER LOOK
Gollmer
PREVIOUS IMAGE
NEXT IMAGE
As the COVID pandemic has taken hold, we have moved into a different way of living.
After initially bracing for a dramatic quick end, we have settled into a sustained long haul that has excited our fears and tested our patience. At least the data analysts are happy, for they have found a treasure trove of new data, and an opportunity to test their mettle at predicting the future with sparse fragments of data. Early predictions were naturally fraught with uncertainty, but that did not stop the pundits from catering to the extremes of public opinion.
Casting the pandemic as an overblown case of the flu and the presence of ever-changing predictions caused us to question the seriousness of the situation.
On the other hand, strict lock downs and the shuttering of “non-essential” businesses are unsustainable.
We now face the choice of putting our lives on hold indefinitely or transitioning to a new normal that balances the needs of public health, economic viability, and the education of our children.
This situation reminds me of the Weather Forecast Game from my days in graduate school. Over nine weeks, I forecasted daily low and high temperatures, wind, and precipitation for two locations within the United States. The following day, my forecast was compared to data released by the National Weather Service (NWS). The difference between my projections and the NWS data was added to my previous days’ score. The winner, as in golf, was the player with the lowest score by the end of the game.
There was one lesson that I learned: the precipitation forecast made the difference between winning and failing miserably.
Our forecasts could be any value between 0 and 100 percent, but the NWS data was either one or the other. If there was some doubt about the chance of rain, only the naïve forecasted one of the extremes. Some days they were lucky and guessed right, but in the end they seldom won the game.
Forecasts are all about weighing risks and benefits.
It takes an expert forecaster to make an informed judgment and then clearly communicate the possible outcomes to the public. Those outcomes are nuanced, reflecting the complexity of the situation. If it were as simple as rain or no rain, there would be no need of a forecaster and everybody would be right once in a while.
The same principle applies to the response to COVID-19, but with an additional layer of complexity. With the weather, there is no need to take public opinion into account. It will rain regardless of what we want. If people didn’t care about COVID-19, then forecasters could make reasonable assumptions about group behavior.
However, informing the public of those predictions, and the corresponding risks, changes those assumptions and, likewise, the prediction.
As a result, public opinion often shifts against the health experts and their warnings for social distancing.
The problem is as complex as predicting fluctuations in the stock market, but in this case, we’re talking about something more serious than financial loss: loss of life.
So here we are, five months later. More reliable pandemic data is available and analysts are providing sophisticated models that take health policies into account, as well as the economic impact of those policies.
It is up to our leaders to ensure clear communication of their informed judgment. It is up to us to listen and respond rationally, rather than blindly following those who want to polarize our opinions.
Let us follow informed science rather than sensationalistic pundits, forecasting either clear skies or a torrential downpour.
Steven Gollmer is a senior professor of physics at Cedarville University.
We now face the choice of putting our lives on hold indefinitely or transitioning to a new normal that balances the needs of public health, economic viability, and the education of our children.
After initially bracing for a dramatic quick end, we have settled into a sustained long haul that has excited our fears and tested our patience. At least the data analysts are happy, for they have found a treasure trove of new data, and an opportunity to test their mettle at predicting the future with sparse fragments of data. Early predictions were naturally fraught with uncertainty, but that did not stop the pundits from catering to the extremes of public opinion.
Casting the pandemic as an overblown case of the flu and the presence of ever-changing predictions caused us to question the seriousness of the situation.
On the other hand, strict lock downs and the shuttering of “non-essential” businesses are unsustainable.
We now face the choice of putting our lives on hold indefinitely or transitioning to a new normal that balances the needs of public health, economic viability, and the education of our children.
This situation reminds me of the Weather Forecast Game from my days in graduate school. Over nine weeks, I forecasted daily low and high temperatures, wind, and precipitation for two locations within the United States. The following day, my forecast was compared to data released by the National Weather Service (NWS). The difference between my projections and the NWS data was added to my previous days’ score. The winner, as in golf, was the player with the lowest score by the end of the game.
There was one lesson that I learned: the precipitation forecast made the difference between winning and failing miserably.
Our forecasts could be any value between 0 and 100 percent, but the NWS data was either one or the other. If there was some doubt about the chance of rain, only the naïve forecasted one of the extremes. Some days they were lucky and guessed right, but in the end they seldom won the game.
Forecasts are all about weighing risks and benefits.
It takes an expert forecaster to make an informed judgment and then clearly communicate the possible outcomes to the public. Those outcomes are nuanced, reflecting the complexity of the situation. If it were as simple as rain or no rain, there would be no need of a forecaster and everybody would be right once in a while.
The same principle applies to the response to COVID-19, but with an additional layer of complexity. With the weather, there is no need to take public opinion into account. It will rain regardless of what we want. If people didn’t care about COVID-19, then forecasters could make reasonable assumptions about group behavior.
However, informing the public of those predictions, and the corresponding risks, changes those assumptions and, likewise, the prediction.
As a result, public opinion often shifts against the health experts and their warnings for social distancing.
The problem is as complex as predicting fluctuations in the stock market, but in this case, we’re talking about something more serious than financial loss: loss of life.
So here we are, five months later. More reliable pandemic data is available and analysts are providing sophisticated models that take health policies into account, as well as the economic impact of those policies.
It is up to our leaders to ensure clear communication of their informed judgment. It is up to us to listen and respond rationally, rather than blindly following those who want to polarize our opinions.
Let us follow informed science rather than sensationalistic pundits, forecasting either clear skies or a torrential downpour.
Steven Gollmer is a senior professor of physics at Cedarville University.
We now face the choice of putting our lives on hold indefinitely or transitioning to a new normal that balances the needs of public health, economic viability, and the education of our children.