WASHINGTON — As much of the world sits at home social distancing and battling the coronavirus threat, it’s important to remember we do so based on the projections of computer models. Britain’s Imperial College expected US deaths from the Chinese flu could range as high as 2.2 million. But what are these projections, based on often imperfect computer models, really worth?
A glance back in history gives us a small but telling clue.
Was Albert Einstein math-challenged?
It’s one of the great ironies of science that the brilliant physicist Albert Einstein had trouble with mathematics. Lucky for him, his good friend Marcel Grossmann helped Einstein synthesize his greatest accomplishment.
“Since the mathematicians have invaded the theory of relativity,” said Einstein, “I don’t understand it myself anymore.”
The same may hold true of at least some of today’s computer-modeling-inspired prophecies of doom as offered by currently projected coronavirus casualty lists.
Battling the coronavirus with computer models
White House coronavirus taskforce member Dr. Deborah Birx recently threw some shade in the direction of her team’s own models. She warned a media prone to sensationalism not to put too much credence in the dire computer estimates.
“Models are models,” said Dr. Birx, “When people start talking about 20% of a population getting infected, it’s very scary, but we don’t have data that matches that based on our experience.”
Her colleague and fellow task force member Dr. Anthony Fauci, on the other hand, is more prone to believe the grim prognostications.
While appearing on CNN, the good doctor said the following.
“Looking at what we’re seeing now, I would say between 100,000 and 200,000 [American] deaths. We’re going to have millions of cases.”
At a recent White House briefing, President Donald Trump attempted to put some daylight between himself and his White House team’s forecasts based primarily on computer models.
“You know what I want to do? I want to come way under the models. The professionals did the models. I was never involved in a model… at least this kind of a model.”
The quip no doubt brought a smile to First Lady Melania Trump, a former fashion model.
Death by numbers
Why this disparity in opinion? Where computer models are concerned, an old adage comes to mind.
“Garbage in, garbage out.”
Wielding a hockey stick
Back in 1999, Dr. Michael Mann, current director of the Earth System Science Center at Pennsylvania State University, a climate change champion, published computer modeling results representing 1,000 years of Earth’s climate history. These results showed a sudden spike in temperatures beginning in the late 20th century.
Presented in the form of a chart known for the steep, upward rise of its fever-line, this graphic image became known simply as the “hockey stick” chart.
But Canadian scientists Stephen McIntyre and Ross McKitrick found a flaw in Mann’s mathematical modeling.
Richard Muller of the MIT Technology Review reports the Canadian team says Mann’s standard modeling method, the principal component analysis (or PCA), was not employed.
“Not only does… [Mann’s] program not do conventional PCA, but it handles data normalization in a way that can only be described as mistaken.”
“Now comes the real shocker. This improper normalization procedure tends to emphasize any data that do have the hockey stick shape, and to suppress all data that do not.”
In other words, Mann’s computer model showing a dramatic, man-made spike in atmospheric temperature. The hockey stick was mathematically baked into the cake.
Not enough information
Coronavirus computer models have a significant flaw. They use data from China. And the communist giant has been anything but frank when it comes to providing accurate information on the effects the virus has had on her population.
The Imperial College projection of 2.2 million US deaths is significantly higher than the recent White House estimate of 100,000 to 200,000.
The wild difference in numbers shows that no one really knows what’s going on. There isn’t enough verifiable data. And the least reliable measure on which to gage government policy consists of flawed computer models.
Battling the coronavirus with the numbers game
Scholars at Heriot-Watt University of Edinburgh, Scotland noticed problems with computer projections involving fire safety design. In “Problems with Computer Models,” the study discovered:
“Different people or groups, using the same model and applying it to the same case, may well produce different results. Indeed, this seems to be the norm. The employment of computer-based models in decision-making may well lead to totally unacceptable, or even dangerous, options being accepted for design. It is not impossible to use models in a way which may be valuable rather than misleading; however, it is not an easy matter to do so.”
This may remind us of the glowing and menacing red eye of the HAL 9000. This menacing symbol of modern technology had a central role in Stanley Kubrick’s classic film “2001: A Space Odyssey.” In the film’s primary plot thread, this digital villain attempts to murder each crew member aboard Discovery One least anyone report his logical error. This fatal AI error was due to a programming glitch.
As we continue battling the coronavirus, we may find our government pushing the United States into what looks to be a second Great Depression. So it’s worth remembering that the trigger for current actions may well stem from a simple computer error. Like forgetting to carry the two. Something to think about.
Top Image: Computer via pixanio,