Tuesday, June 9, 2020

The perils of following computer models intead of evidence based science...

Canada's response to the novel coronavirus, like much of the world, was based not so much on science as on computer generated model predictions.  Predictions of 200,000 people dying with no action taken, and even still 100K with mitigation efforts meant our health care system was forecast to be overrun with Covid-19 patients in either case.  

Our ICUs didn't have the capacity to deal with all the patients computer models said would be resulting.  Ontario premier Doug Ford  released more recent computer modelled predictions in early April.  

The model said Ontario would need an additional 900 beds for Covid patients, on top of the existing roughly 415 spaces that were available before the pandemic.  And this was called the "Best Case" scenario, if we were lucky.  In other words if a bunch of yahoos didn't protest at Queen's Park and thousands of people didn't gather in the province's parks ignoring social distancing directives, then just over 1,300 critical care spaces might get us through.


Well, the computer models were wrong.  But of course true students of science know that computer models aren't real science.  Since at least the 17th century true science has involved testing a hypotheis against imperical evidence with rigorous methodology in place.  Creating a computer program that generates results based on incomplete data and assumptions is no more scientific than making a weather forecast based on a trick knee.  

And because of faulty computer models Canada has wreaked untold misery on countless people, inclduing being directly responsible for lives lost.  We've killed people by cancelling life saving cardiac surgery and cancer screenings.  Studies show that unemployment increases the risk of suicide by 1.7%, and Canada has put ~4 million Canadians out of work.  You don't need a PHD in the social sciences to know that job loss can lead to depression, and then to substance abuse.  Closing schools means that countless kids have been forced to stay home with out of work alcoholic parents, with liquor stores deemed an essential service.  Alocholics Anonymous meetings are non essential, LCBO outlets are, great move.  

Government action and policy needs to be driven by evidence based science, not computer modelled simulations.  




4 comments:

The Disaffected Lib said...

Fair enough, Gordie. Show us your "evidence based science." If you've got it that should easily refute all the brainwashing you claim we're getting from the medical community. So, where is it, Gord? A few cherry picked data points, just like the evangos cherry pick the Old Testament to defend their derangement?

C'mon, Gord. I told you how you can easily monetize this shit if you can only prove it. You and your next four generations would be on Easy Street if there was anything to what you espouse.

Gordie Canuk said...

Sorry DisLib....the onus is on you to prove that the computer models that predicted 200,000 Canadians dying and Ontario's later modelling that showed we'd need "best case" 1,300 critical care beds.....you have to prove how those models were accurate. I can't prove a negative....you probably were with the crowd that wanted Saddam Hussein to prove he didn't have any WMD.

Anonymous said...

I recently found your blog and really enjoy your analyses. Thanks for your insightful commentary on the important topics of the day.

On another subject, I've noticed The Disaffected Lib is trolling your blog. You shouldn't have to put up with that abuse. He is not debating your ideas, but trying to put you down. My advice is to delete his comments or block him.

Gordie Canuk said...

I'm a classic liberal, which means I cherish free speech....so long as there's no profanity its all good. DL is more of a gnat...I don't even think he really knows what strawman arguments or ad hominem attacks even are...

How can one dispute that the models were wrong? You can't because its right out there for anyone with eyes to see. Ontario's model said this province's hospitals were going to be overrun with Covid-19 patients by mid April, and it didn't happen...hence the model was wrong obviously.

But again, models are not science....they're a guess based on estimations and assumptions. DL obviously likes computer models for some reason.