Forgive the click bait title.
FiveThirtyEight recently published an excellent article: Why It’s So Freaking Hard To Make A Good COVID-19 Model
If you're frustrated by conflicting media reports about infection rates or fatality projections, I'd encourage you to give this a read. There are also a ton of interesting links embedded in the article, including a story of the one woman who single-handedly caused an outbreak in Korea. You can go down quite the rabbit hole here, if you're so inclined.
My main takeaway: treat any reporting of coronavirus data and forecasts with extreme skepticism. Anyone who makes blatant statements like "the fatality rate is x%" or "the risk is y% higher for those over 60" doesn't really know as much as they think.
This could be interesting and useful information.
I predict that this thread will quickly devolve, but good luck here to all.
I've noticed that there are three tiers of takes when it comes to COVID-19.
The least sophisticated is the Average Joe take. These people simply look at the current death or infection numbers to judge the seriousness of the problem, and can't think beyond that. Typical quote: "LOL, there's only X cases in the whole country....what a nothingburger!"
The next is the Midwit take. These people are a little smarter than the Average Joes, and understand concepts like compound interest. They rely on simple projection models with fixed values for R0, speed of transmission, mortality, etc. that project high death tolls. They talk excitedly about "flattening the curve", but don't understand https://medium.com/@joschabach/flattening-the-curve-is-a-deadly-delusio…</a">the math involved. Typical quote: "LOL, average people are too dumb to understand exponential growth."
The smartest people are thinking along the lines of the linked FiveThirtyEight piece. They know that this virus may well be extremely bad and cause millions of deaths in the US alone, and that it's worth taking extreme precautions for that reason. But they also know that the model inputs are generated by very inadequate data, and that the range of potential outcomes is huge. Most posters here know that the output of a fancy, complex model can be highly sensitive to one little variable. And if you don't have a good handle on that variable, your model is basically fancy garbage.
Governments around the world are making huge decisions that are tanking their economies. They may be overreacting or they may not be. We need better data to find out.
Repudiandae voluptatibus consectetur sit veniam ut qui. Corporis at omnis nulla ea dolor. Quia architecto neque molestiae. Alias placeat officia quia ea possimus culpa. Dolor tempora repellendus pariatur dolorem omnis. Non et qui at totam.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...