COVID-19 Pandemic Statistical Madness – Whose Numbers Are Correct?
The COVID-19 Pandemic statistical madness has reached epic proportions. I usually do not stray from Workers Comp articles. That is why you read this blog and the newsletter. However, the recent numbers flashed on screens across the world gave me pause.
I usually quote a large number of statistics such as X-Mods, LDF’s, and other organization’s studies. Those numbers, except for a few organizations, are usually static. They vary a small amount. Typically, the rating bureaus and the BLS provide great numbers.
The White House’s press briefings were based in solid numbers until the models were unveiled over the last 10 days. Models and predictive analytics go together. I am a fan of neither as the numbers can change quickly when covering any type of datasets.
One example of models that are supposedly magically correct is the hurricane models. The graphic below shows the variance in the predictions of hurricanes from 2012-2016. If the forecasters were accurate, then the line would be flat-lined at zero. This graphic is not the case.
The associated article to the above graph can be found here. One quote from the article that jumped off the screen was:
…..hurricane models are not good at predicting rapid intensification events such as Maria because so few of them occur.
You are asking, wait – your article title refers to the coronavirus, but now we are on hurricanes? Take a look at the same organization’s article on COVID-10 pandemic statistical madness. Check it out here, no really – take a look and my article will start to make sense.
Chart after chart in that article shows that the COVID-19 experts disagree significantly on the Coronavirus statistics. COVID-19 remains a rarity, so knowing the outcomes compares to forecasting a hurricane.
My suggestion is to keep an open mind when the barrage of numbers starts every morning on COVID-19. Check out the Google Doodle on any Google Search Page
for a great mini-guide on and avoid the COVID-10 Pandemic Statistical Madness.
©J&L Risk Management Inc Copyright Notice