Failed COVID-19 Models Show We Need a More Targeted Approach: Rob Bluey on Newsmax

Rob Bluey joined Newsmax on June 16, 2020 to discuss how The Imperial College model fails the test of robustness, and it fails miserably. The model was not made publicly available, and after parts of it were, numerous problems were found including lack of reproductive and numerous errors. Sadly, its model was one of the inputs relied on as the basis for locking down two countries.

Heritage researchers used a publicly availability model to illustrate wide range of results that these models produce when users make reasonable changes to just some of the many assumptions that the models require.

For example, under reasonable ranges of the rates of communicability of COVID-19, estimates of mortality have varied since our April analysis, ranging from 44,000 dead to 1.05 million dead by August 1st. Estimates diverge considerably based on other assumptions as well.

While statistical models can be useful tools for tracking COVID-19, they are only as accurate as the input assumptions, which depend on continually changing data. Models used to make policy decisions should be made publicly available, have assumptions clearly stated, and have their robustness to changes to these assumptions tested.