Doug64 wrote:@Pants-of-dog Yes, the models were clearly wrong, because the numbers of predicted deaths kept being revised downward.
I can think of more than one hypothesis for decreasing predicted numbers of deaths that do not involve “wrong” models, whatever you mean by “wrong”.
If you are arguing that he models were imperfect at the very beginning because we did not know much about the Trump virus at that point, you are correct. But that is just a way of saying that humans do not have good knowledge of things they are encountering for the first time.
The fact that model output changed as our understanding of the Trump virus improved means that the models improved over time, which in turns means our understanding of the Trump virus also improved.
And yes, it is exactly those incorrect predictions that make models useful to scientists, because they can demonstrate how the theories they are based on are flawed and need to be revised. So they make the revisions, run the models again, and see how the newest predictions compare to what actually happens, make more adjustments to the underlying theories and adjust the models accordingly, wash-and-repeat. Eventually, for modeling really complex systems, you can end up with a model that might be as accurate as, say, your average weather report.
So we agree that when a model is wrong, that is both commonplace and an essential part of the scientific method, and that does not actually discount the findings of scientists.
And what is interesting is that we might be able to reach herd immunity for the Wuhan CCP virus (since you seem to want to use the name to assign blame) more quickly than expected. Note that possibly, just possibly, New York City might be reaching effective herd immunity already.
Define herd immunity, please. I want to make sure we are discussing the same thing. @Hindsite does not use the term correctly, for example.