I certainly agree that a self-published paper lacks a signal in terms of quality in general that would otherwise be present after the peer review process. That said, the peer review process is not bullet-proof and many unpublished manuscripts are cited in published work based on the underlying strength and reputation of the authors. This reputation is established based on articles that have been published in peer reviewed journals. Dr. Wittkowski has dozens of publication in peer reviewed medical journals covering a pretty broad range of topics from infectious diseases and immunizations to emerging medicine in cancer and HIV treatments. He is a data scientist and a medical scholar. Calling him a "fake expert" because his prediction was low or because he has put up a strong fight when pressed to back off his theory doesn't impact his reputation that has been established over decades.

I recognize that many studies were released with confidence intervals and various assumptions that a data scientist can look into and interpret, but I thought the early projections were pretty irresponsible (failure to account for selection bias in testing and upward bias in projections by ignoring reliable data points plus failing to take obvious social response over time into account). I was very disappointed in the public health community that adopted a sort of gentleman's agreement to apply sloppy science to mislead the public based, I think, on the logic that it's okay to mislead if doing so saves lives through improved social response. I think that strategy will present a short-term gain and a long-term loss as better data will come out later and make the public health community look pretty silly.

Layered into this is the fact that most people don't have the expertise to do dig into a study and so they'll rely on the popular press to spoon feed that data. There was a lot of information being released without proper interpretations by people who should have known better. For example, an early NY Times article that was written by a Pulitzer Prize winning MD/PhD from Stanford released some very troubling numbers and left out sufficient data to interpret those predictions. Here's a link:



She cites an unreleased CDC report as her source, which combined with her credentials would appear credible, and her prediction on deaths ranged from 200K to 1.7M. The article would lead a regular reader to believe there would be over a million deaths directly attributable to this sickness. She also states that there will be 2M-20M hospitalizations and then tosses in the point that we have only 925K staffed hospital beds. An average reader of this article would get the sense that even the best case scenario will result in over a million people not being able to get the care they need with mass over-runs of hospitals and people being left to die from other ailments, but even if 2M were hospitalized for COVID-19, that wouldn't be at the exact same time and hospital stays vary based on severity and so her best-case scenario is misleading as presented. Everything about that article was meant to frighten people into submitting to the distancing measures. Most of the articles that were published at that time were very similar to this one and people who expressed skepticism were shouted down or met with some anecdote of a healthy person who died or a sob story by an ER doctor on the ground who "really knows" what's going on (as if such experiences tell us anything about whether 10K, 50K, 200K, or 1.7M will die).