Funny how the super-rational HN crowd throws science out the window as soon as it starts to actually matter. I guess there really are no atheists in foxholes, nor statisticians in a pandemic.
The authors of this study excluded one patient from the study because they died, and several other because they were admitted to ICU.
Assuming the dead patient did not get run over by a car, not considering death as relevant information for your study seems... bold.
They switched outcomes from what they pre-registered.
They did not even measure viral load, their primary outcome, in the control group. They looked at the patient and went like “she’s really sick, that’s a 23 at least”.
This, and the other problems mentioned, make this study suspect. And at the point where actual fraud is a possibility, the top line numbers stop being meaningful.
So in that sense your comment is like my Grandmother’s insistence on Pascal’s wager: “I know there’s no God, but paradise sound too good not to try”
The HN crowd is not super-rational. For the most part, they are smart people with a contrarian bent, which is very dangerous when combined with topics outside of the area of their expertise. We’ve got a lot of armchair epidemiologists and doctors in these threads.
The problem with Pascal's wager (finite loss/infinite win) is that we can't calculate the probability of the existence of a God. If it is infinitesimally small, then there is no paradox.
By the same logic, you must stop eating carrots if I tell you that doing so will be infinitely rewarded after you die.
The authors of this study excluded one patient from the study because they died, and several other because they were admitted to ICU.
Assuming the dead patient did not get run over by a car, not considering death as relevant information for your study seems... bold.
They switched outcomes from what they pre-registered.
They did not even measure viral load, their primary outcome, in the control group. They looked at the patient and went like “she’s really sick, that’s a 23 at least”.
This, and the other problems mentioned, make this study suspect. And at the point where actual fraud is a possibility, the top line numbers stop being meaningful.
So in that sense your comment is like my Grandmother’s insistence on Pascal’s wager: “I know there’s no God, but paradise sound too good not to try”