|
> that's on the order of a .001% chance of dying. Not exactly, there's factors pushing that way down, and some up. We shouldn't look at [mortality /infected numbers], but rather at [mortality / resolved cases]. Given the exponential growth, most infected occurred very recently, and their death may still be pending, bound to occur in the coming weeks. For example, suppose an infection takes 1 month to resolve. If in January 100 get infected, in February 1000, and in February all the 100 infected from January have died, the mortality rate isn't 100/1100, it's 100/100. Instead, looking at just the resolved numbers: either death, or recovery, are more accurate. This is still not fully reliable given the cases resolving in death or recovery may have different timing, but it's better than just looking at gross infected rates. This pushes up the 0.001% figure. More importantly though, our infected numbers are way, way off. Probably 10 times higher. Most countries only test people who are hospitalised. The rest must stay home. There's no country that I know of that is going to people's homes to test people at scale. And there's only a few countries with a drive-through testing kind of model. Given there is evidence that substantial percentages (perhaps 50%) are asymptomatic, and that weak and mild cases aren't being tested either, the true infected numbers are way worse. (which pushes down your 0.001% figure). |