> Putting it in concrete terms, if you have two versions of the virus, X which for each infected person tends to infect 1.8 others, and Y which tends to infect

> Putting it in concrete terms, if you have two versions of the virus, X which for each infected person tends to infect 1.8 others, and Y which tends to infect 2.8 others, then after ten days there might be more people infected with Y than X, but X would still be around. This is actually an observed effect in many transmittable illnesses; over time, they become more transmittable, and less deadly. Extremely deadly forms kill their hosts, ending the ability to reproduce via transmission, and the more transmittable strains reproduce effectively. This type of optimization leads to things like the common cold and flu. The scary thing about COVID19 is that it has a very long lead time between infection and symptoms, if they develop at all, meaning it can circumvent some of those "normal" optimization peaks; it can be both very deadly, and very transmittable, because of the latency in the signal.