[blattimwind]

I don't think a sensible mortality rate can be computed at this point, because we have no idea, not even good estimates, how many people contracted the pathogen. 10 dead out of 1000? Or 10000? Or 20000? In Wuhan they apparently checked ~almost everyone for symptoms - temperature - but showing symptoms != infected. So far it seems that there is an exceptionally large proportion of infected people showing no or barely any symptoms.

[admax88q]

doesn't that mean that we can only compute an upper bound on the mortality rate?

We will never know for certain how many cases there are, but _of the cases we know about_ we can compute a reasonable upper bound on mortality. Were more likely not to learn of cases than were not fatal than we are of cases which are fatal as severe cases are more likely to seek treatment.

[marcosdumay]

With the measurement people are using, yes, you can only know an upper bound.

It's possible to discover an actual estimate, but it takes time and resources. Researchers may care about doing that after the crisis passes, but it's very unlikely anybody will care about this right now.

[admax88q]

Which is fine, people keep pointing to all these unknown as "we can't possibly compute the morality rate!" Which leaves the door open for assuming its way higher than estimated, when in reality current estimates are likely an upper bound.

[marcosdumay]

Although I think the official estimates are severely overestimating the lethality (by an order of magnitude), I also think there isn't enough information for establishing an upper bound yet.

Current estimates are bad.