They do not say death rate, but they do mention the case fatality rate, which is exactly what it sounds like, and as you have described. It’s the number of peo

They do not say death rate, but they do mention the case fatality rate, which is exactly what it sounds like, and as you have described. It’s the number of people who have died, of the known cases. Not everyone who is infected becomes a case, for one reason or another. Maybe their symptoms did not arouse enough concern, or maybe testing was refused to them since they didn’t fit an erroneously narrow list of testing requirements. Another thing to keep in mind is that people calculating the CFR from public data on a napkin typically don’t distinguish between survivors and those who just haven’t died yet. So the other side of the CFR (the people who haven’t died) tends to include both the known survivors and will-be-survivors, as well as the people who are in the ICU hanging on for dear life and could be dead tomorrow. For this reason, unless it explicitly states that it’s only counting survivors and deaths, the CFR is more reliable once the spread has began to decline, when there are fewer new cases each day than the day before.