In the long run, it's going to be excess deaths that do the best job of showing how many people died. If your deaths per year go 10, 11, 14, 27, 17 then you can have a pretty good guess that something happened in the fourth year and how many deaths that thing caused. That's happening already (I...