Norbert
TSLA will win
Why would you take it at face value? I don't. It does say what they think Rt is, until about a week ago, past that it wouldn't have much effect on current death numbers. We would expect that deaths go down, and maybe they will. Or maybe not, rt.live can't really tell us, even though we would expect it to.I just take it at face value - it's a formula based on the number of cases (specifically the change in cases over time). And they apply some correction now based on the testing levels. That's just going to produce a curve of Rt retrospectively. The site doesn't say anything about what Rt is now, and it is not a reflection of how many cases exist (in an absolute sense there is no such relationship), or how many deaths will occur.
I think it's a reasonable guide for what Rt has been, but I'm sure it's not 100% accurate. It's just a formula and if you feed it incorrect data the answer won't be right - and in all cases in all states with significant disease burden, they are guessing at the actual number of cases. So there is definitely going to be error, and test methodology changes over time will introduce additional errors.
If you want to feel good about distancing measures, it does that. But it doesn't show you the lack thereof. Too much green.Basically, just a website that allows you to see the impact of distancing measures and will tell us how things went, a couple weeks after we have taken actions. Fortunately states are looking at this data in closer to real time (but there is definitely always a delay)!
I haven't looked at the formulas, but who did? If the death numbers don't do what we would expect, then the graphs are not as meaningful as we want them to be. They don't say what we think they say.I can't prove or disprove this. The relationship between cases and mortality is complex. There are counting issues, treatment changes, dependence on exposed population, etc. I think it's hard to conclude that if the death numbers don't do exactly what you expect that there is some flaw in their formulas.