Thanks for chiming in. As I said in another thread I think at least this is the kind of framework for a discussion. Want to add more columns? Great. Want to double a number or halve a number, great.
I personally think there is zero downside to publishing the numeric targets. Of you get more data in to change them, just change them (it's at least better than now where they have dates that they just keep changing). I don't see how having a date that you keep changing is locking political leaders more than metrics that you change.
What was touched on in the reply, things like the amount and distribution of testing, the compliance and effectiveness of measures like wearing masks, and other information and treatments could change the criteria marked in green--which means they'd have to be revised publicly causing confusion. Kept internal they can revise it as things change and just announce Level changes publicly.
All of those numbers in green have been called out as likely incorrect at the moment. You can see positive tests jump as more are done in the daily numbers. In situations where people are dying they're often foregoing tests to save them for the living. At some point we'll hopefully have wider testing which will change the numbers and likely the criteria for level changes.
Thanks for replying! I agree I would like to see better numbers. However, I don't think that things like mask compliance would change the criteria marked in green, rather it would change the actual tracked metrics and require another lockdown.
While I am highly confident I didn't get everything right, I did account for the testing angle because I think that while tests are coming in on average (something like 7 days) very high positive, it is a very good sign that test quantity is wildly off. That is why I personally look at that metric. If 95% of tests are coming back negative (assuming good tests), that is a sign testing is going well. If 40% are positive, we aren't testing enough and should be conservative.
If our testing goes up which dramatically shoots up cases, even though its is a good thing I think its prudent to potentially escalate measures to be sure.
It should only be when everything is at least as good as a certain level should things decline. So for example if there are 1000 new cases in a week in california, with 2% growth rate, but 25% positive tests for some reason, then lock it down again (this is a very made up example).
I personally think there is zero downside to publishing the numeric targets. Of you get more data in to change them, just change them (it's at least better than now where they have dates that they just keep changing). I don't see how having a date that you keep changing is locking political leaders more than metrics that you change.