I’m trying to understand what, if anything, SDI is useful for. It’s quite obvious that you believe it’s useful. I’m just not sure why. The goal of my post was to better understand if there’s anything I’m missing about SDI’s potential usefulness. Nothing you said in this giant post provides any evidence that it’s useful.
Uh, yes. My concern is that you decided that SDI must be useful, and you went looking for anything that would confirm this belief.
I know you’ve made this assertion a lot, but I haven’t seen any evidence that SDI is predictive of an entrance to a spike or an exit from a spike. Do you have any evidence of that?
The second piece of this quote is the one that leads me to often dismiss your arguments - you have consistently asserted that testing data is completely unreliable, but you haven’t offered any real evidence for that assertion, other than the fact that the testing data doesn’t comport with your own beliefs. That’s not a scientific mindset, that’s a mindset of someone who is only willing to entertain data as long as it matches their priors.
The search function sucks here, but when I look back to your big post containing your original report, I see a bunch of graphs that provide absolutely no indication of a consistent pattern between SDI and new cases. As far as I can tell, you eyeballed these graphs to come up with some “target” SDI and then claimed that breaching the target SDI had very obvious implications. But those graphs look crazily inconsistent and it’s hard to believe the relation you’re seeing is anything but noise.
This is the post I’m referring to:
It contained graphs like this:
Looking through those graphs, I don’t see any particular link between measured SDI and new cases. Is there another post that you’ve made that shows a link?
Again my question is this: What do you believe the SDI data can be used for, and what evidence is there that it can be used that way?