the initial modelling had a death rate around 1% of those infected. we know now that that's about ten to twenty times too high.

it also had r-not values around 1 or 2. we know now that that's 2-3 times too low.

those are the sources of error in the models, and we can fix that to come up with better projections.

but, the focus of these policies was never to burn the virus out, it was only to slow it down.

we were never going to save anybody, we were only going to prolong their death by weeks or months.