re: Fauci predicts millions of coronavirus cases in US and 100,000-200,000 deaths

quote:

---

Originally top end was 141,995. Current top end is 123, 157. That isn't a factor of 3; It's not even a factor of 1.5.

---

And it looks like using the current reporting, we’ll probably push 120,000 deaths, and there is more upside potential than downside potential given the distribution.

quote:

---

Feel free to rummage through my post history but I can save you the time and assure that I never quoted the Imperial College model in support of anything.

---

And the Imperial model includes a much of different scenarios, from doing absolutely nothing to mitigate the spread to lockdowns.

At the time, the UK was going for “herd immunity,” which was very little mitigation (before trying to quarantine the at-risk) and from what we’ve found is that even extreme mitigation has made that difficult (nursing homes) so their strategy was effectively no mitigation in practice. That was what their really high figure related to, even though we weren’t going for that strategy here (already locking down), they provided it as well for the US.

That being said, the UK did go on lockdown and their median estimate of deaths under a lockdown was 24,000, they are already at nearly 19,000 and will probably end up well above 30,000. And due to their limiting testing, their excess mortality data indicates that there are probably more than twice the total deaths occurring, up to 3 times.

So if anything, the imperial model probably underestimated the impact, under all scenarios because the appeared to underestimated the reproduction rate.

quote:

---

Before you report with "muh population density," those numbers can be adjusted and have been adjusted by a mathematician to account for this. Wilfred Riley did statistical regression math and came to the conclusion that lockdowns had virtually no effect on how severe or not severe the outbreak has been.

---

As someone who feels that the path taken thus far is unreasonable, and decided to read that article out of curiosity: at first glance, that analysis is horse shite.

quote:

---

The ‘p-value’ for the variable representing strategy was 0.94 when it was regressed against the deaths metric, which means there is a 94 per cent chance that any relationship between the different measures and Covid-19 deaths was the result of pure random chance.

---

quote:

---

The only variable to be statistically significant in terms of cases and deaths was population (p=0.006 and 0.021 respectively).

---

See the problem yet? The fact that population was the only statistically meaningful variable in his regression should be a pretty big hint. The second hint should be that he found no statistical impact from median age or population density. Take two states for example:

Rhode Island
COVID-19 deaths: #26
Population: #44
Population density: #2

Texas
COVID-19 deaths: #16
Population: #2
Population density: #26

If you just plugged these two states into his regression, it would find the following:
- COVID-19 deaths increase with increasing population.
- COVID-19 deaths decrease with increasing population density.

Which is fine I guess, until you look at the deaths per million population:
- Rhode Island: 179 deaths / MM pop.
- Texas: 20 deaths / MM pop.

His entire analysis hinges on the idea that the virus impact are more severe in Texas (in this example), while ignoring the fact that the per capita death rate is 9x higher in Rhode Island.

Since the only thing he correlated against were total cases and total deaths, of course population was the only meaningful variable he found. The fact that total population correlates strongly to total cases and total deaths should be obvious to anybody with a pulse. Because that correlation is so strong, it makes all of the other variables look random / inconsequential.

Had he chosen to correlate against per capita cases / deaths we might actually be able to tell how population density, median age, and government actions affect the spread of the virus. But he didn’t, so all of his regression results are meaningless.

And again, I say this as someone that thinks we need to open shite back up.