Thursday, April 16, 2020

so, the models in new york drastically overestimated the problem. the politicians are claiming they overestimated the transmission rate, and social distancing slowed the spread; i am skeptical of this claim, and would suggest instead that they overestimated the mortality rate.

in the end, the death toll is going to be equal to the infection rate multiplied by the mortality rate. so, if your death toll is lower than expected, you have three options to correct for:

1) you can hold your mortality rate steady and decrease your infection rate, which is what the government is pushing.
2) you can hold your infection rate steady and decrease your mortality rate, which i suggest is closer to the truth.
3) you can decrease both factors by differing amounts, which is going to inevitably be reality, but isn't an idea that is being advocated by anybody right now. so, it could be that social distancing is working a little bit and the virus is a little bit less deadly than expected. some separation will eventually be required; right now, the debate is which factor is dominant.

we've seen the opposite in canada, but i should caution that the data is national while the reality is local. why then did the models in canada underestimate the number of deaths?

we could just adjust our reasoning, and this would create some problems for people pushing this widely embraced and rapidly developing pseudo-science of social distancing, to the extent that it doesn't turn into a process of collectivist masochism (if that isn't a redundancy).

then, it would follow that one way to explain why canada underestimated the death toll is because it underestimated the transmission rate, and social distancing did not reduce the spread - or wasn't followed. you'd have to conclude, then, that new york actually managed to do more social distancing, and more effectively than canada....despite the fact that it's actually the baseline used to calculate the worst case scenario. that's not quite a contradiction, but it's certainly inconsistent, and the results are definitely incoherent. if you base your models on the assumption that new york is what happens when no social distancing happens at all, and then new york beats your prediction due to overperforming you on metrics of social distancing, all you've managed to produce is nonsense. which is what i said from the start. so, if we're going to conclude that the numbers are wrong because we underestimated the transmission rate, we should really be deducing that social distancing was less effective in reducing the spread than the modelers initially guessed it might be - not that we're not doing it right. and, this is consistent with my analysis of new york.

the second option would be to argue that they underestimated the mortality rate, which would not be consistent with data we have elsewhere in the world. this is so unlikely that i'm going to discard it offhand.

so, then we're left to conclude that the case study comparison of state data in new york (where they were late to act, and overestimated the death count) and federal data in canada (where they reacted very early and then underestimated the death count) actually produces a consistency - in both cases, the modeling seems to have overestimated the effects (or lack thereof) of social distancing. it's actually pretty much perfectly consistent.

oddly, the provincial governments also seem to have largely overestimated rather than underestimated the toll, so it's starting to look like my analysis of ontario is going to be very similar to my analysis of new york. and, this difference is reflected in the public policy, as well.

trudeau has a knack for knowing the right things to say to persuade secularists to vote for him, but he also has a long history of falling head over heels into complete pseudo-science. this was a big issue under stephen harper, who muzzled scientists and tried to reorganize (perhaps succeeded in reorganizing) the research potential of the state as a propaganda tool of the oil industry. they ran heavily on this point; they were supposed to bring back science to the canadian government, to reintroduce the centrality of evidence-based decision making.

this is what we actually got, instead:
https://www.theglobeandmail.com/opinion/trudeaus-deliverology-on-the-verge-of-becoming-a-punchline/article37023338/

...as it became increasingly clear that this guy is his mother's child, and he is more into yoga and vague concepts of "spirituality" than he is into science, reason and math.

so, it shouldn't surprise anybody that this is the government that bit the hardest, or that there might be people inside of this government that are legitimately shattered by the failure of the policy.

they put everything on this - the prime minister, the government, the party and maybe even the entire society. in the short run, they're more likely to blame it on us for being stupid than they are to admit they were wrong, despite that position having absolutely no empirical support. whatever they do, you can be sure that it will be full of copious attempts to uphold their point and plead that we stay the course. but, in canada, they can't do what cuomo is doing - because the numbers are worse than predicted.

i don't know what it's going to take for them to admit they were wrong, and this isn't working.

i just hope that, in the end, we all realize that there have been people like me pointing to actual science since day one. the worst thing that can come out of this is a retreat from science, or a return to religious backwardsness.