Comment by Myrmornis | Hacker News Reader

Comment by Myrmornis | Hacker News Reader

Parent Full thread

No, I don't think that's right. That was my initial impression also, but what's going on is that people are ridiculing the academic advisor who is working at the Council of Economic Advisors at the Whitehouse because his agency (a) released a tweet about Covid with childishly absurd mistakes in it and then (b) he vigorously defended it in a follow-up tweet.

The problem they are ridiculing is mainly that an observed time series that is going up and down with no clear trend is "modeled" with a polynomial that necessarily goes down, and that the tweet claims there is some value in the fitted polynomial.

Secondarily, the fact that the polynomial is described as cubic, when it appears to be quadratic.

Added together it makes it extremely explicitly clear that at the highest levels of American government, even in situations with the involvement of elite academic advisors, the actual technical content is utterly incompetent; childish; totally illiterate from a statistical point of view; bad even at high school.

I came to the same realization, but it was unnecessarily difficult to see through all of the snark. If all parties had simply focused on the salient points of disgreement rather than snarking at each other, it could have been cleared up much more quickly. The GP is right; all sides should be embarrassed.

It would not have "been cleared up", because the CEA didn't make an honest mistake, it was a deliberate attempt to seed disinformation in order to support the President's political agenda.

There's a subset of people that would prefer to think that the halls of power are manned by morons, rather than the criminally selfish.

I don't know what that says about any of us.

Malice and stupidity are not mutually exclusive.

Exactly. It was the epidemiological modeling equivalent of suggesting we all inject disinfectant. Only it was worse because it was politically motivated instead of solely stupid.

Sorry but you’re completely wrong.

If an analysis is making a reasonable effort to use the best or even some sort of relevant methodology in the field maybe. There is an entire field built around the learnings and failures of previous work that informs how we do things now.

In this circumstance the economist came in with no understanding or desire to understand any of that and through some random excel function at the problem to get the answer out they wanted for political expediency.

Not to mention the prediction was ridiculously stupid and anyone with common sense let alone an epidemiology degree could see it was going to be wildly incorrect within days.

THAT is the model being talked about at the highest levels of government when you have the entire field of the world class epidemiologists at your fingertips.

Trying to “both sides” this one because scientists who have spent their whole lives trying working on this got a little snarky is just missing the boat.

There are plenty of other places where epi Twitter is having productive cordial discussions.

Yes, the prediction was clearly stupid. No one is arguing that. But claiming it was the "worst day in the history of the CEA," or words to that effect, amounted to a long and pointless distraction that could have been avoided, and both sides were clearly guilty of it. The whole exchange would have been rightly flagged into oblivion if it appeared on HN.

>Secondarily, the fact that the polynomial is described as cubic, when it appears to be quadratic.

wut. a quadratic only has one inflection point. that plotted curve clearly has 2.

Yes the curve has 2 inflection points, so if it were a polynomial is would have to be quartic. (Cubic polynomials have 1 inflection point at most). Buz I dont think this is a polynomial.

Yep you’re right, I said something daft. It looks pretty much like a Gaussian doesn’t it?

small tails. maybe. it could be anything though with the right parameters.

Definitely not a cubic function though, you can't get a polynomial to hit 0 that nicely. Any polynomial's leading order term will dominate as x increases. Unless they're using a super high order polynomial and hiding the blowup off the edge of the plot.

Reply on news.ycombinator.com