The person who wrote that article knows very little about statistics. The model would predict a range of possibilities, and the possibility right next to the mean of all possibilities is as likely as any other. An exact match with mathematical expectation is unlikely but far from impossible.
Clear as mud...
None of us have a clue.
What Barron's reported: "A statistical analysis of China’s coronavirus casualty data shows a near-perfect prediction model that data analysts say isn’t likely to naturally occur,"
What poor old DonDiego commented: "Barron's is suggesting that China is reporting numbers of cases to match a statiscal model; and the numbers are too close to the model to represent what real data usually looks like, . . . at least a bit messy, like the real world."
What Kevin Lewis opined: "The model would predict a range of possibilities, and the possibility right next to the mean of all possibilities is as likely as any other. An exact match with mathematical expectation is unlikely but far from impossible."
It seems Barron's, DonDiego, and Kevin Lewis are in agreement. DonDiego fails to see any area of dispute.
Not exactly. Barron's opined that "real data" would not show results that matched what the statistical model suggested. That is incorrect. Such results were simply one of the possible data sets that could have been produced.
Let's say I predict that the Dodgers will win 102 games this year--and they do, in fact, win 102 games. It was unlikely that, whatever predictive mechanism I used, the total would fall exactly on 102--but that result was as likely as any other result within the normal range of variance (so, not as likely as, say, 20 or 150).
I am reminded of the British expedition to climb Mt. Everest and measure its height in 1953. They measured its height at exactly 29,000 feet but reported it as 29,002--because they thought that the accurate number would be regarded as an estimate. Sometimes the results are indeed exactly as estimated or predicted.
To put it another way, if the data were indeed "manufactured" by China, they might have introduced some variation from the model's prediction for the same reason--to make it more "believable." I don't doubt that they're massaging the truth on all this to at least some extent. But then, I also don't believe that their numbers could be 100% accurate under any circumstances.
The problem here is that the real aspects of the situation are being drowned in a torrent of misinformation and disinformation. Those who wish to depict China in the worst possible light are giving birth to internet memes that have no basis in reality but sound good to those already hostile to China. And many right-wing fringe organizations, perceiving hostility to China to be an essential part of Party orthodoxy, support and disseminate those conspiracy theories.
Originally posted by: Don
"Scientists believe killer disease may have begun in research facility 300 yards from Wuhan wet fish market"
Ref: Daily Mail
Correction....." Conspiracy theorists in desperate need of psyochosis therapy believe killer disease may have begun in research facility 300 yards from Wuahn wet fish market"
A subset of that group is sitting United States Senator, Tom Cotton, who is publicly giving creedance to this narrative that has as much basis in fact as the one that told us "911 was an inside job".
(FYI - for those who are not familiar with DonDiego's source. It is essentially the UK equvailent of the National Enquirer. )
PJStroh comments: "(FYI - for those who are not familiar with DonDiego's source. It is essentially the UK equvailent of the National Enquirer. )"
Well, . . . apparently the word is finally seeping into more mainstream sources.
From CNN/New York Times journalist Ezra Cheung:
"South China University of Technology biologist Xiao Botao exposed that the #COVID19 might have originated from an animal research lab 280m away from the epicentre of the outbreak. The scientists were reportedly bitten by bats used for a test on louse.
The Washington Post has initiated a page "Mapping the Spread of the New Coronvirus" which is to be updated periodically so long as someone at the Washington Post remains alive.
It also contains data addressing the number of folks infected in each Country - latest post: 69266; 68551 in China alone . And a smaller map dedicated to infections in the USA - now at 15 in 7 states.
In related news: ". . . a renowned law professor at Tsinghua University in Beijing, Xu Zhangrun, is known to be missing after publicly condemning Chinese President Xi Jinping for failing to contain the spread of the virus at an early stage."
As PJ points out, a major source of all the conspiracy theory crap about the coronavirus is (sigh...) conservatives who feel they are serving Trump by painting the virus as the product of some malevolent experiment gone wrong deep in the secret labs of the Chinese equivalent of Dr. Evil.
As I've noted before, a virus that kills fewer than 3% of the people it infects--a percentage even smaller outside China--makes for a pretty lousy weapon. It's like those laser rifles the Imperial stormtroopers use in Star Wars.
DonDiego reports that a law professor in China who criticized the government's handling of the crisis has gone missing...but in China, criticism of the government carries a mortality risk much higher than 3%. That's the real pandemic there.
For what it's worth:
"Don't Buy China's story: The coronavirus may have leaked from a lab"
Ref: New York Post - 22 Feb 2020
"China Knew of Coronavirus in December, ordered cover-up, report says"
Ref: New York Post - 29 Feb 2020
