He’s backing away from his — to me unfathomable — earlier belief in the accuracy of official Chinese government statistics:
My willingness to believe the official line didn’t stem from any credulity about what the Chinese government would do if it believed the truth wouldn’t serve. As Communists they are lying evil scum pretty much by definition, and denial would have been politically attractive for as long as they thought they could nip the pandemic in the bud. I thought their incentives had flipped and they would now be honest as a way of assisting their own countermeasures and seeking international help.
My first clue that I was wrong about that came from a friend who is plugged into the diaspora Chinese community. According to him, there is terrifying video being sent from Chinese clans to the overseas branches they planted in the West to prepare a soft landing in case they have to bail out of China. Video of streets littered with corpses. And of living victims exhibiting symptoms like St. Vitus’s Dance (aka Sydenham’s chorea), which means the virus is attacking central nervous systems.
My second clue was the Tencent leak. […] the takeaway is that there is now reason to believe that as of Feb 1st the actual coronavirus toll looked like this: confirmed cases 154023, suspected cases 79808, cured 269, deaths 24589.
Compare that with the Johns Hopkins tracker numbers for today, a week later: Confirmed cases 31207, cured 1733, deaths 638. Allowing for the Tencent leak being roughly one doubling period earlier, the official statistics have been lowballing the confirmed case number by a factor of about 8 and the deaths by a factor of about 80. And then inflating cures by a factor of about 12.
Even given what I’d heard about the video, I might have remained skeptical about the leak numbers if someone (don’t remember who or where) hadn’t pointed out that the ratio between reported cases and deaths has been suspiciously constant in the official Chinese statistics. In uncooked statistics one would expect more noise in that ratio, if only because of reporting problems.
So my present judgment, subject to change on further evidence, is that the Tencent-leak numbers are the PRC’s actual statistics. And that has a lot of grim implications.
