The utility cyclist wrote:mjal wrote:I have been refraining from commenting on this video in the hope that someone else would do so ; there has been a partial rebuttal but not as clear as I had hoped. So, here goes...
Has The Utility Cyclist actually watched/listened to the video? Either he has done neither or has no understanding whatsoever of statistical method/sampling etc. Mind you, the speakers in the video seem to have no grasp of these concepts either...
Within the first 5 minutes Dr Erickson (main speaker) makes the most basic and outrageous error ; he declares that the day prior to the seminar "California had 33,865 Covid cases out of a total of 280,900 tested...that's 12% of Californians are positive for Covid". This remarkable sleight-of-hand then allows him to conclude that there must then be a total of 4 million plus cases in California and therefore... a very low death rate.
Am I alone in declaring this conclusion to b a complete farrago? Can Erickson (and TUC) not see the obvious fallacy? Were those tested a random sample of the population? I suggest that many/most were already ill (perhaps in hospital) or were contacts of known cases. Dr Erickson, as I recollect, goes on to work a similar "magic" on the New York State figures ; I have not checked as I really cannot bear to watch and listen again.
This utterly nonsensical use of the "raw data" surely means that any other statements in the video are extremely suspect as, by extension, are TUC's opinions if he regards this seminar as "supportive"of his case.
I look forward to TUC's reply to my points about misuse/misunderstanding of the statistics.
Let's see your test results then?
Yes they were a random selection, listen to John Ioannidis the co director of meta research at Stanford Uni who specifically talks about this testing, here's his initial thoughts about a month ago
https://www.youtube.com/watch?v=d6MZy-2 ... e=emb_logo and then his update post the testing
https://www.youtube.com/watch?v=cwPqmLoZA4s&t=109s he concludes that mortality rate is 0.1% give or take.
Maybe you are more qualified than one of the worlds most eminent persons in his field
You saying something is nonsensical and yet produce zero evidence to back yourself up, it makes you look silly at best.
have another go champ
I really wonder where and how to begin. Shall I try a series of easy-to-understand statements / questions which lead to a logical conclusion?
1. You use the video in support of your opinions re Covid-19 : yes/no?
2. Do you understand the concept of random sampling in statistics : yes/no?
3. Do you really believe that, at a time of scarcity of coronavirus tests, that ALL of the 280,000 tests were randomly administered : yes/no?
4. Do you understand the basic statistical concept that one cannot extrapolate from a particular case (a relatively small number of persons tested) to a general one (the population as a whole) unless those tested are representative of the total population i.e (usually) randomly selected : yes/no?
Your answers should (hopefully) be : yes, yes, no, yes. If questions 2-4 are a problem, please do a little reading online.
Another deep breath...
Erickson abuses numbers (3) and (4) and possibly has no grasp of (2) either. This is why I call the video a farrago and nonsensical ; there is no need to reference any external sources.
Your offering of this seriously flawed video therefore casts great doubt on your opinions on this pandemic.
If you cannot understand my points, please seek advice.
PS: Ioannidis ,incidentally ,is quite clear that the tests in question were NOT administered on a random basis. He also reckons that 10,000 American deaths would be a "good outcome" ; today's figure is 58,640 (ref : Worldometer).