The Unz Review: An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 TeasersiSteve Blog
How Tell from a Graph if We're Winning
🔊 Listen RSS
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeThanksLOLTroll
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

This graph from March 27 uses averages log scales on both Y and X axes, with time displayed not on X but as animation. It shows average over last seven days so it’s not as sensitive to finally turning the corner.

 
Hide 66 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
  1. Is there anyone left, besides their stenographers in the media who believes the numbers out of China are accurate?

    • Replies: @anon
    Yes. But. they began a program of coercive containment fairly soon. That is, forced isolation of all positives, lockdown of the epicenter, extensive testing, masks, etc. We are not approaching this in a remotely similar fashion. So it is a moot point.

    I could list additional reasons I believe the data is real, but will just say that confirmed cases are actual cases that have been identified and subsequently tested positive. As long as R(0)<1, they don't need to test, so probably haven't extensively.

    People that doubt their data aren't consistent about their skepticism. Their data ended with a high CFR. Did they lie about it being too severe or not severe enough?
  2. Who’s “we”, kemosabe?

  3. In order to know if we’re winning, we need to know what is going on. While I’m willing to believe South Korea’s numbers, is there a valid reason to hold China’s in suspicion?

    • Agree: SINCERITY.net
    • Replies: @SINCERITY.net
    Our Western PC societies, media, politicians, police and courts lie about affirmative action, crime statistics by race, IQ, disparate impact, antifa, leftist faults.
    China lies with equal ruthlessness about everything. So their numbers are probably doctored. But at some time exponential growth will catch up with their lies, when reality is 100 times their lying data
    Korea is interesting. And it seems that most of their success is not just for being a disciplined homogenous nation, but radical approach to testing. The US and Europe could just try to get 500 million cheap test kits, for multiple tests of most of the population.
    Also, is Africa protected by their hot climate. Will it get worse in Southern hemisphere Winter? Especially in South Africa?
  4. As long as we maintain our diversity, we will have won!

    • Replies: @Abe

    As long as we maintain our diversity, we will have won!
     
    “We interrupt our regularly-scheduled Clown World programming to bring you this special announcement...”

    Seriously, why are the people who just LAST MONTH were still egging us on into a live-ammo incident with the world’s only other nuclear superpower over America’s right-to-protect obligations toward Syrian ISIS head harvesters and Ukrainian neo-Nazis allowed any credibility as the serious-minded adults in the room now that a bit of non-socially constructed reality has hit them square over the head?

    Why the desperate measures to save the people of a fundamentally illegitimate country (1619 PROJECT?), unless it’s to preserve us for a suitably “redemptive” finale? To die from a snotty hyper-flu is too shabby, but to expire on the Molson brewery floor after the shooting rampage of the next Omar Thornton/Anthony Ferrill with nothing but words of utter devotion to Diversity-Inclusion-Equality on one’s D-I-E’ing lips, that is divine?

    Shouldn’t we demand apologies from the usual suspects as well as their oligarch overlords for the lies, hate, and misinformation they spread over the many years?

    To get things rolling may I suggest:

    “I, Jeff Bezos, apologize for letting my newly acquired media toy lie to millions of young women that becoming a 260 lb. land-whale at age 22 has zero medical consequences.”
    “I, Bill Gates, apologize for letting my vast media empire defame wonderful, public-spirited health professionals like Dr. Antony Fauci with clickbait about tearing down ‘dude walls.’”

  5. Anon[424] • Disclaimer says:

    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors, giving to Hillary Clinton, Emily’s list, the Democratic Congressional Campaign Committee, etc.

    https://freebeacon.com/latest-news/woman-who-ingested-fish-tank-cleaner-was-prolific-donor-to-democratic-causes/

    One of those causes was a pro-science pac. You’d think someone who respected science would bother to learn a few things about chemicals, first, like finding out if it was safe to eat fish tank cleaner.

    The Democrats must be proud of her and her husband. This couple were willing to commit seppuku (sort of) in the Democratic cause just so they could go out Blaming Trump.

    • Replies: @Cortes
    Fish and Wanda.

    Could be the makings of a madcap movie.

    Or a classic female poisoning trial.

    , @Hypnotoad666

    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors,
     
    They were supposed to drink the Kool-Aid, but couldn't even get that part right.
  6. As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus — and huge numbers of patients have pre-existing conditions — but we’ll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let’s get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    • Agree: Ron Unz, TomSchmidt
    • Replies: @trelane

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.
     
    I've looked too and can't find that data. Apparently it's top secret.
    , @UK
    What are you talking about mate?

    No one here could possibly be as bright as Ron Unz and know what exponential is.

    Nor do the government's legion of science advisers.

    Only Ron gets this incredibly complex concept. Anyone who disagrees with him, simply isn't smart enough.

    It may be in the standard Maths curriculum for every single schoolchild in Britain to learn at about age 14 (for a test at 15/16) but then not everyone has decided that China, a country with a 99.99% criminal conviction rate, is entirely honest - this means they must just not understand experiential growth. Isn't that totally obvious?

    The Chief Medical Officer and Chief Science Officer of the UK could never be expected to get a passing grade at GCSE Maths. That is ludicrous.

    ...

    The above line of reasoning reminds me of when I was a little child and I had solved the problem of making money. I could just keep doubling up bets until I won on almost evens bets at a casino. Didn't the adults understand exponential betting? What a bunch of doofuses! It could never be that I was simply entirely ignorant of a bunch of real world factors that massively complicated my juvenile and very simple mathematical model....
    , @Kratoklastes
    Good data sources -

    covidtracking.com is easily the best for US data, since it set up a half-decent data API.

    US Daily (incl test numbers)
    Latest US State level data

    For Non-Exceptional countries, worldometers has a simple interface, but no 'Tests' data and no API (you can get the data out of the charts with 3 limes of Javascript in the Console in Firefox or Chrome - see the bottom of the page)

    Worldometers country table

    South Korea country page

    For Straya, the direct data page is a shitshow - this amateur effort, which has hard-wired image-based charts (so no drill-down, no tooltips, no data extractability... typical .gov low-quality overpriced shit.

    The CDC's page is so shit and so outdated that it's not worth posting the link, and OurWorldInData is slipping terribly - it hasn't updated test counts for most countries since March 20th.


    Footnote: Getting data out of Highcharts and into CSV-ish format from Worldometer

    • Right-click on the page
    • Select 'Inspect Element' from the Context menu. This will open the Developer Tools window
    • Go to the 'Console' or 'Debugger' tab in the Developer Tools window
    • at the prompt (>> - at bottom of page, usually), type

    csv = 'date,cases\n';
    Highcharts.charts[0].series[1].data.forEach(function(d){ csv += d.category + ',' + d.y +'\n' });
    console.log(csv);

    This will produce a little CSV object that can be pasted into Excel. It's a kludge, but it'll give a quick and dirty way to get the data out (I have something similar scripted to scrape every page, but mine gets the result as JSON and saves it to an SQL table).

    charts[0].series[1] means it will go to the first [0] chart, and get the data for the second [1] series. Change these as you see fit.
    , @Almost Missouri

    "the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread."
     
    I too thought that would distort the results, but as you can see, almost every country travels along a nearly identical path (the Japanese, as always, being the significant outlier), despite the many variations in testing, categorization, etc. So apparently, all of the local variations balance out.

    What the chart does not equalize, however, is national population. So larger countries will naturally travel higher up the scale before their curves collapse. The US should be embarrassed, I suppose, to be travelling higher up the scale than China did with triple our population. OTOH, there is some question about the accuracy of the Chinese numbers, and China is bouncing back somewhat in the last frames.

    Something the chart does not show is whether any given curve collapse is due to countermeasures or simply due to the virus playing itself out in that population, though at this stage, countermeasures seems the more likely explanation.
    , @TomSchmidt
    You read my comment directly from my mind.

    Numbers from Korea seem to show a death rate of infected people south of 1%, about .6%, though they of course also missed some infected. Call it .5%.

    So showing an exponential death chart would not be as dramatic as a case chart. Assume we get to 70% of the population infected. That's 200mm people in the USA. At a .5% death rate (and NY State is currently about 1% of known cases), that's 1mm people. Shockingly high number, but much lower than Spanish Flu as a percentage, which we survived without the sorts of measures we undertook.

    In all this, I keep thinking of the examples from Tufte's Visual display of Quantitative Information, especially his examples of lying, as in lying by omission. See:
    http://davidgiard.com/2011/05/11/DataVisualizationPart4Context.aspx
    , @another fred
    Data linked below. In .csv format. It is the data used in the Johns Hopkins dashboard.

    Click on a file link, hit ctrl A, ctrl C, then past in a spreadsheet (suggest you "paste special" >> text to drop the internal links.

    Once in a spreadsheet it is pretty easy to sort. The format changes somewhat as they developed more detail, but the "Country/Region" column remains as does "Province" to sort by. US counties appear around 3/22 which greatly expands the data.

    You can also copy from CDC (also linked) by sweeping across the data below the graph of cumulative cases. CDC and Github track pretty closely most of the time, not always.

    https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_daily_reports

    https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fcoronavirus%2F2019-ncov%2Fcases-in-us.html
  7. Alas, I don’t watch videos so will someone tell me if Taiwan and Singapore are accounted for?

    https://www.unz.com/isteve/social-distancing-hasnt-caught-on-yet-in-south-central-l-a/#comment-3804985

  8. Don’t use log(days) for x; keep days linear.

    Instead of using a seven day average, try to calculate a least squares exponential fit whenever a run of four or more days shows a new trend.

    Avail yourself of https://www.padowan.dk/ to acquire Graph v4.2.2 iffn you hate doing the calulations manually. If you can’t remember the calcs. just assume a form for the function, write a series for f(x) – data(x), take the derivative. You know, it’s been so long since I actually did that! (could be wrong)

    • Replies: @James Speaks

    Instead of using a seven day average, try to calculate a least squares exponential fit whenever a run of four or more days shows a new trend.
     
    The reason for this is you can spot the effect of change in policy. Eg, muh state instituted changes 12 days ago and a new trend line evidenced.
    , @Rob

    Don’t use log(days) for x; keep days linear.
     
    It’s not log(cases) vs log(time). It’s log(new cases) vs log(total cases). Change over time is shown by animation. It’s a clever way of showing that the epidemic has functioned similarly in a bunch of countries, until they’ve done something to drastically reduce transmission. It’s pretty disturbing that the epidemic has been pretty similar everywhere at similar numbers of total cases. One would hope that warm, wet countries were on a different line than temperate places. Cuts down the hope of spring and summer saving us until next winter.
  9. NEWS FLASH

    President Xi, now appearing in public without a mask in a photo op, has found a way to stop the spread of the virus!

    If you take one step too many near him, his masked body guards shoot you dead.

    Long Live Emperor Xi! When he says China has turned the corner, I believe all he says (so long as he doesn’t breathe on me).

  10. Better still is the reason why this happens.
    Growth of 1.25 doesn’t sound incredible to the ordinary person’s ears especially when initial cases are reported under 500 but as readers here know it means a doubling almost every two weeks.

    There’s a flip side to this: what we attack is the “growth curve”, the growth curve is the 0.25 part, not the 1.0 part. Merely drawing the growth curve down to 1.0 and holding it there causes viral collapse.

    • Replies: @James Speaks

    Better still is the reason why this happens.
    Growth of 1.25 doesn’t sound incredible to the ordinary person’s ears especially when initial cases are reported under 500 but as readers here know it means a doubling almost every two weeks.
     
    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.

    Try this: n-tupling every D days.

    n = x^D

    ln n = D ln x

    ln x = (ln n)/D call this c

    x = e^c

    ***

    If you have access to a calculator and don't need to use your hand calculated natural log tables ...

    x = Dth root of n

  11. @Anon
    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors, giving to Hillary Clinton, Emily's list, the Democratic Congressional Campaign Committee, etc.

    https://freebeacon.com/latest-news/woman-who-ingested-fish-tank-cleaner-was-prolific-donor-to-democratic-causes/

    One of those causes was a pro-science pac. You'd think someone who respected science would bother to learn a few things about chemicals, first, like finding out if it was safe to eat fish tank cleaner.

    The Democrats must be proud of her and her husband. This couple were willing to commit seppuku (sort of) in the Democratic cause just so they could go out Blaming Trump.

    Fish and Wanda.

    Could be the makings of a madcap movie.

    Or a classic female poisoning trial.

  12. Mr Sailer! Mr Sailer! America has turned the corner!

    Yes, we’ve turned the corner- and are now steadily going down China St.

    It is true at the end of the street you do come up face to face with a row of tanks.

    But at least you don’t die from the virus.

    Save Steve Sailer! Martial Law now!

  13. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    I’ve looked too and can’t find that data. Apparently it’s top secret.

    • Replies: @res
    This page has graphics which support displaying death data.
    https://91-divoc.com/pages/covid-visualization/
  14. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    What are you talking about mate?

    No one here could possibly be as bright as Ron Unz and know what exponential is.

    Nor do the government’s legion of science advisers.

    Only Ron gets this incredibly complex concept. Anyone who disagrees with him, simply isn’t smart enough.

    It may be in the standard Maths curriculum for every single schoolchild in Britain to learn at about age 14 (for a test at 15/16) but then not everyone has decided that China, a country with a 99.99% criminal conviction rate, is entirely honest – this means they must just not understand experiential growth. Isn’t that totally obvious?

    The Chief Medical Officer and Chief Science Officer of the UK could never be expected to get a passing grade at GCSE Maths. That is ludicrous.

    The above line of reasoning reminds me of when I was a little child and I had solved the problem of making money. I could just keep doubling up bets until I won on almost evens bets at a casino. Didn’t the adults understand exponential betting? What a bunch of doofuses! It could never be that I was simply entirely ignorant of a bunch of real world factors that massively complicated my juvenile and very simple mathematical model….

  15. –uses averages log scales
    Not a math person but this looks like a typo. It’s showing an average from both log scales, with one log scale on each axis?

  16. Steve, I appreciate your way to insert math and science into topics. Maybe you can explain
    a) why the growth rate curve would not immediately explode, as soon as quarantine gets lifted.
    b) why other flu viruses finally fizzled out and don’t re-start the following years. Is that because most people have caught it, or are there more pleasant ways
    c) hot air seems to kill the virus even in the body. Should we all go to Saunas, or breathe steam from hot water or breathe our hair dryer air?

    • Replies: @Charon

    Steve, I appreciate your way to insert math and science into topics.
     
    LOL

    hot air seems to kill the virus even in the body. Should we all go to Saunas, or breathe steam from hot water or breathe our hair dryer air?
     
    Too easy. Not touchin' it.
  17. anon[117] • Disclaimer says:

    Reposting this link: new Petri dish

    https://www.msn.com/en-us/news/us/four-dead-138-sick-on-holland-america-s-ms-zaandam-cruise-in-limbo-amid-coronavirus-crisis/ar-BB11Oc2j?ocid=msedgntp

    Departed Argentina March 7, made one port call but apparently no one removed.
    It’s data.

    Viet Nam has almost no ICU capability, but so far their numbers are low. It’s not so difficult in an East Asian police state.

    https://www.dw.com/en/how-vietnam-is-winning-its-war-on-coronavirus/a-52929967

    They got one thing going for them: a handwashing song. Not quite K-pop but…

  18. @Half Canadian
    In order to know if we're winning, we need to know what is going on. While I'm willing to believe South Korea's numbers, is there a valid reason to hold China's in suspicion?

    Our Western PC societies, media, politicians, police and courts lie about affirmative action, crime statistics by race, IQ, disparate impact, antifa, leftist faults.
    China lies with equal ruthlessness about everything. So their numbers are probably doctored. But at some time exponential growth will catch up with their lies, when reality is 100 times their lying data
    Korea is interesting. And it seems that most of their success is not just for being a disciplined homogenous nation, but radical approach to testing. The US and Europe could just try to get 500 million cheap test kits, for multiple tests of most of the population.
    Also, is Africa protected by their hot climate. Will it get worse in Southern hemisphere Winter? Especially in South Africa?

  19. @SINCERITY.net
    Steve, I appreciate your way to insert math and science into topics. Maybe you can explain
    a) why the growth rate curve would not immediately explode, as soon as quarantine gets lifted.
    b) why other flu viruses finally fizzled out and don't re-start the following years. Is that because most people have caught it, or are there more pleasant ways
    c) hot air seems to kill the virus even in the body. Should we all go to Saunas, or breathe steam from hot water or breathe our hair dryer air?

    Steve, I appreciate your way to insert math and science into topics.

    LOL

    hot air seems to kill the virus even in the body. Should we all go to Saunas, or breathe steam from hot water or breathe our hair dryer air?

    Too easy. Not touchin’ it.

  20. From the news stories I can find, Santa Clara county, CA (where I live) is not publishing the ages or prior health conditions of the 25 dead from COVOID19.

    I have to assume it’s because this data would undermine the panic they are attempting to induce in order to keep us locked up for another month.

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19.

    And yes, I understand exponential growth, but the cases here are not growing exponentially anymore–the curve is starting to bend down.

    My hunch is that this is all overblown B.S.

    • Replies: @James Speaks

    My hunch is that this is all overblown B.S.
     
    My hunch is that you like to say "my hunch" to assert intellectual dominance without expending any effort.

    The world is increasing about 16% per day, the US about 18% per day, and muh state (thank you Ron DeIdiot) currently at 28% per day.
    , @Ron Unz

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19...My hunch is that this is all overblown B.S.
     
    Well, here's an interesting news item for the (hopefully) dwindling ranks of the Coronavirus Hoaxers who plug their ears and shout "It's Just the Flu!!!"...

    Italy's medical association has now reported that 61 Italian doctors have already died of Coronavirus during the current outbreak:

    https://www.unz.com/runz/the-government-employee-who-may-have-saved-a-million-american-lives/#comment-3806303

    If that’s the typical mortality rate for Italian doctors every flu season, their medical schools must have a hard time recruiting applicants...
  21. @James Speaks
    Don't use log(days) for x; keep days linear.

    Instead of using a seven day average, try to calculate a least squares exponential fit whenever a run of four or more days shows a new trend.

    Avail yourself of https://www.padowan.dk/ to acquire Graph v4.2.2 iffn you hate doing the calulations manually. If you can't remember the calcs. just assume a form for the function, write a series for f(x) - data(x), take the derivative. You know, it's been so long since I actually did that! (could be wrong)

    Instead of using a seven day average, try to calculate a least squares exponential fit whenever a run of four or more days shows a new trend.

    The reason for this is you can spot the effect of change in policy. Eg, muh state instituted changes 12 days ago and a new trend line evidenced.

  22. @SimplePseudonymicHandle
    Better still is the reason why this happens.
    Growth of 1.25 doesn't sound incredible to the ordinary person's ears especially when initial cases are reported under 500 but as readers here know it means a doubling almost every two weeks.

    There's a flip side to this: what we attack is the "growth curve", the growth curve is the 0.25 part, not the 1.0 part. Merely drawing the growth curve down to 1.0 and holding it there causes viral collapse.

    Better still is the reason why this happens.
    Growth of 1.25 doesn’t sound incredible to the ordinary person’s ears especially when initial cases are reported under 500 but as readers here know it means a doubling almost every two weeks.

    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.

    Try this: n-tupling every D days.

    n = x^D

    ln n = D ln x

    ln x = (ln n)/D call this c

    x = e^c

    ***

    If you have access to a calculator and don’t need to use your hand calculated natural log tables …

    x = Dth root of n

    • Replies: @AnotherDad

    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.
     
    James, the period here isn't days, it's essentially the mean time between being infected and infecting other people.

    That time is all over the map, but the period between infection and symptoms seems to be about four days. Probably the infection time would be a little long as it is once you show symptoms that you are spewing the virus out.

    If you pick some number, say six days, then at a 1.25 replication, the doubling time would be 18 days. Chop it back to 5 days and you get your two weeks.

    ~~

    My bet is that the replication is below 1 for the US. Certainly in the communities that have locked down. I think you have high transference within households, but i doubt that there's transference between households that is near one.

    As i said in a previous comment, i think the stats will start to show some "community divergence" as different "population groups" have different levels of compliance with hygiene and lockdown directives.
  23. Don’t want to be negative, these guys went out and did something that will be informative to some people.

    But the big problem with it is #4 “testing”.

    Of more specifically the fuzziness of “cases”. During this entire deal, we haven’t had any handle on what the real “cases” are. A) Because testing lagged (particularly in the US with the CDC debacle). But B) because unless you test everyone in some sample you have no idea because so many people (half?) are asymptomatic or just have something that’s indistinguishable for a typical chest cold.

    This is why i’ve consistenly kept quoting the Diamond Princess numbers. Because it’s the only place we actually have numbers. (And guess what? This thing isn’t the end of the world.)

    ~~

    So, how do we tell if we’re winning?

    Deaths.

    And … this weekend the US death rate … stalled–at 450ish.

    Now i don’t know if this is because the death panels that allocate flu and heart disease and cancer deaths to Covid-19 were off and the B-team just didn’t know what they were supposed to do?

    Or … it could be that the social isolation measures put in place a couple weeks back are now showing up in the death rate. Winning!

    What i do know is if this keeps up … we’re going to find that compliance and ergo lower death rate is better for some “population groups” than others.

    And then we’re going to have a spasm of articles explaining that flyover country white gentiles have been hoarding all the social isolation to themselves. Racism!

  24. @Farenheit
    As long as we maintain our diversity, we will have won!

    As long as we maintain our diversity, we will have won!

    “We interrupt our regularly-scheduled Clown World programming to bring you this special announcement…”

    Seriously, why are the people who just LAST MONTH were still egging us on into a live-ammo incident with the world’s only other nuclear superpower over America’s right-to-protect obligations toward Syrian ISIS head harvesters and Ukrainian neo-Nazis allowed any credibility as the serious-minded adults in the room now that a bit of non-socially constructed reality has hit them square over the head?

    Why the desperate measures to save the people of a fundamentally illegitimate country (1619 PROJECT?), unless it’s to preserve us for a suitably “redemptive” finale? To die from a snotty hyper-flu is too shabby, but to expire on the Molson brewery floor after the shooting rampage of the next Omar Thornton/Anthony Ferrill with nothing but words of utter devotion to Diversity-Inclusion-Equality on one’s D-I-E’ing lips, that is divine?

    Shouldn’t we demand apologies from the usual suspects as well as their oligarch overlords for the lies, hate, and misinformation they spread over the many years?

    To get things rolling may I suggest:

    “I, Jeff Bezos, apologize for letting my newly acquired media toy lie to millions of young women that becoming a 260 lb. land-whale at age 22 has zero medical consequences.”
    “I, Bill Gates, apologize for letting my vast media empire defame wonderful, public-spirited health professionals like Dr. Antony Fauci with clickbait about tearing down ‘dude walls.’”

  25. @Anon
    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors, giving to Hillary Clinton, Emily's list, the Democratic Congressional Campaign Committee, etc.

    https://freebeacon.com/latest-news/woman-who-ingested-fish-tank-cleaner-was-prolific-donor-to-democratic-causes/

    One of those causes was a pro-science pac. You'd think someone who respected science would bother to learn a few things about chemicals, first, like finding out if it was safe to eat fish tank cleaner.

    The Democrats must be proud of her and her husband. This couple were willing to commit seppuku (sort of) in the Democratic cause just so they could go out Blaming Trump.

    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors,

    They were supposed to drink the Kool-Aid, but couldn’t even get that part right.

    • Replies: @anon
    Well, sure, that's because it was Flavor-Aid not Kool-Aid.
  26. @Hypnotoad666

    It turns out that lady and her husband who drank the fish cleaner were big Democratic donors,
     
    They were supposed to drink the Kool-Aid, but couldn't even get that part right.

    Well, sure, that’s because it was Flavor-Aid not Kool-Aid.

  27. tl;dw

    What’s the gist or point of the thing?

  28. @Spud Boy
    From the news stories I can find, Santa Clara county, CA (where I live) is not publishing the ages or prior health conditions of the 25 dead from COVOID19.

    I have to assume it's because this data would undermine the panic they are attempting to induce in order to keep us locked up for another month.

    If 30K Americans die annually of the flu, we'd expect Santa Clara county to have ~150 deaths from the flu each year. We're at 25 from COVID19.

    And yes, I understand exponential growth, but the cases here are not growing exponentially anymore--the curve is starting to bend down.

    My hunch is that this is all overblown B.S.

    My hunch is that this is all overblown B.S.

    My hunch is that you like to say “my hunch” to assert intellectual dominance without expending any effort.

    The world is increasing about 16% per day, the US about 18% per day, and muh state (thank you Ron DeIdiot) currently at 28% per day.

  29. @Spud Boy
    From the news stories I can find, Santa Clara county, CA (where I live) is not publishing the ages or prior health conditions of the 25 dead from COVOID19.

    I have to assume it's because this data would undermine the panic they are attempting to induce in order to keep us locked up for another month.

    If 30K Americans die annually of the flu, we'd expect Santa Clara county to have ~150 deaths from the flu each year. We're at 25 from COVID19.

    And yes, I understand exponential growth, but the cases here are not growing exponentially anymore--the curve is starting to bend down.

    My hunch is that this is all overblown B.S.

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19…My hunch is that this is all overblown B.S.

    Well, here’s an interesting news item for the (hopefully) dwindling ranks of the Coronavirus Hoaxers who plug their ears and shout “It’s Just the Flu!!!”…

    Italy’s medical association has now reported that 61 Italian doctors have already died of Coronavirus during the current outbreak:

    https://www.unz.com/runz/the-government-employee-who-may-have-saved-a-million-american-lives/#comment-3806303

    If that’s the typical mortality rate for Italian doctors every flu season, their medical schools must have a hard time recruiting applicants…

    • Replies: @JohnnyWalker123
    What's your opinion of the effectiveness of hydroxychloroquine?

    Trump keeps touting this drug. The FDA just issued emergency approval to hydroxychloroquine for its usage as an anti-Corona drug.

    French microbiologist Didier Raoult has done two studies of the drug. Both studies had very promising results.
    , @TomSchmidt
    According to this article, there are 336,000 medical doctors inItaly. I'm going to assume that the minimum age for a medical doctor is 30 years old.

    https://www.ncbi.nlm.nih.gov/pubmed/12050938

    Last year, the death rate in Italy was 10.7/1000. So I would expect medical doctors to be at least as bad as that rate, and probably worse, being older. That means about 3500 doctors should die in Italy every year. Call it 300 a month, no?

    So, 1/5th of the expected doctor cull this month was fromCorona? And this is outrageous... why?

    You see a problem with math affiliated and think you can run in and help. I think, Sir, that you are fundamentally honest, maybe too honest to think ill of the people who are lying, spinning, and obfuscating with bad data. But they're a malicious ruling elite, in the country whose genius political adviser told the Prince: don't keep your promises. No one else will, and no one expects it of you.

    Of course, that guy also famously wrote that people will more easily forgive the murder of their parents than being made poor. If he is right, the elites made a bad choice in tanking the economy In the hopes of saving people's parents.
    , @TomSchmidt
    If COVID were as deadly as the flu, I'd be worried. But it's not. That's not me saying it, it's the Italian government. But what would those guys know?

    Go through and check the graphs yourself:
    https://www.zerohedge.com/geopolitical/whats-behind-italys-outrageous-10-mortality-rate-covid-19

    Italy is just now catching up to the flu-caused over-65 death rates of the previous two winters. COVID still hasn't made things worse than three years ago, when the country wasn't locked down at all.
  30. @James Speaks

    Better still is the reason why this happens.
    Growth of 1.25 doesn’t sound incredible to the ordinary person’s ears especially when initial cases are reported under 500 but as readers here know it means a doubling almost every two weeks.
     
    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.

    Try this: n-tupling every D days.

    n = x^D

    ln n = D ln x

    ln x = (ln n)/D call this c

    x = e^c

    ***

    If you have access to a calculator and don't need to use your hand calculated natural log tables ...

    x = Dth root of n

    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.

    James, the period here isn’t days, it’s essentially the mean time between being infected and infecting other people.

    That time is all over the map, but the period between infection and symptoms seems to be about four days. Probably the infection time would be a little long as it is once you show symptoms that you are spewing the virus out.

    If you pick some number, say six days, then at a 1.25 replication, the doubling time would be 18 days. Chop it back to 5 days and you get your two weeks.

    ~~

    My bet is that the replication is below 1 for the US. Certainly in the communities that have locked down. I think you have high transference within households, but i doubt that there’s transference between households that is near one.

    As i said in a previous comment, i think the stats will start to show some “community divergence” as different “population groups” have different levels of compliance with hygiene and lockdown directives.

    • Replies: @Rob
    I’m not one to defend them, but why should the blacks playing ball stop so that white grandmas doen’t die? From their perspective, I mean. Their grandmothers are in their thirties. Sure, you might say that as citizens of the same country they have a duty to protect their fellow citizens, but I doubt they see it that way. The Hispanics? Are the illegals getting the federal payout? I could see it having gone either way. Even if Congress didn’t include them, are you sure a judge won’t rule that they’re entitled? Even the legal ones are far younger on average, and their grandmothers live in Mexico and Squatemala. They don’t have much to lose from working and going out. We can hope that the sudden economic turn sends lots of them home.

    Don’t get me wrong, I’d repatriate blacks and Hispanics if I had the ability, but I don’t see how any voluntary measures will interest them in any significant numbers. They aren’t good citizens in the best of times. Will governments compel them? Send out more cops to break up birthday parties?

    Finally, what are the odds of reimportation? East Asian countries have imported cases from aliens going home. Is there any chance the third world will be able to have everyone shelter in place for weeks? Will they even want to, since they don’t have many old people compared to civilized countries? Maybe the coronavirus won’t spread much in warm, humid countries. Without borders, what will keep it out if it does? If aliens had enforced quarantine for a couple weeks coming and going, that would cut down on international travel, so it would be a net positive.
    , @James Speaks

    James, the period here isn’t days, it’s essentially the mean time between being infected and infecting other people
     
    Nope. The period here is days, as in 25% more each day.

    The US is running about 18% but NYC exceeded 30%. Ergo doubling in three days as in actual data.

    Wishful thinking otherwise.
  31. The sign of winning is that we won’t have to be graphing future outbreaks on log scale, simple linear scale will do.

    At that point it won’t be an, oh yet again, that word, exponential threat. As long as we have to represent it log style, we face an exponentially varying situation, which is inherently unstable and not winning.

    Next question? You’re making this too easy, Steve.

  32. corona theatre is really a good show! lol

  33. The reasons for a change in the South Korean trajectory are interesting, but one key thing over the past 2 weeks is a 20% decline in tests per day.

    They’ve gone from testing ~11.6k people a day (the average daily for weeks ending March 2 to March 16), to testing 9k people a day on the week ending March 23 and 8165 people over the the last week (and fewer than 2,000 yesterday).

    Nothing will make ‘new confirmed cases’ decline faster than reducing the number of people you test: stop testing altogether and ‘new confirmed cases’ goes to zero the same day.

    However the much more important contributor, is that the early testing was far more selective than recent testing.

    As we all know, since mid-March testing in SKor has had the admirable goal of getting a genuine ‘society wide’ sample, as opposed to just testing sick people who turn up at hospitals or show symptoms which is what everyone else has been doing. Australia also allows really insistent hypochondriacs to get themselves tested, but not in numbers large enough to de-bias Australia’s <2% positives-to-tests ratio, which has been fairly stable since well before any 'social distancing malarkey.

    .

    The big SKor numbers prior to early March were largely from the Shincheonji megachurch in Daegu, which is even more lunatic and End-of-Time-y than Yank Rapturetards. It also had a bunch of people in the congregation who spent the last weeks of 2019 frolicking[1] in a little town called “Wuhan”.

    The first death was a member of the congregation, and the South Korean government marked its ~200,000 members as priorities for testing. That cohort contributed a very large proportion of the early ‘confirmed cases’.

    By early March the SKors had done ~147,000 tests, and South Korea’s total positive tests stood at about 5,000. The Shincheonji ‘supercluster’ had contributed 60% of all positive tests at that stage, and a whole fuckload of them refused to be tested.

    There have been another 210,000 tests since, and the positives-to-test ratio has dropped back down towards what looks like a constant.

    Obviously if your primary testing cohort focuses on plague-infected retarded jackoffs, you run out of them eventually.

    By March 12/13 the SKors were pretty much only testing non-retard jackoffs.

    Ça explique tout (à peu près).

    One thing to note is that over the last 2 weeks, the drop in the total number of tests means that the denominator of the positives-to-tests ratio has fallen by ~20% but the number of positives has stayed (about) the same… in other words the positives-to-tests ratio is actually slightly higher (but it’s statistically unchanged).

    So to the extent that SKor’s policies did anything, they made risks for non-retards very slightly worse (but not in a way that is statistically meaningful).

    .

    Bear in mind, too, that more than half of SKor’s 9500-ish positive tests have since been released – and never exceeded 7500 ‘currently infected’ at any time, and 99% of all symptomatic-infected have been ‘mild’.

    SKor has only had 162 deaths – 1.7% of all detected positives (and so strictly less than 1.7% of all positives, and strictly much less than 1.7% of the population).

    Almost all the deaths were in old, infirm people with multiple existing ailments, and there might be a touch of that “with covid19, not of covid19” in the numbers.

    [1] I am assuming that retarded Armageddon-tards frolic – I’m not a theologian.

    • Replies: @TomSchmidt
    You got a blog where you've assembled all these comments together? You and Westhunt ought to debate this. Masks on, please.
  34. So all these high highfalutin folks get the chinkyflu: do we get to know how their doctors are treating THEM? lol

  35. @trelane

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.
     
    I've looked too and can't find that data. Apparently it's top secret.

    This page has graphics which support displaying death data.
    https://91-divoc.com/pages/covid-visualization/

    • Thanks: trelane
  36. anon[594] • Disclaimer says:
    @Barnard
    Is there anyone left, besides their stenographers in the media who believes the numbers out of China are accurate?

    Yes. But. they began a program of coercive containment fairly soon. That is, forced isolation of all positives, lockdown of the epicenter, extensive testing, masks, etc. We are not approaching this in a remotely similar fashion. So it is a moot point.

    I could list additional reasons I believe the data is real, but will just say that confirmed cases are actual cases that have been identified and subsequently tested positive. As long as R(0)<1, they don't need to test, so probably haven't extensively.

    People that doubt their data aren't consistent about their skepticism. Their data ended with a high CFR. Did they lie about it being too severe or not severe enough?

  37. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    Good data sources –

    covidtracking.com is easily the best for US data, since it set up a half-decent data API.

    US Daily (incl test numbers)
    Latest US State level data

    For Non-Exceptional countries, worldometers has a simple interface, but no ‘Tests’ data and no API (you can get the data out of the charts with 3 limes of Javascript in the Console in Firefox or Chrome – see the bottom of the page)

    Worldometers country table

    South Korea country page

    For Straya, the direct data page is a shitshow – this amateur effort, which has hard-wired image-based charts (so no drill-down, no tooltips, no data extractability… typical .gov low-quality overpriced shit.

    The CDC’s page is so shit and so outdated that it’s not worth posting the link, and OurWorldInData is slipping terribly – it hasn’t updated test counts for most countries since March 20th.

    Footnote: Getting data out of Highcharts and into CSV-ish format from Worldometer

    • Right-click on the page
    • Select ‘Inspect Element’ from the Context menu. This will open the Developer Tools window
    • Go to the ‘Console’ or ‘Debugger’ tab in the Developer Tools window
    • at the prompt (>> – at bottom of page, usually), type

    csv = ‘date,cases\n’;
    Highcharts.charts[0].series[1].data.forEach(function(d){ csv += d.category + ‘,’ + d.y +’\n’ });
    console.log(csv);

    This will produce a little CSV object that can be pasted into Excel. It’s a kludge, but it’ll give a quick and dirty way to get the data out (I have something similar scripted to scrape every page, but mine gets the result as JSON and saves it to an SQL table).

    charts[0].series[1] means it will go to the first [0] chart, and get the data for the second [1] series. Change these as you see fit.

    • Thanks: TomSchmidt
    • Replies: @Buzz Mohawk
    Thank you.
  38. @AnotherDad

    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.
     
    James, the period here isn't days, it's essentially the mean time between being infected and infecting other people.

    That time is all over the map, but the period between infection and symptoms seems to be about four days. Probably the infection time would be a little long as it is once you show symptoms that you are spewing the virus out.

    If you pick some number, say six days, then at a 1.25 replication, the doubling time would be 18 days. Chop it back to 5 days and you get your two weeks.

    ~~

    My bet is that the replication is below 1 for the US. Certainly in the communities that have locked down. I think you have high transference within households, but i doubt that there's transference between households that is near one.

    As i said in a previous comment, i think the stats will start to show some "community divergence" as different "population groups" have different levels of compliance with hygiene and lockdown directives.

    I’m not one to defend them, but why should the blacks playing ball stop so that white grandmas doen’t die? From their perspective, I mean. Their grandmothers are in their thirties. Sure, you might say that as citizens of the same country they have a duty to protect their fellow citizens, but I doubt they see it that way. The Hispanics? Are the illegals getting the federal payout? I could see it having gone either way. Even if Congress didn’t include them, are you sure a judge won’t rule that they’re entitled? Even the legal ones are far younger on average, and their grandmothers live in Mexico and Squatemala. They don’t have much to lose from working and going out. We can hope that the sudden economic turn sends lots of them home.

    Don’t get me wrong, I’d repatriate blacks and Hispanics if I had the ability, but I don’t see how any voluntary measures will interest them in any significant numbers. They aren’t good citizens in the best of times. Will governments compel them? Send out more cops to break up birthday parties?

    Finally, what are the odds of reimportation? East Asian countries have imported cases from aliens going home. Is there any chance the third world will be able to have everyone shelter in place for weeks? Will they even want to, since they don’t have many old people compared to civilized countries? Maybe the coronavirus won’t spread much in warm, humid countries. Without borders, what will keep it out if it does? If aliens had enforced quarantine for a couple weeks coming and going, that would cut down on international travel, so it would be a net positive.

    • LOL: TomSchmidt
  39. @James Speaks
    Don't use log(days) for x; keep days linear.

    Instead of using a seven day average, try to calculate a least squares exponential fit whenever a run of four or more days shows a new trend.

    Avail yourself of https://www.padowan.dk/ to acquire Graph v4.2.2 iffn you hate doing the calulations manually. If you can't remember the calcs. just assume a form for the function, write a series for f(x) - data(x), take the derivative. You know, it's been so long since I actually did that! (could be wrong)

    Don’t use log(days) for x; keep days linear.

    It’s not log(cases) vs log(time). It’s log(new cases) vs log(total cases). Change over time is shown by animation. It’s a clever way of showing that the epidemic has functioned similarly in a bunch of countries, until they’ve done something to drastically reduce transmission. It’s pretty disturbing that the epidemic has been pretty similar everywhere at similar numbers of total cases. One would hope that warm, wet countries were on a different line than temperate places. Cuts down the hope of spring and summer saving us until next winter.

    • Replies: @James Speaks


    It’s not log(cases) vs log(time). It’s log(new cases) vs log(total cases.
     
    You are correct. I had responded before watching the video. Bad me. Bad, bad me.

    I still prefer to plot total cases versus time and to use the least squares fit to an exponential plotted semi log to observe nuances. If the lines are parallel ...

    BTW in my graphs time is days since 100 cases.
  40. @Ron Unz

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19...My hunch is that this is all overblown B.S.
     
    Well, here's an interesting news item for the (hopefully) dwindling ranks of the Coronavirus Hoaxers who plug their ears and shout "It's Just the Flu!!!"...

    Italy's medical association has now reported that 61 Italian doctors have already died of Coronavirus during the current outbreak:

    https://www.unz.com/runz/the-government-employee-who-may-have-saved-a-million-american-lives/#comment-3806303

    If that’s the typical mortality rate for Italian doctors every flu season, their medical schools must have a hard time recruiting applicants...

    What’s your opinion of the effectiveness of hydroxychloroquine?

    Trump keeps touting this drug. The FDA just issued emergency approval to hydroxychloroquine for its usage as an anti-Corona drug.

    French microbiologist Didier Raoult has done two studies of the drug. Both studies had very promising results.

    • Replies: @anon
    This has already been used widely in hospitals and ER's. The big change is that the FDA has blister packs ready to distribute widely.

    Prior to this, the limited supply was being hoarded by doctors prescribing large quantities to family and friends with refills.

    “We wanted to try to get out in front of that as early as we could,” said Nicki Chopski, executive
    director of the board in Idaho, where pharmacists began reporting a significant uptick in prescriptions for the medications last week. The prescriptions, she said, were being written by doctors for themselves and their family members, often in large quantities with refills.
     
    https://www.nytimes.com/2020/03/24/business/doctors-buying-coronavirus-drugs.html

    Everyone was already using it and the FDA was holding off on wider approval until they had adequate supplies.
  41. @AnotherDad

    1.25 x 1.25 x 1.25 = 1.95 Doubling every three days.
     
    James, the period here isn't days, it's essentially the mean time between being infected and infecting other people.

    That time is all over the map, but the period between infection and symptoms seems to be about four days. Probably the infection time would be a little long as it is once you show symptoms that you are spewing the virus out.

    If you pick some number, say six days, then at a 1.25 replication, the doubling time would be 18 days. Chop it back to 5 days and you get your two weeks.

    ~~

    My bet is that the replication is below 1 for the US. Certainly in the communities that have locked down. I think you have high transference within households, but i doubt that there's transference between households that is near one.

    As i said in a previous comment, i think the stats will start to show some "community divergence" as different "population groups" have different levels of compliance with hygiene and lockdown directives.

    James, the period here isn’t days, it’s essentially the mean time between being infected and infecting other people

    Nope. The period here is days, as in 25% more each day.

    The US is running about 18% but NYC exceeded 30%. Ergo doubling in three days as in actual data.

    Wishful thinking otherwise.

  42. @Rob

    Don’t use log(days) for x; keep days linear.
     
    It’s not log(cases) vs log(time). It’s log(new cases) vs log(total cases). Change over time is shown by animation. It’s a clever way of showing that the epidemic has functioned similarly in a bunch of countries, until they’ve done something to drastically reduce transmission. It’s pretty disturbing that the epidemic has been pretty similar everywhere at similar numbers of total cases. One would hope that warm, wet countries were on a different line than temperate places. Cuts down the hope of spring and summer saving us until next winter.

    It’s not log(cases) vs log(time). It’s log(new cases) vs log(total cases.

    You are correct. I had responded before watching the video. Bad me. Bad, bad me.

    I still prefer to plot total cases versus time and to use the least squares fit to an exponential plotted semi log to observe nuances. If the lines are parallel …

    BTW in my graphs time is days since 100 cases.

  43. People that doubt their data aren’t consistent about their skepticism. Their data ended with a high CFR. Did they lie about it being too severe or not severe enough?

    I don’t know or care whether Chinese believe the things they say. They publicly said on the 7th January 2020 that it was not easily transmitted from person to person. They were reported as saying the CFR was between 5 and 15 per cent. Spanish flu territory. So the Chinese misled early, often, and consistently. We now know what to expect from them: bullcrap. The one thing to remember is they invited Professor Ian Lipkin over and he got infected with the supposedly lethal disease he was supposed to be investigating. Thank you, masked men!.

    • Replies: @anon
    @ Sean

    What were they supposed to have known by Jan 7? There was no data that early. They thought it was serious enough to shut down their economy.

    Here is their data as reported in JAMA as of Feb 24.

    https://jamanetwork.com/journals/jama/fullarticle/2762130

    Cases plotted by date of symptoms looks quite authentic.

    I would like to see US data plotted by symptom emergence date rather than diagnosed/tested date.
  44. anon[594] • Disclaimer says:
    @Sean

    People that doubt their data aren't consistent about their skepticism. Their data ended with a high CFR. Did they lie about it being too severe or not severe enough?
     
    I don't know or care whether Chinese believe the things they say. They publicly said on the 7th January 2020 that it was not easily transmitted from person to person. They were reported as saying the CFR was between 5 and 15 per cent. Spanish flu territory. So the Chinese misled early, often, and consistently. We now know what to expect from them: bullcrap. The one thing to remember is they invited Professor Ian Lipkin over and he got infected with the supposedly lethal disease he was supposed to be investigating. Thank you, masked men!.

    @ Sean

    What were they supposed to have known by Jan 7? There was no data that early. They thought it was serious enough to shut down their economy.

    Here is their data as reported in JAMA as of Feb 24.

    https://jamanetwork.com/journals/jama/fullarticle/2762130

    Cases plotted by date of symptoms looks quite authentic.

    I would like to see US data plotted by symptom emergence date rather than diagnosed/tested date.

    • Replies: @Sean

    https://www.unz.com/proberts/marc-wathelet-a-virologist-and-specialist-in-coronaviruses-and-respiratory-diseases-explains-why-asia-was-successful-in-containing-covid-19-and-why-the-west-is-not/
    They made two errors with absolutely catastrophic consequences in their management of this crisis. The first was to believe that this new coronavirus was transmitted in the same way as the two recently emerged coronaviruses, SARS and MERS. It is the classic mistake of generals to prepare for the coming war by thinking that it will be a repetition of the previous one

     

    I had already made this point in comments before the quoted post appeared.

    This contagiousness, comparable to that of rubella or mumps before vaccination, implies that this virus can only spread like wildfire in an immunologically naive population. And a virus capable of being transmitted by aerosol can only explain this contagiousness; it is a property of practically all respiratory viruses, SARS and MERS being notable exceptions to the rule.
     
    In 2012 Dr Dr. Anthony S. Fauci of the National Institutes of Health in Bethesda, Maryland, stated that "MERS-CoV does not spread in a sustained person to person way at all". But he added there was a potential danger in that it is possible for the virus to mutate into a strain that does transmit from person to person. ("Fauci: New Virus Not Yet a 'threat to the world' (video)". Washington Times. 31 August 2012.)


    A bat coronavirus virus got into humans through a civet (SARS) and then a bat coronavirus virus got into humans through a camel (MERS). The ability of viruses to recombine is well known. What kind of applied scientists see two novel viruses appear without the aforementioned characteristics (contagious by aerosol, and before symptoms) so while ignoring the caveat of Fauci proceed to assume a third novel virus will be the same as the previous two, and through a proclamation lull the whole world into a false sense of security?

    Chinese applied scientists in their totalitarian state do of course. The announcement on 7th January that the Wuhan (COVID-19) disease was a coronavirus but it did not easily spread between people was an eight-legged essay. China produced a global disaster because of its authoritarian centralised decision making. Yet, that very system meant they were able to cope with what came out of Wuhan better than Western rivals. Not least because the Western experts naively took the the 7th January statement on trust, believing China was a place of free thinking and human solidarity.

  45. @Kratoklastes
    Good data sources -

    covidtracking.com is easily the best for US data, since it set up a half-decent data API.

    US Daily (incl test numbers)
    Latest US State level data

    For Non-Exceptional countries, worldometers has a simple interface, but no 'Tests' data and no API (you can get the data out of the charts with 3 limes of Javascript in the Console in Firefox or Chrome - see the bottom of the page)

    Worldometers country table

    South Korea country page

    For Straya, the direct data page is a shitshow - this amateur effort, which has hard-wired image-based charts (so no drill-down, no tooltips, no data extractability... typical .gov low-quality overpriced shit.

    The CDC's page is so shit and so outdated that it's not worth posting the link, and OurWorldInData is slipping terribly - it hasn't updated test counts for most countries since March 20th.


    Footnote: Getting data out of Highcharts and into CSV-ish format from Worldometer

    • Right-click on the page
    • Select 'Inspect Element' from the Context menu. This will open the Developer Tools window
    • Go to the 'Console' or 'Debugger' tab in the Developer Tools window
    • at the prompt (>> - at bottom of page, usually), type

    csv = 'date,cases\n';
    Highcharts.charts[0].series[1].data.forEach(function(d){ csv += d.category + ',' + d.y +'\n' });
    console.log(csv);

    This will produce a little CSV object that can be pasted into Excel. It's a kludge, but it'll give a quick and dirty way to get the data out (I have something similar scripted to scrape every page, but mine gets the result as JSON and saves it to an SQL table).

    charts[0].series[1] means it will go to the first [0] chart, and get the data for the second [1] series. Change these as you see fit.

    Thank you.

  46. anon[594] • Disclaimer says:
    @JohnnyWalker123
    What's your opinion of the effectiveness of hydroxychloroquine?

    Trump keeps touting this drug. The FDA just issued emergency approval to hydroxychloroquine for its usage as an anti-Corona drug.

    French microbiologist Didier Raoult has done two studies of the drug. Both studies had very promising results.

    This has already been used widely in hospitals and ER’s. The big change is that the FDA has blister packs ready to distribute widely.

    Prior to this, the limited supply was being hoarded by doctors prescribing large quantities to family and friends with refills.

    “We wanted to try to get out in front of that as early as we could,” said Nicki Chopski, executive
    director of the board in Idaho, where pharmacists began reporting a significant uptick in prescriptions for the medications last week. The prescriptions, she said, were being written by doctors for themselves and their family members, often in large quantities with refills.

    https://www.nytimes.com/2020/03/24/business/doctors-buying-coronavirus-drugs.html

    Everyone was already using it and the FDA was holding off on wider approval until they had adequate supplies.

  47. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    “the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.”

    I too thought that would distort the results, but as you can see, almost every country travels along a nearly identical path (the Japanese, as always, being the significant outlier), despite the many variations in testing, categorization, etc. So apparently, all of the local variations balance out.

    What the chart does not equalize, however, is national population. So larger countries will naturally travel higher up the scale before their curves collapse. The US should be embarrassed, I suppose, to be travelling higher up the scale than China did with triple our population. OTOH, there is some question about the accuracy of the Chinese numbers, and China is bouncing back somewhat in the last frames.

    Something the chart does not show is whether any given curve collapse is due to countermeasures or simply due to the virus playing itself out in that population, though at this stage, countermeasures seems the more likely explanation.

  48. What an annoying voice.

  49. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    You read my comment directly from my mind.

    Numbers from Korea seem to show a death rate of infected people south of 1%, about .6%, though they of course also missed some infected. Call it .5%.

    So showing an exponential death chart would not be as dramatic as a case chart. Assume we get to 70% of the population infected. That’s 200mm people in the USA. At a .5% death rate (and NY State is currently about 1% of known cases), that’s 1mm people. Shockingly high number, but much lower than Spanish Flu as a percentage, which we survived without the sorts of measures we undertook.

    In all this, I keep thinking of the examples from Tufte’s Visual display of Quantitative Information, especially his examples of lying, as in lying by omission. See:
    http://davidgiard.com/2011/05/11/DataVisualizationPart4Context.aspx

    • Replies: @Buzz Mohawk
    Thanks. That's a good example you linked to. Another confounding variable is the fraction of a given population that is elderly or has chronic conditions or both. It can change: As Corana-chan reduces their numbers, they make up a smaller and smaller portion of the population in that geographic area, thus reducing the low-hanging fruit for her to kill. Then fewer deaths, and the death curve flattens -- while the case curve does not necessarily.

    Extreme scenario to clarify: Corona-chan kills everybody in Localtown, USA over 85, or with a serious health problem, or both. No more weak members of the herd for her to pick on, and death rate plummets. The death curve flattens. Lots of younger, healthier people get sick, to varying degrees (maybe the case numbers even continue doubling) but those people live and go on with their lives.

    I'm not saying she's culling the herd, but she's culling the herd. Lots of herds, in lots of locations.

    The difference between the case curve and the death curve is potentially dramatic, significant and important.

  50. @anon
    @ Sean

    What were they supposed to have known by Jan 7? There was no data that early. They thought it was serious enough to shut down their economy.

    Here is their data as reported in JAMA as of Feb 24.

    https://jamanetwork.com/journals/jama/fullarticle/2762130

    Cases plotted by date of symptoms looks quite authentic.

    I would like to see US data plotted by symptom emergence date rather than diagnosed/tested date.

    https://www.unz.com/proberts/marc-wathelet-a-virologist-and-specialist-in-coronaviruses-and-respiratory-diseases-explains-why-asia-was-successful-in-containing-covid-19-and-why-the-west-is-not/
    They made two errors with absolutely catastrophic consequences in their management of this crisis. The first was to believe that this new coronavirus was transmitted in the same way as the two recently emerged coronaviruses, SARS and MERS. It is the classic mistake of generals to prepare for the coming war by thinking that it will be a repetition of the previous one

    I had already made this point in comments before the quoted post appeared.

    This contagiousness, comparable to that of rubella or mumps before vaccination, implies that this virus can only spread like wildfire in an immunologically naive population. And a virus capable of being transmitted by aerosol can only explain this contagiousness; it is a property of practically all respiratory viruses, SARS and MERS being notable exceptions to the rule.

    In 2012 Dr Dr. Anthony S. Fauci of the National Institutes of Health in Bethesda, Maryland, stated that “MERS-CoV does not spread in a sustained person to person way at all”. But he added there was a potential danger in that it is possible for the virus to mutate into a strain that does transmit from person to person. (“Fauci: New Virus Not Yet a ‘threat to the world’ (video)”. Washington Times. 31 August 2012.)

    A bat coronavirus virus got into humans through a civet (SARS) and then a bat coronavirus virus got into humans through a camel (MERS). The ability of viruses to recombine is well known. What kind of applied scientists see two novel viruses appear without the aforementioned characteristics (contagious by aerosol, and before symptoms) so while ignoring the caveat of Fauci proceed to assume a third novel virus will be the same as the previous two, and through a proclamation lull the whole world into a false sense of security?

    Chinese applied scientists in their totalitarian state do of course. The announcement on 7th January that the Wuhan (COVID-19) disease was a coronavirus but it did not easily spread between people was an eight-legged essay. China produced a global disaster because of its authoritarian centralised decision making. Yet, that very system meant they were able to cope with what came out of Wuhan better than Western rivals. Not least because the Western experts naively took the the 7th January statement on trust, believing China was a place of free thinking and human solidarity.

  51. @Buzz Mohawk
    As the video explains at one point, the number of cases counted is partly a function of the number of tests given, so it is not possible to know for sure how many uncounted cases exist. Also, the case number can increase at an increasing rate as testing becomes more widespread.

    So, why not list, plot, and otherwise make available for visualization the number of deaths instead? This too is subject to inaccuracies when death is counted as happening with coronavirus, not caused by the virus -- and huge numbers of patients have pre-existing conditions -- but we'll just have to live with that.

    Does anybody here know where one can find death numbers over time, listed or plotted by day and location? That would be more enlightening.

    And BTW, probably everyone reading and commenting here does know what exponential growth is, so let's get that straight once and for all. Our questions, debates and discussions are not a result of a lack of understanding thereof. Key questions rather surround the data themselves. This is why actual deaths are what matter and should be plotted. What is their growth rate, and what proportion of cases do they actually represent?

    Data linked below. In .csv format. It is the data used in the Johns Hopkins dashboard.

    Click on a file link, hit ctrl A, ctrl C, then past in a spreadsheet (suggest you “paste special” >> text to drop the internal links.

    Once in a spreadsheet it is pretty easy to sort. The format changes somewhat as they developed more detail, but the “Country/Region” column remains as does “Province” to sort by. US counties appear around 3/22 which greatly expands the data.

    You can also copy from CDC (also linked) by sweeping across the data below the graph of cumulative cases. CDC and Github track pretty closely most of the time, not always.

    https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_daily_reports

    https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fcoronavirus%2F2019-ncov%2Fcases-in-us.html

    • Thanks: Buzz Mohawk
    • Replies: @another fred
    Buzz, my apologies, today's file (yesterday's data) is a string separated by commas not a comma delimited file. Maybe they'll fix it.

    It's not that hard for Excel to pull apart, but still a pain in the ass.

  52. @TomSchmidt
    You read my comment directly from my mind.

    Numbers from Korea seem to show a death rate of infected people south of 1%, about .6%, though they of course also missed some infected. Call it .5%.

    So showing an exponential death chart would not be as dramatic as a case chart. Assume we get to 70% of the population infected. That's 200mm people in the USA. At a .5% death rate (and NY State is currently about 1% of known cases), that's 1mm people. Shockingly high number, but much lower than Spanish Flu as a percentage, which we survived without the sorts of measures we undertook.

    In all this, I keep thinking of the examples from Tufte's Visual display of Quantitative Information, especially his examples of lying, as in lying by omission. See:
    http://davidgiard.com/2011/05/11/DataVisualizationPart4Context.aspx

    Thanks. That’s a good example you linked to. Another confounding variable is the fraction of a given population that is elderly or has chronic conditions or both. It can change: As Corana-chan reduces their numbers, they make up a smaller and smaller portion of the population in that geographic area, thus reducing the low-hanging fruit for her to kill. Then fewer deaths, and the death curve flattens — while the case curve does not necessarily.

    Extreme scenario to clarify: Corona-chan kills everybody in Localtown, USA over 85, or with a serious health problem, or both. No more weak members of the herd for her to pick on, and death rate plummets. The death curve flattens. Lots of younger, healthier people get sick, to varying degrees (maybe the case numbers even continue doubling) but those people live and go on with their lives.

    I’m not saying she’s culling the herd, but she’s culling the herd. Lots of herds, in lots of locations.

    The difference between the case curve and the death curve is potentially dramatic, significant and important.

    • Replies: @TomSchmidt
    Another good point. Thanks.
    , @TomSchmidt
    The example I linked to was wrong! Sorry, the guy who posted it doesn't understand
    Tufte, who is arguing against the speeding enforcement being effective in that one year in CT. He thinks the speeding enforcement worked.

    See the full context here:
    https://www.anychart.com/blog/2011/07/29/advices-by-edward-tufte-importance-of-context-for-charts/
  53. @Buzz Mohawk
    Thanks. That's a good example you linked to. Another confounding variable is the fraction of a given population that is elderly or has chronic conditions or both. It can change: As Corana-chan reduces their numbers, they make up a smaller and smaller portion of the population in that geographic area, thus reducing the low-hanging fruit for her to kill. Then fewer deaths, and the death curve flattens -- while the case curve does not necessarily.

    Extreme scenario to clarify: Corona-chan kills everybody in Localtown, USA over 85, or with a serious health problem, or both. No more weak members of the herd for her to pick on, and death rate plummets. The death curve flattens. Lots of younger, healthier people get sick, to varying degrees (maybe the case numbers even continue doubling) but those people live and go on with their lives.

    I'm not saying she's culling the herd, but she's culling the herd. Lots of herds, in lots of locations.

    The difference between the case curve and the death curve is potentially dramatic, significant and important.

    Another good point. Thanks.

  54. @Ron Unz

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19...My hunch is that this is all overblown B.S.
     
    Well, here's an interesting news item for the (hopefully) dwindling ranks of the Coronavirus Hoaxers who plug their ears and shout "It's Just the Flu!!!"...

    Italy's medical association has now reported that 61 Italian doctors have already died of Coronavirus during the current outbreak:

    https://www.unz.com/runz/the-government-employee-who-may-have-saved-a-million-american-lives/#comment-3806303

    If that’s the typical mortality rate for Italian doctors every flu season, their medical schools must have a hard time recruiting applicants...

    According to this article, there are 336,000 medical doctors inItaly. I’m going to assume that the minimum age for a medical doctor is 30 years old.

    https://www.ncbi.nlm.nih.gov/pubmed/12050938

    Last year, the death rate in Italy was 10.7/1000. So I would expect medical doctors to be at least as bad as that rate, and probably worse, being older. That means about 3500 doctors should die in Italy every year. Call it 300 a month, no?

    So, 1/5th of the expected doctor cull this month was fromCorona? And this is outrageous… why?

    You see a problem with math affiliated and think you can run in and help. I think, Sir, that you are fundamentally honest, maybe too honest to think ill of the people who are lying, spinning, and obfuscating with bad data. But they’re a malicious ruling elite, in the country whose genius political adviser told the Prince: don’t keep your promises. No one else will, and no one expects it of you.

    Of course, that guy also famously wrote that people will more easily forgive the murder of their parents than being made poor. If he is right, the elites made a bad choice in tanking the economy In the hopes of saving people’s parents.

  55. @Kratoklastes
    The reasons for a change in the South Korean trajectory are interesting, but one key thing over the past 2 weeks is a 20% decline in tests per day.

    They've gone from testing ~11.6k people a day (the average daily for weeks ending March 2 to March 16), to testing 9k people a day on the week ending March 23 and 8165 people over the the last week (and fewer than 2,000 yesterday).

    Nothing will make 'new confirmed cases' decline faster than reducing the number of people you test: stop testing altogether and 'new confirmed cases' goes to zero the same day.

    However the much more important contributor, is that the early testing was far more selective than recent testing.

    As we all know, since mid-March testing in SKor has had the admirable goal of getting a genuine 'society wide' sample, as opposed to just testing sick people who turn up at hospitals or show symptoms which is what everyone else has been doing. Australia also allows really insistent hypochondriacs to get themselves tested, but not in numbers large enough to de-bias Australia's <2% positives-to-tests ratio, which has been fairly stable since well before any 'social distancing malarkey.

    .

    The big SKor numbers prior to early March were largely from the Shincheonji megachurch in Daegu, which is even more lunatic and End-of-Time-y than Yank Rapturetards. It also had a bunch of people in the congregation who spent the last weeks of 2019 frolicking[1] in a little town called "Wuhan".

    The first death was a member of the congregation, and the South Korean government marked its ~200,000 members as priorities for testing. That cohort contributed a very large proportion of the early 'confirmed cases'.

    By early March the SKors had done ~147,000 tests, and South Korea's total positive tests stood at about 5,000. The Shincheonji 'supercluster' had contributed 60% of all positive tests at that stage, and a whole fuckload of them refused to be tested.

    There have been another 210,000 tests since, and the positives-to-test ratio has dropped back down towards what looks like a constant.

    Obviously if your primary testing cohort focuses on plague-infected retarded jackoffs, you run out of them eventually.

    By March 12/13 the SKors were pretty much only testing non-retard jackoffs.

    Ça explique tout (à peu près).

    One thing to note is that over the last 2 weeks, the drop in the total number of tests means that the denominator of the positives-to-tests ratio has fallen by ~20% but the number of positives has stayed (about) the same... in other words the positives-to-tests ratio is actually slightly higher (but it's statistically unchanged).

    https://www.dropbox.com/s/o2f7g3b6e96vywk/SKor%20Daily%20Cases.png?dl=1

    So to the extent that SKor's policies did anything, they made risks for non-retards very slightly worse (but not in a way that is statistically meaningful).

    .

    Bear in mind, too, that more than half of SKor's 9500-ish positive tests have since been released - and never exceeded 7500 'currently infected' at any time, and 99% of all symptomatic-infected have been 'mild'.

    SKor has only had 162 deaths - 1.7% of all detected positives (and so strictly less than 1.7% of all positives, and strictly much less than 1.7% of the population).

    Almost all the deaths were in old, infirm people with multiple existing ailments, and there might be a touch of that "with covid19, not of covid19" in the numbers.

    [1] I am assuming that retarded Armageddon-tards frolic - I'm not a theologian.

    You got a blog where you’ve assembled all these comments together? You and Westhunt ought to debate this. Masks on, please.

  56. @another fred
    Data linked below. In .csv format. It is the data used in the Johns Hopkins dashboard.

    Click on a file link, hit ctrl A, ctrl C, then past in a spreadsheet (suggest you "paste special" >> text to drop the internal links.

    Once in a spreadsheet it is pretty easy to sort. The format changes somewhat as they developed more detail, but the "Country/Region" column remains as does "Province" to sort by. US counties appear around 3/22 which greatly expands the data.

    You can also copy from CDC (also linked) by sweeping across the data below the graph of cumulative cases. CDC and Github track pretty closely most of the time, not always.

    https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_daily_reports

    https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/cases-in-us.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fcoronavirus%2F2019-ncov%2Fcases-in-us.html

    Buzz, my apologies, today’s file (yesterday’s data) is a string separated by commas not a comma delimited file. Maybe they’ll fix it.

    It’s not that hard for Excel to pull apart, but still a pain in the ass.

  57. @Ron Unz

    If 30K Americans die annually of the flu, we’d expect Santa Clara county to have ~150 deaths from the flu each year. We’re at 25 from COVID19...My hunch is that this is all overblown B.S.
     
    Well, here's an interesting news item for the (hopefully) dwindling ranks of the Coronavirus Hoaxers who plug their ears and shout "It's Just the Flu!!!"...

    Italy's medical association has now reported that 61 Italian doctors have already died of Coronavirus during the current outbreak:

    https://www.unz.com/runz/the-government-employee-who-may-have-saved-a-million-american-lives/#comment-3806303

    If that’s the typical mortality rate for Italian doctors every flu season, their medical schools must have a hard time recruiting applicants...

    If COVID were as deadly as the flu, I’d be worried. But it’s not. That’s not me saying it, it’s the Italian government. But what would those guys know?

    Go through and check the graphs yourself:
    https://www.zerohedge.com/geopolitical/whats-behind-italys-outrageous-10-mortality-rate-covid-19

    Italy is just now catching up to the flu-caused over-65 death rates of the previous two winters. COVID still hasn’t made things worse than three years ago, when the country wasn’t locked down at all.

  58. @Buzz Mohawk
    Thanks. That's a good example you linked to. Another confounding variable is the fraction of a given population that is elderly or has chronic conditions or both. It can change: As Corana-chan reduces their numbers, they make up a smaller and smaller portion of the population in that geographic area, thus reducing the low-hanging fruit for her to kill. Then fewer deaths, and the death curve flattens -- while the case curve does not necessarily.

    Extreme scenario to clarify: Corona-chan kills everybody in Localtown, USA over 85, or with a serious health problem, or both. No more weak members of the herd for her to pick on, and death rate plummets. The death curve flattens. Lots of younger, healthier people get sick, to varying degrees (maybe the case numbers even continue doubling) but those people live and go on with their lives.

    I'm not saying she's culling the herd, but she's culling the herd. Lots of herds, in lots of locations.

    The difference between the case curve and the death curve is potentially dramatic, significant and important.

    The example I linked to was wrong! Sorry, the guy who posted it doesn’t understand
    Tufte, who is arguing against the speeding enforcement being effective in that one year in CT. He thinks the speeding enforcement worked.

    See the full context here:
    https://www.anychart.com/blog/2011/07/29/advices-by-edward-tufte-importance-of-context-for-charts/

    • Replies: @Buzz Mohawk
    I see. And Tufte's point about context, "compared to what?" relates well to your point in your comment #60 and link to Italy's data and context when the previous flu is included. I'm not taking sides in your debate with Ron, but if Corona-chan this year is mopping up vulnerable people left by a soft flu season the previous year, well then you have something.

    You have something that even goes right to my point about death curves being partly a function of the fraction: easy targets / total population, and the shrinking of that fraction over time as the numerator is reduced. Placed in a longer, multi-year context, we might even say the mortality rate of flu-like illnesses can vary year-to-year as the vulnerable fraction of the population varies -- as you suggest for Italy.

    Hmm...

    I see everyone forecasting deaths based on constant mortality rates, and now that appears clearly wrong, no matter what rates they choose. Persons likely to die shrink in numbers as the virus spreads, so the death rate has to fall. Every prediction is therefore too high for the scenario chosen and modeled.

    , @Buzz Mohawk
    Guess what, I'm wrong too! Reconsidering death rates and shrinking numbers of vulnerable people: as long as the easy targets are evenly distributed throughout the population, they might make up a constant fraction of the total not yet infected, even as they die off.

    Such things are often not evenly distributed at small scales, but I think my idea of the shrinking mortality rate is wrong. As the virus moves through a population at a large scale, it should encounter a fairly constant proportion that is vulnerable; their number gets smaller, but so does the number of non-vulnerable who are not yet infected, probably at the same rate. Now I wish I hadn't gone down this road; it's embarrassing.

    https://attsucksbigtime.files.wordpress.com/2011/10/186781-817homer_simpson_doh.jpg
  59. @TomSchmidt
    The example I linked to was wrong! Sorry, the guy who posted it doesn't understand
    Tufte, who is arguing against the speeding enforcement being effective in that one year in CT. He thinks the speeding enforcement worked.

    See the full context here:
    https://www.anychart.com/blog/2011/07/29/advices-by-edward-tufte-importance-of-context-for-charts/

    I see. And Tufte’s point about context, “compared to what?” relates well to your point in your comment #60 and link to Italy’s data and context when the previous flu is included. I’m not taking sides in your debate with Ron, but if Corona-chan this year is mopping up vulnerable people left by a soft flu season the previous year, well then you have something.

    You have something that even goes right to my point about death curves being partly a function of the fraction: easy targets / total population, and the shrinking of that fraction over time as the numerator is reduced. Placed in a longer, multi-year context, we might even say the mortality rate of flu-like illnesses can vary year-to-year as the vulnerable fraction of the population varies — as you suggest for Italy.

    Hmm…

    I see everyone forecasting deaths based on constant mortality rates, and now that appears clearly wrong, no matter what rates they choose. Persons likely to die shrink in numbers as the virus spreads, so the death rate has to fall. Every prediction is therefore too high for the scenario chosen and modeled.

  60. @TomSchmidt
    The example I linked to was wrong! Sorry, the guy who posted it doesn't understand
    Tufte, who is arguing against the speeding enforcement being effective in that one year in CT. He thinks the speeding enforcement worked.

    See the full context here:
    https://www.anychart.com/blog/2011/07/29/advices-by-edward-tufte-importance-of-context-for-charts/

    Guess what, I’m wrong too! Reconsidering death rates and shrinking numbers of vulnerable people: as long as the easy targets are evenly distributed throughout the population, they might make up a constant fraction of the total not yet infected, even as they die off.

    Such things are often not evenly distributed at small scales, but I think my idea of the shrinking mortality rate is wrong. As the virus moves through a population at a large scale, it should encounter a fairly constant proportion that is vulnerable; their number gets smaller, but so does the number of non-vulnerable who are not yet infected, probably at the same rate. Now I wish I hadn’t gone down this road; it’s embarrassing.

    • Replies: @TomSchmidt
    I think your original idea is correct. It seemed at first that older people were vulnerable, but it's really older, sick people. That sickness is a result of gradual system collapse which takes time (and the low vitamin D levels of winter.). So the sinusoid always nature of the Italian morbidity curve follows your idea. If COVID were really virulent, it would kill some people BEFORE their system collapses had "ripened," to be morbid about it.

    We aren't seeing that in the data,yet. My guess is your insight is correct, and we won't.
  61. @Buzz Mohawk
    Guess what, I'm wrong too! Reconsidering death rates and shrinking numbers of vulnerable people: as long as the easy targets are evenly distributed throughout the population, they might make up a constant fraction of the total not yet infected, even as they die off.

    Such things are often not evenly distributed at small scales, but I think my idea of the shrinking mortality rate is wrong. As the virus moves through a population at a large scale, it should encounter a fairly constant proportion that is vulnerable; their number gets smaller, but so does the number of non-vulnerable who are not yet infected, probably at the same rate. Now I wish I hadn't gone down this road; it's embarrassing.

    https://attsucksbigtime.files.wordpress.com/2011/10/186781-817homer_simpson_doh.jpg

    I think your original idea is correct. It seemed at first that older people were vulnerable, but it’s really older, sick people. That sickness is a result of gradual system collapse which takes time (and the low vitamin D levels of winter.). So the sinusoid always nature of the Italian morbidity curve follows your idea. If COVID were really virulent, it would kill some people BEFORE their system collapses had “ripened,” to be morbid about it.

    We aren’t seeing that in the data,yet. My guess is your insight is correct, and we won’t.

    • Replies: @Buzz Mohawk
    I guess we'll find out. Thanks for the interesting conversation.
  62. @TomSchmidt
    I think your original idea is correct. It seemed at first that older people were vulnerable, but it's really older, sick people. That sickness is a result of gradual system collapse which takes time (and the low vitamin D levels of winter.). So the sinusoid always nature of the Italian morbidity curve follows your idea. If COVID were really virulent, it would kill some people BEFORE their system collapses had "ripened," to be morbid about it.

    We aren't seeing that in the data,yet. My guess is your insight is correct, and we won't.

    I guess we’ll find out. Thanks for the interesting conversation.

  63. If the recent graph of unemployment numbers are true, then no, we are not winning. People loved bandying around “exponential” curves. Unemployment data took such a bend straight up that it looks more like an outlier got plotted by accident versus anything real. Unfortunately it’s reality.

Comments are closed.

Subscribe to All Steve Sailer Comments via RSS
PastClassics
The unspoken statistical reality of urban crime over the last quarter century.
Which superpower is more threatened by its “extractive elites”?
How a Young Syndicate Lawyer from Chicago Earned a Fortune Looting the Property of the Japanese-Americans, then Lived...
Becker update V1.3.2