Coronavirus - tracking the numbers.11505
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
On facebook I've started posting updates on the numbers, and others have found it useful ... how many infected, how many have died, etc. A lot of people - including myself - don't fully appreciate exponential numbers until they are explicitly laid out. The current theory is, both the number of infected and the number dying from Covid-19 are increasing exponentially. The theory on infections I've seen is a doubling every three days. On the mortality rate I haven't seen a hypothesis laid out, though a few have mentioned it may also be doubling every three days. If you do the math, it gets very stark, very fast. Doubling every three days means 15 days = 5 doublings. That is, 2^5 = 32 times the starting number. However, if you extend that out to 30 days that is 10 doublings. That is 2^10 = 1024. That right ... after 15 days you have 32 times more; but after 30 days you have over a thousand times more. I'll lay out the projections here and in subsequent posts, and we can track the daily numbers to see if those projections are high, low, or just about right. So far, they seem to be low. However, testing has accelerated, as more test kits have become available; that is skewing the numbers, though we don't know by how much. I started this Tuesday night, and will update each night between 9:00 pm and 10:00 pm. To recap, on Tuesday I put down what the numbers were then (between 9:00 pm and 9:30 pm), and updated them Wednesday, and again Thursday and Friday. The prediction was: ~13,000 by Friday night, and 100,000 by April 1st. That would be 3,200,000 by April 16th. Well, we blew by those numbers a day early. Tuesday (~9:00 pm): Johns Hopkins University: 6362 Worldometer: 6468 Wednesday (9.30 pm): Johns Hopkins University: 7786 Worldometer: 9301 Thursday (9:45): Johns Hopkins University: 13,678 Worldometer: 13,847 And Friday (9:30): Johns Hopkins University: 19,285 Worldometer: 19,573 I'll stick with the "doubling every three days" hypothesis, but let's reset the baseline with Friday's numbers. Call it 20,000. That means by Monday night it should be 40,000. And the night of April 1st, instead of 100,000, the prediction is now 320,000. Ug. I hope that is too high, but I'm afraid it isn't. BEND THE CURVE, PEOPLE. Also, once I have some more data I'll try to figure out how to do curve fitting in a spreadsheet. I know there is a way - I've done it before, years ago - and I do have some time on my hands, so maybe we can check the actual numbers and see what we have. Here are the numbers we have on mortality: Thursday as of 9:43 pm: 200 dead in the US (Johns Hopkins) As of 8:15 pm: 209 dead in the US (Worldometer) Friday as of 9:13 pm: 249 dead in the US (Johns Hopkins) As of 9:25 pm: 262 dead in the US (Worldometer) Taking 200 as the start number, that would be 400 by Sunday night, or 3200 by March 31st, assuming a doubling every three days. |
||
Post 1 IP flag post |
Collector | Korptopia private msg quote post Address this user | |
There's unknown bias the world over due to inconsistent testing, but we have to work with what we have. I've been following, among others: https://twitter.com/jburnmurdoch |
||
Post 2 IP flag post |
I bought a meat grinder on amazon for $60 and it's changed my life. | kaptainmyke private msg quote post Address this user | |
8 billion people on the planet. That's over 800,000,000 lives that will be dead if 10% die and 5.6 billion people will survive at a 70% recovery rate of all confirmed cases. There are 327 million people in the US. 32 million lives lost. 229 million recoveries. Right? (assuming there is no vaccine or cure, ever) |
||
Post 3 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
Saturday night update. To recap, on Tuesday I pointed out that preliminary evidence shows the number of infected in the US doubles every three days. I put down what the numbers were then (between 9:00 pm and 9:30 pm), and updated them Wednesday through Friday. The prediction was: ~13,000 by Friday night, and 100,000 by April 1st. Well, we blew by those numbers a day early. Tuesday (~9:00 pm): Johns Hopkins University: 6362 Worldometer: 6468 Wednesday (9.30 pm): Johns Hopkins University: 7786 Worldometer: 9301 Thursday (9:45): Johns Hopkins University: 13,678 Worldometer: 13,847 Friday (9:30): Johns Hopkins University: 19,285 Worldometer: 19,573 And today (9:30): Johns Hopkins University: 25,493 Worldometer: 26,685 I'll stick with the "doubling every three days" hypothesis, but let's reset the baseline with Friday's numbers. Call it 20,000. That means by Monday night it should be 40,000. And the night of April 1st, instead of 100,000, the prediction is now 320,000. After one day, we are on track to hit those numbers. Ug. I hope that is too high, but I'm afraid it isn't. BEND THE CURVE, PEOPLE. I'm also tracking mortality, which I started on Thursday. Here are those numbers. Thursday as of 9:43 pm: 200 dead in the US (Johns Hopkins) As of 8:15 pm: 209 dead in the US (Worldometer) Friday as of 9:13 pm: 249 dead in the US (Johns Hopkins) As of 9:25 pm: 262 dead in the US (Worldometer) Saturday as of 9:13 pm: 307 dead in the US (Johns Hopkins) As of 9:25 pm: 340 dead in the US (Worldometer) Taking 200 as the start number, that would be 400 by Sunday night, or 3200 by March 31st, assuming a doubling every three days. |
||
Post 4 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
Quote:Originally Posted by kaptainmyke I don't think it will be that high, but I'm optimistic. I think at the end of the year it will have been 1% of the sick. Let's call it 60% infected by the end of the year ... that's 200 million people. So 1% of that would be 2 million people. But what strikes me now is, that's the optimistic scenario. Ug. |
||
Post 5 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
Quote:Originally Posted by Korptopia Oh, man ... that's an exponential graph, too, not linear. |
||
Post 6 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
I admit that it is all rather frightening to consider what will result once this gets widely spread in say Africa, or some of the island countries that lack more modern and well developed medical staff and facilities...the figures will see some adjustment then | ||
Post 7 IP flag post |
Collector* | Towmater private msg quote post Address this user | |
The testing data coming in the USA is showing that the mortality rate has been declining since about March 8. Mortality rate is number of reported US infections/US fatalities. We are now at 1.25%. March 21 24218/302 = 0.01247 March 20 19393/256 = 0.0132 March 19 13816/207 = 0.01498 March 18 9261/150 = 0.0162 March 17 6468/109 = 0.01685 I could go on but the decline has been happening since March 3, 2020. The rate that day was 0.07258. I hope the rate continues to decline and we get an accurate mortality rate for this bug. |
||
Post 8 IP flag post |
Collector | Drogio private msg quote post Address this user | |
Quote:Originally Posted by Towmater The mortality rate is not declining, they are just testing more people as tests become available. The fate of those most recently tested is unknown and “tested” stats are skewing the actual mortality rate and that turn is giving many false sense of security, which leads to many shirking off CDC recommendations and guidelines. It’s better to look at China where infections are trending downward. You’ll see he mortality rate is closer to 5%. Italy is 33% as many are not being treated because of the percentage of elderly population and the health care system is overwhelmed...if we head down that path where people ignore the shelter at home and good hygiene then we’ll have rates of 10% or higher as well. |
||
Post 9 IP flag post |
Collector | Drogio private msg quote post Address this user | |
China - 81397 cases, 72362 recovered, 3265 dead. That’s 4.3%. if the additional 5770 survive the rate drops to 3.8%. Italy - 59138 cases, 7024 recovered, 5476 dead. That’s a 44% death to recovery. Even If 95% of the additional 46638 cases (59138 - 7024 - 5476) survive it’ll still be a 13% death to recovery rate. That’s pretty scary. Btw, none of the 24218 have been confirmed as “recovered” in the US...so that’s a bit of a lag as I’m sure it takes a month or more to run it’s course given this has been going on for a month now. So basically nobody has made a full recovery yet and there are only deaths. |
||
Post 10 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
Yes, there are different ways to report the numbers, and each has deficiencies. @Towmater is right about what is reported (and right in hoping the number continues to go down!), while @Drogio correctly points out that increased testing has increased the number of infections. And, it is increasing so rapidly, that most of those cases have not yet be "resolved." One website - Worldometers - updates their data regularly throughout the day. I don't know how reliable it is, but it so far seems solid. They report, of the 32,356 total positive cases in the US, 592 are "closed" ... 30% of those recovered, while 70% (414) died. Clearly, that percentage is too high, indicating that a lot of people were never tested and confirmed to have it, likely because their symptoms were so mild. They also report, of the 31,764 "active" cases, 3% are "serious or critical." That's a better number, and likely more realistic. I still believe, based on multiple sources and my own opinion, that the mortality rate will end up being about 1%, give or take. That's 10 times worse than the normal flu, but not so awful. Of course, if we run out of hospital beds, ICU beds, and ventilators, like they did in Washington state, and like is on the verge of happening now in NYC, then that percentage will be higher, unfortunately. But as @Towmater pointed out, we still need more and better data. |
||
Post 11 IP flag post |
Masculinity takes a holiday. | EbayMafia private msg quote post Address this user | |
We won't know a true death rate until we are in a position to extrapolate...and reasonably estimate the number that never sought care or testing. But death rate is going to be a function of who gets it and the level of care they receive. Basically you can divide it into four quadrants: 1) High risk x low care= High death rate 2) High risk x High care= Less High 3) Low risk x Low Care= Lower death rate 4: Low risk x High Care= Lowest death rate My unscientific expectation for the U.S. and most countries, as testing increases is a "recorded" death rate of 0.8%-1.5% and an extrapolated or model-based estimated rate of 0.4% to 0.6%. |
||
Post 12 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
Even if you had recourse to estimate the number that never sought care, you lack the ability to know how many were deaths that were also never recorded as such ….I have seen a lot of evidence to suggest that while China was trying to hide the information about the virus they were often coding the deaths as secondary causes, rather than listing them as the covid 19 virus. In a country like Iran where the media is state run, I would wager, although I couldn't prove, that deaths and numbers infected are being either falsely inflated to enrage people towards the US , or artificially lowered to hide the truth. I think its entirely reasonable to assume the rate for china at above 3 percent is likely lower than the factual number, If you assume in Italy the country is in full lockdown, I am guessing there are any number of deaths in isolated households that have not been discovered yet as well. This thing is certainy frightening and is redefining so much about our world and way of life right now...be safe all of you and our families. |
||
Post 13 IP flag post |
Collector | Drogio private msg quote post Address this user | |
Russians Hack Johns Hopkins Site International Business Times 13 March 2020 The constantly updated data about the death toll inflicted worldwide by COVID-19 placed online by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University is an invaluable tool in saving lives. Unfortunately, some people see this humanitarian endeavor as a means to hack into the computers of those using the CSSE website and make dirty money in the process. Russian cybercriminals are selling to other cyber crooks a digital coronavirus infection kit that uses the CSSE interactive map as part of a Java-based malware deployment scheme, warns Brian Krebs, the American journalist and investigative reporter best known for his coverage of profit-seeking cybercriminals. This malware interactive map, which looks like the real thing, is sent as an attachment to a victim. Clicking on the interactive map (which is in reality a malware PreLoader), will either deploy password-stealing malware or send itself to the victim's email list. Krebs also believes the CSSE interactive map is also being used as attachments to spam e-mails. He said a member of several Russian language cybercrime forums in late February began selling a digital coronavirus infection kit that uses the CSSE interactive map as part of a Java-based malware deployment scheme. According to Krebs, the kit loads a fully working online map of coronavirus infected areas and other data. The malware map is resizable, interactive, and has real time data from World Health Organization and other sources. "Users will think that PreLoader is actually a map, so they will open it and will spread it to their friends and it goes viral!” the seller of the malware says in a video explaining the exploit. What makes this Java-based map more dangerous is that it can apparently be bundled into a filename most webmail providers allow in sent messages. The source said the user/victim, however, has to have Java installed for the map and exploit to work. He also makes the astounding claim the map will work even on fully patched versions of Java. “Loader loads .jar files which has real working interactive Coronavirus real-time data map and a payload (can be a separate loader),” said the seller. “Loader can predownload only map and payload will be loaded after the map is launched to show map faster to users. Or vice versa payload can be predownloaded and launched first.” TOP This new malware kit seems to be drawing a lot of interest among cybercriminals given the dominating presence of news about the COVID-19 pandemic on the internet these days. Earlier this week, cybersecurity experts began warning of new malicious websites using interactive versions of the same map to distract visitors while the sites tried to infiltrate the password-stealing AZORult malware. Online scams and malware linked to COVID-19 has become so prevalent the U.S. Department of Homeland Security (DHS) is warning individuals to remain vigilant. The DHS' Cybersecurity and Infrastructure Security Agency (CISA) warns that cybercriminals are sending emails with malicious attachments or links to fraudulent websites to trick victims into revealing sensitive information or donating to fraudulent charities or causes. It said people must "exercise caution in handling any email with a COVID-19-related subject line, attachment, or hyperlink, and be wary of social media pleas, texts, or calls related to COVID-19." CISA also warns people not to reveal personal or financial information in email, and not to respond to email solicitations for this information. |
||
Post 14 IP flag post |
Collector | poka private msg quote post Address this user | |
I am going to make the argument that death rate in average will be about 1% among those contracting the virus. The basis is that there a very large population who have/will contract the virus, but have not/will never be tested or required to be hospitalized whereas the death numbers will be about the same. |
||
Post 15 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
Illogical statement...the death rate, whatever it is, will remain constant. If its 3.0 that's where it will stay...trying to factor in those that will never be tested/non hospitalized but not allowing for those that will die and likewise not be tested, autopsied nor associated with the virus is faulty. Most people that die from this virus do not succumb to the virus itself, but rather secondary causes complicated by the virus. You have to assume in many older people, that die from the disease while not being diagnosed, they would offset the people who are infected but never discovered. Understanding that idea, it is highly likely there were deaths already within the US due to the virus that went undiagnosed, just as there was already massive community infection widely occurring prior to the discovery. A true measure of this would be to understand how few tests are being done for the living that may/may not be exposed....imagine the lack of testing done for those passing in what would appear innocent circumstances |
||
Post 16 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
I keep studying that world map and realizing it says 350,000 + infected, most of which are recently exposed. Comparing that with the recovered and deceased column, it is obvious very soon we are going to see some grim figures for deaths, even at 1 percent many are stating as their given estimate for the loss of life..whatever the percentage, its going to be devastating and shocking | ||
Post 17 IP flag post |
If the viagra is working you should be well over a 9.8. | xkonk private msg quote post Address this user | |
I think I posted this link in the other thread. We can see how the predictions are looking on the 29th, since they made a concrete prediction for that date. https://fivethirtyeight.com/features/infectious-disease-experts-dont-know-how-bad-the-coronavirus-is-going-to-get-either/ | ||
Post 18 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
the worst case numbers are rather grim for sure, lets hope it doesn't come to that | ||
Post 19 IP flag post |
Collector | poka private msg quote post Address this user | |
Quote:Originally Posted by Darkseid_of_town we definitely disagree with the logic |
||
Post 20 IP flag post |
Collector | poka private msg quote post Address this user | |
Quote:Originally Posted by Darkseid_of_town and the real number is a multiple of this because most people who have contracted the virus will never be tested. there are currently identified 3 different kind of the virus. the largest virus group is the one where people do not have any symptoms but are carriers |
||
Post 21 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
Can you provide or link anything that supports three different kinds of virus please? I have only seen documentation for the two known so far...interested in anything related to a possible secondary mutation of the virus.... Generally, the people who are infected by a virus but display no signs are a natural part of any strains normal infection range rather than a separate strain...so definitely interested in anything that can document or shed further light. |
||
Post 22 IP flag post |
Collector | Darkseid_of_town private msg quote post Address this user | |
Quote:Originally Posted by pokaCorrect, using only the exposed unknowns but refusing to allow for the deceased unknowns would of course intentionally skew the figures, so yes I do disagree with the supposed logic given. |
||
Post 23 IP flag post |
Collector | etapi65 private msg quote post Address this user | |
But you're misunderstanding Poka's point. Likely, everyone with symptoms severe enough to die will be tested now. Even though early data would have been skewed. I imagine a LOT of people everywhere (not just China) died from this without it being identified. THat is very unlikely to continue to be the case. However, those tested will continue to skew further since...and I'm not sure I have this 100% accurate...20% have no symptoms, 60% have mild barely register as symptoms, symptoms and 20% will need medical intervention. Of those x% will die (somewhere between 3-4 at present seems likely). The 20% asymptomatic group will likely NEVER be tested. The 60% group will have greatly diminished to no testing. The 20% medical intervention will 100% be tested as will the 3-4% death (from this point...or probably last week, on). Now I wouldn't put a % on this like poka's 1%, but you're logic doesn't follow what's actually happening and being implemented at the present. In essence, by the very nature of how it's being monitored and tracked the death rate will skew higher than "actual", moreso than total infections. | ||
Post 24 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
Quote:Originally Posted by etapi65 Making it even more difficult, a friend of mine - a nurse - reported to me that in some places they are not testing really sick people "Unless a positive test would alter the prescribed treatment." That is to save on test kits, yes, but mainly to save on PPE and other equipment/time needed to take the swab and get the results. So unfortunately, there is some anecdotal evidence that not all who are sick will get tested, even if REALLY sick. There is also a report I read - not sure where - that early cases in Italy were 'broadly' attributed to Covid-19, even if they died for other reasons. Likely also due to lack of testing ability, but who knows? The article said it was not continuing, and it was more consistent now, just earlier cases/numbers affected. |
||
Post 25 IP flag post |
Collector | poka private msg quote post Address this user | |
@Tedsaid @etapi65 you are saying it better than i can | ||
Post 26 IP flag post |
Collector | poka private msg quote post Address this user | |
Quote:Originally Posted by Darkseid_of_town https://www.scmp.com/news/china/society/article/3076323/third-coronavirus-cases-may-be-silent-carriers-classified |
||
Post 27 IP flag post |
Rock, Paper, Scissors, Lizard, Spock | Tedsaid private msg quote post Address this user | |
The 538 page is interesting. One expert is quoted: "Andrew Lover, an epidemiologist from the University of Massachusetts who took the survey, said his estimates were 'semi-quantitative' and based on the virus’s progression in other countries. 'The doubling times have been 5-8 days most places.' " My projections and numbers are based on the hypothesis that, in the US, both infected and deceased are doubling every three days. Now, even I thought that was high, but so far the numbers are being borne out. And that could be because, concurrently, the amount of testing is accelerating in the US daily. Anyway, it appears the predicted numbers in the 538 article about March 29th are already too low. Currently, there are 41,708 cases in the US (from Johns Hopkins), and my "double every three days" numbers say there will be 160,000 by March 29th. |
||
Post 28 IP flag post |
Masculinity takes a holiday. | EbayMafia private msg quote post Address this user | |
Quote:Originally Posted by Tedsaid I think the doubling will come less often as the base number grows. Looks like a fairly consistent 7,000 to 8,000 per day in the US at the moment. I've heard the positive rate on tests is approx 10% which would indicate about 75,000 test currently being done daily. Of course they are only testing people who show symptoms or have at-risk situations but I'm guessing that 10% rate would go up if all testing were done in NY at the moment. |
||
Post 29 IP flag post |
Thread locked. No more posts permitted. Return home.