3Blue1Brown: "In most towns, people don't actually spend their days drunkenly wandering around the city." New Orleans: "Let me tell you about Mardi Gras."
Another way of saying it: "Math is a shorthand for a theory but does not prove the theory (as such)." In this case the parameter values are assumptions and the various results are useful only to the politicians, which, moreover, strip or ignore the feedback/validation.
Very much agreed, but more so for models involving humans, and less so for natural phenomena we've studied for a very, very long time (think N-body simulations). Making this assessment is important.
@eric I like that angle much better, for two reasons. First, mistaking a representation for the thing it represents is a widespread issue across many domains, not just mathematical modeling. Second, "all models are wrong" sounds confused to me - like you should come up with a better definition of "right" and "wrong" in the contexts of models. It's silly to say "all maps are wrong but some are useful." Many maps are "right" but the threshold for a map being "right" is not as high as reflecting literally everything about the area being mapped.
@@eleanorhuxley6959 Why was it wrong? Semicolons are used to separate complete sentences that are related. These are both complete sentences that are related in content, so I don't see why you're saying it's wrong.
I'm an Emergency Medicine Physician with a PhD background in modelling. I am absolutely blown away by your work combining phenomenally effective graphics with practical modelling and a patient methodical, logical argument. Even Edward Tufte would be jealous a what you've done here. Truly one of the best presentations on any topic I've ever seen in my entire life. You deserve a medal and Isaac Newton's chair.
I don't know how it panned out in Redlands, CA, but the UK screwed up by politicians overruling logicians, I'm afraid. Even now, mostly young male narcissists shop unmasked, as civil liberties don't allow questioning. A solution is they must visor up - we did a unique community visor acceptability study, and everyone can tolerate these alternatives - for the mask intolerant. Boom, as young folk say. Retired old UK medic here. Keep safe :)
@@tim40gabby25 scared people is the problem. Older people ruining it for the young by promoting this fear. I’m disgusted by what my generation have done to the future of my kids and grandchildren.
and absolutly unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory desease can travel over kilometers.
Didn't the US suffer from like, five distinct waves? We failed to learn our lesson four different times. Amazing. edit: and as I recall, they were like, the minimum viable definition of a wave, where we let up as soon as R dipped below 1, instead of actual cases dropping below some number.
Looks exactly like a metal lattice, with slight lines of dislocation - which would represent grain boundaries in metal. If you think about how bonded atoms are basically dots that attract each other but repel each other at closer distances so can never touch, it's pretty analagous to a bunch of repelling dots confined in a space together. This is why you get the same structures and properties in both. Funny how maths do that sometimes.
My high school chemistry teacher tried to explain entropy in a similar way. A real-world example he gave was people choosing seats in a movie theater that’s initially empty and gradually fills up. When the occupancy is low, people are unlikely to sit down right next to a stranger, etc.
11:30 The way that the social distancing makes the dots in the top left box arrange themselves in a periodic/crystalline way, even forming grain boundaries, is so satisfying to watch!
@@3blue1brown It appears that the infection rate is different within a grain vs at the boundary. I was looking at this and wondering whether any effect from this or the geometry of the box could have any physical interpretation, or whether it is an artifact of how this was simulated.
That's known as a Wigner crystal. First observed in experiments in which electrons were poured onto the surface of liquid helium. The crystal becomes unstable at high density or as the level of fluctuations increases. Nicely explained in the Wikipedia article.
this series of animations were so stunning. Never before had I seen such a great educational content. I am a general physician and I had to go through a lot to remind my patients and families to practice social distancing. I broke my back, trying to simulate the epidemic during the recent 3-weeks . and here you did it perfectly. We owe you all!
@@x_x5009 This is the most unrealistic part of Plague, Inc. unfortunately. Diseases mutate, yes, but they mutate independently, and IRL every mutation needs to (re-)infect people with the newer version in order to have the "new" effects. Whereas in Plague, Inc. all your previous infections mutate at once, which makes the game fun and easier, but also completely wrong.
@@amunak_ It would be interesting to make a scenario, in which you cannot evolve (or devolve) symptoms manually, just increasing and decreasing mutations and wait for it.
@@morroghaiky6580 or a scenario where every time you mutate it creates a new patient zero which has to infect everyone again, immune suppression could make it easier to infect pre-infected people.
You guys are forgetting biowarefare! I know that none of us can be certain of where this virus came from, and either way, we can't trust the CCP or anything they tell us or their people. They're already telling them that the virus came from America. So that aside, it's my personal belief that this is a virus that's been created in a lab (in 2013- theres records of it, and it won awards). The genetic technical capabilities of the higher ups is insane, if you dont believe in aliens and backwards engineering secretly happening, then this concept will be implausible to you. The details, proofs, and reasons behind this happening dont matter for now, but if you want to learn more, Edge of Wonder has tons of videos on this and related topics. Anyways. This virus either gets out by accident, or on purpose. The CCP is now taking advantage of the situation, and wants to use it to decimate everything, for it's own advantage. Since it's been created in a lab, I dont know what its capable of. I dont know if the virus itself is nanotechnology, and if those nanoparticles are spreadable from person to person. Maybe the virus is programmed to replicate unnatural human designed particles, and then is able to essentially have an on and off switch that can be turned remotely. Another one of my beliefs is that, due to chinese prophecy (I /think/ that it's in the book, tui bei tu, but you'll have to check what edge of wonder said in their video), basically, this virus will do what it's doing now, and then months later, just stop out of nowhere. No trace of it, and we all go back to normal. Then, 10 years from now, it will come back again, then disappear forever. The only way this can make sense is if the virus has an on off switch like what i described. Theres a lot more details to this that I would like to explain but people dont like reading long walls of text for some reason.
It's funny because the mortality rate is actually rising like crazy so yes you're right it's getting more and more lethal. It used to be about 3% and now it's already 4.5%
absolutely unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory disease can travel over kilometers.
@@igorpashev Respiratory diseases are not always airborne if thats what you mean. Also, I need sources on the last claim because ive only known airborne diseases to remain for mere hours. Airborne diseases also infect efficiently through moist droplets from one person, also made more effective within 6 feet. Wear masks, distance yourselves. Thx champ.
Creamy Pasta Camper van owners just spread the disease to different communities, though. It'd be like the quarantine box simulation, but instead of it being a quarantine box, you just send them all to the same community box.
This shouldn't be a PSA on "Social Distancing," this should be required viewing for Pandemics 101. Great introduction on the thought process and problem solving that can go into epidemiology. As usual, very approachable, 3B1B. Thanks again.
This is calm and reasonable and smart. A video like this is not only informative, but reassuring as well. Thanks for weaponizing your math brain to help people Grant.
And thank you for doing exactly the same thing as Grant does, but with good natured snarky commentary about the Bible and history and Christianity things. You sir, are a treasure, juat like 3 blue 1 brown.
and absolutely unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory disease can travel over kilometers.
There is an interesting tweak that can be done: Separate the "removed" in dead and recovered. Every infected has a probability of becoming dead or recovered. Recovered become Susceptible after a long time (Due to losing the immunity or the virus mutating)
I'd go even further and separate the recovered into immune and susceptible categories. Might also include non-human factors like animal reservoirs and mutations.
if you really want to be pedantic you could make the recovery/death ratio dependent on the current number of active cases (like if hospitals are overrun). Then you could put numbers to flattening the curve lowering the death toll even if the same total number of people got infected.
I know this comment is two years old but I think not distinguishing between recovered and dead was the right call. The probability of an infected person dying depends on a lot of factors that would make the video a lot more complicated, the general rule followed was that having a flatter curve reduces the number of deaths, which is simple and accurate enough for the pourposes of the video.
This is a really good video and I really like how you've gamed out many different scenarios and shown their outcomes. My main criticism, and it's a pretty big one, is that you've given your preferred control mechanism, isolation of infected individuals, a huge advantage by not varying the lag between infection and isolation. I don't think isolation one day after infection is realistic for nCoV2 even in a scenario of very widespread testing. I would have liked to see how quickly the benefit of isolation falls as that lag grows to 2, 5, or 7 days.
A well-structured, polite and respectful criticism that tries to constructively bring more options to the table of an important discussion. Surely a material to be explored in a similar video. Thank you
Exactly. In our country, no protocols were made to address the pandemic until many confirmed cases were already reported and thus no efforts were made for social quarantine. It took weeks until the government responded. There's also the concern of the cases being way higher than actually reported.
This is the problem that we have. It seems that covid-19 may be infections for a number of days before becoming symptomatic in the patient. Some estimates up to 14. Even if we can have proximity testing and contact tracing it would be insufficient. With any significant number of asymptomatic infections - whether or not they remain so - only full population testing *and* isolation of those testing positive could be effective in preventing further waves.
THIS! Also, I have the feeling that people are being "removed" way too quickly in his simulation, plus that the R0 for the real disease was around 3... It's a good video about infectious diseases in general, and about the effects of different measures against the spread of it. However, my fear is that people would underestimate the neccessity for social distancing/hygiene as they might think that isolating would have been the best anyway. Anyway, I still think this should be shared globally (I mean the video, not the disease), so that people really understand why the should keep their distance and keep their hygiene up!
2 years ago, in a completely different simulation, I had my circles escaping the box when I added a repulsive force to solve collisions! I laughed a lot here. Nevertheless, many thanks for this great video (as usual). Cheers from Italy.
Where and how are these simulations done? Can you please tell? Is is just python with some libraries or is there some other specific software for this purpose?
Here’s another idea to simulate: homes There’s small boxes outside of the main box, and each of those small boxes has 3 dots assigned to them. Each dot has a ‘role’ 1 dot has a ~25% chance to travel to a main area (grocery stores, restaurants etc.) on any day. Make three ‘hubs’ in the big box, with an equal chance for the ‘wonderers’ to travel to The second dot has a ~75% chance on any day to leave their house, but stay away from the ‘hubs’ and try to avoid other dots. The third dot has a ~10% chance to leave their house, but they stay right outside it.
Tactical Lemon Excellent concept. I might disagree with your specs. I'm prejudice towards attempting to model a family. At that, you may have a good beginning with three adults, perhaps a married couple with an aging parent. Now we need to add two or three children, each with various levels of activities of interaction with the adults and some potential for going off short distances with the intent to inter act with other children in somewhat similar situations and possibly those children's parents or guardians. I think I 'be lost focus on modeling a single family and have started a neighborhood. Oh well, in a more comprehensive and complex model, we'd want a central city and suburbs. NYC and SF are exceptional. Urban geographers may have something to add...
I was thinking that social distancing would probably be better modeled using many many small networks (families and tight social circles) as well. His "communities" simulation does seem to result in the most interesting and maybe useful results. Beyond the "test and isolate" result, which has been infectious disease dogma for about 400 years in the West (probably longer in China). Of course, we have much improved abilities to test and isolate now... wish we (the US) would have used them -_-
I wonder if you could take a satellite image and set up the algorithm automatically. Identify structures, and then estimate the number of people who spend time there by the surface area it takes up. In big cities this would have to take into account the height of the buildings somehow I guess.
@@m3po22 great idea but no need for satellite image, the phone network is enough, more accurate and you can track contaminated, that's exactly what they did in South Korea and China
Grant, thank you for this great presentation on simulating an epidemic. These visualizations are outstanding for understanding how the numbers really work.
I ran into this because I was looking at your Bayes' Theorem videos. However, it's now 2 years later. It's remarkable (or, as someone below put it, "ridiculous," how prescient this ended up being. It's so clear which paths we followed and how it ended up being just like you predicted.
Wasn't expecting to feel such a wide range of emotions watching some pi symbols and dots shuffling about the screen. And genuinely funny moments too (the dots leaving the Bay Area lols) Great video!
"even when all the parameters are the same, some runs take three times longer to reach this point than others" This deserves more than just a side note. With agents starting at random locations and then moving around randomly, you are likely to end up with a chaotic system. It cannot be stressed enough how crucial it is to have multiple test runs for each set of parameters. For some parameter sets, you might end up with almost everybody infected in one run and the infection quickly dying away in another run, just due to random fluctuations. Virus simulations (as well as economic simulations and weather forecasts for that matter) can be highly susceptible to random noise.
@@00O3O1B Yes, a nice followup would include a qualitative analysis of the official numbers and adjustments that can be added to the series of algorithms. Even though at some point all you can do is measure the overall statistics, once the fundamental parameters of the virus are accounted for then you can isolate effects of behavioral interventions such as limiting borders, etc. I for one am interested in the cumulative effect of panic shopping and hoarding on the broader population in exacerbating issues. Those who could afford to prepare would do in excess immediately when it would have an exaggerated effect on those with lower income and it took weeks for stores to enact any restrictions to buffer the impacts. Another major factor as this goes further will be housing instability and population movement, as that increases it will negate the effectiveness of travel restrictions.
Excellent point! This is why I started all the actions in this video once there was a sufficient number of cases. For each one I showed, I rand multiple to be sure they were qualitatively the same and tried to comment when they weren't. For proper modeling (as opposed to demonstration videos), you'd, of course, want to run many and find the right way to aggregate the results into your summaries.
@@3blue1brown So does this mean we really had to destroy the American economy to achieve this? Once the tests were available (today being maybe a few days to a week where proper testing has been a bit more consistent) could we have lifted the ban like, this week (03/28), now that we know we can test and isolate effectively? It really looked like once you have a grasp on testing and isolation, you could still live your life as long as you were being super hygienic. But can we count on EVERYONE being hygienic? Did you model only partial hygiene?
@@SeraphX2 I'm not sure the US is able to test and trace it's way out of this yet, but you're right that effective testing and isolation can halt the spread and prevent second waves (South Korea and Taiwan being excellent examples of this) with much less disruption. The disruption (and economic damage) will end when the disease does, and the fastest way to do that is to catch up with the disease as soon as possible. Shutdowns and distancing are important in getting to that point.
@@SeraphX2 the model would perhaps have to take into account hygiene/isolation compliance based on infection prevalence (fear) inorder to measure the effectiveness of public policy.
I love this channel... My university frequently refers to your videos as a good learning resource and I absolutely agree! (Also, I've known this channel for years, so I'm happy to see that my university approves it as a good source :) )
Sir, honestly: the simulations and analyses were so well done and have so much common sense that I tilt my hat to your work in gratitude! I am definitely becoming a patron!
I remember reading an article covering the spanish flu, where cities handled the outbreak differently. The city with strong early measures was accused of being alarmist and such, but many more people survived.
@@nomoretalk2967 That's in part because of the socio-economic factor : A lot of people are being (rightfully or not, I won't weigh in) angry with Macron, and whatever he does, they'll keep being angry. "Gilets jaunes", pensions reforms, all those people have a deep bias against him. Case in point, 2 weeks ago, on Thursday, he declared that schools and universities would close on Monday : He was called an alarmist. The next monday, he put confinement rules in place : He apparently did it "too late".
Brilliant job! Well done my friend. Very informative, nicely illustrated and calmly presented. Straight to the point. No loud music, No commercial. I just wish I could find more channels like this. BRAVO !!!
A "full city" model would be interesting... households with a few people each, workplaces, schools, markets, hospitals, retirement homes with younger workers, quarantine zones, X% asymptomatic cases... toilet paper hoarders...
Imagine all the world's intelligence agency doesn't have a package like that in their tool suite. This means governments are being purposefully dumb to become Thanos.
@@SouthernHerdsman I'm certain some intelligence agencies do have that capability and model it thoroughly before releasing these pathogens against opposing nations. What do you think we're experiencing right now? A mistake, or a cold, calculated attack? There will be be a long train of conflict following this. Whether it was an intentional attack, an accidental release, or a natural occurrence, this will lead to war.
"What if people avoid contact Wich others for a while, but then they kind of get tired and stop?" We don't have to simulate the answer to that question anymore, we have the real life data right here.
18:00 When the model is so realistic, it includes existing real life behavior that's clearly outside the initial assumptions that were strictly programmed in.
As a high school math teacher, I'm making this video a recommended watch by all my student (especially those in Algebra 2, who were just learning about exponential growth functions. THANK YOU!
I'd really like to see a large version of this simulation run with parameters close to what we're experiencing in real life. It'll be like looking into the future. For example: - 32 "communities", 4 of which will contain central hubs and 4 of which will contain a central market - 100% are asymptomatic for 3 days; 80% will show symptoms afterwards - 70% will be quarantined after 1 day of showing symptoms - The infection goes away after 14 days - The infection radius is the same as in this video - Recovered patients are immune to future infections - 0.005% of people can go into total isolation ("leave the box") per day - Travel rate between communities is normal (0.2) until a certain threshold is met (in which case, travel rates will go down by 75%) - Social Distancing rate is low (20%) until a certain threshold is met (in which case, social distancing goes up to 75%, then 95% if things get nasty) - Infection rate is high (0.2) until a certain threshold is met (in which case, infection rates are cut in half) - For the first time only, if the number of active cases falls below a certain threshold, travel rates, social distancing, and infection rates go back to normal - Certain Threshold: 0.5% of the population is infected NOTE: The virus's actual asymptomatic period can go up to 14 days, and according to zemerick13, the infection can last a lot longer than 14 days.
3Blue1Brown is far too smart to have not run that simulation...so I worry that its ("its" being Craftie!'s simulation idea, which I think is brilliant) lack of inclusion might be a sign that the results aren't....uh...valuable....at the moment.....no now that I think about it 3Blue1Brown is likely far too ethical to run that simulation and not include it in the video regardless of what it might indicate. They should!
- Would be interesting to model three stages of contagious 1 - asymptomatic (normal mobility) 2 - symptomatic but home (greater isolation) 3 - hospitilization (isolated)
I think the escaped dots represent the people who saw the direction the world was going and preemptively committed suicide. I'm in college and it happened to two of my friends, one of whom I was pretty close with.
Thank you for putting this out. The effectiveness of how you explained this is the best I have seen, in fact, I had to share it because now is the time to understand the impact of your observations.
im curious how having multiple central locations like a school and a few grocery stores will change this assuming that every dot picks 1 of the 3 to visit also it would be interesting to see how things would change if dots try to stay in clumps similar to a family staying at home
I think it dosn't matter if one person out of the family does the shopping, as he earlier talked about people in close proximity are at risk. Eventually the whole family will get it.
QalinaCom the families are also just a similar case to the 12 different communities. Just apply many communities of size 1-7 within each community and you’ve effectively simulated a family
I built a simulation that takes into account school & work (along with household life): n8python.github.io/epidemicSimulation/ I explain it here: ruclips.net/video/Yo1rL0hqHAQ/видео.html
The Covid-19 is a HOAX. Corona-Virus has never been isolated and checked with the 3 Postulates of Robert Koch. Pasteur has treated the whole lot/world ! ALL HAS BEEN PROVEN!!!! Social distance is DEADLY for society...
Numberphile talked about the underlying mathematics and how to construct the SIR models, 3B1B talked about how to experiment with those models in order to test and predict different containment strategies.
The "second wave" scenario is why it will be very challenging to determine an end date to the social distancing, shelter in place, etc strategies, that is until a COVID-19 vaccine is available to the global population.
9:15 Anyone who has played Pandemic 2 knows this. If you give your victims internal hemorrhaging immediately, you'll lose pretty quickly. If you infect the entire world with no symptoms at all, then instantly mutate organ failure, you'll win (assuming you started in Madagascar which, as we all know, is the safest place on Earth). Thank god diseases don't suddenly mutate kill-switches like that in real life.
Thankfully this is not exactly the case in real life. While the coronavirus already has a few slightly different strains, viruses aren't hiveminds that can mutate all at once.
@@Thisisahandle701 He just live in UK, so he say that he is on legal lockdown, on UK. What's the problem ? You think you're special ? In India population is on lockdown, and so for the majority of people in the world. As far as I know, there is 3 billions of persons in lockdown( ~ 700 millions people in Europe)
@@a1mforthetop Doctors are of course well aware of the flaws of tests. Tests are not perfect, so they are never performed only once. The common practice is someone would be diagnosed only when multiple tests turn positive, sometimes a combination of different tests (RNA, viral protein, etc), & hence different sensitivity & specificity, would be used. If you don't trust one particular coin to be a fair one, then flip multiple ones multiple times to be sure.
@Zwenk Wiel if you could test most people and quarantine only them, social distancing is not a necessity anymore, which would preserve millions of jobs and livelihoods. Also, loads of people still won't isolate and stay away unless actually put in a hospital quarantine even with no to mild symptoms. Individualistic societies are not good at following expert advice...
incredible. not only super fun to watch, but also simple to understand. such a shame that so many cannot wrap their heads around such a concept when this video exists.
This channel is one of my favorite channels. It treats maths in a way that I could only wish someone taught me during my school life. I love this channel. Thanks for working so hard 3 blue 1 brown! You're the best!
I am more amazed to how you put up all this into scintillating presentation rather than the content itself. We already got tons of scientific journals and publications but good visualizers are what the world lacks now.
What if you added the role of 'medical staff': - After a certain threshold, the quarantine box is made. - Some susceptible dots of different colour are moved to the quarantine box - They move frequently from one dot to another - Have a lower chance of getting infected because they have protective gear (but does not mean they are completely safe from the virus) - Decrease the 'infected' time of the infected dots by a bit - After the dots recover, they move back into the big society box - Some of these medical staff stay in the big society box (medical staff not in quarantine areas). They interact in a similar way with other dots but with less protection. - These non-quarantine hospitals are also centralised areas people go to.
@@AngleSideSideThm Preppers heading for bunkers are less risky dots than dots isolated at home. They won't infect anyone or get infected, because they are REALLY staying put. They aren't even going to grocery stores.
I find the spread of viruses easier to understand on a macro scale when it's described with math, rather than from some 374560893745 contradicting politicians and what you read in the news. Thank you for this beautiful piece of mind. Cheers!
Thanks Grant. The “visualization” makes this whole thing more understandable to the layperson. Wondering if it’s possible to see if any of your scenarios match the data we’re seeing in the US, as a whole and maybe New York as a “worst case”?
We don’t have the actual parameters to do that. Also simulating 10+ million dots would be a bit tough I assume. But the parameters that he tweaked were arbitrary and nobody knows exact values at this point. It would be incredibly difficult to simulate such a case. You also don’t know how many people actually got sick and who recovered before testing. So many unknowns. This is more of a representative case to show effectiveness of different techniques. Just imagine, some people are dumb enough to break social distancing and it makes every effort go to waste. A person with confirmed covid19 went to a funeral after being ordered to self-isolate and because of that there are 99 new confirmed cases. We went from 35 to 135 in a matter of 4 days. Cant’t simulate all that :)
3B1B: I'll emphasize these are TOY models and I leave it to intelligence of the viewer to determine what it would mean for you. Me, the viewer: But if I have a TOY brain?
@@theavo I would imagine messier. I imagine the models are to account for every factor encountered in reality, and like in one of the models we saw the outcome can vary widely due to randomness.
This is excellent. As a member of a high risk group, all I can do is hygiene and isolate. My trips to the central market - grocery store- are first thing in the morning, I know what I need, and I head to the self checkout to avoid any conversation with the cashiers. No one visits my home. No one gets into my car.
This is me too. Isolating except for trips to the supermarket every 1.5-2 weeks or so. Haven't figured out a way to avoid that (no delivery where I am), so I've also been going early in the morning, using self checkout and not interacting with anyone. Hopefully that's enough.
@@LisaMichele I'm ordering groceries, and the delivery guys are wearing masks 100% of the time + almost everything that enters the house gets disinfected or washed with soap. So, only leave home once in two weeks to take out the trash (I only leave late at night, when streets are empty)
@@LisaMichele I do this! And wash clothes after coming from the market. Even disinfecting shoes. Then use the same clothes next time in 15 days. This is my day 10, only went out for the trash late night. So cool to know same approach is used. But I'm jobless and don't know how much time my borrowings will last...
Mrrossj01 same here. I go in the early morning if at all. Wash hands, change clothes, disinfect shoes and leave shoes in the garage. Wash and disinfect all items from the store. No chances!
Random fun math fact about maths: works no matter what language you do it in and in principle you can communicate via mathematics even if you share no common languages.
Sorry to be so pessimistic .. but i'm so irritated with this idea that WE are supposed solve the problem and ACT upon it. You want to know what the real.. true problem is that makes all these other factors seem irrelevant ? Its so hard to get tested ! If you're not in an area where the virus is already doing serious damage.. They wont do it unless you are BRUTALLY ILL. I want to go post this on every single corona virus video out there because i don't think people really know what is going on. I am male in early 30s living in the US, small town with only 1 confirmed case so far. I have GERD but otherwise no major underlying health conditions. My symptoms : Dry cough for several days follow by sudden Chest pain & Fever Local hospital has a big TENT set up in front of the main entrance. If you have flu like symptoms you may not enter the hospital before being examined in the tent. I call upfront, speak with the hotline, give them all my symptoms and explain that i am concerned because : 1. I was in the same office a couple weeks as a person who just recently tested positive. 2. I already had the flu.. like a month ago and ive had the flue every dam year its come in season and i never had this kind of pain in my lungs. They tell me to come by when the TENT opens up ( 09:30am) They examine me.. do a chest panel ( test for Influenza A & B and other KNOWN respiratory viruses ) and tell me i can go inside where the doctor will need to do an X-RAY and an EKG I don't have insurance, so i explain that i did not come to get treated, i came to get tested. Their reply : You do not meet the "Risk Factors " and i can leave but i need to sign something saying i left against medical advice. I tell them that i don't want to go inside and rack up thousands of dollars in hospital bills i just want to know if i have the virus so i can STAY AWAY FROM EVERYONE. They tell me.. stay away from everyone anyway for at least 2 weeks. I do a little reading and see that people suggest you see your family doctor.. get a referral and then you can get tested. Family doctor is booked solid. Call half a dozen other places - Not taking new patients. Next step : Physicians immediate care (walk in clinic). At these places you pay a lot just to be seen ($250) but they have some kind of rule in place where your bill can never exceed $500, no matter what kind of tests they do ( including X-RAY & EKG) So i go there. They have people outside the facility with masks and they asked about my reason for coming and if i have flu like symptoms. I respond truthfully. They tell me that unfortunately.. they cannot let me in. By the way.. the chest panel came back negative for all of it. There is my story. Now if you will excuse me , i'm going to rage THIS IS SO FUCKED UP !! THEY DON'T HAVE ENOUGH OF THESE DAMNED TEST KITS AND THEY SAVE THEM FOR THE ELDERLY AND PEOPLE WHO BASICALLY REPORT THEY ARE HAVING TROUBLE BREATHING !!!! BY THAT POINT.. YOU HAVE INFECTED GOD KNOWS HOW MANY PEOPLE !! ALL THEY DO THE TESTING FOR IS SO THEY KNOW HOW TO ALLOCATE HOSPITAL RESOURCES AND WHERE TO PLACE THE PATIENT WITHIN THE BUILDING. ITS COMPLETELY POINTLESS ! INSTEAD OF TESTING YOUNG , ACTIVE PEOPLE WHO ARE GOING TO WORK AND SOCIALIZE AND BUZZ AROUND (WHETHER YOU LIKE IT OR NOT ) SPREADING THIS FUCKING BUG , THEY SAVE IT FOR LIARS, PARANOID HYPOCHONDRIACS AND SICK OLD PEOPLE WHO BARELY EVEN LEAVE THE FUCKING HOUSE ANYWAY. HOW AM I SUPPOSED TO COMPLETELY ISOLATE MYSELF ANYWAY ?????? SHOULD I BUILD A KITCHEN AND SHOWER IN MY BEDROOM ?? WHERE DO I GET FOOD ? GO TO A GROCERY STORE (EPICENTER) OR ORDER IN WHICH BASICALLY EXPOSES SOME POOR DELIVERY KID WHO WILL GO ON TO VISIT 50 HOUSES IN THE NEXT 2 WEEKS AND HAND THEM THAT SAME FUCKING PEN TO SIGN THE BILL AND GIVE HIM A TIP. HOW WILL YOU INCORPORATE THAT IN YOUR SIMULATIONS ? OR 20 OTHER SIMILAR SCENARIOS OF CONTACT I CAN COME UP WITH RIGHT NOW ? SOCIAL DISTANCING MY ASS. TEST AS MANY PEOPLE AS YOU POSSIBLY CAN. THATS HOW YOU KILL THIS THING OFF AND NOT BURDEN THE MEDICAL SYSTEM. TESTING TAKES 2 DAYS. A SERIOUSLY ILL PERSON CAN BE HOSPITALIZED FOR WEEKS. WHO IS DOING THE MATH IN ALL THIS ? I'm sorry yall.. and i really like your video bud.. im just so stressed out. it just doesn't make any sense to me man. How the fuck do you not have enough testing kits. Or if you do, why are you being stingy with them ? Btw.. just for some further reference, since i cant stop ranting apparently, im an operations manager at a transportation company. Thats how i came in contact with an infected person. Our drivers are over the road and go through half the country visiting dozens of crowded shippers and receivers and those filthy fucking truck stops before they come into our small office to drop off paperwork and decide to take a shit in the one bathroom we all share. I just can't believe this shit.. its like some daydream nightmare JOKE. 80% of our contracted accounts are pharmaceutical freight and due to this trump bullshit our volume in those commodities went down 70%. We move over 500 TONS of contrast MRI agents - PER MONTH and twice that amount of over OTC and prescriptions medication. Wanna know what we're doing to make up for those losses ? DESTROYING all the costcos , walmarts, wholefoods.. you name it. Any refrigerated commodity that is in scarcity right now. This past month we have charged more money on the spot markets to haul frozen chicken than we have in the last 12 years of moving $1.2Million value truckloads of gadolinium medium. Did i say this was all a big JOKE ? This whole corona virus thing will go down in history as a testament to how incompetent we are as a human race and how stupid the world leaders are... AND A wild conspiracy about some serious masterminds out there behind a very ... very BIG SHORT followed by an obvious clearance sale of stocks that will inevitably go parabolic once this whole debacle is over. I think im gonna go lay down now.
Peter Jones i agree, anyways, i reccomend you search up event 201. It relates to your last statement about masterminds. It’s pretty scary and proves that this is no mere coincidence
Just wanted to say thank you. I do not mean this in terms of intelligence (as for which I've no doubt you lack), but you truly are a genius in what you do. All of your videos are wonderfully intuitive and provide a view of the most elegant aspects of not only mathematics, but also of problem solving and analysis in general. This video is truly excellent; you have provided new, original information to a community who craves for deeper understanding in the underlying dynamics of situations, specially of such relevance. Best wishes for the future, stay safe!
Max I think you maybe meant to say "... intelligence (as for which I've no doubt you have in abundance)" As it reads to me, you are sure there is a lack of intelligence
Beautiful take near the end of the video For the under appreciated heroes on the front lines during this pandemic and the ones that never happened because of you, thank you!
The fact that this channel is so good and doesn't have ads makes me really happy. Also, great job giving recognition to the people working in public health: have to ask, where in Bay Area do you live?
I would like a simulation where: A. Every new infection starts out as undetected for a time period, and B. where once a detection happens, 1/3 of the population immediately self quarantine's, and 1/3 begins social distancing. and I'd play around with those percentages, and see how the curve might (or might not) be affected... This seems like the reaction in my locality. EVERYONE is experiencing an almost forced social distancing because stores have closed either of their own accord or because of city order, but not everyone is taking it seriously and life seems to be going on as though nothing were happening.
One thing that might improve the quarantining model is to move people back into the population after a while, and simulating different time periods of "looks safe" (no symptoms anymore) vs time infectious. This might show the effect of people self-quarantining, but not realizing that they might be infectious before/after symptoms are visible.
Please, Grant, be very proud of yourself. This is a time when many people reveal their nihilism. You have chosen to be rational and make an honest contribution. Thank you.
@@_-Montana-_ I don't see how either of us claimed humans to have purpose or meaning. It just cheers me up to see that people are showing solidarity rather than despair.
@@not3lliott I may well be misunderstanding the ideas of nihilism, but it is the word I chose for the attitude that makes people respond to a crumbling society by saying 'Let it fall!'. This video is an example of someone saying 'No, if we can understand this problem, we can mend it.'
@@nathanielpranger7370 There might be other words to describe the attitude opposite the helpful, optimistic one in the video, but you're correct, as nihilism (Latin _nihil_ = nothing) has developed an added sense over the centuries of _destruction for destruction's sake is preferable to current conditions._
One of the first things that got me worried about COVID was how early it had reached Greenland and Iceland. But it really struggled for a while with Madagascar.
I always started my games off in China and with high inefectivity and low leathality for optimal strategy. It seems mother nature can play the game as well...
@@austinbryan6759 But what's funny is when China gets the virus under control, the whole world starts going much worse.... especially unexpectedly the US.
I can't believe this has been up since March and I've just now found it. I've actually been toying with the idea of creating the same kind of simulation since covid started, only trying to make it as close to reality as I reasonably can. I wonder why he didn't post any follow ups on this, introducing more and more features like putting people in families/houses, creating different centers where a lot of people go with different rules (like a hospital in particular is a great way to spread disease) and giving people profiles like essential workers and such. And most important of all, and much easier, get age into the equation and have a vulnerable population of people who will actually die if they get the disease and see how isolating those people or reducing their movement would fare. I will try to get this project to run and see if I can modify it. With not a lot of extra variables and abstractions I believe this would tend to be an actual scientific way of testing different real strategies. In any case, great job and amazing content, thumbs up all the way.
Calculating abilities of his Computer might be a reason. I can imagine these simulations getting more and more complicated the further you take them. But I agree, would have been nice. And yeah, I also just found this
@@ChrisEffpunkt if you mean that he may have or knew that he was going to hit a limit of how complicated the interactions could get and keep a constant rate to simulate in real time, that sounds like it would definitely happen as the complexity adds up, and it will depend on the complexity and the amount of particles. The good news is the particles just need to be statistically significant, so that number can be kept pretty low. That's an easy enough problem to get around tho, by not simulating in real time, but instead running a simulation which takes all the time it needs to figure out a timeline of events and store them somewhere. Then to show them happening in real time you just "play" the interactions in order and there's only the limit of the ability of the rendering engine to manage all the movements of the pixels or dots, which should handle a huge amount of movement. Of course as complexity continued to rise another limit would arise which would be how long the computation will take to produce the timeline (i.e if it takes a year to compute it, it may be futile). But that limit nowadays can be surpassed easily with distributed computing which can take what would take a year for one computer, and compute it in seconds. Only then you may run into a hard limit with an amount of interactions and complexity which would take all of the world's computers together too long to run it, but that limit should allow for many many magnitudes higher of complexity than the one for this simulation. I suspect the current state of the simulation is very far from being limited by his computer's resources and there's still a lot of leeway there, but if not, running the simulations and then playing them should be enough for the most complexity we could think of with a low number of particles. And furthermore, if we wanted to actually do it for a huge amount of particles, the strategies mentioned could be leveraged to compute at the actual limit of all the computing power we've got. Doing something like this, Google and Apple joined together for instance could use all of the data they have on mobile phone locations for different countries and simulations can be run using the actual movement of people to spread the disease. Sorry for the long explanation, I've thought about this a lot lately, and by the way I don't mean to put you or him or anyone down.
I am a 66 years old physical chemist and quantum chemist and from that also familiar with differential equations. So I've also been playing around with all kinds of SEIRS models. That was quite satisfactory but frankly speaking, I began to lose the overview. I refer to Chinese simulation programs with up to 12 or 15 coupled differential equations that also accounted for all the measures reported from 3Blue1Brown. But never in such a pedagogical superb way as 3Blue1Brown did it. I'm also blown away by that! Just 3 hours ago I rechecked it again with special consideration of the urban centers since especially in large German cities the infection figures per one hundred thousand inhabitants go through the roof and the second wave comes. With some goodwill now easy to understand, provided you invest some brain grease. Unfortunately, we only hear babbling in Talk-shows. Please Emiliano, try to adapt the simulation closer to reality and let me know. If you're interested I could provide you with helpful links to really interesting Chinese and Russian programs. I have also done some basic translations of Chinese to german or English. I wish I were younger and more familiar with modern programming and could do the job on my own, LOL. Greetings from Germany
In these simulations you have serious boundary effects, especially in the social distancing case. You should use PERIODIC boundary conditions, so that when one particles reach one side of the box, it would exit one side and enter the opposite side, thus effectively resulting in a volume (or surface, in this case) that is finite but with no boundaries.
people have boundaries. they generally stay within certain confines. i think this is a decent model. the main issue is the spacing relative to the size of the individual. i think the model needs to be bigger with more motion. this is supposed to be the entirety of movement of a person in a day, sped up and presented over the course of months
The scary thing is that's what coronavirus seems to do, it's like fucking Plague Inc. players designed a bioweapon. This shit is scary yet a lot of communities don't take it seriously.
Wouldn't work in the Plague Inc way because viruses can't mutate everywhere at once in the real world... but just imagine a disease like AIDS that has a years-long incubation time before showing... except it spreads like the flu or, well, covid-19, rather than only by sexual contact. The good news is that there's various reasons why something like that is very unlikely to actually happen.
Awesome video! I’d like to comment on the central hubs and the asymptomatic people. - Central hubs: I see that there are fundamentally different types of hubs (let me call it social gathering hot spots): markets (with limited intimate contact between participants); sport events (stadiums, where after a game supporters will hug each other and go out to drink some biers in large groups); and pubs, bars,discotheques (where people get physically very close to a large number of other people). I’d believe that the R on every of these types of hot spots is very different, so to determine what kind of event should be allowed at a given phase of a lockdown phase, it would be interesting to simulate them separately. - Rate of infection: I wonder if asymptomatic people (some of the yellow dots in your simulation), are as infectious as those that have symptoms (and sneeze and caught). Not sure how much this would change the simulations though... Finally, I have the feeling that mobility is very hard to model accurately. When I look at the concrete cases around me, I see that people do not move socially randomly like mostly made in simulations I’ve seen. Rather, we tend have a large number of contacts inside our immediate family, then a bit less with the colleagues in same the office, finally much less frequently with our friends/sport pals, finally with the rest of the community (in supermarkets, restaurants, bars, etc). In the of the simulations, it’s as if you create another layer of boxes within each community box. Inside those smaller boxes it is very difficult to enforce social distancing, but between them it’s easier to do it (eg close all restaurants and stop sport events, etc). I would expect that one could relax the social distancing rules inside a “micro box” (eg family, close friends, maybe office colleagues) while enforcing strict rules for other interactions (eg in those social hot spots that I mentioned before) and still get similar results that one obtain by asking every single person to observe strict social distancing to absolutely everyone (outside own’s household). (Off course, risk persons need strict social distancing from everyone, in my opinion). The purpose of all this talk... is that it’s easier for a society to follow some sort of tolerant social rules for a long time rather than very stringent (and certainly very effective) rules. If the rules are very stringent, humans can only follow them for weeks, not months, and start being careless after some time. I live in Switzerland and I can observe that after several weeks of carefully the following social distance rules, recently almost everyone started not to follow them at all... In any case, huge appraisal for your work !
3Blue1Brown: "In most towns, people don't actually spend their days drunkenly wandering around the city."
New Orleans: "Let me tell you about Mardi Gras."
😂
@@lbognini Ireland: ... saint Patrick's day
Me : "Wait they don't?"
😂😂😂
That's just daily life over there.
"All models are wrong; some are useful." Thanks for presenting this *very useful* model!
Another way of saying it: "Math is a shorthand for a theory but does not prove the theory (as such)." In this case the parameter values are assumptions and the various results are useful only to the politicians, which, moreover, strip or ignore the feedback/validation.
Very much agreed, but more so for models involving humans, and less so for natural phenomena we've studied for a very, very long time (think N-body simulations). Making this assessment is important.
@eric I like that angle much better, for two reasons. First, mistaking a representation for the thing it represents is a widespread issue across many domains, not just mathematical modeling. Second, "all models are wrong" sounds confused to me - like you should come up with a better definition of "right" and "wrong" in the contexts of models. It's silly to say "all maps are wrong but some are useful." Many maps are "right" but the threshold for a map being "right" is not as high as reflecting literally everything about the area being mapped.
Bruh, why use a semicolon if you're gonna use it wrong?
@@eleanorhuxley6959 Why was it wrong? Semicolons are used to separate complete sentences that are related. These are both complete sentences that are related in content, so I don't see why you're saying it's wrong.
I'm an Emergency Medicine Physician with a PhD background in modelling. I am absolutely blown away by your work combining phenomenally effective graphics with practical modelling and a patient methodical, logical argument. Even Edward Tufte would be jealous a what you've done here. Truly one of the best presentations on any topic I've ever seen in my entire life. You deserve a medal and Isaac Newton's chair.
I don't know how it panned out in Redlands, CA, but the UK screwed up by politicians overruling logicians, I'm afraid. Even now, mostly young male narcissists shop unmasked, as civil liberties don't allow questioning. A solution is they must visor up - we did a unique community visor acceptability study, and everyone can tolerate these alternatives - for the mask intolerant. Boom, as young folk say. Retired old UK medic here. Keep safe :)
Damn that was the nicest compliment I’ve ever seen. From a fellow MD myself, thank you for taking your time to write this representing our profession.
@@tim40gabby25 scared people is the problem. Older people ruining it for the young by promoting this fear. I’m disgusted by what my generation have done to the future of my kids and grandchildren.
I read that as "Isaac Newton's hair"
and absolutly unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory desease can travel over kilometers.
21:10 this aged really well. in that this is exactly what happened, in the US and Europe and nearly everywhere that initially suppressed it
Didn't the US suffer from like, five distinct waves? We failed to learn our lesson four different times. Amazing.
edit: and as I recall, they were like, the minimum viable definition of a wave, where we let up as soon as R dipped below 1, instead of actual cases dropping below some number.
@@ferociousfeind8538 R
@@honestwong8942 well, yeah, five instances of the U.S. almost getting a handle on the pandemic, and then seeing that R
*fact
4th wave 🔥🔥🔥
"Kissing, sneezing on each other, all that good stuff."
Sneezing on each other, as you do.
"Sneeze on me like one of your French girls."
I cracked up at that line hard. Lol.
@@NoriMori1992 I was cracking up too hard
The 100% social distancing model looks oddly crystalline and satisfying.
with the typical errors ^^
Looks exactly like a metal lattice, with slight lines of dislocation - which would represent grain boundaries in metal. If you think about how bonded atoms are basically dots that attract each other but repel each other at closer distances so can never touch, it's pretty analagous to a bunch of repelling dots confined in a space together. This is why you get the same structures and properties in both. Funny how maths do that sometimes.
My high school chemistry teacher tried to explain entropy in a similar way. A real-world example he gave was people choosing seats in a movie theater that’s initially empty and gradually fills up. When the occupancy is low, people are unlikely to sit down right next to a stranger, etc.
ruclips.net/video/0U_FBHKYqRk/видео.html
I would like to see a similar video about the supply of goods relating to differing levels of governmental price controls.
11:30 The way that the social distancing makes the dots in the top left box arrange themselves in a periodic/crystalline way, even forming grain boundaries, is so satisfying to watch!
I agree, that was deeply satisfying.
I NOTICED THAT TOO OMG the lil dots were forming crystals it was adorable
@@3blue1brown It appears that the infection rate is different within a grain vs at the boundary. I was looking at this and wondering whether any effect from this or the geometry of the box could have any physical interpretation, or whether it is an artifact of how this was simulated.
Oh, a chemist in the comment section)
Hello there))
That's known as a Wigner crystal. First observed in experiments in which electrons were poured onto the surface of liquid helium. The crystal becomes unstable at high density or as the level of fluctuations increases. Nicely explained in the Wikipedia article.
Its ridiculous how accurate most of this ended up being
coronavirus 101
right ?
I think the word is frightening tbh.
@@andyjackson3663 its not scary its math but its stupid how we could not stop this
we're even seeing a second wave in places where people are beginning to ignore social distancing rules
this series of animations were so stunning. Never before had I seen such a great educational content. I am a general physician and I had to go through a lot to remind my patients and families to practice social distancing. I broke my back, trying to simulate the epidemic during the recent 3-weeks . and here you did it perfectly. We owe you all!
"If they remain unnoticed and spreadable in everyone before becoming lethal."
Laughs in Plague Inc.
@@x_x5009 This is the most unrealistic part of Plague, Inc. unfortunately. Diseases mutate, yes, but they mutate independently, and IRL every mutation needs to (re-)infect people with the newer version in order to have the "new" effects. Whereas in Plague, Inc. all your previous infections mutate at once, which makes the game fun and easier, but also completely wrong.
@@amunak_ It would be interesting to make a scenario, in which you cannot evolve (or devolve) symptoms manually, just increasing and decreasing mutations and wait for it.
@@morroghaiky6580 or a scenario where every time you mutate it creates a new patient zero which has to infect everyone again, immune suppression could make it easier to infect pre-infected people.
You guys are forgetting biowarefare! I know that none of us can be certain of where this virus came from, and either way, we can't trust the CCP or anything they tell us or their people. They're already telling them that the virus came from America.
So that aside, it's my personal belief that this is a virus that's been created in a lab (in 2013- theres records of it, and it won awards). The genetic technical capabilities of the higher ups is insane, if you dont believe in aliens and backwards engineering secretly happening, then this concept will be implausible to you. The details, proofs, and reasons behind this happening dont matter for now, but if you want to learn more, Edge of Wonder has tons of videos on this and related topics.
Anyways. This virus either gets out by accident, or on purpose. The CCP is now taking advantage of the situation, and wants to use it to decimate everything, for it's own advantage. Since it's been created in a lab, I dont know what its capable of. I dont know if the virus itself is nanotechnology, and if those nanoparticles are spreadable from person to person. Maybe the virus is programmed to replicate unnatural human designed particles, and then is able to essentially have an on and off switch that can be turned remotely.
Another one of my beliefs is that, due to chinese prophecy (I /think/ that it's in the book, tui bei tu, but you'll have to check what edge of wonder said in their video), basically, this virus will do what it's doing now, and then months later, just stop out of nowhere. No trace of it, and we all go back to normal. Then, 10 years from now, it will come back again, then disappear forever. The only way this can make sense is if the virus has an on off switch like what i described. Theres a lot more details to this that I would like to explain but people dont like reading long walls of text for some reason.
It's funny because the mortality rate is actually rising like crazy so yes you're right it's getting more and more lethal. It used to be about 3% and now it's already 4.5%
18:00 "It's not a bug, it's a feature!"
No Clip
Lmao GG
Its a bug *AND* a feature.
Thats us the introverts
Myer Loeb it’s not a bug it’s a virus
When your model starts rebelling…
“just like the simulations!” has never felt so real
absolutely unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory disease can travel over kilometers.
@@igorpashev so that’s why u should wear a mask
@@igorpashev Respiratory diseases are not always airborne if thats what you mean. Also, I need sources on the last claim because ive only known airborne diseases to remain for mere hours. Airborne diseases also infect efficiently through moist droplets from one person, also made more effective within 6 feet. Wear masks, distance yourselves. Thx champ.
b
17:58 literally "it's not a bug, it's a feature"
Louis Auffret
“It’s a bug, but I liked it so much that I made it into a feature.”
Creamy Pasta
Camper van owners just spread the disease to different communities, though. It'd be like the quarantine box simulation, but instead of it being a quarantine box, you just send them all to the same community box.
Minimizes regard for pretentiousness.
"maybe we don't value the heroes behind them the way we should"
Well said
ruclips.net/video/0U_FBHKYqRk/видео.html
Capitalism doesn't value anything that can't turn a profit today or tomorrow. That is why real heroes like that don't get the attention they deserve.
It's like IT: do your job right, and no one will know you've done anything at all.
Yes, They are the unsung heroes
I like how the Pis are side-eyeing each other at 10:40. Like they're saying "Ew, it's YOU."
Hey Cary, nice to see you around (though not surprising).
Love your content, hope you're doing ok!
Aye it's Cary!
Yo, what is up my dude didn't expect to see you here, it seems like you're putting a lot of time into abacaba which is cool.
Nice to see my daddy in 3b1b
I do that at the grocery store now...
This shouldn't be a PSA on "Social Distancing," this should be required viewing for Pandemics 101. Great introduction on the thought process and problem solving that can go into epidemiology. As usual, very approachable, 3B1B. Thanks again.
Yes
This is calm and reasonable and smart. A video like this is not only informative, but reassuring as well. Thanks for weaponizing your math brain to help people Grant.
And thank you for doing exactly the same thing as Grant does, but with good natured snarky commentary about the Bible and history and Christianity things. You sir, are a treasure, juat like 3 blue 1 brown.
I never expected to see you here. Love Mom
The pandemic isn't real.
@@gloekgloek3046 Damage control rabbi detected.
and absolutely unrealistic: how to detect cases? where isolate them (at site)? R depends on absolute humidity. Distancing is bullshit, because respiratory disease can travel over kilometers.
"nope, I"m out" 11/10 would buy happily
Already feeling the "Nope" desire.
18:00 Now that’s called thinking outside the box.
I'm fresh lolwalker
those guys are us gamers having no problem sitting at our pc for weeks with literally no real life interaction
ruclips.net/video/0U_FBHKYqRk/видео.html
the box can always be expanded
There is an interesting tweak that can be done:
Separate the "removed" in dead and recovered.
Every infected has a probability of becoming dead or recovered.
Recovered become Susceptible after a long time (Due to losing the immunity or the virus mutating)
I'd go even further and separate the recovered into immune and susceptible categories. Might also include non-human factors like animal reservoirs and mutations.
if you really want to be pedantic you could make the recovery/death ratio dependent on the current number of active cases (like if hospitals are overrun). Then you could put numbers to flattening the curve lowering the death toll even if the same total number of people got infected.
I know this comment is two years old but I think not distinguishing between recovered and dead was the right call.
The probability of an infected person dying depends on a lot of factors that would make the video a lot more complicated, the general rule followed was that having a flatter curve reduces the number of deaths, which is simple and accurate enough for the pourposes of the video.
This is a really good video and I really like how you've gamed out many different scenarios and shown their outcomes.
My main criticism, and it's a pretty big one, is that you've given your preferred control mechanism, isolation of infected individuals, a huge advantage by not varying the lag between infection and isolation. I don't think isolation one day after infection is realistic for nCoV2 even in a scenario of very widespread testing. I would have liked to see how quickly the benefit of isolation falls as that lag grows to 2, 5, or 7 days.
A well-structured, polite and respectful criticism that tries to constructively bring more options to the table of an important discussion. Surely a material to be explored in a similar video. Thank you
Exactly. In our country, no protocols were made to address the pandemic until many confirmed cases were already reported and thus no efforts were made for social quarantine. It took weeks until the government responded. There's also the concern of the cases being way higher than actually reported.
This is the problem that we have. It seems that covid-19 may be infections for a number of days before becoming symptomatic in the patient. Some estimates up to 14. Even if we can have proximity testing and contact tracing it would be insufficient.
With any significant number of asymptomatic infections - whether or not they remain so - only full population testing *and* isolation of those testing positive could be effective in preventing further waves.
...And in some cases may remain asymptomatic and untested. Hence the importance of acting as though we might be infected.
THIS!
Also, I have the feeling that people are being "removed" way too quickly in his simulation, plus that the R0 for the real disease was around 3...
It's a good video about infectious diseases in general, and about the effects of different measures against the spread of it.
However, my fear is that people would underestimate the neccessity for social distancing/hygiene as they might think that isolating would have been the best anyway.
Anyway, I still think this should be shared globally (I mean the video, not the disease), so that people really understand why the should keep their distance and keep their hygiene up!
2 years ago, in a completely different simulation, I had my circles escaping the box when I added a repulsive force to solve collisions! I laughed a lot here. Nevertheless, many thanks for this great video (as usual). Cheers from Italy.
All the best wishes to Italy!
Stay healthy in Italy!
Take care and stay healthy
Almost like it was designed
Where and how are these simulations done? Can you please tell?
Is is just python with some libraries or is there some other specific software for this purpose?
Here’s another idea to simulate: homes
There’s small boxes outside of the main box, and each of those small boxes has 3 dots assigned to them.
Each dot has a ‘role’
1 dot has a ~25% chance to travel to a main area (grocery stores, restaurants etc.) on any day. Make three ‘hubs’ in the big box, with an equal chance for the ‘wonderers’ to travel to
The second dot has a ~75% chance on any day to leave their house, but stay away from the ‘hubs’ and try to avoid other dots.
The third dot has a ~10% chance to leave their house, but they stay right outside it.
Tactical Lemon Excellent concept. I might disagree with your specs. I'm prejudice towards attempting to model a family. At that, you may have a good beginning with three adults, perhaps a married couple with an aging parent. Now we need to add two or three children, each with various levels of activities of interaction with the adults and some potential for going off short distances with the intent to inter act with other children in somewhat similar situations and possibly those children's parents or guardians. I think I 'be lost focus on modeling a single family and have started a neighborhood. Oh well, in a more comprehensive and complex model, we'd want a central city and suburbs. NYC and SF are exceptional. Urban geographers may have something to add...
I was thinking that social distancing would probably be better modeled using many many small networks (families and tight social circles) as well.
His "communities" simulation does seem to result in the most interesting and maybe useful results. Beyond the "test and isolate" result, which has been infectious disease dogma for about 400 years in the West (probably longer in China). Of course, we have much improved abilities to test and isolate now... wish we (the US) would have used them -_-
I wonder if you could take a satellite image and set up the algorithm automatically. Identify structures, and then estimate the number of people who spend time there by the surface area it takes up. In big cities this would have to take into account the height of the buildings somehow I guess.
yes
@@m3po22 great idea but no need for satellite image, the phone network is enough, more accurate and you can track contaminated, that's exactly what they did in South Korea and China
Grant, thank you for this great presentation on simulating an epidemic. These visualizations are outstanding for understanding how the numbers really work.
Until now, I admired 3B1B for his amazing illustration. Now I also admire him for his social contribution.
9:52 is the main strategy in Plague Inc.
When he talked about how diceased individuals will not stay infective, I immediately thought "true, unless I evolve necrosis for my plague".
I was also thinking about that
Exactly
You mean the bad strategy.
Countless pie people were murdered in the making of this video.
*removed
Infinte RUclips Commentators cannot write a Shakespeare Play, only Monkeys
Pi people? Pie people.
we will never forget the fallen
haha
I ran into this because I was looking at your Bayes' Theorem videos. However, it's now 2 years later. It's remarkable (or, as someone below put it, "ridiculous," how prescient this ended up being. It's so clear which paths we followed and how it ended up being just like you predicted.
Wasn't expecting to feel such a wide range of emotions watching some pi symbols and dots shuffling about the screen. And genuinely funny moments too (the dots leaving the Bay Area lols) Great video!
18:00 It's funny how a bug made the model more realistic.
It's the preppers bugging out ;)
After all human mind is not perfectly programmed codes.
it was weird to even experience that , i guess the math package has some huge bugs in it
yeah, world of warcraft all again... (en.wikipedia.org/wiki/Corrupted_Blood_incident)
"It's not a bug, it's a feature" - every programmer at some point.
"even when all the parameters are the same, some runs take three times longer to reach this point than others"
This deserves more than just a side note.
With agents starting at random locations and then moving around randomly, you are likely to end up with a chaotic system.
It cannot be stressed enough how crucial it is to have multiple test runs for each set of parameters.
For some parameter sets, you might end up with almost everybody infected in one run and the infection quickly dying away in another run, just due to random fluctuations.
Virus simulations (as well as economic simulations and weather forecasts for that matter) can be highly susceptible to random noise.
@@00O3O1B Yes, a nice followup would include a qualitative analysis of the official numbers and adjustments that can be added to the series of algorithms. Even though at some point all you can do is measure the overall statistics, once the fundamental parameters of the virus are accounted for then you can isolate effects of behavioral interventions such as limiting borders, etc. I for one am interested in the cumulative effect of panic shopping and hoarding on the broader population in exacerbating issues. Those who could afford to prepare would do in excess immediately when it would have an exaggerated effect on those with lower income and it took weeks for stores to enact any restrictions to buffer the impacts.
Another major factor as this goes further will be housing instability and population movement, as that increases it will negate the effectiveness of travel restrictions.
Excellent point! This is why I started all the actions in this video once there was a sufficient number of cases. For each one I showed, I rand multiple to be sure they were qualitatively the same and tried to comment when they weren't. For proper modeling (as opposed to demonstration videos), you'd, of course, want to run many and find the right way to aggregate the results into your summaries.
@@3blue1brown So does this mean we really had to destroy the American economy to achieve this? Once the tests were available (today being maybe a few days to a week where proper testing has been a bit more consistent) could we have lifted the ban like, this week (03/28), now that we know we can test and isolate effectively? It really looked like once you have a grasp on testing and isolation, you could still live your life as long as you were being super hygienic. But can we count on EVERYONE being hygienic? Did you model only partial hygiene?
@@SeraphX2 I'm not sure the US is able to test and trace it's way out of this yet, but you're right that effective testing and isolation can halt the spread and prevent second waves (South Korea and Taiwan being excellent examples of this) with much less disruption. The disruption (and economic damage) will end when the disease does, and the fastest way to do that is to catch up with the disease as soon as possible. Shutdowns and distancing are important in getting to that point.
@@SeraphX2 the model would perhaps have to take into account hygiene/isolation compliance based on infection prevalence (fear) inorder to measure the effectiveness of public policy.
I love this channel... My university frequently refers to your videos as a good learning resource and I absolutely agree! (Also, I've known this channel for years, so I'm happy to see that my university approves it as a good source :) )
No blue dots were harmed in the making of this film.
Sir, honestly: the simulations and analyses were so well done and have so much common sense that I tilt my hat to your work in gratitude! I am definitely becoming a patron!
is that hat a fedora by any chance?
I'mma steal a comment I saw elsewhere:
Actions taken before a pandemic are always "alarmist," and actions taken afterward are always "inadequate"
So true, and it's soooo easy to criticize (and profitable if you are a journalist)
I remember reading an article covering the spanish flu, where cities handled the outbreak differently. The city with strong early measures was accused of being alarmist and such, but many more people survived.
tbh it depends. in France it was 1st inadequate then alarmist + inadequate again then nothing we'll surely be all dead by the end of the pandemia
@@nomoretalk2967 That's in part because of the socio-economic factor : A lot of people are being (rightfully or not, I won't weigh in) angry with Macron, and whatever he does, they'll keep being angry. "Gilets jaunes", pensions reforms, all those people have a deep bias against him.
Case in point, 2 weeks ago, on Thursday, he declared that schools and universities would close on Monday : He was called an alarmist. The next monday, he put confinement rules in place : He apparently did it "too late".
Better to be an alarmist than a Trump. I mean, inadequate.
Brilliant job! Well done my friend. Very informative, nicely illustrated and calmly presented. Straight to the point. No loud music, No commercial. I just wish I could find more channels like this. BRAVO !!!
Man 3Blue1Brown coming in clutch during these quarantine hours
RUclips likes to recommend me news videos instead. RECOMMEND THIS GUY
Watch more of his videos, at some point it'll start recommending his videos
I was recommended this video, but then again im subscribed as well so that kinda makes sense.
A "full city" model would be interesting... households with a few people each, workplaces, schools, markets, hospitals, retirement homes with younger workers, quarantine zones, X% asymptomatic cases... toilet paper hoarders...
Imagine all the world's intelligence agency doesn't have a package like that in their tool suite. This means governments are being purposefully dumb to become Thanos.
simulation like that would be very time and computationally costly.
@@SouthernHerdsman I'm certain some intelligence agencies do have that capability and model it thoroughly before releasing these pathogens against opposing nations. What do you think we're experiencing right now? A mistake, or a cold, calculated attack? There will be be a long train of conflict following this. Whether it was an intentional attack, an accidental release, or a natural occurrence, this will lead to war.
@@slartybarfastb3648 Cheer up
But for it to be even more accurate, surfaces touched and air that was coughed in would have to be added.
"What if people avoid contact Wich others for a while, but then they kind of get tired and stop?"
We don't have to simulate the answer to that question anymore, we have the real life data right here.
18:00 When the model is so realistic, it includes existing real life behavior that's clearly outside the initial assumptions that were strictly programmed in.
This video should be on every TV channel around the globe for at least one week. It would save thousands of life.
18:04 "It's not a bug, it's a feature" you nailed it
Right about now you can be saying "I TOLD YOU SO!!!"
As a high school math teacher, I'm making this video a recommended watch by all my student (especially those in Algebra 2, who were just learning about exponential growth functions. THANK YOU!
I'd really like to see a large version of this simulation run with parameters close to what we're experiencing in real life. It'll be like looking into the future.
For example:
- 32 "communities", 4 of which will contain central hubs and 4 of which will contain a central market
- 100% are asymptomatic for 3 days; 80% will show symptoms afterwards
- 70% will be quarantined after 1 day of showing symptoms
- The infection goes away after 14 days
- The infection radius is the same as in this video
- Recovered patients are immune to future infections
- 0.005% of people can go into total isolation ("leave the box") per day
- Travel rate between communities is normal (0.2) until a certain threshold is met (in which case, travel rates will go down by 75%)
- Social Distancing rate is low (20%) until a certain threshold is met (in which case, social distancing goes up to 75%, then 95% if things get nasty)
- Infection rate is high (0.2) until a certain threshold is met (in which case, infection rates are cut in half)
- For the first time only, if the number of active cases falls below a certain threshold, travel rates, social distancing, and infection rates go back to normal
- Certain Threshold: 0.5% of the population is infected
NOTE: The virus's actual asymptomatic period can go up to 14 days, and according to zemerick13, the infection can last a lot longer than 14 days.
That .5% threshold doesn't sound good. There would be 3 000 infected people in my city, for instance.
@@supernovaw39 if 3000 were infected that would mean only 12 dead, 11 of which were about to die shortly anyway
His source code is in the description, if anyone wants to try
3Blue1Brown is far too smart to have not run that simulation...so I worry that its ("its" being Craftie!'s simulation idea, which I think is brilliant) lack of inclusion might be a sign that the results aren't....uh...valuable....at the moment.....no now that I think about it 3Blue1Brown is likely far too ethical to run that simulation and not include it in the video regardless of what it might indicate.
They should!
- Would be interesting to model three stages of contagious 1 - asymptomatic (normal mobility) 2 - symptomatic but home (greater isolation) 3 - hospitilization (isolated)
Day 69:
Dots became sentient, and worked together to escape
Lutfi H nice
They are PI not dot
Day 69 is always particularly auspicious for working together towards a common goal
I think the escaped dots represent the people who saw the direction the world was going and preemptively committed suicide. I'm in college and it happened to two of my friends, one of whom I was pretty close with.
Pi-Net is coming!!
Thank you for putting this out. The effectiveness of how you explained this is the best I have seen, in fact, I had to share it because now is the time to understand the impact of your observations.
Everyone is playing checkers while Grant is playing 23 dimensional Plague Inc
Ha
It was a good game.
1000th like
in 100D chess in minecraft using only basic redstone
Do you mean he was playing : “PARADOX-BILLIARDS-VOSTROYAN-ROULETTE-TENTH DIMENSIONAL-HYPERCUBE-CHESS-STRIPPOKER!”?
This video was so popular they started doing this in real life
Uderated comment
go up my friend
I hope nobody tried to reanimate the control case scenario lol
im curious how having multiple central locations like a school and a few grocery stores will change this assuming that every dot picks 1 of the 3 to visit also it would be interesting to see how things would change if dots try to stay in clumps similar to a family staying at home
I think it dosn't matter if one person out of the family does the shopping, as he earlier talked about people in close proximity are at risk. Eventually the whole family will get it.
Just imagine the dot being a family, if one gets sick probably everyone else will as well.
QalinaCom the families are also just a similar case to the 12 different communities. Just apply many communities of size 1-7 within each community and you’ve effectively simulated a family
Model with visiting "city centers" works that way
I built a simulation that takes into account school & work (along with household life):
n8python.github.io/epidemicSimulation/
I explain it here:
ruclips.net/video/Yo1rL0hqHAQ/видео.html
Amazing work. This needs to be mandatory for everyone to learn.
This video answers literally all the math related questions I've been having about this whole thing.
The Covid-19 is a HOAX. Corona-Virus has never been isolated and checked with the 3 Postulates of Robert Koch. Pasteur has treated the whole lot/world ! ALL HAS BEEN PROVEN!!!! Social distance is DEADLY for society...
Numberphile: Who are you?
3B1B: I'm you but more polished.
Numberphile talked about the underlying mathematics and how to construct the SIR models, 3B1B talked about how to experiment with those models in order to test and predict different containment strategies.
nope
They are completely different
Different videos. You could actually follow along Numberphile's model, meanwhile this is just a showcase.
AHAAHA
17:57 The dots are going to Greenland
Covid managed to reach Greenland (and Iceland) surprisingly early. Madagascar managed to keep it out for quite a while.
Mikołaj Kuziuk only OG’s will understand.
good joke man
I'd say those dots are going to North Korea
Plague Inc player clearly invested in cold resistance and water transmission early. Well played
Outbreak: *starts*
Dots: Ight imma head out
Some people have a temptation to travel on airplanes when there an epidemic at 17:50
Just look at how damn fast it spread to the four corners of this earth.
Someone: *coughs* Guess I'll explore the world!
@@kaboomwinn4026 #epidemic
Pandemic starts
Medics: It's hell on Earth
Mathematicians: It's free real estate
People dropping dead from the virus.
Mathematicians: They are removed
LMAO why not right?
Bartooc ruclips.net/video/I6HM12hpblI/видео.html
The "second wave" scenario is why it will be very challenging to determine an end date to the social distancing, shelter in place, etc strategies, that is until a COVID-19 vaccine is available to the global population.
Well....this will change our lives for the comming year atleast. There is not going to be an end very soon.
It takes like a year to make a vaccine.
@@JohnnyProductionsOfficialTM depends how many rules you want to break!
We need to just man up and all get it, so we can get back to work and the economy can recover
@@VBYTP
Sure. The millions who will die that way are all old or weak or poor or unlucky. I'm sure you won't know any of them
Thanks!
Awesome unintended visualization of grain boundaries in cristalline structures at 11:45 in the top left.
11:31 Top left is a nice visualization of boundary layers in crystal structures
9:15 Anyone who has played Pandemic 2 knows this. If you give your victims internal hemorrhaging immediately, you'll lose pretty quickly. If you infect the entire world with no symptoms at all, then instantly mutate organ failure, you'll win (assuming you started in Madagascar which, as we all know, is the safest place on Earth). Thank god diseases don't suddenly mutate kill-switches like that in real life.
Well, diseases *can* mutate and suddenly become more deadly, but that just makes a new strand that doesen't affect the already infected people.
Thankfully this is not exactly the case in real life. While the coronavirus already has a few slightly different strains, viruses aren't hiveminds that can mutate all at once.
you still lose because of Greenland
How about make a virus with a long incubation and non-symptomatic spread time, with a 20x flu chance of death at the end?
Greenland is the safest not Madagascar
17:58 "It's not a bug, it's a feature"
Storytelling done by an error in your code. Nice. 👌🏽
This is pure gold!
Absolutely fantastic video
Gives an insightful outlook on ways to control a pandemic.
Just loved it
Please keep up the good work!
Oh yeah 24 Karats BABY!
21:43 : "I'm writing this during a pandemic." That just hit hard.
Why, you didn't know already? We're on legal lockdown here in the UK
@@StevenMartinGuitar You think you're special? All of Europe is in lockdown. -That's how you sound, by the way.
@@Thisisahandle701 He just live in UK, so he say that he is on legal lockdown, on UK. What's the problem ? You think you're special ? In India population is on lockdown, and so for the majority of people in the world. As far as I know, there is 3 billions of persons in lockdown( ~ 700 millions people in Europe)
@@miguelgazquez5717 not the point lmao
"Test, Test, Test!"
A sage has told us...
Grant has just demonstrated using a simulation how true these words that were.
The test might be flawed. Do you absolutely know that it works? I think not
@@a1mforthetop No test like these are ever 100% accurate, mainly because you can't calculate human error.
@@a1mforthetop Doctors are of course well aware of the flaws of tests. Tests are not perfect, so they are never performed only once.
The common practice is someone would be diagnosed only when multiple tests turn positive, sometimes a combination of different tests (RNA, viral protein, etc), & hence different sensitivity & specificity, would be used.
If you don't trust one particular coin to be a fair one, then flip multiple ones multiple times to be sure.
@@a1mforthetop That scenario is essentially encapsulated in the models as a fraction of people who are not identified as infectious.
@Zwenk Wiel if you could test most people and quarantine only them, social distancing is not a necessity anymore, which would preserve millions of jobs and livelihoods. Also, loads of people still won't isolate and stay away unless actually put in a hospital quarantine even with no to mild symptoms. Individualistic societies are not good at following expert advice...
incredible. not only super fun to watch, but also simple to understand. such a shame that so many cannot wrap their heads around such a concept when this video exists.
I love how the 100% social distancing case created a crystal
This channel is one of my favorite channels. It treats maths in a way that I could only wish someone taught me during my school life. I love this channel. Thanks for working so hard 3 blue 1 brown! You're the best!
It’s not just a video about math, it’s a video about the Coronavirus.
Everyone: It's a Pandemic.
3B1B: Well, It's also math.
surprise surprise, the best tool human invented is related to everything
*Inhale*
MATH RULES
*Exhale*
ruclips.net/video/0U_FBHKYqRk/видео.html
math is art
I am more amazed to how you put up all this into scintillating presentation rather than the content itself. We already got tons of scientific journals and publications but good visualizers are what the world lacks now.
cant agree more
What if you added the role of 'medical staff':
- After a certain threshold, the quarantine box is made.
- Some susceptible dots of different colour are moved to the quarantine box
- They move frequently from one dot to another
- Have a lower chance of getting infected because they have protective gear (but does not mean they are completely safe from the virus)
- Decrease the 'infected' time of the infected dots by a bit
- After the dots recover, they move back into the big society box
- Some of these medical staff stay in the big society box (medical staff not in quarantine areas). They interact in a similar way with other dots but with less protection.
- These non-quarantine hospitals are also centralised areas people go to.
The "escaped dots" around17:50 are the people leaving big cities.
They are the people leaving where ever they are currently. You shouldn't move anywhere for any reason.
I'd say they're more like preppers retreating to their bunkers.
@@AngleSideSideThm Preppers heading for bunkers are less risky dots than dots isolated at home. They won't infect anyone or get infected, because they are REALLY staying put. They aren't even going to grocery stores.
Going to Florida
@@Ricardo-C Your fascism is showing.
1:32 I like how those Pi's become sad whenever they get infected.
Mustafa Malik
Most of us feel miserable when ill, don’t we?
☹️
Thanks
I find the spread of viruses easier to understand on a macro scale when it's described with math, rather than from some 374560893745 contradicting politicians and what you read in the news.
Thank you for this beautiful piece of mind. Cheers!
Thanks Grant. The “visualization” makes this whole thing more understandable to the layperson.
Wondering if it’s possible to see if any of your scenarios match the data we’re seeing in the US, as a whole and maybe New York as a “worst case”?
We don’t have the actual parameters to do that. Also simulating 10+ million dots would be a bit tough I assume. But the parameters that he tweaked were arbitrary and nobody knows exact values at this point. It would be incredibly difficult to simulate such a case. You also don’t know how many people actually got sick and who recovered before testing. So many unknowns. This is more of a representative case to show effectiveness of different techniques. Just imagine, some people are dumb enough to break social distancing and it makes every effort go to waste. A person with confirmed covid19 went to a funeral after being ordered to self-isolate and because of that there are 99 new confirmed cases. We went from 35 to 135 in a matter of 4 days. Cant’t simulate all that :)
what Data ?
ruclips.net/video/bM9aZflBoOU/видео.html
3B1B: I'll emphasize these are TOY models and I leave it to intelligence of the viewer to determine what it would mean for you.
Me, the viewer: But if I have a TOY brain?
@Whats Thepoint what exactly do you mean by that? How does it look in practice, then?
@@theavo I would imagine messier. I imagine the models are to account for every factor encountered in reality, and like in one of the models we saw the outcome can vary widely due to randomness.
One of the three best videos on Covid19 on RUclips (Edit:The other's are Kurzgesagt's and your other video)
Thank you 3B1B
This is excellent. As a member of a high risk group, all I can do is hygiene and isolate. My trips to the central market - grocery store- are first thing in the morning, I know what I need, and I head to the self checkout to avoid any conversation with the cashiers. No one visits my home. No one gets into my car.
This is me too. Isolating except for trips to the supermarket every 1.5-2 weeks or so. Haven't figured out a way to avoid that (no delivery where I am), so I've also been going early in the morning, using self checkout and not interacting with anyone. Hopefully that's enough.
@@LisaMichele I'm ordering groceries, and the delivery guys are wearing masks 100% of the time + almost everything that enters the house gets disinfected or washed with soap. So, only leave home once in two weeks to take out the trash (I only leave late at night, when streets are empty)
@@LisaMichele I do this! And wash clothes after coming from the market. Even disinfecting shoes. Then use the same clothes next time in 15 days. This is my day 10, only went out for the trash late night. So cool to know same approach is used. But I'm jobless and don't know how much time my borrowings will last...
Mrrossj01 same here. I go in the early morning if at all. Wash hands, change clothes, disinfect shoes and leave shoes in the garage. Wash and disinfect all items from the store. No chances!
@ Good idea, I'll tell the delivery guys to drop off stuff at my front door next time.
Social distancing leads to crystallisation 😂
Looked familiar to me also.
he just simulates gases. some even defuse out the box
@@pythoncake2708 Darn helium.
This is the part that I love about maths the most: when we use it to understand a problem and make decisions how to act upon it.
Random fun math fact about maths: works no matter what language you do it in and in principle you can communicate via mathematics even if you share no common languages.
Sorry to be so pessimistic .. but i'm so irritated with this idea that WE are supposed solve the problem and ACT upon it.
You want to know what the real.. true problem is that makes all these other factors seem irrelevant ?
Its so hard to get tested !
If you're not in an area where the virus is already doing serious damage.. They wont do it unless you are BRUTALLY ILL.
I want to go post this on every single corona virus video out there because i don't think people really know what is going on.
I am male in early 30s living in the US, small town with only 1 confirmed case so far.
I have GERD but otherwise no major underlying health conditions.
My symptoms :
Dry cough for several days follow by sudden Chest pain & Fever
Local hospital has a big TENT set up in front of the main entrance.
If you have flu like symptoms you may not enter the hospital before being examined in the tent.
I call upfront, speak with the hotline, give them all my symptoms and explain that i am concerned because :
1. I was in the same office a couple weeks as a person who just recently tested positive.
2. I already had the flu.. like a month ago and ive had the flue every dam year its come in season and i never had this kind of pain in my lungs.
They tell me to come by when the TENT opens up ( 09:30am)
They examine me.. do a chest panel ( test for Influenza A & B and other KNOWN respiratory viruses ) and tell me i can go inside where the doctor will need to do an X-RAY and an EKG
I don't have insurance, so i explain that i did not come to get treated, i came to get tested.
Their reply : You do not meet the "Risk Factors " and i can leave but i need to sign something saying i left against medical advice.
I tell them that i don't want to go inside and rack up thousands of dollars in hospital bills i just want to know if i have the virus so i can STAY AWAY FROM EVERYONE.
They tell me.. stay away from everyone anyway for at least 2 weeks.
I do a little reading and see that people suggest you see your family doctor.. get a referral and then you can get tested.
Family doctor is booked solid.
Call half a dozen other places - Not taking new patients.
Next step : Physicians immediate care (walk in clinic).
At these places you pay a lot just to be seen ($250) but they have some kind of rule in place where your bill can never exceed $500, no matter what kind of tests they do ( including X-RAY & EKG) So i go there.
They have people outside the facility with masks and they asked about my reason for coming and if i have flu like symptoms.
I respond truthfully. They tell me that unfortunately.. they cannot let me in.
By the way.. the chest panel came back negative for all of it.
There is my story.
Now if you will excuse me , i'm going to rage
THIS IS SO FUCKED UP !!
THEY DON'T HAVE ENOUGH OF THESE DAMNED TEST KITS AND THEY SAVE THEM FOR THE ELDERLY AND PEOPLE WHO BASICALLY REPORT THEY ARE HAVING TROUBLE BREATHING !!!! BY THAT POINT.. YOU HAVE INFECTED GOD KNOWS HOW MANY PEOPLE !! ALL THEY DO THE TESTING FOR IS SO THEY KNOW HOW TO ALLOCATE HOSPITAL RESOURCES AND WHERE TO PLACE THE PATIENT WITHIN THE BUILDING.
ITS COMPLETELY POINTLESS ! INSTEAD OF TESTING YOUNG , ACTIVE PEOPLE WHO ARE GOING TO WORK AND SOCIALIZE AND BUZZ AROUND (WHETHER YOU LIKE IT OR NOT ) SPREADING THIS FUCKING BUG , THEY SAVE IT FOR LIARS, PARANOID HYPOCHONDRIACS AND SICK OLD PEOPLE WHO BARELY EVEN LEAVE THE FUCKING HOUSE ANYWAY.
HOW AM I SUPPOSED TO COMPLETELY ISOLATE MYSELF ANYWAY ?????? SHOULD I BUILD A KITCHEN AND SHOWER IN MY BEDROOM ??
WHERE DO I GET FOOD ? GO TO A GROCERY STORE (EPICENTER) OR ORDER IN WHICH BASICALLY EXPOSES SOME POOR DELIVERY KID WHO WILL GO ON TO VISIT 50 HOUSES IN THE NEXT 2 WEEKS AND HAND THEM THAT SAME FUCKING PEN TO SIGN THE BILL AND GIVE HIM A TIP.
HOW WILL YOU INCORPORATE THAT IN YOUR SIMULATIONS ? OR 20 OTHER SIMILAR SCENARIOS OF CONTACT I CAN COME UP WITH RIGHT NOW ?
SOCIAL DISTANCING MY ASS.
TEST AS MANY PEOPLE AS YOU POSSIBLY CAN.
THATS HOW YOU KILL THIS THING OFF AND NOT BURDEN THE MEDICAL SYSTEM.
TESTING TAKES 2 DAYS. A SERIOUSLY ILL PERSON CAN BE HOSPITALIZED FOR WEEKS.
WHO IS DOING THE MATH IN ALL THIS ?
I'm sorry yall.. and i really like your video bud.. im just so stressed out.
it just doesn't make any sense to me man.
How the fuck do you not have enough testing kits. Or if you do, why are you being stingy with them ?
Btw.. just for some further reference, since i cant stop ranting apparently, im an operations manager at a transportation company. Thats how i came in contact with an infected person.
Our drivers are over the road and go through half the country visiting dozens of crowded shippers and receivers and those filthy fucking truck stops before they come into our small office to drop off paperwork and decide to take a shit in the one bathroom we all share.
I just can't believe this shit.. its like some daydream nightmare JOKE.
80% of our contracted accounts are pharmaceutical freight and due to this trump bullshit our volume in those commodities went down 70%.
We move over 500 TONS of contrast MRI agents - PER MONTH and twice that amount of over OTC and prescriptions medication.
Wanna know what we're doing to make up for those losses ?
DESTROYING all the costcos , walmarts, wholefoods.. you name it. Any refrigerated commodity that is in scarcity right now.
This past month we have charged more money on the spot markets to haul frozen chicken than we have in the last 12 years of moving $1.2Million value truckloads of gadolinium medium.
Did i say this was all a big JOKE ?
This whole corona virus thing will go down in history as a testament to how incompetent we are as a human race and how stupid the world leaders are...
AND
A wild conspiracy about some serious masterminds out there behind a very ... very BIG SHORT followed by an obvious clearance sale of stocks that will inevitably go parabolic once this whole debacle is over.
I think im gonna go lay down now.
Peter Jones i agree, anyways, i reccomend you search up event 201. It relates to your last statement about masterminds. It’s pretty scary and proves that this is no mere coincidence
16:17 up to 16:30 This comparison is information I was looking for some time. Thank you 3Blue1Brown.
Just wanted to say thank you. I do not mean this in terms of intelligence (as for which I've no doubt you lack), but you truly are a genius in what you do. All of your videos are wonderfully intuitive and provide a view of the most elegant aspects of not only mathematics, but also of problem solving and analysis in general. This video is truly excellent; you have provided new, original information to a community who craves for deeper understanding in the underlying dynamics of situations, specially of such relevance. Best wishes for the future, stay safe!
Max I think you maybe meant to say "... intelligence (as for which I've no doubt you have in abundance)"
As it reads to me, you are sure there is a lack of intelligence
No, he meant 3Blue is dumb as a rock
Beautiful take near the end of the video
For the under appreciated heroes on the front lines during this pandemic and the ones that never happened because of you, thank you!
The fact that this channel is so good and doesn't have ads makes me really happy. Also, great job giving recognition to the people working in public health: have to ask, where in Bay Area do you live?
Chris Owens to be fair, RUclips isn’t putting any ads on videos about COVID 19
Thank you for making this video. It probably has saved some lives.
I would like a simulation where:
A. Every new infection starts out as undetected for a time period, and
B. where once a detection happens, 1/3 of the population immediately self quarantine's, and 1/3 begins social distancing. and I'd play around with those percentages, and see how the curve might (or might not) be affected...
This seems like the reaction in my locality. EVERYONE is experiencing an almost forced social distancing because stores have closed either of their own accord or because of city order, but not everyone is taking it seriously and life seems to be going on as though nothing were happening.
Interesting
One thing that might improve the quarantining model is to move people back into the population after a while, and simulating different time periods of "looks safe" (no symptoms anymore) vs time infectious. This might show the effect of people self-quarantining, but not realizing that they might be infectious before/after symptoms are visible.
This is fantastic, you have basically answered every question I have had for the past three weeks!
A remake of this including re-infections would be super interesting
Please, Grant, be very proud of yourself. This is a time when many people reveal their nihilism. You have chosen to be rational and make an honest contribution. Thank you.
To me, nihilism is a mix of ignorance, despair and cynicism. This video brings knowledge and hope.
@@_-Montana-_ I don't see how either of us claimed humans to have purpose or meaning. It just cheers me up to see that people are showing solidarity rather than despair.
@@nathanielpranger7370 And just what does nihilism have to do with that?
@@not3lliott I may well be misunderstanding the ideas of nihilism, but it is the word I chose for the attitude that makes people respond to a crumbling society by saying 'Let it fall!'. This video is an example of someone saying 'No, if we can understand this problem, we can mend it.'
@@nathanielpranger7370
There might be other words to describe the attitude opposite the helpful, optimistic one in the video, but you're correct, as nihilism (Latin _nihil_ = nothing) has developed an added sense over the centuries of _destruction for destruction's sake is preferable to current conditions._
Plauge Inc players: we've... seen things...
One of the first things that got me worried about COVID was how early it had reached Greenland and Iceland. But it really struggled for a while with Madagascar.
I always started my games off in China and with high inefectivity and low leathality for optimal strategy. It seems mother nature can play the game as well...
@@austinbryan6759 But what's funny is when China gets the virus under control, the whole world starts going much worse.... especially unexpectedly the US.
@@johnyang799 Hard to know since China likes to lie about their numbers. We have to use other mean to roughly guess the number.
Pandaxtor That’s a valid point you got there it’d be a shame if you had an anime profile picture you fricken weeb. *yare yare*
I love the style of your visual interpretation. I just love it!
I can't believe this has been up since March and I've just now found it. I've actually been toying with the idea of creating the same kind of simulation since covid started, only trying to make it as close to reality as I reasonably can. I wonder why he didn't post any follow ups on this, introducing more and more features like putting people in families/houses, creating different centers where a lot of people go with different rules (like a hospital in particular is a great way to spread disease) and giving people profiles like essential workers and such. And most important of all, and much easier, get age into the equation and have a vulnerable population of people who will actually die if they get the disease and see how isolating those people or reducing their movement would fare. I will try to get this project to run and see if I can modify it. With not a lot of extra variables and abstractions I believe this would tend to be an actual scientific way of testing different real strategies. In any case, great job and amazing content, thumbs up all the way.
Calculating abilities of his Computer might be a reason. I can imagine these simulations getting more and more complicated the further you take them. But I agree, would have been nice. And yeah, I also just found this
@@ChrisEffpunkt if you mean that he may have or knew that he was going to hit a limit of how complicated the interactions could get and keep a constant rate to simulate in real time, that sounds like it would definitely happen as the complexity adds up, and it will depend on the complexity and the amount of particles. The good news is the particles just need to be statistically significant, so that number can be kept pretty low.
That's an easy enough problem to get around tho, by not simulating in real time, but instead running a simulation which takes all the time it needs to figure out a timeline of events and store them somewhere. Then to show them happening in real time you just "play" the interactions in order and there's only the limit of the ability of the rendering engine to manage all the movements of the pixels or dots, which should handle a huge amount of movement. Of course as complexity continued to rise another limit would arise which would be how long the computation will take to produce the timeline (i.e if it takes a year to compute it, it may be futile). But that limit nowadays can be surpassed easily with distributed computing which can take what would take a year for one computer, and compute it in seconds. Only then you may run into a hard limit with an amount of interactions and complexity which would take all of the world's computers together too long to run it, but that limit should allow for many many magnitudes higher of complexity than the one for this simulation.
I suspect the current state of the simulation is very far from being limited by his computer's resources and there's still a lot of leeway there, but if not, running the simulations and then playing them should be enough for the most complexity we could think of with a low number of particles. And furthermore, if we wanted to actually do it for a huge amount of particles, the strategies mentioned could be leveraged to compute at the actual limit of all the computing power we've got. Doing something like this, Google and Apple joined together for instance could use all of the data they have on mobile phone locations for different countries and simulations can be run using the actual movement of people to spread the disease.
Sorry for the long explanation, I've thought about this a lot lately, and by the way I don't mean to put you or him or anyone down.
I am a 66 years old physical chemist and quantum chemist and from that also familiar with differential equations. So I've also been playing around with all kinds of SEIRS models.
That was quite satisfactory but frankly speaking, I began to lose the overview. I refer to Chinese simulation programs with up to 12 or 15 coupled differential equations that also accounted for all the measures reported from 3Blue1Brown. But never in such a pedagogical superb way as 3Blue1Brown did it. I'm also blown away by that! Just 3 hours ago I rechecked it again with special consideration of the urban centers since especially in large German cities the infection figures per one hundred thousand inhabitants go through the roof and the second wave comes. With some goodwill now easy to understand, provided you invest some brain grease. Unfortunately, we only hear babbling in Talk-shows. Please Emiliano, try to adapt the simulation closer to reality and let me know. If you're interested I could provide you with helpful links to really interesting Chinese and Russian programs. I have also done some basic translations of Chinese to german or English. I wish I were younger and more familiar with modern programming and could do the job on my own, LOL. Greetings from Germany
18:00 They were just epidemiologists thinking outside the box.
In these simulations you have serious boundary effects, especially in the social distancing case. You should use PERIODIC boundary conditions, so that when one particles reach one side of the box, it would exit one side and enter the opposite side, thus effectively resulting in a volume (or surface, in this case) that is finite but with no boundaries.
I guess that the bounds of the video are appropriate for cities, while the fact of there being multiple cities is reflected in the video too
people have boundaries. they generally stay within certain confines. i think this is a decent model. the main issue is the spacing relative to the size of the individual. i think the model needs to be bigger with more motion. this is supposed to be the entirety of movement of a person in a day, sped up and presented over the course of months
"Unnoticed and spreadable in *everyone* before eventually becoming lethal."
Another Plague Inc. employee I see? Mwahaha!
The scary thing is that's what coronavirus seems to do, it's like fucking Plague Inc. players designed a bioweapon. This shit is scary yet a lot of communities don't take it seriously.
@@ChronicVision1955 🤦Let me humour you for a moment. What is supposed to be the symptoms for those who die of a vaccine?
@@ChronicVision1955 Would you care to explain further as I am interested in hearing what you mean?
Wouldn't work in the Plague Inc way because viruses can't mutate everywhere at once in the real world... but just imagine a disease like AIDS that has a years-long incubation time before showing... except it spreads like the flu or, well, covid-19, rather than only by sexual contact.
The good news is that there's various reasons why something like that is very unlikely to actually happen.
@@asdfghyter the symptoms are not getting a pay-cheque in rubles
Awesome video!
I’d like to comment on the central hubs and the asymptomatic people.
- Central hubs: I see that there are fundamentally different types of hubs (let me call it social gathering hot spots): markets (with limited intimate contact between participants); sport events (stadiums, where after a game supporters will hug each other and go out to drink some biers in large groups); and pubs, bars,discotheques (where people get physically very close to a large number of other people). I’d believe that the R on every of these types of hot spots is very different, so to determine what kind of event should be allowed at a given phase of a lockdown phase, it would be interesting to simulate them separately.
- Rate of infection: I wonder if asymptomatic people (some of the yellow dots in your simulation), are as infectious as those that have symptoms (and sneeze and caught). Not sure how much this would change the simulations though...
Finally, I have the feeling that mobility is very hard to model accurately. When I look at the concrete cases around me, I see that people do not move socially randomly like mostly made in simulations I’ve seen. Rather, we tend have a large number of contacts inside our immediate family, then a bit less with the colleagues in same the office, finally much less frequently with our friends/sport pals, finally with the rest of the community (in supermarkets, restaurants, bars, etc). In the of the simulations, it’s as if you create another layer of boxes within each community box. Inside those smaller boxes it is very difficult to enforce social distancing, but between them it’s easier to do it (eg close all restaurants and stop sport events, etc). I would expect that one could relax the social distancing rules inside a “micro box” (eg family, close friends, maybe office colleagues) while enforcing strict rules for other interactions (eg in those social hot spots that I mentioned before) and still get similar results that one obtain by asking every single person to observe strict social distancing to absolutely everyone (outside own’s household). (Off course, risk persons need strict social distancing from everyone, in my opinion).
The purpose of all this talk... is that it’s easier for a society to follow some sort of tolerant social rules for a long time rather than very stringent (and certainly very effective) rules. If the rules are very stringent, humans can only follow them for weeks, not months, and start being careless after some time. I live in Switzerland and I can observe that after several weeks of carefully the following social distance rules, recently almost everyone started not to follow them at all...
In any case, huge appraisal for your work !
I think one of the most Important things is to make sure a hospital never becomes "the one central location"!