Hacker News new | past | comments | ask | show | jobs | submit login
Brains are bad at big numbers; it's impossible to grasp a million Covid deaths (theconversation.com)
9 points by deegles on May 7, 2023 | hide | past | favorite | 17 comments



2022

It seems covid hysterics feel they are much better at understanding numbers, and if only the rest of us did too, we'd be just as hysterical. I also remember being repeatedly told that only the enlightened can understand what exponential means.

To be a bit more serious, not reacting in the same way as you is not the same as not understanding. This smug "if you were as smart as us you'd agree with us" attitude causes a lot more problems than it solves, it's just playing to the base.


As rebuttals go, it's pretty empty. You basically just asserted "I don't agree" and then dived into ad hom against a straw man.

My dad wrote out a million by hand as marks in the 30s or 40s to teach himself what it felt like (he was a mathematician) but ultimately it's still just "the one (1) book I wrote a million marks in"

I tend to think of a million in terms of stadiums of people which are usually 50,000 or so, and then try to imagine 20 of them. To me, this also means aside from an inductive quality of counting sheep and inferring a million is out there, I fail to grasp it except as a meta set of something I know I can visualise.

Which is pretty much what the articles author is on about. We use scientific mega/giga prefixes to reduce large numbers to instances of smaller ones our brains are built around. One, two.. lots isn't far off how we're wired.

Also.. Bedford's law. https://en.m.wikipedia.org/wiki/Benford%27s_law


I tend to feel this is why biologically we were not ready for the Internet.

Dunbar's number [1] makes us feel as if there's only 150 people in our monkeysphere [2]. So when someone's yelling at us from the other side of the Internet, our brains aren't equipped to say "that's one in a million" - it feels like there's one in 150 and suddenly we've got SIWOTI Syndrome [3] going.

It really showed up during the pandemic - our brain couldn't process getting immediate notice that there was 150 people (among 8 million in our county) with a deadly disease.

[1] - https://en.wikipedia.org/wiki/Dunbar's_number

[2] - https://www.cracked.com/article_14990_what-monkeysphere.html

[3] - https://www.explainxkcd.com/wiki/index.php/SIWOTI_Syndrome


This also assumes people care.

The gun deaths pile up. Nothing is done.

The health care system deaths pile up. Nothing is done.

The AIDs deaths piled up years ago. A collective shrug was all there was.


The way I approach big numbers is to use it in a representative Major League Baseball stadium with 40,000 people, because even though it's hard to grok I at least a have a felt sense of what that is like. 1M COVID deaths, then, becomes 25 entire MLB stadiums (there are only 30 of them by the way) getting vaporized. I also use this for things like chances: a 1-in-400 would be the equivalent of 100 tickets in the stadium being selected for something, so I ask myself, "if I were at a game how excited or nervous would I be if they surprise-announced this?"


Humans can get used to a lot. There was an article about nurses in Baghdad hospital. They we joking and lough with hands covered in blood. When asked why, they said if they would be crying for everyone, they would be crying all days an nights. Life goes on. Not much can we do about covid, only check how the numbers were counted and compare with others. Like flue, crime, car crashes. These are more important because we can do something.


Setting aside the number issue, it's nearly impossible to grasp one death. All that we can say about a million Covid deaths is that there is something horrible beyond comprehension there. Or, I suppose, we can denigh that there is anything there to grasp.


It’s a terrible thought, and no disrespect is intended, but a large football stadium holds roughly 100,000 people. So a million people is about 10 large football stadiums. That seems like a reasonable way to comprehend the size and the amount of loss.


Yes, but what we can do is compare, as the article states.

In the UK, since early 2020 there have been 220K deaths from coronavirus in a population of 70 million. That's approximately 0.3% of the population. (My understanding is that the US has roughly the same proportion).

I feel very able to grasp that, you can just look at, say, ordinarily, very roughly 1% of the population dies per year.

edit: the actual percentage is slightly lower, but this is a good ballpark.

If we front load all of the coronavirus deaths into year 1 (this is pretty much true, it's really tailed off post vaccine), and we assume they're all "excess" deaths, then it's a bumper year in which we have 30%, maybe 40-50% more deaths than normal.

Whether you feel that's a lot is an opinion. I feel it's significant but didn't warrant the response it got.

I actually think that what people don't realise intuitively is that over half a million people die annually in the UK. We are actually indeed mortal.


> I feel it's significant but didn't warrant the response it got.

Worth keeping in mind that this would have been significantly worse had we not had the response we had (primarily because the NHS would not have been able to cope). Plus if our response was faster we likely would have saved even more lives.


Do you have an upper bound on this?

Something like - how many excess deaths as a percentage of population did we see in countries without functioning healthcare systems?

I'd probably have agreed with a year of lockdown/restrictions if we could avoid approx. 2% mortality with it.

It has to be at least more than one year of ordinary mortality to be worth significantly curtailing our daily lives for. Obviously that's just my opinion but it feels like a basic cost-benefit analysis.


Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020–21 - https://www.thelancet.com/article/S0140-6736(21)02796-3/full...

> The table provides summary statistics for each country and territory, including reported deaths due to COVID-19, estimated excess deaths due to the COVID-19 pandemic, the ratio of the two, and reported and excess all-age COVID-19 mortality rates. The magnitude of the excess mortality burden has varied substantially between countries. The highest estimated excess mortality rate due to COVID-19 was 734·9 deaths (95% UI 594·1–879·2) per 100 000 of the population in Bolivia (compared with the global rate of 120·3 [113·1–129·3]), whereas negative excess mortality rates were estimated in Iceland, Australia, Singapore, New Zealand, and Taiwan (province of China), representing a large range. Excess mortality rates exceeded 300 deaths per 100 000 in 21 countries. At the regional level, the highest estimated excess mortality rates were in Andean Latin America, eastern Europe, central Europe, southern sub-Saharan Africa, and central Latin America, with several locations outside these regions having similarly high rates, particularly Lebanon, Armenia, Tunisia, Libya, several regions in Italy, and several states in southern USA (figure 2, table). Rates varied substantially in western Europe; some countries had rates nearly as high as the locations listed previously, while others, such as Iceland, Norway, Ireland, and Cyprus, had some of the lowest rates in the world, at less than 50 deaths per 100 000. Mortality rates varied across north Africa and the Middle East, with high rates in Lebanon (416·2 deaths [347·4–515·5] per 100 000), Tunisia (324·2 [265·4–376·4] per 100 000), Libya (292·1 [232·2–358·8] per 100 000), and Iraq (280·1 [216·7–362·5] per 100 000). Low excess mortality rates were estimated across sub-Saharan Africa, with the notable exception of four nations in southern sub-Saharan Africa: Eswatini (634·9 [491·8–775·4] per 100 000), Lesotho (562·9 [460·8–691·4] per 100 000), Botswana (399·5 [329·1–482·1] per 100 000), and Namibia (395·6 [340·2–448·8] per 100 000). In south Asia, some states in India had excess mortality rates due to COVID-19 similar to those of some high-income countries in the northern hemisphere.

> In terms of cumulative number of excess deaths due to the COVID-19 pandemic from Jan 1, 2020, to Dec 31, 2021, the highest numbers of deaths were estimated in India (4·07 million [95% UI 3·71–4·36]) and the USA (1·13 million [1·08–1·18];

---

For the morality within the United States https://www.cdc.gov/mmwr/volumes/72/wr/mm7218a3.htm and https://www.cdc.gov/nchs/data/databriefs/db395-H.pdf https://www.cdc.gov/nchs/products/databriefs/db427.htm

On average, about 3 million people die each year. In 2019, there were 2,854,838 deaths. 2020 had 3,383,729 and 2021 had 3,464,231 deaths. I would be very concerned with the proposal that we need 12x the rate before curtailing various activities in daily life especially considering that increase was with mitigating efforts in place.


So if we take Bolivia as an upper bound, we can assume that approximately 0.7% of the population would die to coronavirus in the absence of restrictions and with a fairly poor healthcare system. (Of course there will still be some level of individual change, e.g. those at risk are definitely not going to be going about as usual - they're still not in many places).

My feeling then is that if we have a course of action that could have completely eliminated that 0.7%, e.g. it would result in zero covid or a perfect vaccine by the end of the course of action, assuming that it significantly curtails normal life (e.g. lockdown, no socialising, forced remote learning for schools, forced home working etc) we should spend approximately 0.7% of the average human lifespan doing it.

So, a lockdown of 6-7 months could be justified to completely eliminate the virus. It was fairly quickly obvious that eliminating the virus was not possible, so we clearly shouldn't go for that long, but that's our starting point.

The UK suffered approx 0.3% mortality that's about 0.4% better than Bolivia.

The UK spent 24 months in various states of restrictions, and perhaps 12 months in fairly severe lockdowns to the extent that normal socialization (for example, going to a group or class to meet new people) was legally prohibited.

Depending on how you define it, we spent 1.25-2.5% of the average lifespan in lockdowns for the sake of avoiding the deaths of 0.4% of the population.

That's a horrendous trade off.


I believe that you are missing a number of dimensions involved with the transmissibility when comparing the population and the population density which also factor in.

It would be akin to comparing the mortality rate for Vermont and New York (which hold the economic level relatively constant).

Vermont was 15.8/100,000 people while New York was 18.5/100,000.

Things like how much international travel and how much intra-region travel? How many cases were there before vaccination was possible?

Changing this to other units and using New York City in March of 2020 - living in NYC was 50 micromorts / day based on the excess deaths. So, each day have everyone in NYC go skydiving (8 micromorts per jump) 6 times and you'll have some people die from that... about 24,000 over the course of a month (data from https://www.nytimes.com/2020/05/22/well/live/putting-the-ris... ). And that was with mitigation efforts in place.

I am not sure that it is fair to do a one dimensional "Bolivia was 0.7% mortality and UK was 0.3% mortality" because Covid hit at different times and had different amounts of transmissibility because of the different population densities which in turn impacted the amount that hospitals were able to care for people.


Covid reactions have likely destroyed or ended more lives than were saved. There's no evidence that an even bigger overreaction would have been anything but even more destructive, if it was even possible. I'd like to think, despite what pushovers people turned out to be, that governments were very close to the maximum of what people would tolerate without open revolt, so there's not much more theatre they could have put on anyway.

If you meant some sort of sane triage, where old folks homes got extra attention and kids going to the park were left alone, then I agree. For context, I live in Quebec, where the government let thousands of our elderlies die due to poor institutional infection controls, while instead focusing their efforts of violating ordinary people's civil rights however they could.


Very reasonable and accurate interpretation. What I'm trying to figure out now is how so many brains could have fallen for the lies and exaggerations.

I think I greatly underestimated how much fear dominates most people. My models of disease were spot on but my models of human behavior are definitely terrible.


> "This pandemic has been full of hard-to-comprehend numbers."

Alternative explanation "hard-to-comprehend numbers" => "lies". I saved a news clip of the Governor of Hawai'i claiming on prime time news (unquestioned) that Covid had a 10% fatality rate (https://docs.google.com/spreadsheets/d/1cMdfajHRjvciOkROPDEA...), in August of 2020. The only thing hard-to-comprehend is how so many government leaders could repeat such outright lies over and over again.

It's not that "brains are bad at big numbers", it is that "brains are bad at big lies". When you try to make sense of it you can't, unless you realize that leaders lie and people believe them.

(On the 1M deaths lie: old age and underlying conditions actually caused most of these "Covid deaths")




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: