Bikes, Covid, and Relative Risks (keep on biking!)

Thanks to Raghu Parthasarathy of UO Physics and http://eighteenthelephant.wordpress.com/ for sending in the following:

..I’ve wondered about this exact comparison [of the relative risks of Covid and of bicycling] for a while. Here are some very rough numbers — I encourage someone to put together better ones.

TL/DR: My estimate is that the annual risk of dying by bike is 10,000 greater than of dying by Covid. Nonetheless I happily bike to work.

Let’s say about 9 people per year in Oregon die in bike accidents — from https://towardsdatascience.com/oregon-bicycle-motor-vehicle-collisions-33af6d4a47a9 ; that’s 9 per 4 million. (I’m surprised it’s not higher.) 

What fraction of Oregonians commute by bike? Even in Eugene, I doubt it’s over 5%. Let’s say the death rate per cyclist is therefore 9*20 = 180 per 4 million. The biking death rate per year is therefore about **50/million.**

In the past two years 6000 people have died of Covid in Oregon, so 3000 per year per 4 million. As we all know, the risk is extremely age dependent, and the demographics of bicyclists will be very different from that of Covid victims! (The fatality rate is *100 times* smaller for 40 year olds than 75; 1000 times for 20 year olds; https://pubmed.ncbi.nlm.nih.gov/33289900/) Very roughly, as a 45 year old, my Covid fatality risk (without vaccination) is therefore around 30/4 million; I’m vaccinated and boosted, so this becomes ~0.3/4 million — I think it’s 100x smaller, though I can’t find the reference now. Plus, I’m in good health. So at most about 0.1/million, and probably more like 0.01/million. The Covid death rate per year, for a healthy and vax’ed 45 year old, is therefore about **0.01/million.**

I happily bike to campus every day; my kids bike to school. We live with such risks; they make life worth living. It is wrong / ignorant / neurotic to think that our Covid risks are anywhere close to the risks that many people perceive them to be.

If the numbers are wrong, I welcome corrections. I suggest, however:
(i) snarky comments without numbers are not constructive
(ii) put your name on your posts. (The internet would be a better place if this were more common…)

Bookmark the permalink.

49 Responses to Bikes, Covid, and Relative Risks (keep on biking!)

  1. Dog says:

    For regular human beings, not physics professors, I think perceived risk is emotional and more powerful than any objective risk
    assessment in terms of on the ground decision making. Wanna change the world, get started with better
    science education early and often.

  2. thedude says:

    I happily bike to work to. I did wish fewer people were on cell phones though. I happily ride my bike on dirt or gravel just for fun too!!

    Risks are everywhere. Including in many of things I love most.

  3. Anonymous says:

    I’m not going to put my name on my post, because harassment for social media posts is real and no thank you.

    Biking has many benefits, which help outweigh the potential risk. It’s healthier, it saves money, it saves resources. I don’t see the same benefits in getting Covid.

    And check your assumptions about ” the demographics of bicyclists will be very different from that of Covid victims!” I’m in my 60s and immune-compromised, and I bike to the university every day. In addition, the assumption that those who survive Covid, like our biking children, have sailed through without harm, may well be premature. There is the fact that Covid seems to increase the risk of developing diabetes, for starters.

    Basically I’m tired of being told that X is riskier than Covid “for the average person,” and so I’m stupid for worrying about Covid. A whole lot of us are not that average person, and are pretty angry about being informed that our risk is worth dismissing.

    • thedude says:

      The Covid diabetes study is a pretty awful one.

      There’s can be long term injuries from biking other than death. I had a friend get by a car that drove over the top of him and parked. This burned his leg on engine. His head injury permanently changed his sense of humor (he went from being sarcastic and the CEO of an expanding company, to a token executive with no real responsibilities for years and NO sense of humor at all). He finally became his self again mostly a 3 or so years after the crash.

      I still bike to work.

    • Also Anonymous says:

      It strikes me that you’re not doing the right comparison: it’s not the benefits of “getting Covid” that we’re trading off; it’s the benefits of having human interactions like hugging, gathering over a shared meal, or having an impassioned debate over issues across a cafe table. I think these are worthwhile benefits, and deserve to have some weight in the risk/benefit analysis.

      Similarly, though, the folks arguing that biking and COVID pose similar risks are making the wrong comparisons. When I get on a bike, I assume all the risk. With a pandemic illness, the consequences of our collective risk decisions to overall public health are much more important than the risk assumed by any individual.

    • “check your assumptions about ‘the demographics of bicyclists will be very different from that of Covid victims!'” For example, “1.0% of workers 16-24 years old bike to work, while 0.7% of workers ages 25-44 and 0.4% of workers 45 and older commute by bicycle.” (American Community Survey data; https://www.census.gov/library/stories/2019/05/younger-workers-in-cities-more-likely-to-bike-to-work.html) I hope it is obvious that no one claims that zero 60-year-olds bicycle, nor that zero 40-year-olds die of Covid. Your commute is not a refutation of the overall demographics; if I state that most birds fly, the statement “ostriches don’t fly” doesn’t contradict it.

      Also, I am not “dismissing” your risk. Your assessment is your own to make! I do believe that *comparing* risks is important, and comparison without numbers is meaningless. You’re free to disagree with this statement; the anti-vaxxers would as well, and would perhaps argue that I’m “dismissing” their feelings when I point out that vaccines are safe and effective.

  4. Trevor Brunnenmeyer says:

    Oregon Heath Authority’s Breakthrough case report from last week claims 718 total breakthrough deaths, with 10 of them in the 40-49 age range. Given the vaccine has only been available for about a year, it would seem a reasonable risk for a person in that age range would be 10/(# of vaccinated 40-49 year olds). I don’t know what that number is, but it’s less than the population of the state. This puts the risk of dying in the last year of COVID slightly above the risk that a random person will die in a bicycle accident. This risk is significantly higher than your back of the envelope calculation. (https://www.oregon.gov/oha/covid19/Documents/DataReports/Breakthrough-Case-Report.pdf)

    Additionally, I think that there are multiple, fundamental problems with doing a simple risk comparison like this. This analysis only worries about the possibility of *death* from both cycling and COVID. The side effects of cycling to work are, in general, positive, whereas the side effects of COVID are not. Around half of people who get COVID experience some form of long COVID. (https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2784918, and https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003773#:~:text=The%20research%20found%20that%20over,significantly%20higher%20than%20after%20influenza.)
    You also can’t spread cycling accidents to other people if you get one yourself. (Well, provided they aren’t cycling really close to you) A large part of trying to prevent the spread of COVID is to prevent other people from becoming infected, not just personal risk.

    • thedude says:

      Most long COvid is psychosomatic. There’s also long term injuries from biking (concussions can permanently change people’s IQ and personality, cause brain fog, and create other long term conditions). There’s hundreds of those types of injuries for every fatal bike crash.

      I still bike.

      • just different says:

        “Most long covid is psychosomatic”? You know this how exactly?

        _
        I’m not going to waste time on the rest of this asinine argument. Whether or not you bike has absolutely nothing to do with whether or not there should be public health measures against covid.

        • thedude says:

          Are you not reading the actual medical literature? Most studies of long covid don’t have a control. The few that day might find for people that have had Covid, 52 percent report long Covid symptoms, and for people without Covid, 48 percent have report long covid symptoms.

          So do viruses have long lasting symptoms? Yes this has always been true. Venus Williams had her tennis career slowed down by a complications from a viral infection. But if you’re going to say I’m making stuff up by raising the issue of psychosomatic attribution bias as a primary issue with long Covid then you’re ignoring most of the medical literature. Read some papers instead of BS scare pieces in the NYT.

          • just different says:

            I can’t tell from the nonsense you wrote what exactly you’re referring to, but your hypothetical would only mean that long covid symptoms need not be caused by long covid. It doesn’t follow that most long covid cases are psychosomatic. And it really doesn’t follow that long covid isn’t something to be concerned about from a public health standpoint.

      • Alan Rempel says:

        Chutzpah! Begin with an unsupported assertion that’s very likely wrong and in any case irrelevant. Long Covid can be debilitating.

      • Compulsory Pessimist says:

        thedude, your assertion that “most long COVID is psychosomatic” is one of the most disgusting comments I have ever read on this blog.

        https://www.nature.com/articles/s41591-021-01292-y?ftag=YHF4eb9d17

    • Dog says:

      In addition, bicycle deaths seems to have gone up during the pandemic.

      https://www.advocacyadvance.org/2021/05/bicycle-injury-and-fatality-statistics-during-the-pandemic/

    • james says:

      >The side effects of cycling to work are, in general, positive, whereas the side effects of COVID are not.

      This isn’t really a fair comparison – yes, the side effects of COVID are obviously negative, but the side effects of behaviors that somewhat increase your chances of catching COVID are overwhelmingly positive. Long term isolation is awful for mental health, and human connection is important. When doing these kinds of risk calculations, whether about biking, COVID, or anything else, it’s probably worth being more explicit about the benefits of taking the considered risks.

    • Mitch Conner says:

      Trevor, the argument of Raghu’s post was that if you are young/healthy/vaccinated and ride a bike, you could reduce your risk of death much more by quitting biking than you could by taking any further COVID precautions. The rate of biking deaths per *cyclist* (i.e. the rate Raghu uses) is the relevant rate for such a person considering their risk (since they’re a cyclist, not a random person), not the rate for the entire population.

      As for side effects: pointing out effects a simple model neglects without quantifying their importance shouldn’t persuade anyone to ignore the model. If long COVID or the benefits of biking reverses the conclusion, it should be easy to do the math and show it.

      But even that wouldn’t alter what I think is the main argument: that many people take much greater risks in other arenas (biking, driving, drinking, sports, etc.) than they do with COVID. Many people overestimate the risk of COVID because they haven’t attempted to put a number on it in the way Raghu has. This is worth pointing out, and should encourage these people to relax.

      • Dog says:

        make sure you specifically identify death by covid in your rebuttal, rather than using the word COVID to cover it all. Personally, I have never worried about death by covid (fully vaxed and boosted), but
        I do worry about being infected by covid since that is statistically orders of magnitude higher than death by covid.

        • Mitch Conner says:

          Thanks, Dog. However, I really did mean COVID generally, lethal cases and otherwise. You are certainly more likely to catch COVID than die from it, but the cost of merely catching it is much lower than the cost of dying. Indeed, most of the expected cost of catching COVID comes from the death risk, not side effects; to be very concrete, a representative COVID case is associated with a loss of 0.055 quality-adjusted life-years, 0.048 of which are due to the risk of death (see table 3 here https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7938736/). If you aren’t worried about the lethal risk, I would relax and not worry about the things that won’t kill you either.

          • Dog says:

            hey, I work at the UO, the most wonderful/excellent place on the planet. What do I have to worry about?

    • Thanks for writing!
      About bicycling having benefits and Covid not: this is of course true, but bringing back more normal activities also has benefits, especially to mental health. (As others have noted.) Comparing those would be another topic in itself!

  5. Mediocre Opinion Man says:

    If you’re only discussing the personal fear of death by COVID that many may have, then I think you have a good point.
    However, the possibility of spreading the disease to others at high risk and the strain on our hospitals is the big scare for me.
    Also, it’s only a fair comparison when biking accidents start to increase exponentially without intermittent cycling bans and restrictions.
    Therefore, I don’t think it’s excessive to collectively keep our heads low until we as a society get better at controlling and treating COVID.

  6. Josh Kashinsky says:

    I think the basic numbers regarding fatality rates and bicycling are mediocre to adequate, but it isn’t like much better data exist. As noted in the crash data source you link, bicycle crashes are only reported if they involve a motor vehicle. “Single vehicle” bicycle crashes or bicycle-pedestrian crashes aren’t reported, although those have a significantly lower risk of fatality. Most bicycle safety efforts use severe injury as well as fatality numbers, but I suppose in this sort of apples to baseballs comparison you want to limit what differences you can. I’m not sure how I feel about comparing bicycle deaths to total population. By this methodology the >400 fatal motor vehicle crashes per year would 40 times more dangerous than cycling, which I don’t think holds up that well. (https://www.oregon.gov/odot/Data/Pages/Crash.aspx)

    Your commute mode share numbers are probably within the margin of error. We mostly get that from ACS, which only counts commute trips. Given the totality of trip types people take (even under COVID induced reduced mobility), I think you’re undercounting the amount of bicycling that occurs, and definitely the number of cyclists. (https://www.lcog.org/thempo/page/commuter-mode-shares) I’m not sure that comparing the total number of cyclists to total population adequately captures risk.

    I have no comments on the math after that. I would agree that we take many risks each day, regardless of travel mode. I think the risks of severe injury leading to death are exacerbated when we stress our healthcare system to the breaking point. And “Allez Opi-Omi” incidents aside, bicycle crashes do not result in more bicycle crashes. In fact, evidence suggests that the more bicyclists you put on the road, the fewer fatalities you see. (https://read.oecd-ilibrary.org/transport/cycling-health-and-safety_9789282105955-en#page1)

    • Thanks for writing! I certainly agree that better bicycling numbers would be great to read, though I doubt that considering them would span the orders of magnitude chasm I estimated. (Even if I’m off by a few orders of magnitude.) But, I’m happy to read a more thorough assessment. A lot depends on how fine-grained one wants to be.

  7. New Year Cat says:

    How nice for you and those younger than you that the risk from covid is mostly lower. What am I, chopped meat? I am tired of hearing it implied that people over 60 don’t matter and that the immune-compromised don’t matter. Just what is your point in dong this comparison? That we should not care for our community as a whole and for your parents’ and grand-parents’ generations because….. why? Just get back to our own lives as they were in 2019 because no one else matters? That seems to be the message behind your comparison. As pointed out above, bicycle accidents are not contaigious.

  8. Anas clypeata says:

    It’s a cute straw man. The math no doubt goes wrong somewhere, but let’s just stick to qualitative analysis:
    .
    Side effects of bicycling: better mental health, better physical health, less war in the middle east, a more livable planet for us and our descendants, less traffic, lower local emissions that harm human health, less land used for parking spaces, less noise pollution, etc. etc.
    .
    Side effects of COVID: long COVID, killing grandma, supply chain disruptions, cancelled airline flights, business closures and bankruptcies, economic turmoil, hospital workers with PTSD, worse availability of child care than usual, not getting to hug your friends, wrecked mental health from isolation and anxiety and reduced human interaction, etc. etc.

    • thedude says:

      The side effects of distancing and lockdowns, remote schooling etc.

      Depression, anxiety, inflation, suicidality, weight gain, delayed medical care, missing out a variety of life celebrations, learning less at school (0.3-0.8 standard deviations less)….

      I’m not saying get Covid. I’m saying avoiding life to avoid Covid has costs too (and normally this means you’re just exporting risks of going to the store or restaurants to high school educated people or drop outs because we’re upper class people who can work remotely i.e.the laptop class).

  9. Dog says:

    Like most discussions on this blog, the original thread has now
    been twisted out of control into something that was never said.

    The original thread was simply about statistical risk management. You could have made the exact same arguments by comparing risk of covid death to swimming, lightning strikes, getting on an airplane, teaching a class at the UO, drinking beer , posting on the UO matters forum, etc.

    Criticism of the original post would consist of faulty assumptions, and poor statistical analysis. Not the benefits of bicycle riding ,,,

  10. honest Uncle Bernie says:

    “The Covid death rate per year, for a healthy and vax’ed 45 year old, is therefore about **0.01/million.**”

    Now, I consider myself a humanistic scholar, a man of letters on a grandiose day, not a man of physics, let alone statistics, but way back in the day when I did take math and physics type stuff, I remember a professor who told us something like “It’s a good idea to learn to judge whether your computation makes sense or not. If it doesn’t make sense, if it doesn’t seem plausible, then you should be suspicious that you have gone wrong in your fancy calculation. At least, you should check things out carefully.” I’ve heard this advice is even more apt in the age of calculators.

    Now, there are probably about 4 million 45 year olds in this country (I will assume crudely that the covid odds in Oregon are about the same as in the country as a whole, it doesn’t affect the conclusion). Maybe 40 million people in their 40’s. Then the probability given by our esteemed physics colleague implies only about 4 chances in 10 that a SINGLE vaxxed, healthy person in their 40’s died from covid in 2021. Yes, 4/10. Yes, odds are that not a single healthy vaxxed person in their 40’s died of covid.

    I invite our esteemed colleague to reconsider his computation.

    • That’s a good point, and my estimate does seem low from that perspective. I’m happy to revisit these numbers, and I hope others do as well.

      From the Nov. 2020 CDC data (the most recent full month) there are 67 deaths among 34 million fully vaccinated people aged 30-49, so about 2 per million, or about 20 per million per year. (https://data.cdc.gov/Public-Health-Surveillance/Rates-of-COVID-19-Cases-or-Deaths-by-Age-Group-and/3rge-nu2a/data) This doesn’t differentiate boosted vs. twice vaccinated, but if the booster reduces risk by 90% (my vague memory; seems consistent with https://www.nejm.org/doi/full/10.1056/NEJMoa2115624), this puts it (roughly) at 2 per million per year — higher than my initial estimate. (It’s still lower than my bicycling estimate, but I suspect that that’s too low also.)

      I’m glad you consider me “esteemed.”

    • Mitch Conner says:

      That isn’t how you calculate that probability; the number you’ve given is the expected number of COVID deaths for that group: https://en.wikipedia.org/wiki/Binomial_distribution#Expected_value_and_variance
      In any case, it’s certainly possible that 0.01/million is an underestimate. But I don’t think that relaxing the assumptions made in the calculation will change the basic result that biking is riskier than COVID for the demographic at issue.
      To check, let’s try a more direct estimate of the COVID risk: there have been 10 breakthrough deaths in Oregon (https://www.oregon.gov/oha/covid19/Documents/DataReports/Breakthrough-Case-Report.pdf) and 40-49 year-olds are about 13% of the Oregon population and have a vaccination rate of 67.8% (see https://censusreporter.org/profiles/04000US41-oregon/ and https://usafacts.org/visualizations/covid-vaccine-tracker-states/state/oregon). So of the 4.218 million people in Oregon, ~372,000 are between 40 and 49 and vaccinated, implying a COVID breakthrough death risk of 10/372,000 = 27/million. Even in this case, the risk from COVID is still lower than that from biking by a factor of 2. Now, that was only vaccinated 40-49 year-olds, never mind if they were boosted and healthy (and they likely were not). Being boosted lowers your odds of death (compared to vaccination alone) by a factor of 5 to 7, and having no comorbidities lowers it by a further factor of 4 (see https://ourworldindata.org/covid-deaths-by-vaccination and https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8209444/), so those odds are probably closer to 1/million, so about 50x lower than the estimated risk of dying from biking.

      • honest Uncle Bernie says:

        Mitch, I thought it was clear I was keeping it simple. In any case, your calculation has some gross oversimplifications — the boosted are about as well off as the formerly fully vaccinated, and in any case, your 1/million just happens to be 100 times larger than the number given that Iaddressed. I think it’s more like 10/million, and see below about the 45 year old boosted healthy smart demographic. I happen to think the personal covid and bike risks are similar for that group. But of course covid has external risks too. Somehow bike wrecks are not filling up ICU’s, preventing stroke treatment, cancer surgeries…

  11. Observer says:

    “Dude” claims the side effects of taking Covid precautions are “Depression, anxiety, inflation, suicidality, weight gain, delayed medical care, missing out a variety of life celebrations, learning less at school.”

    Personally, I prefer to miss out on various in-person life celebrations and risk all these other dire consequences in order not to get Covid. Apparently the GTFF union and many members of the faculty union, who are advocating for the power to make their own decisions about going online, also prefer avoiding illness to missing “life celebrations.” You’d think if we were suffering that much from Covid precautions, we’d all be insisting on packing back into crowded rooms with our students without wearing any of those cumbersome masks.

    So Dude and his ilk may prefer to resume life as usual and hang the consequences (or as a certain Victorian gentleman put it, “if they are going to die, they had better do it, and reduce the surplus population.”). The fact that so many in our society insist on taking the precautions he finds silly suggests that the rest of us actually do prefer not getting Covid.

    • just different says:

      I emphasize that I do not agree with thedude’s general mentality about covid, but in fairness these are all legitimate public health concerns. The central point–which a few people here have made implicitly–is that public health precautions are collective measures and designed to protect our entire system, which is not the same as the sum of optimal outcomes for each individual within the system.

      _
      Those who are only assessing their personal risks are also assuming that the system will keep chugging along normally for them to rely on to keep their risks low. This is badly mistaken. Public health experts aren’t omniscient, but they’re the best information we have, so we should listen to them.

      • thedude says:

        Who says I’m not taking reasonable precautions.

        I’m vaccinated. My kids are vaccinated. We’re not having big indoor parties. I’m not exercising at the gym or going to indoor restaurants or going to concerts.

        My kids got Omicron last week (picked up through the voluntary public health screenings done by 4J each week). 1 probably picked it up a friend’s house who they went to play with on a Saturday. They are not going back until tomorrow. 1 had no symptoms, the other had a fever which broke after a day. I’m not teaching in person again until next Monday out of an abundance of caution.

        So I think I’m taking reasonable precautions. Should I have taken more to prevent them from getting Omicron? I guess I could have zoom schooled them this quarter and isolated them from all friends. Maybe that would have delayed it. I saw their misery and their complete lack of learning for a year and half. I’m not choosing to isolate them from the world to maybe delay it for a vaccine they may or may not prevent infection that may or may not be safe for them that won’t be available for kids for probably another year.

  12. not a UO Physics Prof says:

    In the midst of an attempt to appear quantitative and rational, why in heavens name did you include the massively qualitative “It is wrong / ignorant / neurotic to think that our Covid risks are anywhere close to the risks that many people perceive them to be.”? It doesn’t bother me to have silly calculations referring to personal circumstance of a healthy 40-something who bikes, although the assumptions can make it misleading to those who don’t read carefully. It *really* bothers me that you make that blanket statement denigrating ‘many’ people’s perceptions of risk (who is many? CDC? Fauci? Lane County Health Authority?) and talk about ‘our’ COVID risks (whose? your own self-centered world of a healthy 40-something? or ‘our’ society risk, where healthier 40-somethings infect myriad seniors who could have lived a bit longer and not died under horrific and isolated mass death circumstances? nobody who has been close to that situation would ever post such a glib little thing.) In second order, it also bothers me that because you included the qualitative judgement, someone, somewhere, is going to say ‘UO Physics Professor says it’s wrong/ignorant/neurotic to worry about COVID’.

  13. Dogmatic Ratios says:

    Something’s wrong with any analysis that calculates the 3rd leading cause of death as ‘less risky’ than a cause of 400 times fewer deaths.

    The ‘death risk’ of cycling could be compared to the ‘death risk’ of ways to not cycle: stay at home, travel by car, travel on foot, etc. It can be compared to cycling in places with better cycling infrastructure. But it can’t be compared to the ‘death risk’ of cancer, heart disease, or covid. The operational complexity behind the two numbers is vastly different, so they don’t measure the same thing.

  14. cdsinclair says:

    My math brain objected at 180/4 \approx 50

    Less snarky though: The relevant calculation is not the probabilities but the expected value—much more difficult to quantify because people will disagree on the value and cost of catching COVID and riding bikes.

    If the beginning of the pandemic was due to people not understanding exponential growth, then the long crawl out of COVID will be due to people not understanding very small numbers.

  15. cdsinclair says:

    I mean 45 is just sitting there.

  16. cdsinclair says:

    Let D be the event “dead”. Let C be the event “got COVID”. Let B be the event “biked”. Then on one hand is the calculation \P(D | B) = “probability of death by bike, given you biked”. This should be compared to \P(D | C) = “probability of death by COVID given you got COVID”, but the calculation (if I am understanding it correctly) instead compares it to \P(D intersect C) = “probability of death by COVID”. The author then goes on to condition the second probability on various events that the first conditional probability is not conditioned on. I suspect that being a 45 year old, vaccinated individual in good health also decreases the conditional probability of dying by bike given you are a biker.

    • honest Uncle Bernie says:

      bingo with your last sentence. The bicyclists that I see at risk all seem to be reckless UO students, mentally ill tent campers at Needle Park, old guys who can barely stay on their bikes. People with the mental acuity of 45 year old healthy boosted professors aren’t getting hit.

      • uomatters says:

        I love bicycles – I worked in a bike shop, sold them, raced them, still fix them. But I don’t ride them anymore. I knew one professor who was run over and left a wife and two kids. Another professor who needed extensive rehab to walk again. Sorry, I know this adds nothing to the covid/cycling probability dialectic. Please be careful either way.

        • cdsinclair says:

          The Pedestrian Lemma says that if a pedestrian crossing the street has probability p > 0 of being hit by a car each time they cross, and they cross the street every day, with probability one they’ll eventually get hit by a car. (In fact, it’ll almost surely happen infinitely often!). So your contribution has extended this lemma to professors on bicycles. Want to coauthor?

          • uomatters says:

            Assuming continuity, as the probability of car drivers dying of Covid increases the probability of being killed by a car will go to zero. The extension to Parthasarathy’s Covid/Bicycle problem is obvious. Sure, do you have a good grad student?

        • thedude says:

          I know a couple bikers who died recently.

          One was David Schumann, who I luckily had the pleasure of meeting at UO. He had a heart attack (or maybe it was a stroke) while biking.

          Chuck Long was an experienced local mountain biker who died after crashing while biking by himself (folks found him unconscious on a trail) and he never awoke from his coma months later.

          Life is precious. Always has been. Always will be.

  17. I appreciate the chance to write (thanks, UOM!) and to stimulate discussion. To be clear: I do not claim that my calculations are infallible — I note a correction in one of my responses above, and I could probably write more. (I’d be appalled if anyone *does* think my calculations are infallible, especially in a blog post!) My point, nicely articulated by Mitch Conner, above, is that one can and should articulate risks (and benefits) and assess them, quantitatively, as best as one can. It’s tempting for all of us to rely on vague feelings of unease. (Anti-vaxxers do this as well.) I think we should try to do better.

    I realize there are more comments I haven’t specifically responded to; sorry.

    P.S. To Chris: I always tell my students that pi = 3. I’m probably not allowed in math buildings…

    • uomatters says:

      And with that I officially call the comments for this post closed. Thanks Raghu!

  18. Ben Young says:

    For what it’s worth, I have always used “chance of getting in a car wreck” as a baseline level of personal risk I’m willing to assume.

    I’m honestly much more worried about neurological damage / “long covid” than I am about dying (and honestly that goes for car wrecks, too). Have you looked at the chances of such outcomes, inasmuch as they can be assessed? Some numbers I’ve seen are very high.

    Moreover I’m even more worried about the collective risk (risk to humanity) than personal risk. I think it’d be interesting to try and calculate what effort we collectively undertake to limit car crashes, and compare that to our epidemiological efforts with covid.

    • thedude says:

      About 35-40k people a year die in car crashes. There’s an age gradient too that that reflects experience, age, and driving amounts. You can almost cut that in half if you don’t drunk drive.

      If you look at aggregate deaths for the 2020-2022 period, and it’s clear covid deaths outnumber traffic risks for the 30-40 age group, by maybe 100-200 percent.

      So once you’re vaccinated, I think it’s entirely reasonable to conclude totally traffic accident risks cumulatively are more than Covid if you are under age 55. A lot more if you are under age 30.