‹ BackHN Continuity

Thread

Burning Man death rates – A short lesson in statistics

179 points · 243 comments · viraj_shah

  1. Aurornis · · focus · HN ↗
    Looking at population-level statistics, even adjusted for age, is useless. Going to burning man requires a level of health, energy, free time (not dealing with health issues), and freedom from chronic illnesses that make traveling to a desert a bad idea.

    Most people at high risk of random death are filtered out before they arrive.

    There are so many better comparisons that could have been chosen. Comparing Burning Man to outdoor music festivals would be so much more accurate.

    1. p-e-w · · focus · HN ↗
      > Comparing Burning Man to outdoor music festivals would be so much more accurate.

      It probably wouldn’t be very interesting though, since those are too similar. One or two deaths over a three-day festival with some tens of thousands of people attending is completely normal even in highly regulated environments like the German mega-festivals.

      1. Aurornis · · focus · HN ↗
        The similarity is precisely what would make it an interesting comparator.

        I don't have time to research this fully, but the first few papers that popped up showed that a lot of the casualty events from music festivals are due to incidents like mass stampedes or even motor vehicles, which I suspect would be much lower at Burning Man. On the other hand, deaths from drug overdose or "structural collapse" are probably higher at Burning Man.

        Comparing like-for-like, or as close as you can get, is what enables these interesting comparisons. Pointing out that you're less likely to die of cancer or car crashes while attending burning is not helpful when you compare against population level statistics.

        1. swerner · · focus · HN ↗
          The kinds of drugs people at burning man prefer to take are not very likely to lead to death from overdose.
    2. alexgoodhart · · focus · HN ↗
      Yeah this article is oblivious in such a painful way. The author should really consider their relationship with statistics, which they seem to think govern the world rather than are an abstract extraction from it.

      Burning Man has a hugely skewed age demographic, one that has a significant share of deaths due to car+other accidents, suicide, overdose, homicide, and cancer. Like nearly 7/10 chance of dying from once of those.

      Beside cancer those are all strongly linked to poverty, and cancer death in younger adulthood may be linked to poverty.

      The place certainly has extremes but it’s also a destination where people go with groups and have basically a nanny state watching them.

      Inane work, op.

      1. glitcher · · focus · HN ↗
        Article mentions age demographics and "selection bias of healthier attendees and other confounding demographic factors" right at the top.

        Oblivious in such a painful? I'm still open to hearing a sound argument to back that up. Your arguments are more dismissive and did not address anything specific in the article you disagreed with.

      2. golfer · · focus · HN ↗
        Did you bother to read the article? This is covered.
    3. js2 · · focus · HN ↗
      The post admits this, first in the summary: "However, Burning Man’s near zero death rate over its existence is abnormal, and likely due to selection bias of healthier attendees and other confounding demographic factors".

      Then again in the final sentence: "The more likely explanation for the remaining gap is something no demographic slicing can capture. Burning Man selects for healthier people who are willing and able to spend a week in the desert heat, regardless of their age, race, or income bracket."

      1. fhn · · focus · HN ↗
        so the comparison is pointless by their own admission
        1. TheCoelacanth · · focus · HN ↗
          It's not pointless to figure some things out just because you can't figure everything out.
        2. gwern · · focus · HN ↗
          No, it's an upper bound, which is quite useful. And the more adjustments and demographic covariates you make, the better your bound gets.
    4. bawolff · · focus · HN ↗
      on the other hand, burning man seems to select for people who want to win a darwin award, so i'm kind of surprised at the low death rate.
      1. fingerlocks · · focus · HN ↗
        Not really. Around 50% of attendees have graduate or doctorate degrees. It selects for people that are smart, curious, wealthy, and not risk-adverse. This is very obvious if you’ve ever been.
        1. bawolff · · focus · HN ↗
          That is exactly the group i would expect to do stupid things. People with graduate degrees tend to be intellectually smart but not street smart. People who are wealthy tend to grow up being shielded from the consequences of their actions.
          1. willtemperley · · focus · HN ↗
            You make some wild generalisations there.

            Can you back up your claims?

            Have you actually been to Burning Man?

            1. bawolff · · focus · HN ↗
              well, yes. That was the entire point of my original comment, that the data is not matching my biases, which surprised me.

              I'd also point out that saying 50% of attendees are X doesn't really mean much. Hypothetically, If 0.1% of normal society take absurdly stupid risks, and 0.3% of burning man people take absurdly stupid risks, that would still be selecting for people who take absurd stupid risks.

          2. RugnirViking · · focus · HN ↗
            would you expect that statistics would back you up on this? that people with graduate degrees are more likely to say, die in an accident?
            1. bawolff · · focus · HN ↗
              That would be inconsistent with my comment, so no I would not expect that. (my bias would be that grad students on average lack street smarts, so those that engage in activities requiring street smarts would get into trouble. I would not expect the average grad student to be more likely to die in an accudent)
    5. michaelt · · focus · HN ↗
      I don't necessarily agree.

      Marathons select for people with health, energy etc and put them through a 6+ month cardio training regime. But the 5 hours of a marathon have a greatly elevated rate of cardiac arrest - because the strain of running a marathon more than compensates.

      Personally, I'm quite surprised to learn the week-long desert drug orgy with experimental buildings and art vehicles and 70,000 attendees regularly has no deaths.

      1. grogenaut · · focus · HN ↗
        I would guess marathons select for people trying to push themselves hard, in a short period of training time and for a relatively short period of time. I also wouldn't be surprised if marathon attendance skews upper 40s / low 50s, male, and recently got in shape. That's a bit different than sitting in the desert doing drugs. the risk there would be respiratory distress / fent od is my guess.

        it's a lot easier to get someone out of an opioid od in the field than it is out of cardiac arrest, or at least the treatment protocol is.

      2. Aurornis · · focus · HN ↗
        The comparison for marathons would be other athletic events or sports.
    6. gosub100 · · focus · HN ↗
      Also, car accidents are probably in the top 5 causes of death for that demographic. And there's almost no driving at BM. But lots of driving to and from, so perhaps that smooths out the curve while keeping deaths off the event?
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.