They don’t understand probability

  • Kazumara@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    25
    arrow-down
    1
    ·
    2 days ago

    What’s with the random comparisons? Sometimes it’s the whole US, sometimes specific parts. What does 500 rides mean? Annual deaths per 100’000 population per 500 rides?

      • finitebanjo@lemmy.world
        link
        fedilink
        English
        arrow-up
        4
        ·
        edit-2
        19 hours ago

        Although saying per capita here would be misleading, because it’s not like those people ride the subway 500 times AND drives a car everywhere. The ones who take the subway are the only ones at risk of subway homocide, the ones who drive are actually less likely to die in a traffic accident than people outside of their car like pedestrians and bikers, and the two groups probably are much different in size.

      • Kazumara@discuss.tchncs.de
        link
        fedilink
        English
        arrow-up
        10
        ·
        2 days ago

        Yeah but I don’t understand how it would combine with the 500 rides. Were they counting only people who rode 500 times a year? Or is the implication that for this line they didn’t go by 100000 people, but instead by 500 rides? That would make it incomparable. I just don’t get it

        • prole@lemmy.blahaj.zone
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 days ago

          Oh I had to go back to see what you were referring to with the 500 rides thing… Maybe that’s the sample size for NYC?

    • infinitesunrise@slrpnk.net
      link
      fedilink
      English
      arrow-up
      4
      ·
      edit-2
      2 days ago

      Yes 500 is about the average number of rides that an NYC subway commuter takes per year. Since all the other stats are also per year.

      • Rooster326@programming.dev
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        4 hours ago

        So you need to double it at the absolute minimum?

        There are 52 weeks in a year. 2 trips per day, 5 days/week and you’re at 520 rides already.

        That is with absolutely zero rides to the grocery store, literally anywhere but work and home. And that’s assuming it only takes 1 ride to get to work - which we’ll… Good luck with that anywhere in America even in NYC.

      • Kazumara@discuss.tchncs.de
        link
        fedilink
        English
        arrow-up
        5
        arrow-down
        2
        ·
        2 days ago

        So they only counted those who did 500 rides a year? Or they counted 50’000’000 rides as a proxy for 100’000 subway users?

        • elderorb@feddit.nl
          link
          fedilink
          English
          arrow-up
          11
          ·
          2 days ago

          I am guessing they took the average homicide rate per ride in the metro, and used those rates to create the per capita rate assuming they take 500 rides a year.

          • Kazumara@discuss.tchncs.de
            link
            fedilink
            English
            arrow-up
            2
            ·
            edit-2
            1 day ago

            Oh yeah I can see it, that would make sense! Thanks!

            Since a lot of tickets will be bought anonymously you can’t easily count unique users of the subway system, so the raw data probably consisted of numbers of homicides, and number of rides taken by all customers combined.

            But the rate between those wouldn’t fit the per person risk they want to be comparing. So to calculate a risk to a person they must have assumed a realistic case towards the higher end of ridership, that is, a commuter commuting 5 days a week for 50 weeks a year, to work and back, that would give the 500 rides number.