• maegul (he/they)@lemmy.ml
    link
    fedilink
    English
    arrow-up
    23
    ·
    3 days ago

    From the article

    In a new update, the space agency has increased the chances of asteroid 2024 YR4 colliding with Earth, with the probability of impact rising to 3.1 per cent or one-in-32 odds of impact — the highest probability of a collision yet.

    IE - 3%.

    3% events happen all of the time!

    The article stresses that this probability has been going up over the past year or so, which is likely neither here nor there, but I can totally understand how it’s alarming in a post-COVID world.

    • laranis@lemmy.zip
      link
      fedilink
      arrow-up
      3
      ·
      3 days ago

      Statistics isn’t my strong suit. What’s the probability of the probability increasing given that it has been increasing over time? Should we project the future probability to include this growth?

      • bandwidthcrisis@lemmy.world
        link
        fedilink
        arrow-up
        5
        ·
        2 days ago

        Scott Manley said in his video that it would almost certainly increase right up to a point where it would (probably) drop to 0.

        Imagine the asteroid position is going to be in a range 1 to 100. Earth will be at position 31 when passes. That’s a 1% chance.

        Now I tell you that it’s actually a 10 to 90 range ( like they predict the asteroid a bit more precisely).

        The chance is now 1 in 80 or 1.25% It went up!

        Now I say I’ve got more info, the number is 30 to 70. The chance is 1/40, 2.5% time to panic, it’s getting worse!

        Next it’s 40 to 70. That doesn’t include 31. The chance just dropped to 0.

        So the increase is showing a more accurate prediction still includes hitting earth, but at some point that prediction might show that it will miss.

      • maegul (he/they)@lemmy.ml
        link
        fedilink
        English
        arrow-up
        2
        ·
        3 days ago

        Really no idea personally. My hunch would be that it’s technically a fuzzy problem (what’s the system being measured here exactly?) but also one around which we have some experience and wisdom established by now. Otherwise, the probability has changed like twice or three times, so any statistical inference would likely be close to meaningless with that little data.