• FourteenEyes [he/him]
    link
    fedilink
    English
    342 months ago

    tbh he was probably not far off other statistical analysis in the past but I recall that on the 2020 election his website was extremely wrong

      • ChestRockwell [comrade/them, any]
        link
        fedilink
        English
        43
        edit-2
        2 months ago

        Ok so this is weird revisionism, because I remember 538 was the least bullish on Clinton the entire time. The pre election podcast had her odds at 70% or something, which was way better than NYT, etc.

        Not to say Nate was “right” about 2016, but compared to other outlets, his actually was closer to the statistical average.

        The 93% might have been one of the forecast models, but the day of the election it was only 71.

        He still sucks, but he was actually better than other pundits in 2016. They actually got tons of shit from libs because they couldn’t comprehend that Trump had 1/3 odds

        • HumanBehaviorByBjork [any, undecided]
          link
          fedilink
          English
          20
          edit-2
          2 months ago

          this is also my recollection. i was checking that shit weekly. i didn’t want hillary to win, but her losing to such an obviously awful candidate seemed so unbelievable that the odds they were giving were incomprehensible to me.

        • JohnBrownNote [comrade/them, des/pair]
          link
          fedilink
          English
          92 months ago

          and even if it said 99% that doesn’t necessarily mean the model was wrong. Sometimes you roll a 1 but the odds of rolling not-1 are still 95% or 83⅓% or whatever

          fuck nate but also most of us, including me, don’t have math expertise

      • FourteenEyes [he/him]
        link
        fedilink
        English
        342 months ago

        And I believed it because the Adults In The Room knew What They Were Doing.

        8 years later I’m a rabid communist who counts down the hours until the West collapses