On average, the D less R margin in the early vote mispredicted the final Clinton/Trump margin by 14 points! Pollsters get yelled at when their polls are off by even 3 points, and anything more than that is considered an absolute disaster. Imagine if a poll was off by 14 points: no one would ever listen to it again! And yet we get the same frankly amateurish analysis of the early vote in every election.

  • foggy@lemmy.world
    link
    fedilink
    arrow-up
    27
    arrow-down
    9
    ·
    edit-2
    10 days ago

    Nate silver,

    A man who gained his fame through truly superhuman statistical analyzes of sports events,

    For some fucking reason, decided to go balls deep into politics and has been an absolute utter fucking failure ever since. 0% accuracy.

    I have absolutely no interest in hearing anything this dipshit has to say about politics.

    Give me a prediction on next year’s March madness brackets, Nate. Stay in your fucking lane. Politics != Sports.

    • geekwithsoulOP
      link
      fedilink
      English
      arrow-up
      16
      arrow-down
      10
      ·
      10 days ago

      2008

      Silver’s final 2008 presidential election forecast accurately predicted the winner of 49 of the 50 states and the District of Columbia, missing only the prediction for Indiana.

      2010

      His 2010 congressional mid-term predictions were not as accurate as those made in 2008, but were still within the reported confidence interval. Silver predicted a Republican pickup of 54 seats in the House of Representatives; the GOP won 63 seats. Of the 37 gubernatorial races, FiveThirtyEight correctly predicted the winner of 36.[71]

      2012

      At the conclusion of that day, when Mitt Romney had conceded to Barack Obama, Silver’s model had correctly predicted the winner of every one of the 50 states and the District of Columbia.[79][80] Silver, along with at least three[81] academic-based analysts—Drew Linzer,[82] Simon Jackman,[83] and Josh Putnam[84]—who also aggregated pollsfrom multiple pollsters—thus was not only broadly correct about the election outcome, but also specifically predicted the outcomes for the nine swing states.[85]

      https://en.m.wikipedia.org/wiki/Nate_Silver

      I’d list others but I doubt you’d read it anyway

      • foggy@lemmy.world
        link
        fedilink
        arrow-up
        17
        arrow-down
        9
        ·
        edit-2
        10 days ago

        🙄

        I think the youth call what you’re doing “glizzy” or something?

        Nate Silver’s Political Prediction Misses:

        • 2008 Democratic Primaries – Miscalculated early on, adjusted projections as race continued.
        • 2016 Republican Primaries – Low initial odds for Trump’s nomination.
        • 2016 Presidential Election – Predicted Clinton over Trump with 71.4%.
        • 2020 Presidential Election – Projected Biden 89% to win; closer in swing states.
        • 2018 Midterm Senate Races – Overestimated Democratic chances in key states.
        • 2014 Midterms – Underrated Republican gains, particularly in the Senate.

        Nate Silver’s Sports Prediction Successes:

        • PECOTA Model – Accurate MLB player performance forecasting.
        • 2008 MLB Playoff Predictions – Success with playoff team forecasts.
        • March Madness – Generally accurate bracket predictions for early rounds.
        • 2012 MLB Season – High accuracy in team win predictions.
        • NBA Forecasts – Reliable win projections using Elo ratings.

        You wanna keep going tit for tat, homie? It won’t shake out in your favor. Let’s see you show me some sports failures. I’ll keep showing you successes.

        You keep showing me political successes. I’ll keep showing you failures.

        Let’s see who burns out first.

        Ask his wife or kids. Or just check their social media.

        The dude is talented. Not in politics.

        • mildlyusedbrain@lemmy.world
          link
          fedilink
          arrow-up
          3
          arrow-down
          2
          ·
          edit-2
          9 days ago

          I don’t think you understand statistics well enough to speak on this.

          2016 Presidential Election – Predicted Clinton over Trump with 71.4%.

          The above for example was showing outcomes of simulations. Many polls reported way higher outcomes for Clinton and he was one of the few models showing a trump outcome and Trump’s path to victory was in the simulation results. This wasn’t a fail but a success of his model

          Like when you are told that in a coin flip, heads is the outcome 51% of the time and hear every coin flip will be heads

        • itsJoelle@lemmy.world
          link
          fedilink
          arrow-up
          1
          ·
          edit-2
          9 days ago

          Do you mean “glazing”? Cause the homie reads as such.

          I haven’t come across the term “glizzy” before so it might be new.

        • geekwithsoulOP
          link
          fedilink
          English
          arrow-up
          10
          arrow-down
          14
          ·
          10 days ago

          Weirdly antagonistic tone and not sure when Silver pissed in your Wheaties, but you obviously have a hang up about him. No desire to go tit for tat, other than to say he’s been more reliably accurate over time than anyone else when it comes to politics. It’s like baseball - if you lifetime hit for .300, everyone is going to know your name.

          Also, the whole point of the article is that early voting patterns are not indicative of final results. That’s not polling analysis or data modeling, that’s just historical fact. I don’t think Silver is perfect, and he’s got problematic issues, but on this point he’s just pointing out the thing the media ignores because it gives them headlines galore for the last two weeks before the election.

          • flicker@lemmy.world
            link
            fedilink
            arrow-up
            11
            arrow-down
            2
            ·
            9 days ago

            I gotta weigh in on that “weirdly antagonistic tone,” too. You literally said you’d list more but you doubted that person would read it? Just antagonistic all over.

            You started it!

  • jeffw@lemmy.world
    link
    fedilink
    arrow-up
    9
    arrow-down
    2
    ·
    10 days ago

    No offense but wtf Nate? Why would you use 2016 data when early voting was massively expanded since 2020?

    • geekwithsoulOP
      link
      fedilink
      English
      arrow-up
      9
      arrow-down
      6
      ·
      edit-2
      10 days ago

      He literally talks about that at several points. 2020 is a horrible baseline for looking at anything analytically because it was such an outlier because of COVID. Too many other variables in 2020 for it to be applicable for anything

      • Ultraviolet@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        ·
        9 days ago

        Not just COVID, Trump’s main strategy in 2020 was to turn “taking the pandemic seriously” into a partisan issue in order to make it so early and mail voting become disproportionately blue (there was no significant correlation in the past) while poisoning the well on those ballots’ legitimacy.

      • jacksilver@lemmy.world
        link
        fedilink
        arrow-up
        1
        ·
        9 days ago

        Yeah, but have we really returned to what 2016 was like? This is the issue with a lot of analysis like this, lots of internal bias in the underlying approach.

        He could very well be right, but it’s just as likely we’re closer to 2020 than 2016. I know a lot of people that have gone to full mail-in-ballot (since covid) and I know others concerned the mail-is will be tampered with (given some have been set on fire). Not sure either of those things were in play back in 2016.

        • geekwithsoulOP
          link
          fedilink
          English
          arrow-up
          1
          ·
          9 days ago

          I think that’s sort of the point - if 2016 was our last “normal” election and early voting wasn’t prognosticative of election results then, there’s no hope it would be anything other than more variable and chaotic now.

          The point wasn’t about a “return to normal” or else he would be saying it was an indicator.