He got 2000 “wrong”… Or did he?

  • @lennybird
    link
    English
    27
    edit-2
    5 months ago

    And yet, he hasn’t predicted yet because there are many other “keys”. Case-in-point: see how incumbency worked for Trump.

    Also should be noted other reputable science-based algorithm designers like Nate Silver advises Biden to step down.

    Finally, the unprecedented nature of an open convention also means this guy has nothing to go on for extrapolation.

    • @UnderpantsWeevil
      link
      85 months ago

      Also should be noted other reputable science-based algorithm designers like Nate Silver advises Biden to step down.

      Nate’s algorithm is just a poll of polls. And his reasoning is incredibly short term and superficial.

      Nate wasn’t suggesting Biden drop out back in January when other candidates could run to replace him. He’s only saying it now, because Biden’s polling is at an all time low.

      If Biden recovers (likely, as the memory of the debate fades behind other current events) the pundits will start singing a different tune quickly enough.

      • @lennybird
        link
        English
        3
        edit-2
        5 months ago

        Nate’s algorithm is just a poll of polls. And his reasoning is incredibly short term and superficial.

        That isn’t true. Far more involved than that. https://fivethirtyeight.com/features/how-fivethirtyeights-2020-presidential-forecast-works-and-whats-different-because-of-covid-19/

        Nate went further post-debate performance, knowing where Biden stands in a variety of polling and that this was the worst debate perhaps in the history of debates they was a make-or-break moment for the campaign desperate to reach a widespread audience. Biden capturing the attention of 50 million people will not happen again between now and November. For many Americans this debate, which Republicans will never let anyone forget, will be the last thing they remember.

        More importantly there will be no major positive event that overrides it. That event, if it existed, already passed with Trump’s conviction.

        Nobody can provide me a single data-point where Biden isn’t performing significantly worse than his 2020 race where he won by merely 40,000 votes across 3 battleground states. Time to face hard truths.

        • @UnderpantsWeevil
          link
          15 months ago

          That isn’t true. Far more involved than that

          Step 1: Collect, analyze and adjust polls

          The entire baseline for his predictions, plus or minus some additional adjustments. Everything after that is rooted in the original poll-aggregate foundation.

          Nate went further post-debate performance, knowing where Biden stands in a variety of polling and that this was the worst debate perhaps in the history of debates

          He was not telling Biden to drop out the day before the debate. He was telling Biden to drop out the day after, but before he actually inserted “bad debate performance” into his model and rerun a thousand model elections. This is what always gets Nate in trouble. He shoots from the hip on hot-button issues, rather than remaining academic.

          More importantly there will be no major positive event that overrides it.

          Trump tanked two debates against Hillary and still squeaked through on election day. Nobody is going to be thinking about this debate by the time the Olympics are over.

          Nobody can provide me a single data-point where Biden isn’t performing significantly worse than his 2020 race where he won by merely 40,000 votes across 3 battleground states.

          2020 had some of the highest turnout in US history, thanks to mail-in voting and quarantine. 2024 is going to see a huge drop off in participation. It isn’t immediately clear which candidate is going to suffer the worst from the deficit in support, as Biden has banked hard on appeasing moderate conservative voters while Trump trundles further and further out into right-wing.

          They’re both deeply disliked candidates.

          • @lennybird
            link
            English
            2
            edit-2
            5 months ago

            lol my guy, you are reaching for some serious straws. If you’re not going off of this data, what ARE you going off, exactly…? Vibes? This is more denial than I can handle.

            He was not telling Biden to drop out the day before the debate. He was telling Biden to drop out the day after,

            Well no fucking shit! That’s called adapting to new information. Any good scientist or analyst does it. When an unprecedented event like THAT debate performance occurs, then yes, that means you must go back and readjust the model. Models only work if they actually factor in the latest information — you get this, right? Have you SEEN the post-debate polling coming out? It’s TERRIBLE:

            Post-Debate: “72 Percent Say Biden Unfit Mentally, Cognitively.”

            Post-Debate: “64% of Independents want Biden replaced on the ballot”; that’s more than they want Trump replaced on the ballot by 1%, by the way.

            Post-Debate: “Undecided voter focus group leans toward Trump after debate”

            Trump tanked two debates against Hillary and still squeaked through on election day. Nobody is going to be thinking about this debate by the time the Olympics are over.

            You live in this magical fairy-tale world where Trump is held to the same standard as Biden when he clearly is not. If voters were as informed as you and me then we wouldn’t have either of these fucking candidates. Neither of those poor debates come remotely close to what we saw yesterday from the person espousing to be the fighter to take on Donald Trump. It doesn’t change the reality reflected in every single piece of data we have in battleground states. Now you can choose to bury your head in the sand and go off vibes if you want, but good luck with that.