There are, fundamentally, two different parts to public electoral projections and analysis - model making, and myth making. Whether you are doing quantitative modelling or just sorting through a series of data points to come to your conclusion, you're model making. You're choosing how to analyze data, how to deal with problems and edge cases, all of the slings and arrows that model building comes with.
I haven't usually done that - at my previous stop, I was merely the snarky commentator who functioned as the occasional check on the quantitative work done, basically a qualitative check to make sure the model wasn't working wrongly because of an understandable, but flawed premise. My local knowledge averted a couple of errors, mostly in the 2019 UK election, like the time we privately ran the model and suddenly the Greens were winning an extra seat. It was a simple fix, but it was made slightly easier by the fact that I knew immediately where the seat was and why the model was acting odd.
In the Georgia runoffs, I didn't offer a projection myself publicly - gunshy from the events of November 2020 - but I did serve much the same purpose for the (smashingly successful!) model that Lak and Ali built - including being a driving force (or, at least, nagging annoyance) for the changes that made the model more aggressive - and, therefore, more right - towards the end. Without seeking to overstate my involvement - I didn't build the thing, and I deserve absolutely zero credit for the brilliant work it did in projecting the Early vote perfectly - I'm proud to have helped, in some small way, with the best publicly available forecast of the race.
My purpose wasn't to code, or to necessarily offer answers, but to serve as a sounding board and a reality check. The original version of the model was merely a replication of turnout levels by geography, without accounting for the nature of that turnout - they were, bluntly, missing the Democratic turnout enthusiasm advantage - which the Black electorate surge was but the best representation of - that was so crucial to the election. Sure, it's helpful to know that Democrats had a lead based on 2020 turnout levels, but given the fact that the electorate was Blacker and Democratic in 2021 than 2020, it was clear to me that the GOP were in a worse shape in the early vote than the model they originally had shown. And, eventually, they were able to figure out how to show at a micro level what I was seeing as a potential weak point.
This isn't, again, to self-aggrandize, but I'm thinking of all of this as Rachel Bitecofer makes an ass of herself once more, with a tone deaf and offensive tweet about how if Jamie Raskin had followed her on Twitter, he wouldn't have thought it safe to bring family with him to the Capitol on January 6th, an act he did because he didn't want to be away from family the day after he buried his child, who died of suicide.
Representative Raskin gave a speech on the Senate floor today as House manager for the impeachment trial where he talked about how his daughter told him that she never wanted to go back to the Capitol, and he was at many moments near tears - as were we all who watched. It was the most heartbreaking speech I have ever heard. And Bitecofer used the opportunity to self promote and to admonish Raskin.
I don't even know what to say about the level of crassness and idiocy that tweet embodies, but regardless, it highlights that Bitecofer isn't in this for the model making. She isn't in this to do good psephology. She is here to myth make. She is here to create a myth around herself, and to perpetuate that myth to anyone who will listen. She has no interest in introspection, no interest in humility, and no interest in learning from her mistakes - as plentiful as they've been. She's here to get a following and to be able to tell others that she knows all. Everything is an opportunity for self promotion, and everything is a game.
I'm thinking about the small part I played in helping Lak because it's clear that there is nobody around Bitecofer able or willing to be that check. It's an important responsibility of anyone in public-facing life - the ability to have someone check you when you lose the plot is important, and one that's very important. Is it always fun to have someone poking holes in your work or your ideas? No, of course not. But it's brutally important, and a lot better than keeping yourself isolated from the problems with your work. Do that, and then you get burned in public.
That tweet - which I refuse to link, for obvious reasons - is still up, and she's been on Twitter in the interim. She should know that literally everyone is dunking on her for the tweet, but it's clear that she doesn't care. She doesn't care that it looks just so incredibly crass, she doesn't care that it's a disgraceful look, she doesn't care. She is insulated from the consequences of her actions, pushing a Super PAC that she hopes will have an annual budget of $3M - a laudable goal if your goal is to knock out two House members, but a laughable one otherwise. She is here for the mythmaking, as is her right. Just don't confuse that with either psephology or intellectual honesty. A search for either will show there's none.