9 Comments
User's avatar
⭠ Return to thread
JanusIanitos's avatar

Going into the final weekend thread I know we're going to see a flurry of predictions or people asking for people's predictions. It's all part of the fun, especially with this community!

I have a bit of a meta-question instead: how does everyone feel about the accuracy of their predictions this year? I know I'm the curmudgeon that refuses to make predictions, but even within my internal thoughts I find this election cycle harder to guess at than usual.

If my typical internal-prediction confidence in past years was 75%, I'd be at 50% or possibly even lower this year. I know part of that is the shock of 2016 combined with 2020 being closer than expected.

There's more to it than those elections. I could go through my list but I don't know that doing so would add much. The simple summary is polling points at a close finish that could see things unresolved for a while, while the non-polling data is largely in our favor, all in an era where polling is increasingly fallible but not so fallible that we should dismiss it. There's an easy emotional appeal to dismiss polls out of hand because that's the worst data point for us, but polls are still valuable and informative.

End result is I think the error bars are larger this year on this what to expect for outcomes. Which leads to me wondering how everyone here feels about the error bars on expectations for this election, both with the media/populace at large, with the community here, and with themselves specifically.

Maybe I'm overthinking things and others feel this election is as knowable and no more prone to surprise than usual, or maybe others feel similar to me.

Expand full comment
Mark's avatar

I'm about 51% confident about my predictions. More states than usual are well within the margin of error, meaning that already unreliable polls are that much more indeterminant. Meanwhile, plenty of tea leaves point to an electorate once again realigning itself just as it did in 2016.

Will the managerial class shift to Harris outnumber the nonwhite shift to Trump or vice versa?

Will the male shift to Trump outnumber the female shift to Harris or vice versa?

Should we really expect more than 155 million people to vote as they did four years ago?

And has the broken state of modern polling somehow managed to capture in any way the correct modeling needed to qualify these shifts?

My lesson from 2016 (and to a lesser degree in 2020) is that our community got lucky in 2012. We mocked the Romney supporters who "unskewed the polls" before the election and we mocked them after the election when the polls vindicated us. But it could have just as easily have been us with the egg on our face if the polling error had ticked their way that year as it did the next two cycles.

My pre-emptive lesson this year is to take warning signs seriously. And right now, the situation in Nevada is a warning sign.....and the reason I'm still 51% confident of my predictions from seven days ago rather than 49%.

Expand full comment
Henrik's avatar

I don’t make predictions as a general rule.

It’s hard to assess, and I’ve said consistently this year that it feels like we’re flying a bit blind. It seems - from stuff like Nate Cohn’s article on poll adjustments, and Marquette running a R+5 sample in Wisconsin this week - that pollsters are trying to adjust for 2016/20, but who really knows. Other indicators seem a bit more bullish in our direction, but they’re rarely totally dispositive.

I’d say I’m cautiously optimistic, but bracing myself for disappointment

Expand full comment
safik's avatar

I'm definitely far more open to a lot of different possibilities than I was in previous elections. Being wrong about the last 2 elections has opened my mind to a lot of different possibilities about how it could play out.

Expand full comment
BigGame's avatar

I don't make specific predictions, but I will say I'd much rather be us than them. I felt awful late in the game in 2016. Growing up in western Iowa and living in Des Moines, I could see these communities where Obama was mostly competitive looking like total wipeouts and knowing we were in trouble in the Midwest. The real key for me has been in the House polling. The warning signs started blaring in early to mid-October that we were losing a ton of ground in Midwest battlegrounds. But it was slow to show up in state polling so people dismissed it. I hadn't discovered DKE yet, which was probably for the best because I was very down on our chances. This year looks wildly different. I'm getting the sense the bottom could fall out in the suburbs for MAGA candidates, and I don't think the state polling is accounting for that. Trump has also had the opposite of 2016 in that all of the news down the stretch has been awful for him. It's either been about his mental decline, racist rallies, taking away health care, and so much more.

Despite all of the worry, EV numbers mostly look good in my opinion when I look closely. Lot of work to got to turn out our voters, so there's always unease. If we don't turnout the story ends poorly. But I'm mostly happy. Even Nevada is looking better this week to me than it does to some, though I'll admit I'm no expert on the Sunbelt states. NC is the state I'm most uncomfortable with guessing. I get the sense there's more room to drop in the rurals there than some of the Midwestern states, but the Triangle has grown. Again though, feel less comfortable there. But I like what I'm seeing out of Milwaukee, and as a Dem consultant told me this week, "you can't swing a dead cat in Philly without hitting someone canvassing for Harris." She's run a great campaign given the unprecedented circumstances. Not perfect, but damn good. Hope I'm right. A few more days of hard work and we'll know.

Expand full comment
safik's avatar

I will on occasion slip back into a pre-2016 thought process of how to look at campaigns. A few days ago I saw a headline that "Harris challenges Trump to take cognitive test" and for a few seconds it made me a little worried but then I saw the context that she was asked about Trump calling her low iq and this was in her response and I thought that was fine. The reason I got a little worried because I thought if you're talking about Trump's cognitive ability unprompted in the last week of the campaign, that's probably not a good sign.

I bring that up because Trump's such a weirdo that we don't attribute that type of traditional political strategy to his behavior. But if you were to do that, the things he's done this week would suggest that he's headed for a bad loss. Really trying to turn Biden's gaffe (a story that hasn't broken through the political bubble) into the predominant story in the days immediately before the election would only be done by someone that thinks they're about to lose. And on a much lesser note him going to Dearborn, and courting the votes of a community that I belong to, but is also about 2% of the electorate in Michigan and many of us may not vote for Harris but she's still going to get way more votes from us than Trump is also loser behavior.

Expand full comment
sacman701's avatar

Less confident. I have six states I'm not sure about going into this. The last few times, I don't think I had more than two question marks although I was wrong about some of them (e.g. the blue wall states in 2016). I'm confident that Harris will win Michigan. I'm about 80% sure she wins Wisconsin and Pennsylvania. I think she's slightly favored in Georgia and Nevada, but they're still basically tossups. I have no idea about Arizona and North Carolina.

I'm fairly confident that the polls are either basically right or underestimating Harris, but that leaves a lot of room for uncertainty. Why I think Harris might do better than the polls suggest:

(1) Spending patterns and pundit ratings on congressional races suggest an environment close to 2020, while polling suggests something in between 2020 and 2022.

(2) By all accounts, Harris has a much better ground game than Trump does. Even if polling accurately reflects intentions, ground game can turn intentions into votes.

(3) There is some evidence that late undecideds (not just Puerto Ricans) are mostly breaking for Harris, and that she's won a higher share of early voters than registration splits would suggest.

Expand full comment
Zero Cool's avatar

Nevada is likely going to be another repeat of 2022 where the election will be really close. Hard to say right now officially who will win NV but instinctively, I have a gut feeling neither Harris nor Trump is going to win by a sizable margin.

Expand full comment
Aspi's avatar

Equally confident. Two states I'm undecided on POTUS, zero undecided on US Senate, and eight or ten undecided on US House.

Expand full comment