Why were people writing Fetterman off so bluntly near the end?

[wbrocks67]

The polling trends and assumptions about the national environment favored Oz, and coverage of the debate seemed to help him as well.

*polling trends among GOP outfits. Oz only led one nonpartisan poll, and it was Emerson by 1.

[I spent the winter writing songs about getting better]

https://twitter.com/NilesGApol/status/1595546911800434706?t=i83dvUtIxXwVWII1Ifb5...

[I spent the winter writing songs about getting better]

2016 and 2020 PTSD.

If that's the answer people need to get past this type of thinking. This is just as dumb and unfounded as the notion that Trump meant some random celebrity was likely to or even favored to jump in and win the Democratic nomination in 2020 or that Kanye West ever posted a threat to Biden.

[DaleCooper]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

[Pollster]

It's true that low-quality Republican-aligned pollsters flooded the zone here and in many other competitive races - I personally was conflicted on this because I knew those polls were nonsense yet their topline numbers looked very similar to the numbers we and other Democratic pollsters were pulling out in our own surveys (except in Nevada). Simultaneously, many media and university polls did wind up doing a good job but folks were skeptical with good reason because these were the exact outlets that badly missed in recent years.

Making sense of all of this required everybody - myself included - to make a lot of blanket assumptions which is something that personal preferences and partisan leanings easily cloud decision-making about. Very easy to Monday morning quarterback and the "correct" assumptions always become obvious/clear in hindsight but at the time there were good arguments in a lot of directions - all you have to do is look at many users' old pre-election posts to see them. Arguments about future predictions can still be good and well-reasoned even if they don't wind up being "accurate" per se - and I generally think its tacky to speak down on people for thinking something would happen differently than it did after the fact even though they may have had good reason to beforehand.

If, however, their reasoning was impractical/evidence-free/unserious then they should be mercilessly mocked.

[wbrocks67]

It's true that low-quality Republican-aligned pollsters flooded the zone here and in many other competitive races - I personally was conflicted on this because I knew those polls were nonsense yet their topline numbers looked very similar to the numbers we and other Democratic pollsters were pulling out in our own surveys (except in Nevada). Simultaneously, many media and university polls did wind up doing a good job but folks were skeptical with good reason because these were the exact outlets that badly missed in recent years.

Making sense of all of this required everybody - myself included - to make a lot of blanket assumptions which is something that personal preferences and partisan leanings easily cloud decision-making about. Very easy to Monday morning quarterback and the "correct" assumptions always become obvious/clear in hindsight but at the time there were good arguments in a lot of directions - all you have to do is look at many users' old pre-election posts to see them. Arguments about future predictions can still be good and well-reasoned even if they don't wind up being "accurate" per se - and I generally think its tacky to speak down on people for thinking something would happen differently than it did after the fact even though they may have had good reason to beforehand.

If, however, their reasoning was impractical/evidence-free/unserious then they should be mercilessly mocked.

You guys were seeing Oz up in PA towards the end?

I mean I think one of the biggest red flags for many of the low quality pollsters was their samples of minorities and young people. IA for example was finding the GOP getting 25-30% of the black vote, 70% of the Other/Hispanic vote, and GOP sometimes winning 18-29/18-34. Even with sample variation, there's no way that the GOP would be getting anywhere close to those numbers. It was a consistent trend too!

[Tekken_Guy]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

[Spectator]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

Masters was leading the final polling average.

[Tekken_Guy]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

Masters was leading the final polling average.

I don't remember Masters ever leading a polling average. Maybe it was RCP, because he definitely never led on 538.

[Skill and Chance]

The debate.

[wbrocks67]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

This wasn't CW though? Most people still thought Warnock would lead in the end with or without a runoff.

[cg41386]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

Oh?

[Tekken_Guy]

For me it was all the fake polls that flooded the discourse in that last month. I changed my prediction about a week before the election in both Pennsylvania and Arizona Senate to favor the Republicans.

Conventional wisdom still had Kelly winning. But most people also thought Laxalt would win NV, and while a runoff in GA was considered more likely than not it was expected that Walker would lead the first round.

We also saw this with Hassan's race. And in the governor's races a lot of people believed that Whitmer was now in a dead heat with Dixon.

Oh?

Yeah I meant fake polls showing Whitmer and Dixon roughly tied.

[Pericles]

I thought Fetterman would win until I saw the polling average was tied so I assumed Oz would overperform his polls. I did the same thing in Nevada though-I thought Laxalt would win until the polling average stayed tied so I assumed Cortez Masto would overperform.

[Sestak]

I think there could be a potential for some interesting anaylsis on how pundits or places like this forum perceive races that the media "gets bored of" in the closing weeks of the election.

In this case, the media decided that after the debate there wasn't much more here to get interested in, and they quickly were drawn to (and this forum followed them to) shinier objects in Nevada, Arizona, and Florida as we began to get early vote counts plus Kari Lake being Kari Lake. I think people just started to assume that "since we haven't heard anything, what was happening must be continuing to happen" and assumed that Oz had since fully closed the gap and taken a slight but notable lead. Didn't help, of course, that the fake polls completely mirrored this idea as well.

[I spent the winter writing songs about getting better]

I think I have the right to gloat about this and this is why:

[TwinGeeks99]

When they did so (especially after Ralston predicted Masto would win Nevada) I had a hunch they just didn't want to look dumb after predicting a red wave for basically the entire cycle. That's why you suddenly had Oz and Walker winning in the final predictions despite the data not showing any significant or sudden shift towards either of them.

[Amenhotep Bakari-Sellers]

I think I have the right to gloat about this and this is why:

Lol it's a 303 map but Johnson barely survived, SCOTUS did it's best to gerrymandering Johnson for reelection and Rs blocked Jt Committee so he didn't have to testify on what he did during the Insurrection

[South Dakota Democrat]

I can’t speak for others but I was worried about Republicans doing better than their poll numbers again given it’s happened in this area of the country in every election since 2012 and it was an open race in a Trump/Biden state.

Yeah, I think the long, well-documented history of polling misses in the Midwest and the “red wave” narrative being pushed the last two weeks caused me to go against my priors (originally having Dems as favorites in the Senate due to the extreme candidate quality differences)

Pennsylvania is not in the Midwest.

[South Dakota Democrat]

I don’t think anyone was actually writing Fetterman off, so you’re mostly just making up things here. There’s a difference between predicting a Republican will win a Toss-up race (which even I rated it in my final prediction) and outright rating a race Safe or even Likely R.

I pointed out repeatedly that Oz was far less likely to benefit from a late R surge among undecideds or to consolidate the undecided vote than a generic R due to his striking unfavorability numbers and his remarkable (and atypical) baggage as a candidate. While I thought a red wave environment would be enough for him to eke out a win, I certainly wouldn’t have predicted an Oz win if I had known that we were headed for the biggest D midterm overperformance since 1934.

Your gotchas are getting a little tiresome, honestly.

This is straight up false, actually.  You may not have been writing him off, but I can recall many posts where people were.

[politicallefty]

I can’t speak for others but I was worried about Republicans doing better than their poll numbers again given it’s happened in this area of the country in every election since 2012 and it was an open race in a Trump/Biden state.

Yeah, I think the long, well-documented history of polling misses in the Midwest and the “red wave” narrative being pushed the last two weeks caused me to go against my priors (originally having Dems as favorites in the Senate due to the extreme candidate quality differences)

Pennsylvania is not in the Midwest.

I do the same thing sometimes even though I know it's wrong. Rust Belt doesn't quite fit all the states you want to encompass either. I would say Michigan and Ohio are a lot more like Pennsylvania than Iowa or the Dakotas, but we lump the former states in with the latter. Great Lakes encompasses too much when you just want to refer to Wisconsin through to Pennsylvania.

[wbrocks67]

Pennsylvania is in a weird position, because while I do agree that we are not midwest, I would say if you're from Pittsburgh or Erie or the SW then you probably feel closer to the midwest label. However, for those of us on the East side / Philly area, I've always considered myself to live in the Northeast area of the country, never the midwest.

[tagimaucia]

The biggest takeaway for me from this election cycle is that when the polls are all-over-the-place or saying different things, you should really calibrate your expectations by looking at recent meaningful results that involve actual voters (the Washington State primary, post-Dobbs special elections). 

IMO, those indicators were always screaming that it would be close to a neutral year and that a red wave was definitely not in the cards.  The Washington primary pretty much perfectly predicted the national environment in 2020 as well, suggesting that it would be a few points closer than the average of the polls.

[MT Treasurer]

It's true that low-quality Republican-aligned pollsters flooded the zone here and in many other competitive races - I personally was conflicted on this because I knew those polls were nonsense yet their topline numbers looked very similar to the numbers we and other Democratic pollsters were pulling out in our own surveys (except in Nevada). Simultaneously, many media and university polls did wind up doing a good job but folks were skeptical with good reason because these were the exact outlets that badly missed in recent years.

Making sense of all of this required everybody - myself included - to make a lot of blanket assumptions which is something that personal preferences and partisan leanings easily cloud decision-making about. Very easy to Monday morning quarterback and the "correct" assumptions always become obvious/clear in hindsight but at the time there were good arguments in a lot of directions - all you have to do is look at many users' old pre-election posts to see them. Arguments about future predictions can still be good and well-reasoned even if they don't wind up being "accurate" per se - and I generally think its tacky to speak down on people for thinking something would happen differently than it did after the fact even though they may have had good reason to beforehand.

If, however, their reasoning was impractical/evidence-free/unserious then they should be mercilessly mocked.

I was (as I’m sure you all know by now) a lot more surprised by the NV loss. Any input on what went wrong for Laxalt/any campaign missteps on the Republican side?

Sisolak lost and Republicans won the popular vote for U.S. House 51-47.5, so the votes were clearly there for him (although there was no margin for error). While Laxalt was tainted by Trump, the party's poor national brand, and (perhaps) some of his previous runs for public office, he was effective as an attack dog and had no glaring issues as a candidate like Oz/Walker/Masters. Then again, neither did Budd, and he underperformed rather noticeably as well.

It really wasn’t a race that should have been lost by Republicans. I wish (and in fact already wished back in September) that Laxalt had created/would create some distance between himself and the national party on select issues, but other than that, it seemed like it was mostly external factors (and a far superior D turnout operation) which caused his defeat.

[Pollster]

It's true that low-quality Republican-aligned pollsters flooded the zone here and in many other competitive races - I personally was conflicted on this because I knew those polls were nonsense yet their topline numbers looked very similar to the numbers we and other Democratic pollsters were pulling out in our own surveys (except in Nevada). Simultaneously, many media and university polls did wind up doing a good job but folks were skeptical with good reason because these were the exact outlets that badly missed in recent years.

Making sense of all of this required everybody - myself included - to make a lot of blanket assumptions which is something that personal preferences and partisan leanings easily cloud decision-making about. Very easy to Monday morning quarterback and the "correct" assumptions always become obvious/clear in hindsight but at the time there were good arguments in a lot of directions - all you have to do is look at many users' old pre-election posts to see them. Arguments about future predictions can still be good and well-reasoned even if they don't wind up being "accurate" per se - and I generally think its tacky to speak down on people for thinking something would happen differently than it did after the fact even though they may have had good reason to beforehand.

If, however, their reasoning was impractical/evidence-free/unserious then they should be mercilessly mocked.

I was (as I’m sure you all know by now) a lot more surprised by the NV loss. Any input on what went wrong for Laxalt/any campaign missteps on the Republican side?

Sisolak lost and Republicans won the popular vote for U.S. House 51-47.5, so the votes were clearly there for him (although there was no margin for error). While Laxalt was tainted by Trump, the party's poor national brand, and (perhaps) some of his previous runs for public office, he was effective as an attack dog and had no glaring issues as a candidate like Oz/Walker/Masters. Then again, neither did Budd, and he underperformed rather noticeably as well.

It really wasn’t a race that should have been lost by Republicans. I wish (and in fact already wished back in September) that Laxalt had created/would create some distance between himself and the national party on select issues, but other than that, it seemed like it was mostly external factors (and a far superior D turnout operation) which caused his defeat.

We did a bit of work in Nevada this year - not enough to see a trendline emerge and not close enough to the election to avoid late movers tripping us up, but we never saw the Laxalt advantage that public polls showed but did show the picture most polls painted and that wound up being correct of Masto running slightly ahead of Sisolak (our "final" poll - almost a month out - had Masto up 1 and Sisolak down 4). The difference was almost entirely Hispanic voters, and some non-college white women. Laxalt seemed to have a particularly strong anti-abortion brand (a plurality of voters said they believed him to be "strongly opposed" to abortion being legal) and based on those two subgroups being the ones he underperformed with I'm guessing that's what made the difference.

One of my personal hunches that I have no hard data to back up with is that Sisolak always seemed like an "accidental" governor - he put up an unimpressive showing in a primary in 2018 that didn't attract any genuinely impressive candidates like most Dem primaries in 2018 did and then pretty clearly rode the wave to a similarly underwhelming general election victory against Laxalt himself who clearly has issues statewide here. He didn't seem to have a base and never really cultivated one beyond the state's core Democratic constituencies. That he came as close as he did to being re-elected in a tougher year against a tougher opponent can probably be chalked up to incumbency.