FiveThirtyEight Gets the Ax as ABC News Slashes Jobs in Major Shakeup
mediaite.com
mediaite.com
I can only hope this finds another home.
He's hardly the only center left writer who dislikes the far left these days though, quite a few are furious at progressives for tanking the Democratic brand and throwing a winnable election to the current administration. Nate has still been far more vocal against Trump and all sorts of right wing nonsense.
This is such a perfect summary of contemporary American society with regards to politics.
I heard one person argue that as the far left progressives endlessly pushed "purity tests" against more extreme positions and became intolerant of any other opinions, it actually fed into why people tolerated or were willing to forgive bad things from the MAGA crowd. Many people felt like the Republicans accepted diverse viewpoints within the party, and people who'd made mistakes in the past, but the Democrats would cancel and exile anyone who wasn't perfectly woke.
It's not to excuse people who voted for Trump, but it does help explain the election results.
The idea Nate Silver took a far right turn is laughable. Saying the democratic candidate is going to lose or that some D positions are clearly unpopular doesn’t make one hard right, just pointing out what the numbers show.
Saying the Dodgers have good chance of winning World Series doesn’t make one a Dodgers fan. Or saying most people like iPhones make one an Apple fan.
Link to such tweets? I've gone through two pages of his most recent tweets and couldn't find anything obviously matches "argues on Twitter in favor of whatever Peter Thiel is doing, even as Peter Thiel endorses a completely illegal coup aimed at destroying the USA". The closest I could find were
"Don't neglect how tedious this platform is becoming. Democrats "won" the battle of being the most annoying party for several years running, but their grip on the title will probably be broken soon."
and a series of tweets talking about how smart Musk is.
Is that what you meant by "very far to the right"? Is not being infinitely rabid against Trump/Vance/Musk count as "very far to the right" now?
Yes, political discussion has ceased to be rational and has degraded into merely emotional outbursts today. I'm trying really hard to stay away from any political discussion. They're all very predictable and I feel like engaging in them is making me dumber. I see a political post with 500 comments, I don't even have to read it to know the main arguments.
It amazes me that in this day and age people still (500 as your example states is actually quite low for some of the political shit on HN lately) continue to "argue" and discuss politics. It is 2025, literally everyone is in their own echochamber.
minds of people that still use "X" are basically being ran by state-sponsored media (amazing no one realizes this, would wonder what would happen if say Hillary bought X tomorrow...lol), minds of other people are being ran by whatever other place they do their own research - in 2025 no one will ever be able to "convince" someone else about anything or "open their eyes" to what they are being subjected to - it is suuuuuuch a waste of everyone's time to discuss politics. then you have droves of people complaining that many posts get flagged - the flaggers I believe are just people trying to save 500+ people from wasting their precious time :)
Silver hasn't been on the pod in a long time so you can't say his politics, whatever they are, have anything to do with the pod.
Since Silver took his models with him, and there wasn’t that much time after he left, there wasn't much track record to judge the post-Silver 538, but I think that it remained a first-rate poll aggregator and from all appearances a reasonable poll-based-predictor.
Based on that you'd expect The Economist to be center right but when people try to compare it with other news sources it is usually rated left-of-center [1] which seems odd to me.
On the other hand, the one thing the center-left and center-right have in common is "the center cannot hold", look at how centrist parties are eroding in places like Germany where the usual government coalition now is the CDU (center right) and the SPD (center left) which steadily loses ground to the far right because, even though the center might have its facts straight, it is emotionally false for most people.
The Economist is almost the archetype of economic center-rightism, but IIRC it tends not to fall even slightly on the right-wing side of social/cultural issues (trans rights excepted, which I’ve seen suggested is largely connected to one particular member of the editorial board), which is pretty unsurprising because right-wing social policy is a state-imposed inefficiency to the pure economic neoliberal. In societies where moderately right stances on social issues are common even on the economic left, when you try to construct a single social/economic left-right axis, I can see this making The Economist seem center-left on that composite axis.
https://www.theatlantic.com/ideas/archive/2024/01/media-layo...
https://projects.fivethirtyeight.com/polls/approval/donald-t...
This was probably something internal to ABC, because if it came from Burbank it would have been in the works for two quarters and rumors would have been published in the media for the same amount of time.
I can't find any way to read this other than that you think that Kamala was guaranteed to win, and are still angry that someone told you otherwise?
What you have to do is simulate all the states and DC and run a monte carlo simulation and you will always get an equivocal answer if you do that.
Which is why any model worth its salt only uses national polls to predict state results. (State polls are more meaningful. But the headlines they generate aren't as nationally clickable.)
There aren’t that may states. You don’t need partisan polling in Vermont and Wyoming, for example.
> you have to infer state-level results based on something other than state-level polls
Yes. As I said. That is what you use national polls. “But the electoral college” isn’t a valid argument against election models. The reality is paucity of granular data and complexity of predicting swing states.
You may be thinking of Nate Silver's Silver Bulletin [1]. He took the original 538 model with him when he left. The model 538 runs now is something they made up afterwards.
Of course, this isn't a linear game where scores are alternating, so my line of thought isn't completely fair.
538 almost never got the polls wrong, they simply reported the results the pollsters got. But you probably meant something else was wrong, and not their reporting of the polls. Like, maybe, the poll-based predictions that the polls supported.
But their poll-based predictions of individual results across all of the individual elections they predicted tended to be reasonably accurate in the Silver era, in that results predicted as, say, 70% chances were actually correct about 70% of the time, etc. In the post-Silver Morris period, I haven't seen any after-the-fact analysis, but the universe is a lot smaller, so you'd also expect the actual results on that limited universe to be less accurate in representing the long-term accuracy of the modelling approach.
The main problem, though, is that people want predictions to have probability 1 or 0, and don't deal well with things in between.