FiveThirtyEight is no more.
March 6, 2025 11:39 AM Subscribe
ABC shuts down FiveThirtyEight and pulls the plug on its website. The last 15 or so employees of the once influential data aggregator are set to be laid off by Disney’s ABC News Group, according to a Tuesday report by the Wall Street Journal. (Archive.ph) The main URL already redirects to ABC News.
In honor of its demise, an all timer tweet from the 2012 election. That election cemented 538's place in the world, so closely predicting what would happen
Drunk Nate Silver is riding the subway, telling strangers the day they will die
https://xcancel.com/DanLevitan/status/266374102334451713
posted by macrael at 11:46 AM on March 6 [15 favorites]
Drunk Nate Silver is riding the subway, telling strangers the day they will die
https://xcancel.com/DanLevitan/status/266374102334451713
posted by macrael at 11:46 AM on March 6 [15 favorites]
Whatever you may think of it, it's an incredibly grim sign of how bad the media industry's situation is when even a major high-name recognition property like 538 can't survive, even with a tiny staff of just 15 employees.
posted by star gentle uterus at 11:55 AM on March 6 [25 favorites]
posted by star gentle uterus at 11:55 AM on March 6 [25 favorites]
Can we go conspiracy theorist on this? ABC is owned by Disney, Vanguard Group and BlackRock are significant shareholders.
posted by subdee at 12:02 PM on March 6 [9 favorites]
posted by subdee at 12:02 PM on March 6 [9 favorites]
Whatever you may think of it, it's an incredibly grim sign of how bad the media industry's situation is when even a major high-name recognition property like 538 can't survive, even with a tiny staff of just 15 employees.
Or worse, that 538 in particular was an easy layoff target because future elections won’t be meaningfully competitive.
posted by jedicus at 12:06 PM on March 6 [44 favorites]
Or worse, that 538 in particular was an easy layoff target because future elections won’t be meaningfully competitive.
posted by jedicus at 12:06 PM on March 6 [44 favorites]
I also see this as a bad sign. 538 was really influential when it came out and was seriously helpful in the run-up to Trump's election. It's not good if you ask me, that a site centered on data and election polling just couldn't survive in the current climate.
posted by julianeon at 12:10 PM on March 6 [3 favorites]
posted by julianeon at 12:10 PM on March 6 [3 favorites]
538 has been on the chopping block for an age, ever since the brand got folded in by ABC. We’re years out from a major election. Seems like an easy cut, really. Rest in peace, you once-beautiful-king.
posted by Going To Maine at 12:10 PM on March 6 [4 favorites]
posted by Going To Maine at 12:10 PM on March 6 [4 favorites]
Sounds like those who were let go found out first via the WSJ.
posted by They sucked his brains out! at 12:10 PM on March 6 [5 favorites]
posted by They sucked his brains out! at 12:10 PM on March 6 [5 favorites]
Is there any example in the last decade of a successful site being acquired by a major media company and surviving in both name and function? At what point will The Athletic be renamed to NYT Sports and Wirecutter to Times Reviews?
posted by thecjm at 12:29 PM on March 6 [12 favorites]
posted by thecjm at 12:29 PM on March 6 [12 favorites]
La Taqueria really is the best burrito, I'll give them that one.
posted by Space Coyote at 12:30 PM on March 6 [9 favorites]
posted by Space Coyote at 12:30 PM on March 6 [9 favorites]
We’re years out from a major election.
This is not true though, and won't be true unless control of Congress stops mattering. Right now the GOP controls both houses. Here are some upcoming House elections. And next year (fingers crossed) is when the midterm elections happen.
(Plus that's just the federal elections. State and local elections might matter even more now.)
posted by trig at 12:38 PM on March 6 [18 favorites]
This is not true though, and won't be true unless control of Congress stops mattering. Right now the GOP controls both houses. Here are some upcoming House elections. And next year (fingers crossed) is when the midterm elections happen.
(Plus that's just the federal elections. State and local elections might matter even more now.)
posted by trig at 12:38 PM on March 6 [18 favorites]
I never used 538 very much except to monitor their election poll averages and lately their presidential approval/disapproval trend lines based on multiple independent polls. The strength of the latter was that they incorporated some weighting related to the reliability factor (past performance, etc.) of each of the polls. So now we are left with the Real Clear Politics poll average, which used most of the same polls but excludes any reliability weightings. Besides 538 and RCP I'm not aware of any other poll aggregators publishing something similar. It would be nice if electoral-vote-com, which does an admirable job every four years aggregating presidential polls at the state level, would assume 538's role in this. (But they are only a couple of guys, who do publish excellent (left-leaning) political news roundups every single day.)
posted by beagle at 12:40 PM on March 6 [6 favorites]
posted by beagle at 12:40 PM on March 6 [6 favorites]
Or worse, that 538 in particular was an easy layoff target because future elections won’t be meaningfully competitive.
We could still use data-driven analysis of court intrigue and which factions are currently in favor with the king.
posted by star gentle uterus at 12:41 PM on March 6 [4 favorites]
We could still use data-driven analysis of court intrigue and which factions are currently in favor with the king.
posted by star gentle uterus at 12:41 PM on March 6 [4 favorites]
Nate Silver has posted a piece on the demise of 538 on his Silver Bulletin site.
posted by tommasz at 12:45 PM on March 6 [5 favorites]
posted by tommasz at 12:45 PM on March 6 [5 favorites]
My guess is that it's a combination of things:
-the brand isn't what it used to be culturally or financially
-cutbacks at ABC in general
-the current POTUS likes to claim the polls show he has high approval ratings; the group that compiles the data proving he's full of shit brings heat that ABC doesn't want
posted by DirtyOldTown at 12:45 PM on March 6 [4 favorites]
-the brand isn't what it used to be culturally or financially
-cutbacks at ABC in general
-the current POTUS likes to claim the polls show he has high approval ratings; the group that compiles the data proving he's full of shit brings heat that ABC doesn't want
posted by DirtyOldTown at 12:45 PM on March 6 [4 favorites]
I'm not sure 538 had the credibility it once did; I feel bad for its employees, but saying this is a bad sign for the future of polling is like saying film criticism is dead if they shutter Rotten Tomatoes.
posted by kittens for breakfast at 12:50 PM on March 6 [26 favorites]
posted by kittens for breakfast at 12:50 PM on March 6 [26 favorites]
Was reading through the comments to see if anyone had posted Silver's take on it yet and tommasz beat me by two minutes. Thank you! He may have a greatly inflated ego, but I felt his take on that situation was spot on.
I do wonder if poll aggregators can really survive in a world of polymarket and predictit. Betting markets have shown themselves to be much closer to the truth the last few election cycles. (this is far from an endorsement and they have many drawbacks not the least of which is an aggressively toxic culture and user base within them).
posted by robot_jesus at 12:51 PM on March 6 [1 favorite]
I do wonder if poll aggregators can really survive in a world of polymarket and predictit. Betting markets have shown themselves to be much closer to the truth the last few election cycles. (this is far from an endorsement and they have many drawbacks not the least of which is an aggressively toxic culture and user base within them).
posted by robot_jesus at 12:51 PM on March 6 [1 favorite]
Bad things at 538:
--While he was there, Nate S himself turned into that stereotype of the aging male projecting his weaknesses and regrets onto those "woke kids".
--538's pollster ratings promised to label partisan polls as such. They were inconsistent on this at best. Many obviously right-wing, or even fly-by-night, right wing outfits got a pass that they didn't deserve.
--Some of the predict-o-matic type widgets could lead people to embarrassing failures. You could argue that there is an explicit or unspoken understanding that all the predictions are forecasting probability, not certainties...but the presentation of the data in the fancier widgets could promote wishcasting if you weren't careful.
Good things at 538:
--Honestly, I did like some of their interfaces. Some of that might have been my personal familiarity--but they worked for me.
--A couple of their pages with historic data were really good, like the page with presidential approval graphs going back to Truman.
--I thought they did a good job at tracking redistricting in the last cycle. That by itself was a lot of work and a nice public service.
A lot of the data will be out there next time around (with the disclaimer other people have added: assuming we still have elections in the future). Public polling is, well, public, and there are lots of aggregators. Ballotpedia and even Wikipedia have a lot of useful overlap with what 538 did, although with more mouseclicks and hunting around to get to it. It's probably not a good thing to lose a site like this that isn't paywalled; a lot of the stuff at traditional pundit mills is subscription-only.
posted by gimonca at 12:51 PM on March 6 [12 favorites]
--While he was there, Nate S himself turned into that stereotype of the aging male projecting his weaknesses and regrets onto those "woke kids".
--538's pollster ratings promised to label partisan polls as such. They were inconsistent on this at best. Many obviously right-wing, or even fly-by-night, right wing outfits got a pass that they didn't deserve.
--Some of the predict-o-matic type widgets could lead people to embarrassing failures. You could argue that there is an explicit or unspoken understanding that all the predictions are forecasting probability, not certainties...but the presentation of the data in the fancier widgets could promote wishcasting if you weren't careful.
Good things at 538:
--Honestly, I did like some of their interfaces. Some of that might have been my personal familiarity--but they worked for me.
--A couple of their pages with historic data were really good, like the page with presidential approval graphs going back to Truman.
--I thought they did a good job at tracking redistricting in the last cycle. That by itself was a lot of work and a nice public service.
A lot of the data will be out there next time around (with the disclaimer other people have added: assuming we still have elections in the future). Public polling is, well, public, and there are lots of aggregators. Ballotpedia and even Wikipedia have a lot of useful overlap with what 538 did, although with more mouseclicks and hunting around to get to it. It's probably not a good thing to lose a site like this that isn't paywalled; a lot of the stuff at traditional pundit mills is subscription-only.
posted by gimonca at 12:51 PM on March 6 [12 favorites]
It's an exaggeration and I also think media cuts are bad but, seems like the bed they made.
posted by latkes at 1:10 PM on March 6 [3 favorites]
posted by latkes at 1:10 PM on March 6 [3 favorites]
Looks like Clare Malone is still listed as a writer for The New Yorker. The podcast during the late 10s was generally pretty good.
-the brand isn't what it used to be culturally or financially
-cutbacks at ABC in general
-the current POTUS likes to claim the polls show he has high approval ratings; the group that compiles the data proving he's full of shit brings heat that ABC doesn't want
Yeah, all that. With the wholesale capitulation of Corporate Everything as the backdrop this should be completely unsurprising. There's basically no reason for ABC to have kept it. If it wasn't a profit center, and it was poking a certain fragile ego in the eye? What, you want provocative journalism? From ABC? In 2025? That's directly at odds with what ABC wants, which is shareholder value, zero risk and literally nothing else.
posted by howbigisthistextfield at 1:33 PM on March 6 [8 favorites]
-the brand isn't what it used to be culturally or financially
-cutbacks at ABC in general
-the current POTUS likes to claim the polls show he has high approval ratings; the group that compiles the data proving he's full of shit brings heat that ABC doesn't want
Yeah, all that. With the wholesale capitulation of Corporate Everything as the backdrop this should be completely unsurprising. There's basically no reason for ABC to have kept it. If it wasn't a profit center, and it was poking a certain fragile ego in the eye? What, you want provocative journalism? From ABC? In 2025? That's directly at odds with what ABC wants, which is shareholder value, zero risk and literally nothing else.
posted by howbigisthistextfield at 1:33 PM on March 6 [8 favorites]
There is a separate and important discussion about whether Republicans will allow free and fair elections going forwards, and what place those people think legitimate polling has (if any). Our democracy is on its last legs giving the Trump cult power again, certainly.
But what has 538's record at prediction been the last two cycles? Nate Silver and 538 have been ridiculed for getting Trump and MAGA very wrong since 2016 — and rightly so. If they don't have a good reputation and don't drive traffic that enhances ad revenue, then maybe it was a business decision by ABC, above all else.
On the other hand, Disney (the parent company of ABC) has long been friendly to the people that Trump and his voters are now targeting for discrimination — and worse — namely trans, gay, and non-white Americans. Pixar recently adjusted storylines in Disney+ programming to remove a trans character and replace them with Christians. Maybe there is an internal shift in corporate to appease or placate Trump/Project 2025?
It would be very useful to the public to hear more reporting about what is going on internally at the company, in that its apparent realignment may be a bellwether for how quickly the state is falling into fascism.
posted by They sucked his brains out! at 1:58 PM on March 6 [4 favorites]
But what has 538's record at prediction been the last two cycles? Nate Silver and 538 have been ridiculed for getting Trump and MAGA very wrong since 2016 — and rightly so. If they don't have a good reputation and don't drive traffic that enhances ad revenue, then maybe it was a business decision by ABC, above all else.
On the other hand, Disney (the parent company of ABC) has long been friendly to the people that Trump and his voters are now targeting for discrimination — and worse — namely trans, gay, and non-white Americans. Pixar recently adjusted storylines in Disney+ programming to remove a trans character and replace them with Christians. Maybe there is an internal shift in corporate to appease or placate Trump/Project 2025?
It would be very useful to the public to hear more reporting about what is going on internally at the company, in that its apparent realignment may be a bellwether for how quickly the state is falling into fascism.
posted by They sucked his brains out! at 1:58 PM on March 6 [4 favorites]
La Taqueria really is the best burrito, I'll give them that one.
100%, that was the first place in SF that gave me a burrito without rice, and it changed my life.
posted by grumpybear69 at 2:00 PM on March 6 [4 favorites]
100%, that was the first place in SF that gave me a burrito without rice, and it changed my life.
posted by grumpybear69 at 2:00 PM on March 6 [4 favorites]
With its diversified institutional ownership, in a couple of years I think Disney/ABC will be the second-most progressive of the legacy broadcast/cable properties. MSBNC will stay on brand of course at the left-most, but the billionaires who control / will control NBC News, CBS News and CNN are going to move that product the same direction Bezos has moved the Washington Post.
posted by MattD at 2:18 PM on March 6 [1 favorite]
posted by MattD at 2:18 PM on March 6 [1 favorite]
538's data/modeling stuff was interesting (and I hope someone else picks up that mantle), but their political journalism was just replacement-level, and the fit between the two sides always struck me as awkward.
posted by kickingtheground at 2:26 PM on March 6 [1 favorite]
posted by kickingtheground at 2:26 PM on March 6 [1 favorite]
Nate Silver and 538 have been ridiculed for getting Trump and MAGA very wrong since 2016 — and rightly so
I honestly think this is a false narrative. Leading up to 2016 538 gave Trump like a 26% chance of winning, which they were roundly mocked for being way way way too high. 538 were the ones that said it was not at all in the bag for Clinton.
I feel like people conflate this kind of data science, which is really just poll aggregation with some selective weighting, as something more than it is. I think it's been pretty useful as a data exploration tool, but frequently pilloried for not being Nostradamus.
posted by mcstayinskool at 2:34 PM on March 6 [14 favorites]
I honestly think this is a false narrative. Leading up to 2016 538 gave Trump like a 26% chance of winning, which they were roundly mocked for being way way way too high. 538 were the ones that said it was not at all in the bag for Clinton.
I feel like people conflate this kind of data science, which is really just poll aggregation with some selective weighting, as something more than it is. I think it's been pretty useful as a data exploration tool, but frequently pilloried for not being Nostradamus.
posted by mcstayinskool at 2:34 PM on March 6 [14 favorites]
Seems like any of the major foundations working on civic health and democracy could easily pick up the valuable functions of 539—the poll-tracking, weighting, and analysis—without the need to constantly post traffic-generating stories or provide space for the boring opinions of a middle-aged gambling addict.
posted by Just the one swan, actually at 2:59 PM on March 6 [1 favorite]
posted by Just the one swan, actually at 2:59 PM on March 6 [1 favorite]
[...saying this is a bad sign for the future of polling is like saying film criticism is dead if they shutter Rotten Tomatoes...]
Is it though?
Rotten Tomatoes is and always has been a simple aggregator of reviews. 538 is and always has been an analytical site focused on weighted aggregation of data, PLUS its own forecasting and analysis.
posted by yellowcandy at 3:17 PM on March 6 [2 favorites]
Is it though?
Rotten Tomatoes is and always has been a simple aggregator of reviews. 538 is and always has been an analytical site focused on weighted aggregation of data, PLUS its own forecasting and analysis.
posted by yellowcandy at 3:17 PM on March 6 [2 favorites]
Yeah, I think it’s worth noting that the personal stock of Silver went way down late last year, as it came out he’d been betting huge amounts of money. And then put out a book saying, actually, betting is great. If this is the stuff he’s saying/doing in public, he might be pretty difficult to deal with.
posted by The River Ivel at 3:19 PM on March 6 [4 favorites]
posted by The River Ivel at 3:19 PM on March 6 [4 favorites]
will allow free and fair elections going forwards
Oh, no. They absolutely will allow "free and fair elections" going forward. But after cutting all the security around elections and giving the ol' wink-wink-nudge-nudge-say-no-more those elections are going to be subject to a miasma of bullshit that has to be untangled. The results of which are... well... it's only 2025. I don't want to shit on the robin's egg of hope some people might still have for 2026.
But, in line with the techdirt article from earlier... even if those elections have armed trump brownshirts wearing I'm A Trump Goon silk-screened on their brown shirts while they "monitor" the election they will very explicitly be called Free And Fair Elections. Because that's what they do. They co-opt language. Free and Fair as in Free Speech.
tl;dr - Go ahead and stop saying Free And Fair Elections. Stop giving that phrase legitimacy. Because in twelve months' time, it'll be subverted just like Free Speech. Start figuring out the non-fascist "stop the steal" branding, language and memes now. That'll be necessary, and more effective.
posted by howbigisthistextfield at 3:24 PM on March 6 [3 favorites]
Oh, no. They absolutely will allow "free and fair elections" going forward. But after cutting all the security around elections and giving the ol' wink-wink-nudge-nudge-say-no-more those elections are going to be subject to a miasma of bullshit that has to be untangled. The results of which are... well... it's only 2025. I don't want to shit on the robin's egg of hope some people might still have for 2026.
But, in line with the techdirt article from earlier... even if those elections have armed trump brownshirts wearing I'm A Trump Goon silk-screened on their brown shirts while they "monitor" the election they will very explicitly be called Free And Fair Elections. Because that's what they do. They co-opt language. Free and Fair as in Free Speech.
tl;dr - Go ahead and stop saying Free And Fair Elections. Stop giving that phrase legitimacy. Because in twelve months' time, it'll be subverted just like Free Speech. Start figuring out the non-fascist "stop the steal" branding, language and memes now. That'll be necessary, and more effective.
posted by howbigisthistextfield at 3:24 PM on March 6 [3 favorites]
Didn't this guy give Clinton an 80 percent chance of winning? Or is there some reason I've forgotten which makes me instantly think, Fuck that guy!, when I hear his name?
Anyway, Fuck that guy.
posted by dobbs at 3:25 PM on March 6 [3 favorites]
Anyway, Fuck that guy.
posted by dobbs at 3:25 PM on March 6 [3 favorites]
Rotten Tomatoes is and always has been a simple aggregator of reviews. 538 is and always has been an analytical site focused on weighted aggregation of data, PLUS its own forecasting and analysis.
They're not the only forecaster in town, and I don't think anyone has taken them that seriously in a while, for all the reasons stated above.
posted by kittens for breakfast at 3:51 PM on March 6
They're not the only forecaster in town, and I don't think anyone has taken them that seriously in a while, for all the reasons stated above.
posted by kittens for breakfast at 3:51 PM on March 6
Didn't this guy give Clinton an 80 percent chance of winning?
71.4% for however long that 538 link stays up. They revised their forecast language for 2020 and 2024 to things like:
Harris wins 50 times out of 100
Trump wins 49 times out of 100.
Based on "we simulated the election X times, and Y won Z times." They really wanted to get away from "71% likely to win!" headlines.
As far back as 2012 Silver pointed out that Obama was within a standard polling error of losing, though I can't find a 538 link now to where he said that. And so was dipshit in 2016. In an odd way, they weren't really wrong in either case.
I think pretty much since then those models with a +/- X% margin of error have been off X% in one direction or the other. Which makes selling a horse race really, really hard to do. If your 89% and 49% "predictions" get the same electoral results back-to-back, what's the value in your predictions?
posted by howbigisthistextfield at 3:52 PM on March 6 [6 favorites]
71.4% for however long that 538 link stays up. They revised their forecast language for 2020 and 2024 to things like:
Harris wins 50 times out of 100
Trump wins 49 times out of 100.
Based on "we simulated the election X times, and Y won Z times." They really wanted to get away from "71% likely to win!" headlines.
As far back as 2012 Silver pointed out that Obama was within a standard polling error of losing, though I can't find a 538 link now to where he said that. And so was dipshit in 2016. In an odd way, they weren't really wrong in either case.
I think pretty much since then those models with a +/- X% margin of error have been off X% in one direction or the other. Which makes selling a horse race really, really hard to do. If your 89% and 49% "predictions" get the same electoral results back-to-back, what's the value in your predictions?
posted by howbigisthistextfield at 3:52 PM on March 6 [6 favorites]
In a better world its demise would have been a side effect if the US finally abandoning the electoral college. But alas, we live in the darkest timeline.
posted by Pemdas at 3:54 PM on March 6 [8 favorites]
posted by Pemdas at 3:54 PM on March 6 [8 favorites]
I always harbored concern that 538 was bad for the discourse and the information environment. Like what is the value of an extremely pro-status quo, anti-earnest-political-sentiment sports betting addict treating US politics like it is baseball statistics? None of it was really journalism, and journalism would have maybe better prepared us to know--and even influence--what was going on in the world of politics, in my view. Rather than just checking Nate Silver's ticker every day after looking at your 401k to calm or excite your nerves and then head to the water cooler or whatever.
posted by kensington314 at 4:23 PM on March 6 [14 favorites]
posted by kensington314 at 4:23 PM on March 6 [14 favorites]
100%, that was the first place in SF that gave me a burrito without rice, and it changed my life.
That's almost every kind of burrito at every taco shop in San Diego.
Lousy timing with shutting down, since the NCAA basketball tournaments are coming up and I think 538 typically had a huge influx of traffic after the brackets come out. Granted, that's kind of the filter I see them initially joining Disney/ABC through, but I thought their initial merger was largely based on sports stuff like the tournament for the benefit of ESPN, similar to how Grantland (basically the Ringer now) was originally a ESPN spinoff and did long-ish form sports writing but also a lot of pop culture stuff. But ESPN has been shrinking and putting more and more stuff behind paywalls for quite some time now. I've recognized one of 538's old sports guys showing up with bylines on ESPN.com lately.
posted by LionIndex at 4:27 PM on March 6 [1 favorite]
That's almost every kind of burrito at every taco shop in San Diego.
Lousy timing with shutting down, since the NCAA basketball tournaments are coming up and I think 538 typically had a huge influx of traffic after the brackets come out. Granted, that's kind of the filter I see them initially joining Disney/ABC through, but I thought their initial merger was largely based on sports stuff like the tournament for the benefit of ESPN, similar to how Grantland (basically the Ringer now) was originally a ESPN spinoff and did long-ish form sports writing but also a lot of pop culture stuff. But ESPN has been shrinking and putting more and more stuff behind paywalls for quite some time now. I've recognized one of 538's old sports guys showing up with bylines on ESPN.com lately.
posted by LionIndex at 4:27 PM on March 6 [1 favorite]
It seems like forever and a day that Nate was a statistical hero...
I long ago lost faith in "Polls" as they mutated into not much more than news rating devices devoid of external validity...
"There once was a statistician who drowned in a lake with an average depth of 2 feet."
posted by Fupped Duck at 4:59 PM on March 6
I long ago lost faith in "Polls" as they mutated into not much more than news rating devices devoid of external validity...
"There once was a statistician who drowned in a lake with an average depth of 2 feet."
posted by Fupped Duck at 4:59 PM on March 6
It seems like whenever a shuttering like this happens there’s absolutely no time wasted in pulling the plug on the website and the years of work archived there. Surely it’s pittance for a company the size of ABC News Group (not to mention the parent Disney) to keep the servers running for a little while longer.
In this particular case you’d think a news organization that still has some respect for the practice of journalism (and despite how cynical I’m inclined to be, I know they still do!) would grant a little of that respect to the writers who’ve published their stories on 538 over the past decade and a half.
posted by theory at 5:22 PM on March 6
In this particular case you’d think a news organization that still has some respect for the practice of journalism (and despite how cynical I’m inclined to be, I know they still do!) would grant a little of that respect to the writers who’ve published their stories on 538 over the past decade and a half.
posted by theory at 5:22 PM on March 6
This is about as surprising as someone shutting down a Siskel and Ebert site or finding out that ticket sales are weak for some 70s rock band featuring only members who joined after 2000.
posted by snofoam at 5:26 PM on March 6 [4 favorites]
posted by snofoam at 5:26 PM on March 6 [4 favorites]
It feels like another nail in the coffin for the interactive visualization and data journalism boom of the early 2010s.
In retrospect a lot of the media ecosystem of that era—FiveThirtyEight, BuzzFeed, The Awl, Gawker—was really geared toward white collar workers goofing around on their work computers before everyone had Instagram on their phones. But drill-down-into-the-data widgets and other "interactive" features like those fancy articles that interlace videos, graphics, and text just don't work as well on smartphones as they did with mice and big monitors.
And as Silver mentions, maintaining databases and keeping them up to date takes a fair bit of work, and so does keeping interactive elements working as the internet and web browser standards evolve. I'm not sure if this affects FiveThirtyEight as much, but a lot of data journalism of that era was also effectively venture capital-subsidized, because visualization software providers (including Google Maps) would offer discounted or even free plans to newsrooms. Many of these have dried up over the years.
Not to mention the world is a lot less simple than it seemed to a lot of us in the Obama era. I was just looking back through the FiveThirtyEight burrito bracket stories, and it seems like a relic of a simpler time, when you could report on food from around the country and not have to talk about politics or inflation or egg shortages or trade policy.
posted by smelendez at 5:31 PM on March 6 [11 favorites]
In retrospect a lot of the media ecosystem of that era—FiveThirtyEight, BuzzFeed, The Awl, Gawker—was really geared toward white collar workers goofing around on their work computers before everyone had Instagram on their phones. But drill-down-into-the-data widgets and other "interactive" features like those fancy articles that interlace videos, graphics, and text just don't work as well on smartphones as they did with mice and big monitors.
And as Silver mentions, maintaining databases and keeping them up to date takes a fair bit of work, and so does keeping interactive elements working as the internet and web browser standards evolve. I'm not sure if this affects FiveThirtyEight as much, but a lot of data journalism of that era was also effectively venture capital-subsidized, because visualization software providers (including Google Maps) would offer discounted or even free plans to newsrooms. Many of these have dried up over the years.
Not to mention the world is a lot less simple than it seemed to a lot of us in the Obama era. I was just looking back through the FiveThirtyEight burrito bracket stories, and it seems like a relic of a simpler time, when you could report on food from around the country and not have to talk about politics or inflation or egg shortages or trade policy.
posted by smelendez at 5:31 PM on March 6 [11 favorites]
Nate Silver looked increasingly out of touch during the early pandemic and accompanying social-justice movements. He succumbed to that classic tendency to try to be a pundit about topics where he simply has no expertise and can only go by opinion. Links from a Twitter thread I kept up about this for a while a few years ago.
He had an interesting and novel shtick back in the mid-2000s, but that time has clearly passed. I'm not at all surprised they're moving on from his brand.
posted by limeonaire at 5:46 PM on March 6 [5 favorites]
He had an interesting and novel shtick back in the mid-2000s, but that time has clearly passed. I'm not at all surprised they're moving on from his brand.
posted by limeonaire at 5:46 PM on March 6 [5 favorites]
To be fair, analysing elections is not going to be a growth industry in the US.
posted by pompomtom at 7:00 PM on March 6 [4 favorites]
posted by pompomtom at 7:00 PM on March 6 [4 favorites]
2008 feel like ancient history.
It was good for a while but it turns out politics isn't just like baseball.
posted by zenon at 8:09 PM on March 6 [1 favorite]
It was good for a while but it turns out politics isn't just like baseball.
posted by zenon at 8:09 PM on March 6 [1 favorite]
Can we go conspiracy theorist on this? ABC is owned by Disney, Vanguard Group and BlackRock are significant shareholders.
Vanguard and Blackrock are significant shareholders in every major public company because the standard way for Americans to invest in stocks long term is now to buy them all at once via funds managed by Blackrock and Vanguard.
That’s also a bunch of extra steps to put together a working “theory” when “media companies are scared of Trump” is right there.
posted by atoxyl at 8:27 PM on March 6 [4 favorites]
Vanguard and Blackrock are significant shareholders in every major public company because the standard way for Americans to invest in stocks long term is now to buy them all at once via funds managed by Blackrock and Vanguard.
That’s also a bunch of extra steps to put together a working “theory” when “media companies are scared of Trump” is right there.
posted by atoxyl at 8:27 PM on March 6 [4 favorites]
People are taking the requisite shots at late period Nate Silver but I think the truth is that 538 did lose a lot of its identity, and probably of its audience, with him departing and continuing to do the marquee prediction bit on his own. New upstarts got into that game, too, and meanwhile didn’t the 538 folks take their whole model offline and replace it mid-campaign? They were still very nice for stuff like poll averages and I hate to see them go but I never really felt like they found their footing after the split.
posted by atoxyl at 8:42 PM on March 6
posted by atoxyl at 8:42 PM on March 6
The age of Nate Silver is over. The time of Nate Cohn has begun!
https://x.com/Nate_Cohn/status/1225298054049357826
posted by weard_beard at 9:26 PM on March 6
https://x.com/Nate_Cohn/status/1225298054049357826
posted by weard_beard at 9:26 PM on March 6
> 538's data/modeling stuff was interesting (and I hope someone else picks up that mantle)
Nate was still running his US election model through via silver bulletin newsletter
the Economist runs an election model developed with Andrew Gelman's group
posted by are-coral-made at 12:10 AM on March 7
Nate was still running his US election model through via silver bulletin newsletter
the Economist runs an election model developed with Andrew Gelman's group
posted by are-coral-made at 12:10 AM on March 7
the Split Ticket crew also do the election modeling thing and some other interesting analyses
posted by atoxyl at 12:27 AM on March 7
posted by atoxyl at 12:27 AM on March 7
May I point to a good alternative: https://electoral-vote.com/
posted by DreamerFi at 2:45 AM on March 7 [1 favorite]
posted by DreamerFi at 2:45 AM on March 7 [1 favorite]
Didn't this guy give Clinton an 80 percent chance of winning?
The common discourse of the day in 2016 was that Trump had 0% chance of winning. I remember a New York Times prediction widget that led you through a decision tree to figure out which candidate would win, the conclusion you'd get from that app was "all roads lead to Hillary". If Silver gave Trump a 20 to 25 percent chance, that's 20 to 25 percent more than a lot of other people were offering.
posted by gimonca at 4:30 AM on March 7 [3 favorites]
The common discourse of the day in 2016 was that Trump had 0% chance of winning. I remember a New York Times prediction widget that led you through a decision tree to figure out which candidate would win, the conclusion you'd get from that app was "all roads lead to Hillary". If Silver gave Trump a 20 to 25 percent chance, that's 20 to 25 percent more than a lot of other people were offering.
posted by gimonca at 4:30 AM on March 7 [3 favorites]
Nate Silver looked increasingly out of touch during the early pandemic and accompanying social-justice movements.
posted by RonButNotStupid at 6:19 AM on March 7 [9 favorites]
Silver responded that, "If Kamala Harris were to say I want to ban poker, I would feel targeted as a poker player, and clearly this is a community that wouldn't be interested in the votes of people like me, so why would I want to be in their coalition?""But you wouldn't then go and vote for Trump?" Konnikova responded.I don't think Silver's necessarily wrong about Trump voters here, but identifying with them in this manner was absolutely a dick move.
"Yes, I would," Silver said.
Konnikova and Silver had previously been discussing the issue of voting in your own self- interest. Silver said that it is in some people's "narrow self-interest to vote for Trump."
Konnikova responded, and referenced a small portion of Trump voters, who she described as "Those cynical, piece of s*** human beings, who vote for Trump, just because of cynical self interest and don't give a f*** what happens to the rest of the world and the rest of the country."
"I just have no sympathy for people who don't care if the world burns, as long as they, personally, have money."
"I disagree with that characterization," Silver responded. "I don't think it's a bad thing if people vote in terms of their narrow self-interest," he said. "That's a part of a political economy that can be healthy."
"Trump understands that you do want to appeal to people's sense of identity," he said.
posted by RonButNotStupid at 6:19 AM on March 7 [9 favorites]
Is there any example in the last decade of a successful site being acquired by a major media company and surviving in both name and function?
This is what happens when you sell out. The media ghouls who own your ass now have you by the balls and can shutter the whole operation at their whim, on Day 1 or on Day N.
I'm sick of Planet Sell-Out, someone get me off this rock. Everything decent sells out and turns to shit. Everyone has a price, it seems. Ugh. Would that it were not so. Would that someone would have integrity and want to keep their special thing that they worked hard on that shows their unique vision and that is successful, and *not* try to cash in, for fucking once.
Too many examples to list.
posted by cats are weird at 9:27 AM on March 7
This is what happens when you sell out. The media ghouls who own your ass now have you by the balls and can shutter the whole operation at their whim, on Day 1 or on Day N.
I'm sick of Planet Sell-Out, someone get me off this rock. Everything decent sells out and turns to shit. Everyone has a price, it seems. Ugh. Would that it were not so. Would that someone would have integrity and want to keep their special thing that they worked hard on that shows their unique vision and that is successful, and *not* try to cash in, for fucking once.
Too many examples to list.
posted by cats are weird at 9:27 AM on March 7
I think it's been pretty useful as a data exploration tool, but frequently pilloried for not being Nostradamus.
posted by mcstayinskool
100% agree. The US public is not the.most literate when it comes to mathematics.
Unfortunately, now that there are Billions in Venture Capital sunk into the premise that "aggregation with weighting actually is Nostradamus, just give us 5 years", people are going to be less likely to understand the value of polls.
posted by eustatic at 10:13 AM on March 7
posted by mcstayinskool
100% agree. The US public is not the.most literate when it comes to mathematics.
Unfortunately, now that there are Billions in Venture Capital sunk into the premise that "aggregation with weighting actually is Nostradamus, just give us 5 years", people are going to be less likely to understand the value of polls.
posted by eustatic at 10:13 AM on March 7
May I point to a good alternative: https://electoral-vote.com/
I followed Tanenbaum's (yes, CS nerds: that one!) electoral-vote.com closely in the 2004 election, and 2006 mid-terms.
I recall him getting it pretty wrong in 2004, and don't recall much of the mid-terms. According to wiki he did much better? By 2008 I think Silver had basically stolen the show.
posted by howbigisthistextfield at 10:24 AM on March 7
I followed Tanenbaum's (yes, CS nerds: that one!) electoral-vote.com closely in the 2004 election, and 2006 mid-terms.
I recall him getting it pretty wrong in 2004, and don't recall much of the mid-terms. According to wiki he did much better? By 2008 I think Silver had basically stolen the show.
posted by howbigisthistextfield at 10:24 AM on March 7
100% agree. The US public is not the.most literate when it comes to mathematics.
Oh come on. For discrete events like Presidential elections, then 80/20 is meaningless. It's not the public's fault they present these as legitimate predictions, and not mathematical abstractions. If they meant them to be mathematical abstractions, they'd keep them to themselves and only academics interested in the data would use them.
posted by The_Vegetables at 11:15 AM on March 7
Oh come on. For discrete events like Presidential elections, then 80/20 is meaningless. It's not the public's fault they present these as legitimate predictions, and not mathematical abstractions. If they meant them to be mathematical abstractions, they'd keep them to themselves and only academics interested in the data would use them.
posted by The_Vegetables at 11:15 AM on March 7
This is probably just a semantic point but “discrete event” tends to mean something when you’re talking about probability that does not imply that assigning probability is meaningless. What elections are is infrequent events, so it’s kind of hard to claim convincingly that your model is well validated.
posted by atoxyl at 1:40 PM on March 7 [1 favorite]
posted by atoxyl at 1:40 PM on March 7 [1 favorite]
« Older The Smell of Success | Their brief moment in the sun foretells our own... Newer »
posted by torokunai2 at 11:41 AM on March 6 [4 favorites]