Editor-in-Chief, @FiveThirtyEight. Author, The Signal and the Noise (amzn.to/QdyFYV). Sports/politics/food geek.
RT @NateSilver538: Where's the evidence that Trump's digital ads were very effective? Most of it seems to be a lot of bragging by Brad Pars…
RT @SophiaArmen: all I see is hummus https://twitter.com/reuters/status/1185298138053185537
Implicit in what Kevin's here: these stories are often too credulous. They purport to take you (sort of literally) inside the machine, which is an inherently appealing premise. But they tend to take claims made by campaign operatives too much at face value.
This is a great list of questions to ask when you read stories on campaign technology. https://twitter.com/kwcollins/status/1186033813417279488
A decent hypothesis is that it's about as effective as traditional advertising per dollar spent. I don't know wtf "I thought you were a data guy" means but l I'd like to have seen one piece of evidence in that story, empirical or otherwise, that Trump's FB ads had a big impact. https://twitter.com/MikeIsaac/status/1185948721504292864
Trump spent $44 million on Facebook ads (and Clinton spent $28 million) in 2016. That's small in the context of a presidential campaign! By comparison, Trump received $2.6 *billion* worth of earned online media, and about $5b worth of earned media overall. https://bit.ly/2mECmcg
Still no data on your assumption about Facebook's algorithmic impact.
In 2018 Republicans lost 40 House seats because, while turnout among the GOP base was high, it was just as high among the D base and the D base is larger, plus independents swung strongly to Democrats. It nicely illustrates the shortcomings of Trump's base-focused strategy. https://twitter.com/RachelBitecofer/status/1185946318306512897
How Facebook affects what news people consume is almost certainly more important to election campaigns, probably by an order of magnitude, than anything having to do with advertising on Facebook.
You realize campaigns can pay to place stories in people's feeds, right?
I do, and I'd expect that they represent a small proportion of all the stories that people read. In fact, that's close to an axiom in elections that receive extensive media coverage, such as presidential campaigns: earned media is vastly more important than paid media.
What's also weird is the same tech-savvy writers who would happily acknowledge the limitations of digital ads in any other context (i.e. they're only marginally effective despite the ability to target customers, etc.) uncritically assume they're incredibly effective for politics. https://twitter.com/NateSilver538/status/1185947922933993473
you sure do like talking to yourself nate not sure why i respond to the "dipshit quote dunk twitter method" anymore
RT @NateSilver538: Where's the evidence that Trump's digital ads were very effective? Most of it seems to be a lot of bragging by Brad Pars…
Trump won independents by 16 points in Michigan, 11 points in Wisconsin, 7 points in Pennsylvania, 4 points in Florida. Those swing voters were super important! If Clinton splits indies with Trump in those states, she somewhat comfortably wins the Electoral College. https://twitter.com/davidshor/status/1185959484809723911
RT @brendanNyhan: @kwcollins @mattGrossmann @jbview @NateSilver538 There's no magic. Remember when Karl Rove was an unbeatable political ge…
Far more likely is that they've studied this stuff with *some* empirical rigor, rather than buying into the largely evidence-free media mythology about the effectiveness of digital campaign ads.
We're supposed to believe that the Trump campaign is a bunch of super-geniuses for spending extensively on Facebook ads, whereas everyone who works for one of the 19 Democratic campaigns is an idiot?
No Nate, they're not idiots, they're spending their money on trying to win the Democratic primary, as they should be. But that gives Trump an early advantage that third party groups on the Dem side should help mitigate.
It's not baffling at all! It has an incredibly simple explanation, in fact, although one that implicitly undercuts the article's thesis: There's nothing magic about digital ad spending. https://twitter.com/MikeIsaac/status/1185944131731935232
so your point is that facebook ad spend is a waste? i thought you were a data guy
RT @RachelBitecofer: @NateSilver538 There hasn't been a single district yet in which Ds managed to out perform Rs on turnout. And I argue t…
Where's the evidence that Trump's digital ads were very effective? Most of it seems to be a lot of bragging by Brad Parscale, amplified by media that would like you to ignore more obvious explanations for how Trump won, such as obsessive media coverage of Clinton's emails. https://twitter.com/jonfavs/status/1185934416117293059
Plus, shouldn't one of the lessons of 2016 be that spending more money on advertising *doesn't* make a lot of difference in a presidential race? (Hillary spent 2x as much on ads as Trump.) https://twitter.com/daveweigel/status/1185884524330323968
Money isn't definitive. George Nethercutt was outspent 10-1. When the tsunami hits nothing can save you
DJ (Derek Jeter) LeMahieu https://twitter.com/Travis_Sawchik/status/1185256289632477185
Anyone who goes with the "Hillary Clinton's gonna jump into the race and win!" theory doesn't really understand how Democrats think, nor are they looking at the relevant data that shows how Democratic voters are extremely averse to her running again. https://bit.ly/2P1NjWD https://twitter.com/JenniferJJacobs/status/1185632670132375552
Seems fair for the media to ask more questions about the health of a guy who just had a heart attack. https://twitter.com/rubycramer/status/1185620238605438978
I mean, you could have written a similar story, but I think it would have been wrong. Those events weren't piling on top of one another in the way these are, and Ukraine and Syria aren't one-off stories where Trump can just reverse course and undo the damage. https://twitter.com/BenjySarlin/status/1185570518767935491
Wrote this 2 years ago on how one might go about analyzing the chances of impeachment & removal. Trump still has some protections: Congress is very partisan; he retains high approval among GOP voters. But many of these indicators have turned south for him. https://fivethirtyeight.com/features/chance-donald-trump-impeached/
2-3 (or 3-2) elections usually mean big changes in Congress. (Exception was 1972, which is why the Congress considering Nixon crimes did so.)
People who are like "oh, Republicans will never break with Trump" should read this story. The past few weeks have brought a rapid escalation of GOP criticism, as Trump has piled the G7 and Syria on top of an impeachment inquiry. https://www.washingtonpost.com/politics/growing-number-of-republicans-struggle-to-defend-trump-on-g-7-choice-ukraine-and-syria/2019/10/18/20e56612-f1b8-11e9-89eb-ec56cd414732_story.html
This was a great discussion for people who want a very deep dive into our new NBA metric. https://twitter.com/ElGee35/status/1185336745459503104
RT @NateSilver538: It's also not without irony that Zuckerberg is gaslighting people about Facebook's origin story in a speech where he exp…
Finally, it's not obvious that it's preferable in a representative democracy to have a politician who sticks to his or her personal beliefs at all times. Nor do you want someone who's always flip-flopping. But you probably do want some responsiveness to one's constituents.
Furthermore, quite often the notion of "authenticity" has been wielded in a way to make it harder for women and people of color to get elected, e.g. they're described as too ambitious or not true to themselves. That's another reason to be wary of it.
With few exceptions—Bernie may be one—I'd say it's extremely hard to tell how much of a politician's positions reflect their "authentic" personal beliefs vs. things they say "to get elected". So when people try to make those distinctions, your BS detectors should be set to high.
The past 5 people nominated for president are: McCain, Obama, Romney, Hillary Clinton, Trump. That's quite a lot of voters choosing "pragmatic, deliberate" restraint over demagogy ⁠— with one big exception. But Trump is unpopular and his presidency is in jeopardy.
Also, like, Michael Bennet, really? Even if D voters were to prefer temperamental moderation, there are more well-credentialed (Biden) and/or diverse (Booker) and/or interesting (Buttigieg) choices, all of whom got into the POTUS race earlier than Bennet's half-assed bid.
This whole article on Michael Bennet is sort of a fascinating j-school case study on the ideology of the political press. But it's main thesis seems to be pretty flawed in that it treats n=1 (Trump) as a new paradigm. https://politi.co/2J0P0zE
RT @Klonick: Dear @NateSilver538, You are awesome, but the First Amendment pertains to what the government controls, NOT what Facebook co…
Also, as a semi-related aside, it seems weird that the Warren campaign is now refusing donations >$200 from Facebook executives when such donations would sort of represent a middle finger to Zuckerberg.
The thing I wonder about is how sustainable this is given their mostly liberal workforce. https://twitter.com/nickconfessore/status/1185193580241084416
Biden is not a great Iowa candidate for a number of reasons. But in some of these polls, he's roughly tied for the lead with only around ~22 percent of the vote because Buttigieg is eating into Warren's vote in what *should* be a great state for her.
Some important context for why pro-Warren folks seem testy about Buttigieg is that he is indeed a bit of a problem for Warren in Iowa, where he's likely drawing from the same college-educated white base despite their policy differences. https://projects.fivethirtyeight.com/2020-primaries/democratic/iowa/
Mayor Pete's path to the nomination comes from knocking out Biden, not Warren
RT @NateSilver538: It's also not without irony that Zuckerberg is gaslighting people about Facebook's origin story in a speech where he exp…
It's probably too early in the day for wonky NBA content but was looking at our depth charts and holy hell is the Warriors' roster a dumpster fire after the top ~5 players. https://projects.fivethirtyeight.com/2020-nba-predictions/warriors/
Yeah, this is a bad trend. I'm generally of the view that campaigns don't do themselves much, if any, good by talking about their polls. At least Hogan's numbers are uniformly pretty good, whereas some candidates are extremely fond of cherrypicking the numbers they tweet out. https://twitter.com/byelin/status/1185183930053476352
RT @Klonick: Dear @NateSilver538, You are awesome, but the First Amendment pertains to what the government controls, NOT what Facebook co…
It's also not without irony that Zuckerberg is gaslighting people about Facebook's origin story in a speech where he explains why Facebook won't enforce any standards on the truthfulness of claims made by political actors. https://twitter.com/mattyglesias/status/1184919834771017728?s=19
If you don't take responsibility for the medium you create someone else will. And they won't have your interests at heart.
RT @FiveThirtyEight: The best NBA players of the last 6 seasons: https://53eig.ht/2VRQxx6
Also it seems effed up for Zuck to cloak these convoluted standards in the guise of free speech and the 1st Amendment when, at least to my layperson's understanding, false statements of fact aren't protected speech in the US.
False statements are absolutely protected speech in the US under many, many circumstances. “Misinformation” as complained of absolutely cannot be restrained by government, although a victim could sue for defamation.
IDK it seems revealing that there are a lot of things Facebook expressly says it won't allow -- claims that produce "voter suppression" or which constitue "bullying" -- but that fact-checking claims in politician's ads is somehow a bridge too far? https://www.wsj.com/articles/facebook-stands-for-free-expression-11571336089
Every candidate's online supporters are special in their own way, and the Warren supporters on here are mostly pretty cordial, but there's a part of Warren Twitter that's like "everything must be EXTREMELY good news for ELIZABETH WARREN at ALL TIMES with NO AMBIGUITY whatsoever".
Isn't that the case for any candidate gaining traction? There are fans and then there are FANS. See #Trump
You may have also run into the #YangGang2020 They're pretty intense. For math people.
Buttgieg and Klobuchar did gain in this poll, though. They also did well in the 538/Ipsos poll. The punditry may have been wrong in thinking Sanders had a better night than Warren, as both post-debate polls show Warren with better results than Sanders so far. https://twitter.com/AJentleson/status/1184927340624781312
First fully post-debate horse race poll is ... pretty boring. Biden down 1 point, Warren steady, Bernie down 1, Harris up 1, Buttigieg up 1, Klobuchar up 1, Steyer up 1, Beto down 1, everyone else steady it looks like. https://morningconsult.com/2020-democratic-primary-2/
RT @pareene: the craigslist casual encounters section was also invented to give people a voice during the iraq war
You're thinking of *TheFacebook*, Matt, not Facebook! (This is a joke.) https://twitter.com/mattyglesias/status/1184919834771017728
RT @NateSilver538: Indeed, polls find that many people assume that "Medicare For All" means something like a public option or "Medicare-for…
Then argue that single player is a better system! But don't play gotcha with candidates who used the M4A label to advocate for something different (public option) or to preserve their options, when progressive groups were explicitly encouraging the M4A label to be used that way. https://twitter.com/_waleedshahid/status/1184837993837858819
RT @_waleedshahid: @NateSilver538 @NationalNurses Even Jesse Jackson was calling for a single-payer system when he ran for President. A lot…
It's also noteworthy that the press now seems bored by the Warren-emerges-as-frontrunner story and is looking for new angles. They may be quicker to jump on a Bernie/Harris comeback or Buttigieg/Klob/Booker surge story than they were before. 3/3
But I'm more agnostic how much of chance the rest of the field has *collectively*. I think it's under 50%. But if it's 20% or 30% or something, that's still quite material. 2/3
Folks should probably make a subtle distinction here. Are Warren and Biden considerably more likely to win the nomination than the 3rd most likely candidate, whoever that is? (IMO still probably Bernie?). Yes, IMO; there's a big drop-off after the top two. 1/3 https://twitter.com/jmartNYT/status/1184632891101368322
RT @aaronecarroll: @NateSilver538 We did a whole thing on this @UpshotNYT @nytimes https://www.nytimes.com/interactive/2019/02/21/upshot/up-medicareforall.html
Indeed, polls find that many people assume that "Medicare For All" means something like a public option or "Medicare-for-all-who-want-it". When polls specify a version of M4A that would involve eliminating private insurance, support goes way down. https://53eig.ht/2SDL9fy https://twitter.com/sahilkapur/status/1184836192883433474
I think polls are pretty much only good for highly salient, extremely simple ideas. Anything more complex is either a terrible poll or to be brazenly misused as a marketing tactic
Employers would love Medicare if it were an option. Stop saying these policies belong to employees, or that they choose them. Employer-paid policies are chosen by employers
If somehow the Syria and quid-pro-quo stories fuse together, e.g. if Trump's decisions in Syria are later shown to result from various sorts of untoward interests he has in Turkey, Russia or another country, IMO we'd be looking at a pretty good chance of removal from office.
But Democrats for some reason don’t want to expand the scope of the inquiry.
Not “if”. “When”.
RT @NateSilver538: So Harris doesn't get very much talking time throughout the debate, which isn't her fault — but then she decides to take…
© 2009 creamsocial