Friday, October 25, 2019
The maybe-you-aren't-special heuristic
I'm going to make this one quick and messy (I really need a better phone dictation app) because my schedule is getting tight and when this happens, my idea-to-attention span ratio gets even worse than normal, but I want to make sure to get this point in print reasonably quickly because it's likely to feature in some upcoming and ongoing threads.
Everyone is aware of the dangers of assuming that your perspective is representative. There is, however, an opposite error in reasoning which largely goes undiscussed despite occurring frequently and being often just as dangerous.
If we limit ourselves to extremes, assuming that your experience is completely unrepresentative makes even less sense than assuming it's completely representative. If you picked a person at random out of the population, you wouldn't start out with the assumption that he or she was an outlier. You would certainly admit it was possible, after all, n = 1, but you would probably consider it more likely that he or she was somewhere closer to the mean.
At this point, we need to make the distinction between assuming a fact vs considering a possibility. While you shouldn't assume that what you see is what everyone else sees, unless you have some reason to think otherwise, you should always allow for the possibility.
One of the things that struck me listening to This American Life’s coverage of the 2008 financial crisis was that, at every stage of the process, people observed that their part of the system had extremely troubling problems but they were sure that this wasn’t representative of the system as a whole. They were, of course, wrong.
In an age of hype and AstroTurf, it becomes even more important to remember that you may not be the exception. When you find yourself having what seems to be an unusual reaction or holding a minority opinion you should remember that what seems to be a popular consensus is often a facade based on massive marketing and PR campaigns with budgets sometimes hitting the multi billion dollar mark.
I'm drawing a blank for musical accompaniment, so here's a catchy non sequitur for your weekend.
Thursday, October 24, 2019
Though we have to draw some inferences to get there, it appears that most, quite possibly all terrestrial superstations have a higher profit margin than Netflix
Another data point in our long-running thread.
Busy week, but I want to make a few quick points on this one.
Six years ago, just as terrestrial superstations were taking off, Nielsen released a study claiming that the over-the-air market was small and shrinking. The National Association of Broadcasters said their data showed just the opposite. Every bit of news we've seen since then suggests that Nielsen was wrong.
Not only has the industry (despite a near complete lack of hype) grown at a rapid and, more importantly, sustained clip, but the companies pushing hardest have been the owners of large numbers of TV stations. Not only are they the ones with skin in the game; they also have access to the most complete proprietary data. Pretty much every company in a position to know what's going on has expanded their presence in the market.
Where we do have head to head comparisons, the granddaddy of the industry, MeTV routinely dominates its much better positioned direct competitors like TVland. That's probably why CBS went with Weigel and passed over Viacom when starting Decades.
In addition to more stations constantly popping up, the individual superstations are becoming more ambitious, which brings us to a couple of points specifically about CourtTV. First, I believe this is the first case of a cable channel (even a dormant one) making the transition to OTA. CourtTV is still a brand of some value and considerable name recognition. The fact that it's showing up over the air rather than on cable tells you something about the growth and profit potential of the two media.
Second, this was not done on the cheap by Scripps. They not only acquired the original name and talent rather than putting together a knock-off; they went one step further and bought the original content library. That's a strong indicator of a serious long term commitment.
(By comparison, at least one very heavily hyped media company own far less of its content than most people including journalists and investors realize, but that's a topic for another post.)
On December 10th, 2018, Katz Broadcasting (owned by the E. W. Scripps Company) announced that they would relaunch Court TV as an over-the-air network following the acquisition of the intellectual property rights to the Court TV name and the pre-2008 Court TV original programming library from Turner Broadcasting System and Warner Bros. Entertainment. Scripps announced affiliation deals with Tribune Media and Univision Communications at that date, in addition to existing Scripps-owned stations. Further deals with Meredith Corporation, Nexstar Media Group (which was in the process of acquiring Tribune; the deal closed in September 2019), Tegna, and Quincy Media were announced on May 2, 2019.
The relaunched Court TV features live court coverage with original Court TV anchor Vinnie Politan as lead anchor, Court TV and CNN producers John Alleva and Scott Tufts as vice presidents and managing editors. The network began broadcasting on May 8, 2019.
Busy week, but I want to make a few quick points on this one.
Six years ago, just as terrestrial superstations were taking off, Nielsen released a study claiming that the over-the-air market was small and shrinking. The National Association of Broadcasters said their data showed just the opposite. Every bit of news we've seen since then suggests that Nielsen was wrong.
Not only has the industry (despite a near complete lack of hype) grown at a rapid and, more importantly, sustained clip, but the companies pushing hardest have been the owners of large numbers of TV stations. Not only are they the ones with skin in the game; they also have access to the most complete proprietary data. Pretty much every company in a position to know what's going on has expanded their presence in the market.
Where we do have head to head comparisons, the granddaddy of the industry, MeTV routinely dominates its much better positioned direct competitors like TVland. That's probably why CBS went with Weigel and passed over Viacom when starting Decades.
In addition to more stations constantly popping up, the individual superstations are becoming more ambitious, which brings us to a couple of points specifically about CourtTV. First, I believe this is the first case of a cable channel (even a dormant one) making the transition to OTA. CourtTV is still a brand of some value and considerable name recognition. The fact that it's showing up over the air rather than on cable tells you something about the growth and profit potential of the two media.
Second, this was not done on the cheap by Scripps. They not only acquired the original name and talent rather than putting together a knock-off; they went one step further and bought the original content library. That's a strong indicator of a serious long term commitment.
(By comparison, at least one very heavily hyped media company own far less of its content than most people including journalists and investors realize, but that's a topic for another post.)
Wednesday, October 23, 2019
Read the whole thing. Make notes. We'll be coming back to this.
Jeffrey Funk has a new essay up ("What’s Behind Technological Hype?"), and if you've been following any of our hype economy threads, you definitely need to check it out.
Online tech-hyping articles are now driven by the same dynamics as fake news. Journalists, bloggers, and websites prioritize page views and therefore say more positive things to attract viewers, while social media works as an amplifier. Journalists become “content marketers,” often hired by start-ups and universities to promote new technologies. Entrepreneurs, venture capitalists, university public relation offices, entrepreneurship programs, and professors who benefit from the promotion of new technologies all end up sharing an interest in increasing the level of hype.
Sometimes the beliefs behind political and technology hype merge. Think of libertarians who love cryptocurrencies, defense hawks who love new fighter jets, adventurers who think space travel is human destiny, train buffs who love hyperloop, health care professionals who love any technology that might prolong lives, anticorruption crusaders who love blockchain, social entrepreneurs who love financial technology (fintech), and environmentalists who love renewable energy and electric vehicles. Many of these special interest groups often believe their overall goals are far more important than more practical issues such as cost, performance, economic feasibility, and profitability, a problem made worse by the increasing polarization of the American public along ideological lines. As these special interests push their technologies on social media sites such as Twitter, LinkedIn, and Facebook, they create echo chambers in which people repeat the same message until it becomes an unquestioned mantra, even though few economic details are presented.
And when one overhyped technology fails to solve the world’s problems, there is always another waiting to be hyped. Train buffs replaced magnetically levitated trains with hyperloop, social entrepreneurs replaced microfinance (remember Nobel Peace Prize winner Muhammad Yunus?) with fintech, drug companies replaced stem cells with gene editing, and environmentalists have forgotten about nuclear fusion, solar water heaters, hydrogen vehicles, and cellulosic ethanol. Hype-driven economic disappointments seem never to dampen enthusiasm for new cycles of irrational exuberance.
Tuesday, October 22, 2019
Tuesday Tweets
Check out the thread
[1/4] Ok this is really funny, check this out.
I was in the process of booking a flight via @OneTravel. Trying to make me book ASAP, they claimed: "38 people are looking at this flight".
Whoa, 38 is a lot, I have to hurry up. But first I have to check how they came up with 38 >> pic.twitter.com/UaGhaiCQrR— Ophir Harpaz (@OphirHarpaz) October 16, 2019
This is a real chyron, I promise I did not photoshop it, I feel like that's an important thing to point out in this situation pic.twitter.com/hj6Tm77gZu— Andrew Lawrence (@ndrew_lawrence) October 18, 2019
More of a retcon than a retrofit, but still...
Mark Zuckerberg just publicly retrofit the reason for founding Facebook from a tool that could rate the attractiveness of female college students to an attempt to right the wrongs of the leadup to the Iraq War. You can't make this up— Joe Bernstein (@Bernstein) October 17, 2019
The fake origin story of Facebook written for Mark Zuckerberg by the team of GOP political operatives who runs their DC office versus contemporaneous coverage of Facebook’s origins. pic.twitter.com/dUo7Yw4PrT— Matthew EEEEK!-lesias (@mattyglesias) October 17, 2019
This excerpt from an occurrence in Peter Thiel's Stanford class on startups is a classic example of how "culture fit" is just tech company code for discrimination.
Any rejection based on "culture fit" when candidate is competent is usually sexism/racism/classism. https://t.co/0Sr84HbEnf— Dare Obasanjo (@Carnage4Life) October 19, 2019
Inside TurboTax’s 20-Year Fight to Stop Americans From Filing Their Taxes for Free https://t.co/T88U3WbA9i— Mark Palko (@MarkPalko1) October 19, 2019
Disney remains one of the best arguments against media consolidation.
ESPN is of course, owned by Disney, which has huge commercial interests in China. A senior ESPN executive specifically instructed ESPN staff to play down the NBA/China controversy when it first broke. https://t.co/yIDG7S8npn— Matt Schrader (@tombschrader) October 19, 2019
There is no defensible journalistic rationale for how differently these stories were played. pic.twitter.com/hodQ0H4AwU— Dan Lavoie (@djlavoie) October 19, 2019
Buried in this piece about the streaming rights for South Park likely selling for $500 million (!) -- Apple isn't bidding because they don't want to upset China. https://t.co/7OSVNrndYT— Adam Conover (@adamconover) October 19, 2019
Monday, October 21, 2019
Twilight of the Unicorns -- Blue Apron and the importance of unasked questions
Blue Apron IPO'd in June 2017 at a $2bn valuation
Yesterday, it closed at a stock price of $9.01, representing a market cap of just under $120mm
What happened and what can we learn?— Adam Keesling (@adam_keesling) September 25, 2019
When business historians try to make sense of the rise and fall of companies like Wework and Uber, the biggest challenge will be trying to reconstruct the thought processes of seemingly rational investors and analysts. While there will probably be an inclination to dismiss the whole thing as a madness of crowds, if you followed the events closely in real time, the decisions are a bit more explicable if not defensible.
Perhaps the first thing these future historians will have to understand is the extent to which influential people were both invested in and protective of the age of innovation narrative and the hype economy. If you'll pardon the expression, the "thought leaders" circa 2010 to 2020 firmly believed that they were in the center of a period of wondrous developments where all of the old rules had gone away and the next big thing was just around the corner. This narrative didn't just give them an excuse to dismiss what should have been glaringly obvious flaws in these overhyped business plans; it made them openly hostile to the very mindset which raised doubts and ask disturbing questions.
Even after one of these multibillion-dollar borderline Ponzi schemes collapsed, there was a real reluctance to acknowledge the herd of elephants that had been stomping about the room from the beginning. With Blue Apron, the lead elephant in the metaphor was competition. Meal kits (the culinary equivalent of a canned hunt) are very probably limited to a niche market, but more importantly, it is a market with no barriers to entry and a truly daunting queue of well positioned potential competitors with deep pockets, economies of scale, and well-established brands. There's Amazon/Whole Foods, of course. Add to that Walmart, the Kroger and Safeway groups and other major grocery store chains all of which have been edging into the home delivery space for a while, and fast casual places like Macaroni Grill which would be a natural fit. Then there are celebrity chefs and Food Network personalities who might partner with companies such as Uber Eats and that's far from an exhaustive list.
Blue Apron spent hundreds of millions of dollars trying to popularized the idea of home delivery meal kits, but even if they had succeeded, they never could have cashed in on anything more than a limited scale. The market would have immediately become so intensely competitive that the profit margins necessary to justify Blue Apron's valuation would have been impossible.
These issues have been obvious from the beginning, but even now analysts are reluctant to bring them up. When you start pulling one that thread, who knows what will unravel.
Friday, October 18, 2019
Thursday, October 17, 2019
Disappointed by the omission of MoviePass, but...
When we try to make sense of the unicorn delusion years from now, we'll want to revisit this passage by Derek Thompson. [emphasis added]
Several weeks ago, I met up with a friend in New York who suggested we grab a bite at a Scottish bar in the West Village. He had booked the table through something called Seated, a restaurant app that pays users who make reservations on the platform. We ordered two cocktails each, along with some food. And in exchange for the hard labor of drinking whiskey, the app awarded us $30 in credits redeemable at a variety of retailers.
I am never offended by freebies. But this arrangement seemed almost obscenely generous. To throw cash at people every time they walk into a restaurant does not sound like a business. It sounds like a plot to lose money as fast as possible—or to provide New Yorkers, who are constantly dining out, with a kind of minimum basic income.
“How does this thing make any sense?” I asked my friend.
“I don’t know if it makes sense, and I don’t know how long it’s going to last,” he said, pausing to scroll through redemption options. “So, do you want your half in Amazon credits or Starbucks?”
I don’t know if it makes sense, and I don’t know how long it’s going to last. Is there a better epitaph for this age of consumer technology?
Starting about a decade ago, a fleet of well-known start-ups promised to change the way we work, work out, eat, shop, cook, commute, and sleep. These lifestyle-adjustment companies were so influential that wannabe entrepreneurs saw them as a template, flooding Silicon Valley with “Uber for X” pitches.
But as their promises soared, their profits didn’t. It’s easy to spend all day riding unicorns whose most magical property is their ability to combine high valuations with persistently negative earnings—something I’ve pointed out before. If you wake up on a Casper mattress, work out with a Peloton before breakfast, Uber to your desk at a WeWork, order DoorDash for lunch, take a Lyft home, and get dinner through Postmates, you’ve interacted with seven companies that will collectively lose nearly $14 billion this year. If you use Lime scooters to bop around the city, download Wag to walk your dog, and sign up for Blue Apron to make a meal, that’s three more brands that have never record a dime in earnings, or have seen their valuations fall by more than 50 percent.
Wednesday, October 16, 2019
Companies with monopsony power abuse monopsony power
More competent evil from Amazon reported by Jay Greene of the remarkably independent Washington Post.
SEATTLE — When Jeff Peterson’s Amazon seller account was hacked recently, he frantically tried to reach Amazon’s customer service for help restoring access to his sports memorabilia store.
As nearly 4,000 fraudulent orders rang up, the Garden Grove, Calif.-based seller called Amazon’s seller support line, phoned its main customer service number, reached out via a separate account on its Canadian site, and even sent an email to chief executive Jeff Bezos. Nothing worked.
“I can’t get any answers from Amazon at all to fix this,” Peterson said, as negative reviews of his service accumulated, decimating his business.
One thing he hadn’t done was pay as much as $5,000 a month for a program Amazon offers sellers as a way to get quick help from a real person.
Amazon has become a powerful marketplace alongside its role as an online retailer, with more than 2.5 million third-party sellers who have become global businesses on its platform. Early on, Amazon compelled sellers to use its warehouses to guarantee speedy Prime shipping, in addition to other programs that largely benefited consumers. But now, sellers and former employees familiar with Amazon’s internal strategy say the company is increasingly focused on boosting its profits on the backs of its sellers — often without any clear upside for customers.
The services include charging sellers thousands of dollars to speak to account managers, as well as making it necessary to purchase ads to guarantee the top spot on a search page. Plus, Amazon is aggressively pushing its own brands — something that may be cheaper for consumers in the short run, but demonstrates its overall power over pricing and merchandise on the site. That gives it an advantage over rival products and sellers who rely on Amazon for their livelihood and have few alternatives if they want to thrive selling online.
Amazon says its success is dependent on those sellers and insists it always prioritizes shoppers.
As much as a third of every dollar merchants make goes back to Amazon, according to consultants and sellers. That helped Amazon generate $42.7 billion in revenue from seller services such as fees and commissions last year, a number that has nearly doubled in two years.
That has drawn the attention of regulators and lawmakers both in the U.S. and abroad, who are investigating Amazon and other large tech companies for potential violations of antitrust law and abusing dominant marketplace power. Traditionally, U.S. regulators have focused on consumer harm, but officials recently emphasized the need to look at the way several tech giants are using their market clout to lower quality, reduce innovation and diminish consumer privacy as officials consider regulating giants of the digital economy.
Tuesday, October 15, 2019
Tuesday Tweets (mostly non-political edition)
I once told someone from Uber that I was a transportation writer, and they said, "So why are you writing about us? We're a tech company." https://t.co/0LUddvONCa— Laura J. "𝔣𝔞𝔦𝔯 𝔠𝔬𝔫𝔱𝔯𝔞𝔠𝔱 𝔫𝔬𝔴" Nelson (@laura_nelson) October 14, 2019
Maybe Malcolm Gladwell can work this into his next book.— Mark Palko (@MarkPalko1) October 12, 2019
Today at @theprospect we have the very first excerpt from @matthewstoller's stunning book Goliath: The 100-Year War Between Monopoly Power and Democracy.
This one is all about Andrew Mellon, who ran the country and the economy in the 1920s. https://t.co/FhYkRaAxEo— David Dayen (@ddayen) October 14, 2019
This is a lie: “Testing is done. It works. We are still working out funding and regulatory issues, but we are excited about it,” said Kristen Hammer, Manager of Business Development, with Virgin Hyperloop One.
https://t.co/xwt4RCzHVo— Mark Palko (@MarkPalko1) October 6, 2019
Instead we're getting this PR war of attrition against expert opinion culminating in this endless roadshow. They've already got us talking about public private partnerships and started tax money flowing for a system that most engineers say will never be viable.— Mark Palko (@MarkPalko1) October 6, 2019
In a rare show of panic, other candidates are “running”.— (((Brian Phillips))) (@blessdb1963) October 7, 2019
Voom pilot told me that our 30 mile, 15 minute helicopter trip burned 10 gallons of fuel. An entire tank of gas for a small car is 12. So, generating learnings, and lots and lots of greenhouse gas! https://t.co/ambfJoKtUs— undeath & afterlife of ghastly american cities (@mslaurabliss) October 4, 2019
NEW: Facebook CEO Mark Zuckerberg has been hosting secretive talks & off-the-record dinners with conservative influencers in recent months to discuss free speech and partnerships. Among the participants: Tucker Carlson & Lindsey Graham. scoop w/@dlippman:https://t.co/n5IV0N66X4— Natasha Bertrand (@NatashaBertrand) October 14, 2019
—
Mark Palko (@MarkPalko1) October
6, 2019
This is an extraordinarily important point. Even taking into account the difference between primaries and general elections, this comparative disinterest in a candidate's heart attack compared to that prompted by a fainting spell is striking. https://t.co/Y2mjNTOWBd— Mark Palko (@MarkPalko1) October 5, 2019
Today is also one of those days when making sense of events actually distorts them, because sense is what they fundamentally do not make. https://t.co/mqwZy5B8Yr— Jay Rosen (@jayrosen_nyu) October 14, 2019
Monday, October 14, 2019
Trump as stressor -- Shep Smith edition
This is another one of those threads that goes back three of four years. The conservative movement's media strategy proved to be remarkably effective for a long time, but there waere always tensions and potential instability.
It only worked as long as things weren't pushed too far. With the mainstream press this meant keeping the spin just within the bounds of plausibility.
Over the past few years, the two versions have drifted so far apart that something has to break. Breakage, sharp and sudden, is what we saw last week. To get a sense of just how unexpected, check out the reactions in this clip. Cavuto and Roberts were professional enough to keep talking but they couldn't hide their shock.
It only worked as long as things weren't pushed too far. With the mainstream press this meant keeping the spin just within the bounds of plausibility.
If you remember the elections of 2000 and 2004, you will probably recall talk of Karl Rove and his mastery of "political jujitsu." It was generally discussed as if it were some sort of mystical Jedi mind trick that allowed Rove to make strings into weaknesses and weaknesses into strengths. Mainly, it came down to the realization that most reporters would respond to obvious lies with straight faces and no follow-up questions.Another component was keeping a veneer of respectability on Fox News. This rested primarily on two journalists, Chris Wallace and Shep Smith. For decades, you could find actual news on Fox, even though the majority of the programming consisted of spin, propaganda and often outright disinformation. These two types of shows could coexist because there was some overlap, or at least proximity, between the two narratives.
In 2004, I remember Republican operatives making the argument that George W. Bush's military record compared favorably with that of John Kerry. Just to review, Kerry was a legitimate war hero in terms of courage, sacrifice, and effectiveness. On the other side of the ledger, even if we push aside all of the accusations and contested points about favoritism and completion of requirements, there is a relatively cushy stint in the National Guard.
These and other clearly untrue statements were usually allowed to stand largely because this was a symbiotic relationship. It was in both the source's and the journalist's interests to keep this relationship going and not to push the boundaries in either direction.
Over the past few years, the two versions have drifted so far apart that something has to break. Breakage, sharp and sudden, is what we saw last week. To get a sense of just how unexpected, check out the reactions in this clip. Cavuto and Roberts were professional enough to keep talking but they couldn't hide their shock.
Neil Cavuto and John Roberts clearly blind-sided by Shep Smith departure pic.twitter.com/36ijhC8th6— Nicole Lafond (@Nicole_Lafond) October 11, 2019
Friday, October 11, 2019
Cars of futures past/Class Foghorn Leghorn (Google it)
You probably haven't seen this 1948 short on the cars of the future...
But you may be familiar with this Tex Avery parody of the genre.
Tex Avery - MGM 1951-09-22 - Car of Tomorrow
Thursday, October 10, 2019
Four years ago (more or less) -- Wishful Analytics
You would be hard-pressed to find a group of reporters and analysts who have more to answer for than data journalists on the political beat in 2016.
The dean of the group, Nate Silver, turned in an embarrassingly bad series of analyses during the primary, but to his considerable credit, acknowledged his errors and greatly upped his game during the general election. For this reason, he was perhaps the only major poll watcher to have reasonable forecasts on the eve of the election. The rest of the group, however, learned absolutely nothing from the nomination and continued making the same egregious errors through the general.
Since then, the people who screwed up have pushed the self-serving idea that, while there may have been mistakes, they were things that are only apparent in hindsight.
In cases like this, you simply cannot beat contemporary accounts.
The trouble is that the New York Times is very much committed to a style of political analysis that takes the standard narrative almost to the formal level of a well-made play. The objective is to get to the preassigned destination with as much craft and wit as possible. Nate Silver's problems at the NYT generally came from his habit of following the data to conclusions that made his editors and colleagues uncomfortable (by raising disturbing questions about the value of their work).
Cohn's articles on Trump have been an extended study in wishful analytics, starting with a desired conclusion then trying to dredge up some numbers to support it. He really, really, really, really, really wants to see Trump as another Herman Cain. Other than both being successful businessmen, the analogy is strained -- Cain was a little-known figure who surged well into the campaign because the base was looking for an alternative to an unacceptable presumptive nominee – but Cohn brings up the pizza magnate at every opportunity.
In addition to reassuring analogies, Cohn is also inclined to see comforting inflection points. Here's his response to the McCain dust-up.
Paul Krugman (like Silver, another NYT writer frequently at odds with the paper's culture) dismantled this argument by immediately spotting the key flaw.
Cohn's initial reaction to his failed prediction was to argue that the polls weren't current enough to show that he was right. When that position became untenable, he shifted his focus to the next inflection point:
This attempt to prop up that narrative have become increasing strained and convoluted, as you can see from the most recent entry
First off, notice the odd way that Cohn discusses influence. If I asked if you would like to “play a bigger role in shaping the outcome” of something, you would naturally assume I meant would you like to have more of a say, but that's not at all how the concept of influence is used in the passage above. Cohn is simply saying that a world where Trump was behind in one of the first two primaries might have a different nominee but since Trump wouldn't get to pick who would beat him, it's not clear why he would care and since there's no telling who would win in Cohn's alternate reality, it's not clear why anyone else would care either.
But even if we accept Cohn's framing, we then run into another fatal flaw. Put in more precise terms, “harming all candidates fairly equally” means that each candidate's probability of becoming president would have been the same had Trump not entered the race. This is almost impossible on at least three levels:
Trump has already produced a serious shift in the discussion, bringing issues like immigration and Social Security/Medicare to the foreground while sucking away the oxygen from others. This is certain to help some candidates more than others;
For this and other reasons, the impact on the polls so far has been anything but symmetric;
And even if Trump's support were coming proportionally from each of the other contenders, that still wouldn't constitute equal harm. Primaries are complex beasts. We have to take into account convergence, feedback loops, liquidity, serial correlation, et cetera. The suggestion that you could remove the first two primaries from contention without major ramifications is laughably naive.
Finally there's that “only.” Even if Trump isn't the nominee (and I would certainly call him a long shot), he can still influence the process as either kingmaker or spoiler.
While Cohn's work on this topic has been terrible, what's important here is not the failings of one writer but the current culture of journalism. This is what happens when even the best publications in the country embrace conventional narratives and groupthink, adopt self-serving but silly conventions and let their standards slip.
The dean of the group, Nate Silver, turned in an embarrassingly bad series of analyses during the primary, but to his considerable credit, acknowledged his errors and greatly upped his game during the general election. For this reason, he was perhaps the only major poll watcher to have reasonable forecasts on the eve of the election. The rest of the group, however, learned absolutely nothing from the nomination and continued making the same egregious errors through the general.
Since then, the people who screwed up have pushed the self-serving idea that, while there may have been mistakes, they were things that are only apparent in hindsight.
In cases like this, you simply cannot beat contemporary accounts.
Friday, August 21, 2015
Wishful Analytics
As mentioned previously, Donald Trump's campaign has definitely strained the standard assumptions of political reporting, Though this is an industry wide problem (even Five Thirty Eight hasn't been immune), it is nowhere more severe than at the New York Times.The trouble is that the New York Times is very much committed to a style of political analysis that takes the standard narrative almost to the formal level of a well-made play. The objective is to get to the preassigned destination with as much craft and wit as possible. Nate Silver's problems at the NYT generally came from his habit of following the data to conclusions that made his editors and colleagues uncomfortable (by raising disturbing questions about the value of their work).
Cohn's articles on Trump have been an extended study in wishful analytics, starting with a desired conclusion then trying to dredge up some numbers to support it. He really, really, really, really, really wants to see Trump as another Herman Cain. Other than both being successful businessmen, the analogy is strained -- Cain was a little-known figure who surged well into the campaign because the base was looking for an alternative to an unacceptable presumptive nominee – but Cohn brings up the pizza magnate at every opportunity.
In addition to reassuring analogies, Cohn is also inclined to see comforting inflection points. Here's his response to the McCain dust-up.
The Trump Campaign’s Turning Point
Donald Trump’s surge in the polls has followed the classic pattern of a media-driven surge. Now it will most likely follow the classic pattern of a party-backed decline.
Mr. Trump’s candidacy probably reached an inflection point on Saturday after he essentially criticized John McCain for being captured during the Vietnam War. Republican campaigns and elites quickly moved to condemn his comments — a shift that will probably mark the moment when Trump’s candidacy went from boom to bust.
Paul Krugman (like Silver, another NYT writer frequently at odds with the paper's culture) dismantled this argument by immediately spotting the key flaw.
What I would argue is key to this situation — and, in particular, key to understanding how the conventional wisdom on Trump/McCain went so wrong — is the reality that a lot of people are, in effect, members of a delusional cult that is impervious to logic and evidence, and has lost touch with reality.
I am, of course, talking about pundits who prize themselves for their centrism.
...
On one side, they can’t admit the moderation of the Democrats, which is why you had the spectacle of demands that Obama change course and support his own policies.
On the other side, they have had to invent an imaginary GOP that bears little resemblance to the real thing. This means being continually surprised by the radicalism of the base. It also means a determination to see various Republicans as Serious, Honest Conservatives — SHCs? — whom the centrists know, just know, have to exist.
...
But the ur-SHC is John McCain, the Straight-Talking Maverick. Never mind that he is clearly eager to wage as many wars as possible, that he has long since abandoned his once-realistic positions on climate change and immigration, that he tried to put Sarah Palin a heartbeat from the presidency. McCain the myth is who they see, and keep putting on TV. And they imagined that everyone else must see him the same way, that Trump’s sneering at his war record would cause everyone to turn away in disgust.
But the Republican base isn’t eager to hear from SHCs; it has never put McCain on a pedestal; and people who like Donald Trump are not exactly likely to be scared off by his lack of decorum.
Cohn's initial reaction to his failed prediction was to argue that the polls weren't current enough to show that he was right. When that position became untenable, he shifted his focus to the next inflection point:
Mr. Rubio, the senator from Florida, has a good case to be considered the debate’s top performer. A weaker Mr. Bush probably benefits Mr. Rubio as much as anyone, and if Mr. Bush raised questions about whether he would be a great general election candidate, then Mr. Rubio added yet more reason to believe he could be a good one. Mr. Rubio still has the challenge of figuring out how to break through a strong field in a factional party.You might want to reread that last paragraph a couple of times to get your head around just how wrong it turned out to be. Pay particular attention to the statements qualified with 'probably' both here and in the McCain piece. The confidence displayed had nothing to do with likelihood – all were comically off-base – and had everything to do with how badly those committed to the standard narrative wanted the statements to be true.
...
Mr. Walker won by not losing. In a lot of ways, the moderators’ tough, specific questions played to Mr. Walker’s weakness. He didn’t have much time to emphasize his fight against unions in Wisconsin. But he handled several tough questions — on abortion; on relations with Arab nations; what he would do after terminating the Iran deal; race; and his employment record — without appearing flustered or making a mistake. His answers were concise and sharp.
...
Mr. Kasich also advanced his cause. He entered as a largely unknown candidate outside of Ohio, where he is governor. But he was backed by a supportive audience, he deftly handled tough questions, and he had a solid answer on a question about attending same-sex weddings. His answer might not resonate among many Republicans, but it will resonate in New Hampshire — the state where he needs to deny Mr. Bush a path to victory and vault to the top of the pack.
It was Donald Trump, though, who might have had the weakest performance. No, it may not be the end of his surge. But he consistently faced pointed questions, didn’t always have satisfactory answers, endured a fairly hostile crowd and probably won’t receive as much media attention coming out of the debate as he did in the weeks before it. If you take the view that he’s heavily dependent on media coverage, that’s an issue. Whatever coverage he does get may be fairly negative — probably focusing on his unwillingness to guarantee support for the Republican nominee.
This attempt to prop up that narrative have become increasing strained and convoluted, as you can see from the most recent entry
Yet oddly, the breadth of [Trump's] appeal and his strength reduce his importance in shaping the outcome of the race.
If Mr. Trump were weaker, or if his support were more narrowly concentrated in either New Hampshire or Iowa, he would play a bigger role in shaping the outcome. In that scenario, a non-Trump candidate might win either Iowa or New Hampshire — and he or she would be in much better position than the second-place finisher in the state where Mr. Trump was victorious.
If Mr. Trump were to win both Iowa and New Hampshire, the second-place finishers would advance as if they were winners. Assuming that one or both of the second-place finishers were broadly acceptable, the party would try to coalesce behind one of the two ahead of the winner-take-all contests on March 15.
In the end, Mr. Trump almost certainly won’t win the Republican nomination; the rest of the party will consolidate around anyone else. He can influence the outcome only if his support costs another candidate more than others. But for now, he seems to be harming all candidates fairly equally.
First off, notice the odd way that Cohn discusses influence. If I asked if you would like to “play a bigger role in shaping the outcome” of something, you would naturally assume I meant would you like to have more of a say, but that's not at all how the concept of influence is used in the passage above. Cohn is simply saying that a world where Trump was behind in one of the first two primaries might have a different nominee but since Trump wouldn't get to pick who would beat him, it's not clear why he would care and since there's no telling who would win in Cohn's alternate reality, it's not clear why anyone else would care either.
But even if we accept Cohn's framing, we then run into another fatal flaw. Put in more precise terms, “harming all candidates fairly equally” means that each candidate's probability of becoming president would have been the same had Trump not entered the race. This is almost impossible on at least three levels:
Trump has already produced a serious shift in the discussion, bringing issues like immigration and Social Security/Medicare to the foreground while sucking away the oxygen from others. This is certain to help some candidates more than others;
For this and other reasons, the impact on the polls so far has been anything but symmetric;
And even if Trump's support were coming proportionally from each of the other contenders, that still wouldn't constitute equal harm. Primaries are complex beasts. We have to take into account convergence, feedback loops, liquidity, serial correlation, et cetera. The suggestion that you could remove the first two primaries from contention without major ramifications is laughably naive.
Finally there's that “only.” Even if Trump isn't the nominee (and I would certainly call him a long shot), he can still influence the process as either kingmaker or spoiler.
While Cohn's work on this topic has been terrible, what's important here is not the failings of one writer but the current culture of journalism. This is what happens when even the best publications in the country embrace conventional narratives and groupthink, adopt self-serving but silly conventions and let their standards slip.
Wednesday, October 9, 2019
Hyperloop as terrorist target
There is an extraordinary disconnect between the reporters covering the hyperloop and the scientists and engineers who actually work in this field. With the possible exception of self-sustaining Martian colonies, there is no area where what you read in mainstream publications like the New York Times or the Atlantic will so completely contradict what you hear from independent experts. For the most part, the reporters aren't even aware of the questions they should be asking.
Case in point.
Other than this one somewhat superficial but otherwise pretty good article in the National Interest, there has been virtually no coverage of the hyperloop's unique vulnerability to terrorist attack. This is an enormously complicated issue, but there are certain aspects of it which are obvious to anyone who has seriously looked at the problem even though they very seldom make it into the breathless puff pieces we've come to expect.
Near top of mind for any engineer who has given the question of safety any thought is what will happen if there is a breach in the tube and the vacuum is compromised. At that point, a wall of air traveling at the speed of sound goes barreling down the tube in both directions. For those pods traveling toward the wall at nearly the speed of sound, it will certainly not turn out well. All of the occupants of the first pod will die instantly. What happens next is impossible to determine without extensive tests, but even allowing for the best-case scenario, you will have hundreds of people who need to be evacuated
From the tube which needs to be repressurized immediately. The passengers will then need to make their way to the closest escape hatch. The hatches themselves will present a considerable engineering challenge since they will need to be not only airtight but completely inaccessible from the outside lest they become yet another point of attack.
After you get these people out of the hundreds of miles of tube, you have to do something about the pods themselves. Keep in mind, the primary propulsion and braking of the pods is provided by linear induction motors at either end. I'm not going to spend a lot of time speculating on the best way of getting the pods out of the tube, but it is safe to say it will be neither trivial nor quick.
Then comes the massive job of getting the system back up and running. Not only will you have to replace the section with the original breach, you will have to completely clean the tube of debris, quite a bit of which will undoubtedly be sucked in. Then you will have to inspect every inch of the tube for structural damage and double-check every seal, escape hatch, thermal expansion joint, and other potential point of failure. You will have to make sure that there was no damage done to the magnetic levitation track. You will have to make absolutely certain that the pods are in perfect working order since, when running through the system, a loss of cabin pressure in one of these things means instant death for everyone inside.
There are few systems of Transportation infrastructure where a single attack can produce this level of damage and economic disruption. Of these, none offers the variety vulnerable points that the hyperloop does. If we are talking about elevated tracks, which seems to be the primary approach being proposed, every section of tube and every pylon supporting it is a potential point of attack. Hardening all of those targets might be possible, but it would be unimaginably expensive.
Case in point.
Other than this one somewhat superficial but otherwise pretty good article in the National Interest, there has been virtually no coverage of the hyperloop's unique vulnerability to terrorist attack. This is an enormously complicated issue, but there are certain aspects of it which are obvious to anyone who has seriously looked at the problem even though they very seldom make it into the breathless puff pieces we've come to expect.
Near top of mind for any engineer who has given the question of safety any thought is what will happen if there is a breach in the tube and the vacuum is compromised. At that point, a wall of air traveling at the speed of sound goes barreling down the tube in both directions. For those pods traveling toward the wall at nearly the speed of sound, it will certainly not turn out well. All of the occupants of the first pod will die instantly. What happens next is impossible to determine without extensive tests, but even allowing for the best-case scenario, you will have hundreds of people who need to be evacuated
From the tube which needs to be repressurized immediately. The passengers will then need to make their way to the closest escape hatch. The hatches themselves will present a considerable engineering challenge since they will need to be not only airtight but completely inaccessible from the outside lest they become yet another point of attack.
After you get these people out of the hundreds of miles of tube, you have to do something about the pods themselves. Keep in mind, the primary propulsion and braking of the pods is provided by linear induction motors at either end. I'm not going to spend a lot of time speculating on the best way of getting the pods out of the tube, but it is safe to say it will be neither trivial nor quick.
Then comes the massive job of getting the system back up and running. Not only will you have to replace the section with the original breach, you will have to completely clean the tube of debris, quite a bit of which will undoubtedly be sucked in. Then you will have to inspect every inch of the tube for structural damage and double-check every seal, escape hatch, thermal expansion joint, and other potential point of failure. You will have to make sure that there was no damage done to the magnetic levitation track. You will have to make absolutely certain that the pods are in perfect working order since, when running through the system, a loss of cabin pressure in one of these things means instant death for everyone inside.
There are few systems of Transportation infrastructure where a single attack can produce this level of damage and economic disruption. Of these, none offers the variety vulnerable points that the hyperloop does. If we are talking about elevated tracks, which seems to be the primary approach being proposed, every section of tube and every pylon supporting it is a potential point of attack. Hardening all of those targets might be possible, but it would be unimaginably expensive.
Tuesday, October 8, 2019
Tuesday Tweets -- Great Unwinding* edition
*See yesterday's post.
And from the wages of Strauss thread.
And one just for the hell of it.
This is our next big political thread. Sane, mainstream conservatives are starting to think about the potential consequences to the GOP if the unwinding of the Trump cult of personality goes badly. https://t.co/Y6Ba9EW3uo— Mark Palko (@MarkPalko1) October 3, 2019
“At the end of this process Trump will probably be acquitted by the Senate. He will declare himself vindicated and victorious in his battle against The Swamp. An ugly backlash could ensue — in both parties.” https://t.co/TlcQKpx2js— David Brooks (@nytdavidbrooks) September 26, 2019
A really good pt from @BrendanBuck in this @CatieEdmondson joint - Hill Rs can’t pivot to “inappropriate but not impeachable” bc Trump won’t brook *any* criticism
https://t.co/O0njHcbgJW— Jonathan Martin (@jmartNYT) October 5, 2019
This is our next big political thread. Sane, mainstream conservatives are starting to think about the potential consequences to the GOP if the unwinding of the Trump cult of personality goes badly. https://t.co/Y6Ba9EW3uo— Mark Palko (@MarkPalko1) October 3, 2019
Never gets old. https://t.co/0ucnTckdcs— Rick Wilson (@TheRickWilson) October 2, 2019
So here’s the answer: silence from more than 60 GOP Senators and Members of Congress. https://t.co/kjfKqcDglf pic.twitter.com/GiWmufsdHo— Jim Sciutto (@jimsciutto) October 4, 2019
lol https://t.co/X97erAk5yU via @TPM— Josh Marshall (@joshtpm) October 6, 2019
The Clinton impeachment debacle created a CW that impeaching *always* backfires, which is really coming apart this week https://t.co/Dv7ayqkTx0— Dave Weigel (@daveweigel) October 4, 2019
And from the wages of Strauss thread.
Not quite sure how society survives a significant portion of the ruling generation believing in Baby Boomer Slenderman https://t.co/HDFGyJnM3g— Mark Agee (@MarkAgee) October 5, 2019
And one just for the hell of it.
When your committee has its own ideas about your dissertation proposal pic.twitter.com/kJlXChNUyC— Sanjay Srivastava (@hardsci) October 4, 2019
Monday, October 7, 2019
Out with the Wages of Strauss, in with the Great Unwinding
We have reached a point in the show which always makes the fans a little nervous. we have decided that one of our oldest and biggest storylines is starting to come to a natural conclusion so we need to begin wrapping up the loose ends and introducing the next one.
For years now, when it came to politics, the big recurring story was what you might call the wages of Strauss. we pushed the we pushed the idea that either the main cause or the essential context of almost every major political development over the past couple of decades came from the conservative movements relatively public conclusion that their agenda, while it might hold its own for a while and perhaps even surge ahead now and then, was destined to lose the battle of public opinion in the long run.
This left them with two choices, either modify their ideas so that they could win over the majority of the public, or undermine the Democratic process through a Straussian model, an approach based on controlling most of the money and increasing the influence that could be bought with that money, changing government so that an ever smaller part of the population had an ever-larger role in governing the country and creating a sophisticated three-tiered information management system where trusted sources of information were underfunded and undermined, the mainstream press was kept in line through a combination of message discipline and incentives with special emphasis placed on working the refs, and the creation of a special media bubble for the base which used spin, propaganda, and outright disinformation to keep the canon fodder angry, frightened, and loyal.
For a long time this approach worked remarkably well, but you could argue that the signs of instability were there from the beginning, particularly the difficulty of controlling the creation and flow of disinformation, the vulnerability to what you might call hostile take over, and the way the system lent itself to cults of personality.
We've had a good run with this storyline for a long time now, but it seems to be coming to a resolution and it has definitely lost a great deal of its novelty. (Lots of people are making these points now.)
The next big story, but one which we believe will dominate American politics for at least the next decade or so will be how the Republican party deals with the unwinding of the Trump cult of personality. Dismantling such a cult is tremendously difficult under the best of circumstances where the leader can be eased out gently, but you have with Donald Trump someone who has no loyalty to the party whatsoever and who is temperamentally not only capable but inclined to tear the house down should he feel betrayed.
If Trump continues to grow more erratic and public disapproval and support for his removal continues to grow, then association will be increasingly damaging to Republicans in office. However, for those same politicians, at least those who come up for election in the next two to four years, it is not at all clear that any could survive if the Trump loyalists turned on them.
But this goes beyond individual candidates. Trump's hold on the core of the base is so strong and so personal that, if he were to tell them directly that the GOP had betrayed both him and them, they would almost certainly side with him. They might form a third party, or simply boycott if you elections, or, yes, even consider voting for Democrats.. I know that last one sounds unlikely but it is within the realm of possibility if the intraparty civil war got bitter enough.
Obviously, if Trump survives this scandal and is reelected in 2020, all of this is moot, but if not, then how things break will be a story we’ll be glad to have been following.
Subscribe to:
Posts (Atom)