Thursday, August 23, 2018
Hark the Herald Reporters Sing – – another magical heuristics post
I was reading yet another piece talking about how some Silicon Valley/Silicon Beach company was about to do some impossible thing. As I was mentally compiling a list of factual errors, fallacies, and improbabilities, it suddenly struck me that I was approaching the piece from a completely inappropriate framework. This wasn’t an analysis of a business model; it was the foretelling of the coming conquests of great men. The numbers and anecdotes given in the piece were not evidence and context; they were portents.
We’ve certainly discussed the idea that the discourse of the tech industry is rife with the language and imagery of myths and magic, and we haven’t been the only ones to notice (Paul Krugman recently described the rhetoric of crypto currency supporters as “messianic”), but reading that piece I was struck by an implication which I believe had evaded me up to this point. What are the psychological effects of these stories on the people who tell them?
If tech journalists are increasingly inclined to see figures like Elon Musk as messiahs (and if you think this is over the top, go back and read the Rolling Stone cover story), isn’t it inevitable that these journalists who are announcing the coming of these messiahs will start to think of themselves as heralds?
Up until now, we have tended to describe these reporters as conmen or suckers, but just as two categories are not mutually exclusive, neither is this one. If you dig into the history of charlatans, you’ll have no trouble finding promoters who were both willing accomplices and true believers.
I’ve been struggling to explain the seductiveness of the bullshit narratives which has come to dominate our discussion of science and technology. Part of this unquestionably comes down to the appeal of the pitch: the future is not only brighter than you could imagine, it’s also going to be cheap and easily attained and it’s right around the corner. It’s not difficult to imagine people wanting to believe this, but that still doesn’t fully explain how so many smart and skeptical people have fallen for the obvious flimflam and have rejected the rational for the magical heuristics.
I suspect the psychology of the herald plays a large role here. While it certainly doesn’t hurt that these stories are good copy or that these “messiahs” have appropriately majestic lifestyles (sometimes, the next best thing to being rich is hanging out with the rich), but the reporters covering the stories also have the sense of being part of something great even if their role is only to witness and record.
Wednesday, August 22, 2018
The complicated statistics and conflicting objectives of video recommendation algorithms
What is the objective of the algorithm and how do we know if it's achieving that goal? Well, that depends. The things we would really like to know like impact on long-term customer loyalty and retention, particularly in the event of price increases and/or loss of popular content. Unfortunately, getting an answer to this question that does not entail waiting five or 10 years is extremely difficult.
I suspect the most popular alternative would be simply measuring how often customers accept the recommendation, but that metric is deeply problematic even under the best of circumstances and is almost worthless if approached naïvely.
A big part of the problem is that customers almost always walk into the door with some kind of mental queue, albeit often a vague and incomplete one. A recommendation algorithm that does not change that queue in a nontrivial way is a failure, a complete waste of time and money.
[I almost veered off into a discussion of how models and algorithms interact in this situation, but I'm trying to stay on topic and that's a subject that really needs a serious post or perhaps even thread of its own.]
Now we get to the next level of complexity. If you fail to change the queue you accomplish nothing, but if you do change the queue you still don't necessarily accomplish anything of business value. This is where we need to start getting specific about our objectives because there are some perfectly reasonable but contradictory choices to consider
If the sole concern is serving the customer (and serving the customer is never the sole concern) then the goal might be to get the viewers to watch and give a high ratings to shows they were previously unaware of. Another interesting related metric might be to look at before and after ratings, comparing how much they expected to enjoy a program with how much they actually did.
While great for viewers, this can very easily turn around and bite the company in the ass. For example, if Netflix gets viewers interested in classic cinema, they are likely to start migrating to other services with far better cinephile collections. Filmstruck is probably the best-known of these, but here in LA and possibly in your town as well, the public library offers a free streaming service which includes the Criterion Collection, a catalog which Netflix can't possibly compete against.
So chances are the business not only needs to change the queue; it needs to direct the viewers toward certain programs.
Before we pursue this any further, we need to remember that convincing a person to watch a movie or a TV show that he or she is unfamiliar with has never been easy and has grown far more difficult in recent years. For the first three or so decades of television as a mass medium (let's call it 1950 to 1980 just to have some round numbers), you basically had three networks to choose from. Some minor players popped up – – PBS, independent stations, a few abortive attempts at a fourth network – – but most people spent most of their time watching CBS, NBC, or ABC (usually in that order). Simply getting a show on the air guaranteed enough of an audience for word to get around.
If you want to build the IP value of a television show in 2018, the best way to do it is probably still going the route of a prime time network run and a wide syndication release, but that paradigm is clearly fading. Despite what you may have heard, no one has come up with anything close to replacing it. As a result, companies are either relying on established properties that achieved high name recognition status under the old paradigm or are desperately trying to prop up new properties with dump trucks full of marketing and PR money. Under these circumstances, the suggestion that the viewership for the programming being produced by the streaming services is driven by recommendation engines should be taken with extreme skepticism.
Furthermore, at the risk of being cynical, it would probably be a good idea to approach any story about the role of recommendation algorithms in the world of online video with the assumptions that the sources for the stories have a strong incentive spin them a certain way and that there is a very good chance that those sources don't know what they're doing.
I suspect the most popular alternative would be simply measuring how often customers accept the recommendation, but that metric is deeply problematic even under the best of circumstances and is almost worthless if approached naïvely.
A big part of the problem is that customers almost always walk into the door with some kind of mental queue, albeit often a vague and incomplete one. A recommendation algorithm that does not change that queue in a nontrivial way is a failure, a complete waste of time and money.
[I almost veered off into a discussion of how models and algorithms interact in this situation, but I'm trying to stay on topic and that's a subject that really needs a serious post or perhaps even thread of its own.]
Now we get to the next level of complexity. If you fail to change the queue you accomplish nothing, but if you do change the queue you still don't necessarily accomplish anything of business value. This is where we need to start getting specific about our objectives because there are some perfectly reasonable but contradictory choices to consider
If the sole concern is serving the customer (and serving the customer is never the sole concern) then the goal might be to get the viewers to watch and give a high ratings to shows they were previously unaware of. Another interesting related metric might be to look at before and after ratings, comparing how much they expected to enjoy a program with how much they actually did.
While great for viewers, this can very easily turn around and bite the company in the ass. For example, if Netflix gets viewers interested in classic cinema, they are likely to start migrating to other services with far better cinephile collections. Filmstruck is probably the best-known of these, but here in LA and possibly in your town as well, the public library offers a free streaming service which includes the Criterion Collection, a catalog which Netflix can't possibly compete against.
So chances are the business not only needs to change the queue; it needs to direct the viewers toward certain programs.
Before we pursue this any further, we need to remember that convincing a person to watch a movie or a TV show that he or she is unfamiliar with has never been easy and has grown far more difficult in recent years. For the first three or so decades of television as a mass medium (let's call it 1950 to 1980 just to have some round numbers), you basically had three networks to choose from. Some minor players popped up – – PBS, independent stations, a few abortive attempts at a fourth network – – but most people spent most of their time watching CBS, NBC, or ABC (usually in that order). Simply getting a show on the air guaranteed enough of an audience for word to get around.
If you want to build the IP value of a television show in 2018, the best way to do it is probably still going the route of a prime time network run and a wide syndication release, but that paradigm is clearly fading. Despite what you may have heard, no one has come up with anything close to replacing it. As a result, companies are either relying on established properties that achieved high name recognition status under the old paradigm or are desperately trying to prop up new properties with dump trucks full of marketing and PR money. Under these circumstances, the suggestion that the viewership for the programming being produced by the streaming services is driven by recommendation engines should be taken with extreme skepticism.
Furthermore, at the risk of being cynical, it would probably be a good idea to approach any story about the role of recommendation algorithms in the world of online video with the assumptions that the sources for the stories have a strong incentive spin them a certain way and that there is a very good chance that those sources don't know what they're doing.
Tuesday, August 21, 2018
Codetermination
This is Joseph
The new proposal by Elizabeth Warren to introduce codetermination to the United States has certainly engendered a lot of debate. It led to attacks on the plan as a way to nationalize industries. It also led to questions about whether the same goals could be achieved by empowering unions. A lot of this seems to be an innate resistance to any form of wealth redistribution, which is a tough line to hold as wealth inequality increases and we slide into a crisis of affordable housing.
However, I suspect that there is an even bigger piece that is being overlooked. Just consider these two provisions:
Now other policy ideas could achieve the same goals. But this is a rather clever assault on the principle agent problem as regards to corporate free speech.
The new proposal by Elizabeth Warren to introduce codetermination to the United States has certainly engendered a lot of debate. It led to attacks on the plan as a way to nationalize industries. It also led to questions about whether the same goals could be achieved by empowering unions. A lot of this seems to be an innate resistance to any form of wealth redistribution, which is a tough line to hold as wealth inequality increases and we slide into a crisis of affordable housing.
However, I suspect that there is an even bigger piece that is being overlooked. Just consider these two provisions:
40 percent of the directors would be elected by the company’s workforce, with the other 60 percent elected by shareholders.
Corporate political activity would require the specific authorization of both 75 percent of shareholders and 75 percent of board members.This is a nice way to keep corporate free speech but to prevent a small class of managerial executives from controlling the political agenda of the companies. A company is still free to be an advocate for political positions, but they need to build broad stakeholder consensus that this position is in the best interests of the business. For obvious things that impact the business, like bankruptcy laws for banks, this is an easy sell.
Now other policy ideas could achieve the same goals. But this is a rather clever assault on the principle agent problem as regards to corporate free speech.
Monday, August 20, 2018
A few points to remember about the recent (ongoing?) Meltdown of Elon Musk.
[If you haven't been keeping up with the story, here's a good write-up from the superior Times.]
1. First off, we all know that this story about spending every waking hour focusing on Tesla production problems is bullshit on any number of levels but most obviously based on the fixed upper bound of hours in a day. Musk leads a very public life, and even limiting ourselves to that public record, we know that he finds lots of time to support a multi-hour-a-day social media habit, hang out with his popstar girlfriend, play around with miniature flamethrowers and other toys, confidently make dubious proposals involving Hyperloops and tunneling machines and squeeze in a profoundly embarrassing trip to Thailand.
1a. As always, it is important to remember that Elon Musk has a long history of making questionable claims of extraordinary personal accomplishment. From reading the entire encyclopedia as a child to studying martial arts and taking out the school bully with a single punch to spending a month subsisting on hotdogs and oranges to test his self-discipline and ability to live frugally. These stories are difficult to believe individually; taken together, they strain credulity. Despite this, reporters routinely write "Elon Musk did – –" instead of "Elon Musk said he did – –".
1b. Also important to note that, like many conmen and manipulative people, Musk often relies on appeals for sympathy and personal "revelations" to win over listeners and breakdowns skepticism. This is particularly on display in the infamous Rolling Stone cover story.
1c. None of this is all that important for the state of Tesla, but it tells us a great deal about the state of 21st century journalism.
2. Even if the every-waking-hour story were true, it would hardly be addressing the real problems the company is having. Musk has no relevant experience in the field of auto production and he is, as previously mentioned, a terrible engineer. Even as a way of motivating the troops (an area where Musk has shown a natural gift) this is probably a waste of time thanks to his increasingly toxic relationship with Tesla workers. If he actually wanted to solve this problem, rather than pitching a tent on the production line floor, he would be on a plane trying to poach top talent from other companies.
3. The story does, however, make sense when viewed through a framework of magical heuristics. Keep in mind that one of the primary heuristics, especially when dealing with Musk, is the belief that certain chosen ones have the ability to will things into existence. Though this is always couched in the language of science and business, the real precedents are Merlin conjuring a dragon or Yoda levitating a spacecraft. Closely related to this is the heuristic of destiny. Instead of Merlin and Yoda, think Arthur or Harry Potter or perhaps even a more explicitly messianic figure. Musk repeatedly says that, though his burdens are great, he accepts them because there is no one else who can take them from him.
Keeping things in perspective. This interview will not decide Elon Musk's role in Tesla (his tweet about taking the company private might, but that's a topic for another post). The survival of Tesla will not determine the fate of electric vehicles. And while they have an important part to play, electric vehicles will not be the most important factor in the futures of either transportation or climate change. This isn't that big of a story.
It is, however, an instructive one. We live in an age when massive rivers of money are diverted based on hype, credulous journalists and investors, and folklore passing itself off as business plans. The results were never going to be pretty.
1. First off, we all know that this story about spending every waking hour focusing on Tesla production problems is bullshit on any number of levels but most obviously based on the fixed upper bound of hours in a day. Musk leads a very public life, and even limiting ourselves to that public record, we know that he finds lots of time to support a multi-hour-a-day social media habit, hang out with his popstar girlfriend, play around with miniature flamethrowers and other toys, confidently make dubious proposals involving Hyperloops and tunneling machines and squeeze in a profoundly embarrassing trip to Thailand.
1a. As always, it is important to remember that Elon Musk has a long history of making questionable claims of extraordinary personal accomplishment. From reading the entire encyclopedia as a child to studying martial arts and taking out the school bully with a single punch to spending a month subsisting on hotdogs and oranges to test his self-discipline and ability to live frugally. These stories are difficult to believe individually; taken together, they strain credulity. Despite this, reporters routinely write "Elon Musk did – –" instead of "Elon Musk said he did – –".
1b. Also important to note that, like many conmen and manipulative people, Musk often relies on appeals for sympathy and personal "revelations" to win over listeners and breakdowns skepticism. This is particularly on display in the infamous Rolling Stone cover story.
1c. None of this is all that important for the state of Tesla, but it tells us a great deal about the state of 21st century journalism.
2. Even if the every-waking-hour story were true, it would hardly be addressing the real problems the company is having. Musk has no relevant experience in the field of auto production and he is, as previously mentioned, a terrible engineer. Even as a way of motivating the troops (an area where Musk has shown a natural gift) this is probably a waste of time thanks to his increasingly toxic relationship with Tesla workers. If he actually wanted to solve this problem, rather than pitching a tent on the production line floor, he would be on a plane trying to poach top talent from other companies.
3. The story does, however, make sense when viewed through a framework of magical heuristics. Keep in mind that one of the primary heuristics, especially when dealing with Musk, is the belief that certain chosen ones have the ability to will things into existence. Though this is always couched in the language of science and business, the real precedents are Merlin conjuring a dragon or Yoda levitating a spacecraft. Closely related to this is the heuristic of destiny. Instead of Merlin and Yoda, think Arthur or Harry Potter or perhaps even a more explicitly messianic figure. Musk repeatedly says that, though his burdens are great, he accepts them because there is no one else who can take them from him.
Keeping things in perspective. This interview will not decide Elon Musk's role in Tesla (his tweet about taking the company private might, but that's a topic for another post). The survival of Tesla will not determine the fate of electric vehicles. And while they have an important part to play, electric vehicles will not be the most important factor in the futures of either transportation or climate change. This isn't that big of a story.
It is, however, an instructive one. We live in an age when massive rivers of money are diverted based on hype, credulous journalists and investors, and folklore passing itself off as business plans. The results were never going to be pretty.
Friday, August 17, 2018
Shows with legs – – more background on the Netflix thread (and an excuse for a Friday post of MeTV promos)
[corrected -- Dragon hears "nonrival" as "nonviable"]
I've probably spent too much time on this thread already, but one of these days I ought to do a post on just how problematic the Netflix exclusivity model is, how it goes against the well-established but deeply weird economics of certain nonrival goods. (When increasing supply increases demand, things get very strange very quickly. Insert highly appropriate Twilight Zone reference here.)
For now, we'll focus on one specific corner of the topic, television shows that maintain a viable and highly lucrative syndication presence for decades, often actually growing in popularity since their initial run. I'm not talking about programs that form the basis for reboots or reunions or sequels, but of shows where the original episodes continue to draw large viewerships.
We could have a long interesting discussion on the psychology behind the appeal of the familiar. You probably coulld even come up with a few pretty good research topics on the subject, but I want to keep the focus on the business side. Television became a national mass medium in the late 40s. Within its first decade, it started producing shows like I Love Lucy and Perry Mason, now both over 60 years old, which would continue to maintain a surprisingly steady audience to this day.
The return on investment of these programs is stunning. With a handful of exceptions, all of the following shows turned a nice profit during their original network run. Everything since is gravy.
I Love Lucy
Perry Mason
Leave It to Beaver
The Twilight Zone
The Andy Griffith Show
The Adams family/the Munsters
Bewitched/I Dream of Jeannie
Star Trek
MASH
Columbo
Taxi
The Cosby show (until recently)/Roseann (until recently)
Golden Girls
Cheers/Frasier
Seinfeld
Friends
.
The Simpsons
.Married with Children
Law and Order/CSI/NCIS
And many others.
Nobody understands the economics of these shows better than Weigel Broadcasting, the company that almost single-handedly developed the entire terrestrial superstation segment of the industry. One of the keys to their extraordinary ratings success has been their knowledgeable and affectionate treatment of the material and their respect for their audience.
From Chicago Magazine:
Nowhere does this come through more than in their stations promos.
Here's Carl Reiner's reaction to one.
.
Here's more from Feder on Weigel's promos.
And here are a few more favorites to close the week.
[And yes, I believe that may be the same set.]
I've probably spent too much time on this thread already, but one of these days I ought to do a post on just how problematic the Netflix exclusivity model is, how it goes against the well-established but deeply weird economics of certain nonrival goods. (When increasing supply increases demand, things get very strange very quickly. Insert highly appropriate Twilight Zone reference here.)
For now, we'll focus on one specific corner of the topic, television shows that maintain a viable and highly lucrative syndication presence for decades, often actually growing in popularity since their initial run. I'm not talking about programs that form the basis for reboots or reunions or sequels, but of shows where the original episodes continue to draw large viewerships.
We could have a long interesting discussion on the psychology behind the appeal of the familiar. You probably coulld even come up with a few pretty good research topics on the subject, but I want to keep the focus on the business side. Television became a national mass medium in the late 40s. Within its first decade, it started producing shows like I Love Lucy and Perry Mason, now both over 60 years old, which would continue to maintain a surprisingly steady audience to this day.
The return on investment of these programs is stunning. With a handful of exceptions, all of the following shows turned a nice profit during their original network run. Everything since is gravy.
I Love Lucy
Perry Mason
Leave It to Beaver
The Twilight Zone
The Andy Griffith Show
The Adams family/the Munsters
Bewitched/I Dream of Jeannie
Star Trek
MASH
Columbo
Taxi
The Cosby show (until recently)/Roseann (until recently)
Golden Girls
Cheers/Frasier
Seinfeld
Friends
.
The Simpsons
.Married with Children
Law and Order/CSI/NCIS
And many others.
Nobody understands the economics of these shows better than Weigel Broadcasting, the company that almost single-handedly developed the entire terrestrial superstation segment of the industry. One of the keys to their extraordinary ratings success has been their knowledgeable and affectionate treatment of the material and their respect for their audience.
From Chicago Magazine:
Still under family ownership more than 40 years after its inception, Weigel Broadcasting stands as the last independent television outfit in the city and one of the last in the country. So while the network affiliates in town (WBBM, WMAQ, WLS) blare forth with new, expensively created fare, Weigel’s channels beam with Sabin’s intuition and pluck. “Neal is doing the best television in Chicago with the least amount of resources and the toughest obstacles,” says the former Chicago Sun-Times columnist and local television/radio sage Robert Feder.
Nowhere does this come through more than in their stations promos.
Here's Carl Reiner's reaction to one.
.
Here's more from Feder on Weigel's promos.
And here are a few more favorites to close the week.
[And yes, I believe that may be the same set.]
Thursday, August 16, 2018
Trump and testifying
This is Joseph
There have been a lot of calls for president Trump to testify about his actions in office, for some pretty understandable reasons. However, this seems to misunderstand the way that the modern US criminal justice system is designed. Here is a great twitter thread explaining it. My favorite part is:
Making incorrect statements has been a source of many easy convictions from what I can tell.
This is not to say that this is a good system. I am especially sad that federal agents may lie and suspects face criminal jeopardy for any statements that they make that are not perfect. Heck, it may even be an innocent error on the part of the investigator that they think you are lying about. Guess who faces charges for this innocent error?
So I do think that I would like to hear the president defend his actions and I worry greatly about the optics here. But saying "just tell the truth" has the risk of misleading people into thinking that they shouldn't consult extensively with a lawyer first (which to be fair, the president does have one) and that the situation is not one of extreme danger. I was impressed with how Peter Strzok refused to commit to exact numbers without his case files. Because if he had made a mistake, recollecting events over a year ago, he might have been charged with perjury or making a "material false statement". He was a law enforcement professional, which Donald Trump is not. Not are you and I.
So I do think that we should be very clear about how it is not incriminating to avoid interviews with law enforcement (the point of the 5th amendment), especially given the way that the legal system works. Has anybody not seen this video yet?
So I do want to have explanations for all sorts of things. But I can totally understand why agreeing to an interview with a prosecutor is both a) risky and b) enormously time consuming (in terms of careful and relentless preparation).
There have been a lot of calls for president Trump to testify about his actions in office, for some pretty understandable reasons. However, this seems to misunderstand the way that the modern US criminal justice system is designed. Here is a great twitter thread explaining it. My favorite part is:
Making incorrect statements has been a source of many easy convictions from what I can tell.
This is not to say that this is a good system. I am especially sad that federal agents may lie and suspects face criminal jeopardy for any statements that they make that are not perfect. Heck, it may even be an innocent error on the part of the investigator that they think you are lying about. Guess who faces charges for this innocent error?
So I do think that I would like to hear the president defend his actions and I worry greatly about the optics here. But saying "just tell the truth" has the risk of misleading people into thinking that they shouldn't consult extensively with a lawyer first (which to be fair, the president does have one) and that the situation is not one of extreme danger. I was impressed with how Peter Strzok refused to commit to exact numbers without his case files. Because if he had made a mistake, recollecting events over a year ago, he might have been charged with perjury or making a "material false statement". He was a law enforcement professional, which Donald Trump is not. Not are you and I.
So I do think that we should be very clear about how it is not incriminating to avoid interviews with law enforcement (the point of the 5th amendment), especially given the way that the legal system works. Has anybody not seen this video yet?
So I do want to have explanations for all sorts of things. But I can totally understand why agreeing to an interview with a prosecutor is both a) risky and b) enormously time consuming (in terms of careful and relentless preparation).
Wednesday, August 15, 2018
Revisiting the Big Swinging Check Syndrome
From Wikipedia
This is another bit of context to keep in mind when following the Netflix thread.
You may have seen the following news story earlier:
Netflix Acquires ‘The Blacklist’ For $2 Million An Episode
Furthermore, if you hit a few relevant Wikipedia pages and do some quick back-of-the-envelope calculations, you will see it is difficult to see how Netflix can justify this price-per-episode to its shareholders or how Sony could have negotiated it.
It is the nature of television, whether broadcast or streamed, that while the quality has a way of tapering off after a few years, the commercial value tends to increased sharply once a show has established itself. As a rule of thumb, it is not until programs approach 100 episodes that you start talking real money.
Just to put things in perspective, while a long running, syndication friendly, proven hit like NCIS can bring in over $2 million a year. That is very much an upper bound. The Blacklist is years away from having a viable syndication package. Even when it gets there, its serialized elements will probably keep it from making the really big bucks. A forty-four million dollar deal one year into a series run is extraordinary. It is almost inconceivable that Sony would not have settled for much less.
I realize that the following point should be too obvious to bother with, but the object of business is to bring in as much money as possible while sending out as little as possible. If Netflix just paid $44 million for something which they could've gotten for 20 or even 10, this would indicate a fundamental lack of confidence by the management of the company.
Here though, we get into one of the great paradoxes of modern business journalism. From a strictly logical standpoint, the best run businesses are, almost by definition, those which do the most with the least. From an emotional standpoint, journalists are most impressed by those executives who spend extravagantly without apparent hesitation.
For lack of a better word, the willingness to sign large checks is seen as a sign of virility. The bigger the check, the more positive the impression it makes on the reporters covering the story. The soundness of the purchase does not matter, nor does its positive or negative impact on the executive's company.
Netflix has long been something of a joke within the entertainment industry for its tendency to pay more than top dollar for properties that have already been turned down by everybody else and yet Reed Hastings' reputation as a visionary business genius simply grows stronger.
Along similar lines, when Mark Zuckerberg paid an exorbitant amount of money for a company the New York Times simply gushed with enthusiasm, even though it was later revealed that the primary selling point of the company was the fact that the founder threw awesome parties.
Hastings and Zuckerberg may stand out but that doesn't mean they aren't representative. Executives, particularly tech executives, are routinely lauded for big, bold deals, even when those deals make no sense from a traditional business standpoint. Like so much business coverage we see these days, what is presented as rational analysis is a series emotional reactions to charismatic personalities, catchy narratives and the reflected glow of great wealth.
| Season | Episodes | Originally aired | Nielsen ratings | ||||
|---|---|---|---|---|---|---|---|
| First aired | Last aired | Rank | Average viewers (millions) | ||||
| 1 | 22 | September 23, 2013 | May 12, 2014 | 6 | 14.95[14] | ||
| 2 | 22 | September 22, 2014 | May 14, 2015 | 14 | 13.76[15] | ||
| 3 | 23 | October 1, 2015 | May 19, 2016 | 22 | 11.19[16] | ||
| 4 | 22 | September 22, 2016 | May 18, 2017 | 30 | 9.25[17] | ||
| 5 | 22 | September 27, 2017 | May 16, 2018 | 42 | 8.41[18] | ||
This is another bit of context to keep in mind when following the Netflix thread.
Tuesday, September 2, 2014
Netflix and the big swinging check syndrome
Another post in what what was supposed be a fairly brief Netflix thread. I want to move on to other topics, but this latest news item was just too good an example of certain bad trends in journalism to pass up.You may have seen the following news story earlier:
Netflix Acquires ‘The Blacklist’ For $2 Million An Episode
EXCLUSIVE: In what is believed to be the biggest subscription video-on-demand deal for a TV series, I’ve learned that Netflix has acquired the rights to hit NBC drama The Blacklist from Sony Pictures TV in a deal that will net $2 million per episode. I hear Season 1 of the series starring James Spader will debut on the streaming service next weekend. As for future seasons, Netflix usually makes them available shortly after the season finales.For starters, you will notice that the headline is somewhat misleading. Netflix did not "acquire" the Black List in the sense that, say ABC would have. The show will still be running on NBC next year. Nor did it acquire the rights to stream the episodes during the regular season; those will presumably stay with Hulu. What Netflix did acquire was the right to stream the previous year's episodes.
Sony TV first tested the off-network market waters for The Blacklist in March. While other streaming services, like Amazon and Hulu, do joint syndication deals with cable networks, Netflix, which largely pioneered the series SVOD business, insists on getting first dibs. Twentieth Television just recently sold New Girl to TBS and MTV, more than an year after prior seasons of the Fox series landed at Netflix in a rich deal, said to be worth $900,000 an episode. Like was the case with New Girl, I hear Sony TV has the right to also sell The Blacklist in cable and broadcast syndication, with Netflix getting an exclusive first window. The $2 million per-episode fee is said to be the biggest for an off-network series paid by Netflix (or any others streaming company), eclipsing previous record holder, AMC’s The Walking Dead, whose sale price to Netflix is believed to be $1.35 million per episode.
Furthermore, if you hit a few relevant Wikipedia pages and do some quick back-of-the-envelope calculations, you will see it is difficult to see how Netflix can justify this price-per-episode to its shareholders or how Sony could have negotiated it.
It is the nature of television, whether broadcast or streamed, that while the quality has a way of tapering off after a few years, the commercial value tends to increased sharply once a show has established itself. As a rule of thumb, it is not until programs approach 100 episodes that you start talking real money.
Just to put things in perspective, while a long running, syndication friendly, proven hit like NCIS can bring in over $2 million a year. That is very much an upper bound. The Blacklist is years away from having a viable syndication package. Even when it gets there, its serialized elements will probably keep it from making the really big bucks. A forty-four million dollar deal one year into a series run is extraordinary. It is almost inconceivable that Sony would not have settled for much less.
I realize that the following point should be too obvious to bother with, but the object of business is to bring in as much money as possible while sending out as little as possible. If Netflix just paid $44 million for something which they could've gotten for 20 or even 10, this would indicate a fundamental lack of confidence by the management of the company.
Here though, we get into one of the great paradoxes of modern business journalism. From a strictly logical standpoint, the best run businesses are, almost by definition, those which do the most with the least. From an emotional standpoint, journalists are most impressed by those executives who spend extravagantly without apparent hesitation.
For lack of a better word, the willingness to sign large checks is seen as a sign of virility. The bigger the check, the more positive the impression it makes on the reporters covering the story. The soundness of the purchase does not matter, nor does its positive or negative impact on the executive's company.
Netflix has long been something of a joke within the entertainment industry for its tendency to pay more than top dollar for properties that have already been turned down by everybody else and yet Reed Hastings' reputation as a visionary business genius simply grows stronger.
Along similar lines, when Mark Zuckerberg paid an exorbitant amount of money for a company the New York Times simply gushed with enthusiasm, even though it was later revealed that the primary selling point of the company was the fact that the founder threw awesome parties.
Hastings and Zuckerberg may stand out but that doesn't mean they aren't representative. Executives, particularly tech executives, are routinely lauded for big, bold deals, even when those deals make no sense from a traditional business standpoint. Like so much business coverage we see these days, what is presented as rational analysis is a series emotional reactions to charismatic personalities, catchy narratives and the reflected glow of great wealth.
Tuesday, August 14, 2018
What does Netflix really want? Look at the content quadrants
To make sense of the company's approach toward original content, it is useful to think in terms of long-term IP value vs the hype-genic, those programs that lend themselves to promotion by being awards friendly or newsworthy. For example, a talk show would be in the high hype/low IP quadrant. You have famous people saying topical, interesting, sometimes even important things. The articles pretty much write themselves, but in terms of IP, the genre has a shelf life somewhere between a ripe peach and a properly refrigerated gallon of milk. After over 60 years, the best anyone has managed to do is package a few low rated, niche programs for nostalgia channels out of the absolute cream of the genre. The same goes for art films like Beasts of No Nation and documentaries like Icarus (and no, the bump in IP value that Oscar bestows is not worth what Netflix paid to get it). Easy to book the creators on Fresh Air but don't count on any real viewership five years from now.
By comparison, as previously mentioned, the bulk of the Netflix children's slate has no IP value whatsoever. Since journalists outside of the entertainment industry have little interest in the segment and no understanding of its importance, this falls in the low hype/low IP quadrant. Licensing high name recognition kids shows is essential for building the subscriber base, but it is strictly a short-term investment.
When you start categorizing the various shows according to quadrant, keeping in mind the additional goals of building the subscriber base and not spending money too wastefully (even Netflix has its limits), you can't help but notice that the distribution is not at all consistent with what you would expect given the stated goals and strategy of the company.
We can quibble over some of the classifications. Despite its coult status, Stranger Things arguably falls into the high/high quadrant. What about the Crown? Historical costume dramas have sometimes proven to have legs, but the record is mixed and it's difficult to see how Netflix will ever compete with BBC catalog.
Quibbles aside, it is fairly obvious that Netflix has a strong preference for shows that are easy to promote and that a significant portion of their original content budget (and presumably virtually all of their remaining content budget) is going toward shows that contribute little or nothing to the content library. If Netflix really is playing the wildly ambitious, extremely long term game that forms the basis for the company's standard narrative and justifies incredible amounts of money investors are pouring in, then this distribution makes no sense whatsoever. If, on the other hand, the company is simply trying to keep the stock pumped up until they can find a soft landing spot, it makes all the sense in the world.
By comparison, as previously mentioned, the bulk of the Netflix children's slate has no IP value whatsoever. Since journalists outside of the entertainment industry have little interest in the segment and no understanding of its importance, this falls in the low hype/low IP quadrant. Licensing high name recognition kids shows is essential for building the subscriber base, but it is strictly a short-term investment.
When you start categorizing the various shows according to quadrant, keeping in mind the additional goals of building the subscriber base and not spending money too wastefully (even Netflix has its limits), you can't help but notice that the distribution is not at all consistent with what you would expect given the stated goals and strategy of the company.
We can quibble over some of the classifications. Despite its coult status, Stranger Things arguably falls into the high/high quadrant. What about the Crown? Historical costume dramas have sometimes proven to have legs, but the record is mixed and it's difficult to see how Netflix will ever compete with BBC catalog.
Quibbles aside, it is fairly obvious that Netflix has a strong preference for shows that are easy to promote and that a significant portion of their original content budget (and presumably virtually all of their remaining content budget) is going toward shows that contribute little or nothing to the content library. If Netflix really is playing the wildly ambitious, extremely long term game that forms the basis for the company's standard narrative and justifies incredible amounts of money investors are pouring in, then this distribution makes no sense whatsoever. If, on the other hand, the company is simply trying to keep the stock pumped up until they can find a soft landing spot, it makes all the sense in the world.
Monday, August 13, 2018
"Hype is the fog of business" or "you should've known something was up when Netflix bought the billboard company"
How do you decide if a business scenario is viable? What heuristics do you use? How do you form your informal informative priors? How do you decide the necessary conditions for success have been met? Obviously, there are endless possible answers for these questions, but most probably fall under the general categories of looking for things you would associate with success, growth, drive, and competence.
Consider Netflix. The scenario that has been put forth to justify the extraordinary market The company has recently reached is that it has a reasonable chance of achieving a near monopoly of online media distribution. This would seem to be an unbelievable claim but it does have what we might call heuristic support. Things we can all observe which make the arguments seem somewhat more credible.
Though Netflix is a notoriously secretive company, there are still a number of established facts that back up its case. The subscriber base is undeniably large and growing. The company has an excellent reputation and fantastic name recognition. Its shows generate a tremendous amount of buzz and you can't argue with all those Emmys (actually, you can, but more on that later).
There is, however, one piece of context which is absolutely essential for understanding these indicators of success and yet which is routinely underplayed or omitted entirely from the conversation. Netflix has taken hype to a new level.
To be clear, no one would ever suggest that the entertainment industry is a PR virgin.
Planted news stories, awards campaigns, "creative decisions" designed solely to get attention, the fairly open quid pro quo that drives almost all entertainment journalism. None of the techniques that Netflix uses to promote itself are new, but the scale is unprecedented.
Obviously, some of this has to be inferred, but the inferences are all straightforward and largely undeniable. To live in LA particularly west of the 110 and north of the 10 is to be besieged by outdoor advertising for Netflix, particularly around the longer and longer Emmy season when it seems that every available surface will bear the letters F YC.
Likewise, you can draw fairly reliable inferences about PR spending by looking for certain kinds coverage. If you see a cover story, and interview, a what's on [client's name] tonight article or anything that reads like a press release, the odds are very good that it was either initiated or nurtured by a PR agent. Here too, Netflix has taken old approaches and pushed them to a new level.
We also need to take into account indirect PR, business decisions that are nominally made for some other reason, but have the real purpose of generating buzz. Everyone does this, but, once again, Netflix goes much bigger. The company spent hundreds of millions of dollars promoting art-house fare like Beasts of No Nation and documentaries like Icarus (ass far as I know, Netflix was the first to mount “best picture”level Oscar campaigns in the documentary category). Though thrse are deserving films, in terms of viewership, their marketing budgets are impossible to justify, but if your objective is getting journalists to talk about your company, it's money well spent.
Netflix also has a history of commissioning award bait shows going all the way back to House of Cards. Emmy awards play a special role in this process. Within the industry, they aren't taken all that seriously. Their impact on viewership has long been question and increasingly who gets the nominations and awards is seen as a function of who's willing to pony up the big campaign budgets. (This point was beautifully illustrated when Tatiana Maslany couldn't get a nomination despite incredible buzz and reviews. It was only after the snubs became news that she broke through).
None of this is intended as a criticism of Netflix. Marketing and self-promotion are a part of the game and you can't blame the company for being so good at it. The ads and carefully cultivated press coverage help drive subscriber growth and support the narrative that makes the skyhigh stock prices possible. You can hardly fault management for increasing revenue and maintaining market cap. You can, however, blame analysts and journalists who fail to recognize the impact of this unprecedented marketing and PR push and who casually throw out references to buzz and Emmy awards as if they meant anything at all in this context.
Friday, August 10, 2018
A few moments with Dick Cavett
I've got a post coming up about the IP value of various TV genres. At the very bottom are talk shows which can pull great numbers initially but which have been almost worthless when syndicated or repackaged as anything more than low rated nostalgia fare.
That's not to say that some of these shows aren't worth watching (lots of good TV isn't particularly marketable TV). Cavett holds up remarkably well, especially when he had a geenuinely interesting guest.
Here are a few notable examples.
The Dick Cavett Show Richard Pryor 12 16 85
Richard Burton on The Dick Cavett Show July 1980
This Marlon Brando interview is extraordinary and is one of the best examples you'll ever find of handling an uncooperative subject.
Truman Capote
Thursday, August 9, 2018
Health Care and the poor
This is Joseph. And I am confused.
Jon Chait:
It also suggests that we are heading towards some sort of tipping point in the United States. Health care costs are getting higher and higher. Regulations generally prohibit cheap substitution (you can't create a clinic with non-MD/RNs to service those who can afford nothing more) in the health care market and pricing transparency is low, making comparison shopping hard. It seems like a mild regulatory approach is unpopular and unsustainable, given the political polarization.
One way or another, there are some interesting times ahead.
Jon Chait:
The state-level Republican crusade to deny the Medicaid expansion also hurt insurers. Medicaid wound up soaking up costly patients, freeing insurers to cover a healthier population. (Two studies found this result.) That’s why, Solomon confirmed to me, “in most states [insurers] do support expansion in my experience.” The clear and consistent pattern is one of Republicans repeatedly threatening insurers, to the point of withholding payments they were legally owed, in order to prevent poor and sick people from getting insurance. It is bizarre that Ackerman concludes that the GOP doesn’t actually care about denying insurance to the poor and sick (a goal it has in fact pursued fervently) and instead cares about profits for insurers (a goal it has in fact undermined relentlessly).This really does seem to be correct. If profits were the goal, then this strategy seems to be an odd way to go about it. It is true that there could be a larger goal in mind that suggests short term pain to keep health care profits high, but it sure is not a direct link.
It also suggests that we are heading towards some sort of tipping point in the United States. Health care costs are getting higher and higher. Regulations generally prohibit cheap substitution (you can't create a clinic with non-MD/RNs to service those who can afford nothing more) in the health care market and pricing transparency is low, making comparison shopping hard. It seems like a mild regulatory approach is unpopular and unsustainable, given the political polarization.
One way or another, there are some interesting times ahead.
Wednesday, August 8, 2018
The dangers of Twitter
This is Joseph, stealing Mark's normal beat.
It seems Elon Musk tweeted that he had buyers lined up to take his company private at $420 per share. Really.
Josh Marshall points out that this could go wrong.
I think that this is a challenge of social media platforms, where this type of news can be rapidly sent out without the normal slow vetting of traditional media outlets. This might all be fine, but it seems to create a lot of possible problems for Tesla, which may not have been obvious at the moment that the tweet was made.
It seems Elon Musk tweeted that he had buyers lined up to take his company private at $420 per share. Really.
The surprise tweet comes as Mr. Musk’s long-combative stance against Tesla’s short sellers has grown testier in recent months. He has repeatedly used Twitter to chide investors who are betting against his company, sometimes offering vague positive outlooks for the company that seemed to boost the stock, hurting short sellers’ positions.This highlights one of the challenges of social media, as it is not impossible that this news could affect short sales. So it is very, very important that he really have this funding lined up.
Josh Marshall points out that this could go wrong.
I think that this is a challenge of social media platforms, where this type of news can be rapidly sent out without the normal slow vetting of traditional media outlets. This might all be fine, but it seems to create a lot of possible problems for Tesla, which may not have been obvious at the moment that the tweet was made.
Tuesday, August 7, 2018
SCOTUS game theory
This is Joseph
I want to talk about this tweet in terms of game theory:
What I am interested in is whether this would ever be a good strategy.
So, first of all, would this work. Well, if the "doctrine" was followed, it would allow whoever won the 2020 election to name Ruth Bader Ginsburg's replacement. So the main advantage, in this scenario, would be to make the Republican party look good, by showing how they are a party of principal. That seems an odd goal for a sitting justice to try and make a partisan organization look good as a goal. The open seat would motivate liberals, but it would also motivate conservatives and act as a unifying force in both coalitions. I am not sure that this helps liberals, on net. It might hold together the conservative coalition, under tough conditions, even more so.
If the "doctrine" was ignored then there would be another Republican nominee to the Supreme Court, meaning six of nine justices were nominated by Republicans. Insofar as this matters, and people act like it matters a lot, then it is a major win for the Republicans. After all, what is the point of delaying the appointment of a new Supreme Court Justice at the end of Obama's term if there wasn't some sort of benefit?
It also seems to deeply misread the psychology of the justice involved. If she wanted to lock in her side, why would she not have retired at 81 (like Anthony Kennedy) and given Obama a clear shot at naming a successor? Clearly, she feels competent to do her job (and the evidence that she can is compelling) and doesn't like to play these games with timing her retirement. Fair enough. If she did decide to play games, why would she do it in order to maximize the gains of the Republican party at the cost of liberals? You can say what you want about Donald Trump, but he has not been a noted feminist firebrand and a lifetime of fighting for gender equality seems to position one to not want to go out of one's way to hand him another nomination.
Finally, the five year plan actually makes a lot of sense for a justice who is trying to be non-political. This has her retiring in 2022 or 2023. Long enough before the end of the next president's term that confirmation should happen if it is possible at all. It's well before anybody has any idea who the next president will be and so it cannot be motivated by partisan considerations. The opening is still part of the election discourse, but not in an immediate way that pulls oxygen away from the issues. It's a good plan.
So I don't see how switching to a 2020 retirement would be a more optimal strategy.
Mark?
I want to talk about this tweet in terms of game theory:
What I am interested in is whether this would ever be a good strategy.
So, first of all, would this work. Well, if the "doctrine" was followed, it would allow whoever won the 2020 election to name Ruth Bader Ginsburg's replacement. So the main advantage, in this scenario, would be to make the Republican party look good, by showing how they are a party of principal. That seems an odd goal for a sitting justice to try and make a partisan organization look good as a goal. The open seat would motivate liberals, but it would also motivate conservatives and act as a unifying force in both coalitions. I am not sure that this helps liberals, on net. It might hold together the conservative coalition, under tough conditions, even more so.
If the "doctrine" was ignored then there would be another Republican nominee to the Supreme Court, meaning six of nine justices were nominated by Republicans. Insofar as this matters, and people act like it matters a lot, then it is a major win for the Republicans. After all, what is the point of delaying the appointment of a new Supreme Court Justice at the end of Obama's term if there wasn't some sort of benefit?
It also seems to deeply misread the psychology of the justice involved. If she wanted to lock in her side, why would she not have retired at 81 (like Anthony Kennedy) and given Obama a clear shot at naming a successor? Clearly, she feels competent to do her job (and the evidence that she can is compelling) and doesn't like to play these games with timing her retirement. Fair enough. If she did decide to play games, why would she do it in order to maximize the gains of the Republican party at the cost of liberals? You can say what you want about Donald Trump, but he has not been a noted feminist firebrand and a lifetime of fighting for gender equality seems to position one to not want to go out of one's way to hand him another nomination.
Finally, the five year plan actually makes a lot of sense for a justice who is trying to be non-political. This has her retiring in 2022 or 2023. Long enough before the end of the next president's term that confirmation should happen if it is possible at all. It's well before anybody has any idea who the next president will be and so it cannot be motivated by partisan considerations. The opening is still part of the election discourse, but not in an immediate way that pulls oxygen away from the issues. It's a good plan.
So I don't see how switching to a 2020 retirement would be a more optimal strategy.
Mark?
Monday, August 6, 2018
Kevin Drum on Medicare for all
This is Joseph
This as a remarkable development:
There is some nuance here:
At a certain point you need to wonder just how large would be disincentive effects be (it is a big world), could we change drug patent rules to mitigate the impact, and could we invest in drug research directly, say via the national Institutes of Health. Because half of that 2 trillion dollars in savings might well fund the best research environment imaginable . . .
The biggest problem is structural -- how do you redirect private health care spending into taxes?
One thought that I increasing wonder about is whether this should be a state level program. Canada requires all provinces to have a Medicare program, but allows significant differences between provinces. I am not convinced a single federal program will drive innovation as quickly as 50 separate states all trying to puzzle out the best way to make it work.
This as a remarkable development:
The libertarians at the Mercatus Center did a cost breakdown of Bernie Sanders’ Medicare for All plan and concluded that it would save $2 trillion during its first ten years(via Kevin Drum)
There is some nuance here:
There is the rub. The federal government is going to spend a lot more money on health care, but the country is going to spend about the same.
“Lower spending is driven by lower provider payment rates, drug savings, and administrative cost savings,” Yevgeniy Feyman at the right-leaning Manhattan Institute told me. “It’s not clear to what extent those savings are politically feasible, and socially beneficial.”
(One concern is whether cuts to prescription drug spending would discourage medical innovation. It’s simply hard to know — Mercatus projects a $61 billion drop in drug spending in one year, but there would still be hundreds of billions of dollars spent annually on medications.)But, that said, it is a remarkable inflection point. Sure, making health care less profitable would slow the pace of innovation but there are already some issues with how the market focuses innovation. There is a lot of social good in new generations of antibiotics, but these tend to be underdeveloped in a system that rewards chronic disease medication discovery.
At a certain point you need to wonder just how large would be disincentive effects be (it is a big world), could we change drug patent rules to mitigate the impact, and could we invest in drug research directly, say via the national Institutes of Health. Because half of that 2 trillion dollars in savings might well fund the best research environment imaginable . . .
The biggest problem is structural -- how do you redirect private health care spending into taxes?
One thought that I increasing wonder about is whether this should be a state level program. Canada requires all provinces to have a Medicare program, but allows significant differences between provinces. I am not convinced a single federal program will drive innovation as quickly as 50 separate states all trying to puzzle out the best way to make it work.
Friday, August 3, 2018
Really? This was actually advanced as an argument?
This is Joseph
How does one even parody this?
What if we reduced parking in cities? Would that not make people drive less? Or banning cars altogether?
I mean if the goal is to get people out of cars, why would decreasing fuel efficiency standards (to make the gas cost per mile less) even be in the top ten reason?
Now I know that the argument is also to make the cars cheaper (so people buy kore modern cars) but the wonderful thing about a super-high gas tax is that you could use some of that money to reduce tariffs and sales tax on new cars. Everyone wins.
The thing with cars is that there is an efficiency (in terms of transport) and safety trade-off. We can rethink it, but this seems like the worst policy decision to make in order to accomplish the stated goal.
How does one even parody this?
At the same time, the draft says that people will drive less if their vehicles get fewer miles per gallon, lowering the risk of crashes.If the goal is to discourage driving by making it more expensive, there are an entire suite of carbon tax, gas tax, and congestion tax proposals that I think bear looking into. These will recue driving and make everyone safer. Or what about cheap bus service? These big and heavy vehicles are pretty safe for their passengers and tend to have better safety records than single occupancy vehicles.
What if we reduced parking in cities? Would that not make people drive less? Or banning cars altogether?
I mean if the goal is to get people out of cars, why would decreasing fuel efficiency standards (to make the gas cost per mile less) even be in the top ten reason?
Now I know that the argument is also to make the cars cheaper (so people buy kore modern cars) but the wonderful thing about a super-high gas tax is that you could use some of that money to reduce tariffs and sales tax on new cars. Everyone wins.
The thing with cars is that there is an efficiency (in terms of transport) and safety trade-off. We can rethink it, but this seems like the worst policy decision to make in order to accomplish the stated goal.
Subscribe to:
Posts (Atom)