Girls on HBO

The Web was so abuzz with commentary on the new HBO series Girls, I decided to watch the first episode. (The folks at Vulture have a good summary of the madness.) The show had its moments, though I’m not going to become a devoted fan. I can see why the show appeals, at least in theory, to a certain subset of television viewers. The characters struck me as exaggerated versions of a certain subset of the people I went to college with.

More than anything, Girls was funny. Not in a laugh-out-loud way; it was on the edge of satire. On one level the two protagonists are convincing representations of real people: children of upper-middle class parents, unaccustomed to adversity and struggling to find a professional foothold with their degree from an expensive East Coast liberal arts college. On another, though, their personalities are so exaggerated they almost come across as satiric. Lena Dunham’s character seems so utterly clueless and so utterly unable to make good decisions, it almost reads as a warning not to major in English.

Circling back to the reason I bothered to sit through the first episode, none of the glaring injustices cited by critics stood out to me. Had I not read — or, more accurately, read of — one critic’s whining about the show’s lack of ethnic diversity, I doubt I would have noticed the lack thereof. At least in my limited experience, it is not terribly unusual for a pair of two very close female friends not to include someone black or latina.

As I say, I’m not going to become a regular viewer. It was, however, very self-affirming to watch Girls drinking tea in my prewar Back Bay apartment. My personal and professional deficits could be much worse.

Tightwad Zeros

The Wall Street Journal has an article in tomorrow’s paper about employees who go beyond using their expense accounts judiciously, to the point of being dubbed “tightwads” by the author of the piece. While I’m not going to come out in favor of unchecked spendthrift debauchery, it drives me crazy when people go beyond sensible budgeting and pinch pennies. Yes, one needs to maintain some savings for future goals, emergencies, retirement and all the rest. But, at least for people comfortably ensconced in a professional position — such as most readers of the Wall Street Journal — the bizarre, ostensibly ingenious methods of financial salvation most “experts” preach drive me crazy.

An especially tall person flying from New York to India multiple times a month in coach probably suffers mental and physical losses whose monetary value far exceeds the dollar difference between a coach and roomier business class seat. His employer might save a dollar or two on travel expenses, but they probably lose it to the hit to the employee’s productivity. Or, for someone making a $100,000 salary, indulging his or her base instincts and finally springing for that Netflix subscription for $8 a month isn’t going to tip the balance from financial success to financial ruin.

Maybe I’m being unfair. The credit card industry has good reason to call people who pay their balances in full every month “deadbeats”, and I recognize many average people in this country haven’t got a job, let alone such vexing tradeoffs as the one between a fancy sweater and a piece of home theater gear. But even for people who have net negative cash flow every month, just cutting out a few trips to Starbucks every week isn’t going to solve the problem long term. More likely they need to tackle something bigger and more structural. Perhaps that $100,000 summer rental in Montauk, or the Mercedes-Benz GL550 for hauling home groceries.

Money may not necessarily buy happiness. But it certainly can, or, at the very least, has far more potential to do so when exchanged for goods and services than it does sitting in a bank account. I’m not saying go crazy. My apartment still has no sofa. This is just my plea that penny pinchers ought acknowledge that an extra $5 in one’s checking account might be less utile than a $5 carton of artisan slow-churned chocolate hazelnut ice cream.

Measuring Vagueness

Links to interesting articles, products or services seem to make up the bulk of what I post to Twitter. And, as someone with a quantitative bent, I use a bit.ly account to keep tabs on which links I post people click on relatively more often.

As I’m not an especially popular Internet personality, most of my links receive a modest number of clicks. The mean clicks per link over the past 30 days is a mere 3.4; I’m too lazy to write a script to compute the standard deviation, but I’d guess it’s something on the order of three or four.

A few minutes ago, after posting a series of links to my Twitter stream, it occurred to me that, based on my anecdotal observations, the more vague my Twitter posts are, the more clicks the associated links receive. Consider the two most recent links I posted on Twitter. The first, a link to an article about the fast fashion retailer Zara, I teased on Twitter as, “The Economist covers Zara/Inditex. Interesting they’ve suffered in the US because the clothes aren’t ‘roomy’.” It’s fairly specific: one has an idea what to expect after clicking. The second, a link to an article about a weak study tenuously linking chocolate consumption with lower BMI, I previewed with, “This headline needs more qualifiers.” Fairly vague.

It’s far too late for me to test my hypothesis now, but I did start thinking about how to quantify the vagueness of a piece of text. My first thought was to try using the length, but that’s pretty obviously a bad idea. Something like “amazing shark” is far more specific than “something visual”, despite being shorter.

My current thought is to use the norm of the tf-idf weighted vector for each Twitter post. Vagueness stems from the fact the word choice leaves a relatively larger number of possible signified concepts for a given signifier (word) than another. For instance, “he” could refer to all the males one knows, whereas “Eduardo” could refer only to the (highly probably) smaller set of people one knows with the first name Eduardo. A priori, I’m guessing this will be manifested in the frequency with which terms occur in a corpus. Consider “place”, “city”, and “Chicago”. At least according to Google’s book data, their frequency ordering follows my vagueness ordering for the last 200 years.

Tomorrow I’ll see if I can’t give this a go. Maybe I’m wrong in thinking vagueness drives clicks.

Upstairs Downton Titanic

Looking for a National Geographic special about new findings related to the sinking of the Titanic, I inadvertently downloaded the first episode of Julian Fellowes’s latest miniseries project, Titanic, which started airing in Canada last week and premières in the US on ABC sometime in April.

Especially after the 1997 James Cameron film with Leonardo Di-whatever-his-name-is, I have been put off all dramatic adaptations of the Titanic story. But, given Fellowes’s work on Gosford Park and Downton Abbey, I decided to give the first episode the benefit of the doubt.

Fellowes’s multiple perspective narrative is clever. Everyone knows the Titanic will ultimately sink, which has the potential to rob the story of any drama. In light of this, each of the first three episodes follows a different subset of the passengers on board from embarkation at Southampton to the moments after the Titanic has a huge gash ripped into its side by an iceberg. The viewer, then, is left hanging until the fourth episode resolves who lives and who goes down with the ship. Unfortunately, at least in the first episode, the boat hits the iceberg just as the characters begin to develop and become interesting.

The first episode spends most of its time on the Earl and Countess of Manton, and their daughter Lady Georgiana. Lord Manton is a reasonable, wisely paternalistic upper-class gentleman. Lady Manton is an uppity snob. Lady Georgiana is a rebellious teenager looking to cast off the stiffness of the Edwardian upper class. But, without spoiling anything, just as the characters begin to reveal the nuances of their personalities, the boat hits the iceberg. The story becomes nothing more than how what little we know about the characters translates into their scramble to board lifeboats or otherwise escape death.

It feels rushed, and leaves one unsatisfied. It’s still worth watching, but it seems a shame something with the potential to actually be quite interesting is shortchanged by such a short run.

Please Track Me

A few moments ago an email from the folks at J Crew landed in my inbox. I could unsubscribe, but the emails are fairly infrequent, and I shop there enough that notice of a good deal has some value. What does baffle me is the fact they don’t use the massive amount of personal information they have about me to target me. This latest message, for instance, was an advertisement for their “crewcuts” line for children. Considering that most four year olds probably don’t shop online, I have to think this is geared at parents of young children, a group to which I do not belong.

It’s one thing if companies don’t collect data about me, and are then forced to send me everything and see what sticks. In many cases, I would find this preferable. But the proliferation of quantitative analysis — something I must confess to being a part of — and the concomitant desire on the part of companies to collect as much data as possible about their customers, means that most companies — especially those I interact with frequently — do “know” quite a bit about me.

Speaking from experience, it can be difficult to pin down the exact preferences of a given person. It might not be clear from my J Crew purchase history that I dislike green shirts. At the same time, though, given that I have never bought anything from their womenswear collection, it would probably be trivial enough to know not to send me advertisements for a new line of pencil skirts.

So here is my plea to retailers: if there’s no way for me to opt out of your tracking me, please do something with everything you know about me!

Birthday Hacking →

I also have a script that reads the feed of birthdays off my Facebook feed and automatically writes the birthday emails for me, which is just a hack. I feel like leaving the “happy birthday” comment leaves you partial credit, but when you write the email, you get full credit. But I can still automate it off the same data source.

I thought about doing this, but put myself off it thinking it lacked a certain something. But now that at least one person more powerful than me cops to it, maybe I’ll take the idea up again. Of course, I prefer paper cards, so what I really need is a greeting card service with an API for ordering and card creation.

Finding a Router

My wireless router kicked the bucket last week, and I’ve been meandering around the Web ever since trying to find a replacement to order. The hunt has been illuminating, though not in the most important way.

First, only through this process have I come to see just how polluted Google’s search results are. Querying for “wireless router reviews” or something more specific like “linksys e3200 review” yields lots of spam, and very little substance. It is possible no good reviews exist, but that seems unlikely. I played around with DuckDuckGo and some other upstart search engines, but they suffer from the same problems. (My best guess is that their core Web search uses something similar to the PageRank technology Google uses, which companies game to target Google search.) The new Web search product from bit.ly looks very promising in online demos; hopefully it delivers when the product launches to consumers later in the year.

Second, it occurred to me Amazon.com and other high-volume online retailers could leverage their customer data to tease out revealed preferences to make recommendations. Lovely as customer reviews on Amazon.com et al. are, they have serious shortcomings. A “helpfulness” score gives the only measure of reputation, derived from the questionable opinions of other Amazon.com shoppers. No measure of reviewers’ expertise is presented, other than what the text of their reviews projects. Further, most reviewers consider the product in isolation, and, especially in consumer electronics, reviews become stale very quickly.

Armed with product data like return and exchange rates, and consumer data like product discovery paths and purchases histories, it seems plausible Amazon.com and other very high volume online stores could bubble up and promote good products.

Many online stores have the beginnings of this idea in place, such as the “buyers who viewed this also viewed that” feature or popularity charts on Amazon.com. Exploiting more data and a more sophisticated model, though, offers a level of nuance such blunt metrics do not. Consider, for instance, a product that’s more hype than substance. Using something simple like a popularity ranking (i.e., how many units were purchased), this hypothetical product would rocket to the top of charts, with an implicit recommendation from consumers, leading more unsuspecting people to return and become frustrated with the product. Conditioned on return data and the profile of the people buying such products, though, an online retailer could do other consumers a service by recommending a quantitatively superior alternative — and thus make more money since they would be less likely to return it — or, in a worst case, engender trust by helping shoppers steer clear of a bad egg.

Were the data publicly available, it’s certainly how I would go about making my decision about which router to purchase, especially given the dearth of high-quality reviews I’ve managed to find.

Stocks and Flows

Now that Apple has a $500 billion market capitalization, everyone and their mother has decided to find points of comparison. One, though, takes the cake. A story on CNN Money proclaims that “Apple is worth more than Poland”. While Poland is not the United States, it’s a relatively developed economy. If indeed the market valuation of Apple were greater than the total wealth of the Polish population, its enterprises and its government, it would be a remarkable feat indeed.

Some further investigation, however, reveals that the author of the CNN Money piece in question makes some serious errors. The article bases its claim that Apple is worth more than Poland on the fact Apple’s total market capitalization is greater than Poland’s gross domestic product, or GDP.

This comparison makes no sense.

Market capitalization is what’s known in the trade as a stock variable. That is, it represents an accumulation of something over time, for instance a bank balance or the level of water in a bathtub. A company’s market capitalization roughly represents how much value a corporation has “accumulated” over time, at a given moment in time.

GDP, by contrast, is a flow variable. That is, it represents a rate at which something is added or subtracted from a stock. So while a bank balance might represent the stock of accumulated wealth, income represents the flow of monthly income. GDP, then, as a measure of the total value of goods and services produced by a given economy in a year represents the flow of wealth into a given economy.

If it is not already clear, one cannot draw the kind of conclusion the hapless (or attention-seeking) reporter makes comparing a stock and flow. It’s akin to saying Gwen is richer than Steve because Gwen owns $100,000 worth of property, whereas Steve only makes $90,000 in salary each year. Is it still possible Steve is poorer than Gwen? Certainly. But it seems unlikely Gwen is richer than Steve, as he earns almost as much in a given year as Gwen has managed to accumulate in all the time up to now.

A better comparison would be either Apple’s yearly revenue to Poland’s GDP, or Apple’s market capitalization to the total wealth of Poland. Finding data on the total wealth of the Polish people is tricky, but we can take a look at Apple’s revenue compared to Poland’s GDP.

Unsurprisingly, Apple’s income is smaller than the total output of the Polish economy in a given year, though only by a factor of about 3.6. For the 12 month period ending in December 2011, Apple’s revenue was a healthy $128 billion, or so. In 2010, the latest year for which data is available, Poland’s GDP was $467 billion.

Let this serve as a reminder to pay attention to economics professors.

(Source: CNN)

What’s in a Cup?

As anyone who has spent any time with me knows, I really dislike the Imperial units used here in the United States. Converting feet to miles is far more mentally taxing than converting meters into kilometers, assuming that one knows offhand that a mile is equivalent to 5,280 feet to begin with.

Recently, though, I’ve been working in my spare time on a super secret project that involves statistics, recipes and general deliciousness. For fun, I decided it might be interesting to explore the most popular ingredients in my rather modest, but growing, corpus of recipes (on the order of thousands). While these patterns may not be borne out as the corpus grows to a more respectable size, my sampling is random, so these conclusions are not entirely baseless.

While I didn’t find anything particularly telling about ingredients themselves — the top three were black pepper, salt and butter — one trend did leap out at me. Among the set of ingredients listed in my corpus, the top four word bigrams were: “2 tablespoons”, “1/2 cup”, “1 cup” and “1 tablespoon”. Likewise, the most frequently occurring word unigrams were: “1”, “cup”, “2” and “1/2”. This leads me to wonder what we might then be able to say about the creation of recipes, or perhaps the evolution of weights and measures.

On the one hand, perhaps the cup, teaspoon and tablespoon evolved over time to represent the volumes they do because they are ideally suited to the scale at which most people cook (i.e., for two to six people in their nuclear family). I happen to like this theory, if only because it would somewhat lessen my level of frustration with cookbooks and recipes that don’t list ingredient amounts by weight, let alone in metric units. On the other hand, perhaps the causality goes the other way. Maybe people scale recipes they commit to paper (or a digital text file) to use about 1 cup or 1 tablespoon of given ingredients because that’s the unit people expect.

It’s hard to say which theory I give more credence to on the basis of the evidence, especially given how tiny my corpus is. (I will revisit this analysis at a later date!) That said, the dominance of measurements around 1 cup, 1 tablespoon and 1 teaspoon, despite the widespread availability of finer scoops and, to a lesser extent, kitchen scales, suggests there might be something to the “evolution” theory.

This does make me wonder how recipes written for audiences in Continental Europe, where the metric system is widely used, differ from those in my corpus, which draws from exclusively American sources. If the total calories produced by recipes remains more or less the same, and the European recipes use quantities that more or less correspond to 1 cup, 1 tablespoon and 1 teaspoon, it would suggest the cup, tablespoon and teaspoon aren’t totally arbitrary. By contrast, if recipes from European sources tend to choose easily measured quantities — say measurements in round 100s of grams — then perhaps the other direction of causality is more plausible.

I haven’t done any literature review, and I can’t even pin down the field that would study the evolution of units of measure over time. This could be completely baseless and an utterly useless addition to the world’s body of knowledge. But I found it interesting, and hope to follow it up with more.

Coming Apart

Charles Murray’s new book Coming Apart: The State of White America has inspired a fair bit of analysis and a fair number of critiques. It’s one of the more interesting books I’ve read in the last six months, so let me take this as my moment to re-enter the world of long form writing.

As a work, the book has some compelling and not-so-compelling arguments. I won’t dissect the book in great detail, though some of the arguments merit a critique. I do want to spend some time discussing one of the later ideas in the book: that of niceness and moral superiority. Skip the next two paragraphs if you’re only after my thoughts on nonjudgementalism.

First, though: there’s much to laud in Murray’s book. He observes that the “new upper class” has become isolated in expensive, exclusive areas he terms “SuperZIPs”, surrounded largely by other members of the new upper class. This, he argues, diminishes leaders’ capacities to empathize with the rest of the population: the very people they affect through business or legal decisions. As someone likely to become a member of this new élite, and who very often feels utterly out of touch, this resonated. While I opine from time to time on public health issues like obesity and nutrition, it is very difficult for someone who shops at Whole Foods and eats arugula regularly to really grasp the difficulty of feeding a family of five sitting on the edge of the poverty line.

Other arguments are weaker. In one of the later chapters, Murray paints an unflattering picture of the typical inhabitant of a European welfare state. His less-than-rosy outlook on European societies is hardly baseless, but much of it comes across as extreme. Murray spends a great deal of time in the book as a whole bemoaning the secularization of America. Needless to say, he does not have warm feelings for the even more virulent secularization bug sweeping through Europe, writing: “Europeans have broadly come to believe that humans are a collection of activated chemicals that, after a period of time, deactivate” (284). While not overtly derogatory on face, the rather sterile prose and context make such a view on life sound troubled. Given the more plausible argument he makes that community linkages — such as those one might build through religion — Murray makes throughout the book, his focus on religiosity as such an important part of what constitutes a “good” society is confusing. None of the people I’ve met in or from Europe have struck me as quite as nihilistic as Murray makes them out to be. That they may believe their souls won’t ascend into some next world seems rooted more in logic than pessimism or an attitude that does not have any concern for the future.

Now to the meat of my thoughts. Near the very end of the book, Murray begins discussing the ideas of Arnold Toynbee’s book, Civilization on Trial. In particular, Murray latches onto Toynbee’s notion that great civilizations decline because a creative ruling class has its values and ideas diluted by those of the proletariat. Out of this, Murray explores the rise of “nonjudgementalism” — the name more or less contains the definition — among the new upper class. More than anything else in the book, this idea struck a chord.

While my views skew toward the liberal on social issues and neoliberal on economic ones, I have recently come to realize that I do favor some degree of judgementalness (to coin another word), be it in legislative policy or, even better, in acceptable social norms.

Consider the practice of chewing with one’s mouth open. Doing this is so gross and annoying, that no one should ever do this, ever, full stop. Yet many people — even in my little élite bubble where most people come from good-mannered élite bubble families — still do. Perhaps less trivially, I see no reason why we ought not judge morbidly obese people. My limited capacity to empathize with the seriously overweight aside, the existence of overweight empowerment groups is troubling.

It is important to stress the nuance here, both in the table manners and life choices realms. There’s nothing wrong with choosing to act a certain way if thought is involved. I would never judge someone for chewing with his or her mouth open if, for example, he or she genuinely believed in a religion where chewing with one’s mouth open formed an essential part of religious communion. In general, though, I suspect people who put their mastication on view do it out of some combination of laziness, obliviousness and insouciance. Likewise, with respect to nutrition, I’m scarcely advocating we put runway models from Paris Fashion Week on a plinth and declare anyone with a higher BMI embarrassingly fat. Instead, it seems absurd that it might be socially frowned upon for me to judge someone who, despite weighing 400 pounds, continues to eat at McDonald’s three times a day.

Why must I hide my disdain for people who wile away hours obsessing over American Idol under a veneer of niceness? It’s as if we, members of the educational and social élite, have taken ideas of equality so far they no longer represent equality, but rather blind inclusion. Some things are not equal, and deserve to be treated as such.

And, to agree with Murray’s conclusion on this subject, this comes as a detriment to society. While no small amount of effort has put me on a track for membership in the top rungs of the economic and social order, members of this class ought to live up to their position: Zynga CEO Mark Pincus did not make millions of dollars selling virtual crops only to the partners of top law firms. Anyone who has realized any measure of success has society to blame to some degree. It strikes me as irresponsible to blithely stand by and live a socially bankrupt life, or watch the social fabric crumble for fear of being seen as mean or judgmental.