The Mathematical Improbability Of Being The Best

The unheralded gem Eric Falkenstein muses that,

The key is doing the best with what you can, the self-awareness and motivation to develop one's strengths so that your hard work generates a maximum payoff going forward. As Muhammad Ali once said, "You can be the best garbage man or you can be the best model--it doesn't matter as long as you're the best." 'The best' is mathematically improbable, 'really good' generates the same result. If you are really good at your job your day is filled with sincere gratitude by colleagues and customers....

There are two reasons that being "really good" generates the same outputs as being "the best." First, in professions that scale really well, randomness is going to play by far the largest role. This is the implication of Dean Simonton's equal odds rule (here) that the average publication of any particular scientist does not have a statistically significant chance of having more citations (a proxy for impact) than any other scientist's average publication. Furthermore, in Csikszentmihalyi's interviews of highly creative individuals (here), the most commonly mentioned explanation for success was luck. Second, in professions that don't scale well, it is really hard to determine who is the best anyways, because context becomes so important.

So, it's just as useful and probably less stressful to just try to be really good than to try to be the very best.

In Defense Of Searching

Robin Hanson's interesting post today distinguishes between reading that actively attempts to answer a question ("chasing") and reading that broadly looks for new questions ("searching"). I have three thoughts:

1) Let's think of chasing vs searching as a continuum. Reading further towards the chasing end has the goal of finding specific facts, while reading further towards the searching end has the goal of exposing oneself to general thought processes. Most reading modes will not fall on either end of the extreme but instead will be somewhere in between. You might have more conscious critical thoughts about the subject matter towards the chasing end of the spectrum, but whatever, conscious thinking is overrated anyways.

2) Thought processes matter, because having more ways to approach problems is useful. Exposing yourself to different thought processes can help you find ways to test for causality (e.g., dose-dependence), see different frames for thinking about the world (boosting divergent thinking), and etc. As evidence for this, Dean Simonton notes in Creativity in Science (here) that the best scientists tend to read broadly outside their particular field and maintain active correspondence with people outside their specific pursuit. Sure, this could merely be due to an openness to experience / IQ selection bias, but I'd bet there is at least some treatment effect of searching mode reading on improving cognitive style.

3) As Hanson notes, a reading mode towards the chasing end of the spectrum has advantages, as you are more likely to put down something that is impressive but not actually useful. What he underplays is that reading mode towards the searching end of the spectrum also has advantages, as it can allow you to think critically about general thought processes without getting bogged down in the details. In general, a more chasing reading mode should be more helpful in the short run, while a more searching reading mode should be more helpful in the long run. Another reason to prefer at least some searching is that without it you never would have gotten so far down in this post. And then where would you be? Screwed.

Solemn Higher Ed Vows

Tomorrow I will take my last exam in college. It is on math stats. In case I ever end up teaching in higher ed, I want to make a record of which rules I would need to follow in order to respect myself as a teacher. There are two main points, to which, in publishing this blog post, I hereby solemnly swear to follow:

1) I will grade essays / short answers / papers based on the content of the writing, including the precision of the claims made, the relevance of the evidence cited, and, depending upon the context, its originality. To give the content more weight, I will eschew the importance of spelling, formatting, the proper dictionary use of phrases like "begs the question," the alphabetization of references, and other irrelevant tics that apparently haunt so many professors. Whenever possible I will encourage or at least accept the use of words that I have to look up in urban dictionary, or its equivalent thirty years from now, however horribly newfangled they may appear to raggedy old me.

2) Whenever possible, I will attempt to channel students' competitive urges for the good of the world by having them play non-zero-sum games. That includes collaborating on unsolved problems, writing wikipedia pages (or its equivalent), and articulating hitherto unarticulated or poorly articulated sentiments.

All praise be to the god of integration by parts, in whose holy presence I tremble with mortal fear, amen.

Failure And Creativity In Lost

In his podcast yesterday with Bill Simmons, writer Carlton Cuse explains a bit of the driving philosophy behind the composition of Lost. First, he argues that the premise of a TV show is much less important than its execution. Seinfeld, the "show about nothing," is his example here. I do wonder whether this is will be different for comedies, though. The preponderance of movies based on previously successful books / plays in imdb's top 250 suggests to me that the plot is critical for a drama.

Second, he argues that to be successful creatively in any endeavor, one must risk failure. Apparently, Cuse and Lindelof (the other writer) basically expected to fail. Their goal in writing the first twelve episodes in 2004 was to make them as cool as possible so that if / when the show was canceled, it could become some sort of cult classic that they would be proud of. That would make it like Firefly, which only aired in 2002 for eleven episodes but killed it in DVD sales. The writers being cool with conventional failure for such a plausible and specific reason could explain the show's success and its love of ambiguity.

This post is dedicated to Colin Marshall, who ostensibly hates Lost (see here, here, and here), but who loves the idea that risking and / or embracing failure is essential to success in creativity (see here, here, here, and here).

Where Are Other Cross Subsidies?

In his article on the future of news, James Fallows discusses the model wherein certain high-yield aspects of newspapers subsidize less profitable aspects. The net is cutting this down:

“Newspapers never made money on ‘news,’” Hal Varian said. “Serious reporting, say from Afghanistan, has simply never paid its way. What paid for newspapers were the automotive sections, real-estate, home-and-garden, travel, or technology, where advertisers could target their ads.” The Internet has been one giant system for stripping away such cross-subsidies. 

So, in which other industries do such cross-subsidies exist? The most obvious one is higher ed. Students and / or their parents pay for some variable combo of signaling, learning, and networking. Yet, colleges and (especially) universities use this money on tons of other things that benefit students only indirectly.

One particularly interesting use of student funds is in supporting the research pursuits of faculty members. To the extent that higher ed is mostly about classroom learning, research support is merely a cross-subsidy and should eventually be cut out by the growth of the internet. But, to the extent that paying to participate in higher ed is mostly about associating with high status professors, supporting faculty research is not a cross-subsidy and should remain despite the internet. Time may tell which factor is more important.

Two Ringing Bullshit Meters

1) Vaughan Bell chides certain psychologists for erroneously claiming that cortisol rises during crying can be damaging to the brain:

I have a bullshit switch. It gets triggered when I hear certain phrases. 'Neuroplasticity' is one, 'hemisphere' is another and 'raises dopamine' is a regular button pusher. That's not to say people can't use these phrases while talking perfect sense, but I find it useful that they put me on my guard. Most recently, I've found the phrase 'raises cortisol' to be a useful way of alerting me to the fact that the subsequent words may be a few data points short of a bar graph - potentially some poorly understood drivel....


These claims both reflect one-dimensional thinking about how the brain works. Yes, stress tends to raise cortisol levels and there is good evidence to suggest that chronically high levels of stress and cortisol may be detrimental to brain, but this conclusion is typically drawn from people who have been through some fairly serious shit, wars, deprivation, trauma, or have specific hormone problems. There is remarkably little research on cortisol, everyday stresses in young children and none to suggest normal variation damages the brain in any way.

2) Andrew Sullivan elucidates his qualms about US supreme court nominee Elena Kagan:

Her life, so far as one can tell, is her career, and her career has been built by avoiding any tough or difficult political or moral positions, eschewing any rigorous intellectual debate in which she takes a clear stand one way or the other, pleasing every single authority figure she has encountered, and reveling in the approval of the First Class Car Acela Corridor elite....

Kagan strikes me as the Democratic elite's elitist: free of any conviction that is not caged in a web of Clintonian caution, punctiliously diligent in every aspect of her career, motivated by a desire never to offend those with power, and rewarded in turn by the protection and praise of these elites. Here is Walter Dellinger's almost comically balanced, well-polished, piece of bullshit.... David Brooks calls this generational elite pattern - which is far broader and wider than Kagan's lone example - "disturbing." I find it depressing.

One question is whether these people are deliberately attempting to fool others or are merely deluding themselves. In the case of the cortisol experts, sure they might be telling only their side of the story, but might not they also merely be the victims of confirmation bias? In the case of Kagan, might not she have convinced herself that she merely didn't hold any stances on controversial opinions?

In achieving one's own ends through nefarious means, there is a spectrum between the extremes of total self-delusion and total intentional fooling. Fall closer to the self-delusion side, and you'll seem stupider. Fall closer to the deliberately fooling others side of the spectrum, and you'll seem eviler. Ah, the quandaries and trade-offs of careerism.

Alcohol And Vocab Aptitude

Vocabulary, which seems to be a pretty good proxy for general intelligence, shows a positive and dose-dependent correlation with being an alcohol drinker, among Americans: 

Woah. Razib first found this relationship (here; HT R Wiblin) and we both used UC Berkeley's awesome SDA to do the crunching. The error bars are 95% confidence intervals, and their non-overlap between groups means most of these differences are very unlikely (< 5%) to be due to random chance. The general trend holds for all kinds of different age groups (18-30, 30-50, 50-60, 60-70, 70-100, etc.).

Allow me a couple stabs in the dark as to the relationship here:

1) Alcohol reduces anxiety (see here) and, as Steven Pinker has speculated, "people with higher intelligence are better at overcoming their anxious temperament and more likely to see their own worry list as a problem to be solved, minimizing unnecessary anxiety while still being anxious enough to get things done.” So, people with higher intelligence may be more likely to be self-medicating their anxiety with alcohol.

2) People with higher vocabs are more likely to be more highly educated, and thus been introduced to the drinking culture that is commonplace in institutes of higher ed. It may be a part of the culture there because it is more impressive (see bottom here) to be able to succeed in school and party on the weekends. And most people seek to maximize their relative impressiveness.

Two Irony Models

Dan Owen critiques Colin Marshall thusly:

So much of the content of your work -- blog mostly, but even the radio show and now this video -- is ironic, self-deprecating, light and somewhat dismissive in tone. I like that, and I assume that's your natural voice, but if I were working with you I'd want to know whether you're hiding from failure by adopting a posture of dismissive, self-deprecating humor.

The problem with irony as an artistic strategy -- like sarcasm as a conversational strategy -- is that it's a one-trick pony. It can make one point well, but it can't go deeper, it can't explore an idea metastatic-ally, if you will (I just made that up, but you get the idea I hope).

Ben Casnocha, who has mined a veritable cottage industry out of irony, thinks that: 1) irony can be "an act of self-protection, and can be a sign of insecurity" (here), 2) the "shift to earnestness [over irony] in communications represents a milestone in a romantic relationship" (here), and 3) " too much non-seriousness [read: irony] is hard to take. But in small doses, I find it endearing and funny. It's a balance." (here). Elsewhere, Kelly Stout argues that "sincerity is the post-irony irony" in her lucid exposition of what I have previously called the conformity theory.

Here are my two models for thinking about irony:

1) Irony and earnestness are the two opposite ends of a spectrum that exists in an equilibrium. If you imagine a world full of irony, earnestness would be highly valued (when recognized). In a world full of earnestness, irony would be especially funny and also highly valued. These two factors inevitably push the world to someplace in the middle of the extremes, with people employing various personality styles at various times. Sometimes people get locked into a specific personality type due to inertia, so personalities types can exist even when they are no longer optimal. The more lock-in, the more people will expect to be able to take advantage of the "market failure."

2) Individuals use irony when they are afraid of the prospects of being earnest. This decision could be more rational, such as wanting to defend one's ego by not publicly pre-committing to a goal which might fail. Or the choice of irony over earnestness could be less rational, such as unconsciously avoiding a anxiety-conditioned fear. Either way, the use of irony by any one person will be less tied to other people's use of irony and more based on their own personality and mixture of conscientiousness / neuroticism.

Any Rating System Trumps None

"Nihilists! Fuck me. I mean, say what you like about the tenets of National Socialism, Dude, at least it's an ethos." - Walter, The Big Lebowski

There are lots of lists that attempt to describe the top movies, with various methodologies. Here are three of the major ones:

1) The American Film Institute determined its top 100 list (here) by having film "experts" create their own top 100 list from 400 nominated movies. Movies were ostensibly judged based on winning awards (read: the Oscars), popularity (box office, syndication, home video sales), historical significance, and cultural impact. It's by far the #1 most cited list that people mention when I bring up top movie lists, probably because AFI's lists have been on TV a decent amount and when it comes down to it Americans watch a shocking amount of TV.

2) Metacritic compiles its top 200 list (here) by averaging the subjective ratings of various movie critics. They ask for user votes but don't actually count them towards the top 200. The big supposed upside of their list is that the ratings should be higher quality because they are based on published reviews. The main disadvantage of their list is the low sample size, which leads to more random noise. For example, Superman II is #2 on their all time list on the basis of a whopping 7 critic votes, while it has the class average of a 6.7 on imdb based on 20,000+ votes. So, Metacritic needs to convert to a Bayesian system that punishes low sample sizes in some way. Unfortunately, they also don't include many older movies, as most of their reviews are from the past 10 years.

3) The internet movie database determines its top 250 (here) via user ratings. Anyone with a valid email address who can pass a CAPTCHA test can rate an individual movie, but you have to have a certain amount of votes and various other qualities for your vote to count towards the top 250. Qualities which imdb doesn't disclose. They use a system that punishes low sample sizes, avoiding the Superman II problem. Compared to the other two lists, theirs is more diverse, either via old movies (as compared to metacritic) or foreign movies (as compared to AFI). Their big problem is recent movies, which start off much higher than they end up as (see: the Dark Knight), but are not punished as such. Admittedly, the fact that many others are against imdb's ratings probably makes me like it more, and I am also biased because I'm currently watching the top 250 have invested a lot of time into it. But I definitely do think that it's the best.

What are some metrics by which we can compare these systems? One way would be to look at other systems that use fan votes as compared to expert votes. For example, the NBA All-Star game relies on fan votes to determine its starters, while it relies on journalists to vote on the MVP. The fan votes tend to be not highly correlated to the quality of the player that year. Allen Iverson has been voted a starter each of the last three years even though his stats have been awful. MVP votes are probably more correlated to player's statistical success, although experts aren't perfect either: Steve Nash probably shouldn't have won it twice.

So, NBA All Star votes might count as evidence against imdb. And perhaps that kind of example is why people don't trust imdb? I would argue that sports are qualitatively different because most people don't actually watch all of the games, whereas most everyone who votes on imdb has actually watched the movies.

Regardless, I think sober, intelligent minds can disagree about the relative merits of each of these systems. Your personal preference will probably depend based on to what extent you believe quality is universal, and how much you trust the opinion of insider elites as compared to normal folks.

Much more troubling is the lack of any system at all, of just wandering through the world like a little boy, lost, looking for his mommy. For example, critic Johnathon Rosenbaum thinks that presenting AFI's list of movies in order is "tantamount to ranking oranges over apples or declaring cherries superior to grapes." His attitude is just pure nihilism, through and through.

The Happiness Vs Boredom / Depression Vs Mania Continua

Sandy Gautam argues persuasively that happiness and sadness are not on the same dimension:

[D]epression is characterized as a reaction to losses / continuous exposure to stresses that makes goals out of reach / unachievable... One becomes withdrawn from the situation and does not fight the stress, but flights from the stress by withdrawing in a cocoon. The loss of appetite and more sleep can be seen as behavioral counterparts of withdrawing or exhibiting a flight response to stress.

[M]ania is a reaction to a situation similar to depression -- when something is lost / is under threat of losing -- but this time, under stress, one fights and not flights -- thus one becomes energized to right the wrong and may become angry / irritable if the efforts to retain goals / valued entities are frustrated by [the] external world... the focus is preventive and the state is of scarcity.

Contrast this to a state of abundance when ones (life) goals have been met / are within reach. This apparent positive state of affairs may again give rise to different emotions / behavioral manifestations depending on whether one has approach or avoidance dominant reaction. If one approaches the more free time available after goal accomplishment as a boon that can be used to hone ones hobbies / find other meaning in life / build relationships etc and not as a threat (free time can be a threat) then one experiences positive emotion of happiness and behaviorally flourishes.

In contrast consider a similar person who has achieved everything in life... but given the fact that one is living in abundance is frightened or flights from the free time that has been made available. [T]hat person will be listless, will exhibit ennui or boredom and may even exhibit despair as he finds life meaningless. Thus behaviorally he would languish.

The key distinction between these two continua is that they are pre- and post- goal emotions. While working towards a given goal the extremes of the emotional spectrum will be depression and mania. After the goal is met, the extremes of the emotional spectrum will shift towards boredom and happiness.

One of the most well documented biases is that we tend to assume that our post-goal emotions will be further towards the happiness side of the spectrum than they really are (i.e., see here). Perhaps if people had more precise understandings of the distinction between pre and post goal emotions, this bias wouldn't be so common.

Guatam's model also makes clear predictions about the two different types of depression that so troubled Jonah Lehrer's article about the adaptiveness of depression. I like his model because I think that in the real world there are very few thresholds and everything is on a spectrum of some sort.

You Tube Rating Gets Even Worse

In Nov '08 I called You Tube's rating system "a disaster," and in Feb '09 I explained that they don't care. Instead of improving, it has since gotten worse, or, depending on your frame, is no longer really a legitimate rating system at all. From their shoddy explanation:

Ratings have changed from the Star system to a binary "Thumbs-Up ‘Like’" / "Thumbs-Down" system. Anything other than a 1- or 5-star rating is rarely used on YouTube, and so we moved towards a simpler "Like / Don't Like" model.

What's sad is that there is so much potential at You Tube. So many viewers and your typical proportion of willing raters means they could really impact the world. How cool would it be if there were a top 250 for videos, categorized into music videos, activism, stand up comedy, etc?

Sure there might be more 1 / 5 star ratings than you'd like. So why not incentive 2-4 star ratings by weighing them more, throw out some of the extreme ratings like imdb probably does, or better yet, count the rater's deviation from his own average rating? Switching to a 10 star system couldn't hurt.

Instead of a solid rating system, we must rely on recommendations (with small, insular sample sizes) and feedback-propagating lists of "most viewed" videos. With Google's decision, the internet became a little bit less self-aware. I doubt anyone shed a tear. But maybe we should have.

Short / Long Term Effects Of Better Competition

Jonah Lehrer's article in the WSJ presents some counter-intuitive data: golfers tend to do worse when they go up against a great player like Tiger. Apparently, "whenever Mr. Woods entered a tournament, every other golfer took, on average, 0.8 more strokes." This and other one-shot experimental data goes against the common observation that competition increases performance. So how do we resolve the paradox?

Here's one explanation. In the short run (like one golf tournament), better competition makes us overly anxious, sometimes leading to paralysis by analysis. But in the long run this extra focus might make us better, because it forces us to take our games to the next level, assuming that we make it over the "dip." We come to expect the anxiousness in any given tournament and practice harder to overcome it. Any thoughts?

Reward And Interest

Justin Wehr has a cool post summarizing an article by Paul Silvia about what makes certain things interesting.  Silvia argues that interesting things are new / complex / unexpected and, crucially, comprehensible. About textbooks, he writes that:

[T]he largest predictors of a text's interestingness are (a) a cluster of novelty–complexity variables (the material's novelty, vividness, complexity, and surprisingness) and (b) a cluster of comprehension variables (coherence, concreteness, and ease of processing). Intuition tells us that we can make writing interesting by "spicing it up"; research reminds us that clarity, structure, and coherence enhance a reader's interest, too.

On the personality traits that predict high amounts of interest as opposed to happiness:

Interest connects to openness to experience, a broad trait associated with curiosity, unconventionality, and creativity. Happiness, in contrast, connects to extraversion, a broad trait associated with positive emotions and gregariousness.

Surprisingly, neither Silvia's article nor Wehr's summary mentions how the expected reward of an activity correlates with its interestingness. Surely we become more interested in things if we expect that our interest could be paid off in a big paycheck or the opportunity to publish a paper in a prestigious journal. With respect to the latter, we often immediately become less interested in something once we learn that it has already been published. This phenomenon that cannot be explained without resorting to expected reward!

Other papers have not overlooked this component of interest. For example, in Jianzhong Xu's article on predicting student's interest in homework, he notes that,

[Some] theorists argue that significant others (e.g., parents and teachers) may play an important role in enhancing interest, through external support and continuous feedback... [T]he variation in homework interest was positively associated with affective attitude toward homework, motivational orientation toward homework, student initiative in monitoring homework motivation, teacher feedback, and self-reported grade.

The Pearson correlation he found between interest and self-reported grades was not huge (r = 0.12), but still highly significant at p < 0.01. Moreover, the test just asked for overall school grades (either mostly A's, B's, or C's...) and overall interest. I'd expect that if you looked at interest on an individual assignment, there would be a much higher correlation with the grade on that particular assignment.

At one point Wehr quotes Lennart Sjöberg, who is confused by the inability of interest theory to explain his daughter's hobbies:

My 10-year-old granddaughter is extremely interested in horses and riding, like so many girls of her age. Some of that interest possibly can be explained by collative variables and the activity of riding a horse, taking care of it, and so on, but there also seems to be a question of sheer fascination with horses per se, quite regardless of any activity having to do with horses. We may be hardwired to develop a lust for certain types of objects and activities. Genetic determination of part of the interest variance is a very real possibility.

I highly doubt that there is a polymorphism that predicts liking of horses by young girls! Instead, I'd bet that his daughter gets reinforcement from liking something that her friends and cultural role models like, which makes her much more interested in horses qua horses.

However, if Sjöberg is referring to the genetic components of being interested in broad topics in general, he's surely right. Openness to experience is highly heritable, with estimates of the percentage variance explained by genetics ranging from 45 - 61% (here via here), 57% (here), to ~ 40 - 50% (here).

One specific polymorphism that is associated with openness to experience is the number of repeat alleles in the gene coding for the dopamine receptor D4. Comings et al's chart shows the U-shaped dependence of openness to experience on this one polymorphism, with higher repeat allele numbers to the right, and higher scores for openness to experience up:

Comings et al, 1999, PubMed ID 10402503

One study describes "a possible molecular link between [the] dopamine DRD4 receptor, music and autism, possibly via mechanisms involving the reward system and the appraisal of emotions."

Once I started blogging, I started becoming interested in lots more topics. This can be easily explained by the link between reward and interest. Once you are a blogger, everything you read has the potential to be blog-worthy. You are rewarded for finding good stuff that your readers enjoy by e-mails, comments, and silent, knowing nods of approval. This drives you to be interested more by the things that you read. There's nothing particularly magical about it.

It's no wonder AI researchers like Shane Legg study the neuroscience of the reward system. It's probably our most powerful set of circuits. And when explaining things like interestingness, it's hard to ignore.

Why Science Writing Is Painful

Natalie Angier channels Susan Hockfield (HT: Anthony Sebastian here)

..the time had arrived for writing, the painful process, as the neuroscientist Susan Hockfield so pointedly put it, of transforming three-dimensional, parallel-processed experience into two-dimensional, linear narrative. "It's worse than squaring a circle," she said. "It's squaring a sphere."

####

One of my plans for the next few months is to take the lessons from Eric Barker's awesome blog Barking Up The Wrong Tree and apply them to my neuro blog here (subscribe!). The lessons from his success seem to be:

1) Let the original authors of studies do more of the work via quotes.

2) Hitting a home run on any given post might actually be bad in that it increases the self-applied pressure for the next one.

3) There is value in aggregating lots of small ideas into one location.