Popularity is Overrated

My love of The Flamingo Kid may seem misplaced in view of its meager 6.2 IMDB score—with a mere 3900 ratings—its lack of a Blu-ray edition, and unavailability on Netflix or Amazon Instant Videos. But this is not an unfamiliar thing for me, and probably for a lot of people. Actually, TFK is just one of several movies I love and have watched more times than it would be seemly to admit—not art movies (though there are those too), but Hollywood features with A-list stars—but which languish in obscurity and neglect. Why is this? Is my taste so idiosyncratic? Or bad? I want to suggest that the most popular movies—or songs, or novels, or whatever—are not as much better than others as their popularity would seem to imply. Sometimes they may be no better. By the same token, relatively neglected works are not so inferior and may be not at all inferior.

If popularity is determined by recognizable quality, why can’t hits be predicted in advance? You might think they could be, since once a hit is established, its quality seems evident, and people always say the reason they love a hit song or whatever is because it’s “so awesome.” But evidently not. Studies like this one of the book, music, and movie business and this one of the television business note the inability of industry professionals to predict which products will be hits—despite strong financial incentives to do so—and examine their strategies for coping with the problem. (The second article quotes the president of CBS to the effect that, “All hits are flukes.”) And it seems obvious, really, that these businesses would be managed quite differently from the way they are if the sales differentials could be predicted.

Yet hits generate massive sales, vastly outselling the songs/movies/books/etc. that they couldn’t be distinguished from ahead of time. Sting makes $2000 every day from “Every Breath You Take.” That one song generates over a quarter of all Sting’s music publishing income (see here for details). J. K. Rowling is another obvious example. The best selling book series in history has the usual story of early rejection. Twelve publishers rejected the manuscript of the first Harry Potter book before Bloomsbury finally picked it up. The editor who bought it advised Rowling to get a day job, since there was little chance the book would sell. Rowling would go on to become the first person ever to make a billion dollars by writing books (though not a billionaire, apparently, thanks to Britain’s highly progressive tax code). Examples like this couldn’t happen (or would be exceedingly rare) if stellar sales implied stellar quality.

Here is an interesting study of the phenomenon, focusing on the music business. The authors created their own music download website populated with indie songs from obscure indie bands unknown to the study participants. It was a large study, with over 14,000 participants. Participants were invited to listen to songs, rate them, and download them if they wished. There were two basic conditions: one in which the participants had knowledge of how many times a song had been downloaded, and one in which they didn’t. In the knowledge condition, outcomes were more extreme (i.e., there were higher peaks and lower valleys in numbers of downloads between songs) and less predictable from independent ratings of song quality. In the knowledge condition, songs that were independently rated highly almost never did terrible and songs that were rated terrible almost never made it to the top, but for all other songs, the popularity outcome was completely unpredictable: they could wind up at the top, the bottom, or somewhere in the middle. This was not true of the no-knowledge condition.

Thus, mere widespread knowledge of the preferences of others is enough to generate the phenomenon of the “awesomeness” of cultural products that couldn’t be differentiated from their competitors ahead of time even by seasoned professionals. Apparently a “contagion effect” causes chance initial success to snowball and catapult a work or artist to the stratosphere. The works thus left behind, therefore, are not necessarily inferior—or much inferior—to those lucky few that become “awesome,” any more than those few are necessarily superior—or much superior. So if there’s a movie or other work you love that doesn’t seem to have the popularity it deserves, don’t feel bad. Popularity is overrated.

4 thoughts on “Popularity is Overrated

  1. There might be two questions here: 1. why don’t so many good works intended for a popular audience become popular? 2. why do so many mediocre and even bad works intended for a popular audience become popular? If most popular songs and movies were really good, such that 2 weren’t really a question, the answers to 1 might look different, or at least the factors involved might take on different weights. If most hits were really good, then the really good things that don’t become hits might just owe their obscurity to bad marketing or an overabundance of similar hits at the time or other sorts of low visibility. But in fact there seems to be only a very loose relationship between quality and success; lots of mediocre stuff succeeds even as really good stuff passes unnoticed. That’s what makes the contagion effect idea seem so plausible.

    But I’m not quite willing to give up the idea that quality plays an important explanatory role in popularity. But by ‘quality’ I don’t mean how good something is, but what sorts of qualities a movie or song offers to people. In most cases, it seems pretty easy to pick out what it is about successful movies that people were responding to, even when the movies aren’t very good. To take an example, I think the fourth installment of the Die Hard series, Live Free or Die Hard, is a pretty awful movie. But it has some pretty impressive stunts and effects, it revolves around a cyber terrorism scenario that was fairly novel and plays on people’s anxieties about our reliance on technology, and Bruce Willis is just so cool. So it made $385 million. Granted, none of that is sufficient to explain why that movie did so well while others offering broadly similar forms of entertainment didn’t. But it’s not a surprise that it did so much better than a far superior film that came out the same year, 28 Weeks Later, and it seems to me that a good part of the reason for that has to be that far fewer people enjoy post-apocalyptic zombie films (or, well, far fewer enjoyed them before the success of The Walking Dead, anyway; but I’d argue that the success of TWD has a great deal to do with its ability to appeal to a tremendous range of different viewers).

    So my intuition (that’s right) tells me that the qualities that movies offer have a non-trivial role to play in making them popular, even though having certain qualities that people will enjoy is far from sufficient for success. Music might be a different story. But quality in the more normative sense — good quality — seems to me to have very little to do with it. So here’s a question about The Flamingo Kid: do you think it’s the kind of movie that is only accidentally unpopular, or is it more the kind of thing that just won’t appeal to a huge audience?


    • You’re right, and it’s well-observed, that it isn’t just that the success of the successful pop culture works is telescoped, so that a pack of works that do not vary all that much in quality gets artificially spread apart by the popularity process. That is one factor, but a second and more important factor is the randomness of the success. It’s practically a lottery. And this implies not just that good quality works get left behind but that middling works get promoted. The figures in the Salganik et al. study make this crystal clear.

      No doubt also there are many factors at work in determining popularity, not just the contagion effect. One thing I think must have worked against The Flamingo Kid, for instance, is lack of star power. None of the cast was an A-lister when that movie was made. Of the two who are best-known today, Matt Dillon was just starting out and Hector Elizondo was a TV actor who’d never had a hit series. We wasn’t well-known. Richard Crenna and Jessica Walter were great actors who’d been around a long time, but nearly always in supporting roles. Nobody went to see “the new Richard Crenna movie.” Today I would imagine most people have never heard of them. Janet Jones had a very short career (because she had to go off and marry Wayne Gretzky), and I think TFK was practically her first role. So nobody had heard of her when the movie was released, and almost nobody knows who she is today. Star power is the sort of “quality” you’re talking about. It does bring people into the theater. That’s why there is a star system.

      There could be various other “quality” factors working against TFK, I suppose. Perhaps people think it’s weird to make such a big deal out of gin rummy? But I look at the Garry Marshall movies what were high profile, culturally significant hits, in particular Pretty Woman and The Princess Diaries, and I don’t see any difference in “good quality” that explains the difference in popularity outcomes. Pretty Woman in particular I remember seeing in the theater when it was new. I thought it was mediocre and couldn’t understand its huge success. So my inclination is to think that TFK is mainly a victim of losing the contagion effect lottery.


      • Ok, I watched it. It bears out your recommendation. The story is good and Dillon, Crenna, and Elizondo put in good performances. The conflict it sets up is real, and its resolution nicely revolves not just around Jeffrey’s disappointment with what he thought was a fast track to wealth and glamor, but around his coming to see Brody as superficial and dishonest with a life that is ultimately pretty empty and fake. Still, I’m not surprised that it wasn’t a hit and has only a moderately high reputation. Its drama is somehow just not quite gripping, and too broken up by incidental scenes, while its comedy frequently falls a bit flat. Some of the dialogue is well written, but much of it, especially outside the important dramatic scenes, is pedestrian. The romantic subplot feels at times as though it’s just there to take up space, and there’s no real chemistry between Dillon and Jones. Some of the other actors leave a lot to be desired, and while the movie leans a lot on its 1960’s ambiance, it didn’t really do it for me (compare, say, Goodfellas). The lack of star power must have limited its appeal, but I can see why masses of people didn’t and don’t love it. The main story sets up a problem that most people can sympathize with, the plot revolves around the difficult choices that Jeffrey has to make, and they represent the attractions and costs of his options in a realistic, if not extraordinarily complex, way; as with Pretty Woman, the vapidity and absurdity of luxury are impossible to miss, but so are the struggles and limitations of working class life. To my mind one of the best bits of dialogue is the massage scene conversation when Brody talks about his father (and the silliness that ensues when Jeffrey tries to express the same thoughts to his father). That really captures something about the lives of many people and the sort of experience that can turn someone into a Brody. That core of the story is just really good. But I can’t help feeling that the movie as a whole just doesn’t quite do justice to the core, and that ultimately it’s not a great movie.

        It is, however, much better than Pretty Woman, I agree. The two movies seem to have opposite problems: where The Flamingo Kid has a solid core with so-so dressing, Pretty Woman is just all dressing. Neither of them are anything like unpredictable, but the central plot of Pretty Woman just doesn’t have much substance. No doubt Richard Gere and Julia Roberts helped make that movie a success, but what I remember people enjoying most about it was the comedy, which is all pretty much a riff on working-class-girl-enters-into-rich-people’s-world-and-doesn’t-know-how-to-behave. I guess it’s funnier than the parallel stuff in The Flamingo Kid (though I haven’t seen Pretty Woman since I was something like twelve years old, I think). But that’s really all it’s got going for it, and I don’t consider it a good movie. TFK is a good movie. It just doesn’t strike me as a great one.

        I hope that’s not too unappreciative. I’m glad to have watched it, which is, alas, not a given.


  2. I wonder how much the contagion and randomization effects bias product reviews (when everything goes well enough – i.e., there are not paid good reviewers or too many ridiculous one-star sour-pusses)? I would guess quite a bit less. And I would guess the reason is that the purpose of the rating, in that case, is precisely to assess quality (and quality along pretty obvious parameters). There is less an element of how subjectively awesome the product struck you as. If this is right, we might expect some in-between cases. Maybe reviews of iPhones or Gucci bags are more like the popularity of movies.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s