How Bar Bar Bar became a hit
Sep. 21st, 2013 02:21 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
koganbotThe title of this post is a bait-and-switch, actually, since my fundamental motive is to get you to read my old Las Vegas Weekly piece on "cumulative advantage," which is a concept from economics and sociology that I think everyone needs to know.
To summarize in one sentence that combines two ideas: small advantages in fame, wealth, power, etc. can grow into large advantages, large advantages can grow into larger still, and there's always, ineradicably, an element of randomness, of luck, as to what gets the small advantage in the first place, and what gets to leap from one level to the next. So, all fame is viral, and we can never be certain in advance as to which viruses will catch hold and which won't.* But once something is famous, the fame is very hard to dislodge.** So my first point of the day:
(1) There's always an element of luck when anyone or anything becomes well-known. Always. This includes Darwin as well as Rihanna. From the piece:
(2) My second point is about explanations, not about cumulative advantage. Even disregarding luck, people who claim to know the (other) reasons "Bar Bar Bar" became a hit have at least some shit in their thinking: no explanation of such a single social event is testable.**** For instance, take the idea that "Bar Bar Bar" became a hit because Crayon Pop are different from everyone else in K-pop — they have a different look, and "Bar Bar Bar" has a different sound — which seems like a good explanation to me, and is one I myself would give, despite all these caveats. The trouble is that if "Bar Bar Bar" had not hit, I could have used that very same reason to explain Crayon Pop's failure. And EXO, whose music is at least as novel as Crayon Pop's, didn't go top 10 in Korea until they hit with the relatively conventional "Growl."
(3) Putting 1 and 2 together: there's an element of luck as to which explanations themselves become common and accepted. Cumulative advantage applies to ideas as well as to people and songs.
The problem with the phrase "element of luck" is that it doesn't tell us how large the element is. In my limited reading, I've seen no way of specifying the percentage luck plays in cumulative advantage. Is luck 10 percent of the reason one song became a hit, and 80 percent of the reason another one did? How do we quantify this sort of luck? It's not like a coin flip, which we know is 50 percent, or the roll of a die, which is 16⅔ percent. [EDIT: Well, assuming the coin is balanced and the die isn't rigged, the chance of a particular result of a coin flip is 50 percent and of the roll of a die is 16⅔ percent, but the percentage attributable to luck is 100 percent for each: it's entirely chance.]
Certainly "Bar Bar Bar" couldn't have hit if it had no appeal and no promotion and if no one liked the video or the dance. My guess is that statistics and sociology and social psychology are centuries away from giving us much of an idea of the relative importance of any of those factors; and even then, what they'd tell us would be something like "companies that invest in songwriting get a bigger bang overall for their buck than do companies that invest in market research," or vice versa; but these wouldn’t explain a particular hit, since promotion might be more important in one case, appeal more important in another, and so forth. But when it comes to luck, I wonder if specifying the amount is even doable. Statistics has the concept "margin of error," but that doesn't seem to be related to the sort of unpredictability we get in cumulative advantage.
So, while luck is good for making us question other explanations, and we know it's there, it doesn't work as an explanation itself. Maybe the difference between "Bar Bar Bar" hitting and Evol's "Get Up" not hitting is entirely luck (Evol are almost as unknown as Crayon Pop were, and "Get Up" is almost as good as "Bar Bar Bar" is), or maybe Evol need more than good music and a strong performance to even have a chance, needed the sort of "more" that Crayon Pop had. I don't know and neither do you.
This doesn't mean that there's no way whatsoever to test our explanations, or that we're not obligated to try. First off, we can challenge the explanation itself: Among other things, we can ask, is the explanation implausible? Is it incoherent? Does it run counter to well-founded beliefs? Does it fail to eliminate counter-explanations? Would the explanation seem to apply just as well to an opposite event (e.g., what I wrote above about "Because they did something different" being as applicable to one group's failure as to another group's success)? Would the explanation apply to too many outcomes (which as I said is a problem we have with using luck itself as an explanation: if everything that becomes famous is lucky to do so, then "good luck" is not an interesting explanation for why a particular song becomes famous)? Does it fail to apply to something similar? Does it leave a lot of unanswered questions? Is it based on rumor or dogma rather than evidence? Is there lack of general agreement as to what the words of the explanation mean? Is it an explanation at all?
And then of course we can do some research. For instance, instead of guessing, we can ask people why they like something or why they do something, and how they heard of something. Not that this is always easy, and not that it doesn't run into problems. We may not have access to the people we want to ask, we may not know whether we're getting a representative sample, most people (including me) have trouble knowing why they like a song (e.g., I know I immediately found the tune catchy and the dance funny and charming, not just the five-cylinder jump but Crayon Pop prancing in unison and their imitative gestures and their speed racer helmets; but I still have trouble telling you why I like those things, much less figuring out how much of my liking was due to being primed by David Frazer posting in my comments about Crayon Pop whenever he had an excuse, and therefore getting me to relisten to their earlier tracks and thus giving myself a chance to increase how much I liked them).
Furthermore, cumulative advantage will also apply to our reasons. That is, if someone being surveyed hears a lot of other people saying "Crayon Pop hit because they're different," that person may well say "I like Crayon Pop because they're different." This isn't because people are conformist or dishonest, but because as we plumb our own reasons, the ones that take shape will be the ones we have words for, which are likely to be the ones we're hearing. And then we'll genuinely come to appreciate the difference, even if (as for me), it was the relatively familiar catchiness of the tune that grabbed me most at the start; but the catchiness, and the power pop rock of the arrangement, contributed to my sense of the song's and the group's freshness, once people started making a big deal of the freshness.
The obvious thing to do now is to give an example or two of each of the various species of faulty explanation I outlined above. But actually I'd just rather get this posted, and if we want we can talk more in the comments; or I can do another post someday. I've talked negatively here (about faulty explanations) but I think that some of my explanations about [redacted] are good not necessarily because they avoid all flaws but because they explain a lot that no one else has explained.
But back to Crayon Pop: two fairly lousy explanations jump to mind. One is the idea that the Ilbe controversy helped bring attention to Crayon Pop. I believe that the proper response to this is "What controversy?" — i.e., it didn't bring attention to Crayon Pop since almost no one who didn't already know about Crayon Pop ever knew there was a controversy. But actually I don't know this, and this subject needs a modicum of research. If you conducted a survey, would you find a significant number of people saying they first heard of Crayon Pop owing to the Ilbe supposed flap? But face it, no one's going to conduct a survey, so what we (i.e., someone who is definitely not me) might want to do is to go online and see if there are a lot of Crayon Pop fans posting that they first heard of the group because of some Ilbe thing.
Second is the contention among Netizens and trolls that, since we know that for a not prohibitively large amount of money an agency can game the system so as to place a song high on the streaming charts even when a video is not genuinely loved and streamed, then, for all we know, Crayon Pop's agency did this very thing. Now, putting aside the fact that (1) maybe we don't know that streaming charts can be gamed (I certainly don't know one way or another, though I wouldn't be totally shocked to learn they could be), (2) last I looked no one who claimed to know that such gaming has occurred has actually named names or offered evidence (but I haven't looked much), and (3) there's no evidence that Crayon Pop's agency did this, there nonetheless are also other reasons why "they gamed the streaming" is not a good explanation (even if Crayon Pop's agency had tried to game the streaming charts). For instance... but no, if you've been paying attention while reading this post, you should be able to work out for yourself why such an explanation isn't likely to be good.
In any event, I think "Bar Bar Bar" sounds catchy, and I assume some other people do (but do they all find the same parts of it catchy that I do?). I think the dance funny and charming, and I assume other people do. The five-cylinder-jump is fun to imitate. That Crayon Pop are "different" is no big deal to me. (I've heard plenty of power pop in my life; that Crayon Pop all dress the same when doing the song is kind of funny, that the dress doesn't symbolize "sexiness" isn't "refreshing" to me but may be to some people and anyway I don't know that people actually consider Crayon Pop not sexy, which would be pretty stupid). The song, btw, started slow and was falling down the chart in its third week, with Crayon Pop no longer on the big TV performance shows, when the track got its second wind. That wind was already well in effect when they got a break and appeared on popular TV show Sketchbook; that performance probably helped but the track was climbing anyway, presumably from word of mouth about the videos, and people doing dance covers, but I don't know the reasons for sure. And at least some of all this, as well as Crayon Pop's already having a devoted (though small) fanbase, made it possible for the song to hit (as opposed to impossible), but none was a guarantee.
Also, in its little way, the song kicks butt. It rocks!
Footnotes in the comments.
To summarize in one sentence that combines two ideas: small advantages in fame, wealth, power, etc. can grow into large advantages, large advantages can grow into larger still, and there's always, ineradicably, an element of randomness, of luck, as to what gets the small advantage in the first place, and what gets to leap from one level to the next. So, all fame is viral, and we can never be certain in advance as to which viruses will catch hold and which won't.* But once something is famous, the fame is very hard to dislodge.** So my first point of the day:
(1) There's always an element of luck when anyone or anything becomes well-known. Always. This includes Darwin as well as Rihanna. From the piece:
The more people know about each other’s choices, the more likely they are to come to agreement. In retrospect, this strong agreement can make an outcome seem as if it had been inevitable: Look, all these people agree, so this must reflect the taste of the public, or the quality of the merchandise. But in fact the experiment [by Duncan Watts and crew] shows the exact opposite. The more people know about each other’s choices, the less predictable the outcome.***(By the way, the passage would have been just as correct if I'd added "size of the promotional budget" to the phrases "taste of the public" or "quality of the merchandise.")
(2) My second point is about explanations, not about cumulative advantage. Even disregarding luck, people who claim to know the (other) reasons "Bar Bar Bar" became a hit have at least some shit in their thinking: no explanation of such a single social event is testable.**** For instance, take the idea that "Bar Bar Bar" became a hit because Crayon Pop are different from everyone else in K-pop — they have a different look, and "Bar Bar Bar" has a different sound — which seems like a good explanation to me, and is one I myself would give, despite all these caveats. The trouble is that if "Bar Bar Bar" had not hit, I could have used that very same reason to explain Crayon Pop's failure. And EXO, whose music is at least as novel as Crayon Pop's, didn't go top 10 in Korea until they hit with the relatively conventional "Growl."
(3) Putting 1 and 2 together: there's an element of luck as to which explanations themselves become common and accepted. Cumulative advantage applies to ideas as well as to people and songs.
The problem with the phrase "element of luck" is that it doesn't tell us how large the element is. In my limited reading, I've seen no way of specifying the percentage luck plays in cumulative advantage. Is luck 10 percent of the reason one song became a hit, and 80 percent of the reason another one did? How do we quantify this sort of luck? It's not like a coin flip, which we know is 50 percent, or the roll of a die, which is 16⅔ percent. [EDIT: Well, assuming the coin is balanced and the die isn't rigged, the chance of a particular result of a coin flip is 50 percent and of the roll of a die is 16⅔ percent, but the percentage attributable to luck is 100 percent for each: it's entirely chance.]
Certainly "Bar Bar Bar" couldn't have hit if it had no appeal and no promotion and if no one liked the video or the dance. My guess is that statistics and sociology and social psychology are centuries away from giving us much of an idea of the relative importance of any of those factors; and even then, what they'd tell us would be something like "companies that invest in songwriting get a bigger bang overall for their buck than do companies that invest in market research," or vice versa; but these wouldn’t explain a particular hit, since promotion might be more important in one case, appeal more important in another, and so forth. But when it comes to luck, I wonder if specifying the amount is even doable. Statistics has the concept "margin of error," but that doesn't seem to be related to the sort of unpredictability we get in cumulative advantage.
So, while luck is good for making us question other explanations, and we know it's there, it doesn't work as an explanation itself. Maybe the difference between "Bar Bar Bar" hitting and Evol's "Get Up" not hitting is entirely luck (Evol are almost as unknown as Crayon Pop were, and "Get Up" is almost as good as "Bar Bar Bar" is), or maybe Evol need more than good music and a strong performance to even have a chance, needed the sort of "more" that Crayon Pop had. I don't know and neither do you.
This doesn't mean that there's no way whatsoever to test our explanations, or that we're not obligated to try. First off, we can challenge the explanation itself: Among other things, we can ask, is the explanation implausible? Is it incoherent? Does it run counter to well-founded beliefs? Does it fail to eliminate counter-explanations? Would the explanation seem to apply just as well to an opposite event (e.g., what I wrote above about "Because they did something different" being as applicable to one group's failure as to another group's success)? Would the explanation apply to too many outcomes (which as I said is a problem we have with using luck itself as an explanation: if everything that becomes famous is lucky to do so, then "good luck" is not an interesting explanation for why a particular song becomes famous)? Does it fail to apply to something similar? Does it leave a lot of unanswered questions? Is it based on rumor or dogma rather than evidence? Is there lack of general agreement as to what the words of the explanation mean? Is it an explanation at all?
And then of course we can do some research. For instance, instead of guessing, we can ask people why they like something or why they do something, and how they heard of something. Not that this is always easy, and not that it doesn't run into problems. We may not have access to the people we want to ask, we may not know whether we're getting a representative sample, most people (including me) have trouble knowing why they like a song (e.g., I know I immediately found the tune catchy and the dance funny and charming, not just the five-cylinder jump but Crayon Pop prancing in unison and their imitative gestures and their speed racer helmets; but I still have trouble telling you why I like those things, much less figuring out how much of my liking was due to being primed by David Frazer posting in my comments about Crayon Pop whenever he had an excuse, and therefore getting me to relisten to their earlier tracks and thus giving myself a chance to increase how much I liked them).
Furthermore, cumulative advantage will also apply to our reasons. That is, if someone being surveyed hears a lot of other people saying "Crayon Pop hit because they're different," that person may well say "I like Crayon Pop because they're different." This isn't because people are conformist or dishonest, but because as we plumb our own reasons, the ones that take shape will be the ones we have words for, which are likely to be the ones we're hearing. And then we'll genuinely come to appreciate the difference, even if (as for me), it was the relatively familiar catchiness of the tune that grabbed me most at the start; but the catchiness, and the power pop rock of the arrangement, contributed to my sense of the song's and the group's freshness, once people started making a big deal of the freshness.
The obvious thing to do now is to give an example or two of each of the various species of faulty explanation I outlined above. But actually I'd just rather get this posted, and if we want we can talk more in the comments; or I can do another post someday. I've talked negatively here (about faulty explanations) but I think that some of my explanations about [redacted] are good not necessarily because they avoid all flaws but because they explain a lot that no one else has explained.
But back to Crayon Pop: two fairly lousy explanations jump to mind. One is the idea that the Ilbe controversy helped bring attention to Crayon Pop. I believe that the proper response to this is "What controversy?" — i.e., it didn't bring attention to Crayon Pop since almost no one who didn't already know about Crayon Pop ever knew there was a controversy. But actually I don't know this, and this subject needs a modicum of research. If you conducted a survey, would you find a significant number of people saying they first heard of Crayon Pop owing to the Ilbe supposed flap? But face it, no one's going to conduct a survey, so what we (i.e., someone who is definitely not me) might want to do is to go online and see if there are a lot of Crayon Pop fans posting that they first heard of the group because of some Ilbe thing.
Second is the contention among Netizens and trolls that, since we know that for a not prohibitively large amount of money an agency can game the system so as to place a song high on the streaming charts even when a video is not genuinely loved and streamed, then, for all we know, Crayon Pop's agency did this very thing. Now, putting aside the fact that (1) maybe we don't know that streaming charts can be gamed (I certainly don't know one way or another, though I wouldn't be totally shocked to learn they could be), (2) last I looked no one who claimed to know that such gaming has occurred has actually named names or offered evidence (but I haven't looked much), and (3) there's no evidence that Crayon Pop's agency did this, there nonetheless are also other reasons why "they gamed the streaming" is not a good explanation (even if Crayon Pop's agency had tried to game the streaming charts). For instance... but no, if you've been paying attention while reading this post, you should be able to work out for yourself why such an explanation isn't likely to be good.
In any event, I think "Bar Bar Bar" sounds catchy, and I assume some other people do (but do they all find the same parts of it catchy that I do?). I think the dance funny and charming, and I assume other people do. The five-cylinder-jump is fun to imitate. That Crayon Pop are "different" is no big deal to me. (I've heard plenty of power pop in my life; that Crayon Pop all dress the same when doing the song is kind of funny, that the dress doesn't symbolize "sexiness" isn't "refreshing" to me but may be to some people and anyway I don't know that people actually consider Crayon Pop not sexy, which would be pretty stupid). The song, btw, started slow and was falling down the chart in its third week, with Crayon Pop no longer on the big TV performance shows, when the track got its second wind. That wind was already well in effect when they got a break and appeared on popular TV show Sketchbook; that performance probably helped but the track was climbing anyway, presumably from word of mouth about the videos, and people doing dance covers, but I don't know the reasons for sure. And at least some of all this, as well as Crayon Pop's already having a devoted (though small) fanbase, made it possible for the song to hit (as opposed to impossible), but none was a guarantee.
Also, in its little way, the song kicks butt. It rocks!
Footnotes in the comments.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
Footnotes
Date: 2013-09-21 09:00 am (UTC)**Another relevant factor, though one that I'm not going into today, is that if you graph relative fame and lack of fame you get a "power law distribution" or something somewhat like it, anyway ("power law" is one of the many parts of statistics that I don't really understand). What this means for our purposes is that there are only a few superstars, some but not a lot of somewhat famous acts, and a vast number of performers and would-be performers who are basically unknown. So the same acts keep appearing and reappearing at the top. And also, in any given week, there are only a small number of tracks (whether by an established act or not) with really high numbers; so if you look at a chart (e.g., the Gaon chart) that shows you points, not just rankings, you'll see that the track at number 20 (which this week happens to be by the same guy who's at number 1) is closer to the track at 200 than it is to the top.
***My piece was based on one by Duncan Watts and two of his colleagues (Matthew Salganik and Peter Sheridan Dodds), which you can find at Science magazine (the article seems to be there in full even without my logging in, but maybe I've got a longstanding cookie; if you have to register, it won't cost you anything):
http://www.sciencemag.org/cgi/content/full/311/5762/854
[For those of you who've read the piece, which I highly recommend, the one minor flaw in Watts' and crew's experiment is that there's no second "non-social-influence group," so we can't see if the second universe rates the appeal of the tracks more-or-less as the first had.]
****I do have the uninformed opinion that macroeconomics is a partial exception here, and that Krugman and ilk have pretty good ideas as to what causes various financial collapses, the recent one in particular. (Also, what I wrote in the post, that no explanation of such a single social event is testable, is contradicted by my correct assertion later in the post that some explanations are so bad as that we can eliminate them.)
no subject
Date: 2013-09-25 12:14 pm (UTC)but that's not necessarily the general explanation. Without reading anything about it and making assumptions here and now I assume the Paramore single is out in the same version it was out in originally, as a digital download on Itunes et al. I'm sure they've toured and promoted it in various ways, but..
I don't see that happen in 'other countries' (that I follow). Crayon Pop feels rare even if it was just a few weeks old. Some tracks _remain_ popular for a long time, like your favorite Billboard oldie, but very rarely do the general public just warm to a track over time before it floats to the top. It seems to me. And in Japan.
In the UK I am reminded of quite a few cases where mainland Europan singles sneak their way in, like Romanian summer pop singles hitting big the _next_ summer. But those are foreign artists.
no subject
Date: 2013-09-26 01:38 pm (UTC)Not counting airplay suppresses the impact of adult consumers and album artists on the singles chart.* Whereas in the U.S., not only is airplay part of the formula, it's also the major window into what's going on in specialty markets like country, R&B, adult contemporary, rock, and so on. And there's often a time lag between what a song does in a specialty market and its crossing over to the Top 40. Presumably a song by Blake Shelton, who's also a TV personality, will cross from country to Top 40 faster than a song by country newbie Tyler Farr.
Again, I'm being speculative; but the Korean music industry is event-oriented to an extreme, which probably means it's teen-oriented as well (though OST tracks, which I assume (w/out knowing) are timed to TV airings, probably get adults but are also event-oriented). Adults are less urgent in their entertainment consumption.
In the U.S. there are different trajectories depending on the artist. Pink, whose audience is now mostly grownups, almost always has a slow build on her singles. Whereas the recent Eminem single jumped to the top 5, then fell a bit, and is now holding steady and could even rise again as people decide they like it (though that's something I'm failing to decide). The Paramore track may be a bit of an extreme case: the group is middling in popularity, so the song's initial fan surge wasn't going to get it into the Top 40. But "Still Into You" is especially distinctive, and so it subsequently got a slow build from casual fans and people previously indifferent to Paramore. That's my guess, anyway.
I do track what gets into the U.S. Top 40, and over the past ten weeks about three singles per week (2.8) get in that hadn't been in before; these mostly break in between 30 and 40. So in that sense the U.S. pop market isn't very inviting. But countering that, there are a significant number of performers who are relatively new to the U.S. Top 40 (Avicii, for example), on their first or second Top 40 hit here. Except I haven't really researched this, so I can't say for sure. I know that in Korea over the last three years (as opposed to the few years before that), few newbies, especially few idol groups, have gotten into the top ten. My impression is that this is fairly recent in the U.S., that acts are breaking in to such an extent. And again, this is an impression that I haven't checked out.
In Korea no foreigners get near the top of the singles chart, though a track by Maroon 5 or Carly Rae Jepsen may scrape into the Top 40 and then hang around forever in the lower reaches.
*Which is suppressed anyway, I believe; I intend to do a post one of these days about that.
no subject
Date: 2013-10-01 11:08 am (UTC)Well, if the coin isn't balanced but no one rigged it, wouldn't the percentage attributable to luck still be 100 percent, even if the chance of heads is only 43 percent? But if someone were very careful in manufacturing the coin, so that there were far fewer imbalances and imperfections than in the average coin, wouldn't the fact that this was closer to 50-50 than most coins are be 0 percent luck — i.e., entirely the intention of the person who worked carefully on the coin? But what about the fact that this person had the time and the resources to design this coin, and wanted to, and that the technical means existed both to measure and create the precision she's after? That's not 100 percent intention and 0 percent luck — and what are we even measuring? Whose intention? And is the word "percentage" even applicable here? In 1900, the chance that I would be typing these words at exactly this moment (October 21, 4:18 AM Mountain Daylight Time) would be vanishingly small. But, while the chance at 9:00 PM last night would have been far greater, it'd still have been vanishingly small. But the chance in 1900 that someone in the year 2013 would be more or less thinking about these questions would be very high (basically, a certainty minus the chance that there'd have been a civilization-ending event such as a giant meteor strike or a nuclear war or [something possible but in 1900 as unknown to us as nuclear war]). But specifying a number, a percentage, hardly seems meaningful here, and what the word "luck" means is not so clear. That's what I was trying to get at above.**
Assuming that the principle of cumulative advantage is correct (and how could it not be?), the chance that there will be an ineradicable element of unpredictability — of chance — in what becomes well known is 100 percent. That is, unpredictability is a certainty. Random differences will get locked in. Even a change in human psychology won't change this, assuming we still pay attention to and learn from each other's behavior. (If not, we're no longer human, and probably no longer alive as a species.) But I continue to question the word "percent" earlier in this paragraph: it's not as if something's being measured. And the amount we should assign to random differences getting locked in — to chance — is just as elusive to me today as it was ten days ago, as is what my question about "amount" even means, conceptually.
*Or maybe someone who's thought about such questions far harder and better than I have could argue that specifying a vague number here isn't so meaningless. What shape would that person's argument take?
**But cosmologists rightly are concerned with asking what the chance is that our universe has the physical laws it actually seems to have, or anyway, what the chances are that it would have laws close enough to ours that the universe would still contain matter, and life. I believe Susskind and crew do at least try to assign numbers, unless I'm misremembering.
no subject
Date: 2013-10-20 01:04 am (UTC)http://en.wikipedia.org/wiki/Leonard_susskind
http://en.wikipedia.org/wiki/The_Cosmic_Landscape