How Songs Get Popular

An anonymous reader writes "Researchers created an artificial music market of 14,341 participants split into two groups to pick music from unknown musicians. In one group, the individuals had only song titles and band names to go on. The individuals in the other group saw how others had rated the songs. Turns out popularity bred popularity, which explains why there's so much crap on the radio."

Just like /.

[Mrs. Grundy]

I think we all understand this. For instance here on slashdot this is the way the moderation works--things either don't get mod points or get the extreme value (not that I am hinting, dear bearer of mod-points, the you in particular lack independent judgement) . Pretty soon somebody will come along and mod this post down -1 as a troll. Seeing this, the next person with mod points will quickly mod it down as well--a kind of kick in the ./ groin if you will. If, on the other hand, the first person with points happens to have a wit worthy of Falstaff he will see the genuine insightful nature of this post and graces it with a +1. The result will be an avalanche of +1 placing this post among the few of well-meaning ineptitude that rises to empyreal absurdity. I'll leave it to the reader to determine which case illustrates the "Britney effect" mentioned in TFA.

So much crap in the radio?

[MikkoApo]

I blame the recording industry and its marketing. Popularity might breed popularity, but unfortunately marketing bypasses "real popularity". Unfortunately there are still artists making music which isn't spoiled by even if the system tries its best.

Barriers to Entry Falling = More Freedom

[Ritz_Just_Ritz]

Until relatively recently, the barriers to entry of the music business were sky high because of distribution costs. Now that distribution costs are going into a tailspin (iTunes & Bitorrent...Gracias!), the studios are scared out of their wits. Not because they're so worried about piracy, but because they can be cut out of the game entirely.

So I'm quite content to have actual listeners help shepherd in popular bands rather than have mediocre cookie cutter crap foisted on me by megacorps.

Makes sense to me

[SsShane]

When I want a new book and don't know which one, I go to Amazon and read reviews from others who have bought. It works for the most part. Oh well.

People like to talk about music

[AK__64]

Most people don't listen to music in a bubble, they talk to other people about the music and ideas get implanted in their heads. Also the way people talk about music makes a difference. If you say to me, that you LOVE this song and I HAVE to hear it, and download it and listen to it all the time, I'm going to look at you funny. But if you tell me in a laid-back, smooth and cool manner that this song is cool, I'll be more inclined to listen to you and less likely to write you off. It also works backwards. "I'm used to really like that song too, now I'm getting kinda sick of it..." Now you start to feel the same way, even just a little bit.
There are some really interesting studies on how people react in certain situations, responding to peer pressure and all that. Good stuff.

Uh duh.,..

[phaetonic]

Hence the Top 40 stations and lists

What?

[robyannetta]

I don't need anyone's complex mathematical computations to give me algorithims to find what music will be popular and what won't.

Has anyone actually listened to today's music? It sucks!

the variable that was changed

[blue_adept]

Although it's not spelled out, this study tested whether those songs highly rated by group A, would become more popular in group B, WHETHER OR NOT the ratings were actually true; in other words, the truthfullness of the ratings was the variable.

The acticle doesn't really dwell on this, but if that's not what they were doing, then what's so surprising about the fact that both group A and group B found the same songs to be "good". (d-uh, they're actually better songs!)

Re:What else did you expect?

[Tx]

From the mouths of Cowards comes wisdom. You're absolutely right, people find it incredibly hard to be random. Deprived of more meaningful criteria to base a selection on, they will use less meaningful, or even totally meaningless criteria. Ask me to choose between two cocktails that I've never heard of, and I'll probably choose the one whose colour I like best - I know it has no bearing on how good it's going to tase, but you have to choose, right? (Actually I'd ask for a scotch instead, but you get the point).

Not very surprising...

[lordsid]

The results aren't very surprising considering their "virtual music market" consisted of teenagers.

a couple of real-world examples

[zuki]

This has become such a science, there is just too much at stake for people who routinely invest 6 and 7-figure sums of money into a new album. (And I am not necessarily speaking about record labels here, it could just as well be about the associated release tour, which by now generates far more income than the actual CD sales). Focus groups, endless studies of people's buying patterns, major pressure from the 'top' (i.e.: management) to conform to a predictable sound, etc...

Here's a funny one, on a recent flight I was sitting next to the manager for some very well-known heavy metal and rock acts, who flatly declared that if U2 was a new band today, they wouldn't have a chance in hell of getting signed the way they did in 1983 when their breakthrough album propelled them into stardom. The people he deals with both at the label and promotion level would never take a chance on something that original.... Which of course means that after years of this kind of behavior, the general public's ears do not have a desire for anything new or unusual.

I could very well see a broke Jimi Hendrix today, still playing $100 fill-in gigs at Cafe Wah in the Village (still around too) and no one giving a rat's ass about his life-changing guitar playing because it would be too strong and outside of the norm....

Here's another example, last year a major game developer allegedly saw an increase of sales of their flagship PS2 game to the tune of 5,000 more units per week when they tweaked the music on their current TV campaign and featured background music that was more familiar to their target audience.....

This if doesn't seem like a game of chance and talent anymore, that's because because it isn't. Like P-Diddy said, it's all about Da Benjamins.

Still, it comes down to this: if you are going to do it, do it because you like it, not because of the expected returns.
If you actually have talent, you might go a lot further on that than the empty promises and broken stardom dreams most end up shelving when they get their girlfriend pregnant.

On another (closer) note, maybe someone should transpose this study to /. and do a research on what posts get rated and modded the highest, and how this does influence the writers to conform to a certain style that they know will get them modded? ... and does this make their style more boring and predictable?

How Darwinian!! Z.

Re:How ideas on Slashdot get popular

[Anonymous Coward]

Sorry to disagree with you, but I've seen many pro-Microsoft (particular cases of course, or maybe just points of view) or anti-Linux rants get a +5 Insightful.

That is true, but, how many of those moderations are metamoterated as 'fair' as opposed to 'unfair'?

Of course, true statements often get moderated as insightful. Is it our fault that most statements against microsoft happen to be true?

Oh you mean statements like 'Microsoft is a monopoly because they were convicted as such' as well as 'Windows is so unsecure and unstable'?

The first one, the have never really been a monopoly, as they have always had competition in the form of OS/2, Linux, and MacOS. Comparing OS/2 to Windows 95 or even Windows 3.11, it was much simpler to connect to the internet as OS/2 could not handle dynamic IP addresses, only static. Linux is only now maturing into a real product, and even then it's still in its early stages

Microsoft also has competition in the form of Openoffice.org. They also had competition that wasn't nearly as good because wordperfect and lotus both wanted to stick with 'tried and true' dos. Netscape ended up going nowhere whereas Microsoft was continually improving their browser. The reason Microsft came up on top is they knew how to compete viciously in the market. That is what Capitalism is. Then when they thought they had little competition to worry about, the Mozilla project was building up underneath their radar, now they have competition in the form of Firefox, Thunderbird, Openoffice.org, and Various Linux distributions.

As for the second statement, I haven't updated in a while and I have not had to reboot in a while and I have not had any security attacks on my system at all. The deal with security is mainly with attachments. I do not open unknown attachments and yes, I use Internet Explorer and Outlook Express.

IMHO, I think there is moderations and metamoderations that are based on what a person believes rather than whether or not someone makes a good point. Maybe cmdrtaco could implement a metamoderation system where 8 users can metamoderate one moderation, and base whether it fair or unfair on the majority of metamods so that it can be a little more fair.

On music, I have heard several song on the radio that I don't care for, but my taste is different than someone elses. I believe that there are some song that are popular just because it's popular, others are popular because it's good.

saves a lot of time

[DeveloperAdvantage]

Following the crowd probably evolved as a pretty good way of shortening the decision making process. If someone else ate a berry or mushroom and didn't get sick or die, then there was a pretty good chance that I could eat it too and would be ok. This saves a lot of time and energy instead of having to sort through everything by yourself.

Re:A social experiment

[garett_spencley]

"If you anounced on the radio that sales were exploding for an album by an unknown group and that the stores would be sold out before the end of the day people would line up so they wouldn't miss out knowing no more about the group than everyone else wanted the album."

Yup that is also a very common marketing trick too. It is exactly why every single new movie that comes out is "The #1 Movie In America!!!" and why every single new book is "The Best Selling Book" etc.

Re:A social experiment

[brian0918]

"Remember Beanie Babies? People were desperate to get them yet they were nothing more than a small stuffed animal and effectively worthless."

Everything is worthless unless people want it.

Re:So much crap in the radio?

[kfg]

By that logic, you are an ape as well.

I am not an ape by logic. I am an ape because that's what I am and cannot be otherwise.

KFG

Yep, makes perfect sense.

[Javaman59]

The human brain does not have the capacity, or time, to properly process all the information it receives, so it uses all sorts of aids to free it from information overload. Thus it forms habits, learns from repetition, and relies on other peoples opinions. For the average person, finding the "best" music would be a mental task which is not worth the effort, (they've got other things to do) so they use public opinion to find music which is "good enough". Just like buying cars, or software, clothes, or food. Our brain uses "conformity" in most things, to free it to make personal decisions in the things which are most important to us, whatever they may be.

Re:It's the Garmlich effect.

[Overly Critical Guy]

Turns out popularity bred popularity, which explains why there's so much crap on the radio.

Crap to whom? The nerds on Slashdot?

I find Slashdotters' attitudes toward the other 98% of the mainstream population quite condescending. If you don't like popular music, that doesn't make it "crap."

Rose Tinted Glasses

[Jesapoo]

yeah... You realise that this is what people have always said?
"Music in the [20's, 30's, 40's, 50's, 60's, 70's, 80's, 90's - select as appropriate] was way better than the nonsense nowadays"
Think about how many songs from those eras are actually still popular today. 99% of music has ALWAYS been crap, and we only remember the good 1%.

Re:Would the Beatles have made it today?

[Anonymous Coward]

Not to be a troll, but on what grounds do you claim that today's popular music is of "lesser quality"? I don't know a hell of a lot about music, but is popular crap really unique to this day and age? I would suggest that, in the 1800s and the 1960s and every other period of human history, there existed popular artists who weren't that talented and talented artists who weren't that popular.

Maybe you have some recognizable measure demonstrating the declining quality of popular music. But I don't think you can get away with merely stating it as implicitly factual.

Re:Would the Beatles have made it today?

[mblase]

Would the Beatles have made it today?

Of course not; the question is inherently absurd. Their music was popular mainly because it was radically different from anything that people were listening to on the radio at that time. Since then, they've influenced musicians thousands of times over on both sides of the Atlantic.

But new styles still make it big now and then. Think of the fads of ska or swing dancing in the 1990s, or the gradual rise in popularity of rap from a niche in the early 1980s to the mainstream today.

Re:Crap Rock

[FLEB]

There's also the fact that 20-30 year old music has had quite a bit of time to let the cream rise to the top (or barring the cream, at least the most well known). I've found that I can probably sing along to 3/4 of the stuff on a classic-rock station, and that's probably not extraordinary. When you consider the sheer amount of music that was produced in those years, the classic-rock radio stations (or most other genre stations, for that matter) are only giving a sliver subset... there was much more "music of the times".

"-1 troll" utterance gets +5 Insightful

[Anonymous Coward]

Everytime someone utters "This will probably get moderated -1 troll, but here I go anyway" ends up getting +5 Interesting or Insightful. I'm not trying to get off topic here, but I've seen this happen on numerous occasions.

Oh well, this will probably get moderated -1, Troll but it had to be said.

Re:Would the Beatles have made it today?

[westlake]

today stars are foisted, created, presented to the consuming public by fiat, not a great surprise

explain to me how this differs from (and is inferior to) a traditional patronage system in which an aristocratic elite gets to decide who performs in public at all.

No, no no, you've got it all wrong...

[MayorDefacto]

That's not the way music gets popular. Here's how it happens:
1. Hold auditions at local malls, car lots, county fairs, etc. to find hot young white jailbait
2. Tart up aforementioned jailbait and teach them some slammin' dance moves
3. Get a committee of marketing people together to craft some lyrics that are as sexualized as common decency (read: FCC) will allow. Bonus points if corporate sponsors can synergize their product into the lyrics somehow (if not, don't worry, the product placement people will cram as many soft drinks, cell phones, and designer handbags into the video as possible later)
4. Get some underpaid, under-recognized sound engineers (read: geeks) to put together a cathcy little number on the sequencer. Don't worry about horrendous vocals, those can be corrected in the final mix.
5. Shoot video. Don't worry about making it creative, just fill it with Bentleys, Prada, diamonds, and lots of writhing, Cristal-soaked booty. Bonus points if the video is so over-the-top that a controversy ensues (don't worry, MTV doesn't show full videos anymore anyway-- they'll just show the 20 seconds of the video that isn't offensive on TRL and we can make a mint by selling the "uncut" version on iTunes.)
6. This is the most important part: PAYOLA, PAYOLA, PAYOLA! How will your song ever get popular unless all the top-40 stations play it once an hour, every hour? Make your check out to Clear Channel, and they'll take care of the rest.
7. ???
8. Profit!

Actually, no.

[porcupine8]

It's more related to the fact that articles with lots of comments tend to snowball into even MORE comments, into the hundreds. While those that only get a couple dozen tend to stay in the low numbers.

Because the mod ratings are (ostensibly) based on quality, which in this article was shown to have nothing to do with popularity. Group B did NOT download songs based on the quality ratings that Group A gave them - only based on the number of times the songs were downloaded. Popularity was totally independent of rating/quality.

Not really...

[MacDork]

It's just seems that way to you guys who only read +3's,+4's, and +5's. You never see the conflicted mods... For example, I made a recent post [slashdot.org] that defended an unpopular opinion around here. You never saw it because it only scored +1 informative. It got modded 50% informative, 30% Overrated, 20% Flamebait. At least 4 different mods there, -2, +2.

I read +6 Troll, Flamebait, etc... A lot of mods don't know what the hell they're talking about and if it goes against groupthink, it goes down in Flamebaits. When it does, there are people there like me to pick it up and give it an informative, insightful, or interesting boost. Not everyone runs on default mod settings here at /. Genuine flamebaits and trolls are getting much rarer. I see a lot less GNAA and WIIIIIIIIIIIIIIIIIIIDDDDDDDDDDEEEEEEEEE crap here these days. (With the exception of Apple Trolls. They never go away. They even get Cover Storys in Forbes. [forbes.com] "Likely to top 4 Million units" for iPods. Dipshits... they sold 14 Million [apple.com]) Most of the down mods go to people who simply think differently lately.

Now, so that I'm not totally off topic... the article describes a system where one group could only listen, see track title, artist name, and download. The second group could see all that and could see download counts as well. Wow, the ones that were downloaded most got the most attention and additional downloads... Duh. That's not scientific. There's no F'ing experimental group! Why didn't they have a third group that could see everything group #1 saw, and *randomly generated* download counts? If I see a song has been downloaded numerous times, listen to it, and it's crap, I'm sure as hell not downloading a copy to save if it sucks. I don't care how many people listen to something, but I would consider download counts an indicator of what I should try first... At least until I realized the download counts were meaningless. If they repeat the experiment with the third group and that group downloads random crap like lemmings then maybe they have something worth reporting... Otherwise, they've proven nothing.

Re:A social experiment

[bxbaser]

I used the same trick when i used to sell at swap meets and shows. one person from the booth would stand in front of the table when it got slow, never failed to bring a few more people over.

lastfm charts

[comradevik]

I fully agree with the article but then i also think that thsi music gets pushed by producers and record companies. sometimes its "cool" because it keeps airing on MTV and the crowd will follow to watch. But I think that's only for a short time. what is popular sometimes doesnt correspond with what the majority actualy listens to. I was just lookign at the LastFM charts and bands like Radiohead and Beatles are always on top. Sure MTV might push Britney Spears on top but it will never replace the music that is better by quality. I will rush to download a song that's popular to hear it and listen to it for a while but if the music is bad it will eventualy get off my playlist and Beatles will stay.

Just because *you* don't like radio...

[NotQuiteReal]

Just because *you* don't like a radio station doesn't mean others don't like it. Maybe you are listening to the wrong radio station? Maybe your radio station doesn't exist.

Some people like to listen to rock, some like country. Others like "contemporary" or whatever. Others still, listen to NPR and some listen to Rush Limbaugh. Not many listen to all of the above.

Radio is not about pleasing you it is about making money by attracting enough listeners. MP3 downloads not withstanding, you are not entitled to free entertainment that you like.

Just as there are not enough listeners for an all-opera-all-the-time station. Maybe there are not enough listeners to support your odd taste in music (maybe you want all-opera?). If you think most people have crappy musical tastes, what do you think most people will think of your choices in music?

If you don't like the radio, buy your own music. If you don't like the normal labels, try "independent" sellers. I have purchased several albums from "cdbaby.com" - but, you know what? Much of the music is unremarkable... maybe the labels do know something about picking music that people will like?

Sometimes you do find a gem; one indy album I bought was the www.solvingforx.com album. At least I like it, but that's the problem - there is no objective standard to test music. So, you are left with markets, marketers, hucksters and hype. People like what they like, or what they think they like; What's the difference?

Re:Would the Beatles have made it today?

[damiam]

The beatles fucking sucked. Off-tune, off-time.

No one is saying that the Beatles played/sang with perfect technique. That's not necessarily the mark of a good band. Their technique was good enough for them to get their music across.

They became popular mostly due to their haircuts, and the fact that they were doing something new.

Exactly. They were doing something new. They were innovative and creative and they changed the face of music. Take a songwriting class sometime and you'll see how much of modern rock/pop is based on ideas introduced by the Beatles. Even if you don't listen to a lot of their stuff (and I don't), it's stupid to deny that Lennon/McCartney were musical geniuses.

The conclusion is bubcus

[CandideEC]

The average intellegent person uses the work of others to build on when attacking a task. If I where presented with a ton of songs to download it would take quite an investment to download and listen to them all to find out whats good. I would use the songs number of downloads in an attempt to lighten the load. Of course I would prefer to sort by ratings...like I would at amazon or newegg...but if ratings are unavailable I would go with downloads. This has nothing to do with being affected "socially". I am using others time to lighten my work load. Unless there are a very limited number of songs I don't see any other way to go...who is going to base there pick on artist or song name? That correlation sucks compared to downloads. Not that I think downloads is perfect...but it makes a nice razor.

Popularity

[vanyel]

Popularity breeds popularity because it's easy: someone else has done the work for you. If someone else likes something, there's a much better chance that it's good than a random sample of all the music (or whatever you're rating), because 90% of everything really is crap. It takes someone determined to find the jewels to wade through it all to find the new stuff that really is both original and good. If you think something is crap because it's popular, aside from the arrogance and elitist attitude that implies, the same principle still applies, because it scales down to the people who like the same type of <x> that you do: when one of your subgroup finds something, you'll probably like it too (and the multitudes will probably think it's crap in return). And popularity will breed popularity in your subgroup.

Re:"-1 troll" utterance gets +5 Insightful

[gavri]

Maybe it's only the ones that get +5 insightful/interesting that you see. The ones that say this will probably get moderated -1 troll and then actually do get modded down -1 Troll you don't see because they are "beneath your current threshold"

Re:"-1 troll" utterance gets +5 Insightful

[somersault]

I'm pretty sure I browse at -1, I dont get told any of the troll/insightful stuff..

Also have been shocked at being modded a troll once in the past (or was that twice) when I thought I was just expressing my opinion *shrug* maybe I'm a troll. I should do that thing 'oh I'll probably get modded down for this but..'. I dont even know what the point of the whole mod points thing is, seems fairly arbitrary and hopefully most people like to read all posts, not just the ones that are deemed 'interesting'.. >_> otherwise we may all end up missing the posts where people actually spout facts instead of just hearsay and their opinion. My own opinion can swing wildly from one side of an argument to the other while reading /. , and we may never know in some cases which posters have the facts right..

A Poor Study

[Se7enLC]

The study grouped people into two groups "Independant" and "Social Influence". The problem is, they have no control group, as BOTH groups are real people, and thus have social influence already.

The way the study worked (from my understanding of the article) is that one group could pick songs by title and artist and the other could search by title, artist, and popularity. The results were that the same songs were popular in both groups! Wow, Amazing! All you did was prove that the outside influence on the study was the same! People don't need a list of "most recently downloaded songs" to know what they heard on the radio. I imagine that a lot of the people in the study (when given the opportunity to legally download as much as they wanted) went to another site to find what music is popular and looked all of them up. Or asked their friends "what should I download?" thus reproducing the same effect.

What would have made an interesting test is to have NO artist or title information at all (Artist 123 - Song 6) and run the same test. The problem would still exist (when people recognize a song, they would rate it higher or download it more often), but you would have to listen randomly and rate songs based on actual quality, not on popularity. It would be like a radio station but random instead of being force-fed the popular songs 5 times per hour.

Re:Not really...

[radtea]

Why didn't they have a third group that could see everything group #1 saw, and *randomly generated* download counts? If I see a song has been downloaded numerous times, listen to it, and it's crap, I'm sure as hell not downloading a copy to save if it sucks. I don't care how many people listen to something, but I would consider download counts an indicator of what I should try first... At least until I realized the download counts were meaningless. If they repeat the experiment with the third group and that group downloads random crap like lemmings then maybe they have something worth reporting... Otherwise, they've proven nothing.

False. The probability of a song becoming popular in the group that saw the download numbers was poorly correlated with the rating given the song by the group who did not see the download numbers. These two pieces of information are sufficient to untangle the relative importance of perceived quality vs popularity. Any competent statistical analyst would know that.

Simply because you cannot see how to do something does not mean it cannot be done, or that that people who designed an experiment that you are not competent to analyze are stupid. This study has all the information required to show that we do exactly what you say you would never do: we tend to follow the crowd, regardless of the crowd's taste. I have no doubt I do this, and I am equally sure you do. This sort of group-think tendency is one of the fundamental aspects of human behaviour that makes us such successful animals, able to form large social groups and communities spontaneously on the most flimsy bases.

Denial of a scientific result because it runs contrary to what you would like to believe about yourself is very popular in places like Syria and the Whitehouse just now, but it is no sort of behaviour for any self-respecting human being living in a secular age.