Reconciling assholes for nearly a decade.

Is Justin Timberlake a Product of Cumulative Advantage?

http://www.nytimes.com/2007/04/15/magazine/15wwlnidealab.t.html

<quote>
The reason is that when people tend to like what other people like, differences in popularity are subject to what is called “cumulative advantage,” or the “rich get richer” effect. This means that if one object happens to be slightly more popular than another at just the right point, it will tend to become more popular still. As a result, even tiny, random fluctuations can blow up, generating potentially enormous long-run differences among even indistinguishable competitors — a phenomenon that is similar in some ways to the famous “butterfly effect” from chaos theory. Thus, if history were to be somehow rerun many times, seemingly identical universes with the same set of competitors and the same overall market tastes would quickly generate different winners: Madonna would have been popular in this world, but in some other version of history, she would be a nobody, and someone we have never heard of would be in her place.
</quote>

I can't think of any other reason Timberlake is so popular.
Permalink son of parnas 
April 15th, 2007 11:12pm
I imagine Madonna would have been just as popular in this or any other universe, though. That woman had ruthless ambition.
Permalink Colm 
April 15th, 2007 11:14pm
Agreed, love the material girl.
Permalink son of parnas 
April 15th, 2007 11:15pm
Oh, I can't stand her. Her and her fake pompous British accent.
Permalink Colm 
April 15th, 2007 11:18pm
I wish she could sing better. Other than that ...

This post is essentially a technically oriented rephrase of "The Tipping Point", isn't it?
Permalink $-- 
April 16th, 2007 6:55am
It's a technically reworded version of "If you got a thousand stock brokers to flip a coin a thousand times, one of them would get heads in a row a lot of times and be heralded as the new wall street genius."
Permalink Send private email Сергей РахманиноB 
April 16th, 2007 7:19am
I don't really agree. Timberlake or Madonna aren't random events, IMO.
Permalink $-- 
April 16th, 2007 7:23am
Timberlake has talent and charisma, and I'm told good looks. His being a superstar may be random, but it's not _that_ random. It's not "anyone could be Justin Timberlake" or "There were a million Justin Timberlake's out there and he was just the one that got lucky" random.
Permalink Send private email Сергей РахманиноB 
April 16th, 2007 7:28am
so how come the coin thing?
Permalink $-- 
April 16th, 2007 7:30am
I was just explaining the OP. The OP says that a slight advantage (winning a coin toss) turns into a big advantage in the next round because you're already ahead.
Permalink Send private email Сергей РахманиноB 
April 16th, 2007 7:45am
The coin example is wrong.
Permalink  
April 16th, 2007 8:13am
Because coins don't have a memory, while folks who read "People" magazine presumably do.
Permalink Send private email xampl 
April 16th, 2007 9:01am
Hmm!  That 1000 person thing would be a fascinating experiment.

Get 10 people in a room.  Have all 10 flip a coin.  Have the ones who got tails drop out. Repeat until everybody is out.  Record the longest train of heads.

Repeat this trial several times.  I'm sure the 'central limit theorem' has something to say about what you'll find.

From that earlier "interview question" thread on "JoS", I suspect the average length will be a 'log_base2' answer.
In other words, with 10 coins:
5, 2.5, 1.25, 0 == log_base2(10) == 3, so 3 heads in a row.

log_base2(1000) == 9.96, == 10, so 10 heads in a row.

So, how many would you need to get 1000 heads in a row?  2 ^ 1000 == <overflow>.  Well, 2^32 == 4.29e9.  2^100 == 1.26e30.  2^300 == 2.04e90.

We're over a google here, people.  This scam won't work for 1000 heads in a row.  Note that 5 billion years == 157e15 seconds.
Permalink SaveTheHubble 
April 16th, 2007 9:59am
> and I'm told good looks.

Too carefully said. What are you hiding?

> This post is essentially a technically oriented rephrase

Tipping point:  any process in which beyond a certain point, the rate at which the process proceeds increases dramatically.

Which processes does it apply to? As the article says, most people think acts succeed because of individual decision making. Here they show strong evidence it's not true. Good on science.
Permalink son of parnas 
April 16th, 2007 10:00am
Oh, and I misread the question -- "a lot" of heads in a row will happen, if "a lot" is around 10.
Permalink SaveTheHubble 
April 16th, 2007 10:01am
"Tipping point:  any process in which beyond a certain point, the rate at which the process proceeds increases dramatically. "

I'm thinking specifically as in the book.

basically, if you can achieve some popularity with (say) 100 people, and maintain it for maybe a year or two, things are likely to snowball. Word of mouth.

All advertising does is bias that process.
Permalink $-- 
April 16th, 2007 10:24am
I think it's just the Power Law.
Permalink Send private email strawberry snowflake 
April 16th, 2007 10:39am
> I think it's just the Power Law.

The power law is a description of real things, it's not generative. Aren't you curious about what it can describe and why? Or do you think one you label it you understand it?
Permalink son of parnas 
April 16th, 2007 10:58am
Yes, yes. You're right, son. I was being glib (knowingly, but there's no good smiley for that).

I think lekking, and ornamentation as in one of your previous threads, explains Justin Timberlake just fine: teenage girls have rather blunt status detectors. They look around and see what cultural status symbols everyone else is following (*). Music or cars, or whatever. Then they see who everyone is choosing: wisdom of the crowds, they intuit. If other people have taken the time to evaluate a Toyota or a Timberlake, and are not yelling foul, the choice can't be too bad.

Also teenage girls are mate ambitious: they are sill striving high, have the time and can afford to fantasize about obtaining the highest status mate.

One sees far less bunching up at the top among boys.

* Not an insult to women: I think late teenage/early 20's boys are equally herdable. There's just a few year offset.
Permalink Send private email strawberry snowflake 
April 16th, 2007 11:35am
the article is interesting. But there's more here than meets the eye. For instance:

"The song “Lockdown,” by 52metro, for example, ranked 26th out of 48 in quality; yet it was the No. 1 song in one social-influence world, and 40th in another. Overall, a song in the Top 5 in terms of quality had only a 50 percent chance of finishing in the Top 5 of success."

So : if you are talented enough to get into the top 5 of quality repeatedly, you are virtually certain to make the top 5 of success within a few attempts. And then, your name gets known, and it gets easier.

So the experiment is interesting but ignores the effect of branding, which I'd guess has a bit of an averaging effect on the unpredictability.

Hence, one hit wonders and so on.
Permalink $-- 
April 16th, 2007 11:43am
so what that also says, is that an approach of looking around for artists who seem to be able to maintain a higher quality level, and backing them over a longish period, is a better strategy for long term success than seeking the blans and trying to make it look like whatever was popular 6 months ago.

But of course the real reason for the decline of the music industry is actually illegal downloading, not fucked up A&R policies ...
Permalink $-- 
April 16th, 2007 11:52am
Branding is necessary because there are just too many choices. It works the same way with any other area: the time to evaluate N choices goes up O(N) but the quality of the resulting decision goes up at a much smaller rate. At some point one just wants to listen to some music not evaluate 1000's of songs/artists. Branding and cliques and stuff are a usability/optimization feature of social networks.
Permalink Send private email strawberry snowflake 
April 16th, 2007 11:56am
Timberlake has had some good pieces lately, so I don't think he's a good example of someone with no talent succeeding. Perhaps one can say that why does he succeed and not some other guy with the same level of talent, and exposure is the answer - that's why the labels paid radio stations to play certain songs, and then finally just bought all the stations to control this directly. It's just marketting that explains this, not some organic cumulative advantage theory.
Permalink Practical Economist 
April 16th, 2007 1:37pm
> O(N)

It seems a lot of people use a O(1) algorithm: what is the most popular?
Permalink son of parnas 
April 16th, 2007 2:00pm
marketing = status symbol outsourcing/mediation.

Status pimping has always been around, it's just not been a full-time job description for long.
Permalink Send private email strawberry snowflake 
April 16th, 2007 2:01pm
I want your sexy back.  Yes you, SoP.
Permalink Send private email sharkfish 
April 16th, 2007 5:21pm

This topic is archived. No further replies will be accepted.

Other topics: April, 2007 Other topics: April, 2007 Recent topics Recent topics