Categories
Computer Music

What Sharky Laguana gets wrong

Sharky Laguana, if that is your real name, you get at least two things wrong in your article about streaming music.

  1. It’s spelled “aficionados”;
  2. Your argument about inequity depends on facts you don’t have about distribution of choices.

We can do number 2 a couple of different ways.

Your Rdio spreadsheet example only works, with its difference between columns, on the premise that Brendan is the only person paying each of the artists for which the inequity is great.

Let’s do an example. Suppose that everyone, on average, has the same musical tastes, and listens to artists at the same rate. I know that’s not true — but if it were, then clearly there would be no difference between the Subscriber Share and Big Pool methods as you call them:

1/8x +1/8y = 1/8(x + y)

Your argument requires the premise that distribution of artists listened to is very different at different streaming levels. Do you know this? If so, how?

Finally, I find it funny that you think that these companies aren’t aware of the problem of stream falsification, and aren’t working on addressing it.

 

3 replies on “What Sharky Laguana gets wrong”

Sharky Laguana is my real name. Thanks.

1. Thanks for the spell check. 3,000 word piece and you found an extra “f”. Impressive.

2. “Suppose that everyone, on average, has the same musical tastes, and listens to artists at the same rate. I know that’s not true — but if it were, then clearly there would be no difference between the Big Pool and Subscriber Share methods as you call them”

Correct.

“Your argument requires the premise that distribution of artists listened to is very different at different scales of streaming levels. Do you know this? If so, how?”

There has been a reasonable amount of academic research into this topic (explained in the FAQ with links). See papers and research by Rasmus Rex Pedersen, Arnt Maaso (who used actual service data), Chris Mulligan, Anita Elberse and David Touve at Rockonomic.com (who recently proposed co-writing a paper with me before pulling out due to time constraints) to name a few.
Elberse in particular has looked at how different streaming levels impact diversity in individuals. I don’t know of anyone who has done the same research for groups (such as offices and hair salons), so if you are saying you want to jump in and do that research, that’s awesome. When can I read it?

Elberse’s research conducted in 2007 suggested that the more a user streams, the more diverse their choices are. However, some caveats: this was early days of streaming, with a hetregenuous sample that has, to my way of thinking, some issues (such as some subscribers not having unlimited access, either because of the plan, or their own internal bandwidth constraints), and didn’t really look at how user behavior changed over time. Intriguingly her research showed that users with unlimited plans had a lot more diversity than users with limits. I guess they get bored of the old hits, and it doesn’t cost them much to explore so they do.

But for me the bottom line is that music subscribers are deeply heterogenous. You have highly diverse listeners (like me for instance) with comparatively light usage (I’m probably 200-300 streams a month), and extreme super-fans who listen non-stop to single artists (40 years ago it was Elvis. Maybe now it’s Taylor Swift). You also have the opposite: shallow diversity with light usage, and heavy diversity and heavy usage (as Elberse found).

But this is getting needlessly complex. Let’s just look at it as a simple moral issue, which is largely how I present it: is it right that someone else gets to decide which artists receive your money?

My answer is no.

I think if an artist brings a fan to the table, and entertains them (feeds them with food at the buffet if you will), then they should reap the proportionate rewards.

If there’s not much diversity, as you seem to be concerned about, then there won’t be any harm in switching. However as Rasmus Rex Pedersen tweeted to me this morning, Subscriber Share is the fairer solution.

3. “Finally, I find it funny that you think that these companies aren’t aware of the problem of stream falsification, and aren’t working on addressing it.”

Please re-read what I wrote. I didn’t say they were not aware of it, nor did I say they were not working on addressing it. In fact I listed several things they were doing. I did say the problem could be worse than they admit, or *possibly* even realize. Even Spotify’s own in-house data analysis firm The Echo Nest acknowledges that fraud is “becoming a pretty huge problem”
http://www.bdcwire.com/spotify-hacking-how-is-it-changing-the-music-industry-who-is-to-blame-and-what-is-the-solution/

The problem with click fraud, in my mind at least, is that at a fundamental level there’s no perfect solution: you can’t know what a person’s *intention* was when they clicked on something, you cannot see into their heart. You can create algos to detect patterns, and you can get pretty sophisticated with UID’s and other tactical measures, but there are countermeasures for all of these methods, and equally sophisticated click fraud operators know how to probe, evade, and conceal.

It’s the proverbial arms race.

But the best indication of the current inability of the services to deal with fraud requires little or no inference at all, and we can use direct observation: they can’t even stop the basic simple shit. Like I, Sharky Laguana, in my own (real) name, on my own account, streamed my own catalog, which also has my name in plain text ALL over it, 24 hours a day for for an entire month a couple months ago and absolutely nothing happened. Except a slightly bigger check from The Orchard (who is a content aggregator that acquired my original aggregator, IODA, a company whose business plan I checked for spelling errors since my best friend was starting it. Apparently he should have hire you!)

So yeah, what else you got?

Oh and btw, I published the article last night at 8pm, and have been defending it in all corners of the internet for the past 24 hours with one 4 hour break for sleep. So, after repeating myself ad inifinitum in several venues, and also without the benefit of caffeine (on doctor’s orders), I am quite a bit more sardonic and snippy than usual. Please forgive me. Looking forward to discussions in the future.

Thanks for the detailed comments.

Spelling and crooks aside, this seems to me to be the central issue; you say

“But for me the bottom line is that music subscribers are deeply heterogenous. You have highly diverse listeners (like me for instance) with comparatively light usage (I’m probably 200-300 streams a month), and extreme super-fans who listen non-stop to single artists (40 years ago it was Elvis. Maybe now it’s Taylor Swift). You also have the opposite: shallow diversity with light usage, and heavy diversity and heavy usage (as Elberse found).”

So, for all we know, it’s a wash across all listeners.

I agree with your claim that if streaming services move to subscriber share models than we can avoid the (possible) issue altogether.

Your article makes it sounds like you know that there’s this emergency with artists getting screwed, and purport to demonstrate this with charts and figures. You haven’t demonstrated this, and the research you cite is utterly indecisive on the matter — do we agree on that?

I wouldn’t say it’s a wash, and I’m afraid we don’t agree that the research is “utterly indecisive”. That’s far to the left of where I would put the research: some things are known and some things are not, but a fair amount of work has been done and we do have at least a rough sense of where the things that are not known will wind up. Conclusive beyond all doubt? No. But not “utterly indecisive” either.

With regards to it being “a wash”, you are thinking like a scientist, and looking at macro results, and not considering the value proposition at the micro level – at least I don’t get any sense that you are. I think I can illustrate the problem with an analogy: let’s imagine a box with two mice in it. We have our choice of two poision gases which we can pump into the box. As it turns out either gas will kill one of the mice. If all we care about is which gas will promote overall survivability, then the gases are interchangeable. “It’s a wash”

But what if I told you that one of those mice had a rare gene that will lead to a drug that can cure small cell carnicomas?

Well shit, now we care a lot about which mouse will live. And now we have to carefully decide which gas will not kill our precious cancer-curing mouse.

This is what’s happening in music: there are rare beautiful mice, that if we allow them to live, will be the musical equivalent of the cure for cancer. And then there are mice who just don’t matter. The goal here is to save the beautiful mice. These mice tend to collect “fans”. The other mice tend to collect “clicks”. If you favor the mice that collect fans, your odds of curing cancer increase significantly, benefiting the entire cultural ecosystem, even as the overall survival rate remains unchanged.

Now as for the distribution: we know that heavy listeners are more diverse, but also that they listen to more pop. We know that light listeners are less diverse but that there’s a hell of a lot more of them.

So back to our box of mice: let’s say our cancer-curing artist needs $100 to stay alive/not quit music/not give up. And let’s say he has 100 fans.

As a former semi-world famous musician (SPIN magazine best new artist 1998) let me tell you that the first 100 fans are the hardest to get. So this isn’t some silly this-could-never-happen-in-real-life scenario.

In a Big Pool world he needs 14,285 plays to make $100. If his fans are all working people who only have time for about 100 plays a month, there’s no way for him to get there. They will only total 10,000 plays, even if they played his music exclusively he’s only going to make at most $70.

In a Subscriber share world he needs $1 from each fan. Or $2 from 50 fans. Or $3 from 33 fans and so on until we get to $7 from 15 fans.

It doesn’t matter that they only play 100 tracks a month, subscriber share provides a possible path for him to get to his goal, with the fans he has, in each and every one of these scenarios: 100 fans play him 15% of the time, 50 fans play him 29% of the time, 33 fans play him 43% of the time, or 15 fans play him exclusively.

This is what we want: a toehold for those who need it most, and who have the most to offer us in return.

Leave a Reply

Your email address will not be published. Required fields are marked *