Subject: Re: near/medium future digital media economics
From: "Ben Tilly" <btilly@gmail.com>
Date: Sat, 20 May 2006 02:18:00 -0700

 Sat, 20 May 2006 02:18:00 -0700
On 5/19/06, Stephen J. Turnbull <turnbull@sk.tsukuba.ac.jp> wrote:
> >>>>> "Ben" == Ben Tilly <btilly@gmail.com> writes:
>
>     Ben> OK, this is the copy that I intended to send. :-)
>
> Say "hi" to your secretary for me.  The first one seemed very
> familiar. :-)

You have one of those as well? :-)

>     Ben> Of course most of us are interesting to someone.  My point
>     Ben> was that very, very few of us are interesting to  everyone .
>     Ben> Therefore most people do not derive much value from the
>     Ben> potential connections to most other people.
>
> A meaningless qualititative statement.  There are plenty of ways to
> define "most", "much", and "most" to get any result you want between
> badly sublinear and Metcalfe's Law.

Vague I'll grant you, but not meaningless.  As a human being you know
as well as I that you do not value all potential connections equally.
You don't necessarily know how to readily quantify it (truthfully
neither do I, though I've put some energy into it), but it *means*
something to you.  You have a referent in your experience for it.

>     Ben> Bradford's Law says for a large number of journals n, the
>     Ben> number of articles of interest that can be found will be
>     Ben> proportional to log(n).  The number that a researcher will
>     Ben> find using a traditional library search will be fixed because
>     Ben> researchers do not actually search more than a few journals.
>     Ben> So the actual value (measured here in number of articles
>     Ben> found)
>
> I think that's an extremely questionable measure of value.  For
> example, consider a patent search.  The marginal value, as measured in
> probability of a successful application, of finding an additional
> "near-precedent" increases as you near 100% of near-precedents.  It's
> arguable that searches for academic papers show a similar increasing
> marginal return, because (after "your math is wrong") the most common
> reason for rejection is "we already knew that, see the paper by So An
> So in the Singapore J. of Unsearchable Results."  Or, as the proverb
> has it, "you always find what you're looking for in the last place you
> look".

The more quantifiable a measurement of value is, the easier it is to
argue that "this isn't what I mean by value!"  I agree that it is a
debatable measure of value.  But it is not an unreasonable one, and it
has the strong benefit that I at least have some idea how to quantify
it.

>     Ben> Whichever one you argue for [...] you're going to find that
>     Ben> the general scaling behaviour is VERY different than
>     Ben> Metcalfe's Law projects.
>
> You're forgetting that I'm a coarse-filtered economist, who only knows
> of three kinds of scaling behavior: decreasing returns to scale,
> constant returns to scale, and increasing returns to scale.
> Obviously, I will find that O(n^2) and O(n log n) are in the identical
> class.

In which case you don't care about the entire subject and we can skip
this conversation. :-P

>     Ben> However this alternative scaling law (which Bob Briscoe,
>     Ben> Andrew Odlyzko, and I have tenatively nicknamed the BOT law)
>     Ben> says that as long as the smaller network can reach sufficient
>     Ben> size, network effects do not longer pose an insurmountable
>     Ben> obstacle to its continued survival.
>
> Huh?  The alternative scaling law says no such thing in the absence of
> assumptions about costs.  What are those assumptions?

I thought we were in the internet era and everything was free. :-P

Seriously, you're right and I'm wrong on this point.  Costs enter in
in non-trivial ways.  For instance if the cost of delivering your
network is the same whether you're delivering to 100,000 or 1,000,000
people, then the larger competitor has an absurdly large advantage.
If your costs are fixed per member (the case I was actually thinking
of) then the larger competitor has a significant advantage, but it is
possible for the smaller competitor to actually be that much better.

>     Ben> This is a significant prediction.  It can help explain why,
>     Ben> for instance, Apple has been able to survive despite the
>     Ben> positive feedbacks from Microsoft's far larger installed
>     Ben> base.
>
> Help, yes, but how much?  I think "niche market" is a much better
> explanation.  It's very easy to predict who will use Macs vs Windows
> in my experience.

It used to be easy for me to predict who would use Macs.  However
there are now a lot of people using them who didn't not that long ago,
and predicting is getting harder... :-)

As for "niche market", I tenatively think that competing networks
naturally will divide a market into distinct niches (unless one simply
drives the other one out), which state can be surprisingly stable
since members of each niche get reinforcement to stay in their own
network.

Incidentally I'm somewhat amused at the description that while this
was happening with AT&T versus local telcos they naturally divided
themselves into the niches "very rich" and "rich".  (In those days
telephones cost too much for the middle class to use them.)  The way
this fell out is that a local telco would establish itself in a city,
then AT&T entered when they got long-distance lines there.  AT&T
fairly easily would come to own the niche of people who were so rich
that they regularly wanted to talk to people in other cities, while
the local telco would maintain its hold among the merely rich who
wanted to talk to other people in the same city.  Businesses that
needed to deal with both would be on both telephone networks.  And
this division was suprisingly stable.

[...]
>     Ben> I predict, by contrast, is that while a company can establish
>     Ben> themselves, the barrier for a second competitor will be
>     Ben> significant but not insurmountable.
>
> Two is not competition.

It is more competition than one...

> [...]
>     >> It is nonetheless extremely useful, because for any given human
>     >> the set of knowledge for which he has even wannabe status is of
>     >> measure zero.  And it may be successful enough to evoke a
>     >> Gresham's Law of Encyclopedias unless somebody figures out a
>     >> way to take advantage of the medium for high quality
>     >> encyclopedias.  That's where Tom is going.
>
>     Ben> I wish him luck.  Based on my past experience, I think that
>     Ben> high quality collaborative content depends on having some
>     Ben> barrier to entry, which can be subtle or overt.
>
> I believe that that is what Tom means by his talk about "Wikipedia is
> not a true network."

I don't care to speculate on that.

> I wish him luck, too, and he's going to need it.  But he doesn't need
> a miracle.  Remember that the set of networks you can draw on a set of
> nodes is pretty damn big[1], and we have not gotten very far in the
> process of examining how to design "productive networks" to support
> the "social networks" that actually produce "final value".  It is not
> clear to me that we can't find some networks that provide a useful
> compromise between the chimeric promise of Metcalfe's Law and the
> inefficiency of the collection of all pairwise networks.

We disagree on how big a miracle he would need.

Cheers,
Ben