Athabasca University | AU Student/Staff Login | Invited Guest Login
Metcalfe's Law is Wrong - IEEE Spectrum
Compelling argument, 10 years old now, that Metcalfe's Law and Reed's Law are wrong, and that the correct value for a network should be n log (n). The reasoning is good: the problem with Metcalfe's and Reed's laws is that not all nodes in a network are equal. An analogy is made with Zipf's Law ("if we order some large collection by size or popularity, the second element in the collection will be about half the measure of the first one, the third one will be about one-third the measure of the first one, and so on") which reflects the uneven distribution of value in a network.
This makes sense to me, but could be taken further. It seems to me that there is no such thing as an 'average' network, so we must always examine the actual patterns in any given network to see what value individuals add, and we must always be prepared for some serious outliers that can greatly affect the overall network. If, say, a prime minister or president started to use the Landing, the effect would be quite spectacular (and likely catastrophic, for all sorts of technical and non-technical reasons). There are great risks in averaging things out and looking for statistical effects when observing any human system.
- Social Media Use Linked to Developing Depression Regardless of Personality | University of Arkansas
October 4, 2022 - 5:00pm
Jon Dron - The wisdom of crowds versus the madness of mobs: An evolutionary model of bias, polarization, and...
October 29, 2022 - 10:19am
Jon Dron - My latest paper: On the Misappropriation of Spatial Metaphors in Online Learning
January 3, 2023 - 12:49pm
Jon Dron
Jon Dron
Comments
Very interesting...
Feb 09, 2016 – your post based on an article by Bob Briscoe, Andrew Odlyzko, Benjamin Tilly from http://spectrum.ieee.org/computing/networks/metcalfes-law-is-wrong
re: "build it and they will come."
I welcome your view on the Landing as it relates to the statement above -- I have been wondering about this issue for a while now...
@Rita - there's a book or two's worth of answers to that! As a starting point...
In itself, building something rarely achieves much, but you'll generally get more visitors if you build something in an already busy place than if you put it in the middle of an uncultivated forest. On the other hand, if you can tell enough people about it and there's a compelling enough reason for them to go, that uncultivated forest can quickly become a busy place, and so can the soon-to-form roads in between, so you get growth. We shape our dwellings and afterwards our dwellings shape our lives: it's iterative and recursive.
I take a lot of inspiration from Jane Jacobs, who identified short blocks, diverse primary uses, dense population and mix of old and new buildings as the essential characteristics of thriving city areas. Each of these aspects feeds the others - it's a complex adaptive system with an evolutionary dynamic. On the web, short blocks are about quick and easy connection (we're still working on that on the Landing - a bit of a maze, with too many dead-ends). The more connection, the more people pass by and can travel between areas, allowing diverse sub-communities and uses to thrive. A mix of old and new relates to the low-threshold and high-threshold design (we could differentiate these better but there is a fair mix of quick-and-easy and carefully designed here: anyone can have a presence, but it is possible to design quite complex and elaborate spaces). That drives diversity of primary uses. A central issue is density of population - that's the big driver that drives itself once a threshold is crossed, as long as it does not become too high (Facebook and now Twitter face that problem in spades, responding with filtering algorithms that bring a whole new set of problems). People attract people. And density of population, in turn, leads to and is driven by diverse primary uses. The more reasons you have for being somewhere, the more you will visit and the safer the place will become because there will always be others around you. It's a mix of hard (structural) and soft (social) drivers. In the case of the Landing, it is complicated by the overlaid existing structure, especially of courses, student roles and staff roles, which is both a good thing and a bad thing. Good because it creates density and provides reasons to be here, bad because it reduces diversity and can detract from agency and ownership: people see the Landing as part of a different structure, rather than a thing in itself. And ownership - perhaps more accurately, belongingness - really matters. And, as the article bookmarked above shows, the nature of the activities and connections between people makes a huge difference. That is both formed by context and forms it.
I could write (and have written) a lot more on this! Alas, time is at a bit of a premium right now but see both my books for more. In brief, it's a complex adaptive system fed by people and set in an ecosystem that also provides inputs and receives outputs. It's pretty hard to design structures from the top down to support this kind of environment but, as long as they are flexible, adaptable, richly connected, parcellated and a few other things (for a few more, see my paper at http://www.ifets.info/download_pdf.php?j_id=36&a_id=769) with a bit of nurturing they can evolve to fit their communities.
http://www.ifets.info/journals/10_3/5.pdf -- thanks for the link!
Abstract: “…With especial benefits for lifelong learners and those outside institutional boundaries.” -- Re: those outside institutional... -- How so?
Conclusion (I noticed a couple of typos…): “Social software in e-learning offers great potential pedagogic and practical benefits, both through the amplification and creation of social ties, and through allowing learners to choose (to choose) whether they (want?) to control or be controlled in a learning transaction. --- Twitter, Facebook, etc. (staff interacting with clients, with management, etc.)– why not the Landing (staff, faculty, students)?
And…”A self-organizing environment is not necessarily an effective learning environment…”
Good point. Which one do you consider the Landing to be?
P.S. “(we're still working on that on the Landing - a bit of a maze, with too many dead-ends). “ – I would be interested in making this part of a future project…
I enjoyed reading this paper – thank you!