2 ms·
Blogging readership turns out to follow Power Law distributions. That's hardly surprising. It does have significance for anyone promoting online peer-to-peer
by dredmorbius 5y ago
Blogging readership turns out to follow Power Law distributions. That's hardly surprising.
It does have significance for anyone promoting online peer-to-peer or networked activities as a replacement for localised production, services, or labour.
This type of long-tail, mostly-nothing distribution is typical of all media and communications or network phenomena. Ultimately attention is finite. Both in the individual case, as is increasingly well-recognised, but also in the collective sense: A public only has 10 "Top Ten" attention slots, no matter how large that public becomes. And a single unitary public means that there's only one contest, rather than multiple divided ones. That's the future of a media-dominated information world.
Nicholas Nassim Taleb describes these as "extremistan", in the chapter "Travels in Mediocristan" of Black Swans. A good discussion here:
https://people.wou.edu/~shawd/mediocristan--extremistan.html https://people.wou.edu/~shawd/mediocristan--extremistan.html
At Onlyfans, the mean monthly income is only $145, the median $135, and the modal income (that achieved by the largest number of creators) is ... $0/mo.
A popular tech discussion site sees a spread from the 1st to 100th ranked contributor from 345,912 to 34,802 karma. Even 10th stands at 1/3 the karma of #1, with 101,766 points.
In looking at Google+ usage data (something of a past hobby of mine), I found that the overwhelming majority of profiles had absolutely no visible activity. Fewer than 9% had ever posted to the service, and a small fraction of 1% --- still millions of profiles, but a vanishingly small fraction of the site --- were visibly posting over the previous ~2-week period. Other indicators, including followers counts, suggested very little activity amongst most profiles. This was treated by many as an indictment of G+ at the time. In truth it's not much different for any site.
I later gained access to a comprehensive listing of all 8 million Google+ Community summary data. This was retrieved via page-scraping, and gave the URL, name, subscriber count, and recent activity (up to ten most recent posts) on each Community. The resulting data were interesting.
In count vs. size, there's a near log-log relationship, though there are actually far fewer large Communities than one would expect --- there's a significant disadvantage to very large discussions. The largest should be in the range of 1--10 billion, in actuality the maximum was about 10 million members. See:
https://joindiaspora.com/posts/ab6a5470f57001368d400218b79d8c8d https://joindiaspora.com/posts/ab6a5470f57001368d400218b79d8...
In looking at metrics of user engagement (comments, +1's, shares), subscribers turned out to be far less relevant than posting activity, which is what this set of scatterplots shows:
https://joindiaspora.com/posts/23acfd90f05d01368b220218b79d8c8d https://joindiaspora.com/posts/23acfd90f05d01368b220218b79d8...
"Winner take all", or "winner take most", or even "winner take much" systems are going to be inherently highly unequal in reward distribution. Given the many chance, positive-feedback, and manipulative factors involved in such distributions, arguments over their efficiency tend to ring hollow (another aspect Taleb discusses at length).
And whilst it's likely that much of the little-viewed, little-rewarded content has few merits, there's also much gold buried low, but the discovery costs involved in finding it simply exceed the reward. Restoring any measure of justice to such systems is feindishly difficult.
Randomness (sortition and fuzzing of points), "taxes" on high awards, rotation, rewarding of new discoveries, and other elements might help. As would notions of a substantive baseline contribution or award.