Knowledge of how easily statistics can be manipulated is no longer, in this information age, just the purview of researchers. Consumers, too, are probably aware that “four out of five dentists agree” is somehow orchestrated by an eager young PR practitioner.
But the folks with a truly deep sense of the unreliablity of all statistics, particularly in cyberspace, are the techies. Not only can data be tamed for any purpose, but there are also bots, crawlers, spamming mechanisms, etc. that can impact the most basic metrics such as hits and friends, by which web popularity is judged.
Website hits are thought to be governed by two factors acting separately, consecutively, and/or concurrently. First, the rich get richer (e.g., googlearchy, cumulative advantage). Second, popularity is incredibly volatile (e.g., a single post/song/video goes viral, a blog’s obscure niche suddenly becomes a current event, search engines are actually weirdly egalitarian).
Hits are simple and accessible measures of value, but they of course can be corrupted via concentrated campaigns. These range from the benign, such as gratuitously plugging your band, trolling for friends in social networking sites, or even rickrolling, to the well-funded, strategic, and aggressive, as in stealth marketing and astroturfing. Those techniques play the angles to generate artificial web popularity. But they’re generally given a pass because us Americans appreciate hustle, even if it’s corporate.
Inflationary hit count techniques have recently come to light that seem even more insidious, somehow beyond just hustling the system:
> Ticketmaster may have generated imaginary Facebook friends, through some clever use of ones and zeros.
> MySpace artists (and probably their labels), are using software (PaceSys Traffic Master and ToonBoom Pro) to jack up the number of plays so bands and songs appear to be more popular, too.
My guess is this is the tip of the iceberg. (Thanks, Eliot Van Buskirk at Wired ListeningPost, for the excellent reporting.)
Such technology-manipulating methods don’t have a Horatio Alger ring to them because they leverage wealth to purposefully skew the full information and transparency upon which free market decisions are made. I know information is never full or fully transparent, but it is an assumption that drives free market economies and underlies the decisions of actors in cyberspace. And for those who operate with that assumption, the revelation that it is false can be economically and (by extension) personally upsetting.
I contemplated whether or not I thought the pay-per-hit inflation might someday turn basic measures of web popularity into meaningless dross. But I don’t think so
1) because we’re embedded in a market/society/culture that equates success with dominance and obscures real inequality with the delusional rallying cry of meritocracy,
2) because this context I describe has developed out of a basic human penchant for hierarchy/ranking, and
3) because simple counts of visits and friends (very quaint sounding!), despite being shoddy data, are the only information the public has with which to do this business of deciding who wins.