I think it remains appropriate to be skeptical about CacheLogic’s claims that BitTorrent or other P2P applications are taking up 60-80 percent of all Internet bandwidth. Why? Because their “studies” making such claims seem to be nothing more than press releases about their proprietary products that they hope to persuade ISPs and others to buy. So far as I can tell, there has been no CacheLogic “study” in the sense that, say, a peer-reviewed journal would use the term. Nor does there seem to have been any independent verification of their claims about broadband usage. Nor is their proprietary technology itself available for inspection to determine whether their methodologies are sound.

Lots of folks here will remember that it was only a decade ago that Marty Rimm’s “study” of Internet pornography “proved” that the great majority of Internet usage, as measured by some metric or other, had to do with downloading porn. Rimm’s phony statistics (there does not seem to actually have been any “study” as such) were reproduced uncritically in the press and elsewhere online. One lesson to be drawn from that experience is that we all should be skeptical about CacheLogic’s claims at least until both (a) their measuring technology is made available for public inspection and use and (b) their statistical claims are presented in a way that makes them independently testable.

—–