Please stop pretending you know how much bandwidth BitTorrent uses

I’ve seen this mentioned a lot when people talk about BitTorrent, but this bit from an announcement of BitTorrent 4.20 happened to push me over the apathy threshold:

It’s no secret that bandwidth concerns have been one of the more pressing issues surrounding the BitTorrent community. CacheLogic, which provides P2P caching solutions for ISP networks, has previously calculated that approximately 60% of a networks bandwidth is consumed by the BitTorrent protocol. This average varies according to the ISP, as some ISPs report less bandwidth consumption and other reporting more.

They’re completely wrong about what the CacheLogic study says. The most recent numbers I could find from CacheLogic say that “P2P still represented 60% of internet traffic at the end of 2004” and “By the end of 2004, BitTorrent was accounting for as much as 30% of all internet traffic.” Even if P2P has grown in the past 18 months to consume 99.99% of internet traffic, CacheLogic’s own studies show that eDonkey surpassed BitTorrent in P2P traffic in August 2005. If CacheLogic’s numbers are correct, there’s no way that BitTorrent has more than 50% of internet traffic.

But that’s the real issue here-are CacheLogic’s numbers correct? Look at what CacheLogic sells: P2P caching appliances. Their entire business is built around reducing the amount of bandwith P2P applications use. And they are also the sole source of numbers saying that P2P applications are using lots of bandwidth.

I’m not saying that their numbers are all wrong, I’m saying that I don’t know what the truth is. A press release from a company that has a direct and obvious profit motive from over-hyping shouldn’t be treated as a solid fact. Unfortunately a highly-suspect number is far more attractive to a writer than saying “I don’t know what the truth is.”

CacheLogic seem to have been pretty successful at getting their numbers into the collective consciousness. Traditional media like Wired Magazine, BBC and Reuters trumpet the numbers as if they were a fundamental rule of the internet (like Rule #34: There is porn of it. No exceptions.). Then, the numbers are repeated ad nauseam until sites like Slyck News can pepper a story with them without even needing to cite the source, since everyone knows it’s true.

Let’s stop pretending we know things that we don’t. There’s nothing wrong with saying “I don’t know,” there is something wrong with pretending you know what you really don’t. Let’s get our numbers from someone who isn’t trying to sell us a solution to the problem the numbers describe.

(For more skepticism about CacheLogic’s numbers, check out Peter Sevcik’s piece at Business Communications Review)