Study samples files available on torrent sites, find 1% of files to be non-infringing

A census carried out by Princeton senior Sauhard Sahi, revealed that it is very likely that only one in a hundred available files on BitTorrent platforms does not contravene copyright.

The goal of the study was to provide some figures on what kind of files are available on torrent platforms. To do so, the researcher took a random, uniform sample of 1021 files using the trackerless system Mainline DHT. He then categorised the findings according to file type, language and copyright status.

Categorised by filetype, this was the result:

46% movies and shows (non-pornographic)

14% games and software

14% pornography

10% music

1% books and guides

1% images

14% could not classify

The movie category showed a bias towards recent movies, more so than other categories. Around two thirds of the content proved to be in English, slightly more for games and music. Other languages included Chinese, Japanese, Russian, French and German.

Subsequently, for each category, the files were tested for their copyright status. The researcher considered files to be non-infringing if it was 1) in the public domain, 2) user generated content or 3) freely distributed through legitimate channels. The number of works from the sample that fitted in either of these categories was quite low. None of the files from the movie and TV show category nor the music category was likely to be non-infringing. Most legitimate content was found in the games and software category, with 7 out of 148 files likely not contravening copyright. In the books section, two of the fifteen files were thought not to infringe.

In all, ten of the 1021 files will likely not infringe copyright. That is just short of one percent.

10 February 2010