This article was recently published on Techdirt.
Recently, I learned of a trio of researchers who took it upon themselves to objectively study game piracy over BitTorrent and determine some quantifiable data on game piracy. (PDF) The researchers, Anders Drachen, Kevin Bauer and Robert W. D. Veitch, observed a huge lack of objectivity in research from both sides of the piracy debate. They saw a lot of negativity on the side of the industry and positivity on the flip side.
According to Forskning.no, this lack of objectivity has led the games industry to inflate piracy numbers:(from the Google Translation)
These include an influential 2009 report, made by the U.S. trade association for game manufacturers, ESA. It is this which, according Drachen indicate that piracy is two to three times greater - depending on how the ESA press release is interpreted.
The Danish researchers estimate that is 290 million games a year. In comparison indicates ESA figures that 600 million games pirate copies in a year, and because of uncertainties in the measurement method, which is highly secret, the American figures to be considerably larger, according Drachen.
To counteract this subjectivity, the researchers took a sampling of all commercial games from November 2010 to January 2011 and watched the activity of those games on BitTorrent and combined that information with other information on the games, such as genre and review scores. In all, they listed 173 games of interest, of which 127 were found on BitTorrent. Using this data, they came to the following conclusions:
- The majority of games they tracked had fewer than 50,000 unique peers observed on BitTorrent. At the same time, the 10 most popular games had a combined 5.37 million unique peers.
- 'Action' games made up the lion's share of BitTorrent activity, comprising 45.61% of unique peers in all. 'Role-Playing Games', of which only 10 were observed, accounted for 15.58% of unique peers. However, when they looked at individual games, 'Action' games were less popular than 'Racing', 'Role-Playing', and 'Simulation' games, comprising 0.95%, 1.6%, 1.43%, and 1.08% respectively.
- Aggregated review scores correlated positively with the number of unique peers. Meaning, the higher the MetaCritic score, the more unique peers will be found on BitTorrent.
- In all, observed about 12.6 million unique peers accessing the 127 games on BitTorrent.
The researchers plan to continue their research by observing other features of games such as the ESRB ratings, marketing strategies and international release dates. They also plan to observe time-frequency distribution rates as well. They had already tested this by looking at 20 randomly selected games and found that games have a high peak around initial launch and then drift down to fairly low levels of unique peers.
This is a great first step in bringing objectivity into piracy debates. Hopefully, what will come of this is more interest in objective studies on piracy not just for games but also for other entertainment such as movies and music. The more information that is available will help content creators and distributors to make educated decisions on how to minimize the risk of piracy and better connect with their fans. Sadly, this study has made no headway on any of the major games industry news sites. One would hope that objectivity and quantifiable information would make for some interesting news.