Game piracy, the illegal copying and distribution of digital games, is an exceedingly hard-to-evaluate phenomenon across multiple channels. Its magnitude is difficult to estimate, not least because of the lack of clarity on what constitutes illegal copying and copyright infringement internationally. (Different countries, different rules.)
Game piracy is the cause
of heated debate, with
pirates on one side and game developers, game publishers, organizations like the Entertainment Software Association (ESA), and legislators/policymakers on the other.
Despite the massive interest and controversy surrounding software piracy in general and game piracy in particular, there is only some information available on the subject across game titles. Furthermore, that information often comes from industry organizations or operators of peer-to-peer (P2P) networks, which are sometimes accused of bias, as Ben Goldacre of the Guardian noted in an article back in 2009. As far as he is concerned, “everything from this industry is false, until proven otherwise.”
In essence, the piracy debate has suffered from a lack of objective numbers obtained via openly transparent methodologies that can be verified by third parties.
Until now, that is. For months, a team of researchers from Copenhagen Business School, Aalborg University (with which I am affiliated), and the University of Waterloo have been tracking 100-plus games on BitTorrent, arguably the biggest P2P distribution channel for digital content on the Internet. We have provided some numbers that, for once, are not developed at the request of commercial interests. There is no incentive to deflate or inflate these numbers.
From roughly November 2010 to March 2011, 12.7 million unique peers (or users of the P2P network) accessed 127 pirated games on BitTorrent. That's about 100,000 per game, and we registered participants from over 100 countries/states.
This is a lot of unique peers, and the size of the study supersedes previous work (e.g., a report by the ESA from 2009). The numbers show that file sharing via BitTorrent is exceptionally well distributed geographically and across game genres, hardware platforms, and, notably, the ratings of the Entertainment Software Rating Board (ESRB), which provides recommended maturity levels for digital games.
Also, the notion that the pirated games are all hard-core “shooters” (which focus primarily on action and twitch skills) appears to be erroneous; 38 percent of the 127 games found on BitTorrent out of a sample of 173 had an ESRB rating of "E" or "E10+" (i.e., suitable for children and families).
There are a few interesting wrinkles. BitTorrent activity appears to correlate with review scores. Critically acclaimed games get accessed more frequently on BitTorrent than unpopular ones. The interest in digital games is also extremely skewed. Over 40 percent of the 12.7 million peers in the dataset downloaded the 10 most pirated games. Otherwise, peer interest on BitTorrent is fairly well distributed across genres, with puzzle games, role-playing games, and, less surprisingly, shooters being the most popular.
Perhaps the most important thing to come out of this research is that, for the first time (to our knowledge), we actually have objective numbers on BitTorrent activity for illegally copied games -- unless someone finds critical errors in our method, of course! This puts to rest the main question in the debate about the magnitude of BitTorrent game piracy, although it says nothing about piracy through other channels, notably physical copying, a problem of unknown size. Another apparently increasing source of pirated material is file-hosting services.
Additionally, we still do not know how piracy numbers translate into lost sales. It's definitely not on a 1:1 basis, as some industry voices claim. But there is likely some kind of effect, as indicated by the sales drop in the music industry since widespread online file sharing began.
Interestingly, the numbers show that investigations of file sharing via P2P networks should not be conducted as snapshots. The activity levels of different torrents (i.e., files being shared) vary a lot over time. So investigators should take the long view and follow specific files over their entire lifetimes, or at least the period of peak activity, which is usually the first few days of the file´s lifetime. That makes these kinds of investigations cumbersome, but the added precision is worth the effort.
— Anders Drachen is a veteran data scientist, games analyst, and game user research specialist from Game Analytics.