Wareztracker - About

Wareztracker is a research project measuring video game piracy.


Wareztracker does not facilitate piracy. Only aggregate statistics are provided. The website does not publish copyrighted files, torrent files, magnet links, info hashes, URLs or names of pirate websites.

The source data for aggregate metrics is collected by scraping public websites, and using public API of torrent trackers. No limited access data is used. We have no relations with torrent websites or trackers.

The metrics are published without any warranty whatsoever. You are using them at your own risk.


First, the following data is collected and regularly updated:

Second, for each torrent we try to understand which game or games it contains. We match torrent websites topic titles to Steam games, using a fuzzy full-text search algorithm. Some effort is spent on automatically disambiguating different versions of games, games with similar names, or names which are substrings of each other. Contents of torrents, i.e. file names, etc, are not used for matching at the moment. For better matching some additional tricks are employed, such as a blacklist for games with too general names (which tend to match to big number of unrelated torrents), a list of alternative names for games, and so on.

Third, once we have a list of torrents for each game, we aggregate torrent statistics per game. Note that each torrent may have duplicate statistics per day from different trackers. We aggregate daily statistics from multiple trackers using maximum per torrent, and then sum up torrent statistics per game. We assume that torrents and magnet links often list many trackers, and torrent clients report their download to all of them, so we use maximum instead of sum. On the other hand, people will unlikely download multiple different torrents for the same game, so we sum up the statistics per game.


The idea of the project is to provide some hard numbers about piracy, however it is important to understand what those numbers mean and how they are collected and calculated. In general, the numbers must be regarded as a very rough (under)estimation, should be used in comparison and not as precise absolute numbers.


The data is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.