DaciaJC
Gold Member
From the Ars Technica article:
This is how I imagine his analysis worked: he looked at a statistical composite of new Steam reviews and concurrent player counts and found a consistent, significant decrease in those numbers for a given game once its crack was released. You would expect a game's player counts (or number of new reviews) to naturally decline with time after release as early buyers finish their experience and move on, especially for singleplayer games, but that's partially balanced out by people purchasing the game later after release. If, however, pirating the game becomes an option, at least some portion of those who would have bought a legit Steam copy opt instead to download a crack, and therefore you would see a more severe drop in player count coinciding with the crack release compared to the expected natural decline. By looking at a selection of games that varied in how quickly they were cracked after release, he was able to determine the magnitude of the effect the availability of a crack had on these revenue indicators with respect to time.
In "The Revenue Effects of Denuvo Digital Rights Management on PC Video Games," published in the peer-reviewed journal Entertainment Computing, UNC research associate William Volckmann examines 86 different Denuvo-protected games initially released on Steam between September 2014 and the end of 2022. That sample includes many games where Denuvo protection endured for at least 12 weeks (when new sales tend to drop off to "negligible" amounts for most games) and many others where earlier cracks allowed for widespread piracy at some point.
Unfortunately, the lack of good publicly available sales data for most games makes it difficult to measure these revenue effects directly. To estimate a Steam game's relative sales decline in each week after release, Volckmann uses a proxy that combines the number of new Steam user reviews and, for single-player narrative games, the game's average active player count. While Volckmann acknowledges that these imperfect estimates represent "the biggest limitation of this study," any estimated biases away from actual sales data seem likely to cancel out across the various games in the sample.
This is how I imagine his analysis worked: he looked at a statistical composite of new Steam reviews and concurrent player counts and found a consistent, significant decrease in those numbers for a given game once its crack was released. You would expect a game's player counts (or number of new reviews) to naturally decline with time after release as early buyers finish their experience and move on, especially for singleplayer games, but that's partially balanced out by people purchasing the game later after release. If, however, pirating the game becomes an option, at least some portion of those who would have bought a legit Steam copy opt instead to download a crack, and therefore you would see a more severe drop in player count coinciding with the crack release compared to the expected natural decline. By looking at a selection of games that varied in how quickly they were cracked after release, he was able to determine the magnitude of the effect the availability of a crack had on these revenue indicators with respect to time.