Measuring, Characterizing, and Analyzing the Free Web Games Ecosystem

Abstract

Web games, that are directly playable within web browsers, have recently garnered substantial popularity, particularly among younger demographics. The absence of a paywall for these games has raised concerns regarding the potential privacy-compromising monetization strategies. Comprehensive investigations have been carried out into domains, such as paid online games, multiplayer online video games, mobile gaming, and their associated privacy and security concerns, but a significant gap persists in the characterization and examination of freely accessible web games. To address these voids, our research conducts an exhaustive analysis of the web games ecosystem. We rigorously scrutinize 22 distinct web game websites with the goal of understanding player experience and addressing privacy concerns. Our methodology involves simulating player interactions across approximately 100 000 individual web games to extract insights into player behavior and privacy risks. The outcomes of our work demonstrate substantial insights into the popularity, ownership, geolocation, and game environment, including metadata diversity. It also reveals the privacy risks faced by users on these websites, encompassing various aspects, such as the presence of questionable third-party advertisements designed for revenue generation, sporadic instances of objectionable content, and the persistence of tracking mechanisms, such as persistent cookies. Most of all, there is also a discouraging absence of transparent privacy policy statements, a website's privacy policy statement is a legal document to protect users' rights. Also, this deficiency of clear policy information hinders users' capacity to make informed decisions about opting out and understanding the potential consequences. In short, the insights from our study highlight substantial concerns regarding prevalent privacy practices within the domain of free web game websites. Moreover, we contribute our dataset and code, which contain metadata collected from approximately 100 K web games originating from the 22 websites examined in our investigation, thereby we provide a valuable resource for further research and analysis of this less investigated genre of websites.