Data-Driven Game Development: Ethical Considerations

Using Data to "Git Gud": A Push for a Player-Centric approach tothe Use of Data in Esports

“…The second key point, however, is that players' desires and needs are nuanced and dependent on individual differences. This comes as no surprise, as individual differences always impact play [31]. However, these nuances need to be better understood if we intend to design systems to help players gain expertise, and in order to better understand these nuances, we need to better understand how players are using their data.…”

“…Other work has taken a mixed-methods approach to analysis and modeling, combining data visualization with human insight to create human-in-the-loop systems that can help people understand gameplay [1,18,25]. These mixed-methods techniques are able to overcome many of the shortcomings of pure quantitative techniques, specifically, they better account for context and allow domain experts to glean insight from game data in an interactive and transparent manner [31]. Arguably, this would make such systems good candidates to aid players in their study of gameplay data in order to obtain mastery.…”

“…Data is becoming increasingly relevant in the domain of esports. On the academic side, the past decade has seen rapid advancements in player modeling techniques, visualization, and data-driven systems, spurred on by the increased availability of player data [13,16,24,31]. In the industry, data mining and statistics are becoming the backbone of design and business decisions intended to keep games fun and relevant.…”

Player Modeling using Behavioral Signals in Competitive Online Games

“…The majority of such algorithms use neural networks and latent factor models to predict rank and create match-ups. While these models often achieve better prediction results and higher quality matchmakings compared to rating systems, they lack explainability due to the complex nature of their calculations; an important feature for players who are interested in understanding the logic behind their matchups [4,5].…”

“…These systems often require high computation power due to their memory-intensive and complex computations [3]. In addition, compared to traditional rating systems, these systems often lack explainability and are difficult to understand for regular players [4,5].…”