Update: We are merging forces with the Player Modelling Workshop to form the very first Player Analytics workshop – for further detail please see here.
The analysis of player behavior is an old stable in the design, development and research of games and the people who play them, but in recent years the technology has emerged that allows the tracking of detailed behaviors in games from entire populations of players. This has caused a surge in the data available to game developers and game researchers alike. On the academic side of the equation, incredibly detailed records of player behavior has provided researchers with a new range of possibilities for building game AI, build behavioral models or explore relationships between design and behavior. However, the data deluge has also brought new challenges along with the opportunities, notably the fundamental problem of how to derive actionable insights from large amounts of often high-dimensional and time-dependent data which can be captured from a variety of sources. In essence, gameplay telemetry express all the four classical features of big data, i.e. volume, variety, volatility and veracity.
Game Analytics has emerged as the umbrella term for quantitative analysis of behavior – whether from players or production – across academia and industry. The player focus is by far the most common, and aims at exploring player behavior, improve AI, improve game design, ensure optimal user experiences, identifying valuable players or those at risk, assist and inform learning and training games, personalize and adapt gameplay, or assist matchmaking – just to mention a couple of topics. Game Analytics is a domain in rapid evolution, covering a broad spectrum of games, platforms and techniques. In recent years, machine learning has become an important part of both research and practice in Game Analytics. The context of analytics has also emerged as an important topic, e.g. data visualization.
The workshop aims to act as a unifying point for those interested in Game Analytics at AIIDE, and to help connect industry and academia. Given the recent introduction of Game Analytics as a domain of research crossing into many of the traditional areas of games research, there is as yet a limited number of publications in the domain, and virtually no standards or accepted principles in the field, across research and industry. To address this problem, the workshop has the following main goals:
- Discuss and assess the current state-of- the-art in Game Analytics and the current major challenges in the domain across industry and academia, including academia-industry collaboration, documenting ROI, integrating applied and basic research and developing standards.
- Develop strategies for progressing the state-of- the-art and approaching major challenges in the domain.
- Disseminate knowledge about Game Analytics principles and methods to academia and industry.
The workshop is intended as a combination of presentations and discussions, combined with breakout groups. Given the half-day format, the workshop has limited time and the program is therefore relatively condensed:
- Brief paper presentations and discussion: During the initial orientation, workshop organizers and participants will present an overview of their own work, issues facing them, and current best practices in their field.
- Outlining and discussing the key challenges in Game Analytics and the state-of- the-art
- Breakout groups working on topics identified during the initial discussions: The session will provide an opportunity for participants to gather together into smaller teams and have round table discussion.
Joint discussion on topics: The workshop will end with a collection of guidelines and best practices to integrate Games User Research (“GUR”) in the development cycle for low budgeted projects.
Debriefing and thank you.
- Anders Drachen, Aalborg University
- Rafet Sifa, Fraunhofer IAIS
- Nicholas Ross, University of San Francisco
- Julian Runge, Wooga and Humboldt University Berlin
Postscript: Unrelated to this workshop, a special issue is being formed on Visual Game Analytics – check this page for details: https://sivga.wordpress.com/