Workshop on Automated Game Design

November 9–10, 2026 · A workshop of AIIDE 2026 Federal University of Minas Gerais · Belo Horizonte, Brazil

The goal of the Workshop on Automated Game Design (WAGD) is to provide a home for advancing the conversation and state-of-the-art in automated game design (AGD).

Why AGD Now?

OVERVIEW IMAGE

Game design is increasingly becoming malleable — something that players engage with themselves, something that is procedurally generated and varied, and something that is fluid and always changing. The popularity of game-making tools like Fortnite Creative and Mario Maker, as well as the existence of popular ROM-hacking and randomizer communities, has encouraged an influx of people to play with game-making, which makes the study and understanding of game design and its underlying processes more important than ever. Daily challenges and roguelike variation now often reconfigure fundamental aspects of a game’s design as it is played, making it important to build better tools to understand and predict dynamic systems with an incomprehensible number of potential outcomes. The ubiquity of live access, games-as-a-service, and early access models means that even without runtime modifications, game designs are constantly in flux — creating an opportunity for researchers to study how intelligent systems and tools can understand and support designers working in these rapidly-changing game design spaces.

Beyond this, AGD is a forward-looking, exciting blue-sky research space that aims to open up entirely new ways to play and create games. AGD carves a unique space for designers and researchers to explore how we build systems that produce surprising and diverse game artifacts. AGD lets us explore the tooling at design- and run-time that pushes the envelope of what styles of game design and player engagement are possible. AIIDE is an ideal venue for such a workshop: it has historically been a home for AGD research, forming a conversation between maximalist approaches to automated game design and more specific game AI, mixed-initiative tooling, and procedural content generation research. Hosting this workshop can help turn up the volume of this conversation.

Want an easy introduction into the field? Let Mike Cook give you a quick overview:

Get Involved

We are now accepting submissions for publication and presentation! Interested in joining, but not interested in submitting a paper? You are still welcome! Between paper presentations and collaborative design sessions, there is still a lot to offer. In-person attendance is required for workshop participation.

Please send any questions to samuel.m.shields@gmail.com.

Schedule

Monday, November 9th & Tuesday, November 10th, 2026

We plan on having three major sections to the workshop over two days:

  • Keynote Address An invited talk from a leader in the field — see Keynote Speakers.
  • Paper Presentations Presentations of accepted peer-reviewed work.
  • Collaborative Demo/Jam Session Hands-on, collaborative design and demo time.

Stay tuned for specifics closer to the workshop!

Keynote Speakers

Invited talks at WAGD 2026

Michael Cook

Michael Cook

King’s College London, Department of Informatics

Talk: To be announced

Mike Cook is an AI researcher and game designer, currently working as a Senior Lecturer at King's College London. He is the designer of the AGD systems ANGELINA and Puck, and the founder of PROCJAM, the Procedural Generation Jam. He was very briefly a world record speedrunner for a game about flipping coins.

Tanya X. Short

Tanya X. Short

Kitfox Games

Talk: To be announced

Tanya X. Short is the Captain of Kitfox Games and game designer of Streets of Fortuna, Boyfriend Dungeon and Moon Hunters, as well as the publishing head, overseeing publishing games like Dwarf Fortress and Caves of Qud. Her first job in the industry was as a production assistant in the Halo 2 marketing campaign in 2003, “ilovebees”. She’s also the co-founder and co-director of Pixelles, a feminist non-profit aiming to change games culture, and co-founded the Polaris Game Design Retreat. She also edited two textbooks with CRC Press, Procedural Generation in Game Design and Procedural Storytelling in Game Design.

More to Come

Additional keynote speakers will be announced soon — check back!

Call for Papers

Submission deadline: September 18th, 2026 (AoE)

Submitting Your Work

We plan on having accepted peer-reviewed papers published in CEUR-WS. As such, submissions follow the length guidelines of CEUR-WS (5–9 pages for short papers, 10 or more for full, references and appendices included) and can be anything related to the field of AGD. This includes theory and vision papers, lofi/hifi prototype descriptions, and evaluations of AGD systems. We encourage maximalist approaches; AGD systems often sum many disparate pieces together. Don’t be afraid to be messy or over-scoped. Papers focused on large pretrained generative models are discouraged due to cost barriers, replicability concerns, and sufficient existing coverage at other venues. We will review artifact evaluations alongside traditional papers.

All papers must use one of the following templates:

All submissions will be peer reviewed. Works will be judged on their technical and research contributions, accessibility, replicability, and impact. All accepted works will be published in the CEUR-WS proceedings for WAGD 2026. Submissions should be made through EasyChair, under the Automated Game Design 2026 track.

Interested in coming to the workshop but don’t have a submission? Want feedback on a demo, or want to ideate on a new system? Register anyway! All are welcome to come and learn more about AGD.

Examples of Relevant Topics

Below is a (non-exhaustive) list of topics we are looking for:

  • Languages and formalisms for describing game design spaces. See VGDL (Ebner et al., 2013) or Cygnus (Summerville et al., 2017).
  • Automated systems that design physical or digital games. See Ludi (Browne, 2008) or ANGELINA (Cook, Colton and Gow, 2016).
  • Approaches for evaluating unseen games or spaces of games. Includes evaluating large design spaces for playability, balance, or emergence. See Expressive Range Analysis (Smith and Whitehead, 2010).
  • Mixed-initiative approaches to automated game design. See Tanagra (Smith, Whitehead, and Mateas, 2010) or Sentient Sketchbook (Liapis, Yannakakis, and Togelius, 2013).
  • Studies or surveys of automated game design or similar works in commercial games. See Mosa Lina (Steele, 2023) or Caves of Qud (Freehold Games, 2024) as example games to analyze.
  • Novel methods or combinations of methods for automated game design. See WaveFunctionCollapse (Karth and Smith, 2017).
  • Automated generation of game mechanics, rules, or progression systems. See Zook and Riedl’s work on RPG and Platformer games (Zook and Riedl, 2014).
  • Self-modifying game systems that change over time, either through engagement with players, a larger metagame, or story progression. See Galactic Arms Race (Hastings, Guha, and Stanley, 2009).
  • Systems that generate games scoped within a single platform or genre. See Karth et al.’s work on Game Boy RPGs (Karth et al., 2021), or BrawlerAGD (Shields et al., 2022).
  • Orchestration of multiple distinct creative facets or modalities. See Liapis et al.’s work defining orchestration as a central challenge for AGD research (Liapis et al., 2019).
  • Code-level AGD: synthesizing human-readable and human-editable code (within the context of an existing codebase) as an AGD approach. See Butler et al.’s work on program synthesis for boss behavior generation (Butler, Siu, and Zook, 2017) and Cook’s work on AGD-friendly software design patterns (Cook, 2020).
  • Techniques for inventing good game goals. See Davidson et al.’s work on searching for human-like goals in open-ended play environments (Davidson et al., 2025).

Key Dates

Submissions Open June 18th, 2026
Submission Deadline September 18th, 2026
Review Deadline September 25th, 2026
Submission Notifications October 2nd, 2026
Publication-Ready Submission Due October 16th, 2026
Workshop November 9th–10th, 2026

All deadlines are at 11:59 PM AoE (Anywhere on Earth).

Questions on submissions, want guidance, or not sure if you’re a fit? Send an email to samuel.m.shields@gmail.com.

References

  • Browne, C. B. (2008). Automatic generation and evaluation of recombination games (Doctoral dissertation, Queensland University of Technology). [LINK]
  • Butler, E., Siu, K., & Zook, A. (2017, August). Program synthesis as a generative method. In Proceedings of the 12th International Conference on the Foundations of Digital Games (pp. 1-10). [LINK]
  • Cook, M. (2020, August). Software engineering for automated game design. In 2020 IEEE Conference on Games (CoG) (pp. 487-494). IEEE. [LINK]
  • Cook, Michael, Simon Colton, and Jeremy Gow. "The angelina videogame design system—part i." IEEE Transactions on Computational Intelligence and AI in Games 9.2 (2016): 192-203. [LINK]
  • Davidson, G., Todd, G., Togelius, J., Gureckis, T. M., & Lake, B. M. (2025). Goals as reward-producing programs. Nature Machine Intelligence, 7(2), 205-220. [LINK]
  • Ebner, M., Levine, J., Lucas, S. M., Schaul, T., Thompson, T., & Togelius, J. (2013). Towards a video game description language. [LINK]
  • Freehold Games. (2024). Caves of Qud (Version 1.0) [Video game]. https://www.cavesofqud.com/ [LINK]
  • Hastings, E. J., Guha, R. K., & Stanley, K. O. (2009). Automatic content generation in the galactic arms race video game. IEEE Transactions on Computational Intelligence and AI in Games, 1(4), 245-263. [LINK]
  • Karth, I., Duplantis, T., Kreminski, M., Kashyap, S., Kukutla, V., Lo, A., ... & Smith, A. M. (2021). Generating playable rpg roms for the game boy. [LINK]
  • Karth, I., & Smith, A. M. (2017, August). WaveFunctionCollapse is constraint solving in the wild. In Proceedings of the 12th International Conference on the Foundations of Digital Games (pp. 1-10). [LINK]
  • Liapis, A., Yannakakis, G. N., Nelson, M. J., Preuss, M., & Bidarra, R. (2018). Orchestrating game generation. IEEE Transactions on Games, 11(1), 48-68. [LINK]
  • Liapis, A., Yannakakis, G. N., & Togelius, J. (2013). Sentient sketchbook: computer-assisted game level authoring. [LINK]
  • Smith, G., & Whitehead, J. (2010, June). Analyzing the expressive range of a level generator. In Proceedings of the 2010 workshop on procedural content generation in games (pp. 1-7). [LINK]
  • Smith, G., Whitehead, J., & Mateas, M. (2010, June). Tanagra: A mixed-initiative level design tool. In Proceedings of the Fifth International Conference on the Foundations of Digital Games (pp. 209-216). [LINK]
  • Steele, L. (2023). Mosa Lina [Video game]. Lawrence Steele. https://store.steampowered.com/app/2116860/Mosa_Lina/ [LINK]
  • Summerville, A., Martens, C., Harmon, S., Mateas, M., Osborn, J., Wardrip-Fruin, N. and Jhala, A., 2017. From mechanics to meaning. IEEE Transactions on Games, 11(1), pp.69-78. [LINK]
  • Zook, A., & Riedl, M. (2014, June). Automatic game design via mechanic generation. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 28, No. 1). [LINK]

Committee

The WAGD 2026 organizing committee

Samuel Shields (Primary)

Samuel Shields

IDEO; Shields Games and Research LLC

samuel.m.shields@gmail.com

Sam Shields is a Game Design Research Lead at IDEO and a recent PhD graduate from UCSC’s Computational Media department. Sam’s passions lie in understanding, designing, and implementing interlocking game systems that are expressive and surprising. He’s deeply curious about investigating “metadesign” (how we design design systems) for games. Some of his relevant prior work involves generating fighting games while tuning their properties to be competitively balanced (BrawlerAGD) and having an AI director try and tune a turn-based RPG for different audiences (FighterDDA). He’s an Esper mage and is currently going through a ska phase.

Max Kreminski

Max Kreminski

Cornell Tech

mkremins@cornell.edu

Max Kreminski is a creativity support tools researcher and assistant professor of Design Tech at Cornell Tech, making interactive software systems that enable new forms of playful exploration in creative domains. Their AGD work has focused on human-AI interaction design (Germinate); incorporating narrative into AGD (StoryAssembler, GBS); and authoring support for open-ended interactive stories (Dramamancer, Elsewise, Bonsai). Although conceptually skeptical of games and automation, Max is nevertheless a big fan of design.

Lana Rossato

Lana Rossato

Federal University of Rio Grande do Sul

lbrossato@inf.ufrgs.br

Lana Rossato is a PhD student in Computer Science at the Federal University of Rio Grande do Sul (UFRGS), Brazil. Her research focuses on Artificial Intelligence for Games, particularly Automatic Game Design, Procedural Content Generation, and Computational Creativity. She is interested in developing systems capable of generating, evaluating, and adapting games autonomously, with an emphasis on symbolic AI techniques.

Michael Cook

Michael Cook

King’s College London, Department of Informatics

mike.cook@kcl.ac.uk / mike@possibilityspace.org

Mike Cook is an AI researcher and game designer, currently working as a Senior Lecturer at King's College London. He is the designer of the AGD systems ANGELINA and Puck, and the founder of PROCJAM, the Procedural Generation Jam. He was very briefly a world record speedrunner for a game about flipping coins.

Rachel Heleva

Rachel Heleva

University of California, Davis; The Museum of Art and Digital Entertainment

rlheleva@ucdavis.org

Rachel Heleva is a PhD student in Performance Studies at UC Davis and a member of the Social Intelligence and Narrative in Games Lab. Her research focuses on computational poetics and AI systems for storytelling and interpretation, especially playable systems that use rules, constraints, and variation to make interpretive reasoning part of play. She is also the Executive Director of the Museum of Art and Digital Entertainment in Oakland where she leads public work in access to game design education and AI literacy.