1 of 14

Game Design

in Collaboration with the Algorithm

Major Research Project Proposal

MA in Digital Media

Douglas Gregory

219033117

2 of 14

Background

Douglas Gregory: 15 years as a game designer, 5 teaching game design community moderator on GameDev.StackExchange

Objective: Help game creators achieve their creative goals

3 of 14

Topic Area

Procedural Content Generation (PCG) for Games

Using algorithmic processes to produce game content such as environments, items, objectives, etc. Generated artifacts need to serve both functional (playable) and aesthetic (attractive, fun) goals.

Runtime Procedural Content

Algorithms run on the player’s device to create content on demand, expanding replayability, freedom, and surprise

Innes McKendrick - Continuous World Generation in ‘No Man’s Sky’ (GDC 32017)

Mixed Initiative Creation

Algorithms incorporated into developer tools to ease the burden of authoring game content and amplify creators’ power

Jaap van Mujiden - Run-Time Procedural Placement in ‘Horizon: Zero Dawn’ (GDC 2017)

4 of 14

Open Issues

Common procedural systems are difficult for game and level designers to create, tune, and collaborate with

Skill gaps - code vs design

Complex parameter spaces

Limited tool & UX support

There are hundreds of parameters buried in the depths of the algorithm…creating a seething, chaotic pool of coupled variables.

I no longer have any idea exactly how changes to them will manifest in the final product.

The system is too complex. Sometimes, mysteriously, levels will start looking very uninteresting, maybe ugly, or worse: not fun. The resulting debug sessions aren't really about debugging so much as black magic.

- Jordan Fisher

“How to make Insane,

Procedural Platformer Levels”

“

”

But for all the potential, there are still major gaps and challenges in applying these techniques to games.

Most documentation about procedural techniques or middleware to support it is directed at an audience of programmers - who don’t always have the grounding in game design to effectively teach the systems what traits make “good” content. And many designers who have that knowledge lack the programming or mathematical background to articulate that knowledge in code.

These systems often have high complexity, with many parameters to tune and chaotic interactions between them, making them unintuitive or unwieldy to control - as illustrated by this quotation from Jordan Fisher.

And unlike 3D modelling or level design, where tool packages to support creators are relatively mature, few pieces of software exist that help to bridge this gap. Often designers are working with programmer prototypes with little consideration for user experience.

5 of 14

Coursework

DIGM 5010 - Foundations of Digital Media

Path-First Platformer Level Generation

6 of 14

Coursework

EECS 5326 - Topics in Artificial Intelligence

Applying Reasoning Models to Evaluate Puzzle Designs

Hardest

Easiest

Bottleneck Difficulty

7 of 14

Common Threads

Using algorithms to explore the space of possible player actions

Attention to the game designer’s user experience

Play shapes design

How can I use this in my daily process?

8 of 14

Research Questions

How can a procedural system

be designed to work in tandem

with conventional game & level designer skills?

What conceptual models are intuitive for designers to work with?
How can designer intention be mapped to algorithmic rules?
What user interfaces and feedback support fluency?

9 of 14

Related Works

Danesh - Michael Cook

Unity plug-in that helps visualize and explore the expressive range of custom procedural generators

Tanagra - Smith & Whitehead

Mixed-initiative platformer level design tool based around a timeline of “beats” of player action

10 of 14

Expected Output

A software tool demonstrating improved designer user experience applicable to making games and informing the design of future tools

11 of 14

Ideas & Methodologies

User Study / Playtesting - examine how players behave and enjoy content, to inform models and evaluate output

Casual Creators (Compton and Mateas 2015) - design patterns for autotelic creativity tools

Explorable Explanations (Bret Victor, Nicky Case) - using interaction to help understand complex systems

12 of 14

Challenges

Select specific game design task or output to support

Proposed:

Continue case study in level design/generation tools for HexCells puzzles

Conducting user study to gather data and test assumptions about player behaviours
Interviewing or observing designers working on similar tasks to learn about their processes and mental models
Iterative prototyping of tools
Conducting user studies in the use of the tool and appreciation of its output

13 of 14

Selected References

Alvarez, Alberto, Jose Font, and Julian Togelius. “Designer Modeling through Design Style Clustering.” IEEE Transactions on Games, 2022, 1–1. https://doi.org/10.1109/TG.2022.3143800.

Beyer, Marlene, Aleksandr Agureikin, Alexander Anokhin, Christoph Laenger, Felix Nolte, Jonas Winterberg, Marcel Renka, et al. “An Integrated Process for Game Balancing,” 2016. https://doi.org/10.1109/CIG.2016.7860425.

Compton, Kate, and Michael Mateas. “Casual Creators.” Edited by Hannu Toivonen, Simon Colton, Michael Cook, and Dan Ventura. Proceedings of the 6th International Conference on Computational Creativity, ICCC 2015, 228–35.

Cook, Michael, Jeremy Gow, and Simon Colton. “Towards the Automatic Optimisation of Procedural Content Generators,” 2016. https://core.ac.uk/reader/83956140.

Davis, Nicholas Mark. “Human-Computer Co-Creativity: Blending Human and Computational Creativity.” In Ninth Artificial Intelligence and Interactive Digital Entertainment Conference, 2013. https://www.aaai.org/ocs/index.php/AIIDE/AIIDE13/paper/view/7453.

Fisher, Jordan. “How to Make Insane, Procedural Platformer Levels.” Game Developer, May 10, 2012. https://www.gamedeveloper.com/design/how-to-make-insane-procedural-platformer-levels.

Maher, Mary Lou. “Computational and Collective Creativity: Who’s Being Creative.” In In Proc. 3rd Int. Conf. on Computational Creativity, 2012.

Smith, Gillian, Jim Whitehead, and Michael Mateas. “Tanagra: Reactive Planning and Constraint Solving for Mixed-Initiative Level Design.” IEEE Transactions on Computational Intelligence and AI in Games 3, no. 3 (September 2011): 201–15. https://doi.org/10.1109/TCIAIG.2011.2159716.

Bret Victor. “Up and Down the Ladder of Abstraction.” Accessed May 2, 2022. http://worrydream.com/LadderOfAbstraction/.

14 of 14