Reproduce “Evolving the Hearthstone Meta” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Fernando de Mesentier Silva, Rodrigo Canaan, Scott Lee, Matthew C. Fontaine, Julian Togelius, Amy K. Hoover
Researchers from New York University, New Jersey Institute of Technology, and independent collaborators developed a computational approach to balance Hearthstone's metagame using evolutionary algorithms. Their method achieved near-perfect competitive balance while significantly reducing the number of necessary card changes, offering practical tools for AI-assisted game design.
Key insights
- Computational methods can effectively discover near-perfect game balance in complex design spaces, proving difficult-to-find states for human designers.
- Multi-objective optimization can balance the trade-off between achieving game equilibrium and minimizing the disruption caused by card attribute changes.
- Metrics like Win Rate when Drawn provide practical, data-driven heuristics for identifying high-impact cards that disproportionately affect game balance when modified.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.