Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Level the Level: Balancing Game Levels for Asymmetric Player Archetypes With Reinforcement Learning

Florian Rupp, Kai Eckert

TL;DR

Balancing asymmetric multiplayer games is challenging and often relies on manual tuning. The authors extend PCGRL to balance tile-based levels for asymmetric archetypes, introducing a halved swap-based action space and four new agent archetypes, and evaluating against random and hill-climbing baselines. Results show PCGRL balances a larger fraction of levels and trains faster with the reduced action space, though greater initial asymmetry increases training requirements and can reduce final balance accuracy. A notable limitation is that the method can yield unwinnable draws, pointing to avenues for refining the reward structure and balance criteria.

Abstract

Balancing games, especially those with asymmetric multiplayer content, requires significant manual effort and extensive human playtesting during development. For this reason, this work focuses on generating balanced levels tailored to asymmetric player archetypes, where the disparity in abilities is balanced entirely through the level design. For instance, while one archetype may have an advantage over another, both should have an equal chance of winning. We therefore conceptualize game balancing as a procedural content generation problem and build on and extend a recently introduced method that uses reinforcement learning to balance tile-based game levels. We evaluate the method on four different player archetypes and demonstrate its ability to balance a larger proportion of levels compared to two baseline approaches. Furthermore, our results indicate that as the disparity between player archetypes increases, the required number of training steps grows, while the model's accuracy in achieving balance decreases.

Level the Level: Balancing Game Levels for Asymmetric Player Archetypes With Reinforcement Learning

TL;DR

Balancing asymmetric multiplayer games is challenging and often relies on manual tuning. The authors extend PCGRL to balance tile-based levels for asymmetric archetypes, introducing a halved swap-based action space and four new agent archetypes, and evaluating against random and hill-climbing baselines. Results show PCGRL balances a larger fraction of levels and trains faster with the reduced action space, though greater initial asymmetry increases training requirements and can reduce final balance accuracy. A notable limitation is that the method can yield unwinnable draws, pointing to avenues for refining the reward structure and balance criteria.

Abstract

Balancing games, especially those with asymmetric multiplayer content, requires significant manual effort and extensive human playtesting during development. For this reason, this work focuses on generating balanced levels tailored to asymmetric player archetypes, where the disparity in abilities is balanced entirely through the level design. For instance, while one archetype may have an advantage over another, both should have an equal chance of winning. We therefore conceptualize game balancing as a procedural content generation problem and build on and extend a recently introduced method that uses reinforcement learning to balance tile-based game levels. We evaluate the method on four different player archetypes and demonstrate its ability to balance a larger proportion of levels compared to two baseline approaches. Furthermore, our results indicate that as the disparity between player archetypes increases, the required number of training steps grows, while the model's accuracy in achieving balance decreases.

Paper Structure

This paper contains 12 sections, 2 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Background
  4. Game Environment
  5. PCGRL for Game Level Balancing
  6. Method
  7. Improving the Action Space
  8. Asymmetric Balancing
  9. Results, Discussion and Limitations
  10. Performance and Comparison to Baselines
  11. Generated Levels
  12. Conclusion

Figures (2)

  • Figure 1: Training steps required for model convergence compared to the initial imbalance due to agent asymmetries per setup. A value of 0.5 indicates equal wins, while 1.0 means one agent always wins.
  • Figure 2: Generated samples from different models and archetype setups in comparison with the initial unbalanced version. A player of archetype A (red) is paired with different archetypes (yellow).