Ludemic Abstraction for General Game Systems

24 concepts15 papers

Traditional approaches to General Game Playing (GGP) often rely on low-level logical descriptions to define game rules. While precise, these methods—such as the Logic-based Game Description Language (GDL)—are often verbose, difficult for humans to read, and computationally expensive to simulate. A growing body of research proposes a shift toward ludemic representation, which decomposes games into high-level, atomic conceptual units called "ludemes."

The Atomic Units of Play

At the core of this research is the concept of the ludeme. As described by Browne et al. in A Practical Introduction to the Ludii General Game System, ludemes are fundamental building blocks that encapsulate specific game elements, such as equipment (boards, pieces), rules (movement, capture), and victory conditions.

Instead of defining a game through a series of explicit, step-by-step logical instructions, a ludemic system treats a game as a composition of these high-level primitives. For example, rather than writing complex logic to describe how a piece moves across a grid, a designer can simply invoke a "sliding" ludeme. This approach, central to the Ludii system introduced by Sironi et al. in LUDII - Le Système Ludémique de General Game Playing, allows for a library of modular components that can be reused and recombined across thousands of different games.

Why Ludemic Abstraction Works

The power of ludemic representation lies in its ability to map game descriptions directly to a structured grammar. In An Overview of the Ludii General Game System, Piette et al. explain that by mapping ludemes to a Java class grammar, the system achieves significant gains in both human readability and computational performance.

  1. Conciseness and Clarity: By abstracting away low-level implementation details, game descriptions become much shorter. Piette et al. demonstrated in An Empirical Evaluation of Two General Game Systems: Ludii and RBG that Ludii descriptions are approximately three times more concise than those of competing high-performance frameworks like Regular Boardgames (RBG).
  2. Computational Efficiency: Because the engine understands the high-level intent of a rule (e.g., "capture by jumping"), it can use optimized internal code to execute that rule rather than re-interpreting raw logic at every step. This leads to faster state evaluation and simulation, which is critical for search-based AI agents (Piette et al., Ludii as a Competition Platform).
  3. Emergent Complexity: The modularity of this approach means that complex behaviors can emerge from the interaction of simple primitives. Siu et al. illustrate this in A Programming Model for Boss Encounters in 2D Action Games, showing how sophisticated boss behaviors in action games can be constructed from basic behavioral building blocks rather than monolithic scripts.

Applications Beyond Play

The shift toward a standardized, modular vocabulary for games enables research that goes beyond simply playing them.

  • Design Assistance: Machado et al. developed Pitako, a recommender system that uses these structured representations to suggest mechanics to novice designers, significantly reducing their perceived workload (Evaluation of a Recommender System for Assisting Novice Game Designers).
  • Digital Archaeology: By breaking games down into objective ludemes, researchers can compare games across different cultures and historical eras. Browne et al. argue in their Data Note that this provides a rigorous, standardized framework for understanding game evolution.
  • Explainable AI and Transfer Learning: By extracting hundreds of distinct features—such as "stochasticity" or "capture"—from game descriptions, the authors of General Board Game Concepts provide a high-level vocabulary that helps AI agents understand the commonalities between disparate games, potentially facilitating the transfer of strategies from one game to another.

Formalizing Composition

The mathematical foundation of this modularity is also being explored through category theory. Gavranović and Kupke propose a framework in Compositional Game Theory, Compositionally that uses structures like arrows and bimodules to formally define how simpler game components can be composed into complex systems. Similarly, new domain-specific languages like Ludax are being developed to compile these high-level descriptions into GPU-accelerated code, bridging the gap between human-readable abstraction and the high-performance needs of modern reinforcement learning (Todd et al., Ludax: A GPU-Accelerated Domain Specific Language for Board Games).

Key Foundational Work

LUDII - Le Système Ludémique de General Game Playing (2019) — contributed 3 concept elements to this direction.
Ludii - The Ludemic General Game System (2019) — contributed 3 concept elements to this direction.
A Practical Introduction to the Ludii General Game System (2019) — contributed 2 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_ludemic_game_representation_7ef9480e,
  title = {ludemic game representation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/ludemic-game-representation/rdr_7ef9480ec84246a39ad8e022c478d792},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_ludemic_game_representation_7ef9480e_v1,
  title = {ludemic game representation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/ludemic-game-representation/rdr_7ef9480ec84246a39ad8e022c478d792/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}

See something wrong? Report a mistake