# DungeonBench: A Benchmark for Rules-Rich Tactical Reasoning in Dungeons & Dragons Combat

arXiv: https://arxiv.org/abs/2607.29577 · HTML: https://mlsanity.com/paper/2607.29577/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 23 references that are in the map.
- First posted on arXiv 31 Jul 2026.

## Builds on (most cited first, 22 of 22)

- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [The StarCraft Multi-Agent Challenge](https://mlsanity.com/paper/1902.04043/) (2019)
- [Leveraging Procedural Generation to Benchmark Reinforcement Learning](https://mlsanity.com/paper/1912.01588/) (2019)
- [PettingZoo: Gym for Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2009.14471/) (2020)
- [MineRL: A Large-Scale Dataset of Minecraft Demonstrations](https://mlsanity.com/paper/1907.13440/) (2019)
- [Benchmarking the Spectrum of Agent Capabilities](https://mlsanity.com/paper/2109.06780/) (2021)
- [The NetHack Learning Environment](https://mlsanity.com/paper/2006.13760/) (2020)
- [The Hanabi Challenge: A New Frontier for AI Research](https://mlsanity.com/paper/1902.00506/) (2019)
- [OpenSpiel: A Framework for Reinforcement Learning in Games](https://mlsanity.com/paper/1908.09453/) (2019)
- [SmartPlay: A Benchmark for LLMs as Intelligent Agents](https://mlsanity.com/paper/2310.01557/) (2023)
- [SMACv2: An Improved Benchmark for Cooperative Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2212.07489/) (2022)
- [MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research](https://mlsanity.com/paper/2109.13202/) (2021)
- [Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learning](https://mlsanity.com/paper/2402.16801/) (2024)
- [An investigation of model-free planning](https://mlsanity.com/paper/1901.03559/) (2019)
- [Jumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAX](https://mlsanity.com/paper/2306.09884/) (2023)
- [PDDLGym: Gym Environments from PDDL Problems](https://mlsanity.com/paper/2002.06432/) (2020)
- [CALYPSO: LLMs as Dungeon Masters' Assistants](https://mlsanity.com/paper/2308.07540/) (2023)
- [FIREBALL: A Dataset of Dungeons and Dragons Actual-Play with Structured Game State Information](https://mlsanity.com/paper/2305.01528/) (2023)
- [Factorio Learning Environment](https://mlsanity.com/paper/2503.09617/) (2025)
- [VolleyBots: A Testbed for Multi-Drone Volleyball Game Combining Motion Control and Strategic Play](https://mlsanity.com/paper/2502.01932/) (2025)
- [Reinforcement Learning Environment with LLM-Controlled Adversary in D&D 5th Edition Combat](https://mlsanity.com/paper/2503.15726/) (2025)
- [NTRL: Encounter Generation via Reinforcement Learning for Dynamic Difficulty Adjustment in Dungeons and Dragons](https://mlsanity.com/paper/2506.19530/) (2025)

## Abstract (arXiv)

Games and simulators make valuable benchmarks by turning decisions into measurable outcomes, but many current suites under-test rules-rich tactical reasoning: the ability to choose well when geometry, timing, resources, objectives, and rule interactions all matter at once. We introduce DungeonBench, a benchmark for tactical reasoning in Dungeons & Dragons combat, built to cover the vast majority of combat-relevant 2014 System Reference Document content whose effects can be resolved by the simulator while retaining mechanics that simplified combat simulators often abstract away. At each step, DungeonBench exposes a complete tactical observation, a pending decision, and an indexed list of executable options spanning movement, attacks, spells, reactions, objectives, preparation, and scarce resources. The task is to value legal choices whose consequences depend on action economy, creature traits, battlefield geometry, timing windows, and future encounters. DungeonBench has two tracks: Encounter, which evaluates local tactical play in single fights, and Day, which links encounters through persistent hit points, spell slots, consumables, preparation, and short-rest timing, forcing policies to trade off immediate tactical advantage against future survivability. The same engine-generated decision stream supports heuristic controllers, language-model policies, learned option rankers, and masked-action reinforcement-learning agents. We evaluate frontier language-model policies on this shared decision stream. Results show that full tactical observations do not saturate the benchmark: frontier policies often win direct encounters, but linked encounter days expose failures in resource budgeting, rest timing, and rule-aware tactical discipline.