Metraipolis › archive › dc_73664fd006bb

Reinforcement Learning for the Game of Balatro: A Fast Simulator and Two-Policy Training Approach — arXiv:2610.39049

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.39049
License (as archived): CC0-1.0
Archived: 2026-10-31T06:13:03.350Z
SHA-256 of the archived text: 999d62ddf6d5fff8c2c2048bef8a66dd1e57d90251779b6e008076b5f304adde
Stable URL of this page: https://metraipolis.com/documento/dc_73664fd006bb

How to cite

Reinforcement Learning for the Game of Balatro: A Fast Simulator and Two-Policy Training Approach — arXiv:2610.39049 — arXiv (en), https://arxiv.org/abs/2610.39049 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_73664fd006bb (Metraipolis archive dc_73664fd006bb, sha256 999d62ddf6d5fff8c2c2048bef8a66dd1e57d90251779b6e008076b5f304adde)

BibTeX

@misc{dc73664fd006bb,
  title = {Reinforcement Learning for the Game of Balatro: A Fast Simulator and Two-Policy Training Approach — arXiv:2610.39049},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.39049},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_73664fd006bb; sha256 999d62ddf6d5fff8c2c2048bef8a66dd1e57d90251779b6e008076b5f304adde},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39049. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Balatro, a 2024 roguelike deckbuilder, presents three difficulties that have historically derailed reinforcement learning methods: a 3,000-action state-dependent action space, horizons of several hundred decisions with sparse rewards, and pervasive stochasticity in both card draws and shop generation. Published RL results on the game remain scarce. We introduce an end-to-end pipeline comprising (i) a high-throughput Rust simulator that is automatically state-diffed against the live game client via BalatroBot, (ii) a condensed observation-action interface with dynamic per-step masking of legal moves, and (iii) a multi-agent PPO architecture that decomposes play into a tactical blind policy and a strategic shop policy that share the full game state but act in disjoint phases. Long-horizon credit assignment is partially addressed by per-agent reward routing with explicit cross-agent credit. On the main joint run, training metrics show a sharp phase transition in which the first-blind clear rate converges above 99% and mean episode return increases by roughly an order of magnitude, plateauing at a mean maximum ante of 3 out of 8; single-agent baselines under identical rewards oscillated between greedy and hoarding failure modes and did not stably clear the first blind. These results establish the two-policy decomposition as a strong reactive baseline for Balatro and localise the remaining gap to long-horizon strategic planning rather than tactical blind play.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.39049 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-31T06:07:08.692Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).