Metraipolis › archive › dc_a36698dbac60

PureTD: Reinforcement Learning for Backgammon Money Games with No Evaluation-time Search — arXiv:2608.15146

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.15146
License (as archived)
CC0-1.0
Archived
2026-10-18T19:40:05.869Z
SHA-256 of the archived text
7e9d3b5f1d66074afe9a66b3834ee6083531e1371a91555260c5693afcc98549
Stable URL of this page
https://metraipolis.com/documento/dc_a36698dbac60

How to cite

PureTD: Reinforcement Learning for Backgammon Money Games with No Evaluation-time Search — arXiv:2608.15146 — arXiv (en), https://arxiv.org/abs/2608.15146 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a36698dbac60 (Metraipolis archive dc_a36698dbac60, sha256 7e9d3b5f1d66074afe9a66b3834ee6083531e1371a91555260c5693afcc98549)
BibTeX
@misc{dca36698dbac60,
  title = {PureTD: Reinforcement Learning for Backgammon Money Games with No Evaluation-time Search — arXiv:2608.15146},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.15146},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a36698dbac60; sha256 7e9d3b5f1d66074afe9a66b3834ee6083531e1371a91555260c5693afcc98549},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.15146. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We revisit Tesauro's TD-Gammon for backgammon money games in the setting of no evaluation-time search. Both checker play and cube action (use of the doubling cube) are learned from scratch via self-play reinforcement learning (RL), with minimal hand-coded logic and no expert features. In this setting, we demonstrate that pure self-play RL suffices to train models that reach near-state-of-the-art playing strength. Specifically, for cubeful money games, our search-free model evaluates faster and is substantially stronger than the open-source engines GNU Backgammon and Open Sage running a one-move (1-ply) look-ahead search. Our method of training a cubeful model also outperforms the standard approach of training a cubeless model and deriving cubeful equity with Janowski-style formulas, when the two models are equally sized.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.15146 — 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 2026-10-18T19:02:44.077Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).