On the Reproducibility of Technical Game AI Research
Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment · 08.11.2026 · 22 (1) · 192-200
DOI: 10.1609/aiide.v22i1.43577 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, online erschienen am 08.11.2026. Verfasst von Reva Agarwal und 10 weiteren Autor:innen.
Abstract (Originalsprache)
Modern science is often described as being in a reproducibility crisis, where it is often challenging to validate and build upon published results. In a computational field like ours, publications may not provide source code, data sets, or may not provide clear or reproducible instructions on how to run the code, even if it is provided. In this paper, we investigate concrete sources of challenges with reproducing and building upon existing work and discuss a set of case studies in the form of attempts to reproduce eight different papers. We report challenges that were encountered in the reproduction of each, and discuss common themes between the different case studies, as well as which structural challenges may befall our field as a whole.
Zitieren
APA
Agarwal, R., Afreen, A., Chang, K., Duan, J., Ernest, N. J., Martinez-Arias, I., Pelyushenko, A., Pratt, B., Villalobos, O., Wessel, T., & Eger, M. (2026). On the Reproducibility of Technical Game AI Research. Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, 22(1), 192-200. https://doi.org/10.1609/aiide.v22i1.43577
Vancouver
Agarwal R, Afreen A, Chang K, Duan J, Ernest NJ, Martinez-Arias I, et al. On the Reproducibility of Technical Game AI Research. Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment. 2026;22(1):192-200. doi:10.1609/aiide.v22i1.43577
BibTeX
@article{Agarwal2026,
author = {Agarwal, Reva and Afreen, Aamena and Chang, Kai-Hsiang and Duan, Jiaqi and Ernest, Nihal Jonathan and Martinez-Arias, Ivan and Pelyushenko, Alexei and Pratt, Benjamin and Villalobos, Octavio and Wessel, Thomas and Eger, Markus},
title = {On the Reproducibility of Technical Game AI Research},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment},
year = {2026},
volume = {22},
number = {1},
pages = {192--200},
doi = {10.1609/aiide.v22i1.43577}
}
EUVIMED