Pre-registration (frozen 13 Nov 2026): does PDCFR+'s advantage over other CFR variants survive matched hyperparameter tuning? Replication case C34 and M2 analysis plan with Amendment 2
Frozen protocol, decision code, selection and analysis harness, Kaggle driver and planted-effect table for a pre-registered replication of PDCFR+ (arXiv 2404.13891) with matched grids for every tunable comparator, deposited before any run; plus the M2 analysis plan including Amendment 2 (effect size in orders of magnitude for log-axis metrics, the game as the unit). Simulation only.
Authors
- John Goodman (ORCID: https://orcid.org/0009-0008-4080-8043)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-13
- DOI
- https://doi.org/10.5281/zenodo.24183913
- Primary Topic
- Artificial Intelligence in Games
- Type
- article
- Field-Weighted Citation Impact
- 0.00