Replication package for "How Technical Architecture Shapes Platform Emergence: A Multi-Agent Simulation of Data Collaboration Strategies"

Abstract

Replication package for the HICSS-60 paper "How Technical Architecture Shapes Platform Emergence: A Multi-Agent Simulation of Data Collaboration Strategies" (Aoki & Yoshioka-Kobayashi, 2027). The paper formalises self-reliance, raw-data pooling and federated learning as profiles over two architectural primitives — integration capacity (Cap) and the preservation of competitive distinctiveness (Scarcity dynamics) — and studies their population dynamics in an agent-based model of competing firms. This archive contains everything the paper undertakes to release: the simulation code and analysis pipeline; the configuration for every simulation cell; run manifests recording parameters, commit hashes and software versions; replication-level output for the twelve cells reported in Table 3 (30 replications each); a full model description following the ODD protocol (Grimm et al., 2020); the project's pre-specification together with a note reconciling it with the published paper; and the literature-search protocols behind the paper's two scoped novelty claims. No new simulation was run for the revision: every number in the published paper derives from the output archived here. Section 5.4 of the published paper retracts a size-ordered adoption cascade reported in the submitted version; the superseded analysis and the code that produced it are retained in this archive so that the earlier reading can be reproduced and checked. Two archives are provided. The code archive(under 1 MB) holds code, configurations, notebook and documentation; the results archive (~90 MB) holds the simulation output. Unpack the results archive into the package root. Start from README.md, and from docs/ODD.md for the model itself. Code is licensed under the MIT License; simulation output and documentation under CC BY 4.0, as stated in LICENSE-DATA.md inside the archive.