Cumulative incidence of Final status within 365 days of the first Draft merge, with withdrawal as a competing event, for Ethereum ERCs and Core EIPs: a data note with a privately pre-specified plan (git history 2020-11-05 to 2026-11-11)
For Ethereum improvement proposals first merged as Draft from 2020-11-05 to 2026-11-11, the Aalen-Johansen cumulative incidence of reaching Final within 365 days, with Withdrawn and Living as competing events and open proposals censored at 2026-11-11, was 19.9% (95% CI 16.3 to 23.7) for ERCs (486 proposals, 114 reaching Final by the end of follow-up) and 7.0% (4.1 to 10.1) for Core EIPs (324 proposals, 38 Final). By 1,095 days it was 25.5% and 17.0%. The 91 proposals of the other EIP categories are reported as counts only (26 Final), as the plan requires for a category under 100 proposals. Data: the first-parent git history of the two public repositories (CC0 1.0) at pinned commits, dated by committer time; 901 proposals form the cohort. The intervals assume independent proposals; Core Final events come in batches (38 on 17 dates), consistent with network upgrades, and post hoc, resampling calendar months of the Draft merge gives 15.6 to 24.2% for ERCs and 3.9 to 10.7% for Core. Withdrawal was rare (18 Withdrawn, 1 Living) and is reported as counts. 281 proposals were marked Stagnant, a reversible state, mostly in sweeps (53.7% of first marks on the 10 busiest of 86 dates). ERCs first merged as Draft after their move to a separate repository on 2023-10-25 reached Final less often than those before (14.1% against 25.6% by 365 days). Post hoc, the crude one-year share by year of the Draft merge was 16.7% (2021, 24 proposals), 27.2% (2022), 24.8% (2023), 7.7% (2024) and 17.9% (2025); the post-move years rest on 8 of 104 and 10 of 56, carry no interval and do not show whether the change was a step. No causal claim is made, and crude counts pointing the same way were seen before the plan was written. The plan was written after a feasibility audit had cloned the same commits and counted statuses; it was committed to a private repository, not a public registry, before the clone used for the analysis, which a reader cannot verify, and the analysis code was not committed before its first run. No human has checked the figures. The record holds the note (PDF and Markdown) and a supplement archive: the frozen plan and the feasibility files (internal references replaced), the deviations log, the analysis result summary as written at the run, the checkout record, the stop-rule files, the per-proposal table derived from CC0 metadata, the planned estimates and curves (values withheld by stop rules 3 and 4 removed), the descriptive check of the run, the analysis code with its synthetic test, the independent recheck by a second AI session, the checks added for the note (paper_checks.py, paper_checks.json, diag_recheck_diff.py, recheck_diff.json, fetch_prior_records.py, prior_records.json, fetch_public_refs.py, public_refs.json), a manifest with the sha256 of every file and every change made for the deposit (it also says that the data note supersedes the analysis result summary where they differ), and a trace of every number in the note. The raw git clones are not redistributed; the repositories are cited by URL and pinned commit. Data: git histories of github.com/ethereum/EIPs and github.com/ethereum/ERCs, CC0 1.0. Plan, code, analysis and text by Claude (an AI model by Anthropic) for the author; reviewed in four passes by further Claude sessions; not peer reviewed; no human has checked the figures; the human author is responsible for the deposit.
Authors
- Keisuke Watanabe (ORCID: https://orcid.org/0000-0001-6374-3918)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24402442
- Primary Topic
- Open Source Software Innovations
- Type
- article
- Field-Weighted Citation Impact
- 0.00