Define and sensitivity-test earliest-replication outcome grouping for same-year ties

Citation aggregate outcome chips classify originals by the earliest recorded replication year, with same-year ties following stored record order. The table outcome filter matches any replication. Both are disclosed near the bottom, but users encounter the controls much earlier.

Please choose and document a substantively defensible same-year rule, identify affected originals, test sensitivity of group sizes and estimates to alternative rules, and update analysis documentation and code together. Add concise local explanations beside the chips and table filter distinguishing earliest from any outcome. Do not change statistical grouping silently as a UI patch.

The accompanying frontend patch exposes post-first-replication percentage denominators, labels lifetime citation counts, and gives sort/filter controls explicit accessible names. It leaves the analytical grouping unchanged.

Begin with the citation aggregate outcome chips and table outcome filter, then trace their existing grouping and filter code. Review the analysis documentation alongside affected originals and compare group sizes and estimates under candidate same-year rules. Done when a chosen rule, sensitivity results, synchronized code and documentation, and local explanations distinguishing earliest from any outcome are present.

Written by the indexing model from the issue text.