Skip to content

rr_vs_irv_plurality — Ranked Robin vs. IRV vs. plurality (same ballots)

Where the Condorcet vs. Ranked Robin set asks "is Ranked Robin the same as the Condorcet winner?" (answer: yes, until a cycle), this set asks a different question: on one ranked ballot set, do the common methods even agree on a winner? They don't — and the gap is the lesson.

Case (page) What it shows Plurality RCV-IRV Ranked Robin src
BV2131 — Tennessee capital one ballot set, three winners — the classic center-squeeze Memphis Knoxville Nashville .yaml

Tennessee in one line: Nashville is nearly everyone's second choice, so it beats every rival head-to-head (the Condorcet winner Ranked Robin elects) — but it holds few first-choices, so plurality ignores it and IRV eliminates it before the final round.

Each case here is triple-checked: LH native run_ranked_robin, BetterVoting's RankedRobin.ts (frozen _bv_export.json), and pref_voting's independent Copeland — all must agree on the winner. See each page's "Agreement" table.

Concept docs: Ranked Robin (the method) · Condorcet efficiency (topic hub) · why IRV is fragile here: center squeeze in Ranked Robin vs. Condorcet.

file: README.md