election_title: "Reinforcement — Combined, Ben branch (9 voters; both halves say Ben, the whole says Ada)"

scenario_description: |-
  North (6 voters, the symmetric cycle) merged with the BEN-branch South
  (reinf_south_ben_c3_b3_rr.yaml). Ben is a co-winner of North's three-way tie and
  the outright Condorcet winner of South — a winner in both halves. Reinforcement
  (consistency) therefore demands Ben win the union.

  He doesn't. A NEW Condorcet winner appears in the merged electorate:

      Ada beats Ben 5–4 · Ada beats Cara 5–4 · Ben beats Cara 7–2

  so every Condorcet method elects ADA. This is the same reinforcement failure as
  the original Ada-branch trio, rotated one step: the pairwise majorities that
  decide the merged election (Ada's two 5–4 edges) exist in neither half on their
  own — head-to-head majorities do not add across electorates, which is exactly why
  no Condorcet extension can be reinforcing.

  Branch companions: reinf_combined_c3_b9_rr.yaml (Ada branch → Cara wins),
  reinf_combined_cara_c3_b9_rr.yaml (Cara branch → Ben wins).

paradoxes: [multiple-districts]

voting_method: RankedRobin
num_winners: 1

lot_numbers: [Ada, Ben, Cara]

ballots: |-
  3:Ada>Ben>Cara
  2:Ben>Cara>Ada
  2:Cara>Ada>Ben
  2:Ben>Ada>Cara

expected_winners:
  - Ada