election_title: "Condorcet's 1788 rebuttal to Borda — RCV-IRV agrees here"
scenario_description: |-
  The third race of BV2250: the same 11 ranked ballots, counted by instant runoff.

      4 : Peter > Paul  > James
      3 : Paul  > James > Peter
      2 : Paul  > Peter > James
      2 : James > Peter > Paul

  This case exists to keep the comparison honest. Condorcet's 1788 example is a
  counterexample to BORDA (and to plurality) — not to IRV. James has the fewest
  first choices (2) and is eliminated; both of his ballots rank Peter next, so
  they transfer and Peter wins 6-5. RCV-IRV lands on the Condorcet winner here,
  exactly as Ranked Robin and STAR do.

  IRV's own Condorcet failures require a center squeeze — a strong middle
  candidate with few first choices, eliminated before the head-to-head that
  would have shown their strength. This profile does not contain one: Peter is
  not squeezed, he leads the first-choice count among the two finalists' bloc
  and simply starts second. Citing this election as an IRV failure would be
  wrong, and the repo says so on the folder page.
video_script: |-
  The fairness beat in the Borda/Condorcet lesson. When we show that plurality
  and Borda both miss the Condorcet winner, the obvious next question from an
  RCV audience is "and IRV?" — answer it before it is asked, and answer it
  honestly: IRV gets this one right. That earns the credibility to be believed
  later when IRV genuinely does fail (see the center squeeze cases).
bv_test_id: BV2250
bv_election_id: khcwm4
bv_results_url: https://bettervoting.com/khcwm4/results
voting_method: IRV
num_winners: 1
ballots: |-
  4:Peter>Paul>James
  3:Paul>James>Peter
  2:Paul>Peter>James
  2:James>Peter>Paul
expected_winners:
  - Peter