election_title: "Exercise 10 — Later-no-harm: the reticent ballots"
bv_test_id: BV2195
bv_election_id: g6q42v
bv_results_url: https://bettervoting.com/g6q42v/results

scenario_description: |-
  Nine voters, three candidates. Amir's four fans withhold everything
  below their favorite: Amir 5, Bess 0, Cato 0. Scoring round: Amir 24,
  Cato 15, Bess 13 — Amir and Cato advance, and Amir wins the runoff 6-3.
  The companion file ex10_generous.yaml reruns the election with those
  fans scoring honestly (Amir 5, Bess 3, Cato 0): Bess reaches the runoff
  instead and beats Amir 5-4. Scoring a later choice DID harm their
  favorite — the later-no-harm failure STAR accepts by design. The
  exercise asks which reading is right: did the generous 3s harm Amir, or
  did the reticent 0s hide the consensus winner Bess? Exercise:
  ex10_later_no_harm.md. Ballots and cast are this repo's own.
  Live on BetterVoting (Test ID BV2195): https://bettervoting.com/g6q42v/results
  — with an RCV-IRV race (also Amir). No Ranked Robin race on purpose:
  the rank conversion of these sparse ballots yields a Condorcet cycle
  whose 3-way tie BV resolves at random — not freezable.
voting_method: STAR
num_winners: 1
ballots: |-
  Amir,Bess,Cato
  5,0,0 # 4 voters: Amir fans, scoring NOTHING below their favorite
  5,0,0
  5,0,0
  5,0,0
  2,5,0 # 2 voters: Bess first, mild nod to Amir
  2,5,0
  0,1,5 # 3 voters: Cato first, a token point to Bess
  0,1,5
  0,1,5
expected_winners:
  - Amir

# file: ex10_reticent.yaml