election_title: "Same mean, different spread — the consensus candidate and the polarizing one"

scenario_description: |-
  Two candidates, five voters, and an exact tie in the Scoring Round: Alice 15,
  Blake 15, both averaging 3.0. Every difference between them is SPREAD. Alice
  is a flat 3 on all five ballots; Blake is three 5s and two 0s. The Score
  Distribution block is the whole lesson — same mean, variance 0.0 vs 6.0 —
  and 6.0 is the LARGEST variance arithmetically possible at a mean of 3.0 on
  a 0-5 scale, so Blake is exactly as divisive as this ballot allows.
  Because the totals tie, pure Score cannot separate them at all; the
  divergence block's Approval reading (approve = 3+ stars) picks Alice 5-3;
  STAR's Automatic Runoff hands the seat to Blake 3-2. "Polarizing" is not a
  mood — it is variance, and only the second round reads it.
voting_method: STAR
num_winners: 1
ballots: |-
  Alice,Blake
  3,5 # Blake is my champion; Alice is fine
  3,5 # Blake is my champion; Alice is fine
  3,5 # Blake is my champion; Alice is fine
  3,0 # Alice is fine; Blake, absolutely not
  3,0 # Alice is fine; Blake, absolutely not

expected_winners:
  - Blake

# file: same_mean_different_spread_c2_b5.yaml