election_title: "The same seven voters, grouped by the FIZZY label instead — no spoiler"
scenario_description: |-
  Byte-for-byte the same election as 09a. One thing changes: the blocs: key names
  the four FIZZY drinks instead of the four sweet ones.

  Fizzy holds 4 of 7 first choices — the identical headcount the sugar grouping
  holds. (Not spread the same way: sugar divides 1+1+1+1, fizzy 2+1+1+0 —
  Sparkling Water is nobody's favourite. The headcount is what matches.) And the engine reports NO
  spoiler, because the fizzy front-runner (Diet Cola) simply won.

  Run this beside 09a and read the two [Vote-splitting check] blocks together.
  Same ballots. Same winner. Opposite verdict. A category is not a clone set.

  Live results: https://bettervoting.com/8xrpyp/results

bv_test_id: BV2295
bv_election_id: 8xrpyp
bv_results_url: https://bettervoting.com/8xrpyp/results
blocs:
  Fizzy (a label, not a clone set): [Cola, Root Beer, Diet Cola, Sparkling Water]

voting_method: STAR
num_winners: 1

ballots: |-
  Cola,Root Beer,Lemonade,Sweet Tea,Diet Cola,Sparkling Water,Unsweet Tea
  5,4,4,3,1,0,0   # cola, and anything sweet will do
  4,5,3,3,1,0,0   # root beer, and anything sweet will do
  3,2,5,4,0,1,0   # lemonade, and anything sweet will do
  3,2,4,5,0,0,1   # sweet tea, and anything sweet will do
  1,1,0,0,5,4,3   # diet cola - no sugar, thanks
  1,0,0,0,5,3,4   # diet cola - no sugar, thanks
  0,0,1,1,3,4,5   # unsweet tea - no sugar, thanks

expected_winners:
  - Cola

# file: 09b_same_ballots_grouped_by_label.yaml