election_title: One point — a single mark fills one seat and the lot fills the other (Bloc STAR)
scenario_description: |-
  The zero-support ballots with one thing changed: voter 3 gives Ben a 1.

  That single mark is the only support anybody expresses in the whole
  election — one point out of a possible 75 (3 voters x 5 nominees x 5
  points). It is enough to settle the first seat outright, and it changes
  nothing at all about the second, which is still decided by lot.

  The reason this file sits beside the all-zero six is that it is the HARDER
  case to read, not the easier one. The all-zero reports look degenerate, so
  nobody mistakes them for a mandate. This one looks decisive: the runoff
  summary reads "Ben 1 (100%) vs Ada 0 (0%); majority = 1", because the
  decided-voters denominator is the one voter who expressed a preference.
  Every number on that line is correct and the impression it leaves is not.

  It is also where a tempting rule fails. "Refuse to count an election with no
  support" is easy to state while every score is 0 and impossible to state
  here: seat 1 rests on a real, if tiny, preference that a returning officer
  should report, and seat 2 rests on nothing at all. One election, both
  answers, no threshold that separates them without being a policy number in
  disguise. What the engine can do instead is what it does below — count it,
  and say per seat which rung paid for it.

  Same cast and same first two ballots as the zero-support set; see the README
  for the six methods that count the unchanged version.


voting_method: Bloc STAR
num_winners: 2

lot_numbers: [Ada, Ben, Cleo, Dev, Elsa]

ballots: |-
  Ada,Ben,Cleo,Dev,Elsa
  0,0,0,0,0   # voter 1 — turned out, then scored every nominee 0
  0,0,0,0,0   # voter 2 — a deliberate 0 is not a blank ballot
  0,1,0,0,0   # voter 3 — one point for Ben, the only mark in the election

expected_winners:
  - Ben
  - Ada

# file: one_point_bloc_star.yaml