Range / Score voting tabulation engine¶
Tabulates Range (Score) voting: every voter grades each candidate on a fixed 0..max scale; the candidate with the highest total score wins. No runoff, no elimination — just sum the grades.
Engine choice — pref_voting¶
This wraps pref_voting (Eric Pacuit) via its native grade_methods.score_voting. pref_voting was chosen over Votelib and VoteKit because it is already this repo's independent cross-check engine (see the pref_voting engine) — so range tabulation adds no new dependency and reuses an actively-maintained, well-tested library. Range itself is trivial ("sum the scores"), so we also compute the totals by hand and assert the two agree: the result is reproducible and independently verified. If pref_voting isn't installed the hand count still runs (the cross-check is skipped with a note).
Usage¶
python 06_Other/Range/Range_tabulation_engine/range_tabulation.py <election.yaml>
Reads the library's usual score grid (header row of names, then one 0–max row per voter; N × / N: weights and markers handled). It ignores voting_method: — any score ballot can be read as Range. It writes a full-context <stem>_RANGE_tabulated.txt into the source folder's <folder>_tabulated/ mirror (the RANGE suffix keeps it from colliding with a STAR mirror of the same election).
Where Range sits among the methods¶
- Approval is Range at 1-bit resolution (grades restricted to
{0, max}). - STAR is Range's score round plus an automatic runoff — added precisely to fix Range's strategy problem (see below).
- Range is the most expressive single-mark method but the most strategy-exposed: a voter's rational play is often to give max/min only (which collapses Range toward Approval). STAR's runoff is the standard answer.
See the teaching page Range / Score Voting for the method overview, pros/cons, and ballot examples, and the Black Curtain read as Range. Range is a non-EVC method, so its examples live in other methods, not a numbered root folder.