Skip to content

Participation / no-show paradox — one electorate, told twice (BV2174 / BV2175, live)

The tabulatable evidence behind the Participation topic hub. Eight April fans decide whether to vote; three races on each election (STAR, RCV-IRV, Choose-One) show what their sincere ballots buy under each method: Choose-One and STAR reward them (Celia→April and Bruno→April), RCV-IRV punishes them — Bruno→Celia, their second choice replaced by their last. "Vote — it can only help" is a promise RCV-IRV structurally can't make.

▶ Live on BetterVoting: - BV2174 (8 fans stay home) — vote · results ↗ (election yyhr66) - BV2175 (8 fans vote) — vote · results ↗ (election 9dhv8y)

No Ranked Robin races: both electorates are a Condorcet cycle (that's the soil the paradox grows in), and BV resolves a Copeland tie at random — not freezable. The LH engine's margin tiebreak resolves it deterministically (Celia in BV2174, April in BV2175); see the _tabulated mirrors and LH vs BV on RR ties.

Read the reader-friendly pages (participation_no_show_pages/); the .yaml beside each is the tabulatable source.

Page (read this) What it shows src
BV2174 — the baseline (54 voters) RCV-IRV → Bruno; STAR → Bruno via a Runoff Reversal (Celia tops scores 136/122/120, loses the runoff 34–20); Choose-One → Celia; pairwise = a perfect cycle .yaml
BV2175 — the 8 fans show up (62 voters) Choose-One → April (helped), STAR → April (favorite wins), RCV-IRV → Celia (their extra first-choice votes got their own fallback eliminated — the no-show paradox, and the Twin paradox read forward) .yaml

Concept hubs: Participation · monotonicity (same elimination machinery) · the paradox-catalog page: no-show · up: method_comparisons — same ballots, different methods

file: README.md