Minority winner — a third wins Choose-One, but the majority's choice wins STAR¶
▶ Live on BetterVoting: vote · results ↗ (election 2p33qq, Test ID BV2215).
One line: under Choose-One, a candidate can win with a third of the vote while two-thirds wanted someone else. Read the whole ballot instead, and the winner is the candidate a majority is actually glad about. This is the repo's canonical example of the minority (plurality) winner — small enough to hold in your head.
The setup¶
100 voters, three candidates for one seat: Ada, Ben, Cleo.
- Ada has a passionate base — a third of the room loves her.
- Ben has a passionate base too — another third loves him.
- Cleo is nobody's fiery favorite, but everybody's easy second choice — the one the whole room is fine with.
Here's how the 100 voters feel, scored 0–5:
| Voters | Ada | Ben | Cleo |
|---|---|---|---|
| 34 | 5 | 0 | 4 |
| 33 | 0 | 5 | 4 |
| 33 | 2 | 1 | 5 |
What Choose-One does¶
Choose-One counts only your top pick. So it only sees the first choices:
| Candidate | First choices | |
|---|---|---|
| Ada | 34 | ← wins (34%) |
| Ben | 33 | |
| Cleo | 33 |
Ada wins with 34%. But look again at the scores: 66 of the 100 voters rated Ada a 0 or a 1. Two-thirds of the room did not want her — they just split their first choices between Ben and Cleo, so Ada's devoted third was enough to win. That's a minority winner, and it's the everyday failure of Choose-One in any race with more than two candidates.
What reading the whole ballot does¶
Now count the same ballots with a method that reads all of them:
- STAR: Cleo leads the scoring round with 433, then wins the automatic runoff over Ada 66–34. → Cleo.
- Ranked Robin: Cleo beats Ada head-to-head (66–34) and Ben (67–33) — she's the Condorcet winner. → Cleo.
Both find Cleo — the candidate most people are genuinely happy with.
The line-up¶
| Method | What it reads | Winner |
|---|---|---|
| Choose-One (Plurality) | first choices only | Ada — 34% |
| Ranked Robin | every head-to-head | ✅ Cleo |
| STAR | scores, then a runoff | ✅ Cleo |
The one big idea¶
Nobody voted strategically. Nobody changed their opinion. The entire difference is how much of the ballot the method reads:
Choose-One sees only your favorite, so a passionate minority can win while the broadly-liked choice — everybody's easy second — is invisible to it. STAR and Ranked Robin read your whole ballot, so the candidate a majority actually prefers is the one who wins.
That's the case for a better ballot and a fuller count, in one small election. (It can get starker: with a bigger field, Choose-One winners can take office on 10–20%. A third is just the most common, most believable version — see it degrade in the silly, runnable pineapple progression, where a fan-favorite topping wins on 34% → 25% → 11% as the menu grows, while plain Cheese keeps winning STAR.)
See the same lens across every method: same opinions, every method. The concept page: Choose-One / Plurality. Full engine detail: bv2215_2p33qq_minority_winner_tabulated.txt · run it: bv2215_2p33qq_minority_winner.yaml.