What the Condorcet Voting Method Is
Condorcet methods are a family of ranked-ballot voting systems built on the principle named for the Marquis de Condorcet (1785): if a candidate would defeat every other candidate in a one-on-one majority vote, that candidate — the Condorcet winner — should win the election.
“Condorcet” names a criterion rather than one algorithm; the family of methods exists because the principle sometimes yields no winner, and each method resolves that failure differently.
What Condorcet Methods Rank
Condorcet methods rank candidates from ranked ballots by simulating all possible pairwise duels:
- Every ballot is read as a set of head-to-head verdicts (A over B, B over C…)
- Each candidate pair is tallied as a majority contest
- The result is a pairwise comparison matrix — a complete picture of who beats whom
Core Inputs Used by Condorcet Methods
- Ranked ballots — each voter’s full or partial ordering of candidates
- Pairwise tallies — for every pair, how many voters prefer each side
Like Borda, no preference intensities are used — only order.
How Condorcet Methods Work (High-Level)
- Build the pairwise matrix: for every pair, the winner of their simulated duel.
- If one candidate wins all duels, elect the Condorcet winner — in many real electorates, this happens and the method is done.
- When preferences cycle (A beats B, B beats C, C beats A — the Condorcet paradox), apply a resolution rule:
- Schulze method — elects via strongest indirect beat-paths through the graph of victories
- Ranked pairs — locks in the largest victories first, skipping any that create a cycle
- Copeland’s method — simply counts duels won minus duels lost
Conceptual model: Hold a round-robin tournament among candidates inside every ballot — the champion is whoever no one can beat; when the tournament ties, untangle the strongest victories first.
Key Properties
- Majority-consistent — a candidate with a majority of first-choice support always wins (unlike Borda)
- Clone-resistant — nominating similar candidates doesn’t flip the outcome
- Criterion-rich — leading methods (Schulze) satisfy long lists of fairness criteria voting theorists value
- Cycle-dependent — the paradox means the resolution rule, not the principle, decides messy elections
Update Frequency
Condorcet methods compute once per election — they are aggregation rules for a fixed ballot set, not standing ratings.
Known Limitations and Criticisms
- The paradox — cyclical majorities are theoretically possible and occasionally real
- Complexity — explaining beat-paths to a general electorate is genuinely hard
- Later-no-harm failures — ranking additional candidates can hurt a voter’s favorite in some methods
- Low adoption in politics — despite theorists’ preference, major democracies don’t use Condorcet systems for national elections
Where Condorcet Methods Are Used
Condorcet-family elections are used in:
- Open-source communities — Debian’s project leader elections (Schulze), Ubuntu, and others
- Wikimedia Foundation board elections
- Some student governments, societies, and technical standards bodies
- Voting-theory research as the benchmark family of methods
Summary
Condorcet methods rank candidates by every possible duel: the candidate no majority rejects is the winner, and when majorities contradict themselves, graph-flavored rules untangle the cycle. They are the theorist’s favorite family — maximally faithful to majority will, at the price of maximum complexity.
References and Sources
- Condorcet, M. de. Essai sur l’application de l’analyse à la probabilité des décisions (1785).
- Schulze, M. A New Monotonic, Clone-Independent, Reversal Symmetric… Single-Winner Election Method.
- Wikipedia. Condorcet method and Condorcet paradox.