Back to methods

Voting method

Ranked Choice

Rank candidates in order

How it works

Voters rank candidates from first to last choice. If no one wins a majority, the last-place candidate is eliminated and their votes transfer to the next choice. Repeat until someone has a majority.

Best for

Electing a single winner when you want broad consensus

Configuration

Ballot shape
ranking
Minimum options
2
Seats configurable
No
Pass threshold configurable
No

Worked example

See it in action

Eleven neighbors rank four lunch spots. First choice gets their vote - until someone is eliminated.

  1. The ballots

    11 voters rank the 4options. Only each ballot's highest-ranked surviving option counts in any round.

    • 4Pizza > Salad
    • 3Sushi > Pizza
    • 3Tacos > Sushi
    • 1Salad > Tacos
  2. Round 1 - count first choices

    No one has the 6 votes needed for a majority. Salad has the fewest and is eliminated - every Salad ballot moves to its next ranked choice for the next round.

    • Pizza
      4
    • Sushi
      3
    • Tacos
      3
    • Salad
      1
  3. Round 2 - count first choices

    No one has the 6 votes needed for a majority. Sushi has the fewest and is eliminated - every Sushi ballot moves to its next ranked choice for the next round.

    • Pizza
      4
    • Tacos
      4
    • Sushi
      3
  4. Round 3 - a majority

    Pizza holds 7 of 11 active ballots - more than the 6 needed for a majority - and wins.

    • Pizza
      7
    • Tacos
      4

Ready to run a Ranked Choice poll?

Create a Ranked Choice poll

Edge cases

How ties are broken

Sometimes two candidates tie for last place and one of them has to be eliminated. ClearVote first tries the Next-Preference Cascade: look only at the tied candidates on each ballot and ask which of them voters prefer least. If that singles one out, that candidate is eliminated - the decision reflects real voter preference, not arbitrary order.

If voters give no further signal - every still-tied candidate is preferred equally often - the tie is broken by lot: a shuffle seeded from the raw ballot data itself. The draw is impartial and blind to candidate names, yet fully reproducible - anyone re-running the tally on the same ballots gets the same result.