Ranked-choice vote counter

How the count works

The headline result is instant-runoff voting (IRV), the single-winner method most people mean by “ranked-choice voting”. Each round:

1. Every ballot counts for its highest-ranked option that is still in the running.
2. If one option has more than half of the ballots that are still counting, it wins.
3. Otherwise the option with the fewest votes is eliminated, and the next round begins.

A ballot whose ranked options have all been eliminated is exhausted and drops out of the count, so “more than half” is measured against the ballots still in play in that round. Voters do not have to rank every option; a shorter ballot simply exhausts sooner.

Ties for last place are broken by looking backwards: of the tied options, the one with fewer votes in the most recent earlier round where they differed is eliminated. If they were level all the way back to round one, the counter eliminates the first alphabetically and prints a note saying the outcome should be treated as a tie. Alphabetical order is arbitrary, and the note is there so nobody mistakes it for a real result.

A worked example where every method disagrees

Fifteen colleagues rank three lunch options. Six rank Pizza, then Sushi, then Tacos. Five rank Tacos, Sushi, Pizza. Four rank Sushi, Tacos, Pizza.

• Plurality counts only first choices: Pizza 6, Tacos 5, Sushi 4. Pizza wins, even though nine of fifteen people ranked it last.
• Instant runoff: nobody has 8 of 15 in round one, so Sushi is eliminated and its four ballots move to their second choice, Tacos. Round two is Tacos 9, Pizza 6. Tacos wins.
• Borda count gives 2 points for a first choice, 1 for a second and 0 for a third: Sushi 19, Tacos 14, Pizza 12. Sushi wins.
• Condorcet compares every pair: Sushi beats Pizza 9 to 6 and beats Tacos 10 to 5. Sushi is everyone’s acceptable compromise and wins every head-to-head contest.

Three methods, three winners, from one set of ballots. Sushi is eliminated first under instant runoff despite being the option a majority prefers to each alternative. This pattern, where a broadly liked middle option has the fewest first choices, is known as the centre squeeze, and it is the most-cited criticism of IRV. None of the methods is wrong; they measure different things. The practical advice is to agree the rule before the vote, not after seeing which rule favours your preference.

Which method should a group use?

• Instant runoff is familiar, easy to explain round by round, and used in real public elections, which makes it easy for a group to accept.
• Borda favours consensus options and suits choosing something everyone can live with, like a team name or a venue. It is easy to game by ranking a strong rival last.
• Condorcet is the strongest claim to “the group prefers this”, but a winner does not always exist: preferences can go round in a circle (A beats B, B beats C, C beats A).
• Dot voting, where everyone spreads a few votes across options, is a quick way to narrow a long list before a ranked vote. Count dots with plurality.

Where it stops being reliable

This is a counting tool for informal group decisions. It does not collect ballots, check that each person voted once, or keep votes secret: whoever types the ballots in sees them all. Public elections that use ranked-choice voting each have their own statutory rules on tie-breaks, batch eliminations and ballot errors, and those rules can differ from the ones here. Do not use this page to verify an official result.