A standard election allows you to vote for one candidate. Your ballot looks like this:
The results of such an election are easy to understand, too:
Vanilla wins! But hey, that’s not fair! Most people preferred some fruity flavor over Vanilla. But the fruity flavors split the vote, allowing Vanilla to take the lead.
In a ranked-choice voting election, your ballot looks more like this:
Instead of voting for your top choice, you’re voting for your top choices. You want Strawberry, but would rather have any fruity flavor over Vanilla. In this RCV election, Vanilla can’t just have the most votes: it needs more than 50% of the votes. If no candidate receives more than half the votes, the candidate with the lowest number of votes is eliminated. If your top candidate is eliminated, your vote isn’t wasted: your second-choice candidate now gets your vote.
RCV is spreading. In 2018, it was used in Maine for the first federal election in the USA. New York City just adopted it. San Francisco has been using it since 2004. Australia has been using it since 1918.
Ranked-choice voting proceeds round-by-round. How can we visualize each of these rounds? Let’s see some example.
As an interactive bar chart
In the first round, Vanilla had the most votes, but didn’t hit the required threshold of 50%. Banana was eliminated, and its votes were split between the other three candidates (shown in pink). Finally, Strawberry was eliminated, and most of its votes went to Blackberry, making Blackberry the ultimate winner (shown in blue).
As a Sankey diagram
The same election can be visualized as a Sankey diagram. The thicknesses of the lines flowing between rounds shows the number of votes that moved from one candidate to the next.
As a table, round-by-round
Maybe you prefer numbers? Here’s the number of votes and the changes that happened each round.
As a table, by candidate
Perhaps you want to see a breakdown of what happened to a single candidate. You can drill down to see how Blackberry’s votes changed in each round, and where those votes came from:
Learn more about RCV
The visualizer project is hosted open-source on github. Want to contribute? Drop me a line at firstname.lastname@example.org.