Voting system types

You place a single cross next to one name. Whoever collects the most crosses wins — even with just one voter in four behind them. The gesture is unbeatably simple, but everything hinges on that one choice: there's no way to say "I prefer X, but if X can't win, I prefer Y" — that's the trap of tactical voting.

This is the current system. If no candidate clears 50% in the first round, only the top two face off in a second. The final head-to-head gives the winner clear legitimacy, but your favourite can be eliminated as early as round one because of a "spoiler effect" — a similar candidate splitting the vote.

You rank the candidates in order of preference. At the count, the last-place candidate is eliminated, and every ballot that ranked them first transfers to the next candidate on that ballot — and so on until someone clears 50%. You can vote sincerely for a minor candidate without "wasting" your vote: it will transfer. The flip side: every elimination means picking the ballots back up, which makes counting locally, station by station, essentially impossible — it has to be centralized.

You rank the candidates. Each rank is worth a decreasing number of points (1st gets the most, last gets the least), and it's all added up. Unlike instant-runoff, a single pass is enough: the points add up directly, station by station. In exchange, the method is easy to game: ranking your most dangerous rival dead last, even insincerely, hands points to everyone but them.

You rate each candidate independently of the others, like reviewing dishes at a restaurant. Two candidates can get the same score, or none can get the top score — rating isn't ranking. Simple to count (it's all addition), but the optimal strategy — giving everyone either 0 or 5, never anything in between — drowns out more nuanced voters.

You check every candidate you approve of, with no limit on how many. Whoever gets the most checks wins. Just as simple to count as a single-round election, and with no spoiler effect: approving one more candidate takes nothing away from the others. The limit: a plain yes/no expresses no intensity — there's no way to say you slightly prefer X over Y if you approve of both.

You give each candidate a grade, from "Excellent" to "To reject" — like a school report card. The winner is whoever's middle grade (the median) is best, not the average. More robust than an average against isolated extreme opinions, but a coordinated bloc of voters grading everything "Excellent" or "To reject" can, at scale, shift that median.

You rank your 5 favourite candidates. The count compares every pair of candidates like a series of small duels: between X and Y, who was placed ahead by more voters? Whoever beats every other candidate in a head-to-head wins — the method most faithful to the collective preference. The cap at 5 candidates (instead of 12) serves both the voter (ranking 5 favourites is reasonable, ordering 12 candidates one by one is a much heavier cognitive load) and the count (fewer ranks to mark means fewer marks to read and check) — without changing the way the matchup matrices add up, just as easily at 12 as at 5.