Why ranked voting beats a yes/no grid

What a grid of ticks cannot tell you

Suppose five people are asked about four slots and the grid comes back with Tuesday 10:00 ticked by all five. Decision made. Except that three of them ticked it because it was possible, not because it was good: one has to leave another meeting early, one is on a train, one would far rather have met Thursday. The grid recorded "yes" for all three, and the organiser never found out.

A tick box flattens three different answers into one. "I would love this time", "I can do this at a push" and "I will say yes to anything so this ends" all look identical. When a grid shows a tie, which happens often with small groups, the organiser has no information left to break it with except guessing.

The grid also pressures people to overcommit. Ticking nothing feels unhelpful, so people tick slots they would rather not take, and the meeting lands on one of those. Then half the attendees turn up distracted or send apologies at the last minute.

What ranking changes

Ranking asks a richer question in almost the same amount of time: of these slots, which is your first choice, which your second, which your third, which others would you accept, and which can you not make at all? Each voter in FINDaTIME gives at most one 1st, one 2nd and one 3rd choice, and can mark any number of slots Also ok (workable, just not a favourite) or Unavailable. A slot left blank is simply "no answer".

Those limits are deliberate. If a voter could mark everything as first choice, you would be back to the tick grid. Having to choose forces each person to reveal a preference, and the limited number of ranks means the preference is meaningful.

How the score is worked out

A 1st choice is worth three points, a 2nd choice two, a 3rd choice one and Also ok one; Unavailable is worth nothing. Also ok scores the same as a 3rd choice on purpose: it says "this works for me" without claiming a preference, so it adds a little weight and counts the voter as available. But points are not the first thing that decides the order. Slots are ranked by these rules, applied one after another:

1. Fewest people who marked the slot Unavailable.
2. Highest total score.
3. Most people who said they could make the slot, by a rank or by Also ok.
4. The earliest start time, as a final tiebreak.

The first rule is the one that matters most. A slot one person cannot make ranks below every slot that everyone can make, however many first choices it collected. The purpose of the meeting is usually to have everyone in the room, and a popular slot that excludes one person is not a win. The slot is not thrown away, though: it stays in the list in its proper place, so if you decide the missing person can be briefed afterwards, the choice is yours to make with the facts in front of you.

A worked example

Four people, three slots: Monday 09:00, Tuesday 14:00 and Thursday 16:00. Ana ranks Monday first, Tuesday second, and marks Thursday unavailable. Ben ranks Monday first, Thursday second, Tuesday third. Chloe ranks Tuesday first, Monday second, and marks Thursday Also ok. Dev ranks Thursday first, Tuesday second, and marks Monday unavailable.

On a tick grid Monday and Thursday each get three yeses and Tuesday gets four; Tuesday wins without much thought. With ranking, Monday scores 3 + 3 + 2 = 8 points, as much as any slot, but Dev cannot make it. Thursday scores 2 + 1 + 3 = 6 (Chloe's Also ok is the 1) with Ana unavailable. Tuesday scores 2 + 1 + 3 + 2 = 8 with nobody unavailable. Tuesday comes top, and you can see that it is everybody's acceptable choice rather than anybody's ideal one, which is exactly the kind of thing worth knowing before you send the invitation.

Now change one answer: Dev marks Monday as his third choice instead of unavailable. Monday now scores 9 with no one unavailable and takes first place, and the result tells you that Monday is the slot two people actively want. A tick grid would have shown the same four ticks either way.

Why a blank is not a no

People skip slots for all sorts of reasons: they ran out of ranks, they did not scroll far enough, they were not sure. Treating a blank as "unavailable" would punish slots at the bottom of the list for no reason. Treating it as "available" would invent commitments nobody made. FINDaTIME does neither: a blank counts as neither available nor unavailable, and the results page shows how many people responded so you can judge how much weight the numbers deserve.

For the same reason, no recommendation is shown while nobody has said they can make the top-ranked slot. An empty poll is not a decision, and a recommendation built on nothing would only mislead.

Getting the most out of a ranked poll

• Offer four to eight slots. Fewer than three makes ranking pointless; more than ten wears people out.
• Ask people to mark Unavailable honestly. It is the one answer that changes the outcome most.
• Ask people to use Also ok on every slot they could live with. It separates "I would rather not" from "I cannot", and it is what keeps a workable slot from looking unpopular.
• Set a deadline, so the vote closes while the slots are still in the future.
• Read the second-placed slot too. If it is close, it is your fallback when someone drops out.

See also how FINDaTIME works end to end. To put ranking to work on your next meeting, create a poll.