Voting methods

tally-kit supports multiple methods. The right method depends on the kind of decision you’re making. This page explains what “approval voting” means, why we use it for some votes, and when ranked or point-budget ballots make more sense.

This election uses: approval (checkbox). This means every ballot is a simple checkbox ballot.

Approval voting (checkbox ballot)

You approve every option you’d be OK with. You can approve one, several, or all options. The option with the most approvals wins (and the full approval totals are also shown in results).

Tie-break: if there’s a tie for the most approvals, tally-kit breaks the tie deterministically. By default, it uses a deterministic-random order (seeded from the election + ballot id) so ties don’t default to alphabetical order. Some “band / guardrail” ballots may intentionally use candidate-order tie-breaks (for example: Low → Medium → High) to set a conservative default.

Why it’s good

Low friction: no ranking, no math, no “points” budgeting.
Captures acceptability: “I can live with this” vs “no”.
Great for bands / guardrails: if multiple ranges are acceptable, you can express that directly.
Good signal for consensus: high approvals usually means broad agreement.

How to vote (practical)

If only one option is acceptable to you, approve just that one.
If you’re flexible, approve the options you’d accept (even if you have a favorite).
If you truly don’t care, approving all options is a valid “any is fine” signal.

Why not ranked choice for this kind of vote?

Ranked choice is great when you’re picking a single winner from a long list (like naming), because it captures “if my top choice loses, here’s my next best”. But for policy / band decisions (like hourly rate bands or timeboxes), forcing a strict ordering can be misleading: you might genuinely be OK with multiple bands, and approval voting lets you say that directly.

Other methods you may see in tally-kit

Ranked choice (rank your top N)

Use ranked ballots when you want a single winner and voters can express compromise preferences. This is often the best UX for naming votes.

Tie-breaks (ranked): tally-kit uses a deterministic, auditable tie-break chain:

IRV winner: if there’s a unique majority winner in the final IRV round, that wins.
IRV tie: if IRV ends in a tie, we break the tie by Borda score within the tied set.
Still tied: if still tied and a ballot includes prior scores, we use those; otherwise it’s a manual decision.

Note: in earlier IRV rounds, if there’s a tie for last place during elimination, we eliminate deterministically (alphabetical by candidate name) so the process is reproducible.

Point budget / cumulative (allocate a fixed budget)

Use a point budget when you want voters to express intensity (how much more they prefer option A than B). It can be powerful, but it asks more effort from voters and can invite strategy.

Tie-break (point budget): highest total points wins. If there’s a tie, tally-kit breaks ties deterministically (alphabetical by option name).