Ten marks, each one quoted.
The score counts how complete and checkable the published plan on this issue is. Whether the plan would work is a separate question, left to the reader. This page holds the rubric, the searches every candidate was read under, what earns nothing, and the table showing where the method is weakest.
A full score looks like this:
The best available today is 9, because one mark needs a questionnaire that has not gone out:
What earns a mark
- A position in the candidate’s own words2 marks
Two points for a statement on the issue that names the candidate as the person who will act. One point when only the party’s platform covers the issue and the candidate is not quoted.
- Names a mechanism the office controls2 marks
The candidate’s own statement names an instrument the elected body can adopt by its own vote: a bylaw, a budget line, a fee, a program, a land decision, a Board policy.
- States a measurable target2 marks
A number with a unit that describes what will exist or change: homes, spaces, days, kilometres, a percentage.
- States a cost or funding source1 mark
A dollar figure for the measure, or a named source of the money.
- States a timeline1 mark
A date, a year, or a term milestone such as the first budget.
- A dated, sourced action on the issue1 mark
A motion, a vote, a program delivered, a report authored or an organization led, in any public role, with a date and a source.
- Answered the voche.me question on the issue1 mark
The same question went to every candidate in the office. An answer that names the candidate earns the point.
One mark is currently unearnable by anybody. “Answered the voche.me question on the issue” is worth 1, and the voche.me questionnaire has not gone out yet, so no candidate holds that mark and no candidate is disadvantaged by it.
What a score ignores
- A value, a mood or a slogan with no instrument behind it. “I will fight for renters” names nothing the office can adopt by its own vote.
- A party platform with no candidate named as the actor. It earns one position mark and nothing else, and the row says so.
- A promise about something the office has no power over. It is recorded on the profile, marked as outside the office, and earns no mechanism mark.
- Volume. Ten vague commitments score below one costed commitment with a date.
- Endorsements, follower counts, sign counts, money raised, and anything voche.me could not read on a page and quote back.
What was read for every candidate
Every candidate in a published office was read under the same 7 searches, in this order. Each one is logged on the record with the date and what it returned, so a zero can be checked against what was actually looked at.
The candidate’s own website, every page reachable from its menu that concerns the campaign.
The candidate’s page on their party’s site, and every party release that names them as a speaker or an actor.
The party’s platform page, read for this candidate’s name.
The City of Vancouver’s official candidate list, for the ballot name, number, office and party label.
The candidate’s VoteMate profile, where they submitted text themselves.
A name search: the ballot name with Vancouver and 2026, then the name with the office and a topic, opening every result that could hold their own statements.
The public record for any public role on the profile: council, board and committee minutes, motions and recorded votes.
- 100 candidacies have 7 of 7 steps logged.
- 29 candidacies have 6 of 7 steps logged.
Which office is scored on what
An office is scored only on issues it has power over, so a School Board candidate is never marked down for saying nothing about bike lanes.
Mayor
City Council
Park Board
Where this method is weakest
Recomputed from the records on every build and published whether or not it is comfortable. A cell is one candidate on one issue, so an office with ten issues contributes ten cells per candidate.
| Office | Candidates | Cells | Mean | Own words | Party text only | Has a record | Zero |
|---|---|---|---|---|---|---|---|
| Mayor | 8 | 80 | 4.6 of 10 | 100% | 9% | 40% | 1% |
| City Council | 56 | 560 | 1.9 of 10 | 93% | 33% | 15% | 19% |
| Park Board | 35 | 210 | 2.4 of 10 | 100% | 37% | 18% | 5% |
| School Board | 30 | 150 | 2 of 10 | 90% | 36% | 19% | 9% |
The mean runs from 1.9 to 4.6 across the four offices, and the share of cells scoring zero runs with it. The “own words” column is where that gap comes from: it is the share of each office’s candidates quoted in their own words on at least one issue, and it is 100% for Mayor, 93% for City Council, 100% for Park Board, 90% for School Board. The rubric awards marks for what a reader can check, so a candidate who has published nothing scores zero however well they would do the job.
| Party | Candidates | Mean | Own words | Party text only | Has a record | Zero |
|---|---|---|---|---|---|---|
| ABC Vancouver | 24 | 2.3 of 10 | 96% | 36% | 40% | 12% |
| Vancouver Liberals | 19 | 2.3 of 10 | 100% | 31% | 15% | 9% |
| TEAM for a Livable Vancouver | 14 | 1.7 of 10 | 100% | 57% | 10% | 6% |
| OneCity | 13 | 2.9 of 10 | 100% | 22% | 24% | 2% |
| COPE | 12 | 3.3 of 10 | 100% | 23% | 25% | 1% |
| Vote Vancouver | 12 | 2.1 of 10 | 100% | 48% | 15% | 6% |
| Green Party of Vancouver | 10 | 2.5 of 10 | 100% | 31% | 23% | 8% |
| Independent | 9 | 2 of 10 | 67% | 0% | 6% | 49% |
| Bright Future Vancouver | 7 | 1.4 of 10 | 86% | 37% | 1% | 23% |
| Independent Parks Party | 5 | 3.1 of 10 | 100% | 33% | 3% | 0% |
| TrueBlue Vancouver | 3 | 0.9 of 10 | 33% | 40% | 0% | 35% |
| Affordable Housing | 1 | 1.1 of 10 | 100% | 0% | 0% | 80% |
“Party text only” is the share of cells where the party has published on the issue and the candidate has not been quoted on it. A high share describes how a party’s website is built, with one platform and short candidate pages, rather than what its candidates think. “Has a record” favours anyone who has held office before, which is an advantage of incumbency and not of the plan.
How a score changes
A score changes when a candidate publishes something new, when they answer the questionnaire, or when a mark was awarded wrongly. Every change is dated and explained on the candidate’s own page, and the page it was read from was archived as text on the day it was read, so a candidate who later deletes a page keeps the marks it earned.