How We Assign Candidate Stances

What a stance is

For each quiz question, we place every candidate on a six-point scale that mirrors the one you answer on:

A candidate is only matched to you on questions where both of you have a known position. We never impute a stance a candidate hasn't taken.

The "no unsourced claims" rule

Every non-neutral stance (anything other than 0 or no known position) must cite at least one source — a campaign platform page, a candidate questionnaire, a news report, or a specific piece of DC legislation. The site will not publish a stance that lacks one. Most stances also carry a short verbatim quote from that source so you can judge the wording yourself.

How we tell +1 from +2

"Strongly supports / opposes" (±2) is the easiest label to over-assign, because campaign language is emphatic by nature. We reserve ±2 for cases that clear two bars at once:

1. On-topic. The evidence has to address the question's specific claim, not a neighboring topic or a shared keyword. Supporting tenant cooperatives is not the same as pledging to restore and strengthen TOPA; backing "biking and walking infrastructure" in general is not the same as building protected bike lanes and dedicated bus lanes; saying you'll explore ending mayoral control of schools is not the same as committing to it.
2. A strong signal. At least one of:
3. Emphatic framing — the candidate makes it a top-priority or signature initiative, or uses absolute language ("every school," "fare-free for all," "end all cooperation").
4. A concrete commitment — a specific dollar figure, a named program, a measurable target, or a bill the candidate personally authored or championed on exactly that topic.

When a candidate clearly supports something but the language is general, aspirational, or hedged ("modernize," "review," "explore," "supports… but…"), we record it as +1, not +2.

Weighing the evidence

• The campaign's own words are authoritative for what a candidate believes today. A candidate's platform page or questionnaire answer outranks inference.
• Sponsorship is strong evidence; a lone vote is weak. A bill a candidate authored on a topic can justify a ±2. A single floor vote — which is often a package deal and can be cast for procedural reasons — generally supports only a ±1.
• Council records corroborate, they don't override. For sitting and former Council members we cross-check the official DC Council legislative record (LIMS). An authored, on-point bill can sustain a strong stance even when the campaign site is silent; an off-point bill never does.

Translations and language

When a candidate's content is available in more than one language, each language's text is taken from that candidate's own materials in that language — their English campaign page for the English version, their Spanish page for the Spanish version. Where a campaign's own pages differ between languages — a date, a dollar figure, a turn of phrase — our pages reflect that difference rather than silently "correcting" one version against the other. Candidate names, citation titles, and the sources themselves stay in their original language.

Where a campaign has published no materials in a given language, we provide our own faithful translation of the candidate's English content. In that case a direct quote is rendered as an accurate translation of the candidate's words rather than their verbatim original.

How to push back

These judgments are made by a human-supervised review against the sources cited on each candidate's page. If you think a stance misreads a candidate — especially a ±2 you'd grade as a ±1, or a missing source — the quote and link are right there on the candidate card so you can check our work and tell us where we got it wrong.