🎥 Video

A video of the candidate speaking, and every video says who made it, most independent first: Independent (news outlets and nonpartisan series like the League of Women Voters), Civic forum (community groups that invited the candidates), Government (public meetings, where only officeholders and people who spoke appear) and From the campaign (the candidate's own video, from their website or channel). Forums start where that candidate speaks. Debates with everyone in the race show once, above the candidates, under “Watch them together”. Videos load from YouTube's privacy-enhanced player only when you press play.

🚩 Flags

A flag is a fact from a source, found by a written rule, never our own reading of someone. Red flags are judgments against a candidate by a body whose job is to judge that; yellow flags are things worth asking about. Each rule is applied to everyone its source covers, and the card lists which checks ran, so “none found” means none found by those checks, not a clean bill of health.

• Red: Public discipline on their State Bar record. Covers candidates who are lawyers: everyone running for judge or Attorney General, and anyone whose ballot designation says they practice law (California).
• Yellow: Fined by the state's ethics and campaign-finance commission. Covers every candidate, searched by name in the Fair Political Practices Commission's monthly enforcement releases since January 2021 (California).
• Red: Rated "Not Qualified" by the county bar association. Covers judicial candidates the Los Angeles County Bar Association (California) or the King County Bar Association (Washington) rates.
• Yellow: Didn't answer a news outlet's candidate survey. Covers candidates LAist sent its survey to (LACCD trustees).

Each card's “What we checked” lists every check that covered the candidate and what it found, with links. Lawyers who share their name with several others on the State Bar's rolls are marked as checked but unmatched: we won't guess which record is theirs. An ethics fine shared by everyone named in a case (a campaign committee, its treasurer and the candidate) says so.

We don't flag a missing candidate statement: in California, candidates pay to print one, so that would measure money, not candor.

💩 The 💩 meter

How full of it a candidate has been, by the fact-checkers' count: the share of the candidate's claims that PolitiFact checked and rated Mostly False, False or Pants on Fire, across every PolitiFact ruling on them (not only this campaign). Numbers are copied from the candidate's PolitiFact page on the date shown.

💩 % = (Mostly False + False + Pants on Fire) ÷ all PolitiFact rulings × 100, rounded

Each card shows the formula with that candidate's own numbers, the year of their newest ruling, and a link to their PolitiFact page. We find people in PolitiFact's own list of everyone it has rated, and only use a page when its description names the candidate's state (so a same-named actor doesn't count).

• Fact-checkers choose which claims to check, so this is their record of a person, not a measure of everything they've said.
• The meter only appears in races where someone has been fact-checked; in a race nobody has, there's nothing to show. Below a percentage, the icon shows the raw count (“1/5”: 1 of 5 rulings Mostly False or worse).
• Fewer than 5 rulings gets no percentage (“too few to score”), and no rulings is never shown as 0%.
• A race only shows percentages when every candidate in it has at least 5 rulings. Otherwise each card shows the raw counts, so one candidate's number never sits next to another's blank.
• News fact-checks that don't use a rating scale (like CapRadio's check of the governor's debate) are listed under the meter as “claims fact-checked”: the claim and the outlet's verdict, both quoted word for word. We never turn a verdict into a rating, so they don't count toward the %.

Stakes on ballot measures

A long ballot is a lot to get through, so the Expanded view lists each section's measures with the biggest stakes first (turn that off under Filters). The grade never looks at a measure's topic or which side is right. It adds up:

• Money, 0 to 3 points: the official yearly estimate. 3 for $1 billion or more, 2 for $100 million or more, 1 for $1 million or more, 0 for less or for no dollar figure. We quote the official words beside the grade.
• Hard to undo, 2 points: it changes a constitution or a county charter.
• Published guides disagree, 1 point: the guides we use land on both Yes and No.
• Local vote, 1 point: fewer people vote on it, so each vote counts for more.

4 or more points, or $1 billion or more on its own, is High; 2 or 3 is Medium, and under 2 is Lower. Lower doesn't mean unimportant: a measure with no dollar figure can still matter a great deal to you. Retention questions aren't graded, since each is about one judge.

🏅 Badges

Who recommends the candidate (or Yes or No on a measure): party slates, newspaper editorial boards and the League of Women Voters, each pick checked against a quote on the page it was published on. Party and advocacy guides only appear as left/right pairs, so no side is shown alone. Also here: the county bar association's rating for judges, key endorsements as a news voter guide lists them, and the endorsements on the campaign's own site. Hover a badge for the quote; click it for the source.