Awards, Not Scores
Every figure this movement publishes about a politician is a percentage of an award. "A 72% Llama Award" describes something we conferred, and how much of it — not a place on a scale the person carries around. That distinction changes what is being claimed and what a blank means, and it is enforced by a test that refuses to build the software when the wrong word reaches a page.
What it does
Everything published about a politician is an award, a percentage of one, or the stated absence of one. The arithmetic is unchanged; what is fixed is the grammar the result may arrive in. Every published figure is joined to the award it is a percentage of. "72% Llama Award" is compliant. A bare 72 next to an animal is not. One shared function in the code joins the number to the award name, so no page assembles that phrase for itself and the wording cannot drift.
The display rules are fixed too, rather than left to whoever builds the next screen:
- The mark, the number and the word travel together. A mark alone is an epithet. A number alone reads as a place on a scale. Neither ships.
- The percentage is big, the award name small, above or beside it. The reader learns what is being conferred before reading how much.
- A blank renders as a sentence, never as 0%. "No award has been given yet" says what the movement has done, not what the person is. A zero nobody computed is a false claim with a numeral on it.
- The "N% known" figure travels with every percentage, so a confident number on thin evidence argues with itself on sight.
- All three marks render at one shared height, and no mark ever appears on a record the movement never assessed.
Enforcement is a test, not a habit. A permitted lexicon and a forbidden list are checked by a grep across every file that publishes an award about a politician. The forbidden side covers the report-card vocabulary, the words that place a person somewhere on a scale, and words like penalty and demerit that smuggle a verdict back in. A hit fails the build with a message naming what to write instead. Its scope is whole directories rather than a hand-kept list, so a page added next month is covered the day it exists.
Citizen-side figures deliberately keep the word score. Karma, your own trust score and the leaderboard are a citizen earning recognition for their own logged work, not a published claim about somebody else, and the check leaves them alone.
None of this is ever a reason to hold a number back. A 4% award publishes as 4%; a 91% award publishes as 91%, under the person's name, with its acts listed.
The goal it serves
The movement publishes unflattering arithmetic about named, living, powerful people who can afford lawyers. That work only survives if the form it arrives in is honest. An award is a plain act with a stated basis: it was given or it was not, it may have been given in part, and the derivation is published in full with its sources. Nothing is left over for anyone to defend — no implied claim about the human being. A bare number out of a hundred reads as a verdict instead, and invites exactly the argument the record was supposed to end. The doctrine is laid out in Award, not a score.
The second reason is what a blank has to mean. On a scale, no number reads as the bottom of the scale. With an award, not receiving one is an ordinary fact about most people most of the time — which is what lets the movement say "we assessed and gave nothing" and "we have not looked yet" as two different sentences, neither of them an accusation. The wall keeps those findings in separate lanes; this rule keeps them in honest words.
The third reason matters most. The aim is not to insult incumbents; it is to elect people who will work for citizens and refuse to be bought. A shelf of awards a challenger is partway to filling is an invitation to keep going; a report card is a reason to stay out of public life. The case for publishing any of it is in The Problem.
What keeps it honest
The limit written into the rule runs against the movement's own comfort. The vocabulary check may never gate a number: a failure means the copy is wrong, and the fix is the copy, never the removal of a published percentage. Adding an exclusion to a politician surface to make a build pass is a violation, not a workaround.
Two failures are held equally serious, in opposite directions. Withholding a number because it is uncomfortable is a defect. So is a surface that says "no award" where the truth is "we never ran the calculation" — that is a false exoneration, and the same severity of bug. A blank always carries a stated cause in words.
What the rule leaves alone: database columns, code symbols and already-published web addresses may keep older names. A link to a politician's award page that a citizen quoted in an argument must keep resolving to the same content, permanently. Breaking it would pull the arithmetic out from under a published claim.
Works with
- The Monkey Award — the damage finding from the voting record, published as a percentage of an award rather than a verdict on a person.
- The Llama Award — the repair finding, where a partly earned award is an invitation to earn more of it.
- The Flamingo Award — three states that only stay readable because "not given" is an ordinary fact.
- The Wall — keeps the two kinds of evidence in separate awards, which is what makes a percentage of one award mean anything.
- Show The Arithmetic — publishes the derivation behind every percentage, so the award can be checked rather than believed.
Where to go next
- Politicians — everything the movement publishes about a person in or seeking office, and the rules it publishes under.
- The wall — why the voting record and everything else stay in separate awards.
- Not assessed is not a verdict — "we found nothing" and "we have not looked" are two facts, never one.
- How It Works — the whole loop, from recording a position to publishing a finding.
- Take Action — what to do this week, now that you can read what the numbers claim.