Points to monthly karma
Layer 1 — raw points
Every action type in the catalog carries a point value, defined in exactly one place and published. The spread is wide on purpose:
| Action | Points |
|---|---|
| add a politician video | 5 |
| a mass communications post | 10 |
| add a politician channel | 15 |
| identify a challenger | 25 |
| flag a worrisome challenger | 25 |
| a canvass shift | 40 |
| recruit a citizen | 50 |
| vet a challenger | 100 |
| hold a meeting | 100 |
| a media appearance | 150 |
| publish an explainer | 200 |
| recruit a challenger | 250 |
| a press pickup | 300 |
| a challenger files | 500 |
| a challenger wins | 5,000 |
The ratios are the argument. Adding a video is real work and it is not within a thousand of putting a better person in a seat, and a scoring system that pretends otherwise will get you a lot of videos.
Actions are grouped into five tracks — challengers, communications, organizing, lawmaking, participation — so a citizen who does one kind of work well is visible as such rather than being buried under people who do a different kind.
Points are frozen at verification. If a catalog value changes later, already-verified work keeps what it was awarded. Nobody's past is silently repriced.
Layer 2 — the normalized monthly score
At month close, raw points are summarised into a normalized monthly karma score, and everything downstream uses that.
Normalization is what makes the number comparable — across citizens, across tracks, and across communities of very different sizes. A month of organizing in a community of thirty and a month of lawmaking in a community of three thousand should be readable side by side.
The month is sealed
Once closed, a month is sealed. It is not recomputed when the catalog changes or when somebody files something late. The record of what the movement recognised, when it recognised it, stays fixed — the same discipline the sealed editions of every other calculation follow.