How every score is built.
The same framework, applied to every governor, published before we scored anyone. If you think a score is wrong, this page tells you exactly how to check our work, and how to challenge it.
1. Where promises come from.
We track only documented commitments: the published manifesto, the inauguration address, and formal policy documents issued in the first 100 days. Campaign-trail remarks without documentation are excluded. If it isn't written down, it isn't scored.
2. How promises become checkable.
Each commitment is broken into discrete, verifiable units. "We will transform education" is not scoreable; "We will build 50 new classroom blocks" is. Vague promises are logged and labeled unverifiable as stated, visible, but excluded from the index. (Yes, this means vague manifestos produce shorter ledgers. That is itself information about the manifesto.)
3. Status definitions.
Kept: delivered substantially as promised, evidence on file. In progress: verifiable work underway, on a plausible track. Broken: deadline or term milestone passed without delivery. Abandoned: formally cancelled or contradicted by the administration's own actions.
4. How the index is calculated.
Kept = 1 point. In progress = 0.5, but only when backed by physical or budgetary evidence. Announcements alone score zero. Broken and abandoned = 0. The index is points earned over points possible at this stage of the term, so governors aren't punished early for long-horizon projects.
5. Evidence standards.
Primary sources outrank announcements: signed budgets, contract awards, FOI responses, geotagged-then-scrubbed site photography, court records. Press releases are claims, not evidence.
6. Citizen verdicts and what they can't do.
Verdicts never change the editorial score. They are displayed beside it, and they direct our verification efforts. The wall between the two is absolute. It's what keeps the index un-buyable and un-mobbable.
7. Corrections and challenges.
Anyone, including a governor's office, can challenge a status with evidence via the right of reply / corrections form. Upheld challenges change the score and are logged publicly in the changelog. Our mistakes stay visible after we fix them.
8. Who oversees this.
The Editorial Review Board audits scoring quarterly and rules on contested statuses. Board members are named. Scorers apply the framework; the board enforces it.
See the Editorial Review Board.