Accuracy checks
Each layer of the archive is scored against a sample checked by hand against the source pages: motions and votes, split votes and the ordinance pipeline, the payables lists and Treasurer's reports, project facts and the debt figures. The scores below are computed on the live data.
| Area | What is checked | Checked | Errors | Error rate |
|---|---|---|---|---|
| Votes | Member votes on 114 hand-checked motions wrong, missed or invented votes | 485 | 0 | 0.0% |
| Votes | Motion results (carried or failed) | 114 | 0 | 0.0% |
| Votes | Dissents (no votes and abstentions) missed or invented | 30 | 0 | 0.0% |
| Decisions | Split-vote rows (tally, who voted no, result) | 36 | 0 | 0.0% |
| Decisions | Ordinance pipeline rows (reading dates, status, days) | 41 | 0 | 0.0% |
| Spending | Payables line fields (25 of 28 sample lines found) lines recalled: 89% (scans with missing rows) | 45 | 0 | 0.0% |
| Projects | Pay application fields | 90 | 0 | 0.0% |
| Projects | Change order fields | 14 | 0 | 0.0% |
| Funds | Treasurer's report values (5 reports) | 60 | 0 | 0.0% |
| Funds | Summary of Receipts values | 10 | 0 | 0.0% |
| Projects | Status, contractor, contract amount, dates and parties of sampled projects | 52 | 0 | 0.0% |
| Debt | Issue facts and balances (15 issues) | 50 | 0 | 0.0% |
| Debt | City-wide totals and debt-limit figures | 8 | 0 | 0.0% |
What the samples do and do not show
The samples are small (tens of motions, a few hundred votes, a handful of reports per layer), chosen at random across eras where possible, and some were drawn from documents never used while writing the rules. A clean score means the rules work on documents like those; it does not prove every page in the archive was read correctly. Old, faint or handwritten scans are the most likely to be misread. Known gaps and the data checks that flag doubtful rows are on the coverage page.
If you find an error, please report it.
Data last updated October 15, 2026. Built from the archive's documents; see how the data is built.