How the comparison works
Successful governance votes are extracted from transaction messages and checked against emitted vote events. The comparison includes historically registered operator addresses after creation; votes from other addresses are excluded. Historical bonded status has not been reconstructed. Weighted votes are compared as complete allocation vectors. Failed transactions are excluded.
Each pair is compared only on proposals where both voted. First and last recorded ballots are available separately. All delays use UTC block inclusion time, not signing time. Simultaneous inclusion does not establish which person acted first.
Validator addresses survive renames and possible changes of operator. Current names, genesis names and dated name changes are preserved separately. A new name is not evidence of a new owner; an unchanged address is not evidence of unchanged ownership.
The analysis also checks fixed delays, changing lead order, yearly and quarterly patterns, common execution wallets and multi-validator transactions. Proposal-specific timing and choice baselines, exploratory shuffles, and complete exceptions are available in the analytical outputs.