INITIAL
—wrong majority
—non-unanimous
When more votes help, and unanimity fades.
Model majority error under explicit independence assumptions. This is a probability calculator, not a GenLayer protocol replica.
Choose assumptions, then calculate the binomial majority risk.
Panel
comparison
Run scenario
APPEAL
—wrong majority
—non-unanimous
Why risk falls: under independence, more than half the panel must be wrong together.
Why disagreement rises: with more validators, at least one different vote becomes more likely even when the majority improves.
The independence assumption is the whole story.
Identical independent correctness is a model, not observed network evidence. Correlated validators, shared training data, adversarial behavior, abstentions, weighted votes, changing evidence, appeal selection, and costs can dominate the binomial result. This calculator does not reproduce or forecast GenLayer.