Function pick_winner 

pub(crate) fn pick_winner<'a>(
    reports: &'a [Report],
    reference: Option<&str>,
) -> Option<&'a Report>

The best candidate: highest accuracy among models whose top token was an offered label in every case, ties broken by median latency. Fidelity is a gate first — a model cerno cannot read is not a candidate at any accuracy — and the reference is excluded, since it is the yardstick rather than an option. A /v1/systemone row is never a candidate: it measured Ollama’s path, and the verdict names a model for cerno’s.