Every column here is an alliance figure from official results. Three robots share one score, so a team carried by two strong partners reads identically to the team carrying them. Robot‑level performance comes from scouting reports, in the last column — that is the question this table cannot answer and the reason Recon collects data at all.

Evidence, not a verdict. A robot with three shooter issues this event is shown as exactly that — three reports. Recon does not conclude “their shooter is unreliable”; that is a call a human makes with the thread in front of them.

Anyone can report. Changing an issue’s status needs pit_scout, scout_lead, mentor or admin — enforced by the database, not by this page.

Discovery produces candidates, never truth. Nothing here is visible to a scout until an admin approves it, and approving is a separate act from making an image canonical — one tap must not silently do two things. A wrong robot photo in prematch is worse than no photo, because the scout will trust it.

Rejections are kept. Without that the next discovery run re-offers the same wrong photograph, forever.

Roles are granted here and nowhere else. Nobody picks their own at sign-up and nobody can promote themselves — the database refuses it, not this page. Every change below is written to an audit trail with your name on it.

A join code is a bearer token. Whoever has the string gets in, so every code grants scout and nothing more — there is no setting for that, by design. Promote somebody afterwards, by name. Codes expire and can be revoked, and revoking one does not remove anyone who already joined.