About TaterScout

All data comes straight from the official FIRST Tech Challenge Events API. Matches, scores, rankings, alliances and awards are ingested from FIRST; EPA and OPR are computed by TaterScout.

OPR is a least-squares estimate of how many points a team contributes to its alliance, solved over qualification matches and split into Auto (autonomous) and TeleOp(driver-controlled). TaterScout solves OPR from the raw FIRST match scores: the world rankings leaderboard uses each team's best single-event value, each event page shows that event's value, and the trajectory chart computes OPR after every match.

EPAis a point-unit Elo derivative: a team's rating is its predicted point contribution, updated after every match by the gap between actual and predicted scores. A k-factor shrinks as a team plays more matches, and a margin parameter shifts the model from pure offense early in the season toward full win-margin later. Auto EPA is tracked separately and TeleOp EPA is the remainder. TaterScout replays the entire season's matches in order, following the Statbotics EPA model (adapted from FRC's 3-team alliances to FTC's 2-team format). Unlike OPR, EPA accounts for win margin and schedule strength. Event pages are time-aware: they show each team's EPA as of the end of that event, and predictions use a team's rating enteringthe event, so there's no hindsight.

Every event page runs thousands of Monte-Carlo simulations — generating schedules, tabulating ranking points, selecting alliances and playing out the bracket — to estimate each team's win probability, predicted seed and make-playoffs odds. Unplayed matches show a per-match win %. It adapts Statbotics' approach to FTC's 2v2 format.

How lucky was a team's qualification draw? TaterScout compares the real schedule against thousands of random ones to produce three percentiles — Δ RP, Δ Rank and Δ EPA — and a composite, where higher means a harder draw. Toggle between pre- and post-event ratings.

The ↻ refresh button in the header pulls the latest results straight from the FIRST API. It only fetches the events that changed since the last sync and recomputes the affected ratings — so a match that just finished shows up in seconds.