Season
Power ratings
| Event | CCWM | OPR | DPR | Alliances | Fit |
|---|---|---|---|---|---|
| 2026长三角VEX机器人精英赛: MS，Override | 13.5 | 30.2 | 16.7 | 9 | R² 0.327 · ±37.34 pts |
| 2026-2027 VEX机器人竞赛华东大区赛: VEX V5 Robotics Competition | 25.1 | 51.3 | 26.3 | 9 | R² 0.314 · ±33.97 pts |
OPR is what this team contributed to its own alliance's score, DPR what it allowed the opposing alliance, and CCWM the difference. All three are solved across every alliance at that event at once, so they do not depend on who the team was paired with — which win rate and average score both do.
Read them as rankings, not quantities. The fit column is honest about the precision: these explain roughly a third of score variance, and each team's figure rests on about five alliances. DPR can even come out slightly negative, which is the least-squares fit overshooting rather than a team causing negative points.
A team playing pure defense suppresses opponents instead of scoring, so it shows as low OPR and low DPR. That is why CCWM is the column to lead with — at the Mall of America event the undefeated winner ranked 6th on OPR and 1st on CCWM.
Allies
| Team | Name | Together | Eliminations | Record | Win rate |
|---|---|---|---|---|---|
| 10090A | OVERCLOCK | 1 | 1 | 0–1–0 | 0% |
| 2052D | shine | 1 | 1 | 0–1–0 | 0% |
| 10017Z | 中国杭州低碳科技馆 | 1 | — | 0–1–0 | 0% |
| 12731M | 小手宇宙机器人 | 1 | — | 1–0–0 | 100% |
| 12778K | 全域接管队 | 1 | — | 0–1–0 | 0% |
| 18960B | 18960B | 1 | — | 0–0–1 | 0% |
| 1996T | Wuyi Primary School | 1 | — | 0–1–0 | 0% |
| 2052A | 2052A | 1 | — | 1–0–0 | 100% |
| 21131W | iLogicbots | 1 | — | 0–1–0 | 0% |
| 222S | 222S | 1 | — | 1–0–0 | 100% |
| 2832A | Bad Robot Team 2832A | 1 | — | 0–1–0 | 0% |
| 29588Y | Spanner AI | 1 | — | 0–1–0 | 0% |
Top 12 of 18 distinct allies.
Sorted by elimination partnerships first, then by how often they played together. That order is the point: a qualification partner is a random draw, so playing with someone repeatedly says nothing about either team. An elimination partner was chosen in alliance selection — so even one says something, because it is either who picked them or who they picked.
The record is the pair's record in the matches they played together, which is the only record that describes the pairing rather than either team on its own.
Matches
Robot analysis
Nothing here yet, and the reason is worth stating plainly rather than filling the space with numbers that do not mean what they appear to.
Tracking produces four anonymous tracks per match — robot 0 through robot 3 — and nothing in the video says which one is 3658A. Team plates measure about 17 pixels tall in this footage, so reading them is out of reach. Attribution has to come from the match schedule plus each robot's starting tile, which is not built.
Until it is, per-robot measurements — distance travelled, time in each zone, contact events, build-style clustering — live on the match pages, unattributed. They are deliberately not rolled up here, because a season average over tracks that may belong to the opposing alliance would be worse than no average at all.
Robot videos
Not searched yet. The daily search reaches every team in turn, or run python -m integrations.youtube --team 3658A now.
Match footage
No footage linked to this team's matches yet. A match page takes a stream link and the time the match starts in it.