Julia Wan - Week #12 - Offensive Power Rating

Offensive power rating (OPR) is a metric used in First Tech Challenge (FTC), a global robotics tournament that I participate in, to determine a team’s average individual contribution to their alliance in matches at a tournament. 

First, let’s discuss why Offensive Power Rating is needed in the competition. In FTC, matches are played in a 2v2 format, and the “winning alliance” is the alliance of two teams that scores the most points in a match. In the competition playoffs, the top-ranked teams (ranked by number of total wins/losses in random 2v2 matchups throughout the qualification matches) get to choose their alliance partners to go into elimination rounds. This alliance selection is crucial in determining which teams will win the tournament and advance into higher competitions, so teams throughout the day scout the performance of other teams and keep track of which teams will work best with their robot and game strategy.

Because scores are recorded as alliance totals, it’s difficult to estimate a team’s individual scoring ability based on scores because they’re dependent on the strength of the team’s alliance partner. Simply using a team’s average match score across a tournament’s qualifying matches wasn’t an accurate metric for scouting. Therefore, in 2004, many scouting websites began ranking teams with a metric known as Offensive Power Rating (OPR) which approximates a team’s individual contributions to an alliance. 

The OPR metric is calculated using an algebraic system of equations. For example, if team A and team B score a total of 10 points when they alliance, team A and team C score a total of 13 points, and team B and team C score a total of 7 points, it can be algebraically approximated that team A has an OPR of 8 points, B has an OPR of 2 points, and C has an OPR of 5 points. With more matches and more teams involved, not every system of equations will have an exact solution, but OPR is approximated to be the solution that yields the smallest amount of error.

OPR often doesn’t tell the full story about the performance of a team. If a team’s robot breaks during the first few matches of the day but fixes it and maintains high performance throughout the rest of the tournament, their OPR will be low since it only gives the consistent performance of a team throughout a tournament. Furthermore, if one team performs poorly in all their matches except one where they do exceptionally well and Team B does average in all their matches, Team B will receive a higher boost in OPR than Team A does from that match, even if Team A contributed more to the match. However, OPR has made it much easier to approximate a team’s performance/consistency and rank the strength of teams on scouting websites such as ftcscout.org, making it possible to give teams state ranks and global ranks.

(original screenshot from ftcscout.org showing the NorCal team rank by OPR)


Comments

  1. Hey Julia! This is such a solid breakdown of OPR and why it matters in FTC. My team also uses OPR for scouting, like most, and it’s definitely helpful. Yet we know it doesn’t always tell the full story. A team could have an amazing robot but get unlucky with alliance partners or have one bad match that hurts their ranking. It’s kind of frustrating how inconsistent it can be, but at the same time, it’s one of the best tools we have for quickly evaluating teams. It’s truly interesting how alliance selections go beyond just picking the teams with the highest OPR and is such a complex process!

    ReplyDelete
  2. Hey Julia, when I first read your title I thought it said "offensive power rankings," and thought to myself "Dang... I didn't know Julia was into the NFL like that..." As I opened and started to read your blog, I realized that your OPR was about robotics and I was about to click out of the blog before I decided to continue reading and see how similar our two OPRs are to each other. Turns out they're actually really similar. In the NFL, OPR means the rankings of each team's offensive strength, meaning how good each team is compared to each other in terms of when they have the ball in possession (as opposed to DPR which is when their opponents have the ball). The NFL, similar to FTC, grades the OPR in terms of points based on how many touchdowns and field goals teams score with the ball (similar to robotic scoring). I never knew FTC had such a sporty-statistical vibe to it, the competition definitely seems a lot more interesting than I thought.

    ReplyDelete

Post a Comment

Popular posts from this blog

Neev - Week 16: Remembering Junior Year

Pranatee Shah Week 16 - Reflection of the Year

Arushi Naidu - Week 9: Power in Positivity through Adversity