Qualifier Match Randomness

At the FMA Seneca event this past weekend my team, 2720, qualification match schedule had me wondering about how the random qualification matches are made as well as how random they are. And does this randomness help make sure competitions are fair? Due to the pandemic, I am rather inexperienced in this subject but from what I have seen in the past, events that have around 35 teams in attendance, teams can expect to go against another team twice at the maximum during the qualifiers.

At Sencea it felt like we were playing the same teams again and again because, in reality, we kind of were. We played against team 203 SOUPer Bots, 1807 Redbird Robotics, and 5895 Peddie Robotics all 3 times each. Is this normal for a competition or were the random alliances truly random and we just got an odd schedule?


You’re in luck, there’s literature about it!


The FMS runs the matchmaker, and as I understand, it is a big deal and HQ has to be involved (and doesn’t like it) if an event decides to run matchmaker again after it is run the first time. This makes sense, to avoid accusations of bias from the FTA/scorekeeper volunteers.


It didn’t used to be that way (but definitely is now). Back in the good ol’ days, I could go up to the score keeper with a list of teams that are really struggling to pass inspection and say “make sure these team’s aren’t in the first couple of matches”. They’d run it a few times and it would buy is an extra half hour or so in the morning to get them passed so they can get on the field.

Now, they just generate the schedule, review the output to ensure the metrics look good (there’s a status screen that shows things like min/max turnaround time, pairing with/against the same teams, etc), and then hit print. And 75%+ of the events I’ve been to (60+ teams before this year, 55+ this year), the last team to pass inspection was in the first 2 matches, and half that time I’m walking them out onto the field as opening ceremonies conclude.

I really wish we could get that sort of input into the schedule again. Let me tell the score keeper the 3 teams that will struggle to pass inspection before opening ceremonies, and give them a place to put those numbers in so the process can assign them in later matches in the first “round” of matches!


If you really want to dig into the math involved, you may want to check out Caleb Sykes’s analysis:

1 Like

The TL;DR of this, which applies to your situation, is that the algorithm that does the scheduling heavily penalizes having the same partner in multiple games, while only moderately penalizing having the same opponent in multiple games. The penalty is roughly 3x for having the same partner compared to the same opponent, when the numerical ratio of opponents to partners is only 1.5x.


This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.