Thoughts on the new elimination rounds system?


I can see both points of view here, but I do have to agree overall. The math checks out in terms of it not affecting that many people, but it doesn’t change the fact that it’s completely unfair to the teams that have dedicated their whole season to one style of play, whereas blanket-enforcing this next season will affect all teams equally. Imo that’s a lot more fair.

I’d like to add on that the small percent being affected by this change just so happens to be the most invested, and the ones that end up in elims consistently throughout the season anyway, so you wouldn’t be damaging as many as you think if you were to try it for next season. Also a lot more is on the line at worlds than local events, although states is also at risk here.

I don’t understand why a bunch of you are suggesting that the outcome of high-level matches are significantly influenced by luck, when the top seeded teams for these events are determined by perfect or near perfect match records. If the 1st seed’s 10-0 robot was going to fail, it likely would have already.

Couldn’t you just take and apply this logic to qualification rounds and argue that teams are unfairly losing their opportunities to win the event because those aren’t best of three? Isn’t there 10 times more luck involved in being 1st seed and being able to pick the best robot in the division? I’m not buying this concept that these events will be a weaker evaluation of team performance. I think there’s less room for error, which suggests the opposite in my opinion.

I’m still confused on what made the GDC want to change this. Like every single robotics competition other then VEX does the regular 2 wins to advance way. All of this for 8 teams? Really? Getting to the elimination rounds is meant to be tough and an honor, and changing the entire elimination system for 8 more teams (who will in all likelihood get eliminated their first match) into eliminations? Totally uncalled for. This out of all things would have been the one to ask for the community’s approval of. Even if I was one of those 8 teams I would still feel horrible if a higher seeded alliance gets eliminated due to their VEXNet keys disconnecting or something. Please reconsider not making this change in a way it will only effect the TOP teams in the world, and instead letting this apply to the qualification events next season, so teams who don’t put in as much effort will also be effected.


However, the best teams on alliances at worlds are seldom the alliance captains. This is mostly about the amazing first pick, that may have had a hard match schedule, but deserves to win worlds (think 2915A from sack attack). To say that the best teams never fail is just wrong.

10 matches is high enough of a sample size for good teams to do well.
We don’t complain about qual matches because you can still bounce back and do well in elims due to the fact that people can pick you, so your qual record doesn’t decide your fate completely.
However, in elims, that is where your task is to prove yourself. It is also where you face your greatest challenges, and the closer the games, the more influence random events or small mistakes have, making players much more frustrated.
For the best player to become champion, they cannot make a mistake for 10 straight matches against good teams (4 in division, 5 in RR, 1 in finals)

This is not at all what we’re saying. If a team’s robot was improperly functioning due to their own fault, then that’s their problem, and yes, it would have malfunctioned already. What we’re referring to are things that are out of an individual team’s control that could potentially cost them a match, and if it was a single elimination match, the entire tournament. If VEXnet keys disconnect, or static resets encoders, or the field was improperly set up, and the first seed team was affected, their entire day is now over because of something they had absolutely no control over. If it were best 2 of 3, they would at least have another chance or two to prove that their robot can still come out on top, but now with single eliminations they’ll just have to either pack their bags or watch the alliance that beat them on pure luck lose in the subsequent round to another alliance that was also clearly better than them.

I think you’re evaluating team quality based on solely the capability of a robot here. There’s more to a team than what their robot can potentially do. 2915A was having a tough time dealing with coordinating with weaker partners. If I recall correctly, they weren’t even having any technical issues. My team (the alliance captain) was able to coordinate better with all our partners and ended up as the first seed (after quite a few narrow but calculated wins) due to that, and picked them. They had a significantly better performing robot than us, but to say that we weren’t a better team overall is pretty short sighted.

You can say that it was their inability to coordinate with weaker partners, but that can only go so far. At some point, an alliance can be so far under that even the most cohesive match in which absolutely no mistakes were made and the alliance played every single thing perfectly could still lose if the opposing alliance has robots that are that much better.

Yes, dealing with lesser alliance partners is part of the game, but you can’t just say that a team is inferior yours “overall” just because you were able to coordinate with your teammates better. Had 2915A played all of their matches with your teammates and your team with theirs, are you saying you still would’ve come out with a better record than them?

This is suggesting that the referees won’t be considerate of situations outside of the control of the teams. I think that they’re going to be more inclined to replay matches that they think may be out of a team’s control. We [referees] have to make a lot of judgement calls about things like disconnects or other issues, and generally speaking, I think that I’d err on the side of replaying a lot more matches if they weren’t out of threes (I head ref a lot of events in my region, including our regional). I wouldn’t be too concerned about this.

Yes I am. I’ve had a lot of discussions about this specific topic, and I believe that our schedules were close enough in difficulty that it wouldn’t have mattered if we swapped. In my opinion, the outcome would have been that I would have been higher seeded vs them still. If you know what TRSPs are (they’re a synthetic measure of schedule difficulty), they only had slightly more than us (and only a five more objects worth of SPs). And for the record, my team had the highest TRSPs out of all of the first seeds of the five divisions (more than double in two cases).

I don’t think analyzing past competitions is going to provide valuable insights. However, I do believe that we need to show that the single elimination system significantly worsens the competitive experience at Worlds. Ultimately, this change has caused a lot of backlash among teams, and it would be very unreasonable for it to actually be implemented with such short notice at Worlds.

It has been decided.

I would hate to see teams like 8675A, 1970K, 202Z and 929U (Just to name a few, there are several more that I could list) who have put countless hours making some of the best robots in the world be stripped of a hard seasons work and the title of world champion because of a poor connection with VexNet in QF1. I would hate to see that the teams who can’t afford to buy new parts suffer because their 3-year-old motors burn out and the teams who can, win, undeservingly. This season has seen some of the highest level robots and highest level teams. Everyone will be playing at the top of their game. Thus something like vexnet disconnecting for even a second or a motor controller burning out will cost them the match. but it doesn’t matter. The best teams would adapt to the change and fight their way to the finals. the best teams will come out on top. But now if vexnet disconnects or a motor controller burns out, the top team is out. For good. We all make mistakes but it’s the ability to be given a second chance and come back and prove that you are the better team that makes the saying “worlds level” mean best of the best. If you lose twice, then you are not the best team. Losing once says very little about the team, especially with the best of the best competing. No accurate data has ever been obtained from doing an experiment only once. The compromise to adding more teams to compete and creating sudden death elimination rounds is that the quality of the event will decrease. There will always be luck as a factor and the best teams will always control as much of their robot as possible, but the times that bad luck slips through, there will be no compensation of another match to prove who is the best. Some of best teams may be sitting on the sidelines this worlds. I do hope this is not the case. Good luck to all teams.

  • Evan

The whole goal here is to push the GDC to reconsider, although, yes, it is unlikely that they will roll back this part of the manual update.

I think this is a good decision. While no one can say for sure what will happen, and how it will turn out, progress can’t be made by simply doing nothing. I think to improve on the situation in which @Karthik discussed earlier in this thread, it is important to at least attempt to do something about the situation than to do nothing, and let the problem persist. I think that testing this out at Worlds is the best decision simply because it gives them a decent amount of teams to test it on, while not inconveniencing too many teams, and not testing this system with too few teams. I can understand that this will be an inconvenience to some at the World Championship, but like @Paul Copioli said, only 32% of teams at Worlds participate in eliminations, so it is only inconveniencing a small portion of teams. No pain, no gain! Also, I think that this will force teams to make a more consistent robot that has fewer problems during competitions, which will be a good thing. Also, with the soon to be released V5, I think we will possibly see a decrease in problems arising during matches, due to better batteries, remotes, radios, and microcontrollers. I believe that now isn’t the best time to test this due to short notice and the upcoming V5 changes. I think that the 2019 or 2020 VEX Worlds is the best place to test out this new system and give it a shot, rather than dismissing it completely because of the negative aspects. Think about the movie Moneyball with Brad Pitt. He created an amazing team out of “flawed” players that actually gave them the advantage. While something can seem like a bad idea simply because of the negative aspects, it is possible that the positive aspects will outweigh the negative. Food for thought.

Well, to be fair to the GDC, the primary goal here is to spark discussion, get all of our facts, opinions, and assertions out, and converse with them on their decisions so that we can all get a more complete understanding of the situation at hand. After, we can make a collective decision as to whether or not the GDC should reconsider. I’m sure many teams already want them to reconsider, but let’s all take some more time than just a few hours to process all of this information before making any irrational decisions.

Well said!

I hope that everyone’s ready for some statistics so we can back up some claims about BO1 vs BO3, because here we go!

Edit: A “reverse sweep” here would be the case where someone wins the last two matches after losing the first.

Edit 2: I forgot to include divisions before, so have re-generated the numbers. All apart from one of the percentages went up though.

(These numbers were obtained by analysing all elim matches in VexDB)

Statistics for 1-team alliances

Quarter-Finals Matches

  • 498 total matches
  • 47 total reverse sweeps
    • 27 by red
    • 20 by blue
  • 9.44% of matches are reverse sweeps

Semi-Finals Matches

  • 337 total matches
  • 41 total reverse sweeps
    • 19 by red
    • 22 by blue
  • 12.17% of matches are reverse sweeps

Finals Matches

  • 175 total matches
  • 31 total reverse sweeps
    • 15 by red
    • 16 by blue
  • 17.71% of matches are reverse sweeps

Statistics for 2-team alliances

Quarter-Finals Matches

  • 3005 total matches
  • 268 total reverse sweeps
    • 153 by red
    • 115 by blue
  • 8.92% of matches are reverse sweeps

Semi-Finals Matches

  • 1936 total matches
  • 208 total reverse sweeps
    • 131 by red
    • 77 by blue
  • 10.74% of matches are reverse sweeps

Finals Matches

  • 979 total matches
  • 125 total reverse sweeps
    • 71 by red
    • 54 by blue
  • 12.77% of matches are reverse sweeps

Statistics for 3-team alliances

Quarter-Finals Matches

  • 10154 total matches
  • 1136 total reverse sweeps
    • 713 by red
    • 423 by blue
  • 11.19% of matches are reverse sweeps

Semi-Finals Matches

  • 5567 total matches
  • 760 total reverse sweeps
    • 452 by red
    • 308 by blue
  • 13.65% of matches are reverse sweeps

Finals Matches

  • 2837 total matches
  • 469 total reverse sweeps
    • 267 by red
    • 202 by blue
  • 16.53% of matches are reverse sweeps

Please Vex… Don’t test this out on Worlds, you’re going to ■■■■ off 90% of your most invested customers. Giving teams a whole season to adapt would seem much better as many teams now would have to change up their game and scouting strategy that they have been working for months on. I’m not saying there will be a correlation but Snapchat’s stocks are looking pretty bad after they kept an update that didn’t run well the the community.