Best of One is Not Needed for VEX U

  1. 5 weeks ago

    Nathan Rossi

    May 14 The Internet PYRO & 8800X
    Edited 5 weeks ago by Nathan Rossi

    Before the 2018-19 manual for VEX Turning Point becomes official on August 17th, I propose to revert back to the best of 3 system for VEX U.

    Based on the comments made by members of the GDC in this thread and in this video , it can be derived the reasoning behind switching to best of one was made primarily to eliminate the third member of a highschool/middle school alliance. This new system with two alliance partners effectively eliminates the primary incentive for match-throwing. However, by eliminating the third member, more alliances and a larger bracket were needed to allow a sufficient number of teams to play in the elimination rounds. By having a larger elimination bracket, it had the added benefit of allowing 8 more teams to play in the elimination rounds (in both highschool and VEX U).

    While this reasoning makes sense in the context of the highschool and middle school program, it does not hold water for VEX U. The first obvious point is there are no alliance partners in VEX U. There is no incentive to throw matches in order to be the “lucky” third member of a high seed. Throwing matches in VEX U only does harm to the team throwing matches. If the primary reason for switching to best of one was to help eliminate match throwing, the problem is almost non-existent in VEX U.

    Where there is a discussion to be held is around the number of teams playing in the elimination rounds.

    Of the 14 tournaments that ran qualification matches last season (excluding worlds), only 2 had 16 or more teams . Meaning only 12.5% of tournaments would have seen an increase in the number of teams playing elimination rounds as per the new rules.

    While I do not have the projected growth for the number of VEX U teams this upcoming season, I speculate the number of VEX U teams will not increase much (or there could even be a small decline). I speculate this because of the new rule changes to allow 2 robots and requiring teams to use the V5 system. While these are great changes that will benefit the program in the long-term, I suspect there will be first and second year jitters with teams that are unable to compete due the cost increase. Anecdotally, I know of at least two teams in our area that most likely will not continue due to the changes (allowing two trade-ins per VEX U team would really help reduce the cost barrier!).

    Last season, there were 6 tournaments that had 12 or more teams. Because VEX U growth is fairly slow (there are only so many universities) and the previously mentioned cost increase could discourage growth, these tournaments will likely only see about 16 teams attending their tournaments in the near-future. If registration for these events does reach 16 teams on the nose plus one or two more, this would mean close to 100% of the teams at these tournaments would participate in the elimination rounds. If this is the case, what it the point of playing qualification matches? To find the one or two teams that don’t get to play? Qualification matches in VEX U are already fairly low stakes (excluding the world championship) as you don’t need a great win loss record to compete in the elimination rounds. One’s exact rank doesn’t matter because there is no real benefit to being the #1 seed vs the #9 seed. Allowing 16 teams to compete in the elimination rounds would make this problem even worse and make qualification matches essentially pointless. You would only need about 2 wins at most events to still make it to the elimination rounds.

    The one tournament where having 16 teams play in elims makes the most sense is at the world championship. It was pointed out in this thread before the change to best of 1 that the increase to 92 teams from 62 meant a smaller percentage of teams attending would play in the elimination rounds. The important figure from this thread is the RECF aims to “have the top 20-25% of teams (within a division) compete during the elimination matches.” It should be noted that if the elimination system used in 2017 that had the top 10 teams compete in elims was used in a division with 46 teams (2 divisions, 92 team event), 22% of teams in their division would compete in the elimination rounds, falling within the acceptable range laid out by the RECF.

    I don’t see VEX U divisions becoming large enough to fall outside this 20-25% range at the world championship for a couple of reasons. The 92 team capacity was much larger than it needed to be. Before teams outside of the top 5 world skills ranking were invited to the world championship (meaning everyone that earned a spot through traditional means), the VEX U team registration was at approximately 46 teams (I unfortunately do not have a screenshot from the time to support this, but it was around this number). There is still a lot of room for VEX U to grow as a program and fill these spots with teams that earned them through traditional means (like highschool). There is no need to increase the team capacity for VEX U, if anything, it should be reduced (though, I wouldn’t advocate for that).

    If the number of teams per division is the same for 2019 but only the top 10 compete in the elimination rounds, there would be a drop from 35% of VEX U teams competing in elims to 22%. Is this drop acceptable? Again, it falls within the RECF’s accepted range of teams participating. I personally would much rather have the 2017 VEX U system for the elimination rounds than the current system. Though, I am curious about what other VEX U competitors have to say about this.

    In summary, because VEX U does not have alliance partners or an appropriate number of teams, the justification for changing to best of 1 does not hold water. In the future, best of one would devalue qualification matches to the point of being worthless at all but one tournament (the world championship). The solution is to bring back the best of 3 system with 8 teams. A modified bracket like the one used at the 2017 world championship could be utilized for the 2019 world championship and future world championships.

    Thank you for your consideration.
    P.S. Turning Point looks like it's going to be an awesome game! I'm a big supporter of the autonomous line

  2. Complexist

    May 14 My Basement, Michigan BNS / Former 69
    Edited 5 weeks ago by Complexist

    Best of one has actually created an incentive to throw matches in VexU. Because the new format doubled the number of teams in elims, high ranking teams can afford to lose matches intentionally so that they place where their bracket will be easiest.

  3. nenik

    May 14 V5 Beta Tester

    Isn't "easiest" the first seed? Pitched against #16, then #8 or #9 and only semis start to be more interesting.
    At worst, you'll meet #4 in semis, but thanks to the Bo1, you could also breeze through #12 or #13....

  4. OscarMNOVA12

    May 14 Event Partner, V5 Beta Tester Northern VA and Pittsburgh 12
    Edited 5 weeks ago by OscarMNOVA12

    The original vex u format where round of 16 was best of 1 worked well. I'd even say doing best of 1 in quarters is fine.

    After seeing what happened to XD and my friends on MASON, I just wish semis and finals were best of 3.

  5. 9065_Parker

    May 14 Platte City Missouri 9065C

    @OscarMNOVA12 The original vex u format where round of 16 was best of 1 worked well. I'd even say doing best of 1 in quarters is fine.

    After seeing what happened to XD and my friends on MASON, I just wish semis and finals were best of 3.

    The problem that they're suggesting is that most tournaments don't have room for 16 teams.

  6. Nathan Rossi

    May 15 The Internet PYRO & 8800X
    Edited 5 weeks ago by Nathan Rossi

    @nenik Isn't "easiest" the first seed? Pitched against #16, then #8 or #9 and only semis start to be more interesting.
    At worst, you'll meet #4 in semis, but thanks to the Bo1, you could also breeze through #12 or #13....

    Theoretically yes, the #1 seed should have the easiest path. Though, I'm sure you know as well as I that rankings more times than not are not perfectly accurate. For the past two qualifying tournaments, we were the #1 seed and our hardest matches were in the semi-finals, not finals. In the Innovate division this year, the #1 seed was eliminated by the #16 seed. You can find endless examples of inaccurate rankings.

    My comment of "one’s exact rank doesn’t matter because there is no real benefit to being the #1 seed vs the #9 seed" primarily refers to the fact there is no alliance selection and thus no incentive to rank highest and get the best pick. If you can make the cutoff, that's all that really matters. Will you face the most difficult team in quarter-finals or finals? It's entirely determined by luck of the schedule.

    @OscarMNOVA12 The original vex u format where round of 16 was best of 1 worked well. I'd even say doing best of 1 in quarters is fine.

    After seeing what happened to XD and my friends on MASON, I just wish semis and finals were best of 3.

    I'm curious what you mean by "worked well." Why should a team like XJTU1 who lost connection for the entire match for an unknown reason in the round 16 not get to play best of 3? They were a very capable team who could have won the division if best of 3 was in place.

  7. 4 weeks ago

    224x

    May 16 Malibu High School, California 224X

    The best of one makes for a better build quality, it means you can't afford your robot going wrong. Vex U should have had this rule for a while BECAUSE it doesn't have alliances, it's always the same robots going against each other, if all robots work, why play the same match twice?

  8. OscarMNOVA12

    May 16 Event Partner, V5 Beta Tester Northern VA and Pittsburgh 12

    In previous years of VEX U, the early rounds of elims were best of 1. Once you got to quarterfinals, it was all best of 3.

  9. tabor473

    May 16 V5 Beta Tester OYES, WPI

    @224x The best of one makes for a better build quality, it means you can't afford your robot going wrong. Vex U should have had this rule for a while BECAUSE it doesn't have alliances, it's always the same robots going against each other, if all robots work, why play the same match twice?

    I think you might be suffering from a common misconception. More matches is a better measure of consistency than less matches. It is why during scientific tests they recommend a larger sample size. Any robot can work or not work once, over 100 matches you would really see the difference in consistency.

  10. Aponthis

    May 16 Gilbert, Arizona Formerly 127A, 127C, 127X

    @OscarMNOVA12 After seeing what happened to XD and my friends on MASON, I just wish semis and finals were best of 3.

    The problem is, random weird things (or even highly competitive matches) can happen before this. In a single elimination tournament bracket (not specifically best of one, but a bracket where you only need to lose a series to one team as opposed to losing two series), being "finalist" is not a guarantee that you are better than a quarterfinalist, even under ideal circumstances (the best team always wins).

    @224x The best of one makes for a better build quality

    I had my best-built and most consistent robot that I had ever made for Worlds, yet I was still screwed by the best of one format. I would not consider what happened to us to be foreseeable or preventable.

  11. 224x

    May 17 Malibu High School, California 224X

    @tabor473 I think you might be suffering from a common misconception. More matches is a better measure of consistency than less matches. It is why during scientific tests they recommend a larger sample size. Any robot can work or not work once, over 100 matches you would really see the difference in consistency.

    But this means you have to make your robot more consistent, and isn't that better then having it work 2/3 times?

  12. OscarMNOVA12

    May 17 Event Partner, V5 Beta Tester Northern VA and Pittsburgh 12

    I'd rather have it work 100% of the time for 2 matches and win those 2. Or if I get beat by just a better team and everything is working, so be it.

  13. Just gonna throw out that I have no problems losing to a better team, but if I lose a match in BO1 to a silly mistake with the robot, that would kill my inhibitions to compete

  14. 224x

    May 17 Malibu High School, California 224X

    @griecoeric Just gonna throw out that I have no problems losing to a better team, but if I lose a match in BO1 to a silly mistake with the robot, that would kill my inhibitions to compete

    Then don't make silly mistakes :) , but, seriously, you don't see any BattleBots doing 2/3, and they don't make silly mistakes

  15. meng

    May 17 Singapore 8059

    @224x Then don't make silly mistakes :) , but, seriously, you don't see any BattleBots doing 2/3, and they don't make silly mistakes

    I seriously don't think XD made any silly mistakes...

  16. tabor473

    May 17 V5 Beta Tester OYES, WPI

    @224x But this means you have to make your robot more consistent, and isn't that better then having it work 2/3 times?

    If my robot works 80% of the time and yours works 50% of the time a single match will have "the wrong one" of us fail 10% of the time. As compared to playing 2 matches "the wrong one" of us will fail 1% of the time. (fail both)

    Does that make sense? You prove consistency with large sample.

  17. Nathan Rossi

    May 17 The Internet PYRO & 8800X
    Edited 4 weeks ago by Nathan Rossi

    @OscarMNOVA12 In previous years of VEX U, the early rounds of elims were best of 1. Once you got to quarterfinals, it was all best of 3.

    Correct, I acknowledge the modified bracket at the 2017 world championship utilizes best of one for two matches. I think using best of one in this particular case makes sense as only the lowest seeds have to play it (creating an incentive to rank higher!).

    @224x Then don't make silly mistakes :) , but, seriously, you don't see any BattleBots doing 2/3, and they don't make silly mistakes

    Quick story. At our first tournament for ITZ, we were undefeated going into semi-finals. Before the first semi-final match, our robot was sitting right next to the field. I was handed a battery by a team member, I checked the voltage of the battery myself, it was good, so I plugged it straight into the cortex and then placed the robot on the field. Match starts, our lift moves up, but the drive does not move. Turns out the battery went bad in the brief time (15 seconds?) from its voltage being checked to being placed on the robot. The backup battery gave the cortex enough power to stay connected to the field, but we were unable to move the drive because the main battery was not providing power to the cortex. That battery had given no indication it was bad the entire season, this was the first sign of something wrong with it.

    Upon getting our first loss of the season, our thought was "oh well, at least we have the next two matches." But, if we were playing best of one, that would have been the end of the tournament for us. There was absolutely nothing we could have done to prevent that situation. I don't think I would classify that as a "silly mistake."

    All this to say again, why does VEX U need best of one? We don't have alliance partners and 87.5% of the time we don't have enough teams to play a full round 16. The world championship is only time the round 16 potentially makes sense. So it comes down to the question, do we need 16 teams playing in the elimination rounds or is the top 10 enough at worlds?

 

or Sign Up to reply!