SG3 is Broken, Please Consider This Fix

VEX Robotics Competition Discussion VRC > Tipping Point (21/22)
1

Game Design Committee,

As currently written, SG3 is penalizing teams for interactions that should not be illegal, and has resulted in frequent ticky-tacky penalties that are affecting the outcome of matches. Below you will find my rewrite of SG3 and reasoning as to why I believe this version would be much more fair to teams, easy to enforce for referees, and easier to understand for both competitors and spectators.

Platforms are “safe” during the endgame. During the last thirty (30) seconds, Robots may not directly or indirectly contact the opposing Alliance’s Platform. The intent of this rule is to prohibit Robots from inhibiting their opponents’ ability to utilize the Platform at the end of a Match.

Contact with the opposing alliance’s platform that does not affect the outcome of the match will result in a warning. Contact with the opposing alliance’s platform that is match affecting will result in a Disqualification. Teams that receive multiple warnings may also receive a Disqualification at the Head Referee’s discretion.

a. Indirectly contacting the opposing Alliance’s Platform through another Robot or using a Scoring Object per SG10 would be considered a violation of this rule.

b. Robots partially or fully within in the opposing alliance’s Home Zone in the last 30 seconds of the match will be considered acting defensively and will not receive the protections normally given to an offensive robot by G13

c. Placing a Mobile Goal on or under the opposing Alliance’s Platform, at any point during the Match is considered a minor violation of this rule that, at a minimum, will result in a warning.
i. If the placement is accidental, and immediately rectified (i.e. the Mobile Goal is immediately removed), then this violation will be considered a warning.
ii. If the placement is intentional and / or not immediately rectified, as judged by the Head Referee, then it will be considered a match affecting violation.
iii. Repeated, strategic, and / or egregious violations may also escalate to a match affecting violation, at the Head Referee’s discretion

Note 1: Per point “b”, by entering the opposing alliance Home Zone in the last 30 seconds, Robots accept the risk that they are automatically considered acting defensively and per G13, Head Referees will err on the side of the offensive robot during a questionable rules violation.

Note 2: If it is not obvious a Robot is within the opposing Home Zone volume, i.e. only a small portion is partially within the volume, the Head Referee will always err on the side of the Robot being outside of the opposing Home Zone

Note 3: If point “c” has escalated from a warning into a violation, then it will automatically be considered a match affecting violation, i.e. will result in a Disqualification.

Changes:

  • Removed the concept of “interfering with gameplay” and replaced it with “match affecting”
  • Removed the elevation bonus
  • Removed G13 superseding G14
  • Added that in the last 30 seconds, robots in the opposing alliance’s Home Zone are automatically considered acting defensively
  • Replaced “transitive contact” with “indirect contact”

Current problems with SG3:

  • SG3 violations are too frequent
    • This results in many post match discussions with the Head Referee, slowing down cycle times, which is important to keep up at events, especially worlds
  • The zero-tolerance nature of SG3 removes all nuance, and combined with infinite transitive contact, means it is very easy to accidentally commit a violation
  • The automatic elevation penalizes an action that by definition “does not interfere with gameplay” which has a few consequences
    • Changes the outcome of matches for something that should not be a penalty
    • You get penalized if your partner in a qualification match commits this violation, changing the outcome of the match for something you could not control
  • “Interfering with gameplay” violations that result in a DQ are often violations that do not actually affect the outcome of the match
  • Allows teams to intentionally push other teams into an SG3 violation by nulling G14
    • If it is legal, teams will do it. You cannot “hope” teams will not do something if it is legal to do so. I have seen teams take advantage of this rule first hand
    • Potentially having to invoke the Code of Conduct and likely get the EP involved because of a team taking advantage of the nulling of G14 is a waste of time for the EP

Goals for this rewrite:

  • Reduce the frequency of calls to only violations that affect the outcome of the match
  • Close the G13 superseding G14 loophole
  • Ensure the rule is easy to understand and enforce

Explanation:

First, the primary difference between this rewrite and the current version of SG3 is removing the concept of “interfering with gameplay” and replacing it with “match affecting.” If it ain’t broke, don’t fix it. One of the many aspects I always admired about VRC games is how they did not have points associated with violating a rule like other competitive robotics programs. The rules in VRC games were designed such that violations were extremely rare and only came into effect if it actually changed the outcome of the match (or multiple violations occurred).

While the majority of the rules in the Tipping Point game manual still use the match affecting clause, SG3 has instead opted to use “interfering with gameplay,” which to my knowledge, is a new concept to VRC. This has led to scenarios which “interfere with gameplay” but do not affect the outcome of the match. Why penalize a team for something that does not even change the overall outcome?

I speculate this was done in an effort to make it easier for referees to enforce and decentivize teams from getting close to the opposing platform in the last 30 seconds. I get it, simple rules are easy to enforce, but the zero tolerance nature of SG3 is getting teams (rightly) frustrated over penalties that frankly make no sense being a violation. The rule has also not discouraged teams from getting close to opposing platforms in the last 30 seconds. This is because there are so few primary scoring objects that you often have to at least attempt to take a mobile goal near the opposing platform or you are guaranteed to lose.

To reduce frequency and increase fairness, the primary change is quite simple. By directly or indirectly touching the opposing alliance’s platform, did it change the outcome of the match? Yes? Disqualification. No? Warning. The two-tier penalty system of an extra elevation and/or a DQ is gone.

In terms of how easy it will be for referees to enforce this change, it’s quite simple. In the vast majority of cases (there will always be difficult edge cases), it should be easy to judge if contact with the platform was match affecting. Did a robot knock off mobile goals or unbalance a platform changing the outcome of the match? Match affecting. Pushing a mobile goal on the ground such that it is now touching the platform making it unbalanced changing the outcome of the match? Match affecting. An opposing robot is in the way of a team trying to drive up their own platform and in the process the opposing robot touches the platform indirectly or directly causing a long enough delay to change the outcome of the match? Match affecting.

With the switch to match affecting and removal of the bonus elevation, infinite transitive contact no longer poses any real issues. The main issue of accidentally touching the platform through 3 scoring objects for a 30 point penalty that never interfered with gameplay is null. The change to “indirect contact” was to keep the rule concise, contact would be effectively enforced the same.

The other major change is to SG3-b and keeping G14. As mentioned in Q&A 948, the intent of SG3-b as currently written is to remove the need for judgment as to which robot pushed who into the platform and which robot is at fault for it. This is an important aspect of SG3 with regards to enforcement as it can be difficult to judge who is at fault. However, ditching G14 is not the way to do it as it has created a loophole where you can intentionally force your opponent into a violation.

The change to robots acting defensively in the opposing home zone in the last 30 seconds is to provide protection around the platforms. One example of this in action is a red robot attempting to elevate a mobile goal on the red platform. A blue alliance robot is currently stealing a mobile goal near the platform that is still on the foam tiles but is currently in the way of the red robot. The red robot is trying to push through to score and in the process the blue robot contacts the red platform. Although the blue robot is trying to take a mobile goal to score (offensive action), they are in opposing territory and are considered acting defensively. If the interaction delayed the red robot from elevating the mobile goal on the platform and it affected the outcome of the match, DQ. By going into opposing territory, you accept the risk of a ruling against you.

Let’s take another example, a blue robot is attempting to steal a mobile goal from the red alliance corner in the last 30 seconds, they grab the mobile goal, and begin to back away. A red robot t-bones the blue robot and continuously pushes them into their own red platform, causing it to become unbalanced. Although blue is acting defensively by being in the opposing home zone, the red robot continuously pushed blue into the red platform, and since G14 is no longer null, blue is not at fault as they were forced into the penalty.

This change to SG3-b would make it similar to how it worked in past (1) games (2). If it worked perfectly fine then, why not now?

If one of the arguments against this change is “We can’t change how this rule was written just for the last tournament of the season.” I would argue this change simplifies the rule, and would be enforced in a very similar way but would only DQ teams if they actually have done something wrong. Also, a much bigger change was made right before the 2018 VRC World Championship, this is not unprecedented.

I hope you consider these changes. I left out some points of discussion for (relative) brevity. If there is anything I can clarify, please let me know.

@Grant_Cox @Jon_Jack @Sidoti

49 Likes
2

I fully agree. As both a competitor and a referee, SG3 as it stands is deeply and utterly broken. It is nearly impossible to enforce properly, and even when it is, it’s often critically unfair to everyone involved in the competition. It is not a sign of a good rule when a greater infraction can result in less of a penalty. It is unfair to teams when they can get disqualified for something they were forced into doing, it is unfair to referees to force them to waste attention watching for ultimately irrelevant contact with the platforms, and it’s unfair to event partners to delay cycle times with every third match having a rules infraction. SG3 needs to be fixed.

15 Likes
3

I agree that SG3 has many issues, and is very hard to judge. The GDC could do better on making this rule, but some of the proposed solutions will not fix the rule, but just make it harder to enforce. I do think that the GDC should pay attention to this thread, and attempt to change SG3 accordingly to what is said.

Additionally, before reading this post and even while thinking about the post above, think about the intention of SG3. What I believe the intention of SG3 to be is that robots should spend the first part of the match trying to score as much as they can while they are unprotected, and then have 30 seconds to get back to the platform. If everyone was playing as the GDC wanted, this would work perfectly. However, most teams (me included) play with the intention to win. Having teams drop their mobile goals near/on their platform close to the 30-second mark can change the course of the match, so teams will go for them and risk the SG3 violation. SG3 is so harsh to try and prevent this, and thus SG3 violations are very common.

While trying to make the rule simpler may seem like a great idea at first, judging “match affecting” with the platform is very hard. I can totally see situations where teams can claim “but ref if they did not touch my platform I totally could have double parked with all my goals and win.” I think what the GDC is trying to do by making the rule very harsh is make it very easy for refs to dish out this rule. This is one part of the rule that while very annoying for competitors does not really have a better option.

The elevation bonus is very fair and prevents the GDC from having to dish out DQs almost every match. If a robot is near another team’s platform close to the last 30 seconds, they are almost certainly trying to disrupt scoring of the other alliance. The elevation bonus allows for the other alliance to have a “fair” addition of points similar to what they would have gotten if the other team was not interfering.

This is something I would love to see happen in the game manual. The fact that robots can push other teams into their platform is very unfair. If you are close to their platform and they push you in I could see this as being fair like you included in your updated rules. However, if a team is minding their business far from the platform they should not be able to be pushed into it. Having the rule state that being in the opposing alliance zone is a great change.

7 Likes
4

There is an inherent problem with this though. In qualifications, a DQ is only applied to the robot that commits the violation. The point penalty applies to both teams. Therefore, the lesser infraction that requires a penalty could result in a loss for the entire alliance, whereas if a robot had gone further and gotten itself DQed instead, its alliance partner would still get their win for the match.

28 Likes
5

Yes this is an issue for teams, but I wouldn’t say it is a problem with SG3. SG3 is probably the one that gets broken the most, but alliance partners in qualification matches can always be the maker or breaker of winning or losing. This is just a part of vex, and always will be. I don’t think changing SG3 will ever fix having worse alliance partners who may not know the rules as much.

3 Likes
6

Underlying the aggressive nature to elevate and possess as many goals as possible, low points for rings. If rings were worth more, you might see teams actually focused on scoring them, rather than being a nuisance on the field - speed bumps/pot holes :slight_smile: Say make rings worth 5pts for bowl, 10pts for regular branch, and 20pts for high branch would considerably change the dynamics of the game and game strategy.

I do agree SG3 allowing a robot to be pushed across the field into a violation seems an invitation to bad behaviors.

21 Likes
7

To be clear, the goal of the rewrite did not state “Ensure the rule is easier enforce than how SG3 is currently” it stated “Ensure the rule is easy to understand and enforce.” The current problem with SG3 is that it is too easy to enforce and penalizes actions that makes grandma go “Wait, why did they lose over that? That shouldn’t be a penalty.” It currently resides at the extreme end of the “easy zero tolerance” to “everything is an unclear judgement call” spectrum. It is too easy to enforce because there is currently no nuance in interactions involving the platform.

This rewrite will involve the Head Referees needing to make judgement calls, but this is how VRC has historically worked for 10+ years (see original post for Q&As on the matter) without issue. As stated in the original post, the judgement made by the Head Referee ultimately comes down to:

Violations of the rules should only have an impact if they actually affected the outcome of the match. Trust me, as someone who has been a Head Referee for this game, the last thing I want is a confusing rule that is difficult to enforce. Hence the goal of being easy to understand and enforce. Judging whether an interaction changed the outcome of the match will generally be quite simple (see the examples listed in the original post).

In an early draft of this rewrite, I had originally removed SG3-b in its entirety. However, I believe it was important to ensure robots attempting to elevate on their platform still had benefit of the doubt protection from opposing robots, hence SG3-b in its current form. An opposing robot cannot argue they were “acting offensively” in the opposing Home Zone. This eliminates most of the ambiguity as to who is at fault in unclear situations, making it easier to enforce for the Head Referee.

I would also argue certain aspects of the rule are now easier to enforce. Head Referees no longer have to take their focus away from the overall flow of the match and major interactions
to instead look and see if 3 scoring objects create a chain to the platform.

EDIT: The pitch in the first paragraph of the original post did say “easier to enforce for referees” as part of the benefits of the rewrite, I have edited the first paragraph pitch in the original post to reflect the goal of the rewrite.

Teams disagreeing with the Head Referee’s judgement will always occur. This is why it is the responsibility of the Head Referee to clearly explain their reasoning and what rule was violated.

I agree, and as I stated in my original post, I believe the intent of SG3 was to keep robots away from the opposing platform in the last 30 seconds. In practice, however, the math often does not work in your favor and you often have to go for a mobile goal in the opposing corner in the last 30 seconds. These interactions are inevitable and SG3 has failed to deter teams from trying.

My experience Head Reffing this game is quite the contrary. The majority of SG3 violations resulting in an elevation bonus are trivial interactions of teams trying to carefully navigate around the platform to get to a mobile goal but end up accidentally touching the platform (often through a random scoring object) for a fraction of a second. That should not be a penalty. My original post pointed out a couple of reasons as to why the 30 point bonus is problematic.

The rewrite of SG3-b is currently written such that if a team is legitimately trying to disrupt the opposing alliance from elevating, benefit of the doubt is given to the team attempting to elevate so a ruling is likely to be in their favor.

18 Likes
8

Just to add some data about how often SG3 DQs have to be made, I looked at the results from the last three events I’ve been at.

Across 144 qualification matches, there have been 13 DQs (as far as I know, all of which were SG3 violations), or about 9%

Across 39 elimination matches, there have been 3 DQs, or just over 7%

I have no data how often the lesser penalty has been instituted, but the fact that this many matches involve a DQ is quite frankly ridiculous.

17 Likes
9

I don’t either, but at a tournament I refereed yesterday, roughly half the matches involved penalties, and at least a dozen matches had their outcome determined by these penalties.

15 Likes
10

I agree that SG3 remains fraught – it is difficult to referee and unintuitive for the audience. The high level of protection the rule provides to platform interactions often seems unfair. That said, GDC has stated that is the intent of the rule. They want teams to play offence in their own home zone rather than fight for zero-sum points in their opponents zone. If this is the case (and I could certainly be wrong about GDC intents) you are asking to change the intent of the rule, not just clean it up.

I would also like to see changes to these, but your proposal would dramatically reduce the protections of SG3. As noted above, the elevated penalty is particularly troublesome, as it penalizes the alliance partner. I am for eliminating it for that reason. However I would suggest keeping the penalty high (and easier to ref) by clarifying “interfering with gameplay” to mean any action that changes the score of the game. Match affecting goes away, as any action that changes the score gets that team a DQ. Saves refs having to total the score two ways, doesn’t penalize partners, ignores incidental bumps.

I like the idea that being in the opposing home zone in the last 30 seconds automatically makes a team defensive. It adds clarity to the risks of going there and quashes the “But I was trying to get my own goal, so offence” argument. I don’t think this solves the problem of the Q&A you linked – still possible for red to force blue into a penalty by pushing blue into the platform. You could have the edge case where blue successfully recovers their alliance goal and is a wheel roller away from escaping/becoming an offensive robot by entering the neutral zone. The scenario of the Q&A plays out the same.

Is this just a vocabulary change, or are you shortening the chain of objects required for platform contact?

6 Likes
11
1 Like
12

Ive seen this a good bit at events I helped with and at comps I have gone too
complaints about people not reading the manual aside platform and SG3 violations are the most likely to be seen

13

I agree with all of this. It should be well-known, by now, that the Platform Is Lava™ in the last 30 seconds (and often before). A lot of fault can be found with the GDC for the complexity of the rule, but, at the end of the day, the competitors must read and understand the rules. Head Refs during Drivers Meetings must (and often do) reinforce Nothing Good Can Come From Being Near the Enemy Platform in the Last 30 Seconds.

I’d be curious what the repeat offender rate is. I remember reffing one high-achieving competitor’s first tournament where they were DQed 3 times (and should have been DQed again in the elimination rounds, but we missed a call). They understood why they were DQed, and yet kept going back to that well. That’s not on the GDC.

To add data to this, today in 58 Qualification matches, the head ref logged 20 match violations (roughly 8.5% of competitor-matches, though affecting roughly a third of matches), all of which were SG3

5 Likes
14

SG3 is not a problem if competitors practice their driving and just avoid the opposing alliance platform. The rule is clear: Do not touch or interfere with you opponents platform in the last 30 secs. In my opinion, teams cannot complain about this rule anymore. It is cut and dry. If you don’t want to give the other team a free robot park, stay away once 30 secs hits. the one major problem I see with this rule at the moment is even though you yourself can follow the rules, your teammates can’t. The GDC messed up by punishing you for something your alliance partner has done. There is only so much you can do and say to influence your alliance partners descision making.

4 Likes
15

and yet, we have DQs where an opposing alliance pushes bot across the field to force a DQ. So there are still scenarios that are difficult to swallow, and do warrant complaints (about the rule, not really the call made).

10 Likes
16

I completely agree with this. I have gone to 4 tournaments so far this season and competed in over 40 matches (scrimmages included) and not once gotten an SG3 violation. Teams need to get out of the opponents home zone when 40 seconds hits.

1 Like
17

this is either a drivers error by letting yourself get pushed across the entire field, or a robot issue, where the design limits your gameplay enough to wear you have a major flaw in being pushed too easily. any somewhat experienced team should just be able to drive away if they are being pushed or just avoid being pushed in the first place

18

or maybe the rules are flawed that intentionally disqualifying an opponent is accepted.

I understand the frustration on this particular point of the rules.

oh well. not much to do at this point the season.

14 Likes
19

Joey, while at some high level of gameplay your argument might have some merit, when I show up to a local tournament with my 8 motor drive, built-for-defense robot and shove the rookie school team with the omni wheel drive into my platform, no one has fun. At a high level, yes, every bot should be built to resist this, but you need to accept that most aren’t. Don’t pretend like you are just any other competitor, either. You are consistently winning signature-level events, and are clearly on a different level than the vast majority of teams. The rules need to be fair at all levels of competition, not just yours. Have a little empathy for those still learning. How would you like it if it was your first ever tournament and you got DQed because someone shoved you from across the field? You would feel pretty disheartened, no? I know I would.

11 Likes
20

The first thing I want to get out of the way is that I understand my position and my teams skills level compared to most teams. However, you mention to be empathetic to those still learning. I am just stating the facts as they are. Most teams that are falling victim to being pushed into their platform are lower level teams, and the problem for them is that their robot build and driving skills are not as developed as a more experienced team. I am aware of this, which is why we refrain from abusing this rule in matches where it may be very easy to abuse it. I’m on your side here. I don’t like the rule, and even though it is a legal maneuver, that doesn’t mean everyone is doing it and contributing to the problem.

1 Like