Ruling on Robonauts Balance

[wireties]

Originally Posted by bduddy

This is what happens when the GDC (or anyone else) tries to make "common-sense" rules or "simple" rules or anything of the sort. Discussion about "rules lawyering" or whatever misses the point: rules need to be clearly defined and as comprehensive as possible, or there's no point in having them. Otherwise you end up with differing interpretations and that very rarely ends well.

This argues both sides of the issue. "This is what happens when" versus "as comprehensive as possible". One absolute truth is that it is NOT possible (preivald - "they can't possibly anticipate every solution") to construct a set of rules that is 100% comprehensive. So every year there will always be some level of ambiguity. The engineer's response is to calculate the risk and make a decision.

The robonaut's device is ingenious (dare I say awesome). It was risky for consuming mass on the robot, money, time and effort etc. But the ramifications of it not working (or being illegal) was near zero because the robot is awesome without the feature. So risk of effort times risk of result is still near zero and they went for it. It was kewl to see it in action!

[pfreivald]

Originally Posted by Cory

I see no possible interpretation in which this put them in a difficult situation.

It's a difficult situation for all game designers who want a game that is both fun to play and to watch. (IMO Warhammer 40K and D&D are both fun to play... not so much fun to watch. Football is fun to watch, but I have no interest in playing.)

Originally Posted by Cory

There is no way the GDC did not understand that teams wanted to know if they could hang from the side of the bridge. They chose the cop out route of answering with a non-answer.

It's not a cop-out at all. What they did was refuse to get into a cycle where they are forced to define every word in the manual (and then possibly define the words used in the definitions, and then the words in those definitions). These are really smart people, and I'm certain they understand Godel's Incompleteness Theorem as applied to creating rules sets.

They've been using "reasonably astute observer" definitions for years as a way to essentially say, "Look, folks, it's impossible to create a positivist document. Not difficult, not really hard, but actually impossible. So we're not going to try to do that. Be creative, but do so within the spirit of the competition-as-sport that we've set up."

Originally Posted by wireties

The engineer's response is to calculate the risk and make a decision.

And The Robonauts did indeed create an awesome device with minimal risk -- it's not like they can't balance without it.

[Billfred]

Some thoughts:

1) Yes, 118 is taking this design to the very limit of the rules (though, in my mind, never quite over).
2) Yes, the GDC should be more clear in these edge cases. (Not the "Can we use the F-P motor to raise our shooter up to get a closer shot?" questions, but you know what I mean.)
3) Yes, I'd love to read the GDC's position on the whole matter.
4) I'm a touch surprised, but eminently pleased that we made it to Week 1 of competitions before we reached a rule or ruling that made me want to say "Paging the IRI Planning Committee..."
5) No, I still don't want to be on the other side of the glass from these guys.

[Taylor]

Has anybody done a stress analysis on the actual bridge structure to see if it could withstand suspensions from the side? As thorough as they are, I'd assume 118 has done so but I haven't seen evidence.
That's a pretty incredible moment arm for that part of the bridge to support; if the same part of the bridge was used to suspend several times during a regional, and several regionals during a season, it's possible for that component to fail.
Can you imagine the (figurative and literal) carnage that would follow if the bridge broke during eliminations? Possibly leading to field, robot, even site damage? Screams of "We were balanced until the field failed - we should be given those points!"
Are there spare bridges shipped with the field? What if a particularly innovative team coupled 118's design with 1501 - allowed a robot to drive upon it, then balanced on the side of the bridge? There's no way the structure could support that.
The GDC made the ruling that had to be made. Sometimes the game has to be played in real time to really understand the rules - it's happened before. That's why we have team updates on Tuesdays and Fridays.
I agree the timing is poor in this circumstance. Kudos to 118 for everything.

[Jim Giacchi]

Originally Posted by Taylor

Are there spare bridges shipped with the field?

Yes, having set the field up in New Jersey, each field has a spare bridge, spare back boards, spare hoops, spare plexiglass for the player's station and a spare barrier. There is also an entire extra backup field ready to go in the case of some catastrophic accident with a truck catching on fire or freak meteor showing destroying a building.

[cgmv123]

Originally Posted by Jim Giacchi

There is also an entire extra backup field ready to go in the case of some catastrophic accident with a truck catching on fire or freak meteor showing destroying a building.

Just to clarify, that field is not on site. It's in Tennessee, and I think it becomes Einstein after the season.

[O'Sancheski]

Originally Posted by cgmv123

and I think it becomes Einstein after the season.

It only becomes Einstein if it is never used.

[JesseK]

To the Robonauts: you guys took a calculated risk in which the impact meant you potentially wasted build time. Given that the rest of your bot performed remarkably well for Week 1 from what I saw and that you could still balance like an "average" robot, I'd say the risk was very worth it. The evident systematic decision making on your team is something that every team should aspire to. I'd like to thank you for sharing the full extent of your experience.

The good news: The "side-hanger" can be replaced with a "stinger". Heh. I agree that the way the ruling happened leaves some room for improvement, but the ruling itself may put them in a better position than before.

[KrazyCarl92]

I just hope this helps the GDC realize that there is a REASON these specific to design questions are asked in the Q&A isn't because we want to ask for design reviews, we need to know whether or not certain and specific robot-robot or robot-field element/game piece interactions are legal in order to build robots within the rules. Something that is vague and without formal definition is immediately something that needs clearing up. Although I personally agree with the GDC's decision as to the "grappling" associated with 118's mechanism, I would like to thank the Robonauts for a few things:

1. Pushing the envelope
2. Inspiring others to push the envelope
3. Hopefully showing the GDC why this kind of legality is an important thing to decide week 1 or 2 of build; not week 1 of competitions. Would this have gone on til week had that been when 118 first competed? That would be ridiculous... (it already is in my mind).

[Kevin Sevcik]

Originally Posted by pfreivald

They've been using "reasonably astute observer" definitions for years as a way to essentially say, "Look, folks, it's impossible to create a positivist document. Not difficult, not really hard, but actually impossible. So we're not going to try to do that. Be creative, but do so within the spirit of the competition-as-sport that we've set up."

The problem with falling back on "reasonably astute observers" is that, clearly, reasonably astute observers are going to disagree about things. It's a fine guideline for most cases and situations, but it's completely useless in these kinds situation. I mean, I hope we can agree that teams like 1967 and 118 are made up of reasonably astute people. These reasonably astute people are telling the GDC, "We think this is legal, but we thought we'd check with you. What do you think?" And then the GDC mumbles something, shrugs and wanders off. Exactly how are teams supposed to react to this kind of behavior?

One option is to move forward in good faith but with some trepidation that the the GDC will turn around and declare it illegal: "Didn't you hear us mumble that it wasn't legal?" Clearly 118 picked this path and I think they have some cause to be annoyed at the GDC for mumbling and equivocating during build season before finally making a decision over a month later.

I suppose your other option is to start assuming that a "reasonably astute observer" is actually a perverse killjoy that hates creativity and unexpected situations. So any time the GDC falls back to that excuse you just assume that your idea is illegal, but they don't want to outright TELL you so because, you know, that'd be all depressing and would make the GDC feel bad. This is certainly a safer position to take, but you have to admit it makes things a lot less interesting.

[waialua359]

Although I dont agree with the final ruling made on Friday morning, perhaps its a lesson to teams and the GDC in the future where teams should be allowed to submit diagrams and pics in the Q&A asking a question like, "Is this legal?"

[=Martin=Taylor=]

I think the rules keep getting more restrictive each year...

And there were only a few teams who came up with this innovative solution. They should be rewarded.

Are there any good videos of it using the rollers to balance?

[Chris is me]

Originally Posted by =Martin=Taylor=

I think the rules keep getting more restrictive each year...

And there were only a few teams who came up with this innovative solution. They should be rewarded.

I agree with you in principle, but I think the better answer would be "The Q&A should actually answer questions like these in advance so that more teams are willing to know if it's a waste of time or not." Non-answers hurt everyone. If they allowed it, teams would feel cheated when they didn't take a legality risk due to their interpretation of a vague Q&A answer. It would be a firestorm. Remember how bad the perfectly, unambiguously legal 469 robot in 2010 was?

Regardless... I can't help but draw comparisons to the "troll bot" debacle debated here just a few weeks ago. Both were failures of the Q&A system, though this was a non answer gone wrong versus a full answer suddenly reversed... They're the same issue to me. The teams are not at fault in either situation.

I'm interested to see if anyone thinks one was different than the other, and why they believe that to be so... but since that would derail this thread, private message may be a more appropriate medium for that discussion. Or I could start a new thread, I dunno.

[45Auto]

[G10]
Robots may not grab, grasp, grapple, or attach to any Arena structure. Robots may not push or react against the top of the Fender. (Robots may push or react against any element of the Arena that is not protected by another rule.)

A reasonably astute observer would look at G10, and say that 118 was "attached" to the arena structure.

They would say the same thing if the 118 mechanism was used by a robot to lift itself on the rail around the arena, or the gates into the arena, or the backboards behind the goals to create a robot capable of dunking the balls (in other words the rule works the same way for "any arena structure").

[philso]

I had the privilege of speaking with one of the students who designed the mechanism that was ruled illegal. He seemed pretty astute to me. I am very sorry that we will not get to see the improvements to that mechanism that he spoke of. I am sure that even though the mechanism was ruled illegal and will not be used in competition, one of the main purpose of this competition has been achieved. The students learned how to solve a difficult engineering problem and did it in a spectacular way.

Kudos for pushing the boundaries yet again, 118. That is why our civilization advances.

Phil