IRI Rule Changes - 2012

[Travis Hoffman]

Originally Posted by LeelandS

At the same time, IRI is definitely not a place for the weak of heart.

I like that characterization much better.

[EricH]

Originally Posted by Chinmay

IRI isn't an official FIRST event. They are not required to play by the same rules. They can make these changes because it's an offseason event... Street ball has different rules than the NCAA tournaments doesn't it? I think that IRI is essentially some really high class street ball.

Yes and no. It really depends on the street ball you're playing, and where you're playing, and who you're playing with, and which game/game type you're playing. I've seen it at 2/basket, regardless of distance; you could be playing halfcourt, short basket. It really depends on all those factors.

Most common, of course, is that the official scoring rules, height rules, etc. are left alone except as needed for half/full court situations.

Oh, but wait! There are some changes made in official tournaments! For example, in the NCAA, the time is often shortened in the wide spaces of the bracket. This is because there are games on back-to-back days, and having players play tired isn't good for anybody. I seem to recall that there's something similar in international competitions. But the only things messed with are time and possibly number of officials and replay.

I would suggest that IRI isn't street ball. It's the Olympics. You're supposed to tweak small stuff. The problem is, the tweak this year was to change the inbounding rules (or whatever other basketball rule you want to use--slam dunks was a good one earlier), which changes the face of the game.

[JVN]

Originally Posted by Justin Montois

Also...
http://www.chiefdelphi.com/forums/sh...d.php?t=106213

In that poll, the FRC community voted 120-92 to keep the CoOp bridge at at least one point. I think that would have been a fair compromise.

"A vote says more about the voter than the thing being voted on."

-John

[Akash Rastogi]

There is no reason to have co-op points at an event like IRI.

The co-op bridge and points were determined by many to be a way for underperforming teams to have a chance at being seeded high (you may not share this opinion, but I do). No that doesn't mean that it always shot lower performing teams into the top 8, but I don't think that IRI is the place to cater to the underperforming teams. Most of the teams attending are above average or top tier.

I see no place for the co-op bridge or co-op points at the IRI. This rule change, I believe, was a correct one.

+$0.02

[Tristan Lall]

Originally Posted by BJC

I think I fundamentally disagree with you. I believe that every team's final rank should be resultant from how much they win (how good they are) because that is what we are competing to find out. Everyone is competing to see who is the best, so when the ranking system utilizes a metric which has nothing to do with winning it throws a tremendous amount of noise into the ranking system. If the ranking system is supposed to rank teams by how good they are why should hypothetical team 0000 be ranked over hypothetical team 9999 when 9999 worked harder during the build season, can more effectively play the game, and won more qualification matches?

I don't think it follows that how much they win = how good they are. We can measure wins easily, but we factor other things into our assessment of goodness.

Given that the original rules seemed to count the ability to balance the co-opertition bridge as a good thing, I have to disagree with the premise of your argument.

If you use wins as a proxy for goodness, you should expect error because you're neglecting other factors. But that error isn't noise in the conventional sense—it's a collection of unidentified explanatory variables.

[pfreivald]

Originally Posted by Tristan Lall

Given that the original rules seemed to count the ability to balance the co-opertition bridge as a good thing, I have to disagree with the premise of your argument.

I agree 100%. There's a difference between "good at playing any particular match" and "good at Rebound Rumble." There are a lot of commonalities between these two things, but they are distinct -- distinct enough to call these rules changes significant game-changers for which competitors thus disadvantaged could feel justified in their disgruntlement.

[Justin Montois]

Originally Posted by JVN

"A vote says more about the voter than the thing being voted on."

-John

I don't disgree, but why then have the vote...

[Bjenks548]

This is going to put a new strain on Qual partners, someone has to be in charge of calling the balance and when. In Elims, its been fairly eazy as there is always a captain. Now each alliance will have to have someone incharge of balancing because 2 going for a double and 1 going for a triple does not work well. I know many (mostly michigan) teams pretty well and trust their coaches, but there are many more that I don't know. I hope to go into every match with an idea of how my alliance can triple balance, even if we don't use it. Also, really hoping to get paired up with 67 and/or 118 and another wide as I know we can triple quickly and reliably with them. Should be a lot of fun to see how triples progess (or don't) throughout qualifications.

[BJC]

Originally Posted by Tristan Lall

I don't think it follows that how much they win = how good they are. We can measure wins easily, but we factor other things into our assessment of goodness.

Given that the original rules seemed to count the ability to balance the co-opertition bridge as a good thing, I have to disagree with the premise of your argument.

If you use wins as a proxy for goodness, you should expect error because you're neglecting other factors. But that error isn't noise in the conventional sense—it's a collection of unidentified explanatory variables.

Every robot is built to try to win matches in its respective game. Even though there are other factors involved (such as your partners and opponents) I think that how many matches you win overall during the season is still a fairly good indicator of how "good" your robot is. It's not perfect, but its the metric used by competitions everywhere. It is, in any case, a better way to determine robot rank at a competition then the co-op bridge is. So no, it is not the only or best metric to determine how "good" a robot is. Additionally, I believe that my use of "good" has been slightly misunderstood because it is such a vague term. I eariler used "good" to describe how high a robot would seed. While I can't think of a better term, that is probably a poor choice of words for the reasons you described. I hope I explained myself well enough.

Regards, Bryan

[Tetraman]

Originally Posted by BJC

Every robot is built to try to win matches in its respective game. Even though there are other factors involved (such as your partners and opponents) I think that how many matches you win overall during the season is still a fairly good indicator of how "good" your robot is. It's not perfect, but its the metric used by competitions everywhere. It is, in any case, a better way to determine robot rank at a competition then the co-op bridge is.

But that's the thing, the co-op bridge isn't the only way in which your rank is determined. You have to win the match before you can reap the real benefits of the co-op bridge. There is a benefit in earning co-op points when you lose, but the real winner is the robot that wins the match and earns co-op points too. And in order to gain those points, your alliance must still win the match. So yes, you are right, if a robot is deemed "good" it's usually because of how many matches they win, but part of the game of Rebound Rumble and what makes teams (not robots) good in Rebound Rumble is their ability to utilize the white bridge.

Every robot at IRI is going to win their fair share of matches. That's for certain. And the robots that are expected to be in the top 8 will most likely do so. But the co-op bridge allows you to extend your victory and turns it into the same as two victories. Cooperating is a skill of Rebound Rumble, and it is a whole lot more than just another way to rank robots. It's a skill, and an alliance can be rewarded for utilizing and implementing that skill.

I find that removing the co-op points is more than just the removal of a "fairness" system that FIRST implemented because Coopertition and GP and all that, it's also removing a core and fundamental part of the game just so the teams and spectators can watch a few more triple balances.

[EricH]

Originally Posted by BJC

Every robot is built to try to win matches in its respective game.

Do not continue from this point until you define the game and the match.

This is where the people who like the rule change and those who don't like it differ. They have different definitions of the game. Those who like it say that the rule change focuses on winning the match. What match?

First I'll look at the "2:15" match definition. In that match of the game, each match lasts 2:15, and the game is to win each match. People who see the IRI rule change as a good thing will tend to cite the number of what you might call "sub-par" teams in the top 8, fairly high in the top 8, in fact, and the relatively high number of teams who were really good but not in the top 8 as a bad thing. They tend to express the opinion that coop points ruined the game. They play by the Game section of the Manual.

Now, I'll look at another definition. In this, a "match" is about the length of a regional. A game lasts a full season. This is where the people who don't like the IRI change tend to hang out. The game they play is defined by the Game and Tournament sections of the Manual. The game changes every year. This group plays to win events.

In normal play, the two definitions are the same for about 5 hours. Saturday afternoons at a regional or district are full of teams playing to win matches to win events. But before then, there are teams who play to win matches and teams who look at their rankings, knowing that the only way to guarantee a spot in the eliminations is to be top 8. These ranking-watchers know that they can affect their rankings by scoring for their opponent--or, in extreme cases, actively preventing themselves from scoring. Teams like that understand the full game, top to bottom.

IRI is different, however. Normally, the rule change affects the game play as defined by the Game section of the Manual. This year, it affects the game play as defined by the Tournament section of the Manual. Does it penalize teams unfairly? Possibly. But not for sure. Does it take away the meta-game that was discussed during the season? Definitely. Is that meta-game important?

That question is one that each team decided during the season. In my personal opinion, it was very important to play that meta-game well during the season. At IRI, it has no value.


For many sports leagues, there is actually a meta-game. It's called seeding. If you win, you get so many points towards seeding. If you lose, you might not get any--or you might get some based on how much you scored. It is possible, in some leagues, to lose a single game--and yet come out ahead in the meta-game by scoring a lot of points in those losses, forcing your opponents to score even more to beat you.

[Travis Hoffman]

Originally Posted by Adam Freeman

What did any team do in their design process that was specifically pointed at performing a Coop balance? I have seen very few teams that do anything other than drive on the bridge and hope it balances.

Perhaps the skills developed by teams surrounding the "art of the co-op" are not solely centered around the robot?

In addition to a solid drivetrain, perhaps the skills that separate good co-op teams from those who aren't as regularly able to take advantage of it include the ability to effectively strategize with the opposing alliance before the match and communicate an effective co-op plan, and repeat and execute that planning process every match, like clockwork. Knowing how best to:

1. Decide which two robots will co-op (there was a standard we used all season that worked very well for us here).
2. Decide when the teams would head for the co-op bridge.
3. Decide who would tend the bridge.
4. Decide how and where the 2nd bot gets on.
5. Communicate the balancing method - who "leads the dance".
6. Convince the opposition that you would honor the co-op contract and had the driver skills/robot capability necessary to make it work.
7. Make it clear that robots violating the co-op contract by continuing to score instead of balancing would incur your graciously professional wrath.
8. And accomplish all this within a very short amount of time.

I don't think these things were as easy to accomplish as people make them out to be, especially at events where the pool of co-op partners was filled with more robots/teams with various deficiencies. Navigating that minefield successfully at each regional/district was a learned skill.

I understand why the co-op bridge process has been removed at IRI, but let's not trivialize the efforts of those who were able to grasp what the GDC was intending for teams to accomplish at the bridge and use the system to their advantage.

[IndySam]

So with the advent of these rule changes and the quality of the robots at IRI will we finally see the advent of a major ball starvation strategy?

If you have 3 robots that have a more than reasonable chance of tripping and are decent at ball collecting would it be advantageous to collect 15 of the balls before going to triple so the other alliance can't put up big points on you while your trying?

[AdamHeard]

Originally Posted by Tetraman

Every robot at IRI is going to win their fair share of matches. That's for certain. And the robots that are expected to be in the top 8 will most likely do so. But the co-op bridge allows you to extend your victory and turns it into the same as two victories. Cooperating is a skill of Rebound Rumble, and it is a whole lot more than just another way to rank robots. It's a skill, and an alliance can be rewarded for utilizing and implementing that skill.

I don't understand this point. Teams will win/lose just as much as before.

The co-op would've been negligible at IRI anyway with a 90%+ success rate, so why even have it? Teams would've been sorted win/loss pretty much.

The removal of co-op is completely separate from the addition of triples, and shouldn't upset anyone once you analyze what likely would have happened.

[Travis Hoffman]

Originally Posted by IndySam

So with the advent of these rule changes and the quality of the robots at IRI will we finally see the advent of a major ball starvation strategy?

If you have 3 robots that have a more than reasonable chance of tripping and are decent at ball collecting would it be advantageous to collect 15 of the balls before going to triple so the other alliance can't put up big points on you while your trying?

Of course, the same strategy could be played by those expecting an opposing triple, preserving ammo for late in the match for that final scoring + double push.

A lot probably depends on who wins hybrid mode.