[MCC] Minimum Competitive Concept 2014

[BrendanB]

Originally Posted by JesseK

I don't know that the MCC "build" discussion paints the entire picture of what happened in 2014 - at least not from a "what lessons can we carry forward" perspective. I'd like to point out it took quite a while for the GDC to admit that the definition of possession required refs to deduce intent. This means that if the teams with a spring or other such passive intake/passthrough did not actively go talk to the refs there was a very good chance the assist would not be counted. It also means that it is a terrible idea to design for such a concept during the build season given its subjectivity.

Additionally in 2014 there are inherent risks to an alliance associated with a passive device like a lawn chair - it is just as easy for the robot to get an opponent's ball accidentally, considering the lawn chair's inbounding zone is the same zone in which to catch the opponent's truss and/or HP scoring zone inbound.

Agreed. Looking back its extremely easy to see how a few strands of surgical tubing on a robot were a huge asset in the final rounds but I would not encourage a team to build a trampoline bot back in week 1 (which was a concept we penciled out day 1). If a kitbot with a basic structure and surgical tubing is what you can build might I suggest Vex or FTC which might better suit your team? Teams like 5136, 4908, 5112, and 4909 were extremely good at what they did with just intakes and practiced drivers.

Sadly when teams are successful at building a MCC you tend to get picked in tough positions (usually first/second round of the bottom four alliances) which is what happened to us in 2011.

[Lil' Lavery]

Originally Posted by BrendanB

Sadly when teams are successful at building a MCC you tend to get picked in tough positions (usually first/second round of the bottom four alliances) which is what happened to us in 2011.

#8 Alliance for lyfe!

[Nemo]

MCC is one concept that is defined in a certain way, but it brings a different question to mind.

For a team with median resources,
assuming that competitive success is prioritized over cool factor,
given what was known about the rules a few days into build season,
What is the smartest robot design to attempt?

This year, lots of teams prioritized shooting over acquisition and ended up not being able to participate effectively in assist cycles. It wasn't a secret that ball control would be hugely important for every robot. It was right there in the way the game and its scoring system were designed. Why did so many teams miss on that?

Should we have this thread nice and early next year? It will probably be possible to take a decent shot at MCC (or similar) in the first week of the season, and that might be useful to a lot of teams that would otherwise put their eggs in the wrong basket.

[artK]

Originally Posted by JesseK

Additionally in 2014 there are inherent risks to an alliance associated with a passive device like a lawn chair - it is just as easy for the robot to get an opponent's ball accidentally, considering the lawn chair's inbounding zone is the same zone in which to catch the opponent's truss and/or HP scoring zone inbound.

I think 5136 and 4242 got MCC perfect this year. 4242 was even a captain in DC.

Both of them look like amazing robots, but I would probably give a slight advantage to 5136 (more on this below), because 4242 looks like they could get an accidental possession with their completely open inbounder.

5136 was one of my favorite robots this year, probably number 5 in my personal list (Admittedly there may be some bias from CVR). A pure inbounder, with good defense, AND a goalie pole?!?!? They were the part that really scared me about the Newton alliance, because that was their BACKUP, and with some shuffling, could have replaced any of those robots and still have a scary alliance.

[JesseK]

Originally Posted by artK

Both of them look like amazing robots, but I would probably give a slight advantage to 5136 (more on this below), because 4242 looks like they could get an accidental possession with their completely open inbounder.

I was just giving another example of active MCC mechanisms that didn't merely copy the Ri3D/BB stuff for the inbound. 4242 never actually possessed the opponent's ball, iirc. They also had a tough time inbounding 'on the fly'.

[Cory]

Originally Posted by Jared Russell

In other words...

Yeah, the thread was over as soon as 5136 was brought up. The robot every rookie (and many veterans) should have built.

[Caleb Sykes]

Originally Posted by RunawayEngineer

The "whatever reason" is poor scouting. In my experience, any team that significantly contributes will be noticed by the teams with solid scouting and good match strategy.

I would probably chalk up the "whatever" reason to be some combination of poor scouting, poor strategy, poor communication between drive team and scouting team, and the human tendency to judge things based on how they appear on the outside.

I think making a robot that forgoes "primary" tasks and does other tasks well is a great build strategy.

I think it tends to be a good strategy as well, however you have to hedge your bets. Teams like 5288 should have been snapped up for elims at every event, and they were by the 2-3 teams at every event that recognized their capabilities. Unfortunately though, there just aren't enough teams that make this decision out there, which means that lower-ability teams have to hedge their bets if they want a solid chance to make it into elims. I would have loved to have my team make a robot like 5288's, but I don't have enough faith in other teams' scouts to do this. So instead, we added in some other functionality to our robot. I guess the distinction that I am making here is that the MCC for elims would be 5288, but unfortunately you might not make it to elims with this design, so maybe the abilities of the MCC need to be a bit higher.

You will probably be ignored by the captains with poor scouting (whom you generally don't want to partner with anyway) and you can be noticed by the ones that scout and pick strategically.

I would rather make the elims with a partner who I don't want than not make elims at all.

And you can always make sure you are noticed by talking to the teams that will be in a position to pick. I have been on a high seeded team a number of times - never once did I have a team come to me to discuss their value as a 3rd partner. Any team that would commit to working with our strategy and could demonstrate their ability to do it would leap up my pick list.

I've had a couple of cases where teams try to "sell themselves" to my team before the elims, but I agree that the prevalence is far lower than it should be.

[EricH]

The assumptions: First, assuming that most robots have some form of functional shooter. Second, assuming most robots may have trouble with intakes or shooters from time to time, with some types needing more help than others. Third, assuming that 66% or more of the time this robot will be playing defense; the remaining 33% will be taking defensive hits. Fourth, assume that a reliable automode--or adaptability to partners' automode(s)--is important.

To seed high: Score one ball in auto, high or low. Reliably inbound the ball and immediately dump it to a partner that can shoot accurately. Play shutdown defense. Wash, rinse, repeat. In absence of said accurate partner, dump to someone that can drop in the low goal quickly, and revert to defense.

MCC: Robot with a strong drivetrain, a quick-acting acquisition device, and a reasonable target for inbounding. For a slightly more advanced MCC, add a shooter capable of making it into the high goal. (I would not necessarily rate a shooter as a requirement--the assists are more important--but if you run into a pair of BLT robots on your alliance, it's nice to get 10 points as opposed to one as your base.)


I actually have a robot in mind for MCC, believe it or not. And it's a rookie team that's been in FTC for a few years now, with a couple of members off of the team I currently mentor. FRC5124 sported a 2-wheel-over-fork pickup device that could function as an output, an accurate 1-ball automode, and a shooter with enough power and accuracy to score... but dat surgical tubing do. They also played killer defense. Take off the shooter and you've got a low-goal assist specialist.

[mman1506]

Originally Posted by Chris is me

I would argue the minimum competitive concept did not even need floor pickup. The ability to receive a ball from a human player and kiss pass it / spit the ball out is the minimum that was really needed to be competitive.

Hmmm, reminds me of something......

[IKE]

Originally Posted by BJC

Isaac Rife "IKE" usually posts up a thread like this once a year or so. I think its a pretty interesting thought exercise.
...snip...

Cheers, Bryan

I am often amazed at what one might assume as boring that ends up being insightful. Thanks for reviving this one. The previous years, there are usually a few examples of a couple teams having success with an MCC type concept, but most not having success with a jack of all trades.

I specifically did not do one this year, as to me, it seemed very driver centric/referee centric on what an MCC would be for a particular event. Why I mention refs is, what costitues a possession this year ranged wildly between events (though it was usually pretty consistent at any one particular event). Because of that, I would prefer an MCC there was no room for interpretation, they "possessed" that ball. This year did remind me a lot of 2009 in that it was another year that driving wasn't only important, but often a game changer.

The neat thing is, there were a lot of great examples of it this year, and a lot of teams that did well with it.

I also noticed a trend (at least in my opinion) for the "almost any year MCC".

If you have a decent base chassis (4 cim minimum though I prefer 4 CIM with shifting to get a bit more speed/torque like the Kitbot on Steroids concept), and you have a roller collector to manipulate the game piece, and 1 face that is max competition height, I think you can play elims at most any district.

I added the 1 tall face as this was a very beneficial defense item in:
2014, 2013, 2012 (blocking inbound, not shooters), 2007, 2006, 2005... the rest is before my time.

This year, a tall wall on a well driven base often shut down shooters that weren't ready for defense. It also was used very well towards the end of season as a way to block inbounding students. IE, go ahead and truss shot to HP, I am going to put a 60" wall in your face so you have a hard time inbounding...

And an active roller collector to manipulate the game piece was great for:
2014, 2013, 2012, 2011, 2010-ish, 2009, 2008, 2007, 2006...

Ignore fancy arms, and shooters and catapults and.... If you had something that could collect, move around the field, and release a game piece... I would likely want you within the top 24 at most events in any of those years.
This year, that skill was lethal as you were essentially worth 20 points for being the 3rd inbounder. Also, dumping to the 1pt. with a bunch of "passes" under your belt woften defeated good two team combos in elims (not the majority, but there were a lot of "upsets" this year).

My personal favorite MCC this year though was "Fridgebot". They actually picked a poor strategy (catching only), but executed it so well that they had a good deal of success throughout the season.
http://www.thebluealliance.com/team/5084/2014
https://www.youtube.com/watch?v=vgza-vzG2Fs

[Tom Line]

Originally Posted by BJC

Assumptions are that one of the primary goals of the MCC is to play in elims (not necessarily win on Einstein), and your team has mid-pack to lower fabrication resources.
Please list your assumptions, strategy to seed high, estimate of a winning score, and what robot design elements would achieve this score.[/i]

I'll toss my two cents in later.

Cheers, Bryan

I made this post before reading the others in the thread, because I want to put down some unbiased thoughts. I'll make my comments using my frame of reference right now, because there were some assumptions our team was making early in the season that turned out to be not true.

Initial Assumptions from a mid-tier team point of view:
1. It's unlikely I'll be one of the top 4-5 robots at a given event.
2. Those top robots will be excel at the primary game focus.
3. There are secondary attributes to the game that are crucial to a playoff team.
4. If I can excel at those, I will have a competitive robot that might slip into the top 8, but will definitely be picked.

A 'top' robot will almost always be the finisher - the robot that does the actual scoring. So defining the support roles that need to be played is crucial.

Support roles (in no particular order):
1. Receive the inbound pass securely with little chance of a miss.
2. Perform a quick hand-off, or a quick return to the human player.
3. Be able to pick up quickly from the ground.
4. Have a strong enough shot to be able to truss from most places in the middle zone.
5. Be able to hit your ball in auton every time.
6. Be able to play good defense - some height is helpful.
7. Have a fast/quick drivetrain (not necessarily powerful).
8. Durability

When we were doing our scouting list this year, these were our priorities. Things we saw that made us shy away from teams:

Inbounding: Some teams routinely had the ball bounce out or through their robot. In some cases, if an opponent got between the human player and the inbound robot, the ball couldn't be thrown in. We avoided these.

Hand-off: We put a high priority on robots that could accept the ball and hand it off without turning. It was common for defense to be pounding the inbound robot, and we didn't want a robot that was stuck with a ball because they had to face their intake/output toward the human player up against the wall.

Ground Pick up: We put a low priority on this, because we never wanted to see our ball on the ground. A talented defensive driver could shut down a team once a ball ended up there. Look at what team 27 Rush did to our alliance in the State Championships. However, the team did HAVE to be able to pick up from the ground.

Trussing: Getting the ball over easily with a nice arc from pretty much everywhere was quite important. Almost all our strategies centered on the human player catching the ball. A high arc was a plus. A high-er release point or the event to throw while being defended was very important. We also stayed away from teams that had to lower an intake - they could be defended just by not allowing them to lower that intake.

Auto: You know that by the time you're in the final elim rounds or at champs, missing an auto score is often signing your death warrant. The robot has to nail those shots.

Defense: Defense was a matter of positioning, and that came down to drivetrain speed. Power wasn't all that important - you just had to be quick enough to get between your opponent and where they wanted to be. Many robots weren't able to accurately change their shooting locations, and that meant a defender could shut them down just by parking in the right spot, or by driving back and forth just inside the white line.

That split things up a bit for us: our ideal 2nd robot (inbound and truss) wasn't necessarily our ideal 3rd (inbound and defense). However, if I'm designing a bot I want to be able to truss, so I'm going to design our '2nd' bot.

So, if we were doing a decision matrix for that robot, I think it would look like the one attached, and my "MCC" would be a robot designed around the '3rd pick' column priorities.