On Game Design
 

Over the past couple of years, I haved made some observations on the direction FIRST seems to be headed. FIRST has changed the overall direction of some of the rules and the game design in this time span. In some ways these changes have been positive but I've started to believe that there were some unintended consequences that were derived from these changes.

First of all, we've seen that overall the games have gotten more complicated over the past few years (last year with the stacking, this year with the hanging). By complicated and complex, I mean that it requires a complicated or complex robot to successfully compete in the game. This was done with the stated intention of giving more of a challenge to veteran teams but still allowing a "easy" component to the game for rookie teams. While I do find it a bit condescending to leave an "easier" component for the rookie teams, the bigger problem is the fact that it divides the competition into several levels of play. In some ways it's saying that if you're rookie or choose one objective that you aren't eligibile to actually compete.

One of the other major issues that I've seern is tied into the idea of more complicated games. Over the past several years, FIRST has dramatically loosened the parts requirements. This had made it much easier for more complicated robots to be built. Because it's easier to build complicated robots, FIRST has had to make the game more challenging to keep up. While I do appreciate some of the well designed systems that teams have come up with, I think that it helps to further divide teams between those who are able to compete and those who are not.

Because of this added complexity, some teams spend their off-seasons designing two-speed drive trains and other components. While I saluate teams that are able to successfully create a year-round program, it shouldn't be a requirement to be competitive. FIRST knows that many sponsors are not interested in year-round programs and has attempted to address the issue; by making it easy to design systems that can be used on the robot outside of the build period (specifically, I look at various drive trains), FIRST is not helping reduce the reliance on year-round programs.

My proposed solution to these problems is fairly simple. For one, make the complexity of the game less. It's not a problem to raise the bar but don't make it too high for most teams to compete. However, simply making the complexity of the game less would not be enough as it would just allow teams to dominate. The real key is to also drastically restrict the allowable parts. I would highly suggest dropping the allowable spending amount to below $1000 (I think somewhere between $500 and $750 would be a good point). This has many benefits. For one, it does not stop anyone from building a complex robot; it merely makes it harder and would require more ingenious solutions. It would also require more tradeoffs to be made if one complex component were desired by a team. It also drops the required amount of fundraising a team has to do as it wouldn't be possible to spend as much money on the robot.

As for reducing the complexity of the game, I would make several other suggestions to FIRST. For one, keep the methods of scoring to two ways that require specific mechanisms to accomplish it (i.e. moving your robot to a certain area of the field wouldn't count as a method). For games that did this well, I'd look at 2002 and 2000. This makes it reasonable to either make a specialized robot or a generic robot. The key is to also keep the difficulty of the two methods not too divergent or one method will be ignored. The other key is to keep the methods of scoring somewhat similiar in point values (2004 and 2000 did this very well; 2003 did not).

I do not know if it's desired by FIRST for everyone to compete on an equal footing (which, while truly impossible, we can at least attempt to approach it), but I would say it should be. The key to doing this is to make the games such that even a rookie can successfully do any of the tasks. The corollary to this is restricting the parts availability rules such that teams have to engineer from scratch every year to add everything they want to (this means that teams can't carry over designs year-after-year as easily and instead have to tailor their design to each year's game).

Matt

Re: On Game Design
 

Matt-

Good ideas, but as always people will find ways around them. Currently you could have $100,000 worth of machining on your robot as long as it was donated by a sponsor. This would only keep the price of raw materials under $750-$1000

Also, given as FIRST implicity states that the program isn't about the competition, it's about inspiring students, I don't think they will be taking many steps to "level the playing field" beyond what they've done now.

Cory

Re: On Game Design
 

Matt I'd have to debate you on several of your assertions.

First, on the point that the game has become more complex. On the contrary I would say that the requirements have changed such that any robot with a working drive train can score points. Look at '97 (placing inner tubes up on a high goal), '98 (placing balls up on high bars), '00 (placing balls in a high trough AND hanging. If you couldn't do any of these things, your contribution to the score was ZERO! At least with the last 2 games a rookie team can get their robot moving and contribute some points - they can have the goal of building a mechanism to take it a notch further but it allows every team to compete and get some satisfaction.

Second, on loosening the parts requirements I would disagree it allows more complicated robots to be made. I would contend that it allows more teams to make complicated robots. This is very much tied to the allowable cost amount - I think having a larger amount helps the rookie teams and small teams much more than the highly funded big teams since they are usually much more engineering and manufacturing resource limited. Consider if you had to design and build your own winch versus going out and buying one, or how much additional complexity it would require if you had to work around sloppy bearings versus precision ones. And as a veteran if I've already designed a lift or arm before I can crank one out alot faster, regardless of what part limits are in the game.

I sometimes refer to this as the "SECOND" competition, because it takes your entire first year (and then some) to figure out all the lessons you need to be competitive. Are teams quitting after their rookie year because they didn't win the championship? No, they become veterans who do better each year. I loved the game this year. We chose not to build a 2x ball handler this year because of resource limitations, but there were enough aspects of the game where we could focus our efforts and be very competitive.

Re: On Game Design
 

In 2000, robots that were on the ramp, but not hanging, also received points. A robot that moved was capable of scoring points.

Re: On Game Design
 

Matt-

I agree with some of your observations, but I don't think the solution is to give the game an easy element. I think thats what they tried in 2003, and the free market gave us a whole bunch of talented "boxes-on-wheels" and few stackers. In my opinion, when you have experienced teams building not much more than a drivetrain, its gonna be a boring slugfest.

Everyone loves the 2000 game. It was (and still is) my favorite. It did not have an easy element (yes I know I am ignoring simply sitting on the ramp, but there were few boxes on wheels tha did just that). 2004-style herding is easier than 2000-style hanging or 2000-style ball-bin-scoring.

I though this years game came close to 2000 in terms of the entertainment factor. Sure, there were a lot of teams who bit off more than they could chew (we did, the 2X ball grabber is on the "wall of shame" in our shop), but I would say that that penalized the experienced teams (who wanted to do it all) more than the less-experienced teams (who wanted to be great at one or two things).

Drive systems are going to evolve regardless of the game, in my opinion. If a team focuses too much on a specific drive system, it might hurt them when they see in January that the new game makes parts of their idea obselete (for example how important was it this year to have a multi-speed drive as compared to a platform climber?)

Ken

Re: On Game Design
 

My issue with the games isn't that individual components are getting more complex (if anything, the more recent games have one part that's easier than they were in the past). The problem I have is that there seems to be a part of the game which is much harder. The veteran teams (and, unfortunately, I wouldn't include all teams that have been around for a number of years as veteran teams as many don't reach the level of sophisication of other teams) usually have no problems accomplishing these tasks. It's the non-veteran teams who usually aren't able to complete them. You simply aren't able to win by just having a moving robot (or at least you can't with any level of competition). We shouldn't be encouraging rookies to just build a robot that moves. We shouldn't make the tasks so hard that rookies fail. We should instead aim at the level where it's not accomplishing the task that's the differentiator but how well the task is accomplished.

You're right by saying that veterans will always have an advantage over rookies. That's completely true. However, we can work to lower the some of the advantage that the veterans have over rookies.

As for the parts rules, I don't think it will take additional complexity if you have to make more of the things yourselves. Instead, I think teams won't try to build as complicated robots. If I'm not able to use precision bearings where I'd like to, I'll have to design something that doesn't require precision bearings (i.e. a simpler device that doesn't require such small tolerances). Or, I'll have to save money somewhere else by reducing the complexity there so that I can use precision bearings where I want to. By reducing the price limit it requires teams to make more trade-offs over expensive parts so that there will be less expensive parts overall. Less expensive parts generally means less complexity.

I should say that my suggestions by no means will be a silver bullet. They aren't going to fix everything but I think they would be a step in the right direction.

Matt

Re: On Game Design
 

I don't think there should be an "easy" element of the game. Perhaps my original post wasn't clear enough in that regard. I think that there shouldn't be one easy element and one hard element. I think there should be two medium difficulty elements. I don't like the easy and hard element mix because it tends to differentiate teams into those who can compete and those who cannot. Or, worse, you end up in a position like 2003 where you pick one element to design against (stacking) and find out that you can't actually score points that way (reliably) and you're basically done competiting for the year at that point.

I agree that drive systems will advance but there's no reason to make it easier. And there's also no reason to allow someone to easily build a very powerful drive train with additional robot elements. Allow a team to choice either a powerful drive train or additional manipulators.

I single out drive systems because there seems to be a lot of emphasis on them, particularly in the off-season. I also think that it's harder to get students involved when the robots get more complex (although it's not impossible but the more complex the robot, generally the harder it is to get students involved).

Matt

Re: On Game Design
 

This is a little OT, but I feel relevent to the quote...

I believe you're seeing veterans improve their drive trains in the offseason because it's the one thing you can standardize. Most teams run fine with the same drive train every year. By standardizing on mechnisms it gives teams more time to concentrate on the new changes in the game.
You would think that this put veteran teams to a distinct advantage, and in some ways it does. However, FIRST responded pretty well to this by providing a basic drivetrain in the kit of parts.

Right now my team is in the process of standardizing a drivetrain. For three reasons: we found one we like, and we just don't have the people power to re-engineer every mechanism each season. The final reason is because FIRST has raised the bar and requires synergy between teams (mostly drivetrain and control systems) that just can't happen in 6 weeks. Our autonomous was poor the last two years because we haven't put offseason time into it. By standardzining on a drivetrain and complimenting it with a control system we will have a more competitve autonomous mode.

Re: On Game Design
 

To expand in another direction from Ken, from my vantage point I saw the rookies this year do much better overall than in the previous few years (not to offend any rookies that did well in the past). By making the big points scoreable with mechanisms, nearly all of the rookie and 2nd year teams that created a "box on wheels" robot a year ago or the teams that have never built a robot before, attempted and in many cases succeeded at getting their mechanism(s) to work. Heck, there was a rookie in the runner-up alliance in Atlanta. The teams this year stepped up to the challenges placed in front of them, much the way the young teams did back in the late 90's when ChiefDelphi boards were in it's infancy, and design sharing was just getting started. Since life isn't fair and neither is FIRST, I believe the game this year sufficiently challenged the veterans and the young teams alike, and is one of my favorites along with '99. The only way for the young teams to be able to strive towards the harder tasks is to start experimenting early, and continue to innovate new ways to solve the same technical problems. Many of the challenges this year were solved with very simple mechanisms and that is one of the key lessons for any team to learn.

Keep the challenges coming, as long as the game points fit into a reasonable scheme.

Steve

Re: On Game Design
 

I don't see the game as too complex.

This comes from being with a team that is in its second year and could complete all task each of the last two years. This has let us be in the division finals each of the last two years. Fairly good for how untechnically advanced our robots have been. We've used sprocket and chain drive and have not done a lot in the off season.

Hopefully we might consider doing more this summer because a new drive system would be nice, but a very simplistic one has worked great. Our biggest asset is that we try to live by KISS. The only machined parts our the hub for our tires. Our most complex part we've purchased is taking the impeller from a shop vac. With strategy and a fairly robust design you can do well. You may not be able to do all the task well (we only made stacks twice last year, and barely herded any ball this year), but you can compete.

I like the rules how they are now. It's not going to be completely fair but it won't stop people from competing. And if you are only in it to compete and win you've got to realize you can't always but just take advantage of the learning experience.

As to FIRST becoming a year-round program. It is. So much needs to be done before that 6 weeks unless you are just crazy. You can do it in six weeks and be successful but I wouldn't reccomend it. Get fundraising, and paperwork done early. Our team spends a lot of time in the community during the off season.

FIRST is definitely like a sport. I played three sports in high school. I concentrated on basketball during basketball season but that doesn't stop me from shooting around during the rest of the year. Some just take it farther and play tournaments and spend a lot of time praticing their skills. You can be competitive just playing during the season, but those that put in the time all year will have an advantage.

I don't think changing the game to two intemediate task or letting people spend less is going to change anything. IMHO.

Re: On Game Design
 

The complexity of the games doesn't hurt the rookies at all. Even though it does seem that the games are more complex, it also seems that the easier component of the game can beat out the harder one. For this year's game as an example - hearding balls and capping could easily beat out hanging robots. Hearding isn't that hard and capping takes a little more work but isn't as hard as hanging. I did see a lot of rookies do well using these methods. As for last year's game - though stacking was hard to do, it was almost irrelevent. I still remember the match in Chicago when 16 was making a stack. I think they spent a good 30 seconds on that. The second that top box was put on, there was a robot charging at it to knock it down. While rookies didn't have the good drivetrains that were necessary last year, they were still able to give the veterans a run for their money. I don't see the design of the games inhibiting the new teams ability to compete. I think every rookie has a good shot at winning as long as they can do one thing and do it well.
Eric

Re: On Game Design
 

I think this is exactly one of the points Matt is making- it does put veteran teams at an advantage, even over other veteran teams that do not standardize. The advantage is exactly as you say- you spend much less time creating what is arguably the most important aspect of your robot. In most cases I like to think virtually all the design that goes into a robot should be done within the 6 week limit, not the 8 months between competition and the next build season. Otherwise, you may as well fabricate your base and drive train, and simply plug in your motors and electronics as soon as you get the kit. Slap on a goal/bar/ball grabber and you have 5 weeks of practice time.

FIRST relies on our gracious professionalism to ensure we don't do that sort of thing. The 6 week limit is there for a reason, and I feel to work on any part of a robot outside of competition or that period defeats the purpose of having a set time limit. I think that's why in last year's game we didn't see too much stacking- it was a new thing no one has seen before, so no pre-proven or optimized mechanisms were around. Adds to the challenge aspect of the FIRST experience.
That's exactly the purpose of the 6 weeks- it's a challenge.

Re: On Game Design
 

What!??!?!?!? Thats probably our entire budget that we would spend on metal let alone the whole entire robot. And that money is what usually amounts to us build a robot that looks pretty professionally done. It would probably make a lot of teams life's harder.

Re: On Game Design
 

The idea isn't supposed to make lives easier. It's supposed to make everything more balanced. It means that you have to make more tradeoffs on design (which is a key factor in real life engineering).

And while you may think that it's impossible to build a robot for that price, that number is far higher than the amount we were allowed to spend at Small Parts back in 1998 and 1999 (and probably years prior to that). And yes, you had to purchase almost all your materials from Small Parts (including aluminum, etc.; there was a small list of things you could purchase outside of Small Parts).

Matt

Re: On Game Design
 

Don't forget the Human element. The great thing about this year was that a human was the one making the points. Our robot had massive problems (which were finally fixed LAST NIGHT) with the drive train and was never able to run the entire UCF regional. However, the team was able to recruit a high scoring human shooter, and made what little points we started with, and the points our alliance partners generously let us have. With some luck, a good shooter, and no robot, we finished in 21st out of 41 teams. Surprisingly, if our goal was capped, the enemy alliance goal is capped, our alliance hangs one robot, and the enemy alliance hangs a robot...the game comes down to whichever team scored the most 5 point balls. Rookie teams have just as good of chance recruiting a human player as any other highschool.