On Game Design

[Ryan F.]

Originally Posted by Matt Adams

I would say that most typical small-medium teams are running budgets of 20-30k aren't limited by the costs on building the robot. The $5,000 for the first competition and $4,000 for a second regional, plus all of the buses or airplane tickets hotel arrangements... this isn't cheap stuff. This is prohibitive.

Money doesn't help you build a better robot... or even level the field. Great engineers will do more with less.

Totally true...people talk about all these tens of thousands of dollars people will spend on their robot....that might be nice and all...but it's definetly not necessary. Our team this year had the extra equiptment on the robot valued a little over 1000.... 500 being for donated machining of plates for our transmission. Besides those costs....almost no more money went into the robot. But, we were still able to be very competative...and take 7'th place at our regional.

Ellaborate systems are not needed to make a winning robot. Sure, they may be nice....but once they break down your in a very tight spot. We use simple...but effective solutions, that we know will work match after match after match. In the end, what matters is consistency.

[Matt Leese]

While there are many teams there are a few teams that can build anything they want to out of raw materials, they are in the minority and will almost always be so. However, that doesn't remove the fact that it's a legitimate point that they can build any part that they would want to buy. That said, I don't think that the small number of teams that this refers to should be what the decisions are based around.

In fact, it may not be possible for a team to always build all the components that they want. I remember specifics instances in 1998 and 1999 where we built our base out of steel, not because of strength issues, but merely because of cost issues (this was when the cost limit was around $500). When that becomes the case, the cost issues can become quite a limiting factor. (I realize that for a single gear this isn't a large issue but it does add up over time).

As for robots not looking good enough for television, as long as I've been involved with FIRST, there have always been good looking robots (and that includes times when the price limit was much lower). In my opinion, the winning robots have begun to look worse rather than better the past few years (but this is just my personal opinion). In fact, by including more ways to score in the game, it makes it less likely that FIRST will ever get television coverage (which may or may not be a good thing).

While I normally don't like to include more rules and I'm not even sure if this would ever be a good idea, but perhaps FIRST needs to look into a way of recording the costs of machining time? If FIRST is supposed to approximate real-world engineering in any way, this would seem to be an important part of the program. That said, I don't think any simple system would work nor am I proposing any. It's simply a thought.

Matt

[Steve Shade]

I've been reading the posts regarding spending limits and I see 2 separate arguments. The first is about the cost placed on the teams to buy their parts from their sources. The second is the cost to sponsors for use of machining facilities.

In one scenario, if the cost allowed per robot and materials rules remains around where it is now, many parts can be bought from suppliers by the teams. These parts bought by teams are then charged to the cost allowed per robot and assuming they are within the materials limits, don't cost the sponsor any more additional money. Sponsors don't lose any machine time in creation of those parts and can more easily help with other aspects of the robot.

In the second scenario, if the materials rules and cost rules tighten, and assuming Section 5.3.2.2 still is the same, then teams will be limited by their outside selection and will be forced into custom designing parts that can be readily available off the shelf. Now the teams are responsible for buying the raw material and convincing one of their sponsors to place a decent amount of money into allowing the team to access and use their shop. The cost is now placed on the sponsors has increased significantly (in most cases) because the sponsor is now also paying for additional time on machines, employees to help the teams, and is not making any money with those machines while robot parts are created. Plus that sponsor will be asked to help with the things it has provided in years past!

Going back to the way it was is a huge advantage to the well funded teams. Plus there are many areas of this and other countries where manufacturing capabilities are very limited. All sponsors (large and small) are better off having the teams have access to COTS (Commercial Off The Shelf) Parts than to custom design and manufacture everything. I have no problem with reasonable material or cost limitations, and as Paul said, we are always limited by that nagging 130 lb weight thing.

We want to have this competition attract sponsoring companies to build corporate-school relations. The only way to do that is to make the sponsor's time and money worthwhile.

Steve

[Matt Adams]

Originally Posted by Matt Leese

While I normally don't like to include more rules and I'm not even sure if this would ever be a good idea, but perhaps FIRST needs to look into a way of recording the costs of machining time? If FIRST is supposed to approximate real-world engineering in any way, this would seem to be an important part of the program. That said, I don't think any simple system would work nor am I proposing any. It's simply a thought.

Matt, I just read all of the posts that you've written on this topic. I understand the points that you are trying to make, however, I disagree with your FIRST pillisophy. I'll admit that I haven't been involved in FIRST for as long as you have. We're from different eras.

From your posts, What I am seeing is that you would like to have a game that is like the following:
You're in a 24' x 48' arena. 2 blue robots against 2 red robots. There are 30 soccer balls across the field edges. There are two 5' wide, 4' tall stationary octagon-shaped PVC goals, designated blue and red. Each ball is worth 5 points when you get it inside the goal. Most points wins.

This is a straightforward game, the rules are simple, and everyone has the same objective. Get the balls off the floor, put them in the goal. The tasks aren't trivial, but they're not difficult. The machine with the most consistent ball gatherer and unloader should win every time. You don't run the risk of picking the wrong "key of the game."

Personally, I find that boring.

And here's why:

Strategy.

After I saw the game released this year, I knew that there would be no robot that could do EVERYTHING consistently. One or two teams came somewhat close.

Everything:
Grab the goal
Climb the 6" steps
Knock off the 10 point ball
Hang from the bar
Fully secure the big ball from the floor, small goal, and big goal.
Pick balls off the floor
Capture the balls falling from the drop chute.

I personally would rather see teams make their engineering choices on their game strategy as opposed to the costs of components on their machine. You won't be able to control the dollars spent unless you limit FIRST to a single distributor and have a mile of bookkeeping. To be honest, the $3,500 limit appears to be more of a gesture than a true check. I think that the dollar limit really just ensures teams aren't using immense amounts of titanium. Limiting machines down to a much lower dollar amount just won't make the impact that you're looking to see, because the amount of money that teams typically spend on their robots only makes up between 10% and 20% of teams' budgets as it is.

Money is not what limits teams' success. Experience does. It always will.

Originally Posted by Matt Leese

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.

I can tell from this comment that you don't like the idea of hanging, it's very 0 or 1. Hanging or not hanging. 50 points or 0 points.

I'm not so sure that I'm a fan of this either... especially in 2003's game where being on the ramp was the deciding factor on if you won the match or not in 95% of the cases. However, the points were balanced this year. 50 points was appropriate. 75 would have been too many. It should be noted that many competitive alliances had robots that did not hang, and they were very successful.

I do not think this year was complicated. There were no complex math formulas. You could read the points on the field. 5 points for the purple balls in the goals, you double the ball values if you cap the goal they're in, you get 50 points for hanging on the bar. For anyone attending a competition, this was easy to explain.

I think that restricting the "everything" list to one or two things will really limit the objective of this program - inspiration. Seeing robots twice as capable as your own with the same restrictions is awe-inspiring. I think FIRST wants to see more students with their tongues wagging over something incredible.

Loosing to incredible machines is okay in my book.

I think that rookies usually fail when they try to bite off more than they can chew. Hanging was a tough task, and many rookies who tried to do this via climbing up the steps had two seriously challenging things to do. This was perhaps the wrong choice for many rookies.

But I bet they learned a lot.

And I bet they'll be back next year.

Matt

[Matt Leese]

I think I haven't come across clearly enough about my philosophy of how the game should be designed. I honestly thought the points for this year's game were well balanced. However, I do believe that there were still too many ways to score and that the difficulty in the various tasks was too great. I don't like the idea of extremely hard tasks (hanging this year after climbing steps) nor do I like the idea of easy tasks (ball herding). I don't think hanging is bad in and off itself (I have no problems with scoring a lot of points via one method as long as it's not out of balance with other scoring methods).

While I originally (meaning last year when I first started thinking about this) thought that one method to score was optimal, when talking to Aidan Browne at the Championship this year I realized that it wasn't a good idea. I think that there should be two main methods of scoring that have balanced difficulty and balanced points (think the balls and hanging from 2000 or the balls and goals from 2002). Two methods of scoring gives plenty of opportunity for strategy while not making the game too complicated.

As for the idea that most teams don't spend a significant amount of their budget on their robot, I highly beg to differ. Most young teams do not have large financial support. I've been on several of them. Many teams have a budget under $10,000 a year. These are the teams that I think are under a significant disadvantage under the current system that could be rectified. The best way I can think of to make it more fair for these teams is to lower the cost limit.

Matt

[P.J. Baker]

I don't like the idea of extremely hard tasks (hanging this year after climbing steps) nor do I like the idea of easy tasks (ball herding). I don't think hanging is bad in and off itself (I have no problems with scoring a lot of points via one method as long as it's not out of balance with other scoring methods).

I don't accept that herding small balls was easy. Without capping, your HP needed to make 10 shots to offset a hanging robot. Many teams thought they could accomplish this by simply pushing the balls around, but had trouble with it. The only ball machines that I saw make an impact were the catchers or gatherers. I would describe both of these as intermediate tasks

I think that there should be two main methods of scoring that have balanced difficulty and balanced points (think the balls and hanging from 2000 or the balls and goals from 2002). Two methods of scoring gives plenty of opportunity for strategy while not making the game too complicated.

There really were only two ways to score this year since the 2x ball meant nothing without the small balls. There were many tasks (hang, cap, herd, mobile goal manipulation, defense, etc.), which I feel made it more likely that both teams in an alliance would have to contribute for a win, rather than having the "dominant" team in an alliance take over the match. I'll also point out that the 2002 game had three ways to score (robot position, goal position, balls in goal). What it didn't have was very much strategic variety - control the goals and you probably win. I disliked that game, the combination of the "mouse bots" and the fact that a match could be effectively over in the first 10 seconds made it a complete disaster in my book.

As for the idea that most teams don't spend a significant amount of their budget on their robot, I highly beg to differ. Most young teams do not have large financial support. I've been on several of them. Many teams have a budget under $10,000 a year. These are the teams that I think are under a significant disadvantage under the current system that could be rectified. The best way I can think of to make it more fair for these teams is to lower the cost limit.

I hear what you are saying, but the logical conclusion of limiting the build budget is that a very small number of teams (as you pointed out previously) will have a huge advantage. Don't discount the psychological impact of knowing that you simply can not compete with the elite few because they can make parts that you can afford to buy but are not allowed to.

P.J.

[Matt Leese]

Originally Posted by P.J. Baker

I don't accept that herding small balls was easy. Without capping, your HP needed to make 10 shots to offset a hanging robot. Many teams thought they could accomplish this by simply pushing the balls around, but had trouble with it. The only ball machines that I saw make an impact were the catchers or gatherers. I would describe both of these as intermediate tasks

Terming it as "easy" wasn't really my wording. From what I was told by Aidan Brown, FIRST thought that herding balls would be something "easy" that rookie teams could accomplish. I agree that it wasn't nearly as easy as it appears at first. That said, there were a number of teams that made a difference with balls that didn't catch or gather them explicitly (there were some on the Newton Field at least).

There really were only two ways to score this year since the 2x ball meant nothing without the small balls. There were many tasks (hang, cap, herd, mobile goal manipulation, defense, etc.), which I feel made it more likely that both teams in an alliance would have to contribute for a win, rather than having the "dominant" team in an alliance take over the match. I'll also point out that the 2002 game had three ways to score (robot position, goal position, balls in goal). What it didn't have was very much strategic variety - control the goals and you probably win. I disliked that game, the combination of the "mouse bots" and the fact that a match could be effectively over in the first 10 seconds made it a complete disaster in my book.

When I say two was to score I don't mean literally two ways to score (I realize that's rather unclear). What I mean is that there are two "major" ways to score. Putting your robot in a certain position in 2002 I wouldn't qualify as a major way to score as it's a general assumption that your robot needs to be able to move to score. There were definite problems with the 2002 game, I'm just trying to highlight some of the things that went right with it (in my opinion).

While with so many ways to score this year, you'd think that both teams in an alliance would have to compete, it didn't appear that way to me. At least from watching the Championship finals, it seemed that one team of the alliance would immediately go and hang and the other alliance member would then proceed to attempt to herd small balls, cap with the big ball, and then go hang. That doesn't seem like a fair dividing of resources so I wouldn't say that a more complex game helps make all the teams participate.

I hear what you are saying, but the logical conclusion of limiting the build budget is that a very small number of teams (as you pointed out previously) will have a huge advantage. Don't discount the psychological impact of knowing that you simply can not compete with the elite few because they can make parts that you can afford to buy but are not allowed to.

Unfortunately, I think we're already at a point where there are teams that don't feel they can compete with the elite few because they can't afford to purchase the parts or design a complex system that purchased parts would allow. I think there's a wider gulf between teams because of the ability to purchase such parts than there would be otherwise. There would be a more natural gradation as opposed to a huge step between the contending teams and the non-contending teams.

Matt

[Adam Y.]

Unfortunately, I think we're already at a point where there are teams that don't feel they can compete with the elite few because they can't afford to purchase the parts or design a complex system that purchased parts would allow. I think there's a wider gulf between teams because of the ability to purchase such parts than there would be otherwise. There would be a more natural gradation as opposed to a huge step between the contending teams and the non-contending teams.

I think our team is one of those gradiation. Once we got our robot actually moving it was actually quite competitive for such a simple ball hearder design. We could beat teams that hung just because of our one design. That design had to be one of the easiest designs to build. Also, you are doubting the creativity of some of these teams. I saw robots which clearly didn't have the resources that other teams had but still appeared competitive. I remeber seeing one robot almost entirely built out of wood. They said it worked perfectly and could heard five balls and cap a goal.

I don't accept that herding small balls was easy. Without capping, your HP needed to make 10 shots to offset a hanging robot. Many teams thought they could accomplish this by simply pushing the balls around, but had trouble with it. The only ball machines that I saw make an impact were the catchers or gatherers. I would describe both of these as intermediate tasks

That was mainly due to the fact the way the balls and field were set up. My one advisor made the point that the balls were nearly impossible to push in there since they would squish up the lip of the opening and not go in.

[Ryan F.]

Originally Posted by P.J. Baker

I don't accept that herding small balls was easy. Without capping, your HP needed to make 10 shots to offset a hanging robot. Many teams thought they could accomplish this by simply pushing the balls around, but had trouble with it. The only ball machines that I saw make an impact were the catchers or gatherers. I would describe both of these as intermediate tasks

Simple solutions can work wonders. True...those methods of ball collection were great...but there were some simple things that you could do to make your robot herd the balls. This year...we used some 1x1.5 alluminum tubing....and made a bumper that rode on the ground in front of out robot. The balls would hit it....and it would create backspin on them..keeping them with the robot. It worked wonders, and allowed us to herd very well without a complex mechanism.

Sometime I think people really need to remember the difference between theory and reality. Complex mechanisms may work great....but sometimes a bar of alluminum is all you need.