Strategy - Your method??
 

With the season quickly approaching I was just thinking of the best ways to analyze the game and pick the best strategy. Do teams go by the numbers and design a bot that will achieve the task that gets the most points? If so, how do you figure which task is most important? Or do teams analyze the game, try to figure out (using numbers and some intuition) which tasks will be most important for a winning bot to achieve?

Last year my team convinced ourselves that being able to stack would be the most important task, followed by going under the bar and being on the ramp at the end. While being able to stack was important it was also near impossible and in the end the mechanism was replaced with a box pushing arm. Is there any way to avoid making strategical mistakes like this?

Re: Strategy - Your method??
 

Be only as complex as necessary. Don't get me wrong, certain features of your robot might be very complex, but the robot as a whole should be simple. Design it to do what it is meant to do. That being said, from my perspective, there are two types of robots. Robots meant to achieve high scores during the qualifying rounds, and robots meant to be moderately successful in the qualifying rounds but shine through in the elimination rounds. FIRST Robotics, whether purposely or inadvertently, was designed in such a manner that the two robots often complement each other. Sure, the robot that is designed to win during elimination rounds is an obvious pick, but it is being picked by the robots that played well during the qualifying rounds. From this perspective, any robot, whether defensive or offensive, whether designed to achieve high scores or be useful during the elimination rounds, can be successful, as long as it does whatever task it intends to do, and does it well. The task to do is relative to the robot you want to build. If you wanted to score high scores and be a top pick, this year one would want to build a stacking robot. If you wanted to be useful during the elimination rounds, a robot that can clear boxes and play defensively, then take the hill at the end of the match would be a great robot for the task. In conclusion, the task that you choose to accomplish is relative to the robot that you want to build. Be warned, though, it is quite often disadvantageous to build a do-everything robot.

::sigh:: Not quite sure if I nailed that question dead-on or said something completely tangential.

Re: Strategy - Your method??
 

well.. you gave me something to think about. maybe i should restate my question i'll be a bit more specific.

In last years game, how did your team decide to build what you built? Was it a breakdown of points? Or did you try to build something unique so that you'd get picked? Or did you look at the game and say "Well.... this is going to be important, lets build a robot that does it..."? Or did you......?

One thing that I am really interested in is for those teams that go by the points/numbers, how do you do it?

Re: Strategy - Your method??
 

The first thing that I would tell you is that typically in the games that FIRST creates, there are multiple ways strategically to play. Our team was also a stacker bot last year, and we were very proud that we designed a robot to play the game. In every game, there are 2 different kinds of robots that can be designed. Your team can either select to be an offensive or defensive machine. Our team can pretty much always be counted on as one that will be there to be offensive.

The best way that I can give you to help decide what type of a machine to make, is look at the game. Try to have each team member write down on a piece of paper, how in their mind they imagine a typical match being played. Then, go around in a circle and have everyone share their ideas. Once you see how each person imagines a match, then you can see what types of features each thinks are neccessary.

One of the most important things to remember is that no matter how hard you try, you aren't always going to succeed. It is not a surprise to have an extremely high amount of confidence in your robot before competing then get to the competition and realize that, oops, guess we should have done it a little different.

My best advice, find a design you like and do it, but don't be afraid to make adjustments if you find that something you designed didn't work. Last year, we had a box flipper. After 1 or 2 regionals, it turned into a device to pull ourselves up. Good luck! :D

Re: Strategy - Your method??
 

the team i was on decided what we wanted to do in the game and then prioritized those tasks....our opinion last year was that stacking was a suckers bet and would be too ineffecient in the time available, also we decided we wanted to be the first to the bins every match so we designed a method to do that....we then tested and tested and tested, modify then tested more (went through about 9 major revisions of the robot). Only we took our drive train for granted and didn't put enough work into it therefore we had problems throughout the year.

We figured when the whole wall is on our side within the first 6 seconds it more or less put the other teams on the defensive quickly.....

Re: Strategy - Your method??
 

Short Answer:

Decide WHAT you want to be able to do - taking into consideration game rules, scoring , and devising ways to try and ensure a win.

Prioritize those items (we all know how fast six weeks go by)

Make them happen - the HOW comes from your strengths, knowledge, resources, budget

Have FUN

Re: Strategy - Your method??
 

right. What I'm looking for is HOW teams decide what they want to be able to build. The method used on our team is pretty much this: The mentors (FIRST veterans who are now college students) are home for break and go to a kickoff event. We watch it on TV because there is no event close by. Our mentors talk about the game and the robot on the trip home and when we meet on the Monday after kickoff they tell us what they think, it makes sense to us so we build it. This year I am going to try to involve the students more in the design/decision process. I am looking for ways to analyze the game and come up with strategies.
thanks.

Re: Strategy - Your method??
 

It's surprisingly hard to get a grasp on the game. Most people thought that stacking would be huge last year. Though successful for some teams, this generally did not work as well as most hoped. It's hard to analyze the game until you see trends in other robots, matches, and just how the game is played. Try as you might, there's no guarantee that you'll be able to analyze the most effective strategy and build an effective robot to suit it. Your best bet is to build a robot that will be effective in whatever strategy you deem to be most likely to be effective, but have it excel at another simple task. I.e. If you were building a king of the hill bot last year, make it able to plow boxes effectively. It's always nice to have a backup plan.

Re: Strategy - Your method??
 

Our robot was designed to do everything - stack, go under the bar, knock over stacks, knock down the majority of the wall on the ramp during autonomous and king of the hill -ing. Our problem was that we were not able to do one thing really well. We qualified for the finals (our award), but most of our match performances were not very spectacular.

In my opinion, it is VERY important that your robot is able to perform a specific function very well. Afterwards, when you have a solid system designed, then you can flare it up with function. Any robot can knock boxs and take up space on top of the ramp quite simply by being able to drive around.

Picking that one or two function(s) on which your robot design will focus on should be dictated by your teams technical experiance. If a robotic arm will yield the most points and none of your team mates have any experiance with pneumatics or welding then its out of the question, settle for the next best function. 6 weeks is a very short period and it is a time in which you apply your knowledge as oppose to 'learning as you go' with overly ambitious plans. Its the time between September and the kick off where you try out something new [advanced multispeed drive systems - oops!!, guess you'll have to wait till the Canadian Regional (and Atlanta, lol)].

Judging by the past few years, and Houston, a solid drive system is manditory. If your robot can't push its way past opponents or get to the other side of the field quickly, then whats the use of having all the neat functions that you've worked on for the past 5+ weeks.

Our team spends the next 36 hours after the kickoff (the rest of the weekend) at someone's house brainstroming ideas. We give ourselves 5 weeks to design and build the robot. Construction should be finished by the end of the 5 week period and the robot will be tested. You won't know how well your design works untill you test it (preferably on a mock field and with pervious years robots in the way). !!!! Don't wait until your regional to find out that your multispeed transmission is locking up !!!!.

There are many rookie teams and teams which have little technical experiance. Expect that they will build a robot which will perform the simplest task possible - move and and knock over stacks - and not much else. Provided that your team can do it, you would probobly build a stacker. During heats your have your team mate defend you as you stack, avoiding low scoring matches where all the boxes where cleared off the scoring zones by the pushers.

Further advice from our misfortunes:
-when an idea is presented, have the presenter provide scale drawings
-if you need 4 pieces of something, make 6. Making spares at that time saves alot of time and keeps thier dimensions very simialr to the originals.
-have a solid electrical setup, our robot stalled many times and we lost countless heats because of that.
-encase sensative components in plexiglass/lexan.


Good luck this time round and we hope to see you in Atlanta.

Re: Strategy - Your method??
 

Look at all aspects of the game. Try to see what other teams might do. Try to understand the complexity of all parts of the game. Most teams will try to do 1 or 2 things well. Last year we looked at how hard it was to stack and how easy it was to knock over. Why waste 1:30 of a game stacking 5 boxes when another team can knock them over in 1 second. Our human players can stack faster so we let them. As was stated above, a good drive train is VERY important. There is no use in being a bulldozer if you don't have the power to push. We also built a ramp and used humans to simulate designs and test how boxes fall. we found that it was very tough to move around with a lot of bins in your zone. From this we decided we HAD to be first to the top of the ramp.

I hope this process is what you are looking for.

Re: Strategy - Your method??
 

You dont need a bot that scores as many points as possible. Heres how my team approaches it:

1) whats the most useful strategy. Now heres what I mean by "useful". For elimination rounds, teams picking alliance partners wont necessarily want the highest scoring bots, they'll want robots that can do what they cant, and can be the most help in a match. With 2003, a stacker bot isnt much use to someone who can already stack, KOTH bot would be much more useful. So what we do is think what the most popular strategy will be, then do something else. Had we been involved in 2003, our team would have been a box pusher/ stealer that would have gone under that bar. Why? Because we figured most stacker bots wouldnt be able too, and it's quicker than the ramp.

2) we also take into consideration whats strategy is the simplest to build a robot for. A bot that can limbo and flip a bin over the bar is (probably) much simpler than a big stacker bot. And usually, its easier to build a single-task robot thhan one that can do everything. Its easier to one thing well than it is to do everything.

So thats my advice, the most useful, simplest strategy that you can do well.

Re: Strategy - Your method??
 

our number one rule in building a bot is KEEP IT SIMPLE. the simpler the bot is, the easier the repairs and maintaince. of course if your a verteran team and you know what you're doing, then this does not apply as much, but newer teams should keep the design simple.

our second goal to designing a robot is to HAVE A SOLID DRIVE-TRAIN. our first step in building a robot is to make a good drive train, so at least our robot will mobile. even if you have a great design that works great, it is worthless if the robot doesnt move. however, if you have a solid drive train and the rest of the robot isnt so hot, at least it will be able to move durring the competetion, and a robot that can move allways serves a purpose on the playing field.

i truely feel that if you follow these two guidelines in building your robot, you will be sucessful in the playing field.

Re: Strategy - Your method??
 

i forgot to mention a few things in my other post...

FIRST usually creates a game where there are two main strategies on the playing field. for example, in the game last year, the two ways to go were to either be a stacker, or a tote handler. the year before that, you could either be a ball collecter or a goal handler.

my suggestion is to build a robot that specializes in one thing and only one thing. build a robot that can complete one task, and do it very well. for example, in last years game, the teams that specialized in either stacking or handling did better than the teams who tried to accomplish both tasks. if you are a newer team, i strongly suggest building a robot that can complete one task and do it well. trying to do too much with your robot can be overwhelming and lead to disappointmet to the newer teams.

of course if your team has been around for a while and you are confident in your abilities in building a strong robot, go ahead and build a robot that can do multiple tasks. those are the teams that are fun to watch, and those are the teams that add another level of excitement to the arena. :D

Re: Strategy - Your method??
 

I think that our team has (after SEVERAL failures) finally began to get good at this aspect of the competition. Last year we built a stacker that had a quick autonomous mode and had good pushing power to end on the ramp. The only major design flaw we had was not making arms that extend horizontally to hit more boxes in the auton mode.

We've determined that each year there is always ONE thing that all winning robot alliances MUST do. Last year, that task was hitting the boxes at the beginning during auton mode; in 2002 it was being able to grab, maneuver and defend the goals; in 2001 it was balancing the bridge.

Our strategy is simple: Figure out what this thing is and make a robot that does it extremely well. Once that's accomplished, augment the bot with extra features capable of performing other tasks.

Re: Strategy - Your method??
 

Hey Bill,

Good Question. Your team may want to decide how they'll play the game BEFORE they start talking robot design. It is therefore important that your design group know the game rules and restrictions backward and inside out. {As you probably know, a surprising number of team members don't make reading about the game a priority} A long hard discussion about all the possible permutations - all the scoring possibilities, all the 'traps' you may get into, and every possible aspect of the game should take place. Critique all your choices constructively, looking for strengths and weaknesses. Easy to say, hard to do. No magic. No single answer.

With your game strategy in place, choose machine designs based on how much engineering support and student commitment you have. Any number of designs can and will win, some with luck, but if your strategy works, and your robot does what you tell it to do consistently, you'll do well. Again, easy to say, hard to do.

One last comment.....your drivers will be a key factor in the outcome of your season. I have seen great drivers win with an average robot, and poor drivers lose with a great robot. Drivers need to be "strategy-smart", extremely well-versed in the game, able to adapt to anything happening on the field calmly. Some drivers may be fabulous looking in practice, but if they allow unexpected events during competition to mess with their heads, then they are not the right drivers. If your drivers don't want to listen to your field coach, then they are not the right drivers. Sometimes, the "most obvious" drivers will not the the best drivers.