FRC GAME ANALYSIS AND DESIGN STRATEGY Caleb Sykes

FRC GAME ANALYSIS AND DESIGN STRATEGY Caleb Sykes Mentor, FRC 4536

ABOUT ME Caleb Sykes Education Currently an undergraduate student at the University of Minnesota pursuing electrical engineering FRC Student on FRC 2052 from 2009 -2013 Mentor on FRC 4536 from 2013 -present Volunteer at multiple regional and off-season events

PRESENTATION FORMAT This presentation is designed to showcase some useful methods and tricks that I have found over the years to get entire teams involved in strategic design. The presentation will not be all-inclusive. I will leave additional information within this slideshow if anyone is interested, and they can look back on this later. If you need specific help moderating strategy discussions or working through other steps of the engineering design process, contact me and we will work something out.

ENGINEERING DESIGN PROCESS Read this paper, it is really good. Design process steps: 1. Understand the problem 2. Generate Specifications 3. Rank Specifications 4. Generate Concepts 5. Prototyping 6. Choose a Concept 7. Detailed Design 8. Manufacturing/Implementation 9. Test/Tweak/Iterate

UNDERSTANDING THE PROBLEM Decide on a competitive goal Watch animation and read the rules Develop deeper understanding of rules Determine things the “average” team will strive to achieve Develop a general feel for the gameplay Look for “game-breaking” strategies Determine our team’s general strategy

DECIDE ON A COMPETITIVE GOAL Something that should be done by team leadership well before kickoff even happens. Possible goals: *Play in eliminations Be an alliance captain Win the event Win a design award (Quality award, Excellence in Engineering, Imagery, etc…) *We will be assuming this one for the remainder of the presentation

WATCH ANIMATION AND READ RULES Welcome to the 2006 FIRST Robotics Competiton, and this year’s game, Aim High! Game Animation: https: //www. youtube. com/watch? v=Pl. ADdm. CAMB 8 Rules: http: //archive. firstinspires. org/roboticsprograms/frc/2006 -first-robotics-competitionmanual-and-related-documents The sections relevant to us are 3, 4, 5. 3. 2, and 8. 3 -8. 4 To access rules: 1. Go to archive. firstinspires. org 2. On the top, click “FRC” 3. On the left column, under “Game and Season Info, ” click on “Archived Game Documentation and Event Results” 4. Scroll down to “ 2006 -Aim High” and select “Competition Manual”

DEVELOP DEEPER UNDERSTANDING OF RULES There a few key areas that groups will present on so that everyone is completely familiar with them: How are teams ranked in the qualification matches? (8. 3. 3 -8. 3. 7) How are the robot rules different from previous years’ robot rules? (5. 3. 2) What are the general characteristics of the game pieces? (How many, where do they start, what are they) What are the primary characteristics of the arena? (3. 2) What are all of the ways to score/descore in this game? How much are these events worth? (4. 3. 3) What are the time periods in this game? (4. 3. 2) What are the most important game rules? What rule(s) will most often be violated?

DETERMINE THINGS THE “AVERAGE” TEAM WILL STRIVE TO ACHIEVE The “average” team will usually take the game at face value. That is, they will usually try to do the thing in the title of the game, or the thing that offers the most points during the teleop period. On the following page, I have compiled some basic objectives that teams might identify for this game, note that actual robots will not fit nicely into these categories, this is just to get a sense of how the other teams will design themselves.

DETERMINE THINGS THE “AVERAGE” TEAM WILL STRIVE TO ACHIEVE In your groups, determine where you think the average team will go in each of these categories. Feel free to choose options which are not enumerated in this table. Piece acquisition Scoring balls Holding balls Movement Scoring on platform Unable to acquire balls after auto Get balls into high goal one at a time Hold up to 10 balls at a time Has a basic drive Unable to score on platform Acquire balls from ground Get balls into low goal one at a time Hold up to 20 balls Has an extremely Able to score on at a time powerful drive platform Acquire balls from HP Dump balls into low goal Hold up to 40 balls Has an extremely Able to push others at a time agile drive onto platform

DEVELOP A GENERAL FEEL FOR GAMEPLAY One way that 4536 uses to understand gameplay is to simulate it. The important thing to remember when doing simulations is not to exactly mirror in-game conditions, but rather to gain insight about game flow. When simulating a game, you should be thinking about the following questions: Where are the game pieces at which times of the match? How many points are being scored via each scoring method? How long do actions take? How can defense be used?

DEVELOP A GENERAL FEEL FOR GAMEPLAY Use the materials I have distributed to simulate this game. First, assign people to the following positions: Robots: 6 Timekeeper: 1 Scorekeeper: 1 Optional, referees: 1 -3 Optional, human players: 2 Next, decide on the functionality of all of the robots involved. To start, assume each robot has the “average” qualities we defined earlier. Then, play through a match. If you need to pause between periods to track the score, please do so. After each match, spend a few minutes sharing insights about the gameplay.

LOOK FOR GAME-BREAKING STRATEGIES A game-breaking strategy is one that goes outside of the established norms of the game. A strategy that tries to game the game. The most famous example in FRC is team 71 in 2002: https: //www. youtube. com/watch? v=G 0 ubppvlh. Jo Spend a few minutes developing possible game-breaking strategies of your own. It is okay here to be borderline breaking the rules. The point of this exercise is to get you to think more deeply about the game. I don’t generally recommend carrying out game-breaking strategies, but it is still helpful and fun to consider them.

BUT WHAT ABOUT DEFENSE? In most games, if your goal is just to make it into eliminations, you should not worry about the other team defending you. Generally, if you are good enough that the other team is trying to defend you, then you will almost certainly be playing in the elimination rounds. Designing your robot to play defense is another story. I don’t have much personal experience with this, but I have not seen very many successful defense-only robots. Besides making a powerful, robust drivetrain and giving your drivers lots of robot practice, I generally think that making defense-specific mechanisms is not a particularly good use of time.

SEPARABLE VERSUS DEPENDENT GAMES A useful metric when analyzing a game is the game’s separability, that is, how much interaction between teams on the same alliance will factor into the gameplay. Some examples of separable versus independent attributes of games: Qualities of separable games Qualities of dependent games Many game pieces (2012, 2013, 2015) Few game pieces (2008, 2014) Linear scoring (2012, 2013) Exponential scoring (2007, 2014, 2015) Scouts focus on individual robots Scouts must focus on whole alliance Non-functional robots do not change partners’ scoring ability (2011, 2013) Non-functional robots limit scoring opportunities for their own alliance (2012, 2014, 2015) No coopertition (2013, 2014) Coopertition (2012, 2015)

SEPARABLE VERSUS DEPENDENT GAMES 2013 was one of the most separable games of all time, while 2014 was one of the most dependent games of all time. 2015 was somewhere in the middle. How does separability affect design strategy? If your goal is just to make it into the playoffs: If game is separable: Focus on finding ways to augment your alliance without inhibiting partners. Autonomous and endgame are really good times for this. For example, in 2013, a robot with a good autonomous and a consistent 10 -pt hang could play defense during teleop and still be a great contributor If game is dependent: Focus on not being the limiting factor for your alliance. Make sure that you can achieve the basic game tasks quickly and reliably. For example, in 2014, a robot that had great intake and passing abilities was really important for playoff alliances, even if said robot could not throw the ball at all.

DETERMINE YOUR TEAM’S GENERAL STRATEGY Alright, we know a lot about this game, so now we need to apply it. Remember, our goal is just to make it into the playoffs. To do that, we need to stand out amongst the competition, which usually means choosing a strategy that is different than the average teams’. The most important thing is to not do the same thing as everyone else. Specialization is crucial, and the jack of all trades is the master of none. For more specifics on this, watch Karthik’s awesome presentation here. This presentation is one of the cornerstones of my strategic design philosophy.

DETERMINE YOUR TEAM’S GENERAL STRATEGY Go through each category from the table again, but try to select a strategy that is different from the average teams’. Usually, the chosen strategy must yield at least as many points as the average team will earn. Piece acquisition Scoring balls Holding balls Movement Scoring on platform Unable to acquire balls after auto Get balls into high goal one at a time Hold up to 10 balls at a time Has a basic drive Unable to score on platform Acquire balls from ground Get balls into low goal one at a time Hold up to 20 balls Has an extremely Able to score on at a time powerful drive platform Acquire balls from HP Dump balls into low goal Hold up to 40 balls Has an extremely Able to push others at a time agile drive onto platform

DETERMINE YOUR TEAM’S GENERAL STRATEGY strategies that I would choose for this game: Hoarding balls from HP and ground and dumping into low goal Being a good defender that can get up on platform at the end of the match and has autonomous defense, collects balls from HP

DETERMINE YOUR TEAM’S GENERAL STRATEGY Q But what if my goal is to be an alliance captain? A Then instead of focusing on standing out, your focus needs to be on seeding high. For this game, that means finding a high-scoring strategy that will win you the grand majority of your matches regardless of who you are partnered with. Also, pay close attention to second and third order sorts in the ranking criteria.

THAT’S IT FOR STEP 1 But we’re not done, this was only the first step of the design process linked to in slide 4. However, that’s all we’re going to cover in this presentation.

USEFUL RESOURCES Please contact me if you have any questions/comments/concerns. I love talking about this stuff, so you most certainly won’t be a burden. Caleb Sykes, FRC 4536: calebsyk@gmail. com Here is another link to the paper I keep mentioning, please read it: Engineering Design Process for FRC. Here is a link to Karthik Kanagasabapathy’s awesome presentation on a variety of topics, the first 50 minutes are focused on strategic design: Effective FIRST Strategies This is the single most useful site for anything and everything related to FRC: Chief Delphi