Karthik’s Strategic Analysis

Coming Soon - Check back January 5th!

Hey everyone, welcome to Karthik’s Strategic Analysis. In this post I’ll be walking you through the strategic design process I went through to analyze Recycle Rush and to create a strategic priority list for the game. The full process that I used was outlined in a presentation I gave this fall. I highly recommend you give it a read/watch. The video is here:

And the PowerPoint slides can be found here: http://content.vexrobotics.com/buildblitz/Effective-FIRST-Strategies-October-2014.pdf

The strategic design process is essential to ensure that you a build a robot that plays Recycle Rush in a strategically efficient manner. You use this process to create a priority list that defines the Robot you will build. This is what we call a “Strategic Design”. Of course every team will have a different priority list. This is not only because teams may come up with different analyses of the game, but because teams come from varying resource levels with varying goals. This priority list was created from the perspective of team whose goal is to consistently make the elimination rounds at the Regional/District level as an Alliance Captain or first pick. We felt this would be the most valuable frame of reference to use during this prospect, as this applies to the greatest number of teams. If you’re looking for an Einstein level priority list, you won’t find that here. Many tasks which may be very valuable in Recycle Rush may be low or nonexistent reward to effort ratio may be too low for a team with the resources/goals in our assumption. Also, please remember that the content of this blog post is just a start. The goal is here is walk you through the strategic design process, but not to do everything for you. We’ve intentionally not gone fully in depth here to make sure we’re leaving some of the fun to you! We want this to be an aid to help teams in their strategic design process, not to be used in lieu of their own!

The first step in the strategic design process is the game analysis. To start this off you need to fully read and understand the Recycle Rush game manual. If you haven’t we urge you to do this right away. You can find the game manual here: http://www.usfirst.org/roboticsprograms/frc/competition-manual-and-related-documents

The next step is to fully understand the ranking system. We’re going with the assumption that the goal here is to rank as highly as possible and to win the event. Table 5-1 in the game manual indicates how teams are ranked. The primary criteria is Qualification Average, meaning teams are ranked not based on wins and losses, but on how many points they score in each qualification match. This is a very important distinction. The performance of the other alliance on the field during your matches is not directly related to the primary ranking criteria. Teams are almost solely being rewarded for maximizing their own score. Based on the scoring criteria for the game, we can see that match scores will range from about 0 to 250, with a normal distribution centered around the average. Since ranking is based on average match score, this makes the possibilities of ties unlikely. As such this priority list will focus only on the primary ranking criteria of average match score and ignore the subsequent tiebreakers. So our goal is to score as many points as possible!

Now that we know the big picture of what we’re trying to do, we need to brainstorm every possible action in the game that can earn your alliance points or lead to an action that scores your alliance points. If you’re at a loss of where to start, go right to your game manual and the scoring section (3.1.2). This will give you a list of all actions that earn points directly. From there you look at each of those actions and brainstorm actions that lead to those.

Here’s the list I came up with for this exercise. Remember, as stated earlier, please do not consider this list to be all encompassing. It’s important that your team goes through this exercise on your own to make sure you consider all possibilities.

  • Drive to Auto Zone in Autonomous Mode (4 points for 3 robots)
    • Skills Needed: Driving
  • Create a Tote Set in Autonomous Mode (6 points for 3 totes)
    • Skills Needed: Driving, Manipulate Totes
  • Create a Container Set (8 points for 3 containers)
    • Skills Needed: Driving, Manipulate Containers
  • Create a Stacked Tote Set (20 points for a stack of 3)
    • Skills Needed: Driving, Manipulate Totes, Lift Totes, Stack Totes
  • Create a Coopertition Set (20 points for 4 totes)
    • Skills Needed: Driving, Manipulate Totes, Lift Totes
  • Create a Coopertition Stack (40 points for 4 totes)
    • Skills Needed: Driving, Manipulate Totes, Lift Totes, Stack Totes
  • Score Totes (2 points per tote)
    • Skills Needed: Driving, Manipulate Totes, Plow  or Lift Totes onto Scoring Platform
  • Score Recycling Containers (4 points per level of Container)
    • Skills Needed: Driving, Manipulate Containers, Lift Containers
  • Score Litter (1 point in landfill, 4 points for opponent zone, 6 points for container)
    • Skills Needed: Driving(?), Manipulate Litter(?), Deliver Litter(?)
  • Drive
  • Drive over Scoring Platforms
  • Manipulate/Plow Totes
  • Receive Totes from Human Players
  • Lift Totes (in varying amounts and heights)
  • Stack Totes
  • Manipulate/Plow Containers
  • Lift Containers (in varying heights)
  • Manipulate/Plow Litter
  • Deliver Litter
  • Reorient Totes
  • Reorient Containers
  • Remove Totes from Platforms
  • Remove Containers from Platforms

As you can see, there’s a lot of duplication here. That’s okay as it’s good to see which tasks are duplicated, as it will help you realize which tasks are the most important. When forming the final priority list we’ll strip out the duplication. The end goal here is to have a strategic design list of strategic actions to help define your robot design. So the final priority list will be based around strategic actions, but with indications of the overall actions they’re intended for. (An example of a strategic action versus an overall action is Stack Totes vs. Create a Coopertition Stack)

Normally after listing all the ways to score points, the process would focus on ways to deny your opponents points. However as we established earlier that the criteria towards advancing is based on total points scored, there’s no reason to focus on denying your opponent points. This may become something you want to do consider for finals matches, however that discussion is beyond the scope of this analysis.

Now for each strategic action we need to compare the reward for doing so (the point potential) versus the level of difficulty. We want to prioritize actions that have a very high reward to effort ratio. This part of the process is part quantitative based on scoring values, but part qualitative as well. After doing this you can rank the items based on determined ratio. Here’s what I came up with, along with some comments on the analysis that go me there.

  1. Drive

Every task overall action we described relied on the robot being able to move. Although there are creative ways you could play this game without ever moving, that type of design is beyond the scope of initial set of assumptions we’re working with. As such drive is the top priority as it almost always is in any FRC game. The fundamental assumptions about drivetrain design will change in this game since there is no robot interaction. We’ll definitely be talking about this in our Drivetrain Design Guide & Philosophy.

  1.   Manipulate/Plow Containers

Seeing this one this high up on the list may be surprising for many of you. The ability to plow containers without other actions only really comes in handy during the autonomous mode when creating a container set. However, alliances are always looking for simple ways to maximize their autonomous scores. Since points are already awarded for creating a robot set by driving to the auto zone, there will be an impetus for team to take their containers with them. Since this 8 point scoring option requires the involvement of all 3 robots on the alliance, you can see that this will be very high in demand. Combine this with low level of difficulty to plow these containers, it makes sense to put this up high on the priority list. However, prototyping may lead to the discovery that pushing these containers is harder than we think. So you’ll need to do some investigating. We’ll be publishing some resources on Simple Game Piece Manipulation that will dive deeper into this.

  1. Manipulate/Plow Totes

The logic on this one is the same as #2. The reasons it’s lower on the list is because a tote set in auto is worth less points than a container set and because many alliances will not want anyone going for a tote set, so one robot can work on the higher valued stacked tote set.

  1. Plow Totes onto the Scoring Platforms

The main scoring task in the game. Scoring totes by plowing them into the platforms appears to be relatively easy, earns points each time its done, and also allows for the potential scoring of a container. Of course, the disclaimer about the difficulty level that was mentioned in #2 applies here as well. My gut instinct is telling me that plowing Totes and Cans might be harder than our initial impressions.

  1. Plow Litter

The point values are low here, but the difficulty level is also very low. In addition, you may need to be able to plow these just to keep them out of your way while performing other game tasks.

  1. Lift Totes onto the the Scoring Platform or Step

Now we’re getting into the tasks that are more difficult. Being able to lift a tote allows you score on a platform and possibly earn coopertition points on the step. However, lifting these one at a time is not going to be as efficient as plowing en masse.

  1. Stack Totes

A lot of people upon seeing the game might assume this is the #1 task, since it’s the dominant image of the game. However stacking is heavily reliant on driving and tote manipulation, hence it being lower on the list. If you can’t get to, manipulate, and lift the totes, you’re going to have a hard time stacking them. Stacking is higher on the list than lifting/placing cans, because it seems to be fundamentally easier. Prototyping may prove this to be wrong.

  1. Lift and Place Cans

There are huge points for this task, clearly the dominant point value in the game. However it’s reliant on having stacks already created for you, or being done in tandem with stacking. This combined with the potential difficulty have it put it lower on the list.

  1. Reorient Cans

This could be a very useful “helper” task. I envision a lot of robots that can only lift cans that are upright. Having the ability to right cans that are horizontal could be very valuable.

  1. Drive over Scoring Platforms

Most scoring in this game can be done by avoiding the scoring platforms. Even so, being able to drive over these platforms will increase your scoring efficiency in general.

  1. Remove Cans from the Step

Very useful task in terms of giving your alliance more potential scoring objects, however the totes in front of the step are a significant obstacle. This makes this task very difficult. Even though this task will become important in higher level game play, it’s definitely a lower level priority for most teams.

So there it is, a priority list for Recycle Rush. This should be a very good sample to help spur teams to make their own list and do their own analysis. I cannot stress how important this process is for every team. Having a solid understanding of the game and strategic process for design will help teams ensure they build robots that play this game in an efficient manner. If you have questions about this list you can contact me on Twitter @kkanagas, or you can ask them during tonight’s broadcast! I’ll talk to you all then!

– Karthik