Match strategies & more prototyping
Today, we’ve made progress on our game analyses, and we continued creating and testing out several prototypes.
Ranking point analyses
We’ve taken a close look at all possible ways we can score RPs and what would be the best options.
Melody RP
For the Melody RP to be scored, 18 (or 15 with coopertition) notes have to be scored during the match. To determine the difficulty of getting this RP, we’ve looked at cycle times from 2013/2017/2023 since they all have full court cycles.
We converted these cycle times to fit this year’s game and wrote down how many cycles a robot will be able to do during teleop (see table below).
| Bad robot | Average robot | Good robot | Alliance | |
|---|---|---|---|---|
| # Cycles 2013 / 2017 | 3 | 8 | 14 | 25 |
| # Cycles 2023 / 2024 | 5 | 9 | 15 | 29 |
It can be concluded from this table that scoring the RP with average alliance partners is pretty doable. It gets harder to score the RP when scoring it solo. In that case, the chance is great that you’ll have to get the coopertition bonus to reduce the note threshold just enough to solo score the RP.
Ensemble RP
In the table below, you’ll find all options to consider during a match when going for the ensemble RP. Right now, we’re estimating that 50% of the robots will be able to get ONSTAGE during week 1 competition. This means that it will be most likely that during quals, most of the time you’ll have to get the RP with only 2 robots. This presents the one of the most important trade-offs in the game: “Is the TRAP a trap?”. Unfortunately, we don’t have enough proof right now to make a decision. Simple because we can’t estimate how much trouble the trap subsystem will bring. We’ll know more about this later this week when we have physical prototypes to test.
| Spotlit | Not spotlit | |
|---|---|---|
| 2 robots can climb | Same chain | Not possible without trap |
| 3 robots can climb | Different chains | 2 same chain, 1 other |
Ranking point estimations
We have estimated that getting the melody RP should be pretty doable with an average alliance, and using the coopertition bonus will definitely help below average alliances to get just enough notes to score the RP.
On the contrary, the ensemble RP is going to be way harder for average alliances, because of the options being limited too:
- 3 robots ONSTAGE: chance of having 3 robots that can get ONSTAGE is low.
- 2 robots ONSTAGE with spotlit: relies heavily on human player throwing skills
- 2 robots ONSTAGE + trap note: Requires a robot that can score trap. In most cases, an average alliance will not have such a robot.
We will look further into trade-offs tomorrow, and we will use the findings mentioned above to make an informed decision on the robot archetype later this week.
Human player analysis
It’s no surprise that the human player will be very valuable this year. They will have to do the right actions during the match in order to help the alliance perform better. We have created a list of this to think about for both the Amp HP and the Source HP:
Amp HP
- Click the coopertition button (if needed)
- Wait with clicking the amplification button till first shot hits
- Coordinate with drive coaches when to get teams ready to shoot
- Throw High notes onto the Microphone or introduce them close to a robot ready to shoot (still a legal action atm)
HP Source
- Train with signing to get ready
- Train with dropping the notes just on time for the robots to pick it up
- Coordinate with drive coaches to always be ready with a ring
Key takeaway from this: make sure to assign your human player early on in the build season and make them throw 20 high notes every day they visit the workshop.
Match strategies
As you might already know, this game contains a lot of match strategy and have the right strategy can have tremendous impact on match results, even when facing faster cycling robots.
There are a lot of different alliance combinations that could work well. Some examples are:
- 2 robots speaker cycles, 1 robot amp cycles
- 1 robot speaker cycles, 1 robot amp cycles, 1 robot zone defense.
- 1 robot speaker cycles, 1 robot amp cycles, 1 robot feeds notes.
At the moment, it looks like being a more specialized robot that does only 1 thing very well isn’t a bad option for most average teams. The hardest question to answer is, do we need a robot that can do everything? Simplicity is key, when designing robots, but does the game force more elite teams to be more flexible with their scoring options because of alliance partners not having certain skills?
The answer really depends on the knowledge gained from our prototype tests and how complex we expect each subsystem to be. Right now, there isn’t a clear winner in terms of scoring abilities that pops up.
Cycle times
Today, we have also tried a lot of cycle paths with a physical drivetrain. We tested the most important paths 3 times and collected the average cycle times. The list below shows the estimated cycle time based on the driving we did today.
| Starting position | End position | Average time |
|---|---|---|
| Subwoofer | Contested notes | 3 seconds |
| Subwoofer | Source | 5 seconds |
| Amp | Source | 5 seconds |
| Source | Wing | 4 seconds |
Using the collected data from this in combination with the old games analysis mentioned in yesterday’s post, we can make a good estimation on cycle times.
We’re currently predicting 10–12 seconds for Einstein level cycles. After doing a rough calculation of dividing the cycle time at Einstein level with 0.7, we think that 14–17 seconds will be cycle times during week 1 competition level.
Prototype Progression
We’re currently busy with designing and building the prototypes in such a way, that we can test all different kinds of parameters isolated. We do this by for example using extrusions for our shooter prototype so we can test the angle, compression, and height separately without changing the other parameters.
Intake & Trap prototypes
This season, we are lucky to have a CNC machine in our workshop. We’re currently busy with lasering the plates for the intake and trap (aka drainpipe) prototypes. We are assembling them tomorrow evening.
Climber hook
We tested our trident hook concept today using a metal bracket attached to our 2022 climber. The purpose of these tests was to check if our trident hook prototype will hold position on the chain. Check out the videos below to see the results:
Trident hook 1st version
Trident hook 2nd version
Have you spotted a note?
You may have noticed, but we still don’t have any real notes to test our prototypes with. Because of some shipping delays from the USA to The Netherlands, we will receive our notes Friday or Monday, hopefully. This is a major setback for us, since we planned to only prototype for week 1 and start designing the robot in week 2, but without notes it’s hard to test and try what parameters work best for our subsystems. For the upcoming days, we will try to make the best out of it and make some fake notes as good as we can.
Nevertheless, I think we can say that we’ve once again had a successful day of prototyping and strategizing. I can’t wait to see our prototypes come to life and put to action.
Written by:
@Bjorn - Lead Data Driven Decision Making