Tips to make your team a contender?

FIRST General Forum
#27

There are many great suggestions here. This should be a “classic” thread that we steer rookie and young teams to each year. In fact, if someone could compile this into a white paper we could post it here and also get FIRST to post it as a resource on their website.

I’ll describe what I think has led to our success over the last 4 years. Until this year, our “shop” was a shipping container in the field behind Steve Harvey’s math classroom. We had a bandsaw and table saw, and last year we got an old Bridgeport mill. I think many other teams are in similar situations so I don’t believe that resources need be a limiting factor to achieving significant success. Within these constraints, here’s what lifted us since 2012:

  • Fundraising and sponsorship: Our team budget increased from $35k in 2012 to $125k this year. We had to get the money from somewhere. Fortunately we were able to get initial seed money from UC Davis, and we’ve used that to leverage into other sponsors. In a small community like Snohomish you can get to know every business personally–start with your Chamber of Commerce and the city’s economic development office. You also can reach out to large companies in Everett like Boeing. Teams in large cities can contact the large corporations with HQs there. Sell the “program” not just the team aspect. We’re not the local baseball club; we’re an integrated project-based education program that uses a sports-model to inspire students. To achieve this you need non-technical students to lead this effort. Bring in students who are interested in business, media, presentations, even the arts. You can make your team a bigger community.
  • Outreach: We have developed projects outreach both to our community and other FRC teams. Somehow that has provided a catalyst of dedication of our team members. I think it makes the students realize that they are part of bigger effort. I don’t think it’s coincidence that most of the Hall of Fame teams are also competitive powerhouses. Again, you can involve students who are interested in more than engineering. Education, arts, media are all needed for this.
  • Constant improvement: If you look at pictures of our 2013 bot at CVR vs champs you wouldn’t believe that it’s the same machine. We constantly work on improvements throughout the season. Until this year 1671 would build a machine that very good at their first regional but it would be static and other teams would pass them by. This year they were constantly making improvements. They weren’t particularly close to us at CVR but they outscored us as Sacramento and then took a riskier pick with a higher upside to go for the win. Taking risks is part of the constant improvement mindset.
  • Strategic analysis: Karthik’s talk at Champs covered most of this. Think before building. Game out all of the possibilities. Be willing to go down potential dead ends and cast aside your failures. We have developed a braintrust led by our mentor Mike Corsetto. You can’t duplicate Mike but you can get most of the way there by duplicating the process. We make a list of what we would want to achieve and then slim it down to what is really achievable. For example we dropped landfill loading because we realized it was a tradeoff and that we couldn’t do all types of loading plus a fast cangrabber. Last year we decided that being the top finishing shooter was less important in our overall strategy.
  • Scouting: we start developing our system even before build season. We push data via smartphones to our drive team. We’ve integrated our pit scouting into our electronic system so that we can see that data in the pits before we develop our pick list. We game out potential draft scenarios. (We knew what was coming in 2013 on Curie and if the other teams had been able to see it as well they may have made different decisions.) Our scouting system was critical to identifying 840 and 295 at CVR in 2013; we did not have the best robot at that competition (3970 did.) A side benefit is that you use your excess programmers to develop and maintain your scouting system. There’s really only so much programming needed for the robot. Plus you can bring students who are interest in statistics and “Moneybot”. Karthik said this is how he got involved with robotics.

We will post videos from our fall workshop that address some of these topics, and will have another workshop in October in Davis.

#28

Some really good suggestions here. I’m taking notes.

The practice robot thing was one of the biggest takeaways I had from last year, and it served us really well this year to have one. But if I can rant a bit, the fact that you need one to be competitive is quite frankly unfair.

FIRST set the bag rules (and crate rules previously) to put some kind of cap on the labour teams sunk into their bots. This is good for the sanity of the mentors, and was a noble attempt at leveling the playing field. Unfortunately it back fired on both counts. Now all successful teams build two robots so they can continue practicing and developing into the competition season. That puts a doubly high cost barrier between rookie and poorer teams and ever being competitive. Everyone knows this is unfair (not to mention a complete waste of resources), but the image of “stop build day” is so compelling that no one is willing to acknowledge the problem.

As far as I’m concerned, we should stop the expensive farce and get rid of the bag rules. Changing the marketing literature build a robot in 7 weeks, or 8 weeks or 10 weeks is just as impressive as 6. End of rant and derail :stuck_out_tongue:

edit:

We did the same, and that’s good advice. We were lucky to have old drive base parts donated to us by 4039 and 2056 for the job and it certainly saved us money to withhold our elevator for the practice bot. The rest of our mechanisms weighed more than 30 lb so we still had to (pointlessly in the big picture) build a lot of stuff twice.

#29

This and some more of this. Before you even start thinking of designs, read ALL the rules with the entire team, and then decide on your strategy based off the game and scoring methods. For example, this year your priority list for scoring may have looked something like this.

1. Load from landfill
2. Stack to 6 high
3. Cap with a RC
4. Load RC with noodle
5. Load from HP
6. Pick up pool noodles off the ground

Once you set your strategy, you can start looking at designs on how to accomplish it. The important thing is making sure something low on your list doesn’t trump something high on your list. For example using the list above, your ability to load from the HP should never jeopardize your ability to cap with an RC.

#30

Get as many people on your team as possible over the 10,000 hour point.

I believe someone could do an entire paper on the rise/fall of powerhouses within FRC as related to the arrival/creation & departure of these individuals.

#31

Personally, I feel that if a team is low-budget and lacks a large build team, building 2 robots can be extremely difficult and lead to students being burnt out.

However, I completely agree that driver practice is crucial. Drivers need to have complete control of the robot in order to maximize effectiveness on the field. Having a competition-ready robot by week 5-6 ensures that drivers have sufficient time to master the robot.

On the topic of drivers, do driver tryouts before the season starts to make sure that the designated driver(s) get as much practice as possible.

#32

Building a practice robot is one of our team’s goals. However, we haven’t had the resources to build a second robot due to team size and space.

This year, three things helped us make a big improvement in our team. (Our first year to qualify for PNW District Championships and the World Championships)

  1. We upped our outreach program (by several magnitudes)
  2. We sought out advice from upper level teams, 1983 before Central Washington (helped us improve our ability to align to the feeder station and handle totes) and 4488 at Central Washington (helped to convince us that a canburglar would be more important than a ramp at PNW Champs).
  3. We used our withholding allowance to continue to improve our robot. Our previous mechanical mentor could only provide us with his garage shop until the bag and tag day. Our new mechanical mentor committed time beyond six weeks.

On side note, our team really enjoyed working with your team and had fun as an alliance partner at the Glacier Peak event. We’ve seen the improvement of 4309 over the years.

#33

I personally have believed this would make an excellent book. Several years ago, I read Built to Last and thought it would be wonderful for someone to do a study on FRC (or FSAE of Mini-Baja…).
While 10,000 hrs. is on the verge of doable (that is 5 years of working a full time job of 40 hrs/wk), I think that sustainability is likely more about getting a lot of members on a path to 1,000 hours. While and order of magnitude less “experience”, it is not an order of magnitude of less “goodness” that the individual can offer.

A discussion on practice driving:
For those that are interested in watching the progression, grab some new students and have them drive the robot. The first 10 minutes to 1 hour are them basically figuring out the controls. After about 10 hours of active practice, the driver/operator likely get to the “good” level. Around 100 hours of active practice, your drive team should be very very good. In an FRC season, it is hard to get much past 100 hours of practice (not impossible, but very hard).
This is not much different than with video games. The first time a person plays a new type of video game that they have never played that style, their talent is the biggest contributor to initial success. Once they get around 10 hours of play time, most can do a pretty good job of handling the character and moving about the board/world/game space. Around 100 hours, the player is likely to be considered very skilled relative to outsiders, but still a long way from an expert. Around 1,000 hours, they should be incredibly proficient (though experts will still likely be considerably better).
Between that 100-1,000 hour mark, they can switch to a different game of a similar genre, and will likely start out relatively good, and have initial progress that is much more rapid.

This is much the same with CAD, design Calculations, strategic analysis, fabricating parts…

#34

The best thing you can do is understand the game well, and build well within your means. Strive to work smarter, not harder. Make a priority list and stick to it like the law. Have fun. Keep organised.

Do these things and the rest will come. Don’t jump too far into the practice robot camp before you seriously evaluate not only your financial situation (the thing that everyone things about) but your ability to actually make that second robot during the season (the thing that few people talk about). I have seen far too many teams have “practice robots” that the team is too busy fixing/getting ready/building two robots at a time that no practicing actually gets done. Not only do they not practice, it hinders the competition robot as less time is devoted to it. Be very careful with them. That being said, if you can do it, it’s a huge asset.

#35

Play practice match 1. Build a robot that can play practice match 1. A good robot that was finished on practice day will always lose to a mediocre robot that played in practice match 1. If you can get into that match, in most cases you will be able to play at least 4-5 matches back to back without waiting on the filler line. If you do that you’ll be head and shoulders above every team that was working in their pits back then. This means that you should design a robot you can bag completed. Charge your batteries before practice day. Plan hour to hour what you’re going to be doing and when you’re going to be inspected. Finally, don’t forget to program your radio.

Note: if you can get inspected really quickly there won’t be a wait to get an inspector.

#36

Understand your resources. Know your limits, your weaknesses, your strengths, and potential opportunities in all aspects of your team. Your students, mentors, build space, sponsors, local community, finances, and your team experience are just a few aspects of your team that can impact your performance as a team. With the 2015 season still fresh in your mind sit down as a team and discuss how the year went. What did you do right? What did you do wrong? What didn’t you capitalize on? What parts of your team do you see as a strength moving forward? What parts of your team do you see as a weakness? What steps can we take between now and kickoff to better prepare ourselves?

Get the robot done and practice, practice, practice and don’t be afraid to iterate mechanisms to improve your performance. Constant improvement is needed to stay competitive with the game if you come out in week 1 swinging.

#37

Our neighbor FRC team, 1912 Team Combustion, refers to these three components as their “Combustion Triangle”. (Though as I recall it’s sponsors rather than money, but that’s a subtle distinction.)

#38

Thanks! You guys have been awesome the past couple years when we’ve been at the same events. That’s awesome you got to the championship.

#39

While it wasn’t something Karthik said directly at his Effective Strategies presentation last week, two pieces of what he said engaged in some promiscuous correlation.
Key pieces in making a high-scoring robot are:

  • execute plenty of “game cycles”
  • spend a lot of time building and trying things
  • be ready to toss an idea when another works better

These ideas combine into the concept of “innovation cycles” - how many times during build season can you go through the various engineering cycles:

  • redesign (back to the drawing board)
  • major update (rebuild nearly or completely from scratch)
  • minor update (rebuild a component, whether a single piece or subassembly)
  • tweak (minor modification of a component)
  • tune (adjustment of a component)

That is, I have to think that the speed with which you make each engineering cycle (while being sufficiently thorough, of course) increases the value of each hour spent in the design and build process.

#40

At the same time though, you don’t want to be tweaking so much that your drive team never gets to actually, you know, drive the robot.

“Hang on guys… I just want to try adjusting this one other thing over here…”

#41

I note that Karthik also brought up the 10,000 hour myth in his talk on Wed. night. Several authors, including David Epstein in the “Sports Gene” show that Malcom Gladwell was incorrect in developing his anecdotal assessment about needing 10,000 hours to be successful. Experience is very helpful, but there are means of shortcutting it.

#42

Not as subtle as you might think. Nurturing sponsors means that money arrives on a continual basis. Sponsors also can deliver much more than money, including publicity, mentors, community and other support. Never think of your sponsors as “money.” Make them integral to your community.

#43

I think the anectdote makes more sense when you think about an example like someone just practice hitting the ivories on a piano for 10000 hours alone and can become really good at it. Doing FRC for 10000 hours independent of other things can probably still make one person really good at it, but leaning on the processes of other people and getting a relevant education or listening to talks like the strategy discussion can take the hours to mastery count down a lot.

1 Like
#44

The point stands though, whether or not the quantitative metric is correct - if you want to be good at something, you need to spend a lot of time with it. Learning, practicing, improving, talking, teaching, competing. The more you do the more proficient you become.

#45

Practice is indeed how you get to Carnegie Hall.

#46

I don’t dispute it, I was just highlighting shortcuts. There are teams in FRC that are not just powerhouses, they are incubators. A good analogy here is The Daily Show. It takes people with potential, puts them through a proven process, then catapults them to success in a future project. It’s possible Stephen Colbert could have hosted the Late Show without ever working on the Daily Show, but that kind of environment is very beneficial.

I think someone on 1678 said it best that they can’t replicate the great mentors on the team, but they can replicate the processes those mentors work through and educate their students on it.