What separates the good from the great?

FIRST General Forum
1

I’ve been giving a lot of thought/research time into what practices help lead great teams to success that my slightly above average team (1155) is missing and could learn from.

So keeping my question as broad as possible, what do you think separates teams like mine from teams like those on Einstein? What do you think we could learn from/ copy from in order to be more successful in the coming season?

2

Iteration. Always improve the robot to fit the strategy, and the strategy to fit the metagame.

3

Good Drivers. They can make a good robot great.

4

Good teams are good for 4 years when they get a batch of good students. Great teams stay good for decades.

5

I would like to generalize a bit: always improve. Continuously reflect and continually improve in all aspects of what your team does. Be a bit better from event to event and from year to year, and it adds up a ton over time.

Also, document your knowledge so it doesn’t go away when students graduate.

6

Thank you very much for the question! I’ve been rolling it around for an hour or two. Let me ask a counter-question in two parts: How do you identify a good team? How do you identify a great team? The answer to these two questions will greatly affect the answer to your question.

Many of the responses so far seem to have an answer for a good team of “one which wins events with some regularity” and a great team somewhere along the lines of “a team which consistently earns its way to CMP and regularly earns its way to Einstein, and takes home several blue banners from Einstein”. The key ingredients to this sort of success are:

  • Talent *
  • Hard Work
  • Luck

For future reference, I will refer to these as “great FRC teams”.

Instead, I believe that a good team is one which recruits a combination of students who have “already chosen STEM” and those who had not, and inspires these students to become productive leaders and performers in making the world in which they will live. A great team is one which inspires teams outside of its normal (geographic) sphere of influence to be/become good (and occasionally great) teams. For future reference, I will refer to these as “great FIRST teams”.

Many of the great FRC teams are also great FIRST teams, because of their competitive prowess and a willingness to share with other teams. 'Nuff said there.

However, there are definitely teams out there which are great FIRST teams which are not great FRC teams. As a specific case, let me cite team 358, Robotic Eagles of Hauppauge, NY. According to my research, they have been to CMP more years than not in their history, but have never done better at CMP than being eliminated in quarter finals for their division. This is clearly a “good but not great” FRC team. However, for many years, their running description of the FRC control system has been second only to the official FIRST documentation in authority, and has consistently provided details to teams which they could not find anywhere else. I therefore declare 358 to be a “great FIRST team”. There are many other teams which regularly serve as beta teams, or which test components and share results, but which may not be among the competitive elite. These are the teams I consider to be great, which I hope to get my team to emulate.

  • Talent is rather dependent on a team’s situation. A public school in a low-tech area will likely have a limited talent pool. A school in a high-tech area will have a rather greater pool to draw from. A community team in a high-tech area may have access to massive talent.
7

Since you’re asking about teams that play on Einstein, I’ll give a very performance focused answer. Note that the teams on Einstein do a lot more off the field to make them so well respected in the community. Here’s a list of things I’ve observed about the regular Einstein teams from listening to and talking to them.

  1. High Strategic Focus: Identifying very early in the season what strategies can help them reach the number one seed. For example, are autonomous points the first sort after W/L/T? Then you better believe they’re building the bot around maximizing autonomous points first. Identify how many points you need to score to win qualification matches even if your partners can only accomplish the bare minimum. How fast will you need to cycle to score that many points? Keep of all this in mind when you’re prototyping mechanisms and measuring their speed.
  2. Identifying cost v.s. reward: Some of the best robots in FRC choose to not play certain aspects of the game. In 2008 1114 choose to run super fast hurdles instead of being able to place the ball on the overpass at the end of the match. In 2013 all three robots from the world championship alliance couldn’t climb above level 1. This year lots of teams that focused entirely on gears did incredibly well. It’s important to identify when there is an aspect of the game that is too difficult to justify the points it scores.
  3. Constant iteration: The six week build season is a myth, and every single team on Einstein knows it. You need to be working on improving your performance at every event you go to. This includes driver practice, new and improved mechanisms, and additional autonomous routines. This year 254 was already one of the worlds top fuel robots before championships. They still choose to manufacture a new faster shooter for worlds because they knew everyone would be getting faster. Don’t forget about developing cheesecake mechanisms, because your opponents definitely will be.
  4. Reliability: Nothing your robot does on the field can be left to chance. Every mechanism needs to be tested in prototype, then on the practice bot, then maybe it can go on the competition bot. Every system needs to be checked before every match. Every autonomous routine should be run and verified heavily before you ever use it in a match. You should have spares ready to go for all your motors, and mechanisms that could be damaged. Zip tie your battery connector, have contingency plans for if sensors fail, anything you can do to minimize the chance of not contributing in a match.
8

The difference between an 85th percentile team and a 99th percentile team is an immense amount of effort in all aspects of your program There’s not just “one weird old trick” that makes these teams better; it’s a thousand little things all put together.

The biggest things I can think of:

  • Willingness to iterate quickly / “fail faster”
  • Attention to detail; never letting a potential failure mode remain unchecked
  • An honest analysis of the team’s shortcomings and active efforts to improve on all aspects of them, every year.
  • Good use of scouting and other analytic data
  • A practice robot and not being willing to stop work at Stop Build Day
  • A strong team culture that is never satisfied with mediocrity and is willing to do whatever it takes to find success.

Things that are not on the “required” list for sure:

  • A hard requirement to complete all aspects of the game challenge well every year
  • Overworking team members past the point of burnout
  • Ignoring a team’s current limitations in lieu of an optimistic view of their success.
9

One word, mentors.

10

Not just mentors… mentors who are willing to go above and beyond what any sane person would understand to be reasonable.

11

This is a big part of it. If your team is doing things right, mentors exist outside of the four-year cycle of your student team members and therefore can hold and pass along the institutional knowledge that enables sustained success.

In particular, I’ve noticed that MANY (not all) of the most successful teams I have encountered have FRC alum mentors. These are people who have a deep history with the program and went off to college to further build their skills while (usually) staying connected to FRC. In doing so, they have built and bring with them a very specific and useful skill set to a team (what it takes to perform at a high level and how to execute in the context of FRC’s constraints like available/legal parts, weight limit, etc.). If they’ve already spent a big chunk of their lives within this world, they know a lot of shortcuts and what specific skills to build into your current students to achieve greatness on and off the field. It should be noted that this is only my impression from the outside looking in. As a (relatively) young team, we are still waiting for our alums to graduate and/or come home.

You can get a purely industry-based mentor this skill-set but it takes a significant amount of time and effort. [Note: this time and effort is TOTALLY worth it for a lot of reasons]

12

When all team members (mentors & students) are going in the same direction, it’s much easier to iterate with focus. Focus + Drive is what makes a good team great, IMO.

13

AYYYYY

14

I’ll echo IndySam about mentors…but also want to point out how critical school/community support is in growing a great sustainable team.

15

They’re testing more things than you. More potential robot systems, more refinements to the systems they pick, more driver practice against opposition, more time to do it all (through a practice robot). And then they’re setting up a support structure that allows them to do all that sustainably–enough kids, enough mentors, enough funding, enough space, so on so forth.

16

This is probably the crux of it. There’s a saying that the last 20% takes as much work as the first 80% combined, and I think that’s very true in FRC. Getting the really little details right in every aspect of the program is the difference maker.

17