Experienced and accomplished mentors and teams: How did you level up?

I know this is a pretty open ended question so let me set some context.

I’m in my 5th season with a 10 year old team. We went to World’s as a Rookie team and again on an alliance.

We haven’t had a lot of success since 2019. We barely missed World’s in 2020 and since then, have been steadily performing worse with each competition. The school supports us, we have a solid array of sponsors. Student demand is high, we typically have over 50 kids registered each season.

So the ingredients are all there to field a competitive Robot yet lately, we’re just not able to.

I believe every successful team has experienced this moment and must have had a ‘Level Up’ moment. Something they changed about the way they do things that brought with it a higher capability; a better chance at being successful.

Any thoughts?

8 Likes

what does your build season schedule look like?
How often do you change or optimize designs between competitions?
How much drive practice?

3 Likes

One key is being able to do a good retrospective of each event and the season as a whole. Where did you go wrong? Being able to trace your issues back to specific decisions or decision making processes is important. Did your game analysis go wrong, and you focused on the wrong parts of the game? Did you not spend enough time practicing driving? Did you miss out on prototyping, finding out things like critical dimensions too late in the process?

That retrospective is key, because every team is going to have different areas that are an issue each year.

28 Likes

KISS - Keep It Simple Stupid. We make a list of what we want the robot to do by DCPM and a list of bare minimum for Week 1 comp and start with building the bare minimum and working towards or inevitable goal. This year we wanted a cycler that could do speaker and amp as well as climb at the end. Week 1 comp we were just a speaker cycler with a 2/3 note auto. Comp 2 is this weekend and we plan to cycle speaker with capabilities for amp if needed. And at DCMP in ~2-3 week we plan to possible have a climber attached. 2 years ago we were always 30/30 at our events and now we are an alliance captain or high first pick.

4 Likes

We run Mon-Thur from 330-800, eating dinner together every night. As comp gets closer, we come in on weekends.

Last season we made some mods and they yielded some success. This year we were constrained to 1 comp. But we came 31 out of 36 teams after quals. The robot was just kinda…wrong.

With our only comp being week 1, we struggled to get a design built with enough time to get reps on the robot to a level I think we needed…but our issue wasn’t drive, it was the robot.

I really wish we were in a geographic location that allowed us to compete more. We lost our regional when covid hit and too many teams folded, FIRST won’t bring it back. Our only comp is a flight away, it’s week 1.

Adding to Jon’s post, do a retrospective on the robots that beat yours. How did their robots differ from yours? How did those differences make their robots more effective? Can you implement any of those differences into your robot?

That said, the number ONE thing that as made my former and current teams better is just playing more. Do fundraising to go to additional regional and/or offseason events. Yes I know this hard and may not be possible.

1 Like

This is valuable feedback, thanks. We did the retrospective this week. I would say it was partially successful. More than anything, I think it opened my eyes to our team believing we’re better than we are.

I’m doubling down on finding solutions to this before next season.

what the timeline for working on the robot mechanically/prototyping/programming?

we basically work on parallel streams on each aspect of the robot from kickofff right to comp

so programming is working on the bot at the same time as it is being built? or are there times where prog or build has the bot exclusively?

Failing. A lot.

34 Likes

One of the most important things that my team has learned is that if we think we’re ahead of schedule, we’re not.

6 Likes

Not finishing the robot early enough and not getting enough practice can manifest itself as a broken robot. You haven’t broken the robot enough at home so you don’t know how to make it more robust.

We’re not a top team but I’ve watched a lot of videos from other mentors to figure out how to improve our team. The thing that made the biggest difference from this year to last, is that we had a mechanically complete robot by week 5, and a “competition ready” robot by week 6 (competition was on the the 9th weekend of build season). Despite this accomplishment, my number one reflection on our season is that we didn’t build fast enough (along with “do it right the first time”).

The two talks I’d recommend are Mike Corsetto’s strategic design talk and Adam Heard’s themes in simplicity talk.

Talking with 1678’s mentors, they have a practice drivebase built before the season, and a competition drivebase done at around the first week. Everyone’s seen how they have an alpha bot mid week 2. It’s not because they work absurd hours, it’s because they’ve optimized their process around building robots faster.

In Adam Heard’s simplicity talk, he says you should have your drivetrain moving by day 9, and the robot finished by day 29. He also says that if you’re partway through the season and not on track for that, you should figure out how to drop scope and simplify your robot to accomplish that.

You also really need two events to get the most out of FIRST IMO. It’s unfortunate that geography and FIRST not having a universal points system makes that difficult for you.

11 Likes

Having been on every type of team from low-resource to very high-resource and high performance, I kind of disagree with this. My opinion is that most good teams exist in a state of flux, where things are generally good but they’re not going 100% right all at once. Members join, members leave, sponsors come and go. Every team has things they are confident about and things they know they’re not the best at.

I’ve helped three teams (in three different districts) win their first ever events, and in each case it was my first year mentoring them. Maybe that means I’m an expert in leveling up low-resource teams. But here’s the thing: there was no formula. I wish I could say “this is how you - yes, you - can win your first ever event”, but at this point I still can’t do that. So maybe I’ll just tell you the story of each team.

3322’s first event win ever was Livonia 2012. We won that event because we had a really accurate shooter that could shoot from the key. We used PID back when most teams still didn’t know how to use it correctly, and spent hours upon hours dialing in our one shot. It paid off.

2930’s first event win ever was PNW DCMP 2015. We won that event because we followed the meta-game and determined that we had a higher likelihood of getting picked for playoffs at DCMP if we had a can grabber. So we made a PVC can grabber and, through a stroke of luck, figured out that it was the fastest can grabber at the event. We videotaped every can grabber at the event and, through frame-by-frame analysis, confirmed that ours was the fastest. We then sent scouts to the top teams showing this information, got picked, and the rest was history.

3142’s first event win ever was Bridgewater 2022. Our success that year was a combination of what I said in the last two paragraphs. We had an accurate shooter, but the reason we did was because we followed all of the developments being posted here. Some very helpful teams posted that their testing showed that hub bounce-outs were an issue. So we made a shooter with top-bottom rollers to ensure the ball had no spin, and chose a high trajectory as well.

Reading those back, I think it’s really important to follow developments in the game, in order to position yourself correctly. If anything, that should put your team on the right path.

6 Likes

Specifically: Fail Faster

Get a prototype done in days instead of a week. That way, you can find out what you don’t like about it, and get another prototype made in another couple days, and find what you hate about that one. Repeat this 4-10 times for a single mechanism, and you’ll find you have a pretty good mechanism by competition.

19 Likes

Some thoughts from watching your first playoff match - these might be kind of year specific, but maybe there’s good generalities here:

  • Your arm is very bouncy. We’ve not done terribly many arms, but when we do we put a large amount of reduction on it and use a trapezoid profile for the motion. It doesn’t even have gravity correction, but it has so much torque as to just ignore that. It might not be the fastest in the world, but is faster than waiting for the oscillations to settle, but it also doesn’t need the elastic.
  • Several speaker shots fall short. Did you build practice elements and design the shooter to make those shots?
  • You seem to have to move the arm even when shooting right against the subwoofer. For us this year, we don’t have to move anything when lining up for the close shots (and that’s also the only shot we really do).

General stuff that really leveled us up around 2020-ish:

  • Organize stuff in smaller bins and organizers than the FIRST totes, except for really bulky stuff. Actually have spare fasteners at events that match what is on the robot instead of random whatevers.
  • Making sure the robot works by the first event, meaning it drives, scores, scores in auto, scores in end game
  • Prioritization of RPs - 2020, 2022, and 2023 especially with end game RP, 2024 was less so because trap was too hard
  • Somewhat bold design choices - 2019 we were too meek in our strategy choices and took too long to decide on them (like 2.5 weeks)
  • Build robots elements out of wood that can score - and sometimes it’ll just stay on your robot but that’s okay
5 Likes

Sometimes.

I meant more, accept failure as a fact of learning though. It’s through mistakes that you learn the most, pick yourself up, and do it again until you make it right.

5 Likes

This is fantastic feedback, I really appreciate it.

In their defense, FIRST Alberta is doing what they can to get a regional back, it’s just a numbers game. We need 15 local teams, I think we have 7.

3 Likes

Thanks for watching.

Our practice field was accurate, our shooter just started having issues at comp. We chose to accelerate the note from stealth wheels on the sides of our shooter rather than top and bottom. In theory this should have been ok. It wasn’t. If the note came into the shooter more to one side, the energy was enough to really shoot it. We spent the week prior to comp working on that arm bounce you detected and just ‘hoped’ the shooter would be ok.

hope…

oy