Is a robot that does more better?

redesign

#1

So, my team just had a week 2 competition and we won. We have a (working) hatch bot and does a level 2 climb (consistently). Our allainace captain at our first competition had us play as a defense bot and that worked very well for us too. Coming out of that win, my team beleives being able to do a level 3 climb will be better for us in the long run, and help us go tonthe championship. However, to do this we have to add some major changes to our robot (which is hard to add more stuff because we designed a small bot). And by doing that we also risk not improving on what we already have and not testing the new mechanism in time for our next competition. My question is, would it be more beneficial for our team to improve on what we already have or add more so we can do more. Like, do people know from who has been to championship if for championships which is better for us to do and what will more likely get us farther?


#2

From past expierences, I’d say improve on what you have. Most teams go for robots that “specialize” instead of trying to do everything. You have to take reliability into account. If you focus on adding more, will you become unrealiable at what you have? Do you have a set in stone strategy that works for you, even if something on your robot doesn’t work as it should? Keep things like that in mind when making a decision like this.


#3

If you have five tasks and thirty “units” of resource to allocate, it’s better to do three things at 10/10 than all five at 6/10.


#4

Improve on what you have, 100%. This year, 1646 super focused on making low Cargo and hatches our specialty. We focus so much on cargo, we didn’t even run hatches our first competition. But because we became very good at one thing, people noticed us. We won matches that, based on past histories, we should have been absolutely slaughtered in. All because we focused on doing one thing really well.
Growing teams often look at the best and say, “They can do everything great! We want to be like them!” The problem is, the can only choose one of either “everything” or “great”, and they don’t often choose the latter.


#5

I already know I’m going to be the outlier in this thread because my opinion goes against what I would say for 9/10 of the past games.

This game is a little different to me. A level 3 climb is almost required to seed high. If your goal is to be a high seeded captain you’ll need to build one. If you want to be picked by a top seeded bot you need to have what they want. When we seeded first at our regional we picked the best CARGO and HATCH bot, with the bonus that they also had a level 3 climb that we could use that if ours failed. Our second pick was a robot with a great drive team that we trusted to play defense. I unfortunately don’t see a role this year for a HATCH specialist when Cargo is worth 50% more. (that’s not to say you’re not useful. I’d just put you as a defender if you had a great drive team).

This will be our 7th season in a row going to champs, and we had 1 good run to get to Einstein. I do not know your team’s capabilities, but if it is in your ability I would build a level 3 climber or a cargo manipulator. If that’s not possible I HIGHLY SUGGEST you focus on driver practice and be an amazing drive team that a top team would notice and want to select you for their second pick to defend. I don’t see a hatch bot being a first round pick for point scoring, so you’d be my pick for defense.

Just my 2c


#6

It’s better to do one or two things well than four or five things sporadically. There are only a handful of teams who can do everything well for most games, and they often choose to do one or two things spectacularly at the cost of others. For many years, FRC has been played by alliances of three robots, and in eliminations, these alliances have been chosen by a top seeding team to complement/support each other.

If you really do get that new manipulator working well, and you can integrate it with what you have and stay under weight, sure, go for it. But usually you’re going to do better by improving at the things you’re already good at.

Also, if you aren’t already, spend some time on scouting. Even if you aren’t picking as an alliance captain, it helps to know which potential alliance captains might want to select you, and communicate with them while there are still a few rounds of qualifications left. Sometimes they’re looking for something specific, and you can do better at getting picked by proving that capability to them than by improving your seeding rank.


#7

To add on, we were picked by our alliance captain because of our climb.


#8

You might be picked next event because of the defense you played in your last event.

That alone might make your team go far in future events, without adding anything to your bot.


#9

Hi TechTigers!

First off, congrats on your win.

I can give you an account of what 5686 is changing for Hartford if it helps. After missing elims, alliance captains came to us to say that they wanted to pick us, but hatch panels were our weakness. We have maximized the speed of our L2 climb, but there are areas where we can get faster and more efficient.

We are going to dial in our cargo shooter to be quicker, and have more drive practice.

Next we are redesigning our entire hatch panel arm to be able to line up quickly at the feeder station, at various angles, and then score quickly with less driver precision.

We aren’t adding new features because we have many areas we can still maximize by practicing, becoming faster at, and just being more consistent.


#10

Using raw ranking stats (not component OPR), there are exactly 3 teams (out of ~2000 team-event combos) who are > 90th percentile in all 4 ranked game elements (SandStorm, Hatch, Cargo, and HAB Climbs). Those teams are:

330, 525, 1676

OP: If you watch match video of these robots you see just how well-practiced they are in all game aspects (and possibly got lucky with the SS bonuses). Following along those lines, I believe it would be better for your team to do what they can to get more practice on the mechanisms you have rather than adding a brand new mechanism and getting no practice at all.


#11

Two responses to this I’m tempted to give:

  1. No, a robot that does more is not always better. In FRC it is far better to be good at a few things than average at many things.

  2. That said, you should never stop iterating if you can. Continuous improvement is how the powerhouse teams stay competitive all year. Do it with care, and maybe don’t rebuild the whole robot if you’re not sure it will work as well, but constantly look for opportunities to make changes.

Now about the specific question - I think the closer you get to champs the less you will need an L3 climb for the ranking points. L2 climbs will get more common as the quality of events improves, and 2 L2’s plus L1 is a ranking point all the same. That said the L3 has plenty of value as a source of points which is worth considering.


#12

I completely agree about the iteration part. The only reason my team did so well in 2015 is because we iterated and continuously improved our robot.