The Quest for Einstein

[Monochron]

Originally Posted by Citrus Dad

I think we've built niche robots in each year of 2013, 2014 and this year. . . This year we focused on cangrabbing and were not extraordinary at stacking, just very reliable. . . We built our robot to fit a role in an alliance--just good enough.

I think we are talking about two different scales here. To many teams the fact that you were "reliable" at stacking AND you focused on cangrabbing means that you were not a niche robot. To most teams a niche robot is one who does one or two non-main-scoring activities well and typically doesn't do other activities.

Even in 2014 your robot could play most any role very effectively, it is just that at your level of play there were other robots who did other tasks better than your robot could. So naturally you fall into roles. That is just a natural part of competition, that robots will fill roles that they excel at. The fact that you could fill most any role (and do so better than a large amount of teams) means that you weren't really a specialist.

[Chris is me]

Originally Posted by Monochron

Even in 2014 your robot could play most any role very effectively, it is just that at your level of play there were other robots who did other tasks better than your robot could. So naturally you fall into roles. That is just a natural part of competition, that robots will fill roles that they excel at. The fact that you could fill most any role (and do so better than a large amount of teams) means that you weren't really a specialist.

1678 was not particularly good at high goal shooting in teleop, due to the distance away from the goal needed for the shot and the vulnerability to defense. Since anybody that could shoot over the truss was technically able to shoot in the high goal, your caveat as written would eliminate any robot from being considered a "truss specialist". This specialized role was probably the best task for a mid-level team in 2014, as it was significantly easier than high goal shooting, contributed consistent points regardless of number of assists / method of finishing, and was a valuable way to pass the ball to a human player.

[sciencenuetzel]

Originally Posted by Citrus Dad

This year we focused on cangrabbing and were not extraordinary at stacking, just very reliable.

1678 only stacked 5 high with a can. Their main focus was making sure they had those two extra step cans. At the beginning of the season they were working on trying to get all 4 cans from the step before they scrapped that idea.
(Please correct me if I am wrong...) 1671 was probably better at stacking which is one of the main reasons why this alliance was the world champions.

1678 is definitely good at identifying the niche they need to play at the highest levels of competition.

[Noudvanbrunscho]

Originally Posted by sciencenuetzel

1678 only stacked 5 high with a can. Their main focus was making sure they had those two extra step cans. At the beginning of the season they were working on trying to get all 4 cans from the step before they scrapped that idea.
(Please correct me if I am wrong...) 1671 was probably better at stacking which is one of the main reasons why this alliance was the world champions.

1678 is definitely good at identifying the niche they need to play at the highest levels of competition.

You won't become 2nd seeded in a division with 'only' your focus on a cangrabber. It was a perfect robot with the fastest can grabber i've seen. (correct me if i'm wrong)
Indeed, the most deserved champions.

[AdamHeard]

Originally Posted by Noudvanbrunscho

You won't become 2nd seeded in a division with 'only' your focus on a cangrabber. It was a perfect robot with the fastest can grabber i've seen. (correct me if i'm wrong)
Indeed, the most deserved champions.

I think it's worth noting that 1678 was not a perfect robot (and they'd admit this). You don't need to be perfect to be competitive and win the championships.

[Monochron]

Originally Posted by Chris is me

Since anybody that could shoot over the truss was technically able to shoot in the high goal, your caveat as written would eliminate any robot from being considered a "truss specialist".

Not quite, it was possible to carefully lob the ball over the truss at a trajectory that wouldn't work on the high goal. In fact one of first prototypes last year did exactly that. I see your point though, that it is a hazy clarification. And while 1678 didn't have a fantastic, consistent high-goal shooter, they still played that role effectively in matches. I'm guessing they designed their robot to play a role at the high level of play, and be able to still score points at a low level of play. I think the word "niche" in the context of FRC means different things to different people. Team caliber may play into that difference.

[Doug G]

Originally Posted by AdamHeard

I think it's worth noting that 1678 was not a perfect robot (and they'd admit this). You don't need to be perfect to be competitive and win the championships.

But what 1678 was able to do, they did perfectly. Every match.. Seriously, I'd love to see the scout data for the standard deviation of their capped stack points per match. Incredibly consistent at the regionals I saw. And Can Grabbing?? How many can battles did they lose? I know they didn't always need to use them, but did they ever not win a can battle? Maybe they did, but I never saw it happen.

Getting back to the OP... It has been said plenty of times already... Practice, Practice, Practice. Find a way to get the students time with the robot. Ever since we started emphasizing the amount of practice time with our robot (or a 2nd bot), our level of competition skyrocketed. This last year we had students practicing several times a week starting in week 4. Granted, our robots have been pretty simple, but we still can compete at a high level. At some point we will "up" our game when it comes to robot design and when we do, we should be able to get further in division elims. Someday...

[sciencenuetzel]

Originally Posted by Noudvanbrunscho

You won't become 2nd seeded in a division with 'only' your focus on can grabber

My use of the word only wasn't meant to say that they weren't great at it, but to point out that stacking wasn't the first goal they had in mind. The first goal was can grabbing but obviously they were good at stacking as well.

[logank013]

Originally Posted by Doug G

But what 1678 was able to do, they did perfectly. Every match..

Isn't this what Karthik says at his presentations? Don't do 3 things mediocre. Focus on 1 thing you can do really really well. There were only a select handful of teams that could do everything (3 tote auto, can burglars, totes, cans, L and HP) this year. I can only think of 1114 and 2056. Many of the top teams were missing one thing like 254 (Can burglars) and 118 (no hp unless they formed a tethered ramp I didn't know about). And even one of the best teams this year was 148 and, they were missing two qualities (No L and Can Burglars). That's the biggest thing that many teams emphasize and our team has done some emphasizing doing a few things close to perfect is better than doing all things just ok.

[carpedav000]

Originally Posted by logank013

Isn't this what Karthik says at his presentations? Don't do 3 things mediocre. Focus on 1 thing you can do really really well. There were only a select handful of teams that could do everything (3 tote auto, can burglars, totes, cans, L and HP) this year. I can only think of 1114 and 2056. Many of the top teams were missing one thing like 254 (Can burglars) and 118 (no hp unless they formed a tethered ramp I didn't know about). And even one of the best teams this year was 148 and, they were missing two qualities (No L and Can Burglars). That's the biggest thing that many teams emphasize and our team has done some emphasizing doing a few things close to perfect is better than doing all things just ok.

From looking at 148's robot it could probably play landfill. Robin would be dragged around for the ride, though.

[Citrus Dad]

Originally Posted by sciencenuetzel

1678 only stacked 5 high with a can. Their main focus was making sure they had those two extra step cans. At the beginning of the season they were working on trying to get all 4 cans from the step before they scrapped that idea.
(Please correct me if I am wrong...) 1671 was probably better at stacking which is one of the main reasons why this alliance was the world champions.

1678 is definitely good at identifying the niche they need to play at the highest levels of competition.

Yes, 1671 was better at stacking, which is why they outqualified us at Sacramento (yes, they did!). We sold 118 on the cangrabbing niche plus the ability to run auto (another niche), so 118 selected us instead.

[Citrus Dad]

Originally Posted by Noudvanbrunscho

You won't become 2nd seeded in a division with 'only' your focus on a cangrabber. It was a perfect robot with the fastest can grabber i've seen. (correct me if i'm wrong)
Indeed, the most deserved champions.

We got 2nd because of our reliability (which is critical to getting to Einstein). Several teams were putting up more points than us in single matches. In 2014 and 2013, our consistency match to match was the difference as well. We have never been "brilliant" in a Champs match. (Our 2014 Inland Empire finals with 399 counts as our one exception.)

[Citrus Dad]

Originally Posted by Doug G

But what 1678 was able to do, they did perfectly. Every match.. Seriously, I'd love to see the scout data for the standard deviation of their capped stack points per match. Incredibly consistent at the regionals I saw. And Can Grabbing?? How many can battles did they lose? I know they didn't always need to use them, but did they ever not win a can battle? Maybe they did, but I never saw it happen.

Getting back to the OP... It has been said plenty of times already... Practice, Practice, Practice. Find a way to get the students time with the robot. Ever since we started emphasizing the amount of practice time with our robot (or a 2nd bot), our level of competition skyrocketed. This last year we had students practicing several times a week starting in week 4. Granted, our robots have been pretty simple, but we still can compete at a high level. At some point we will "up" our game when it comes to robot design and when we do, we should be able to get further in division elims. Someday...

And 701 lifted there game this year so that they made a Regional final as the captain and qualified for Champs. And we saw that again at Capital City Classic when we combined to score over 220 points

[William Frost]

Originally Posted by Doug G

But what 1678 was able to do, they did perfectly. Every match.. Seriously, I'd love to see the scout data for the standard deviation of their capped stack points per match. Incredibly consistent at the regionals I saw. And Can Grabbing?? How many can battles did they lose? I know they didn't always need to use them, but did they ever not win a can battle? Maybe they did, but I never saw it happen.

Getting back to the OP... It has been said plenty of times already... Practice, Practice, Practice. Find a way to get the students time with the robot. Ever since we started emphasizing the amount of practice time with our robot (or a 2nd bot), our level of competition skyrocketed. This last year we had students practicing several times a week starting in week 4. Granted, our robots have been pretty simple, but we still can compete at a high level. At some point we will "up" our game when it comes to robot design and when we do, we should be able to get further in division elims. Someday...

This is a screenshot of our scouting app that shows a graph of the number of totes we stacked throughout our matches on Newton. The lowest data point is 15 (Three stacks of five) and the highest is 21 (Three stacks of five and one stack of six).

[Citrus Dad]

Originally Posted by MrJohnston

Everybody has their opinions about what it takes to reach Einstein each year. Many teams make that their annual goal. Most don't make it. A few make it regularly. We often refer to those teams "elite." Each year we can't wait to see what sort of " robotic genius" teams like the Cheesy Poofs produce.

Two years ago, 948 made it to St. Louis for the first time ever - in its 11th year. We then had a very strong run through Archimedes, losing only one or two matches and earning an alliance captaincy. Of course, we met the Poofs in the quarter-finals and watch the rest of the event from the bleachers...We had a fantastic year and now realize we *can* compete at the highest levels. This year, our eyes are set on Einstein again... We are not "elite," but would like to develop our team so that we can annually compete at an elite level.

I know there are many teams out there much like ours. So, I'd like to pose the question:

What do folks think it takes to to annually compete at this level? What should clubs like mine do in order to accomplish this?

I brought the OP back to the top again to refocus the discussion. I think the keys are 1) Be reliable and consistent at the task. You almost never see 254 or 1114 fail in a match. The first corollary is to problem solve quickly and efficiently; the second is to institute check lists to avoid the obvious failures (which cost us in 2013.) 2) Identify the highest value added tasks beyond the primary offensive tasks. Every top team can run the auto routine at the highest level in a Regional. Again, being reliable at that task is key. 3) Look for niche roles with high value and how that will fit with the strongest robots.

948 had a great full court shooter in 2013 that was among the most consistent at the task. I think consistency was the issue that kept 948 from going to the next level. In comparison, 1983 was quite as spectacular a scorer, but was more consistent and got 2nd in Curie that year. If we didn't both have ground pick ups, we might have allied that year.