Thoughts on Ri3D and BuildBlitz

[JB987]

+ 1 Adam

[KeeganP]

Originally Posted by AdamHeard

This is a little unfair for you to state. You came into this thread looking for people to agree with you, and are unwilling to reconsider your viewpoint.

It goes both ways though, doesn't it? I have my opinion, and you have yours. Yes, I started the thread by voicing my own opinion and wanted to see if anyone had similar thoughts, but I was also curious as to why people would disagree with me. Is that unfair? I see the advantages of Ri3D and BB, but I also see some disadvantages.

If I were to ask you what you would improve upon for Ri3D, what would you say? I'm not asking why you think it's so great (you've already told me, along with everyone else on the thread -- so I know why, and agree with a lot of the reasons), I'm asking what, if anything, you would do to change it.

So far I've gotten maybe three responses (out of 70+) that answer that question.

[Racer26]

Originally Posted by KeeganP

If I were to ask you what you would improve upon for Ri3D, what would you say? I'm not asking why you think it's so great (you've already told me, along with everyone else on the thread -- so I know why, and agree with a lot of the reasons), I'm asking what, if anything, you would do to change it.

So far I've gotten maybe three responses (out of 70+) that answer that question.

I wouldn't change much. The changes from 2013->2014 made a big difference. More teams, better documentation provided quicker, and better blogging of the process as it was happening are all improvements that happened in 2014.

The only thing I'd like to see that hasn't happened yet, is bringing together 6 (or more in 2015?) 72-hour robots to actually play some real matches of the game, perhaps on the first weekend after kickoff, as that will give us some more information about how the game will actually be played, rather than just what robot mechanisms can solve the game's challenges (without the context of defense and so on to frame it).

[mrnoble]

Originally Posted by AquaMorph

I completely agree.

Actually, you don't. You edited my quote to make me say what I didn't say. I fully support everything that the teams did, in terms of the quality of work, the design ideas, the reveals, and even the CAD and BOM documents. I think these things only help FRC in general.

As I said in an earlier post, I know that my opinion on the inner workings of VexPro and AndyMark is limited and ill-informed. As an outside observer, and since we were asked repeatedly by the original poster what we would change, I stated that it [i]looks[i] to me like there might be a connection between the owners of a business working on a project, and the amount of time and energy they would have for their core service. I could very well be wrong about this, as several of you have pointed out.

[sanddrag]

While this may come across as somewhat arrogant, and I don't intend for it to, but if there's anything I've learned in my 12 years in FIRST, it's that if you're a good competitive team, your partners will let you down, nearly every time. Have we been the team to let others down? Of course (especially 696 in 2010), it happens, but I think more often than not, we've carried the weight of the alliance throughout my teams' histories (696 and 968). It's statistics. At a typical regional, there are perhaps 2 really good teams, 3 pretty good teams, 10-12 mediocre teams, and the rest are rather terrible.

If the 72-hour builds help any of our alliance partners in any match score more points (any points actually), work with us better, and in general just be more competent and competitive, it seems like a good thing. It's no fun to play on alliance where you are the only robot capable of scoring and for years (and even easy years like 2011), we have.

[Anupam Goli]

Originally Posted by Racer26

The only thing I'd like to see that hasn't happened yet, is bringing together 6 (or more in 2015?) 72-hour robots to actually play some real matches of the game, perhaps on the first weekend after kickoff, as that will give us some more information about how the game will actually be played, rather than just what robot mechanisms can solve the game's challenges (without the context of defense and so on to frame it).

Something that really would help every team, rookies and veterans alike, is to see a match in action, showing potential strategies, and mainly just to see the timing of certain tasks. One thing that really influences a lot of design decisions on teams is the (mis)timing of completing objectives in a match. I know this would've helped my team in 2012, we used horrible time estimates for each task to be completed and we underestimated how many points could potentially be scored!

[AdamHeard]

Originally Posted by KeeganP

It goes both ways though, doesn't it? I have my opinion, and you have yours. Yes, I started the thread by voicing my own opinion and wanted to see if anyone had similar thoughts, but I was also curious as to why people would disagree with me. Is that unfair? I see the advantages of Ri3D and BB, but I also see some disadvantages.

If I were to ask you what you would improve upon for Ri3D, what would you say? I'm not asking why you think it's so great (you've already told me, along with everyone else on the thread -- so I know why, and agree with a lot of the reasons), I'm asking what, if anything, you would do to change it.

So far I've gotten maybe three responses (out of 70+) that answer that question.

Sorry, I guess I should have been more clear. My apologies.

I think what they are doing is great, and the change I would suggest if I was forced to suggest a change is do more of what they are doing.

[Dale]

I will say anyone considering getting together with a group to do a robot in 72 hours should think twice about why they are doing it. Just as FIRST has a scalibility issue it is tackling with districts, what would happen if in 5 years we had 50 72 hour robots taking place? Could be a crazy number but I'm sure it would be very easy to find 50 groups who'd be willing to do it. Do you think that would have an impact the community?

Here's how I'd like to see this 72 hour robot process improved in the future:

1) Having more than a couple of these 72 hour projects going has the effect of exploring the solution space more thoroughly than I'd like. Last year with one robot in three days we got to see one good way to build a minimally competitive robot. That was good. Six robots this year explored the options so thoroughly that there is less room to innovate. Can you still do things better? Sure. But I think we've now seen the broad outline of what 90% of the robots at the competitions will look like. I'd like to see the number of teams doing 72 hours robots reduced next year.

2) Don't release CAD files or detailed walkthroughs until after the season is over, if ever. Reasons have been discussed previously in this thread.

3) I'd encourage the 72 hours teams to intentionally keep it simple. Many of these teams are run by competitive, brilliant, type-A personalities who naturally want to do the best they can. What they can do in 72 hours is better than many teams can hope to accomplish in six weeks. We run the risk of demotivating kids when their accomplishment over the build season isn't competitive with what the pros could do in three days.

4) Show robots failing in release videos. Don't just edit it to show the times when the robot worked great but show it not working so students can see the designers are human. It will make them feel better when their robot fails.

[KeeganP]

Originally Posted by Anupam Goli

Something that really would help every team, rookies and veterans alike, is to see a match in action, showing potential strategies, and mainly just to see the timing of certain tasks. One thing that really influences a lot of design decisions on teams is the (mis)timing of completing objectives in a match. I know this would've helped my team in 2012, we used horrible time estimates for each task to be completed and we underestimated how many points could potentially be scored!

That's really interesting. I hadn't thought about that before, but this year -- especially with passing -- it would have been cool to see just how long/easily/consistently robots can realistically pass a ball between them. Our team is hoping to host a day near the end of Build Season that we can assemble a field and test with multiple teams, and that'll be our first time being able to see how the game is played without just guessing.

[KeeganP]

Originally Posted by Dale

Here's how I'd like to see this 72 hour robot process improved in the future:

1) Having more than a couple of these 72 hour projects going has the effect of exploring the solution space more thoroughly than I'd like. Last year with one robot in three days we got to see one good way to build a minimally competitive robot. That was good. Six robots this year explored the options so thoroughly that there is less room to innovate. Can you still do things better? Sure. But I think we've now seen the broad outline of what 90% of the robots at the competitions will look like. I'd like to see the number of teams doing 72 hours robots reduced next year.

2) Don't release CAD files or detailed walkthroughs until after the season is over, if ever. Reasons have been discussed previously in this thread.

3) I'd encourage the 72 hours teams to intentionally keep it simple. Many of these teams are run by competitive, brilliant, type-A personalities who naturally want to do the best they can. What they can do in 72 hours is better than many teams can hope to accomplish in six weeks. We run the risk of demotivating kids when their accomplishment over the build season isn't competitive with what the pros could do in three days.

4) Show robots failing in release videos. Don't just edit it to show the times when the robot worked great but show it not working so students can see the designers are human. It will make them feel better when their robot fails.

+1 on this. I think that, like a pendulum (bad analogy, I know), it's possible to swing too far from one side to another before finally coming to rest at a nice even place between the extremes. Perhaps this year it swung too far towards giving teams 5 options for "robots in a box", and last year we were too close to not being beneficial enough for teams. Maybe there's a common ground to be found?

[AquaMorph]

Originally Posted by mrnoble

Actually, you don't. You edited my quote to make me say what I didn't say. I fully support everything that the teams did, in terms of the quality of work, the design ideas, the reveals, and even the CAD and BOM documents. I think these things only help FRC in general.

As I said in an earlier post, I know that my opinion on the inner workings of VexPro and AndyMark is limited and ill-informed. As an outside observer, and since we were asked repeatedly by the original poster what we would change, I stated that it [i]looks[i] to me like there might be a connection between the owners of a business working on a project, and the amount of time and energy they would have for their core service. I could very well be wrong about this, as several of you have pointed out.

Oh I am sorry. I didn't link the two posts together and thought you were talking about something else.

I'd like to see it taper off from this year. I think they went a little too far, because everyone was excited by last year's result and wanted to participate.

This was very misleading because we were talking about changes to the whole building a robot in 72 hours. Everything after I agree was my opinion. I was not presenting it as yours.

I'm curious how people would feel if FIRST gave out a kit of parts that had instructions to build a fully working robot for the game. If inspiration is best achieved though having a successful robot why not do this? Maybe give it to rookie teams? Maybe this is the next progression from what people have learned from Ri3D.

[magnets]

Originally Posted by sanddrag

While this may come across as somewhat arrogant, and I don't intend for it to, but if there's anything I've learned in my 12 years in FIRST, it's that if you're a good competitive team, your partners will let you down, nearly every time. Have we been the team to let others down? Of course (especially 696 in 2010), it happens, but I think more often than not, we've carried the weight of the alliance throughout my teams' histories (696 and 968). It's statistics. At a typical regional, there are perhaps 2 really good teams, 3 pretty good teams, 10-12 mediocre teams, and the rest are rather terrible.

If the 72-hour builds help any of our alliance partners in any match score more points (any points actually), work with us better, and in general just be more competent and competitive, it seems like a good thing. It's no fun to play on alliance where you are the only robot capable of scoring and for years (and even easy years like 2011), we have.

I agree. Lowering the ceiling is a bad thing, but there's nothing wrong with raising the floor.

[Steven Donow]

Originally Posted by AquaMorph

I'm curious how people would feel if FIRST gave out a kit of parts that had instructions to build a fully working robot for the game. If inspiration is best achieved though having a successful robot why not do this? Maybe give it to rookie teams? Maybe this is the next progression from what people have learned from Ri3D.

That's a completely different situation, though. That would be FIRST directly influencing the way (strategically) the game is played. And there are plenty of resources (even outside of Ri3D/Build Blitz) to help rookie teams have a direction in their overall robot design; resources/fields that FIRST shouldn't do themselves (game-specific, that is).

I'll admit, last year, I was initially against/not the biggest fan of Ri3D for the very reasons you're posting/arguing. Then, I saw that last year was one of the most even-leveled fields in recent memory. And even then, teams still built "boxes on wheels". I initially credited it to the nature of Ultimate Ascent, but I think Ri3D really did have an impact. This year, with the very nature of the game (ie. the abilitites of your partners directly impact your own success), I truly think we'll see a playing field even more level than last year. Maybe you don't see it as much since you're not in districts, but sometimes, at smaller districts, there are not 24 'average' robots that can play the game. This makes for unexciting competition and a low floor of competition. I think with 6 different robots built/designed in the first 72 hours, the floor WILL be raised.

[Tom Bottiglieri]

Originally Posted by KeeganP

I'm asking what, if anything, you would do to change it.

It would be really nice if all the teams used the official control system and did autonomous mode.

Other than that, I'd like to see more teams, more CAD, and more code.

Our community needs as many resources as it can get. Imagine trying to scale this program around the world without Andy Mark, VexPro, WCP, and other suppliers. It's just not going to happen. It's REALLY hard to find effective mentorship for new teams. Ri3D has the potential to help these teams a ton.

[Racer26]

Originally Posted by AquaMorph

I'm curious how people would feel if FIRST gave out a kit of parts that had instructions to build a fully working robot for the game. If inspiration is best achieved though having a successful robot why not do this? Maybe give it to rookie teams? Maybe this is the next progression from what people have learned from Ri3D.

We are essentially handed instructions to build a fully functioning robot to play a FRC game already in the kitbot, and some teams still struggle to build a box on wheels.