Mythical Six Week Build Season

[Jon Stratis]

Originally Posted by techhelpbb

That's fine. You built the robot and it placed highly in competition.
My point is that building the robot is not enough.
The fact you can take longer to build the robot is not always about adding features.

Team 11 & 193 (both in Mount Olive High School) both student led teams build generally at least 3 robots sometimes 4 in 6 weeks. Sometimes we add on a few prototypes as we go. One major reason we split the team was because with so many people on just Team 11 wouldn't get the full experience they could have. Now we facilitate that at the cost of a whole extra team.

So it's not just about building the robot. If it was I wouldn't need FIRST.

I'm glad you build a lot of robots every year. Even if the season was twice as long as it is now, my team (and many, many others, probably well over 1/2 of FIRST) wouldn't have the funds to do so. I still don't see, however, how this plays into the length of the build season or whether we should have a stop build day. Even with a stop build day, what's stopping you from going year-round and building 20 robots each year? If getting MORE the the required number of robots done in 6 weeks is possible, why is that 6 week limitation a problem?

[aldaeron]

For me, the build season is a major selling point like Al mentioned. When you tell the general public that a bunch of high school kids built this fantastic robot in 6 weeks - they are always astonished. I typically correlate this to a real world competitive bid for a project demo - if you don't have a working demo at the deadline, you're out!

The main thing I worry about with regard to the 6 week build season limit is newer teams that compete for one or two season and then fold. The first year I mentored we failed spectacularly (caused by our lack of knowledge of "what works in FRC" training resources). While it was a great "learning experience" for the kids, most of them did not return the following year.

I recall seeing some statistics recently about returning percentage of rookies being quite poor (someone better at searching - help me out here). My experience is that FRC can be quite a steep learning curve and that with more time, newer teams would be more successful and want to keep participating.


My suggestion is to keep veteran teams on the 6 week schedule and allow rookies an extra 1-2 weeks before bagging. My reasoning is that they would get to see other robot reveals, catch up if they are behind, tweak their design slightly based on reveals of top teams or get in some driving practice to ultimately make more competitive robots. Since many regions have pre-ship scrimmages rookies would be able to see their robot in action and then make some tweaks over a longer period (typically there are only 3 days between scrimmage and bag). This would hopefully improve their experience and make them want to return. I realize that in some cases, this could be abused to give these teams an advantage.

Similarly, 2nd year teams that performed poorly in their rookie year could apply for an extra week and FIRST could allow this on a case by case basis.

I am assuming a bad experience is the primary reason teams do not return, though it can also be loss of funding/sponsor/teacher/etc. I also realize the logistics of this proposal could be significant. I am confident CD can help shape this into a better idea.

In my experience the best measure for student success has been the ratio of students to mentors. The more time students get to spend with mentors the more they learn and the more they get out of the program. With a longer build season for rookie teams I think there would be more time for students and mentors to work together. Most new mentors I have met seem to ease into FRC and only come once a week.

Since this discussion comes up every year, I think FIRST should poll teams about build season changes as part of the year end survey (or perhaps a pre-season survey).

-matto-

[Nemo]

Originally Posted by Alan Anderson

So you'd be helping some of Group #2, and you'd be contributing to frustration and/or burnout of the rest. It's a tradeoff, not an obvious win.

Only group #2 (or group #2a if you prefer) is spending more time than they did before. You can call it "increasing burnout" for that group, but I call it making a choice to gain a benefit in exchange for time. Any team that doesn't want to spend additional time can exercise some free will and choose not to spend the additional time.

[techhelpbb]

Originally Posted by Jon Stratis

I'm glad you build a lot of robots every year. Even if the season was twice as long as it is now, my team (and many, many others, probably well over 1/2 of FIRST) wouldn't have the funds to do so. I still don't see, however, how this plays into the length of the build season or whether we should have a stop build day. Even with a stop build day, what's stopping you from going year-round and building 20 robots each year? If getting MORE the the required number of robots done in 6 weeks is possible, why is that 6 week limitation a problem?

You can, and as I pointed out previously the team I mentor soon will be, in a position to go year round in some fashion. However you can't build a robot before the official start of kickoff without knowing the challenge and after that you only have 6 weeks unless you buy enough parts to make at least a second robot.

The issue is we need to build a 2nd robot, and sometimes per team (we have 2) to be competitive and have time to train the drivers on that year's robot even with 20 years of experience in how to build FRC robots. That's at least 2-4 control systems cost and the load of both this work on the mentors and the contributions on the local community. To some level that scales with the number of participants and to some level it does not.

Also it's one of motivation. Once the season ends you've quick burned the time. Mistakes are being made you can't undo and you burn your resources out. Plus when you go to your respective work leaders and tell them it's merely 6 weeks you are not being honest. It's not really 6 weeks. If you go year round you have a side job you probably pay to work. If you go less than year round it's very likely more than 6 weeks. So how would you expect those mentor employers to react to your mere 6 week engagement turning into 10, 12, 16 weeks when they expected it to end?

I know that if I start doing this year round - I am trading the sprint for the long term vision otherwise my coworkers will rightly ask which is my real job.

Originally Posted by aldaeron

The main thing I worry about with regard to the 6 week build season limit is newer teams that compete for one or two season and then fold. The first year I mentored we failed spectacularly (caused by our lack of knowledge of "what works in FRC" training resources). While it was a great "learning experience" for the kids, most of them did not return the following year.

...

Similarly, 2nd year teams that performed poorly in their rookie year could apply for an extra week and FIRST could allow this on a case by case basis.

I am assuming a bad experience is the primary reason teams do not return, though it can also be loss of funding/sponsor/teacher/etc. I also realize the logistics of this proposal could be significant. I am confident CD can help shape this into a better idea.

I know that when we started FRC11 (which was FRC8 the first year) that we all thought that a poor showing would work against our ability to find the funding to continue. A poor showing worked directly against our personal motivations to pour our personal funds in to secure the resources we couldn't get funded by sponsors. It also enabled any detractors to argue we were less than capable making the problem much worse. It is not 20 years ago but these human problems are still the same human problems and that is still the same 6 week build season.

[Jon Stratis]

Originally Posted by techhelpbb

The issue is we need to build a 2nd robot, and sometimes per team (we have 2) to be competitive and have time to train the drivers on that year's robot even with 20 years of experience in how to build FRC robots. That's at least 2-4 control systems cost and the load of both this work on the mentors and the contributions on the local community. To some level that scales with the number of participants and to some level it does not.

This right here is your biggest assumption - that you need to build two robots to be competitive. As I said, my team never has, and except for two years hasn't even designed improvements after stop build, yet we've been to elims a bunch, finalists a bunch, and winners twice. Yeah, we weren't close to Einstein, but you don't have to be up that high to be competitive or successful with your team.

You are choosing to put that added work on yourselves because you feel the benefit outweighs the cost - but it's not something that is required to be competitive.

[Lil' Lavery]

Originally Posted by asid61

The main valid argument I've seen for the "stop build day" is that it would discourage teams from going to Week 1 regionals. However, I argue that teams would go anyway due to locality and the fact that everybody has only had the 6 weeks anyway- it would change very little. However for teams that don't have the funds to build two robots and/or go to only one local regional, the benefits would be immense. Personally I would vastly prefer the extra time to drive, test, and improve things, as we usually go to week 3 and 5 regionals.

There are already teams that avoid week 1 events. Some do it because they don't like being "guinea pigs" for a new game. Others do it because they want more time. 1712 falls in the latter category.

While in a regional format, you can put a bad event behind you and pick up a clean slate in your next event with an improved robot. In the district format, if you have a really bad event, you've essentially condemned yourself to missing DCMP and CMP. 1712 learned that the hard way in 2013. We simply weren't ready for our week 1 event, and missed the eliminations as a result. Despite great improvements at our 2nd event, we missed DCMP because we had put ourselves in a massive points hole. We still competed in week 1 in 2014 because we hate back-to-back events even more than we fear week 1, we skipped out on the very local (and very awesome) week 1 event in 2015 and are doing so again in 2016. It makes much more sense for us to have the extra time to work on our withholding allowance and test our programming on our development chassis (as well as get some lessons learned from watching earlier events). Even if we're competing against teams who are now going into their second event, guaranteeing we have something functional prevents us from prematurely ending our season.

If bag day was removed, I anticipate this aspect would become far more pronounced. Especially in the district format.

[Rangel]

Originally Posted by Jon Stratis

This right here is your biggest assumption - that you need to build two robots to be competitive. As I said, my team never has, and except for two years hasn't even designed improvements after stop build, yet we've been to elims a bunch, finalists a bunch, and winners twice. Yeah, we weren't close to Einstein, but you don't have to be up that high to be competitive or successful with your team.

You are choosing to put that added work on yourselves because you feel the benefit outweighs the cost - but it's not something that is required to be competitive.

To add to this. I know 610 didn't have a practice bot in 2013 but that didn't stop them from becoming world champions and being one of the top robots of that year. Were they the absolute best? I think most would agree they weren't but they were definitely good enough to be competitive and good enough to have a chance at going all the way as they did.

[techhelpbb]

Originally Posted by Jon Stratis

This right here is your biggest assumption - that you need to build two robots to be competitive. As I said, my team never has, and except for two years hasn't even designed improvements after stop build, yet we've been to elims a bunch, finalists a bunch, and winners twice. Yeah, we weren't close to Einstein, but you don't have to be up that high to be competitive or successful with your team.

You are choosing to put that added work on yourselves because you feel the benefit outweighs the cost - but it's not something that is required to be competitive.

Respectfully I disagree.

It may be that your team's circumstances enable you to get this absolutely right and your drivers trained in 6 weeks. However plenty of other teams have this issue beside Team 11 & 193 (who has stuck with one robot as they desire). So I'd love to see a poll of how many teams build a 2nd robot because they feel they need to and as an option because they want to.

It's hardly just an engineering issue. There's weather. There's logistics. There's the 2 pizza problem (remember we are student led and there are a lot of students). The fact we hold the FLL championship for NJ, an FTC competition and an FRC district event. Again when you compare teams you need to really think about what loads are on those resources.

Sure we could make some choices to make the problem smaller - the point is making these choices comes at a price not just for our teams.

Also do not discount luck in the competition itself. Sometimes the difference between success and disaster is just luck. For example Tom posted below that your region matters. That's just luck you happened to have competitors between you and your achievement that weren't better able to stop you or get lucky themselves.

[Tom Bottiglieri]

Originally Posted by Jon Stratis

You are choosing to put that added work on yourselves because you feel the benefit outweighs the cost - but it's not something that is required to be competitive.

The amount of work needed to be competitive varies by region. It would be hard to be a perennial contender at SVR without the use of a second robot unless you got VERY lucky in your early design choices.

[Michael Corsetto]

Originally Posted by Jon Stratis

This right here is your biggest assumption - that you need to build two robots to be competitive. As I said, my team never has, and except for two years hasn't even designed improvements after stop build, yet we've been to elims a bunch, finalists a bunch, and winners twice. Yeah, we weren't close to Einstein, but you don't have to be up that high to be competitive or successful with your team.

You are choosing to put that added work on yourselves because you feel the benefit outweighs the cost - but it's not something that is required to be competitive.

You are assuming that your team's record qualifies as "competitive".

I suppose we (and many others) may have different definitions of "competitive".

-Mike

[zinthorne]

Originally Posted by mrnoble

Awesome that we get to discuss this again.

The unintended consequences will be rampant, but in the end, there will still be three basic categories of FRC teams at competition:

1. top level teams, who work year round and take advantage of every opportunity and loophole available to win more;
2. low resource, low knowledge teams, who will come to competition with poor quality robots, and
3. teams who are willing to work ridiculously hard, but cant yet be called "elite" (if they ever get that far)

In other words, the rich would stay rich, the poor would stay poor, and a bunch of us would just keep chuggin'. So we might as well keep it the way it is.

My opinion.

I agree with this alot.

I like the stop build day. If we were to get rid of it, I feel like it would create an even bigger disparity between the elite and rookies. The elite would have several robots by the end of the season, and there would be a lot of cloning... Also think of all the fun times that we would miss. 1114's harpoons would not be such a cool big deal! They would be pre-build to a specific robot that would play like junk in quals so they would be picked. Part of what i think makes FRC so great is the pressure. The pressure to get your design. It simulates a real world job where you must finish a project by a certain time. Getting rid of stop build day would I believe lower the fun and lower the number of unique designs, and working with what you have.

[AdamHeard]

Removing stop build day won't make the elites better.

The elite teams already are skipping stop build day via practice bots. Elite teams already are doing complete robot rebuilds

[MrForbes]

Originally Posted by Michael Corsetto

You are assuming that your team's record qualifies as "competitive".

I suppose we (and many others) may have different definitions of "competitive".

-Mike

John might have a definition of "competitive" that is relevant to most teams.

[Paul Richardson]

I've seen a few responses mentioning design convergence if we were to eliminate robot bagging. While I'm not necessarily against this, I do see a way to put a bit of a damper on it. We are already required to submit a Bill of Materials, though it's not looked at too closely. If you wanted to limit in-season changes, implement a "feature-freeze" using the BOM.

Teams would submit a BOM online after 6 weeks listing the normal stuff, plus a short list of their robot's subsystems and 'Planned Additions'. These would need to be specific (say "Floor pickup for Game Piece A" and not "Change robot to be better"). Basically, you can make additions, but only if you came up with the general concept of it on your own. This doesn't really prevent copying subsystem designs so much as copying strategies.

This would not prevent can-grabbers from 2015 or minibots from 2011, because everybody would have put those things on their Planned Changes list. It was obvious from the design of the game that those mechanisms would be important. I see this as more of a game design flaw rather than something the rules need to address. Even if we had strict bag rules and no practice robots allowed, teams would just make them at competition.

This would not prevent redesigning existing subsystems (copied or otherwise). Teams routinely redesign intakes, shooters, and such, replacing their original designs under the existing withholding allowance rules. This wouldn't change.

This would prevent a team from copying something like 118's bridge hanger from 2012 (had it been legal). If you don't declare at 6 weeks that you might build a device for hanging from bridges, then you can't add it later. Similarly, probably a lot fewer teams would have had stingers.

This would also prevent a team from copying something like 71's unique drive system from 2002. A team would have to have planned to build a high traction 'walking' drive. If they simply planned to 'drive' they'd be able to change wheel types/sizes, gear ratios, and other tweaks. Switching from tank to H-drive would also need to be declared ahead of time. Teams would be limited to functionality improvements for their existing subsystems, not functionality additions.

Minor Bonus Effects: Encourages stopping design in favor of drive practice/polish, discourages overworking after 6 weeks, and encourages teams to take a better look at the rules.

My Personal Opinion: No bag, no limits. My BOM system would be harder on the bottom group of teams than the top because of the experience difference. Many teams didn't think about can-burglars at all last year because running out of cans was so far off their radar. It'd be frustrating to see all these designs you aren't allowed to build. That's not inspiring or fun.

[AllenGregoryIV]

Originally Posted by AdamHeard

The elite teams already are skipping stop build day via practice bots. Elite teams already are doing complete robot rebuilds

Not even just elite teams. Spectrum is by no means elite but we did a full rebuild this past season and we copied ideas we saw from other teams. (Thanks 118, and others). It was a very hard decision to scrap 7 weeks of work for a new idea but we know our robot wasn't competitive. It took a very long time to get the new "cloned" robot working half way descently. This was definitely a rushed job and better clones could be made with more time but largely it would require you to be be building your clone through most of competition season. You would not have any time to actually test it at a real event. Almost any clone is going to be worse then the orgional (for an entire FRC robot) because you don't have all the experience gained from the prototyping and design phase.

I think 2014 is a good place to look to see if the majority of teams would really clone another robot. It was pretty clear early on that 254 had a very special design once you saw the robot. A lot of teams could have probably used withholding and COTS parts to build something that cloned their robot, but to my knowledge none of the elite teams did that because they believed tweaking and getting better with their own robot had a better likely hood of success. The equation wouldn't change all that much if you got rid of stop build.