paper: Stop the Stop Build - Counterpoint

[John.Antilla]

Thread created automatically to discuss a document in CD-Media.

Stop the Stop Build - Counterpoint by John.Antilla

[BMiller2559]

I am not sure this is actually a counter point. True, you do address his comments point by point and somewhat disagree but in the end you basically agree that some sort of modification would be helpful especially to lower resource teams that cannot build twin robots.

I really liked your questions at the end and wish something like them had been in the poll. Things like; Does you team build right up to bag day with little practice driving time? Does you team often finish their robot at their fist competition? Does your team build a twin robot to continue to prototype and practice after bag day? These would have all been questions to gain insight into how teams at different levels operate.

[John.Antilla]

I see your point. I thought the speculation that teams will develop the behavior of not competing until late in the season to take advantage of the extra time WAS the counterpoint. Maybe I wasn't clear enough. Thanks for the feedback.

[AllenGregoryIV]

Originally Posted by John.Antilla

I see your point. I thought the speculation that teams will develop the behavior of not competing until late in the season to take advantage of the extra time WAS the counterpoint. Maybe I wasn't clear enough. Thanks for the feedback.

That can't happen, there aren't enough events in enough areas for that to happen. How many events each year don't fill? I'll try to pull the data together this weekend, but I'm sure it's not enough to matter. Most areas are struggling to add enough spots to fill demand.

[gblake]

I like what you have written, but I personally would have finished with a differnt, perhaps more sharply contrasting counterpoint.

The on-the-field performance of teams/robots is not the central focus of the FRC program described on FIRST's website, nor is it the thread tying together the many descriptions offered by FRC's founders and current leaders. However, in the day-to-day operation of many FRC teams, and in the minds of many participants, on-the-field performance becomes the tail that wags the dog. With that in mind, and wanting to reduce the amount of dog-wagging done by that tail ...

A) I believe that having (recreating?) an actual 45-day build season would be (restore?) an excellent part of the foundation of wise compromises that FRC rests upon. FRC should push in this direction.

B) I think that the the on-the-field performance exhibited by the teams that are already doing well (in that part of FIRST), is good enough. I don't mean to say that better performance would be harmful; but if I'm right, I do mean that across the globe, for the teams that aren't struggling, improving the on-the-field part of FRC should not be pulling time, attention, and other resources away from the other parts of FRC.

Now, to get to the heart of it:

For the teams that are not doing well, and are consequently exhibiting the trouble-on-the-field symptom, giving them an 8 hour out-of-bag period each week (or not) seems to be a relatively small item a big picture. I think that properly preparing them to make good use of their 45 days will make the 8 hours unnecessary (the extra time won't hurt, but it won't be *necessary*). If a team makes good use of those first 45 days, they will be ready to do well enough when they roll into their first event. If instead they are currently using the 45 days poorly, it seems intuitive to me to expect that after a year-or-two transition they will be using added 8-hour windows poorly too.

So ... Please don't let a post-stop-build-day weekly unbag period rise to the top of the list of ways to assist struggling teams. Instead let's focus on helping them with project management, and everything that is tied to that subject; along with helping them with the soft skills of recruiting students and mentors, finding funding, having effective meetings, etc.

I know the original paper's author isn't focused so strongly on the 8-hour unbag topic that he is ignoring other ways to help struggling teams. I also know he is part of a large, deeply-caring and highly-motivated FiM team that wants all students to succeed.

But, reading this paper, a reader can get the impression that the OPR tail ought to (and will) wag the FRC dog. Instead I think OPRs and similar symptoms/metrics won't change for the long haul until the dogs are healthy. A few unbagging periods will help a little, but still won't make the dogs healthy. The root causes for struggling will still be there.

Given my beliefs, I'm personally led to believe FRC shouldn't be churning its ranks with debates about ways to extend/expand build seasons. There are much more important fish to fry.

Sure, ask permission to try a handful of small well-designed unbag experiments, but worry about them only after getting permission to carry out pre-build-season experiments that will attack likely root causes, before attacking what is likely to be only a symptom. If that is what is already happening, the paper does itself a disservice by not mentioning building on or amplifying those efforts.

Blake

[Richard Wallace]

Originally Posted by AllenGregoryIV

That can't happen, there aren't enough events in enough areas for that to happen. How many events each year don't fill? I'll try to pull the data together this weekend, but I'm sure it's not enough to matter. Most areas are struggling to add enough spots to fill demand.

Certainly there will not be enough available late-season slots for all the teams that would prefer them, if John's speculation is valid. Lack of availability does not refute the speculation, though. If playing late really is an advantage, then how will FIRST decide which teams get that advantage?

My own (FiM) team prefers playing one district early and another late; e.g. Week 2 and Week 5 in 2016. The early one is an advantage for teams that work hard and finish the robot(s) so they can practice. The late one is an opportunity for improvement. So, for us, John's speculation is not an issue.

For a one-and-done Regional team, I think that John's speculation is valid.

[JB987]

Originally Posted by gblake

I like what you have written, but I personally would have finished with a differnt, perhaps more sharply contrasting counterpoint.

The on-the-field performance of teams/robots is not the central focus of the FRC program described on FIRST's website, nor is it the thread tying together the many descriptions offered by FRC's founders and current leaders. However, in the day-to-day operation of many FRC teams, and in the minds of many participants, on-the-field performance becomes the tail that wags the dog. With that in mind, and wanting to reduce the amount of dog-wagging done by that tail ...

A) I believe that having (recreating?) an actual 45-day build season would be (restore?) an excellent part of the foundation of wise compromises that FRC rests upon. FRC should push in this direction.

B) I think that the the on-the-field performance exhibited by the teams that are already doing well (in that part of FIRST), is good enough. I don't mean to say that better performance would be harmful; but if I'm right, I do mean that across the globe, for the teams that aren't struggling, improving the on-the-field part of FRC should not be pulling time, attention, and other resources away from the other parts of FRC.

Now, to get to the heart of it:

For the teams that are not doing well, and are consequently exhibiting the trouble-on-the-field symptom, giving them an 8 hour out-of-bag period each week (or not) seems to be a relatively small item a big picture. I think that properly preparing them to make good use of their 45 days will make the 8 hours unnecessary (the extra time won't hurt, but it won't be *necessary*). If a team makes good use of those first 45 days, they will be ready to do well enough when they roll into their first event. If instead they are currently using the 45 days poorly, it seems intuitive to me to expect that after a year-or-two transition they will be using added 8-hour windows poorly too.

So ... Please don't let a post-stop-build-day weekly unbag period rise to the top of the list of ways to assist struggling teams. Instead let's focus on helping them with project management, and everything that is tied to that subject; along with helping them with the soft skills of recruiting students and mentors, finding funding, having effective meetings, etc.

I know the original paper's author isn't focused so strongly on the 8-hour unbag topic that he is ignoring other ways to help struggling teams. I also know he is part of a large, deeply-caring and highly-motivated FiM team that wants all students to succeed.

But, reading this paper, a reader can get the impression that the OPR tail ought to (and will) wag the FRC dog. Instead I think OPRs and similar symptoms/metrics won't change for the long haul until the dogs are healthy. A few unbagging periods will help a little, but still won't make the dogs healthy. The root causes for struggling will still be there.

Given my beliefs, I'm personally led to believe FRC shouldn't be churning its ranks with debates about ways to extend/expand build seasons. There are much more important fish to fry.

Sure, ask permission to try a handful of small well-designed unbag experiments, but worry about them only after getting permission to carry out pre-build-season experiments that will attack likely root causes, before attacking what is likely to be only a symptom. If that is what is already happening, the paper does itself a disservice by not mentioning building on or amplifying those efforts.

Blake

Blake,
Much of your point here is based on the attribution of poor choice/time management on the part of struggling teams. In many cases, teams are limited in time they can spend each week and are limited as far as access to build facilities by school or district fiat. Additional weeks to work directly on their robot could help them better deal with time limitations...who knows, maybe even give them time to add an autonomous element to their performance

[Brandon Holley]

Originally Posted by John.Antilla

I see your point. I thought the speculation that teams will develop the behavior of not competing until late in the season to take advantage of the extra time WAS the counterpoint. Maybe I wasn't clear enough. Thanks for the feedback.

This is where an experiment (or targeted data collection) can be pretty helpful I think. Is it simply time with the robot that is making teams better later in the season? Or is the act of executing on a field, with tighter constraints and against real competition forcing iteration/evolution that bumps performance later in the season?

FWIW, my personal starting guess is that its probably 80% the latter, and 20% the former.

With my above viewpoint noted, I don't view competing later as an automatic advantage. You may be potentially gaining more time with your robot, but at the same time, so is everyone else. Per Jim's data, you would need to come out of the gate with a stronger performance if your first event was week 5 versus week 1. This also totally passes the eye test I see each season. You get more time with your robot, but that time comes at the expense of HAVING to be better to reach the same end performance.

-Brando

[Siri]

Originally Posted by Richard Wallace

Certainly there will not be enough available late-season slots for all the teams that would prefer them, if John's speculation is valid. Lack of availability does not refute the speculation, though. If playing late really is an advantage, then how will FIRST decide which teams get that advantage?

My own (FiM) team prefers playing one district early and another late; e.g. Week 2 and Week 5 in 2016. The early one is an advantage for teams that work hard and finish the robot(s) so they can practice. The late one is an opportunity for improvement. So, for us, John's speculation is not an issue.

For a one-and-done Regional team, I think that John's speculation is valid.

It's certainly a reasonable question, so let's examine the dissausive factors. District teams would (and are) put in a position to trade off playing late versus the amount of time they have to improve and recoup between events. They also shoulder the same tradeoff as a one-and-done Regional team, below.

A one-and-done Regional team, unless they are somehow in a regional that's all one-and-dones*, is at a disadvantage with that cohort in terms of actual competition experience. Unfortunately, we can't have any data on the exact value of this alone without dropping SBD. Nonetheless, one could speculate that a one-and-done has to choose which week to attend based on their own Performance vs. Week curve. They need to map and decide beforehand which week they're likely to be above the largest number of other teams they would play. Teams already do this, of course, though it's certainly a difficult calculation and not likely even to be undertaken by all, much less calculated correctly.

• We can surmise that N teams won't follow this impetus, either because they can't themselves logistically, the events aren't available, they don't realize it, or they don't want to.
• There will also be M teams that calculate incorrectly, wanting to go late and finding out that they're outmatched.
• P teams will calculate incorrectly the other way, thinking they should go early when everyone is still weak and realizing they'd personally do better at a late event.
• Some set of N-P are likely to adjust next year's calculus based on their experience.
• And then of course there are those that do or will calculate correctly, and we'll get Qs that routinely choose to go early when no one has competition experience and they can play others who have not yet hit their stride. Certainly this also already happens (actually all of these do).
• And we'll get Rs that successfully give themselves more time to improve, copy successful systems, etc, and can correctly sacrifice whatever actual competition does for performance curves.

Overall, I can certainly agree that the ratio of N-Rs is likely to change based on whether there's an SBD, and that an individual team's set is also likely to change. I don't see why this is a negative, however. It's possible the number of teams who don't get their preferred event may even go up, at least until long-term event planning and team densities catch up. On the other hand, it may also go down. I see no good way to set odds on this. The forces and counterbalances on all sides look robust enough, and I'd expect the league to reach an equilibrium of event distribution with or without SBD. Is there a different reason to think these forces will be overmatched?

[gblake]

Originally Posted by JB987

Blake,
Much of your point here is based on the attribution of poor choice/time management on the part of struggling teams. In many cases, teams are limited in time they can spend each week and are limited as far as access to build facilities by school or district fiat. Additional weeks to work directly on their robot could help them better deal with time limitations...who knows, maybe even give them time to add an autonomous element to their performance

I understand.

However, there are skilled and experienced FRC people (who are excellent teachers/mentors) and teams that can show other teams how to turn one long, hard weekend into a decently-sophisticated, working robot.

If instead of using skilled /experienced people and a long, hard weekend, a team that has been struggling learns project management skills, plans and executes pre-season preparation/practice; then during the season, thinks up, designs, and executes a robot (including simple autonomy ) that is within their abilities and that fits comfortably within the 45 days; I predict success.

What will make or break them (for this part of being a team) is whether they can learn to set very modest goals initially, and only incrementally-increase the sophistication of their goals AFTER they have proven they can handle the modest goals.

With that in mind, unless they are simply limited by outsiders to truly ridiculously small amounts of time per week, I'll bet a very nice dinner that any *motivated* team that has learned to manage a project well, can build a good-enough FRC robot in 45-days (and build a better one the next season).

Basically, the struggling team needs to learn to accomplish (on average) in each of the 45 days (on a simple robot+controls), roughly what the fast-and-furious folks accomplish in each hour (on a reasonably sophisticated robot) during their 48 hour sprints. OBTW, to help the struggling team, they can get detailed design/implementation info for previous-season robots and controls from several sources.

If a struggling team's time-available-to-work is so tightly constrained that they can't accomplish what I described, regardless of how well they manage their project/time; then it would be a *good* idea (not a bad idea) to rethink the entire subject of competing in FRC. FTC, VRC, etc. are almost certainly better choices for them in that situation.

Bottom line: How to build an FRC robot+controls in under one week is well known. With preparation, building a rugged, dependable, allies-like-to-see-you, cookbook-ish, simple bot built in 6 weeks (with time set aside for practicing) for a given season has got to be do-able, unless the team is in a situation that is simply incompatible with FRC. The next season they can adjust their goals.

Blake
PS: Again, all of this is in the context of not letting the tail wag the dog. A simple robot that easily understood and altered is an excellent robot, if the team uses it, and the competitions, to inspire people to consider looking into STEM(ish) careers seriously. The team doesn't have make people into STEM graduates, or overwhelm audiences with the FRC equivalent of a space shuttle. Instead they just need to open peers' eyes to the fact that a STEM career can be rewarding, is worth considering, and is within their grasp.

[AllenGregoryIV]

Originally Posted by Richard Wallace

Certainly there will not be enough available late-season slots for all the teams that would prefer them, if John's speculation is valid. Lack of availability does not refute the speculation, though. If playing late really is an advantage, then how will FIRST decide which teams get that advantage?

My own (FiM) team prefers playing one district early and another late; e.g. Week 2 and Week 5 in 2016. The early one is an advantage for teams that work hard and finish the robot(s) so they can practice. The late one is an opportunity for improvement. So, for us, John's speculation is not an issue.

For a one-and-done Regional team, I think that John's speculation is valid.

It't not an advantage though as you just said competing early is often an advantage. Everyone at each event has the same amount of time. Teams competing at late regional if it is their first regional will still be at a slight disadvantage from the teams that may have been able to get more experience at earlier events. Abolishing SBD or giving more robot access time is only designed to improve the lower and mid tier preforming teams not level the playing field.

More money, skill, people, resources will always be an advantage in this sport (and most sports) but we can at least do everything we can to make it easier to feel accomplished and be inspired.

Each event is scheduled mostly on venue availability. FIRST HQ isn't deciding much anything.

[AdamHeard]

Originally Posted by gblake

I understand.

However, there are skilled and experienced people (who are excellent teachers/mentors) and teams that can show teams how to turn one long, hard weekend into a decently-sophisticated, working robot.

If instead of a using skilled /experienced people and a long, hard weekend, a team that has been struggling learns project management skills, plans and executes pre-season preparation/preactice; then during the season, thinks up, designs, and executes a robot that is within their abilities and that fits comfortably within the 45 days; I predict success.

What will make or break them (for this part of being a team) is whether they can learn to set very modest goals initially, and only incrementally-increase the sophistication of their goals AFTER they have proven they can handle the modest goals.

A unless they are simply limited by outsiders to truly ridiculously small amounts of time per week, I'll bet a very nice dinner that any team that has learned to manage a project well, can build a good-enough FRC robot in 45-days (and build a better one the next season).

Basically, the struggling team needs to learn to accomplish in each of the 45 days (on a simple robot+controls), roughly what the fast-and-furious folks accomplish in each hour (on a reasonably sophisticated robot) during their 48 hour sprints. OBTW, to help the struggling team, they can get detailed design/implementation info for previous season robots and controls from several sources.

If their time-available-to-work is so tightly constrained that the struggling team can't accomplish what I described, regardless of how well they manage their project/time; then it would be a good idea (not a bad idea) to rethink the entire subject of competing in FRC. FTC, VRC, etc. are almost certainly better choices in that situation.

Bottom line: How to build an FRC robot+controls in under one week is well known. With preparation, getting a cookbook simple bot built in 6 weeks for one season has got to be do-able, unless the team in a situation that is simply incompatible with FRC. The next season they can adjust their goals.

Blake

The thing is, it takes teams a while to learn this lesson (if ever). No amount of external mentoring, help, etc... will cause the shift until the leadership of the team is ready for it.

Once bag day happens, there is no time machine to go back.

However, if after competing once (at a low level) they see teams they can copy some details from (and perhaps receive from advice from others), they can take their robot that isn't scoring points at all and have it contribute at their second event.

I'm absolutely certain that kids prefer scoring points to not scoring points, and ideally greater inspiration will follow.

Potentially that greater inspiration will lead to them wanting a better plan for season, to get done sooner, to realize those benefits etc... but we can't just state from our high horse that kids should just do it right the first time.

We don't need to get rid of the bag, we just need to allow unbagging windows for all teams all weeks.

[Oblarg]

Originally Posted by AdamHeard

We don't need to get rid of the bag, we just need to allow unbagging windows for all teams all weeks.

Why is this supposedly better than just getting rid of the bag?

This is still regressive, it's just marginally less-regressive than the current policy.

[AdamHeard]

Originally Posted by Oblarg

Why is this supposedly better than just getting rid of the bag?

This is still regressive, it's just marginally less-regressive than the current policy.

It's a more likely compromise.

[gblake]

deleted by gblake