Suggestion for a new overall approach

[Alan Anderson]

Originally Posted by Pault

I absolutely agree that helping other teams is a huge part of FIRST. I just don't agree that including coopertition as part of the game incentives teams to do so.

How does it not? If you think having to rely on your alliance partners in order to win a match isn't a good incentive to help them do well, I don't know what you'd suggest as a better one.

[Citrus Dad]

Originally Posted by Pault

But I still don't see how coopertition among an alliance makes a team more likely to help other teams. The good teams out there always have helped teams, and always will, not at all because they are afraid of getting bad alliance partners but because it is making progress towards the ultimate goal of FIRST. I can't imagine a team saying "I don't want to be with bad alliance partners, so I am going to help all the teams around me to be better alliance partners... Even though it is more likely that they will be competing against me." I imagine them saying "I am going to help the teams around me because it will help to inspire them and their community to take up STEM carreers, as well as increase the competitiveness of FIRST in order to better bring about the culture change which we all desire."

As an economist, I'll say that there is a very big difference between what people SHOULD do, and what they ACTUALLY do. (I think I said this earlier as well.) Right now teams are not ACTUALLY reaching out as well as they could. (And our team was one of them until recently because we were spending out efforts building up internal resources.) I think the GDC came up with this game because they made the same observation and wanted to increase interteam cooperation. I agree with the GDC if that's there objective. I'm making suggestions to further the GDC's objectives.

The status quo isn't working--come up with a new suggestion.

[BeardyMentor]

Originally Posted by Citrus Dad

As an economist, I'll say that there is a very big difference between what people SHOULD do, and what they ACTUALLY do. (I think I said this earlier as well.) Right now teams are not ACTUALLY reaching out as well as they could. (And our team was one of them until recently because we were spending out efforts building up internal resources.) I think the GDC came up with this game because they made the same observation and wanted to increase interteam cooperation. I agree with the GDC if that's there objective. I'm making suggestions to further the GDC's objectives.

The status quo isn't working--come up with a new suggestion.

The status quo has only been one competition season. Give the current system time to sink in. If there is 2 or 3 years in a row that we see strong intra-alliance cooperation required, we will see exactly what you are describing.

[Citrus Dad]

Originally Posted by BeardyMentor

The status quo has only been one competition season. Give the current system time to sink in. If there is 2 or 3 years in a row that we see strong intra-alliance cooperation required, we will see exactly what you are describing.

Why not give a signal this coming season? Why wait until next January? Why wait 2 or 3 years for something that can happen in 6 months?

The status quo I was referencing was the history of FRC in which more experienced teams generally have not shared with and supported newer teams.
(Yes, there are exceptions, but that's not the general case.) The GDC went part way to fix the status quo. I'm suggesting going further.

[BeardyMentor]

I just don't think it is necessary to send out game information before kick-off. The assumption should be every year that robot interaction will be required. Sending out a 'Robot interaction required' signal on certain years will only really help on those years. What happens with rookie team who starts on an 'interaction' year vs one that doesn't?

Your goal is one that I think FIRST shares, but giving game information before kickoff is something that to my knowledge has never been done and is likely not the best way to change how teams work together outside of competition on a sustained time scale.

Maybe the focus should be on encouraging the less able teams to ask for help when they need it. My team has never been refused a reasonable request for assistance. Nor have we ever refused to help anyone who has asked us for help we were capable of providing.

[Jon Stratis]

Originally Posted by Citrus Dad

1) FIRST needs to announce in September, long before Kickoff, that it is planning a game that requires robot interaction with bonus points. This gives all teams a signal that they must rely on their alliance members much more than in the past. The GDC need not reveal anything more so teams are not going to get a jump on design.

2) FIRST needs to provide a list of newest teams (including prospects) to other teams in the region so that the older teams know who they need to contact for step 3). FIRST should try to finalize this list by the end of November.

3) The more experienced teams should start in September to visit the newest teams, both this year's and last year's rookies to start, to explain how they design for different game strategies, including focusing on specific, manageable tasks at the outset, and to train these teams in building robust, reliable robots. And guess what? This program both enhances the FIRST experience AND achieves some of the most important educational objectives of FIRST. It also builds community by bringing together the best teams (which aren't always viewed in the best light) with the newest teams.

FIRST could take this a step further by assigning the top teams a number of new teams to mentor, e.g., 3-5, and start the assignments based on world ranked order. Participating could become a requirement for FIRST membership. Many top teams do this, but it would formalize the process and ease finding the newest teams. FIRST could even create the ability to have "superalliances" that some how play into regional rankings and world championships qualifications.

1. Simply put, I don't see this happening. The best we can hope for is a couple of years in a row of this style game, such that coopertition becomes expected by teams, and not something they complain about.

2. You can already do this. As soon as a team registers for a regional, they're listed on the regional webpage. You can also search for teams in your area, and list them by team number to find both the oldest and the youngest.

3. Many teams already do this. My team has mentored several FRC teams through their first season, we regularly have teams over to our build space or go to theirs to help them improve their programs, and we present every year at a local rookie-centric 1-day conference (I believe we had 6 presentations this year, on a wide variety of topics).

You won't see FIRST assigning specific teams other teams to mentor. That sort of mentorship is part of what Chairman's is all about, and it's expected that the best teams out there will be actively going out and doing it. Further, FIRST has no way of ranking teams in order to determine which ones would provide good mentorship for a rookie team and which ones wouldn't.

FIRST has a steep learning curve. A very steep one. Until we start getting rookies signed up in September so they have 3+ months to train and learn, they'll always struggle a bit. You simply can't get all that knowledge into a rookie team that doesn't sign up until December.

[bduddy]

Originally Posted by Pault

But what incentives would this actually give teams? It is J̶u̶s̶t̶ a̶s̶ more likely that the team you are helping is going to be against you as it is that they are going to be on your alliance. If I magically made everybody in New England 10x better except for my own team, we would have a much harder time, regardless of the game.

It decreases your variance. Let's say that you have a very good robot, better than all of the other teams at your regional. If all of the other teams were at a roughly similar level, you would expect to win every match. However, if some of the other teams were OK and some were very bad, you could lose some matches due to the randomness of the alliance selection algorithm. Of course, this affects everyone equally, but if you have a good robot you want to remove this randomness.

[Citrus Dad]

Originally Posted by Jon Stratis

1. Simply put, I don't see this happening. The best we can hope for is a couple of years in a row of this style game, such that coopertition becomes expected by teams, and not something they complain about.

2. You can already do this. As soon as a team registers for a regional, they're listed on the regional webpage. You can also search for teams in your area, and list them by team number to find both the oldest and the youngest.

3. Many teams already do this. My team has mentored several FRC teams through their first season, we regularly have teams over to our build space or go to theirs to help them improve their programs, and we present every year at a local rookie-centric 1-day conference (I believe we had 6 presentations this year, on a wide variety of topics).

You won't see FIRST assigning specific teams other teams to mentor. That sort of mentorship is part of what Chairman's is all about, and it's expected that the best teams out there will be actively going out and doing it. Further, FIRST has no way of ranking teams in order to determine which ones would provide good mentorship for a rookie team and which ones wouldn't.

FIRST has a steep learning curve. A very steep one. Until we start getting rookies signed up in September so they have 3+ months to train and learn, they'll always struggle a bit. You simply can't get all that knowledge into a rookie team that doesn't sign up until December.

You're making the same mistake that many people make--assume that individuals will somehow decide to do the right thing voluntarily including taking the effort to gather information and to then process it. As a professional economist, I see many studies that show that this is a false assumption. To make a policy work effectively, you need to provide people with information and give them default choices that drive them toward the preferred choice. In fact, taking an opposite tack can drive them toward the opposite decision (and is just as much external decision making as the first case.)

So this includes actively sending a list of new teams, not waiting on existing teams to take the added effort of looking up what new teams are registered. In addition, there's a lag between when a team is actually formed and going and when it shows up on registration--a lag that can be critical. Also, the regional director can inform existing teams about informal new teams to help accelerate the process.

While some teams do reach out and they are rewarded through Chairman's, relying on volunteer charity is not a good policy. As I mentioned earlier, Hal Varian, Chief CIO at Google, wrote a paper in 1986 showing this is not a socially optimal or preferred policy and such support is underprovided by relying solely on voluntary action. The point is that "many" teams is still not "enough" teams. We know that many more teams could be helping, and giving them stronger incentives to do so can boost this. Relying solely on Chairman's isn't enough due to the low probability of winning for most teams and great effort for submission.

BTW, a ranking system isn't too difficult to create. Use regional winners and finalist plus award winners.

[Citrus Dad]

Originally Posted by bduddy

It decreases your variance. Let's say that you have a very good robot, better than all of the other teams at your regional. If all of the other teams were at a roughly similar level, you would expect to win every match. However, if some of the other teams were OK and some were very bad, you could lose some matches due to the randomness of the alliance selection algorithm. Of course, this affects everyone equally, but if you have a good robot you want to remove this randomness.

I agree. Having a poor robot on your alliance hurts more than having a poor one on the opposing alliance helps. In addition at a regional you can have a robot on your alliance several times, depending on regional size.

[Citrus Dad]

Originally Posted by BeardyMentor

I just don't think it is necessary to send out game information before kick-off. The assumption should be every year that robot interaction will be required. Sending out a 'Robot interaction required' signal on certain years will only really help on those years. What happens with rookie team who starts on an 'interaction' year vs one that doesn't?

Your goal is one that I think FIRST shares, but giving game information before kickoff is something that to my knowledge has never been done and is likely not the best way to change how teams work together outside of competition on a sustained time scale.

Maybe the focus should be on encouraging the less able teams to ask for help when they need it. My team has never been refused a reasonable request for assistance. Nor have we ever refused to help anyone who has asked us for help we were capable of providing.

Again, you're relying on individual teams to take actions, and particularly inexperienced teams that probably have no idea which teams are good to ask for help.

And yes, I'm saying the FIRST should make a simple change in its game announcement. To be honest, I don't think that it will affect the surprise element of the new challenge. FIRST has released game hints before (although not always accurate). This would be more explicit.

You don't specify why this isn't the best way to accomplish this goal, so I'm assuming you don't really have a better suggestion beyond relying on what is already occurring and in my opinion is not working as well as it could. I think the complaints about this year's game back up my position.

[Jon Stratis]

Originally Posted by Citrus Dad

You're making the same mistake that many people make--assume that individuals will somehow decide to do the right thing voluntarily including taking the effort to gather information and to then process it. As a professional economist, I see many studies that show that this is a false assumption. To make a policy work effectively, you need to provide people with information and give them default choices that drive them toward the preferred choice. In fact, taking an opposite tack can drive them toward the opposite decision (and is just as much external decision making as the first case.)

So this includes actively sending a list of new teams, not waiting on existing teams to take the added effort of looking up what new teams are registered. In addition, there's a lag between when a team is actually formed and going and when it shows up on registration--a lag that can be critical. Also, the regional director can inform existing teams about informal new teams to help accelerate the process.

While some teams do reach out and they are rewarded through Chairman's, relying on volunteer charity is not a good policy. As I mentioned earlier, Hal Varian, Chief CIO at Google, wrote a paper in 1986 showing this is not a socially optimal or preferred policy and such support is underprovided by relying solely on voluntary action. The point is that "many" teams is still not "enough" teams. We know that many more teams could be helping, and giving them stronger incentives to do so can boost this. Relying solely on Chairman's isn't enough due to the low probability of winning for most teams and great effort for submission.

BTW, a ranking system isn't too difficult to create. Use regional winners and finalist plus award winners.

You can't legislate culture. FIRST is all about changing the culture, inspiring Gracious Professionalism, and provides incentive for teams to do the right thing through awards. There's nothing wrong with providing incentive, but assigning teams to be mentors is not the right way for FIRST to go about it. FIRST provides all the information it can in a way that is easily searchable, and I know a lot of local area teams that help rookies.

As for your ranking system... I know regional winners I would not recommend mentor other teams. They may end up with a robot good enough to be on the winning alliance, but that doesn't mean they have a program that can adequately support and train a rookie team. I also know many teams that win awards but have robot issues every single year - they would not be a good match for helping a rookie team get something built. I also know teams that fall into both these categories, yet would not be willing to do more than a token e-mail to a rookie team if FIRST required it.

Just this past year MN set up a system of "hubs", designed to help rookies, organize team assistance, and otherwise increase the level of competition across the state. The system was set up via an opt-in approach, where a number of teams were invited to an initial meeting and signed up to be "hub leaders". Those teams contacted other teams in their area to start setting up hub activities and get the ball rolling, and it's worked out great - I know I've attended mentor gatherings for two different hubs and been able to give advise on specific issues teams were facing all season. The hub my team is in had several events, including group strategy sessions, group build sessions, and group Chairman's preparation. All on a volunteer basis with no incentive for the teams involves beyond what FIRST already provides with awards at the regionals.

[Citrus Dad]

Originally Posted by Jon Stratis

You can't legislate culture. FIRST is all about changing the culture, inspiring Gracious Professionalism, and provides incentive for teams to do the right thing through awards. There's nothing wrong with providing incentive, but assigning teams to be mentors is not the right way for FIRST to go about it. FIRST provides all the information it can in a way that is easily searchable, and I know a lot of local area teams that help rookies.

As for your ranking system... I know regional winners I would not recommend mentor other teams. They may end up with a robot good enough to be on the winning alliance, but that doesn't mean they have a program that can adequately support and train a rookie team. I also know many teams that win awards but have robot issues every single year - they would not be a good match for helping a rookie team get something built. I also know teams that fall into both these categories, yet would not be willing to do more than a token e-mail to a rookie team if FIRST required it.

Just this past year MN set up a system of "hubs", designed to help rookies, organize team assistance, and otherwise increase the level of competition across the state. The system was set up via an opt-in approach, where a number of teams were invited to an initial meeting and signed up to be "hub leaders". Those teams contacted other teams in their area to start setting up hub activities and get the ball rolling, and it's worked out great - I know I've attended mentor gatherings for two different hubs and been able to give advise on specific issues teams were facing all season. The hub my team is in had several events, including group strategy sessions, group build sessions, and group Chairman's preparation. All on a volunteer basis with no incentive for the teams involves beyond what FIRST already provides with awards at the regionals.

Not true--you can legislate cultural changes. That's what the Civil Rights Act was about 50 years ago, and it was fairly successful. I can name a number of other Awards are not strong enough incentives alone. They haven't worked well enough, and the GDC appears to agree by changing the game design so much this year. The information provided requires SEARCHING--that's an important transaction cost that impedes this activity. (Working on this type of problem is one of my professional focuses.)

You may know lots of teams that help rookie teams, but that is apparently not enough. You haven't suggested an approach that INCREASES that level of assistance--that's what the GDC is aiming for. Instead you're defending a status quo that is not functioning to its best level. You need to prove your basis premise--that the current system is just fine and leaving less experienced teams to swim or sink is the best policy. Instead right now we have a high attrition rate as those inexperienced teams find it too difficult to manage in the current FRC environment. Why not take important steps to help them further?

Yes, there are teams that don't build great robots that win regionals and awards, but to be honest, its much more probable that they will have good robots and programs than teams that haven't won any of those honors--you have to admit that. In your world, rookies are left with NO information about how to screen which teams are potentially good mentors.

That's great that MN set up hubs. (We locally have started a similar program.) Who initiated that process? Was it the state organization? Why leave this to MN? Why not have all states implement such programs? And what proportion of rookie and 2nd year teams participated? Why wait for this to spread slowly through voluntary actions rather than starting a larger program across the board? I expect that participation would increase even more with a more active recruiting process on both sides. Experienced teams would have an even strong incentive to host such programs.

Why are you so resistant to the idea that FIRST should actively reach out to help new and struggling teams?

[Jon Stratis]

There's just no way that FIRST can adequately pair up teams for mentorship. There's no way for someone sitting at HQ to understand the behaviors and desires of a team 1000 miles away in order to know if they would be a good team to select to mentor another team. It takes more than winning a regional or a few awards to be a good mentor.

Here in MN, the status quo is just fine. In the past 9 years (since FIRST got started in the state with 1816), we've had less than 10 teams drop out of FRC (most of them this past year) - out of almost 190. One of those I know merged with another team, and most of the rest switched over to FTC - they were victims of being pushed into FRC when the school simply wasn't ready for that level of commitment, and FTC is a much better fit for those teams. That's adding 20 teams per year, on average, and losing maybe 1.

If we can make it work here, why does FIRST have to step in and be heavy-handed about controlling it? Instead of looking for FIRST to step in and solve the problem for you, why not get things set up in your area to solve the issues you see directly? Let other areas come up with solutions that best work for them.

The people at HQ are already stretched with everything they do, and the last thing we need is to increase costs for everyone by hiring more people to organize something that can be done perfectly well by some volunteers and teams themselves.

[Citrus Dad]

Originally Posted by Jon Stratis

There's just no way that FIRST can adequately pair up teams for mentorship. There's no way for someone sitting at HQ to understand the behaviors and desires of a team 1000 miles away in order to know if they would be a good team to select to mentor another team. It takes more than winning a regional or a few awards to be a good mentor.

Here in MN, the status quo is just fine. In the past 9 years (since FIRST got started in the state with 1816), we've had less than 10 teams drop out of FRC (most of them this past year) - out of almost 190. One of those I know merged with another team, and most of the rest switched over to FTC - they were victims of being pushed into FRC when the school simply wasn't ready for that level of commitment, and FTC is a much better fit for those teams. That's adding 20 teams per year, on average, and losing maybe 1.

If we can make it work here, why does FIRST have to step in and be heavy-handed about controlling it? Instead of looking for FIRST to step in and solve the problem for you, why not get things set up in your area to solve the issues you see directly? Let other areas come up with solutions that best work for them.

The people at HQ are already stretched with everything they do, and the last thing we need is to increase costs for everyone by hiring more people to organize something that can be done perfectly well by some volunteers and teams themselves.

It may be working in MN, but it's not working well elsewhere. And this is a national and worldwide program, not just MN. And the GDC appears to agree with that point by so radically changing the game this year to create these incentives. So I suggest being so parochial and looking at the needs of the broader community. (And I'll make the point that MN is much different than the rest of the U.S. in community involvement and support, in a good way. But don't expect the rest of us to be like you.) The problem is bigger than what can be fixed at the grassroots with the hope that it will spread. Instead it requires a top down approach that creates stronger incentives to create the type of program that MN has. I work in public policy, and I know that trying to rely on self started grassroots programs to change broader community issues is very slow going, on the scale of decades.

I'm still curious how you propose that rookie teams identify competent local teams as useful mentors. Running a workshop isn't any better than relying on honors as a criteria. At least here in California teams that are alliance captains or first picks and are Chairman's, EI or other engineering award winners are generally quite good at what they do.

The fact is that what I've proposed will not be a large burden on the FIRST staff. Here's what I proposed: 1) Tell us whether the game will require cooperative robot play (and nothing more) in August. 2) Provide to experience teams (perhaps those that meet the honors won criteria) the names of new interested FRC teams being formed PRIOR to formal registration as those become available. 3) Ask the experienced teams explicitly to approach the new teams to provide guidance and assistance.

[Citrus Dad]

For those of you who don't think that the Aerial Assist format caused even the power teams to step up their support of other teams this year, listen to EJ from 254 discussing their in-pit strategy on GameSense, starting at about 0:30 (and further discussion of 1678's in-pit scouting). This should end the discussion about whether the cooperative play mode changed how teams interacted.

https://www.youtube.com/watch?v=eQdiGHYcrdQ