Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Championship Event (http://www.chiefdelphi.com/forums/forumdisplay.php?f=12)
-   -   Qualifying for CMP in the future (http://www.chiefdelphi.com/forums/showthread.php?t=95064)

dodar 07-05-2011 16:37

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Lil' Lavery (Post 1059995)
I don't see how it reduces the number of teams going to CMP. If anything, teams going to 3+ regionals makes fewer teams qualify for Championship. For instance, 1114 earned four Championship invites this year by winning 3 regionals and a RCA. If they only attended two regionals, they would have opened the door for another team to qualify for CMP.

Actually if they had won those regionals and RCA or not, those spots would be given to another team(s). Its only difference between the teams winning/earning their spot or giving the spot to a team off the waitlist.

rwood359 07-05-2011 16:50

Re: Qualifying for CMP in the future
 
For those that want to eliminate the #3 alliance member take a look at 2010.
We (359) were 3rd or 4th seed in San Diego and chose 294 with our 2nd choice. Although 294 improved their robot and were a finalist in Los Angeles, this 3rd team position on our winning alliance was their entry to 2010 CMP.

They won Newton as the 1st seed captain and won CMP in a major upset.

Basel A 07-05-2011 16:55

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Lil' Lavery (Post 1059995)
I don't see how it reduces the number of teams going to CMP. If anything, teams going to 3+ regionals makes fewer teams qualify for Championship. For instance, 1114 earned four Championship invites this year by winning 3 regionals and a RCA. If they only attended two regionals, they would have opened the door for another team to qualify for CMP.

I think he means that the same number of teams go to less events each, it decreases the team-events, which theoretically reduces the number of regionals, though not by much. I think we should be trying to get more teams to more events, because that makes the CMP more competitive (more event experience). There just needs to be a way to limit the teams going to the CMP then.

Tetraman 07-05-2011 17:13

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Lil' Lavery (Post 1059974)
Heck, we've barely even scratched the surface of what all the other possibilities are.

Are we even looking for a complete change in the system? Because I'm sure what we want is a Championship more than a Festival.

If innovation to the system is what we want, then at the beginning of the year, lets split all the robotics teams into Divisions right off the bat. You play as either Archimedes/Curie/Newton/Galileo for the entire season, earning Qualifying points based on the results of each regional you go to. At the end of the year the highest 50 robotics teams in each division are invited to the championship to play on their division turf at the Championship, plus all the award specific winners.

The goal is to restrict the number of teams allowed into the Championship while letting the program grow without restriction - maintaining a Championship feel that imitates other High School sports while being its own.

PayneTrain 07-05-2011 18:08

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Lil' Lavery (Post 1059974)
This idea isn't very well thought through.

"All 50 or so teams" is a far underestimate of the number of teams that win those awards. There will be some "doubles" in terms of teams who win two of those awards and there will be many teams that don't make it to Championship. But with 48 regionals and 3 of those awards per event, we're talking well over 100 teams that make it to the Championship event based on those criteria.

Secondly, why are we lumping them in as #8 seeds? Basing it off of these Top 25 rankings, we see four hall of fame teams (16, 67, 111, 254), two regional Chairman's winners (1114, 2016), one Michigan district Chairman's Award winner (33), and four Engineering Inspiration winners (27, 118, 234, 2415). Based on whether or not you accept HoF and DCA winners into the Kamen division, you range from having six to fifteen of the Top 25 teams in one division (of five). You don't have to worry about not having enough talent here to compete against the other divisions, you have to worry about placing all your good teams into one division. Even ignoring the rest of the gripes about separating these teams from the field, you're throwing off the competitive balance by doing this.

I merely forgot to place the 1 in front of the 50.

Also, HOF teams would be dispersed among the other four divisions, and DCA winners don't even qualify for CMP.(33 won MSC, not MSCCA)

If you have five divisions, there would ideally be 5 teams of the unofficial top 25 in each of the 5 divisions. Having 6 in one isn't ideal, but it's obvious that FIRST doesn't split up the divisions in a finite way.

There is also an option for FRC to do a Round Robin on Einstein. Sure, it would take longer, but you also have a guarantee of getting the two best alliances in the final.

Dale 07-05-2011 19:02

Re: Qualifying for CMP in the future
 
Quote:

I don't see how it reduces the number of teams going to CMP. If anything, teams going to 3+ regionals makes fewer teams qualify for Championship. For instance, 1114 earned four Championship invites this year by winning 3 regionals and a RCA. If they only attended two regionals, they would have opened the door for another team to qualify for CMP.
Thanks for helping to clarify my thinking. Limiting teams to two regionals makes more room at the regionals (and hence allows for program growth) but doesn't help at the CMP as far as team numbers. It does mean more qualified teams attend rather than those who just wait list in. The limitation does help level the playing field but that's not what we're talking about here.

I'm not sure letting Rookie All-Stars attend automatically is doing them any huge favor. Sure they'll get to go to the event and see how the big boys play but getting beat up at CMP can be pretty demoralizing. That's one nice thing about a point system, you can give RAS teams a goodly number of points but it could be set up in such a way that they would also need to get to a certain level of competence on the field to qualify for CMP.

A point system could also help with the #2 pick problem being discussed in this thread. Maybe the team captain gets a bunch of points (enough to qualify), the first pick gets enough to normally qualify, and the second pick gets fewer. You'd rig it such that a well qualified 2nd pick would qualify but a box on wheels would normally not. I don't know what all would go into the qualification formula but something like OPR would be an interesting addition.

GaryVoshol 07-05-2011 21:38

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Dale (Post 1060018)
I'm not sure letting Rookie All-Stars attend automatically is doing them any huge favor. Sure they'll get to go to the event and see how the big boys play but getting beat up at CMP can be pretty demoralizing. That's one nice thing about a point system, you can give RAS teams a goodly number of points but it could be set up in such a way that they would also need to get to a certain level of competence on the field to qualify for CMP.

"Getting beat up at CMP" isn't an automatic result for rookies. Look at the rankings - yes, there are some rookies at the very bottom of the standings. But there are some others at the very top, and others scattered all in between.

Now I have no idea which of these teams got to CMP because of the RAS award and which got there by being part of a winning alliance. Your system of points to require a certain level of competence might illustrate the dividing line between those who ended up ranking high vs. low.

Lil' Lavery 07-05-2011 22:39

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by PayneTrain (Post 1060012)
I merely forgot to place the 1 in front of the 50.

Also, HOF teams would be dispersed among the other four divisions, and DCA winners don't even qualify for CMP.(33 won MSC, not MSCCA)

If you have five divisions, there would ideally be 5 teams of the unofficial top 25 in each of the 5 divisions. Having 6 in one isn't ideal, but it's obvious that FIRST doesn't split up the divisions in a finite way.

There is also an option for FRC to do a Round Robin on Einstein. Sure, it would take longer, but you also have a guarantee of getting the two best alliances in the final.

I'm aware of what award(s) 33 won, thus why I labelled them as a District CA winner. I'm also aware that winning a DCA doesn't auto qualify you for CMP, but it's obvious that you can win one and still attend Championship. If you're setting up a division specifically to honor the "big picture" awards, why are we not including teams that won these awards and are at the event? Just because they already had to be vetted again?

Same for HoF teams. Why aren't they being included?

And that 6 number is actually the lower end of the spectrum if you look at previous years. Looking at LF's 2010 Top 25 we see eight teams that won a RCA/MSCCA/EI, one that won a DCA, and four members of the HoF (341 wasn't yet a member). There wasn't any Top 25 list that I know of in 2009, but LF's 2008 list yields seven teams that won RCA/EI and five HoF members.

nikeairmancurry 08-05-2011 00:07

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by GaryVoshol (Post 1060029)
"Getting beat up at CMP" isn't an automatic result for rookies. Look at the rankings - yes, there are some rookies at the very bottom of the standings. But there are some others at the very top, and others scattered all in between.

Now I have no idea which of these teams got to CMP because of the RAS award and which got there by being part of a winning alliance. Your system of points to require a certain level of competence might illustrate the dividing line between those who ended up ranking high vs. low.

3539 was the 2nd seed captain this year on arch... I believe they won a rookie all star award at waterford, along with the district

Richard Wallace 08-05-2011 08:22

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by nikeairmancurry (Post 1060063)
3539 was the 2nd seed captain this year on arch... I believe they won a rookie all star award at waterford, along with the district

3539 also won RA at Troy, where they were a finalist and #2 AC. They were the only rookie team to make the MSC eliminations, and ended up with the eighth highest qualifying point total in Michigan. "Getting beat up at CMP" was definitely not in the plan for 3539.

Frenchie461 08-05-2011 09:43

Re: Qualifying for CMP in the future
 
What if the system was score based:

Winning an event is 10 points
Finalist at an event is 6
Making Elims is 2
Regional Chairmans is 15
Engineering Inspiration is 8
Rookie All-Star is 10
Woodie Flowers Finalist is 4 for the team
Judges is 4
All other design related awards are 2
All other non design awards are 1

top 300 highest scores go to world

Richard Wallace 08-05-2011 10:05

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Frenchie461 (Post 1060093)
...

top 300 highest scores go to world

Compare the point system you described above to the one used by FiM in 2011 (see page 6 at that link). Differences in the number of qualifying points that teams receive for their various achievements can be important. As an example, there were three district RA winners and one district EI winner that did not earn enough points to qualify for the Michigan State Championship this year.

Reaching agreement on how many CMP qualifying points to award for each achievement will be a challenge for FIRST.

Eric O 08-05-2011 10:26

Re: Qualifying for CMP in the future
 
Here is my proposal for Championship Qualification as well as some thoughts on regional play.

1) Championship Qualification needs to be a points based system. This is the only way any further manipulation of "how to qualify" to be discussed. Instead of discussing "Should this qualify a team?" the question becomes "How many points is this worth?". Every award in FRC is improtant, which is why they should all help qualify teams, some may just be worth more points.

2) Past years performance should be taken into account. Zondag's spreadsheet is the answer to how to calculate this, but once again, I think what to set % to is the question. My take, 30% would be the correct number.

3) I have always like the fact that teams can sign up for Championship if the they didn't go the year before. However, this is obviously not sustainable. Here comes the beaty of having a points system. What if every year you didn't go the championship you earned some points. Over X number of years you could add these points to those earned at compeitions and earn the ability to go. These points would see the same reduction value as mentioned above. Again I am sure there are vast opionions on how much not going to championship should be worth, but it's only an option if there is a point system.

4) I think that automatic entry should only be to the origional teams. I think HOF teams should get points on the order of 10 million when they win chairman's, which should enable them to go for quite a few years, but eventually that will run out per the past performance calculation mentioned in bullet 2.


I have further ideas for nation wide district play which I will post later.

-Eric

Nemo 08-05-2011 14:33

Re: Qualifying for CMP in the future
 
Lots of interesting discussion here. After doing more reading and seeing some good ideas, here are my latest thoughts.

I read through the Michigan points system again (thanks to Richard for the link). There is a lot to like about a points system if it is well designed. It takes more information into account, so it can more accurately sort out the most deserving teams. Our existing qualification system places all of the value on four specific achievements (Winner / RCA / EI / RAS) and no value on any of the other achievements (qualifier wins, advancing through elimnation rounds, all the other awards).

Cases when a deserving team might qualify in a points system, but would NOT qualify as a Winner/RCA/EI/RAS:

1) Team has one of the best robots around. Catches a couple of tough breaks and loses in the finals at two separate regionals.

2) Team wins several awards (non-RCA/EI/RAS), compiles a respectable record in qualifiers, and finishes as semifinalists in two regionals.

Cases when a potentially less deserving team would qualify over a more deserving team under the current system, but probably not under a points system:

1) Rookie team with a minimally functional robot and no award submission materials wins Rookie All-Star award by being the only rookie team at a regional.

2) Team wins a regional with a weak robot after being picked by a strong #1 seeded alliance with the last pick of the draft.

If I were designing a points system, I would want to do the following:


i) Use the Michigan points system for qualifier wins, alliance selections, and elimination rounds. A team that goes 10-0 in qualifiers, is 1st seed or 1st pick, wins the regional would earn 56 points.

ii) Give the Chairman's Award winners slightly more points than a top seeded regional winner. Perhaps 60 points. That way the top award winner gets more points than the top seeded regional winner.

iii) Give Engineering Inspiration a lot of points, but not as much as Chairman's (CA is the top award, after all). Perhaps 40-50 points.

iv) Make Rookie All-Star worth some number of points such that rookies could get in by winning that award AND achieving a modest level of competitive success. Perhaps 20-30 points.

v) Increase the value of other awards compared to the Michigan system. 2 or 5 points seems pretty small compared to 50+ for winning a regional. Perhaps 5 and 10 points would be appropriate. Or 10 and 15.

vi) Award points for Woodie Flower or Dean's List awards. I know they are individual awards, but I don't see any harm in awarding a few points for them.

vii) Give teams points if they didn't attend in the previous year or years. Great idea, Eric.

One advantage of a points system is that once it's in place, you don't have to modify the system further when the number of teams and events gets even bigger. You simply set the number of teams you want at the championship, and you invite the top n teams. Teams can join a wait list, and their wait list priority is determined by their point rankings. As the total number of teams goes up, the bar for points goes up.

One Event vs Two Events

A major weakness of a points system is that many teams only play in one regional each year. Things would be a lot more fair if all teams accumulated points in exactly two events, like they do in Michigan.

Art outlines a proposal for a three tiered system that would help this situation. The system would require reorganizing some regional events into a larger number of smaller, less expensive events. I am in favor of that. He also suggests not forcing teams into districts based on geography - also a good idea. He proposes using points to qualify for Tier 2 and using the existing system to qualify Tier 2 winners for champs (tier 3). That's the bit that I disagree with; our existing system for qualifying 6 teams per event is a really blunt instrument that doesn't necessarily qualify the best teams for the Championship.

There is no tidy district / regional model that will work perfectly for everybody. But I would really like to see us and FIRST sort that out, even if we have to settle for just getting a district system to work in more of the higher density regions. And I think part of that has to involve embracing a lower cost event model, which has been discussed elsewhere.

Other Invitations
I think it's fine to invite the previous year's Einstein winners.

On a philosophical level, I see no reason why original and sustaining teams should be auto-invited. This is a minor point since, as was stated above, this only results in a small handful of invitations.

I think Championship Chairman's Award teams should get invited for the next 5 or so years after they win the big award. Then invite them once every several years after that. This way Hall of Fame Row can stay a reasonable size and would be populated by the most recent 5 HOF teams plus a rotating group of teams that won more than 5 years ago.

EricH 08-05-2011 14:57

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Nemo (Post 1060131)

If I were designing a points system, I would want to do the following:


i) Use the Michigan points system for qualifier wins, alliance selections, and elimination rounds. A team that goes 10-0 in qualifiers, is 1st seed or 1st pick, wins the regional would earn 56 points.

ii) Give the Chairman's Award winners slightly more points than a top seeded regional winner. Perhaps 60 points. That way the top award winner gets more points than the top seeded regional winner.

iii) Give Engineering Inspiration a lot of points, but not as much as Chairman's (CA is the top award, after all). Perhaps 40-50 points.

iv) Make Rookie All-Star worth some number of points such that rookies could get in by winning that award AND achieving a modest level of competitive success. Perhaps 20-30 points.

v) Increase the value of other awards compared to the Michigan system. 2 or 5 points seems pretty small compared to 50+ for winning a regional. Perhaps 5 and 10 points would be appropriate. Or 10 and 15.

vi) Award points for Woodie Flower or Dean's List awards. I know they are individual awards, but I don't see any harm in awarding a few points for them.

vii) Give teams points if they didn't attend in the previous year or years. Great idea, Eric.

One advantage of a points system is that once it's in place, you don't have to modify the system further when the number of teams and events gets even bigger. You simply set the number of teams you want at the championship, and you invite the top n teams. Teams can join a wait list, and their wait list priority is determined by their point rankings. As the total number of teams goes up, the bar for points goes up.

Take a look at the Fantasy FIRST points system.

RCA: 42 points
EI: 36
Winner: 30
RAS: 25
Finalist, first seed, RI: 20
15 points for the technical awards
5 points for high score, highest rookie seed, judges award
2 points for win, non-robot award that isn't listed
WFFA 8 points, Dean's List 4 points
1 point for a tie
Loss does nothing to the score
Seeding from 2-16 has varying levels:
2-3, 12 points
4-8, 6 points
9-12, 3 points
13-16, 2 points

Then for the non-qual years, the team gets some fraction of the points they earned (say, 1/2 of the points) or a fixed number, whichever is higher, to be added to the next year's point total.

XaulZan11 08-05-2011 15:11

Re: Qualifying for CMP in the future
 
If you go with a point system, you need to find a way to account for event size and strength of regional. If not, I'd consider going to the 31 team Utah regional or one of the Minnesota ones instead of Midwest. Also, since teams get points for where they are selected, you'd see teams pick their friends in order to get them to the Championship.

dodar 08-05-2011 15:20

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by XaulZan11 (Post 1060133)
If you go with a point system, you need to find a way to account for event size and strength of regional. If not, I'd consider going to the 31 team Utah regional or one of the Minnesota ones instead of Midwest. Also, since teams get points for where they are selected, you'd see teams pick their friends in order to get them to the Championship.

So if you were the 1st seed and you had a choice between 254/148/1114/217/111 or whoever or a friend of yours, you would pick your friend? I highly doubt that. My friend's would understand why I didnt choose them if I had those other options because they would make the same choice.

Tom Ore 08-05-2011 15:25

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by XaulZan11 (Post 1060133)
If you go with a point system, you need to find a way to account for event size and strength of regional. If not, I'd consider going to the 31 team Utah regional or one of the Minnesota ones instead of Midwest. Also, since teams get points for where they are selected, you'd see teams pick their friends in order to get them to the Championship.

Sounds like the strength of regional issue would sort itself out over time. Teams would move to events that they felt gave them the best chance of success - thus leveling the regionals.

Jaxom 08-05-2011 15:32

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by dodar (Post 1059901)
If teams had to support an FLL or FTC team to go to CMP, how would you tell the kids on an FRC team that built that robot over the 6 weeks and poured thousands of hours into the team and made it to the semis at maybe 2 regionals that they cant get into the CMP because they dont have enough team resources to try to support an entirely different entity for their competition year? That would hurtful to those FLL/FTC kids who wouldnt get the full effect of FIRST because their FRC mentors are striving for their results for their team as it was.

What team resources do you think would be necessary to support an FLL team? This year we closely mentored 3 FLL teams that started as a direct result of work our FRC team had done by volunteering & helping run the FLL regional tournament. We also did a bunch of other stuff:
* Volunteer & help run summer camps that teach Mindstorms to kids.
* Volunteer at rumbles (think "practice tournament"). Both rumbles in our area are before FRC build season starts.
* Provide mentors during FLL meetings for those 3 teams, week after week. This interferes the most with FRC build season (which was a small percentage of actual FLL time; most of which was before FRC season), but we had enough FRC students helping that we had enough for each meeting. And it only took about 2 hours out of build time, two days a week, for each FRC student who mentored.
* Volunteer & help run the regional tournament. This happens after FRC ship date.
* Provide other, occasional, mentoring support for a bunch of other FLL teams that we weren't as directly involved with starting.

Does this cover most, or even a large percentage, of situations? Of course not. One big factor is that our regional FLL tournament is very late when compared to most. Every circumstance is different, but my point is that there's a lot you can do that directly impact FLL teams (we did other things for younger kids as well; this isn't an all-inclusive list) that don't really take a lot of money, if any. And not really that much time, either. But for any particular situation, there are ways to influence younger kids.

Having said all that, I don't think that supporting FLL or FTC teams should be a requirement for attending CMP, or for winning EI or RCA. But supporting those teams is a Good Thing To Do, and FRC teams that do support FLL & FTC (and VRC, and other STEM programs for younger kids, and....) are better off for it.

XaulZan11 08-05-2011 15:32

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by dodar (Post 1060135)
So if you were the 1st seed and you had a choice between 254/148/1114/217/111 or whoever or a friend of yours, you would pick your friend? I highly doubt that. My friend's would understand why I didnt choose them if I had those other options because they would make the same choice.

I wouldn't do it, but it happens. I've heard from several different people that it occurs in Michigan after they went with the district format.

Nemo 08-05-2011 15:39

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by EricH (Post 1060132)
Take a look at the Fantasy FIRST points system.

RCA: 42 points
EI: 36
Winner: 30
RAS: 25
Finalist, first seed, RI: 20
15 points for the technical awards
5 points for high score, highest rookie seed, judges award
2 points for win, non-robot award that isn't listed
WFFA 8 points, Dean's List 4 points
1 point for a tie
Loss does nothing to the score
Seeding from 2-16 has varying levels:
2-3, 12 points
4-8, 6 points
9-12, 3 points
13-16, 2 points

Then for the non-qual years, the team gets some fraction of the points they earned (say, 1/2 of the points) or a fixed number, whichever is higher, to be added to the next year's point total.

That's surprisingly similar to what I had in mind. Cool stuff. I am now reading about Fantasy FIRST. That's pretty interesting too - I had told my students earlier this year that we should do that, so I'm glad to see that it exists.

rsisk 08-05-2011 17:04

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Nemo (Post 1060131)
iv) Make Rookie All-Star worth some number of points such that rookies could get in by winning that award AND achieving a modest level of competitive success. Perhaps 20-30 points.

For both RAS and RCA, you should remove robot performance from any consideration in the ranking. Robot performance is no part of either of these awards.

Maybe a better way to solve the issues of winning RAS because the team is the only rookie at the competition is to weight the points earned for RAS by the number of rookies at the regional. I would make it an exponential rating so the point value rapidly increases as the number of teams go up.

The problem with a weighting system is you may have a truly deserving RAS winner as the only rookie at a regional. But there is not enough information captured to determine the relative value of the RAS award across various regionals.

Nemo 08-05-2011 17:36

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by rsisk (Post 1060154)
For both RAS and RCA, you should remove robot performance from any consideration in the ranking. Robot performance is no part of either of these awards.

Maybe a better way to solve the issues of winning RAS because the team is the only rookie at the competition is to weight the points earned for RAS by the number of rookies at the regional. I would make it an exponential rating so the point value rapidly increases as the number of teams go up.

The problem with a weighting system is you may have a truly deserving RAS winner as the only rookie at a regional. But there is not enough information captured to determine the relative value of the RAS award across various regionals.

I suppose you could make it worth a pile of points and then make it an optional award that the judges only give out if there is a rookie team present that has earned it.

More generally speaking, I am a fan of the rookie awards. They recognize rookie teams and the difficulty of being a rookie. I have issues, however, with the idea that we should send lots of rookies to the Championship to inspire them. I think they're better off finding a way to get to a second regional every year instead of going to the championship once as rookies. A second regional is going to do more to build up their confidence in their own abilities. It is also important to note that most regionals have at least a couple of really strong teams present, so it is not usually necessary to go to the Championship to gain exposure to those high caliber teams.

Kpchem 08-05-2011 21:19

Re: Qualifying for CMP in the future
 
On the matter of the inequality created in a points-based system when some teams attend multiple events while others attend only a single event, wouldn't it be possible to divide the number of points a team earns by the number of events they attend? This would level the playing field in that respect, and also reward teams for consistently high performance at multiple events.

nikeairmancurry 08-05-2011 22:19

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Kpchem (Post 1060221)
On the matter of the inequality created in a points-based system when some teams attend multiple events while others attend only a single event, wouldn't it be possible to divide the number of points a team earns by the number of events they attend? This would level the playing field in that respect, and also reward teams for consistently high performance at multiple events.

But this also kills continous improvement... Lets say a teams first events ends horrible and don't make elims, and in there second event are finalist, when you add up and divide the points, they have an average point total of maybe a quarter-finalist.. I don't see that as a solution

1986titans 08-05-2011 22:34

Re: Qualifying for CMP in the future
 
How about this:

Any team may go to however many events they feel like going to. However, only one event - most likely the one that produced the most points as any other wouldn't make sense - goes towards qualifying for Championships.

Siri 08-05-2011 23:27

Re: Qualifying for CMP in the future
 
Aren't you still skewing against teams that can only attend 1 event? Virtually all teams do better at their later regional(s), so comparing a 1-regional Semifinalist someone who doesn't get anything at their first 2 events but wins their last one isn't really valid it seems.

dodar 08-05-2011 23:39

Re: Qualifying for CMP in the future
 
Ok, Im sorry but people are just gonna have to realize that life ain't fair. If a team can attend more than 1 regional a year, would you want to penalize them against a team that cant?

Nemo 09-05-2011 00:06

Re: Qualifying for CMP in the future
 
I would want a points system to push everyone in the direction of attending two events, so I would not be in favor of taking average score or highest score. I really like the Michigan system of adding points from the first two events of the year.

It seems to me that attending only one event is an inefficient use of all of the resources that go into an FRC season. That second event costs marginally less than the first (because robot is already built and the hundreds of hours of time are already spent), and I think the second regional of the year is in some ways more valuable than the first. I really like it when students get a chance to take a working project and then tweak and refine it to take it to the next level. Too often they run out of time on projects (in school and in FRC and elsewhere) and have to settle for something that barely works, and that doesn't give them the same valuable experience or the same confidence boost as working on something until it's actually good.

Lil' Lavery 09-05-2011 00:20

Re: Qualifying for CMP in the future
 
I'm not convinced a point system is the answer. In fact, unless there's some pretty serious schedule changes, I'm convinced it would be a bad idea.

The points system works in Michigan because a vast majority of the teams who qualify are going to be within a few hours of the location. It's a lot easier to set up travel arrangements on short notice when they're within your state. Additionally the registration for MSC is $1000 cheaper than CMP registration.

On the Championship scale, too many teams are going to have to book airplanes, come up with large registration fees, and generally prepare for this event on too short of notice. Yes, many of the teams that qualify in the last couple weeks of regional competitions or come off the wait list already face this problem. But this expands it to all teams going to Championship. Because of the nature of the point system, everyone but the top few teams is going to be in jeopardy of falling out of qualified position until the last week of competition.

Forcing 300 teams to scramble to submit field trip forms, registration fees, transportation, hotel reservations, robot shipping, tool logistics, and everything else associated with coordinating going to an event in a couple week period is pretty ridiculous. And this is generally in the same time of year as standardized testing begins.



On a related matter, in order how to determine how to allocate points, you have to decide the goal of your championship.

Is it to reward the teams who were the most successful? Or is it to put the teams that will compete at the highest level (on the field and/or for awards) into a competition against eachother?
Those are two related, but fundamentally different goals. And there are plenty of other possibilities to consider as well.

Because depending on what your goal is, it will change the weighting of how you compare both awards, matches, and events.
If team A was consistently good across multiple regionals, but never spectacular, should they make it over team B who struggled at their first event but performed spectacularly at their second?
Should you be looking at average performance, peak performance, total performance, earliest performance, or most recent performance?

My personal take, should a point system come around, would be a hybrid of those. Simply put, I'd calculate points at each event, rather than per team. I'd then pick from three different rankings lists in an rotating fashion.
  • Team's average score from all events attended
  • Team's net score from all events attended
  • Team's peak score from all events attended
The first team off the the average list gets invited first. The highest remaining team on the net score list gets invited next. Then the highest remaining team on the peak score list. And it continues rotating.

The concern, naturally, will be that teams "buy" their way into Championship buy attending several events and getting high on the net score list, even if they perform at a mediocre level. My counter is, if a team can afford (both in terms of money and manpower) to attend 4+ events in a year and then the Championship, I think they're clearly doing something right that deserves recognition.


That being said, I still don't like a point system being the main qualifier with this scheduling set-up. Give a bigger space between regionals and championship and I'll change my mind.

JohnBoucher 09-05-2011 06:44

Re: Qualifying for CMP in the future
 
How are we running out of room at champs?
Is it pit space? I was not at St Louis, but was there no room for any additional pits? We can look toward larger venues or be more creative with pit locations.
Is it match count? Match count can vary from regional to regional. Lower the match count.

What are the actual limitations we are being warned about?

thefro526 09-05-2011 08:31

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by JohnBoucher (Post 1060319)
How are we running out of room at champs?
Is it pit space? I was not at St Louis, but was there no room for any additional pits? We can look toward larger venues or be more creative with pit locations.
Is it match count? Match count can vary from regional to regional. Lower the match count.

What are the actual limitations we are being warned about?

From what I could tell, if Two fields are left in the pits, then there isn't all that much room for additional pits. You might be able to squeeze enough pits in to get the number of teams up to an even 400, but anymore than that and you're really pushing it.

If the two pit fields were moved back into the dome, then you could probably put at least another 70+ pits without too much difficulty, but at that point the divisions might be too big. (IMO, any bigger than they are now and they're too big.)

Interestingly enough, it seems that FRC will reach the breaking point as far as the Championship is concerned sometime after next season - which means that it'll be time for FIRST to start looking for a new home for the Championship. I wonder if they'll look for a larger venue to house more teams and run more divisions, or if they'll restructure the Championship (Registration aspect of it) so that the event can remain the same size.

Taylor 09-05-2011 08:57

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Nemo (Post 1060294)
It seems to me that attending only one event is an inefficient use of all of the resources that go into an FRC season. That second event costs marginally less than the first (because robot is already built and the hundreds of hours of time are already spent), and I think the second regional of the year is in some ways more valuable than the first. I really like it when students get a chance to take a working project and then tweak and refine it to take it to the next level. Too often they run out of time on projects (in school and in FRC and elsewhere) and have to settle for something that barely works, and that doesn't give them the same valuable experience or the same confidence boost as working on something until it's actually good.

Speaking from the perspective of a small, money-tight team:

Attending a second event essentialy doubles our budget. To run one season costs us about $10-15k, adding a second regional bumps us to $25k+. Travel, room, and board costs a LOT, especially when the only option for travel is out-of-state and we have to charter our own bus. Couple this with living in a state where the Powers that Be have ... interesting ideas as to what the benefits and priorities of secondary education are, it can be a tough sell to remove some of the best students from school for a cumulative week.
When compared to international teams whose only option for multiple regionals lie in going outside the country, we've got it easy.

I agree that multiple regionals is a great experience, and twice in our seven years we have been able to do that, but it's not always viable. Expanding the sports analogy our Leaders have designed FRC around, I'd be willing to bet there are a bunch of Butlers or TCUs or Boise States out there that could make a bunch of noise at the highest levels, if only given the chance.

I think the idea of recognizing the points earned at a team's single-best, or two-best regionals, or an average of the season, would be a good solution - at least until a district-style model becomes widely available.

JohnBoucher 09-05-2011 09:22

Re: Qualifying for CMP in the future
 
Moving to a points based system will mean that some teams will never go to champs. Having the open slots available, allows those mid-level teams a chance to play against and be inspired by the top level teams that are always at champs. It will raise their game for many teams. Open slots may need to be on a three or four year allocation.

Ether 09-05-2011 09:26

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Taylor (Post 1060335)
Couple this with living in a state where the Powers that Be have ... interesting ideas as to what the benefits and priorities of secondary education are, it can be a tough sell to remove some of the best students from school for a cumulative week.

I don't get the connection between "state ... Powers that Be" and "tough sell to remove some of the best students from school for a cumulative week". Could you please elaborate on that? Has your school failed to meet state standards and been put under state control?



Taylor 09-05-2011 09:51

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Ether (Post 1060340)
I don't get the connection between "state ... Powers that Be" and "tough sell to remove some of the best students from school for a cumulative week". Could you please elaborate on that? Has your school failed to meet state standards and been put under state control?


- While taking care to avoid any deeply political statements -

Indiana has put a priority on Math and English/Language Arts, potentially at the expense of other subjects. It has slashed school funding, practically erasing many transportation budgets. It has placed a premium on 180 full days in school.
The combination of these three things, along with others, potentially make it difficult for teachers and students to get school board approval to miss more than one event's worth of school, even if the experience likely outweighs what that student/teacher may get from the classroom environment.

We are fortunate to have a fantastic administration that is supportive of STEM in general and FIRST & PLTW in particular. I've heard anectdotal stories from other schools that are not so fortunate - one that springs to mind is a team that essentially split itself in half - 1/2 the team attended one regional, 1/2 the team attended another; it's even more difficult for teams whose students don't belong to a single school/system (regional teams, 4-H teams, Scout teams, etc.).

IKE 09-05-2011 10:39

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by JohnBoucher (Post 1060339)
Moving to a points based system will mean that some teams will never go to champs. Having the open slots available, allows those mid-level teams a chance to play against and be inspired by the top level teams that are always at champs. It will raise their game for many teams. Open slots may need to be on a three or four year allocation.

John,
With 2100 teams, it would require 500 open slots/year for every team to have the opportunity to attend the Championship 1 time within every 4 years. If this was the model FIRST wanted, I would recommend changing the name to the FIRST World Invitational as it clearly would no longer be a championship.

The championship has been just barely functioning on an old model that essentially outdated itself 3-4 years ago, but statistically is just now coming to a head. 2012 will unfortunately be a very rough year for many areas of FRC and FIRST HQ. I think there will finally be an additional 1-2 "areas" that adopt the district model, and FIRST HQ will have to adopt some other method of qualifying teams for the Championship. This will upset a lot of people that aren't prepared for the change. And there will be an inordinate amount of people threatening to quit. Ironically their threats will be to join other competitive leagues that already have qualifying systems in place (gotta love that).

All in all, FRC is going to go through a drastic change that will likely take about 3 years to settle out. Personally, I think it will be for the better. 5-10 years from now, the young teams of tomorrow will get to hear stories of back in the day when anyone could buy into a championship, and you would have entire alliances that couldn't even execute the scoring objective.

By 2015, we will have around 3000 teams with 150 district comeptitions and 30 Regional Championships. The big question I have is whether or not the Regions will have theoretical borders or practicle borders.

dodar 10-05-2011 18:30

Re: Qualifying for CMP in the future
 
Or what if FIRST says that they will stop creating new regionals for a few years and just build up the regionals that they have now? Turn the regionals that we have now into multi-field super-regionals.

Mr. Van 10-05-2011 18:36

Re: Qualifying for CMP in the future
 
Some musings...

WHY HAVE THE CHAMPIONSHIP AT ALL?*

I believe that for a while, FRC has done the right thing in awarding "robot performance" and "team performance" nearly equally in guaranteeing championship slots. More importantly, there have always been spaces available for teams to register who had not attended championship in recent years. It has allowed FIRST to keep the focus where it says the focus should be. If FIRST has outgrown the ability to do this, perhaps it should separate the FLL, FTC and FRC events as Madison has suggested, or it should look at creating other opportunities to "change the culture".

What if FIRST took some of the resources spent on the championship and made each regional event have an elevated level of "show"? Invite vendors, bring back the team social, have a series of workshops/conferences, etc?

Over half of the FRC teams do not attend more than one regional. Only one in six go to the championships. Many times, however, the folks in New Hampshire seem to forget about these teams and place all of their focus on the big show. I suggest that instead of focusing on "how do we cut down the number of teams at the championship?" we should focus on "how can we make each regional feel like the championship?"

The alliance competition system of FRC almost guarantees that the winning alliance is NOT made of the three best performing robots. This alone casts doubt on the idea that winning a regional means you are "championship worthy" - as is evidenced by suggestions that 3rd picks not be eligible to attend. The fact that teams compete with alliances throughout qualification rounds and the not-quite randomness of the schedule makes the entire process suspect.

Besides... from what I understand, the real hard-core competition is IRI...

-Mr. Van
Coach, Robodox



* I do know the answer to this - national sponsorship - I know... but still...

Ankit S. 10-05-2011 18:43

Re: Qualifying for CMP in the future
 
Please correct me if needed.

I believe that the number of teams attending Championships should be increased. This can be done by adding a new division. Einstein would need to change to compensate for this. Instead of having elims immediately, have a round robin on Einstein that eliminates the "weakest" alliance, and also allows for a less random 1v4 and 2v3 rather than AvC and NvG.

I have never attended Championships before, so I am not sure about the space issue, but adding an extra field is a plausible solution, although the expanding pit area will be something that will need to be dealt with.

Thanks for hearing me through.

Jeanne Boyarsky 10-05-2011 23:22

Re: Qualifying for CMP in the future
 
Quote:

Originally Posted by Mr. Van (Post 1060720)
If FIRST has outgrown the ability to do this, perhaps it should separate the FLL, FTC and FRC events as Madison has suggested, or it should look at creating other opportunities to "change the culture".

This reminded me of something I noted at CMP. In New York City, we are lucky to have a "super regional" with FLL, FTC and FRC all competing. It's great because the FLL kids can see the "big bots." At CMP, a couple lego kids came by and were amazed at the "big bots." They don't have FRC at all in their city so it was the first time they saw it. I think it's really cool having all three together to inspire the younger kids.

I'd also like to mention that Einstein wasn't idle - it was used for FLL earlier in the competition.


All times are GMT -5. The time now is 13:29.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi