[FRC Blog] Frank Answers Friday: Championship Slots

[orangemoore]

Posted on the FRC Blog, 8/28/15: http://www.usfirst.org/roboticsprogr...pionship-Slots

Frank Answers Friday: Championship Slots

Blog Date: Friday, August 28, 2015 - 16:37


Today’s good question comes from Jon Stratis, a mentor on FRC Team 2177, The Robettes, from Mendota, Minnesota, USA:

Question:

Frank -

With the way championship slots are allocated to District teams versus regionals, we're seeing areas that are still doing regionals vastly under represented at champs when compared to District areas. For example, Minnesota in 2015 had 192 teams, 6.63% of all FRC teams, but only sent 24 teams to champs - 4% of teams at champs. That's 16 teams less than the state would have sent if it was a district! What plans does FIRST have for future District expansion or to remedy this situation and provide regional areas with more equal representative attendance at champs?

Thanks,

~Jon Stratis

Answer:

Hi Jon. Thanks for the question.

One of the advantages of the District system is the ease with which we can proportionally allocate Championship slots to teams. The standard points model, used for ranking, makes this a breeze, and if a District team that earns its way to Championship on points declines the invitation, we’ve got the next ranked team that has not yet received an invitation to make the offer to, and so on down the line, until all allocated slots are full. This is why we call these slots guaranteed.

While the District model has this and even more impressive advantages going for it, it has its downsides also, a major one being the amount of work and planning it takes at the local level to stand one up and keep it running. While FIRST HQ and the local organizations work together to get a District up and going, by far the greatest amount of work is done at the local level. So, while there are a few good candidates for 2017, and we’re certainly encouraging District transition for those areas that are ready, I don’t feel comfortable listing who those candidates are, as we are still uncertain they will become a reality at this point for 2017.

For Regionals, we will be staying with awarding slots based on performance at events themselves. I understand that this does not proportionally award Championship slots based on geographic representation. However, I think to award based on geographic representation under the Regional system we would need to take some type of pseudo-District ranking approach, which I believe is not a good fit for a competition model in which teams may be attending one, two, or more events, and there is no competition level between Regionals and the FIRST Championship. I envision such a system as being complex and getting us away from the immediate satisfaction and excitement teams can get at Regionals of learning they’ve got their golden ticket. However, I’m not closing the door on this forever. I’m willing to take a look at any specific proposal anyone has on this, or any other concern.

Thanks again!

Frank

Frank Answers Fridays is a weekly-ish blog feature where I’ll be answering ‘good questions’ from the FRC community. You can e-mail your questions to goodfrcquestion@usfirst.org. Please include your name, team number and where you’re from, which will be shared, if selected.

[Jacob Bendicksen]

Not a terribly satisfying answer, but a reasonable one.

Does anyone have a proposal that could be debated/sent to Frank?

[Greg Needel]

The only solution to this that I can think of is to offer waitlist teams based on geography and performance in the season, instead of first cone first serve or time since last championship.

There would need to be some sort of ranking system for performance that normalizes the performance within a region ( maybe a modified version of the district point system). There would definitely be some issues as the algorithm would need to take into account quite a few different things. For instince there would always be a debate about which regions get priority as it would be nearly impossible to perfectly balance the regions. It would also be quite a bit of work for FIRST to inform teams and make these decisions, which for the most part only affect a few teams.

Another method of implementation might be to modify the wildcard system so that it continues to pass down until it's used. This would be problematic when you get to the semi finalists as then you would have to bring scores and rank( pre elims) into the picture and this get equally as muddy. For example should the 2nd round pick on the #3 alliance who was a semifinalist advance before the #5 alliance captain that got put out in the quarter finals? Wildcard right now is a good thing that does help cover some inequalities at the region but it also does influence alliance selection as teams who thin they can't win will consider which situation best earns them a wildcard spot, I am not sure what an expanded version of this system would do.

Though question that would take some serious thought to figure out a reasonable solution, if in fact it is worth solving at all because of the small number of teams this effects each year.

[MikLast]

Originally Posted by Jacob Bendicksen

Not a terribly satisfying answer, but a reasonable one.

Does anyone have a proposal that could be debated/sent to Frank?

His thoughts on the PNW's new funding model?

[Brian Maher]

My proposed solution is three-pronged:

• Replace Finalist alliance position with district points for Wildcard distribution. The Stand Points Model was created to be an ideal assessment of team performance, so it seems perfect for this. This creates a system that can scale to include more than three teams, including those not on the Finalist alliance.
• If a team declines a CMP spot, pass a wildcard down the district points rankings until all spots are filled.
• Distribute extra Wildcard slots, which I call Bonus Slots, to large regionals. It seems silly to me that the 31-team Waterloo Regional and the 66-team Palmetto Regional qualify the same number of teams for CMP. This can be done by awarding the remaining "waitlist" slots to regionals based on the number of teams attending minus a certain value, which I used 40 for. While this doesn't create a strictly geographic distribution, it would be a much better approximation than the current system.

I put together a Google Spreadsheet demonstrating my proposal.

[Mr V]

Originally Posted by Greg Needel

The only solution to this that I can think of is to offer waitlist teams based on geography and performance in the season, instead of first cone first serve or time since last championship.

The current method of allocating the number of extra spots at CMP is best described as a lottery. It is not first come first served. If you do however choose to put your hat in the ring the number of lottery tickets you get is based on the number of years since your last CMP attendance. Overall I think it does a pretty good job of allocating spots and achieving the goal of teams making it to CMP on a periodic basis.

Originally Posted by BMSOTM

Distribute extra Wildcard slots, which I call Bonus Slots, to large regionals. It seems silly to me that the 31-team Waterloo regional and the 66-team Palmetto regional qualify the same number of teams for CMP. This can be done by awarding the remaining "waitlist" slots to regionals based on the number of teams attending minus a certain value, which I used 40 for.

I think that this represents the biggest issue. The spots a district has is proportional to their size but it is fixed across a large range of Regional sizes. As nice as it would be I just don't see a good way to have all Regionals have a similar number of teams. So a way to give more spots to larger Regionals is not a bad idea.

Another issue is the fact that a district gets a number of spots proportional to the size of the district and teams in the district are still eligible to enter the lottery. So in at least a couple of instances last season "extra" teams from a district were awarded spots at CMP. Lottery spots were also awarded before the end of the regular season. Now I understand that it was done to ensure that declines were available to be passed down and that teams had more time to make arrangements but it did have some interesting effects.

It made some Districts send more teams than their allotment in addition to those that were pre-qualified. (PNW had two teams that won the lottery, and they earned a spot pushing their earned spot down to the next teams in the rankings) It also may have resulted in more wildcard spots being generated at week 6 Regionals, though I do not know that for a fact. IF that did happen then teams attending week 6 events had an even greater chance at earning a wild card spot than they already had.

One thing that I think might be a good option is to reserve a number of lottery spots and remove them from the available spots when calculating the number of spots for a district. FIRST does want teams to be able to attend CMP occasionally even if they are not able to meet the performance standards necessary. (I know that some people don't like that idea but FIRST has determined that best suits their goals, so it is what it is, and I'm not trying to restart that debate.)


So how about something along these lines, now the numbers are just pulled out of a hat because I really don't know the real numbers for the upcoming year with all the new districts which are causing some Regionals to go away and a couple of new Regionals that are supposed to happen.

Say there are 250 spots that could potentially be earned at Regionals and teams are 50% Regional and 50% District. So that means that there would be 250 spots to divide by the districts based on their relative size. That would mean there would be 500 spots that could be earned subtract that and the number of prequalified teams which I'll call 25 for the sake of this discussion and you have 75 spots reserved for the lottery.

That would mean that the proportion of District teams vs Regional teams attending CMP would be relative to the respective proportion of all teams, less the fudge factor caused by pre-qualified teams and the inevitable randomness of the lottery winners. Of course that would still not make it so individual states got proportional representation due to the different size of Regionals and the fact that sometimes teams earn spots outside of their home state. Of course the fact is that not all states have events and for some locations an out of state event is more convenient as a team's first or second play.

The ultimate solution of course is for all teams to participate in the District system but the District system is more work than the Regional system and some areas are a long way from having the number of teams to make the District system viable let alone having the ability to put the infrastructure in place to make it happen.

[Ginger Power]

One thing I'd really like to see happen is a way for the best individual team at a regional to get to champs. The way things currently stand, districts do a much better job of getting the top robots to champs than do regionals. Whether that happens through an existing slot like Engineering Inspiration, or some other method I don't care.

[Doug Frisk]

Originally Posted by Ginger Power

One thing I'd really like to see happen is a way for the best individual team at a regional to get to champs. The way things currently stand, districts do a much better job of getting the top robots to champs than do regionals. Whether that happens through an existing slot like Engineering Inspiration, or some other method I don't care.

Best robot or best team, and how do you define either of those other than as the winner of the regional or Chairman's/EI?

[ATannahill]

I find it interesting that according to the survey about two championships the most important element of the championship is 'Seeing and competing with the teams with the best robots in FRC' yet we are currently discussing how to distribute the championship slots fairly among district regions and regional regions?

Do people feel that we can have both of these? Can we properly represent all regions while still having the experience of seeing and facing the best robots in FRC?

Please don't turn this into a discussion about two championships. That horse has been beaten twice already.

[EricH]

Originally Posted by rtfgnow

Do people feel that we can have both of these? Can we properly represent all regions while still having the experience of seeing and facing the best robots in FRC?

I think so. The big issue is getting the "representational adjustment" slots TO the best robots in FRC in any given year. For example, getting MN's adjustment slots to the best MN teams.

Of course, solving that issue is a fairly "interesting" "little" problem...

[Richard Wallace]

I wonder if the FRC community should change our expectation of how CMP slots ought to be allocated?

For example, would it be more effective* if CMP slots were allocated to states, or district systems that combine states, in proportion to the total number of official event matches played in each state or district before CMP, rather than to the number of teams?

-------
*I expect that reasonable people will disagree on how "effectiveness" ought to be defined and measured.

[dodar]

Originally Posted by Richard Wallace

I wonder if the FRC community should change our expectation of how CMP slots ought to be allocated?

For example, would it be more effective* if CMP slots were allocated to states, or district systems that combine states, in proportion to the total number of official event matches played in each state or district before CMP, rather than to the number of teams?

-------
*I expect that reasonable people will disagree on how "effectiveness" ought to be defined and measured.

So then you would unjustly give more spots to district areas. Districts, on average, play more matches than regionals.

Until all teams are under the District System, the regional slot allocation should be what we stay with.

[Ginger Power]

Originally Posted by DareDad

Best robot or best team, and how do you define either of those other than as the winner of the regional or Chairman's/EI?

Make it another judge's award. "Best on Field Robot Performance" or something along those lines. Or give EI to the best engineered robot, which is likely one of the be one of the best on the field. There isn't a simple solution like there is with Districts.

[AdamHeard]

Originally Posted by Ginger Power

Make it another judge's award. "Best on Field Robot Performance" or something along those lines. Or give EI to the best engineered robot, which is likely one of the be one of the best on the field. There isn't a simple solution like there is with Districts.

I've judged at an event, and I wouldn't trust Judges to pick the "best robot".

[Ginger Power]

Originally Posted by AdamHeard

I've judged at an event, and I wouldn't trust Judges to pick the "best robot".

My thing isn't that we need to get the absolute best robot out of regionals. Any robot good enough to be in the conversation is better than the status quo.