Universal Point System Proposal

FIRST General Forum
1

Tldr the regional qualification system is broken, and we can do something better while we wait for districts.
I’ve talked about this a bit before, but wanted to write them out more to gather more feedback, especially from people currently in districts.

What
The universal point system would apply the district point model to everyone still in the regional system.

Why

  • Regional teams are currently at the mercy of event assignment, where a couple of powerhouses at a regional can tank their chances of qualifying to champs

  • Regionals don’t incentivize consistent solid performance (ranking high, getting picked, making semis, etc.) or winning non-CA/EI/RAS awards (1)

  • For a given team with the same performance, their chances of making champs varies widely depending on whether they’re in the regional or district system

  • Teams in geographically isolated regions or states/countries with few teams won’t be left behind as more areas (hopefully) switch to districts

How

  • Teams get assigned “district” points (calculations linked above) for the first two regionals they attend, single event teams double the points from that event

  • Teams are grouped by state, country, or geographic region to form a pseudo-district (2)

  • Point and award slots are calculated by pseudo-district, and Chairman’s/EI/RAS judged virtually from regional winners

(1) Teams in California that would have qualified under this model

This is true across all regional teams, but to focus on California in particular, teams that are consistently making semis – sometimes even consistently making finals – aren’t always qualifying due to the wildcard system. From Jared’s CA district point calculations, ~38 teams would have qualified to champs from California (slightly fewer after subtracting awards teams), qualifying the following teams who didn’t attend champs:

3647 (rank 12), 2x regional finalist
8852 (rank 23), regional winner (2nd pick) + rookie
2551 (tied rank 26), 2x regional sf
8521 (tied rank 28), regional finalist + RAS + rookie
3255 (tied rank 30), regional finalist + sf
7042 (rank 32), 2x regional sf
5104 (rank 33), regional qf+ sf
5419 (rank 34), 3x regional sf (only first 2 would count)
6995 (rank 37), regional qf + sf
8 (tied rank 38), regional sf + qf
1072 (tied rank 38), regional sf + qf
2102 (tied rank 38), regional qf + sf + qf (only first 2 would count)

(2) Rough mockup of possible US pseudo-districts

I tried to form theoretical pseudo-districts approximately 100 teams large, since smaller districts are prone to odd effects with point qualifications (see Indiana). In reality these should probably be balanced more with geographical location and which areas play together more often.

California - 260 teams
Minnesota - 189 teams
New York - 135 teams
Group 1- 96 teams (FL - 68, SC - 28)
Group 2- 99 teams (AR - 13, LO - 27, MS - 6, AL - 14, OK - 39)
Group 3 - 107 teams (MO - 64, NE - 1, KS - 20, IA - 22)
Group 4 - 111 teams (WI - 60, IL - 51)
Group 5 - 85 teams (HI - 21, NV - 22, AZ - 42)
Group 6 - 84 teams (CO - 35, UT - 19, ID - 18, MT - 1, WY - 3, ND - 7, SD - 1)
Group 7 - 107 teams (KY - 9, TN - 25, WV - 3, OH - 53, PA (not in MAR) - 17)

85 Likes
2

Proposed changes to the district point system

I didn’t want to get these too mixed in with the original proposal, but I believe the following changes should be applied to both district and pseudo-district teams. I think these will both help with the district qualifying system in general and help bring the two models more in line:

  • There should be no auto-qual spots, winning DCMP or CA/EI/RAS should be rewarded with an appropriate number of points – pseudo-districts have no DCMP, but also the % of teams getting auto-qualed varies widely by the size of district

  • The multiplier on DCMP points needs to vary slightly (probably from 2x to 3x) based on size of district, with the smallest districts weighing it the least to avoid essentially creating a regional at DCMP

27 Likes
4

Thank you for posting this Rachel, I believe changes like these would be an improvement for FRC. This is the direction we should be moving.

11 Likes
5

I like this. I’d also be interested in seeing a breakdown of doing this not geographically tied and seeing how the qualifying teams change. So instead of each district sending their top x%, if the top 400 teams or whatever internationally got championship invitations.

I’m guessing this would create geographic advantages and imbalances, but would be curious to see data to show this one way or another.

8 Likes
6

Fully support a system like you are proposing here @Rachel_Lim ! I think you have a good foundation, but I did want to open a discussion on one sub-topic. The challenge of what to do with single event teams is real. Some situations and “incentives” can appear for a team as a season goes on, but I’m not sure how real these would actually be.

I’m imagining something like a team finishing as a Finalist at their first event, and then deciding to skip their 2nd event (and not risk the “drop” in points from 2 Finalists to 1 Finalist + 1 QF or Not Picked, etc) just to ensure they lock in a points spot for CMP.

17 Likes
7

I wonder if this could even work by just using your best regional performance. Maybe that won’t differentiate teams enough and we would have too many ties. It might have to combine with something like the round robin or double elimination proposals.

4 Likes
8

Although this is a viable proposal for some regions wouldn’t the simpler solution be to get these regions operating on the district model.

2 Likes
9

A couple potential options.

  1. Limit the types of teams that can have their points doubled if they only go to one event. Perhaps non-North American teams and rookies.

  2. Change the multiplier for different types of teams, it’s double points for rookies and international teams but it’s 1.5 or 1.75 for other teams.

  3. Have teams commit to a single event or two events before their first event so they can’t cancel a 2nd event just to get double their first event points after the fact.

11 Likes
10

Cutting awards teams leaves a really sour taste in my mouth. Sure it would increase the competitiveness of Champs, but is that really the end goal of the program?

The place where districts are really falling apart at the seems is just in the smallest districts, the areas that would theoretically be served by ~1-2 regionals. It seems weird to cater the solution to the outliers. The district system felt like it did a great job of qualifying the best teams in somewhere like PNW where the district actually replaces many regionals with one district.

Also - the District model already qualifies way fewer award teams. Why is it even in consideration to cut the <25% of District teams qualifying to worlds on awards when ~50% of regional teams qualify on awards? Indiana is the only district qualifying the right amount of teams on awards if the regional distribution is anything to go by.

12 Likes
11

The reality of that is that some regions have leadership that’s at best apathetic towards implementing districts, or have other situations that would complicate the creation of a district. (such as upstate vs NYC in New York) This proposal is sort of a compromise between implementing full districts and the current regional qualification system to bring the advantages of districts without any action required by RPCs.

1 Like
12

I’d actually propose something different: the “average” district score.

For teams competing at 2 events, you take the sum of their scores. For teams at one event, you take their first score, plus an agreed-upon “average” score. As a theoretical, the “average” score would be defined as “for a 36-team event, the sum of the available points divided by 36” or “for a 36-team event, the score of the #18 team as sorted by points after all awards”.

I suspect “best event” would work, but then if you’ve got the misfortune of being at SVR and Sacramento as your only two events then landing on the #8 alliance at both, that might be not so great either…

The other possible option, to account for teams that do 3+ events per year, would be straight average score. Sum all event scores for one team, divide by events. (Though that does open it up to the shenanigans Brandon was thinking of.)

4 Likes
13

I’m in favor of this, but I’m sure there’s a bunch of geographical quirks within North America. It might be fun to catalog them all.

Here’s British Columbia’s:

  • Can’t compete west: Ocean

  • Can’t compete south: PNW District

  • Cant compete north: Also PNW District & Wilderness

  • Don’t want to compete East: 40 school busses over the Rocky mountains in Feb/March to connect Calgary and Victoria

  • Can’t compete with ourselves: <24 teams & most teams can’t afford a second registration

I hold out hope that one day BC’s free arenas can be combined with the PNW competition fields.

3 Likes
14

I hate to be an overly critical or negative voice in what I believe is an important discussion, but I’ve spent a decent amount of time thinking about this and I am not optimistic that a district-point system will affect much positive change in the current regional system.

First of all, district points benefit higher seeded teams and higher seeded alliances by default. Given that a lot of the complaints leading to this idea stemmed around teams being steamrolled by high-seeded alliances too often, I doubt many teams would be excited when, in the rare instance #7 upsets #1 in finals for example, the 1 alliance still walks away with an advantage on points. What the regional model excels at is giving every elims alliance a reason to give it their all, knowing that there’s no way their season can be over if they just manage to win, or make finals in case wildcards are in play.

The district system succeeds largely because of the hierarchy of qualification and raising stakes - teams don’t have to win each individual district event since some X% of teams qualify for district champs (anywhere between 1/3 to 2/3 depending on district size, roughly), and once you qualify for district champs you don’t necessarily need to win to advance to champs due to the number of points slots available (again depending on district size). At best, such a system for regionals gives a team a second chance to redeem a middling score from the first event (or to do just well enough at both to make the cutoff).

The other major question I have about this system is on the accounting side. The point slots in districts aren’t created out of thin air - they are taken away (roughly) from other slots that would have existed in the regional alternative. Larger districts benefit the most from this at the expense of qualifying award slots (CA/EI/RAS) and event winners (only 3 winners per DCMP vs 3xN for regionals). In fact, the robot-based slots (winners + points) are quite seamless as the point slots benefit teams who performed well both at DCMP and throughout the season. But even then the scores are pulling from 3 events per team including the DCMP which all eligible teams attended (divisions or otherwise). tl;dr the district system utilizes the district/DCMP hierarchy to cull the team list iteratively until there is a list of advancing teams, which would not happen in this projected model.

The problem with single-event teams cannot be understated. A lot of teams for any number of reasons only attend 1 event per year. Doubling points is a band-aid that might work (assuming the loophole is closed for teams dropping the second event after a very strong first event). Taking each team’s best single event instead of best two seems promising, but it ultimately favors the event winners who already qualify in today’s system. I have yet to hear a solution that is fair to all parties.

I also wonder how your qualification method handles regional qualifying awards? If a CA winning team doesn’t qualify for champs on points, will they still be judged virtually for that region and have a chance to advance based on the award? Will the awards be given out much like today’s districts where only a fixed number of qualifying awards is given out for the region? I’m not necessarily against this idea - a couple districts gave out awards pre-DCMP this year and I haven’t heard any major complaints (not to mention all of the 2020-21 remote Chairman’s judging).

To finish this short novella I want to mention that I am cautionary on how to handle the pseudo-district boundaries. The approach of trying to classify by geographic boundary to make same-size regions is a good baseline but in my opinion a fundamentally wrong approach. By lumping in HI with the continental west, for example, you are having two groups compete for slots that have not competed against each other save for at most a handful. I would advocate to keep those regions separate for qualification for this reason, and keep Hawaii as a separate standard regional with the option for NV/UT/etc to be treated as one unit. I think there’s also a lot more nuanced discussion to be had about navigating this part of the conversation since Regionals currently are not geographically bound.

8 Likes
15

image
image974×433 50.8 KB

Plz use the quote feature (or subheadings if you’re feeling fancy) on giant text blocks. It’s incredibly difficult to parse arguments on such a multi-faceted issue without.

1 Like
16

Do not like. Sorry

image
image1298×277 37.5 KB

Just doesn’t compete with the excitement of “win and in” from the regional experience.

Also being a state with few teams means we travel to regionals in different states from year to year, letting us see a wider variety of teams. Seeing the same ~90 teams year after year seems not as inspiring.

And most importantly, you’ve presented no data to show this proposal would materially change which teams would have gone to champs. And would this be a good thing — in 2022 Team XXXX went to champs as a regional winner, but under the pseudo-district scheme they would have been bumped by Team YYYY who squeezed past them on the basis of doubled points on an imagery award.

6 Likes
17

The current system does indeed leave out robots capable of competing at the highest level but get “blocked” by powerhouse teams winning multiple events in elims. Having said that, I propose some additional considerations for teams to qualify for Champs:

  • Award a wildcard to certain engineering design award winning teams such as Industrial Design, Robot Quality, Engineering Excellence, Innovation in Control

  • Eliminate an automatic spot for Rookie All Star (I’m sorry, I see way too many rookie teams at champs that made it there because they were the only rookie at an event, are not in any way prepared to compete there, and should be spending their $$ on team buildup rather than champs registration. Rookies still have ways to get there based on awards merit above)

  • Eliminate wildcards for all but the alliance captain of finalists. This reduces the ambiguity of 8th seed 1st and 2nd picks (if they happen to be finalists) and gets rid of the “hope we get picked by alliance 2,3,6,7 and not 4,5,8” mindset

6 Likes
18

Will someone take on the task of seeing who frome each of these models would have gone to champs this year?

I’d also like to propose that the winners of a regional still auto qualify. If regional teams are still paying what they are to play then there had to be some benefits to the regional model. Also would like to make it so that district teams can’t come in and take spots away from regional teams( but this topic has already been discussed in other threads)

19

Why not use the district points system universally, without defining number of CMP slots by geography? Win a Regional or DCMP, and you’re automatically in. Earn RCA, DCCA, REI, DCCEI, RRAS, or DCRAS and you’re automatically in. The remaining spots go to teams worldwide based on points accumulated in the season (once again, normalizing one-regional teams with multiple-play teams).

Of course, this is all a band-aid until FRC finally makes SuperRegionals a reality.

2 Likes
20

California, Michigan, Texas teams “Yes!”

Indiana teams “Oh the horror!”

I’m in favor of points based qualification. Maybe the waitlist should be emptied based on a points based requirement instead of random.

5 Likes
21

This is highly dependent on your definition of “simpler”. By most definitions, I would say no. Not saying that districts aren’t the “right” option, but I think a universal point system would be significantly more simple for most regions to adopt than moving the entire world to districts. Easier for FIRST HQ? Yeah, fewer edge cases and program philosophy questions to answer. Easier in terms of total person-hours put in to make the change happen? Unlikely.

I like the idea here a lot. One (definitely surmountable) challenge I haven’t seen mentioned is that it effectively forces all teams into a week 6/7 qualification, creating a massive scramble at the last minute for everyone, whereas right now that’s relatively contained.

In MN we have used a slightly modified district point ranking system for qualification to our State High School League tournament and it has worked well for created a highly competitive event. We look at your first event only since a decent number of MN teams attend only a single event.

There are a few other relatively minor changes you’d have to consider for single event teams as well. For example, this year we changed the rookie point bonus from 10 (and 5 for second year teams) to a flat 5 for both first and second year teams since in a district model those 10 first-year points were effectively distributed across two events.