FRC Blogged - 2014 FIRST Championship Eligibility

[Jon Stratis]

Originally Posted by Racer26

The truth of the matter is that we DO need more than 400 teams at Championship.

I have proposed a solution that provides 4 divisions of 140 teams with 12 qualification matches each.

Each division running 2 fields, as the 2004 and 2006 Toronto Regionals were done, provides nearly double match throughput. It requires fewer volunteers than adding more divisions, avoids lengthening Einstein, as well as avoiding noise pollution issues associated with having 8 different divisions crammed too close together.

We've outgrown the old CMP model. Something drastic has to change in the next 2 seasons. It makes the most sense to make such a change in 2015 when the control system is already being revamped.

Objectively speaking, what percentage of teams *should* be able to go to champs, in an ideal world (For example, The state of Rhode Island gets converted into one giant conference center with an attached indoor Olympic stadium that could hold over a dozen FRC fields)?

IMO, the answer isn't 100%. If it was 100%, then why call it champs? There needs to be a selection process, and getting there needs to feel like an elite accomplishment. It isn't 50%. I would even say that it isn't 25%.

How does 15% sound to everyone?

[Hallry]

Originally Posted by DevenStonow

And if I'm not mistaken, I believe in MAR (so it should follow through with FiM history) the last statement does hold true; the only example of this I could find/think of excluding team 341(who have a HoF spot, so it's skipped anyway) was 2590 in 2012, who, along with having enough qualifying points, as well as previously winning the Montreal Regional, were skipped over in giving out point spots according to this document

Yes, this is true in MAR, and I'm pretty sure (not 100%) in FiM too. At least for MAR, some of their St. Louis slots are given to the top ranked teams (via ranking points accumulated over the whole season, see the Mid-Atlantic Robotics website for more details) in the region. In 2012, this sent 5 teams to Champs, and I believe it was increased to 6 in 2013 (along with there being another spot for EI too - lucky us )

In 2012, 2590 was skipped over in the MAR ranking slot distribution since they won Montreal, 365 was skipped over due to HoF, and I'm pretty sure 222 was skipped over since they were preregistered via waitlist for St. Louis. 25 was also skipped over since they won MAR Champs. However, 341, already prequalified for Champs due to HoF, also won the MAR Championships, and were first in points. While they were skipped over in the distribution of the MAR slots given to the top ranked teams, I believe this wasted away a slot given to a MAR Champs winner.

In 2013, 11 was skipped over in MAR ranking slots since they already qualified by winning Palmetto, 1676 was skipped over since they won EI at MAR Champs, and 103 was skipped over due to HoF. 2590, 2729, and 1640 were skipped over since they won MAR Champs.

[Allison K]

Originally Posted by Jon Stratis

Objectively speaking, what percentage of teams *should* be able to go to champs, in an ideal world (For example, The state of Rhode Island gets converted into one giant conference center with an attached indoor Olympic stadium that could hold over a dozen FRC fields)?

IMO, the answer isn't 100%. If it was 100%, then why call it champs? There needs to be a selection process, and getting there needs to feel like an elite accomplishment. It isn't 50%. I would even say that it isn't 25%.

How does 15% sound to everyone?

15% looks good for now, but might be on the high side if FRC grows. My assumptions in this statement are that anything over 400 teams is unreasonable given the logistics demands, and anything under 300 is undesirable as well.

[Alpha Beta]

Originally Posted by waialua359

As the pressure to increase the # of teams at CMP keeps coming up, how about increasing the # of teams that make eliminations similar to offseason events such as IRI? Good teams can continue to play, other than the 8-10 matches in recent CMP events.
The format of just 24 making eliminations out of 100 or more teams, seems a bit too low.

2013 was the perfect example of why such a format would have been useful.
Teams were rushing to get reinspected after they attached a blocker against full-court shooters.
At IRI and TRR, we just selected a robot that could already do it.

I would guess that every team that attends CMP would want a greater chance at making eliminations.

Originally Posted by dodar

I think this might be the best solution. Just allow each alliance to pick their "4th bot."

I'm on record as liking this idea back in 2010. I'm still on board.

[Gregor]

Originally Posted by waialua359

As the pressure to increase the # of teams at CMP keeps coming up, how about increasing the # of teams that make eliminations similar to offseason events such as IRI? Good teams can continue to play, other than the 8-10 matches in recent CMP events.
The format of just 24 making eliminations out of 100 or more teams, seems a bit too low.

2013 was the perfect example of why such a format would have been useful.
Teams were rushing to get reinspected after they attached a blocker against full-court shooters.
At IRI and TRR, we just selected a robot that could already do it.

I would guess that every team that attends CMP would want a greater chance at making eliminations.

Trying to make a picklist for 24 teams is enough, the 32 team picklist takes forever. It only works at IRI because of the ridiculous talent depth. Even then, it began to be a struggle around team 28ish.

The talent depth in the divisions is not there. The upper back half of the draft already starts to see a massive drop in performance. I don't want to have to pick (or be picked as) another barely functional robot. There are always exceptions and outliers (read, the successful alliances) but for the most part the third robots are low enough.

[JB987]

I'm with Glenn and Aaron regarding selection of 4th bot for Division/Einstein elimination alliance. Only "negative' impact I see is a few minutes added to alliance selection time and a little extra cost for extra trophies/banners. There were at least 8 more good robots available at Archimedes last year.

[AdamHeard]

Originally Posted by Gregor

Trying to make a picklist for 24 teams is enough, the 32 team picklist takes forever. It only works at IRI because of the ridiculous talent depth. Even then, it began to be a struggle around team 28ish.

The talent depth in the divisions is not there. The upper back half of the draft already starts to see a massive drop in performance. I don't want to have to pick (or be picked as) another barely functional robot. There are always exceptions and outliers (read, the successful alliances) but for the most part the third robots are low enough.

Looking at actual data from champs, and I don't see 4th robots being substantially worse than 3rd. I'm sure teams would prefer picking their own backup versus the next remaining highest seed.

[Lil' Lavery]

In terms of everyone picking their own back-up/4th bot, I'm not sold on the idea. It makes for an interesting debate on multiple fronts.

Originally Posted by waialua359

As the pressure to increase the # of teams at CMP keeps coming up, how about increasing the # of teams that make eliminations similar to offseason events such as IRI? Good teams can continue to play, other than the 8-10 matches in recent CMP events.
The format of just 24 making eliminations out of 100 or more teams, seems a bit too low.

Why does it seem too low? Just because a larger percentage of teams make eliminations at regionals than Championship? This is the championship event, shouldn't there be a higher grade of exclusivity there, both in terms of who qualifies for the event and who qualifies for eliminations? Why further dilute the field of competitors? We already have a 96 team tournament.

Originally Posted by waialua359

2013 was the perfect example of why such a format would have been useful.
Teams were rushing to get reinspected after they attached a blocker against full-court shooters.
At IRI and TRR, we just selected a robot that could already do it.

This obviously changed the game (and alliance selection) dynamic heavily, and without a clear answer of if it's "better" or "worse." It's obvious how the proposed format encourages both building and selecting "specialized" robots and role players over well-rounded machines. It's clearly not as simple as picking the next 8 "best" robots in the field, and I'd venture to say it would likely also impact earlier parts of alliance selection.

To me, one of my favorite parts of the elimination tournament is seeing robots suddenly change their strategy to match an opponent. Rather than playing "rock, paper, scissors" with which robots to put into a match, teams would have to adapt strategically (and sometimes mechanically, as you mentioned with mounting blockers) on the fly. I find moments when teams like 469, 973, and 1126 suddenly become shutdown defenders or 217 scrambling to attach pool noodles or 1114 and 294 working together to improve mini-bot launchers as some of the coolest and most memorable moments of elimination tournaments. I love the strategic engineering that occurs when you have to figure out how to make your alliance composed of X, Y, and Z beat an alliance of A, B, and C.

Originally Posted by waialua359

I would guess that every team that attends CMP would want a greater chance at making eliminations.

I also addressed the exclusitivity of the elimination tournament. I obviously also see the merit of trying to enhance the experience of many teams at Championship.

The biggest appeal of adding teams to the elimination tournament for me would be mitigating the impact of "weak" alliance captains (assuming they were humble enough to bench themselves). We always see a few alliance captains each year who are borderline elimination teams (or sometimes worse), and rather than damning their alliance partners to a quarter-final exit, a back-up bot may increase their odds.

[Chris is me]

Originally Posted by Gregor

Trying to make a picklist for 24 teams is enough, the 32 team picklist takes forever. It only works at IRI because of the ridiculous talent depth. Even then, it began to be a struggle around team 28ish.

The talent depth in the divisions is not there.

I really disagree. The 24th best robot in a division is far better than the 24th best robot at a regional, and somehow we all manage to make 24 team pick lists at regionals. For example, this year's 24th selections were 217, 862, 4039, and 2959. Yes, it is much harder to make a good pick list this far down the draft list, but it's by no means impossible.

Also consider the teams that missed CMP eliminations this year. 4265, 4124, 230, 2145, 399, 125, 2648, 2485, 842, 180, 79, 191, 503, 2439. Almost all of these teams outright won regionals. Surely there's room in the backup round for them.

[PayneTrain]

Originally Posted by Libby K

By the way, you still haven't answered WHY we need more than 400.

I think the capacity is going to get really finicky as early as 2015, but definitely by the last year under the current venue agreement, unless there are sweeping changes to the eligibility for the event (dropping EI, RAS, Wild Cards, or pre-qualified participants) there are size restrictions for regionals, there are caps for number of regionals, or dropping points-based slots for district systems. Following the trends of event increases, team increases, and the average acceptance rate for merit-based slots (I believe 85%), there will be a problem next year, but that's not factoring in fluctuations of merit based slots of groups of regionals transitioning to a district system which can see an increase, decrease, or stagnation of slots held.

[BBray_T1296]

Originally Posted by Allison K

15% looks good for now, but might be on the high side if FRC grows. My assumptions in this statement are that anything over 400 teams is unreasonable given the logistics demands, and anything under 300 is undesirable as well.

This is absolutely a great point. No matter what static system anyone is proposing (15%, or 400 team limits), they will all be outdated extremely quickly. As the number of teams (and events) increases, the number 15% represents increases, and the number 400 becomes tiny. In 2015 when there is (theoretically) 150 events sending teams, who gets to go and who doesn't? sending 3 teams per event is already over 400, and what about chairman's and rookie all star? 15% becomes a significantly larger number but by definition would scale properly with size. But!, there is no venue with capacity for 8 fields/600 teams. (please, if you haven't already, see this) As you will read, the schedule is stuck, no room either forward or back, none-the-less expand. Districts add a level before the current regional, and allow for a marginally larger coverage for district champs, but if the organization shifts entirely to the district system, the number of district champs rivals that of the current number of regionals. Adding another level of competition after champs (District>District Champs>[then a regional?]> World Championship) has little room a reorganized schedule, unless the number of district events per weekend is dramatically increased. If you read Frank Answers Friday (you can find it here) you should understand why it is logistically impossible to have these progressive events immediately after each other, and thus there must be a weekend between each level. Having dozens and dozens of districts in one single weekend to fit the locked schedule, and necessary timing, requires: several dozen fields, and more challengingly: thousands of (more) volunteers.

Even combinations of these plans would soon fail as FIRST grows so quickly.

I am not proposing a strategy myself, I am simply gathering information and acting as the soothsayer of the impending crisis.

[themccannman]

I definitely agree that 4th picks should happen. Teams that are good enough to be picked as a 4th robot are more deserving of a chance at elims than the team that happened to seed high (since we all know seeding has very little to do with robot performance) and I'm sure that alliances would prefer to choose their own back up robot.

Regarding location and fields, there is more than enough space in the st louis arena for 8 fields (specifically the giant empty space in the middle), the issue is arrangement and volunteers. Twice as many fields translates to nearly twice as many volunteers. The only way to place twice as many fields in the arena is to either arrange them length wise, or put one behind the other. Neither of which is spectator friendly unless you put bleachers in the middle of the stadium facing out. The other issue is scouting. Scouting would become almost completely impossible for most teams as they wouldn't have enough team members to scout 2 fields at once.

After thinking it over, I really don't think it it's a good idea to move to 2 fields per division. In terms of exclusivity I don't think there is a percentage of teams that can really be set to attend CMP. The number of teams attending is really limited by the arena capabilities. There should be no more than 100 teams per division to ensure an adequate number of matches (less than 10 matches per team is undesirable), and you can't add divisions in small increments, you essentially need to add teams at a rate of 100 per addition to maintain the current division layout. With how CMP is run right now I can't see there being any more (or much less) than 400 teams without a major overhaul. I think FRC is staying as is for a while but I definitely think the largest improvement we will see is the increasing prevalence of districts as that seems to be the best system to determine CMP eligiblity.

Districts can't get here soon enough

[Racer26]

Originally Posted by Jon Stratis

How does 15% sound to everyone?

15% is roughly where we are now.

[Kevin Leonard]

Originally Posted by Gregor

Trying to make a picklist for 24 teams is enough, the 32 team picklist takes forever. It only works at IRI because of the ridiculous talent depth. Even then, it began to be a struggle around team 28ish.

I disagree. There were some incredible robots left that didn't get picked.
11, 1625, 1741, 829, 2252, 1676, and many more were all putting up good numbers.
I thought for sure 11 would get picked, since they had such crazy potential.
I also thought that some alliance would select 1640, because although their climber wasn't working very well at the end of quals, if they could get it working by elims, they could provide a climb to an alliance sorely lacking one (Like the 118-469-2590 alliance).

IRI certainly went way different that how I thought it would in more than one regard (How did 1334 slip to the second round? How did 4265 slip to the third round? I just don't get itttttt)

Anyway, it's past midnight and I have school in the morning and this post is barely relevant to the conversation. Good night, Chief Delphi.

[Caleb Sykes]

I don't really want to have to make a 32 team pick list, but it would be nice to have a little control over which robot your backup robot would be. What if alliance selections stay as they currently are, but you can pick any non-selected robot you want to be the backup when/if the need arises? I know that our team has run into instances where the available backup robot is somewhat mediocre.
Just a thought.