6 Div. Round Robin Vs. +2 Wildcard & Standard Playoffs

Between you, me, and the rest of CD, I’m not a fan of the Round Robin format for Einstein.

There are a bunch of reasons but the big ones are the following:
Hard to follow. Yes, FIRST will not doubt have a leader board based on Championship Points but nothing as clear as “You win, you advance. You lose, you go home.”

Finals are a Repeat. I just don’t like the fact that the finals on Einstein are 2 (or 3) repeats of a match up we’ve already seen in the Round Robin. Steals some of the drama in my book.

Potential Skulduggery. In the later rounds of the Round Robin, it is very possible that some teams are going to be mathematically out of it while others may be mathematically in even if they lose a match. This opens up the potential for teams to influence who goes to Einstien in ways that are… …unsavory. I know FIRST and I think it is unlikely that something like this will happen, but why even build a system where the appearance of throwing a match to influence what teams got to the Finals is possible?

So… …what would I put in its place?

I think it would not be too hard to rank divisions. I know FIRST is not particularly happy about the idea of ranking divisions but personally, I think the benefits outweigh the negatives.

Once we have the Divisions Ranked 1-6 then all we need are 2 more Wildcard Alliances and we can have a 8 Alliance Playoff on Einstein. In short, Einstein ends just as we end every other FIRST tournament.

FIRST could pick from a bunch of ways to rank the divisions: Average (or median) score in qualifying matches. Average (or median) number of rotors in Playoffs. Average (or median) pressure in Finals. Total score in Playoffs. Number of Time Outs in Playoffs.

It really doesn’t matter that much to be honest.

Where do we get the Wildcard Alliances?

IN MY DREAM SCENARIO, as each alliance gets eliminated from their division playoffs, they’d go into an “all worlds second chance, single elimination, one and done tournament” that would play until there were only 2 teams left standing.

So, after Quarterfinals, there would be 24 Alliances from the 6 divisions. These teams would play 12 matches of randomly selected pairings. That’s two matches per field. We could squeeze those in just before starting the Semifinals on each field.

After the Semifinals, again, we have 24 alliances (the 12 winners from the last batch and the 12 losing semifinals). Again we have 12 randomly assigned pairings that are again 2 more matches played before we start the Finals.

After the Finals, there are 18 Alliances left in the second chance tournament. 9 matches cuts it down to 9. Two teams at random play to make it to the field of 8 which play 2 more rounds to leave the 2 Wildcards to go to Einstein.

This would add a total of 40 matches on 6 fields, or approximately 7 matches per field. That would be approximately an hour to the play per field (no time outs allowed for the second chance bracket – you either show up on time for your match or you forfeit). The excitement of these matches would be really something to see. FIRST, PLEASE DO THIS.

IN MY SECOND FAVORITE SCENARIO, FIRST publishes a rubric for how the wildcards are selected from among the Division Finalists. It could be that the top two divisions just have their Division Finalists go to Einstein. But I would prefer that it was something particular to that the Finalists that determined their wildcard selection. Again there are many choices: Average (or Median) Score in Playoffs. Average (or Median) number pressure. Average (or Median) Touchpads.

In either scenario, I think a standard 8 team playoff on Einstein is much preferred over the round robin format.

Thoughts?

Dr. Joe J.

I don’t think the divisions are unbalanced enough to merit having a 1-8 seed based on whatever criteria. I think just a regular 6-person bracket system (Winners’ and Losers’ brackets) would work better than the current system.

I 100% agree. Round robin format is just not ideal for a game like FIRST stronghold. Round robin was OK for recycle rush, however in this game, or in games like stronghold, aerial assist, ultimate ascent, etc. this 6 division round robin format on Einstein just makes it awkward IMO. It’s more complicated and IMO it ruins the entire Einstein feel.

Would prefer your dream scenario, however I think FIRST will avoid it because they don’t want to jump through as many hoops as you suggest… but I think the 6 seed elimination bracket is more feasible. We’ll see how Houston champs go though.

One of my big complaints about this format is the loss of continuity in matchups that we’ve come to expect from FRC eliminations. Going from a Bo3 to essentially a bunch of Bo1 matches takes away takes away some of the “Back and Forth” strategy between alliances as they adjust their lineups and strategies between matches.

Also, this puts a lot of pressure on scouting the other divisions ahead of time and being able to plan a strategy and implement it correctly the first time, 5 times in a row. From my perspective (sitting on my couch) it seems like this would be an exhausting experience for scouts and drive teams. I’m interested in hearing some feedback on this from the people competing.

I have mixed feelings on the round robin vs. standard playoffs debate.

On one hand, I agree with your main points. It is harder to follow, less dramatic, and leaves a door open for un-GP actions. But on the other hand, it ensures that the two best alliances on Einstein are the ones to win and get finalist. We’ve probably all seen a set of QF matches where the two best alliances were pit against each other. Sadly, only one of them can advance. Even if the alliance who lost in this QF match could have beaten every other alliance besides the one that they lost to they are still only quarter-finalists. I think that even though it makes it a little less intense for the audience, it decides winners and finalists better.

I agree with you that the divisions are not that unbalanced so I really don’t care that much about how the divisions get ranked. The main reason for ranking them is to fill out a bracket.

The more important ranking is how to select the wildcard finalists. But even that unless FIRST is going to go with my crazy complicated second chance bracket, I would be okay with the teams being drawn out of a hat, to be honest.

It would be better if it was based on some objective metric but it isn’t that important. I just don’t like the round robin thing…

Joe J.

I think they should do two groups of three. Play a double round-robin with alliances in your group, and the top alliance from each group advances to the finals.

This will stop finals from being a rematch and reduce the number of alliances that are eliminated before their last match.

It also avoids the awkward ‘wildcard’ spots in your proposal.

I agree with you. And yet, this is a part of FIRST that I have more or less made my peace with. The way we rank teams is wacky. The way alliances get selected is wacky. It’s wacky turtles all the way down. I’m okay with it.

AND…

I like to see a straight up tournament tree at the end of it all not this round robin thing.

It doesn’t really make sense given that everything else is mess up but I want what I want…

Dr. Joe J.

I don’t understand why they didn’t just have 8 divisions of 50ish teams each. That way more qualifications could be played and they could keep the 8 team bracket for Einsteins.

I imagine adding an extra “bracket” round to find the wildcard alliances wouldn’t be too difficult. I would think a one or two match series where one finalist plays another would suffice (e.g. Carver finalist vs Galileo finalist, Hopper finalist vs Newton finalist, Roebling finalist vs Turing finalist). Out of this, three winners would emerge and some metric could be used to determine which two of the three advance (I think average score would be pretty good for this). I would imagine these wildcard games would have interesting strategies where alliances are weighing whether or not to play offensively or defensively since they have to win the match/series and get high scores.

The harder determination would be deciding which division winners play against the wildcard finalists. Perhaps the divisions could be ranked by highest average playoff score in order to find the top 2 divisions that play the wildcards rather than the other division winners.

I don’t know exactly what method would best determine a good championship bracket, but I agree with you that a bracket would generally be better than a round robin.

The only problem that I see with this is that 8 divisions means 8 fields. 8 fields means more space (not only the field but also 2 more queuing space, bleachers, etc.)

2 more divisions means 2 more fields/stands, 2 more sets of refs/volunteers, and other logistical challenges that make 6 divisions more feasible. Would you rather have to recruit 33% more help or change the Einstein layout? I think FIRST would prefer the latter.

I understand that with more divisions you will need more space but they had 8 divisions last year and all the years previous (I’m pretty sure that Worlds is in the same building as it was last year for St. Louis). I don’t understand why they didn’t have the same setup for Worlds but just add another competition.

Here are some of my scattered thoughts:

FIRST needs 2^n fields at each championship. At this point I don’t even care if there are more than two of the “championships.” The idea of one championship is already fading in the rear view mirror, so let’s just hasten the hopeful transition to super-regionals and one championship again.

If we do have 6 fields, then I could see a couple better ways of doing Einstein. One way is to have two 3-team round robin pools that advance two into semis and then finals. This cuts down on the number of round robin matches—only six—and you can prevent the finals from being a rematch. Round robin isn’t great in general, I guess. The other way is to bring in two wildcards, but I’d like to see them restricted to only coming from the finalist alliances. That would also have to involve a round robin system.

This may be an unpopular opinion, but I want to bring divisions back up to around 100 teams. I think FRC deserves concentrated competition and “best of the best” alliances on Einstein. Look at the third robots that won championships in 2011, 2012, 2013, 2014: 973, 16, 610, 2848. World champion alliances should have serious firepower top to bottom. I know 100 is a lot of teams to scout. Get over it. It’s the championship. It’s hard to build alliances like that when you only have 1/12 of the championship robots available to select from.

As a side note, watching the old Einstein footage on the webcast reminds me that we need Dave Verbrugge and Blair Hundertmark back for Einstein, please.

I would love for 8 divisions but after being at Houston Champs, it’s just not possible. There isn’t enough space unless you kick FTC out of there space and put the 2 divisions on the third floor by themselves. That would then mean 132-4 would have to load in on the third floor. I agree this round robin format will not work in the future unless the games can be reset 100x faster than this.

But it’s not staying there in the future. With no regular tenant at the Edward Jones dome now, don’t expect it to come back either. There just aren’t many football stadiums with attached convention centers to choose from that can accommodate eight divisions. I personally think they should find a way to fit it in just a football stadium or shrink it until it does, but that’s another issue entirely.

I would keep the round robin, use this to rank the teams. Give the top 2 teams byes. Then play an NFL style championship.

EDIT -
Another option would be to use the round robin on the 6 finalist alliances, give the top 2 teams spots in the bracket. Then coming from a hard division would give a 2nd chance.

The bigger problem is 8 50 team divisions would be garbage from a competitiveness standpoint.

+1, this dilutes the field of excellent robots to the point where division playoffs would be potentially weaker than DCMP playoffs.

They are seemingly already weaker than some DCMP’s playoffs.