Round of 16 Tournament

Competition
#4

MSC 2016

2767 seeded first and picked 33. Seemed like a solid strategy, to the alliance and to the audience. Didn’t work out exactly as planned.

204 qualifying matches*, then Octofinals. Long and painful, although it was exciting to see Alliance 16 (5048, 548, 217) upset #1.

The following year we went to four 40-team divisions, followed by FiMstein. Much better. Especially this year.

*To run that many, two fields were used with play alternating, so that one field was being reset while a match ran on the other. So total qual match time was about twelve hours. The fields were informally named Batman and Superman, after our two lead FTAs. Their official names were based on sponsorship, Ford and Consumers Energy, if I recall correctly.

5 Likes
#5

That’s a good point

1 Like
#6

Good thing we have Michigan to test our ideas at scale :slight_smile:

What did you think of the management of the two fields model? Definitely requires a big venue to facilitate two fields, and more field volunteers. It would be more engaging for the audience to have more continuity in match play, assuming the fields could be placed side by side.

It would be interesting if there were two smaller fields playing a two on two match in parallel. (8 teams playing in a cycle vs 6) Team’s individual performance would be better represented in the rankings. Maybe Michigan has already done that too :slight_smile:

Thanks for the feedback Richard!

#7

All of the FTC events I’ve been a part of have done exactly this.

1 Like
#8

Would it be an improvement for FRC to consider? Or should it just remain an FTC thing.

#9

It’s hard to say. There are pros and cons, obviously. FIRST switched from 2 team alliances to 3 team alliances in 2005. I was not around then, so maybe someone who experienced that can explain their feelings about the change. I would guess that because it hasn’t changed back since then, it’s probably been a good change.

My view on 2-team 2-field competitions:

Pros:
Matches can be run near-constantly (field 1 runs while field 2 resets, etc.)

Teams are likely to be better ranked after a fixed number of qualification matches

Teams probably squeeze in one or two extra matches (8 teams playing instead of 6)

Cons:
Alliances (especially alliance 1) can be even more unbalanced (1 picks 2 and that’s it, no snake draft)

You lose some of the (imo, inspiring) scale. Robots on a sports-sized field feels better than robots in a smaller space, for subjective reasons. Perhaps that the current FRC scale is quite analogous to human-game scales?

Setting up and running two fields is a different challenge. Some volunteer and set-up aspects are likely quicker (only need one, albeit slightly more complex AV-system), and some are slower/require more manpower (refereeing/scoring/FTAing, etc.)

1 Like
#10

I really like the continuity and extra match play potential. I also like what it does to eliminations. It would make the bracket expand to a 16 team model with two teams on an alliance like FTC. Scouting at an event becomes significantly easier too. Teams just need to pick their partner, which is likely another team nearby in the rankings.

Personally I’m not one to look for a power balance to be in place, but I understand it can be encouraging for up and coming teams. Its nice to see the top performing machines win an event. Also keep in mind that breakdowns of one machine would be more influential to the outcome of a match. In the case that the 1 or 2 seed breakdown under heavy defense then it would be up to the remaining team to carry them to victory which would be tough for most teams. End game alone would typically swing the match to the lower seed’s favor.

I’m also a fan of the current size of the field. Yet sometimes as a viewer its tough to keep track of all 6 robots operating independently all across the field throughout the match.

#11

To be clear, FTC elims (for regionals/districts, in my experience) are typically four alliances of three teams. In the brackets, alliances play best-of-3, in which each of the three robots on an alliance must appear at least once (i.e., all 3 teams are guaranteed to play).

Defense in FTC (in my experience: refereeing recent years) is very limited. This is a simple consequence of two-robot alliances, because with two-robot alliances it is even more unlikely that you can prevent more points than you can score (similar logic to if you knew your 3rd bot was not going to move in 2019 FRC, you would most likely use both the moving robots on offense). Now that I think about it, this is another downside (if you like the existence of defense) to two-bot alliances.

#12

In that case I guess it wouldn’t be the exact same as FTC after all. Reducing the number of teams in eliminations isn’t appealing, but increasing the elimination capacity with the 16 alliance model from 24 teams to 32 teams would be nice.

Defense would probably be less common, but I still wouldn’t be surprised if teams engaged in heavy defense if their mechanism broke. This would encourage teams to design more reliable mechanisms for elimination play. It would probably lead to more teams to design to be a high performing specialist rather than a mediocre jack of all trades. It would be cool to see two complimentary specialist that other teams over looked win a regional.

#13

The worst thing about it is it effectively doubles the amount of time needed for playoffs. Only one hour extra was added when it was used at MSC, which guaranteed we would run late.

#14

That and the extra three hours of qualifying (comparing 204 matches on two fields against 320 matches on four fields) makes me very unlikely to feel any nostalgia for the good old Octofinals. May they rest in peace.

1 Like
#15

I think a 16 alliance bracket would lead to alliances with a super weak 3rd bot in the higher seeds if it’s run at regionals. With all due respect, (especially since this would include my own team at our first event) some teams barely move until halfway through the event, and they would inevitably end up picked and end up weighing down otherwise strong alliances if you ran this bracket at any regional large enough to run it.

I would much rather see 4th bots at large regionals replacing the automatic wildcard.

2 Likes
#16

NO!!!

Why would you replace at champs a finalist captain with a robot that probably never played in eliminations?
There is no reasonable scenario where a 4th back up bot would be stronger than the alliance captain or first pick of the finalist alliance. The regional qualification is broken enough already and this would make it worse.

3 Likes
#17

I just said I’d want it over a 16 alliance playoff, not that it’s a better system than what we have.

However, I do want to see 4 team alliances at large regionals over the current backup system. I don’t want the finalist wildcard to go away, but it can’t really be done at the same time as having 4 team alliances unless they add more available slots to Worlds or reduce representation from districts.

#18

Or reduce the waitlist.

#19

Per frc.divisons.co, there were only 4 teams at Detroit that qualified via the waitlist.

I’d like to see 4 team alliances but only if it came with a universal point system that would NOT automatically qualify all 4 winning teams.

2 Likes
#20

[question removed because it was answered by adding the word NOT]

#21

Whoops, I missed a ‘NOT’ in that post.

It’s been proposed several times by several individuals, but something like the district point system but for regionals. Winning a regional as 3rd and 4th robot would help but probably wouldn’t automatically qualify them for the championship.

1 Like
#22

One thing that people forget about four-team alliances is that discussion about who to play can get heated. On all three of our championship alliances, we’ve had at least one disagreement with our captain about who should play in what match. While all of those teams were civil and respectful in working it out, these discussions can get hairy and I don’t think thrusting these decisions on every alliance at every event is a great idea, especially if Championship qualification were to become a factor.

1 Like
#23

As for the main topic of the thread, teams struggle enough to make a good pick #24. The average team that doesn’t have a strong scouting team is going to have a hard time composing a 47 team long picklist.

5 Likes