The “best” alliance at the whole event might be a second seed in a division that had a bad match in eliminations. The final finals is definitely not the best x number of teams - its for the drama.
Oh, absolutely, which is why if I had a magic wand it’d be Bo3 8 alliances of 2 robots in division elims.
In my opinion, Round Robin is confusing to follow, but has advantages in terms of ensuring the best and second best team have a match together as the final match of the event. If we’re doing things like playoffs between divisions to get into Round Robin, you’re defeating the purpose of it and just making things more confusing for no reason.
Let’s not forget “having fun!”
My priors would be to maximize the number of plays for the most number of teams. Finding the “best” robot alliance out of 800+ is secondary. Yes, even at an event billing itself as the World Championship.
There are a number of teams that have consistently performed well at Signature Events throughout the season. That, to me, is almost more impressive than winning a single 800 team event
Yep, having fun is #1 also in my book. But I’m a realist that 796 teams will leave having fun, two teams will leave very bitter and two teams will be over the moon.
imagine having fun at a robotics tournament lol what a silly proposition.
That number feels high
That number feels low
This one seems about right though
Because you know what’s happening beforehand vs knowing when you arrive. It’s especially useful for making connections with other teams prior to the event starting. Coordinating Scrimages, learning about other strategies. Basically extending teamwork outside of your own school, something which i think the recf tries to encourage alot
If you would like to answer this yourself, please design and share a single-elimination bracket with 10 teams that will be mostly fair to all teams competing in the overall finals.
(Please note that accurate seeding of division alliances will be extremely difficult)
You are referring to scouting - which is competition-specific.
but what @Foster is talking about is the entire worlds experience. And of course, also the need to have the division lists out 1 month before the event.
inb4 skills-based seeding
Also, does anyone know why the capacity was increased to 820 teams? I am guessing this means 10 divisions of 82 teams as 820 is not divisible by 8.
Naah. We will never do well in any simulation.
Lol… Think we were not even in the eliminations during the Tower Takeover Virtual Worlds.
But really, SG teams will never do well in any true-skills matrix or whatever matrix they are using. Simply because we can’t afford to go for so many official events (which means we will need to travel out) and in local events, the 8059 teams will be killing off each other and the other teams just have that small task of beating 8059A that’s all 
clearly, 41 divisions of 20 teams, duh
The venue plan shown in the recf town hall shows ten sets of fields Whats Worlds Like? - #63 by Connor. I would expect 10 divisions based on that plan.
I don’t care if there are 2 or 41 divisions. You posted that 10 was abnormal and I called you out on why 10 was abnormal. Still need an answer on that.
Much as I’ve called out others for why they needed the number of divisions now. I got back “well to start strategy with them!” Sure, lets go with 100 team divisions, you all are really going to contact 99 other teams and do strategy for the 8 matches you’ll play? Or are you going to really wait for the match schedules to come out and just talk to the 8 teams you need to work with? Or is the harsh truth you’ll wait until the queue table to speak to your alliance partner?
Which reminds me @Sylvie, how are you making out with that really simple program / webpage that you enter your team number into and how many divisions and it returns the possible teams in your division? That would be great then, I can practice going to DougMoyersWorlds and see what teams are in my 20 team division. And it’s a conversation starter for the rest of the thread "Hooo, really want 7 divisions because I’ll be in the same group as 23T, but really don’t want 6 divisions because then I get 23A and they just have a clawbot.
Hey dog, why skills based seeding vs actual match play seeding? Are you seeding divisions that way so there is “Top 80 teams here, bottom 80 teams there” meaning that the lower skilled teams are just not crushed all week long by more powerful teams making their event happier?
Or are you thinking of the classic TSA setup where you take skills runs, that sets up the alliance order, alliances are picked (can you imagine how long to pick 400 alliances?) and then have elimination tournaments. If you did that you could make the entire elimination BO3 since the teams in the Round of 512 would only play 2 matches instead of 10, freeing up valuable match space for the top contenders.
fun fact: at least for texas, we’re not doing seeding off of skills this year, we’re just running a straight tournament. I think they’re moving away from the ways of old, so if anyone wants to seed off of skills you’ll likely be the only ones doing it now.
for the mental sanity of everyone involved, I would hope you cut it at 128-256 alliances max, though 400 alliance elims seems like bracket chaos and could be fun, imagine the qf being seed 244 v 375 after you see a gazillion upsets lol
Let’s be honest, it’s for nerds who love stats to make way too early predictions that won’t remotely be true. It’s fun.
Overall finals. If you have 10 division winners, that doesn’t work out into an equal number of matches, meaning you have to pick certain alliances to either play into the full bracket or you give some a bye to some teams – how do you pick that metric? All the alliances won their divisions, so how you determine who goes where has to picked some other way? Qualification metrics could be skewed, #12 in one division could theoretically be better than #2 in another. So what, we flip a coin and determine it that way?
It makes for a weird/possibly unfair bracket.
Ideally you want a bracket that’s a power of 2 (4, 8, 16) so that each alliance is the same number of elim matches away from winning the event. They got away with 5 because of round robin, but round robin obviously wouldn’t work for 10 alliances.
So what happens is, if you get unlucky and end up in a division that’s in the final elims Ro16 and doesn’t get a by to QF, that’s one more match you have to win where bad luck could knock you out.
Obviously in normal events the bracket isn’t always even, but higher seeds are the one who get a by, so they “earned” it. In this case it’s just luck of the draw with the (in theory) closest and hardest matches of the event, where luck is most likely to tip it.
And yes, preempting the likely response, I’m aware you can also get unlucky with qual schedule and who’s in your division and in about a hundred different ways, but the fact that luck is already a huge factor doesn’t mean it’s ok to make it an even bigger one.
EDIT: oops, beaten to it, this is what I get for trying to write out this wordwall on my phone
Cool, I stopped helping out on TSA events because of the skills runs. For a team that spends 6 months, close to 60-70 hours of work to play 3 minutes of skills and then sit out the rest of the day didn’t seem to very inspiring.
I’m literally not the one making it. It’s the spreadsheet I linked above.
But that has a flaw, the builder needs to know the number of divisions to create it, and it creates all the divisions. People are interested in their division, so being able to enter their team number and how many divisions and get the list of the teams they will be playing with.
you quoted my answer; to make it more clear, you can’t design a fair single-elimination bracket with 2 teams. its abnormal because it’s different than what the RECF has done for the past 2 renditions of worlds overall finals when there were 8 divisions.
obviously not, but i would assume i’d talk to far more than 8 teams in an 100 team divison yes.