|
So, I just wanted to talk a bit about a new potential format that I think might work to make the Dota 2 scene more international while at the same time providing some games between the best teams possible. I got this idea while listening to the current CS:GO major in the background while I write a research paper, so I'll explain what I'm thinking would be a good format and what the possible consequences of that format might be!
Just one aside: I wouldn't use this approach with The International, as I feel the large groupstage at that tournament adds to the fun and prestige of the event. This would only apply for future major tournaments!
8 Direct Invites - These teams are seeded directly into the Major Tournament. This is based on LAN wins / placings, previous major performances, as well as just plain ole' picking good teams. The problem with previous majors is that inviting teams often feels shallow; while 4 invited teams doesn't provide enough depth for invites, 12 is often FAR too many. I think 8 is a good number and it works well with the format I'm about to propose.
16 Qualifier Slots - Now you might be thinking: 16 qualifier teams? That's a huge number! And you'd be right; in dota 2 history the most qualifiers we've ever had from regions for a main event to a tournament is 3. That said, with Valve's expansion to include 2 new regions for qualifying in the next major, I think that this number of teams is almost necessary to make the numbers work. Here's how I think this should go:
EU - 3 slots NA - 3 slots CN - 3 slots SEA - 3 slots CIS - 2 slots SA - 2 slots
This adds up to a total of 16 teams, and allows for some interesting stuff which I'll explain in a minute.
1st Group Stage - This consists of the 16 teams from the regional qualifiers only; the 8 directly invited teams do not participate in this stage. You separate the teams into two separate pools:
1) 1st place from each group and second place for the 2nd place teams from EU and CN (total of 8 teams). 2) All the rest.
You then draw randomly to have one team from the 1st group play one team from the second 2nd group. They play a best of 3 series to determine a winner. Now, you have another two groups:
1) Winner from last matches (1-0) 2) Losers from last matches (0-1)
You then randomly draw opponents from each group to play against each other, so you have all the 1-0 teams facing other 1-0 teams and all the 0-1 teams facing other 0-1 teams in another best of 3 series. This then leaves you with 3 groups:
1) 2-0 teams (4 of them) 2) 1-1 teams (8 of them) 3) 0-2 teams (4 of them)
They then play again in their groups. This pattern keeps repeating until a team either wins 3 games or loses 3 games; teams that win 3 games advance to the next group stage while teams that lose 3 games are eliminated from the competition. It takes a total of 5 rounds of games, and you end up with 8 teams that will have 3 wins and 8 teams that will have 3 losses.
2nd Group Stage - You can do the games in this stage however you like, but you take the 8 teams with 3 wins from the previous stage and you pair them up with the 8 direct invite teams for another 16 team group stage! You can do the same format as above, a GSL format, or anything that really tickles your fancy. The really important part, in my opinion, is the first part. Let me explain why:
Pros
1) Every single team that makes it to the tournament will play 3 games. This is very important, especially when introducing new regions. It allows for new teams to the international stage to get more games than they might otherwise by forcing them to lose 3 times before they get eliminated (compare to TI's 2 for the team that loses in the wildcard), and you never have to be in the trap of a traditional bracket of losing to the same team twice to be eliminated from a tournament in the group stage. In fact, no team has to EVER play the same team twice in the first group stage (and they probably shouldn't). For example, consider this bracket:
A > D, B > C, A > B, C > D, B > C
In this case, team C loses two games to be knocked out of a GSL group, both to team B. If team B is the second best team in the tournament, then C effectively gets no chance to prove themselves against anyone else. The group stage proposed above would allow for no team to play another team twice, which is a huge advantage over the current group stages in my opinion.
2) With the huge number of regions involved, this allows for teams other than the very best team from the region to play games against international opponents, benefiting the whole scene! For example, it is highly likely that Team Infamous wins the SA qualifier. However, if only Infamous gets international experience, then the scene becomes even more polarized around that team. It takes bringing their biggest competition along with them to help maintain a healthy and competitive scene in a new region. This new format would allow for at least two teams from every region while still giving more slots to regions that are stronger and more developed!
Cons
1) This group stage is LONG. It will take time.
2) You have to fly out 24 teams rather than 16 or 18.
In the end, here's what a group stage like this could look like:
EXAMPLE FIELD FOR THIS FORMAT
8 Direct Invites - EG, OG, DC, Ad Finem, iG.Vitality, Newbee, TnC, Team Liquid
3 EU - Team Secret, Na'Vi, Alliance 3 CN - iG, VG.J, LFY 3 SEA - Team Faceless, Warriors Gaming, XTCN 3 NA - NP, Complexity, Team Freedom 2 SA - Team Infamous, SG eSports 2 CIS - Virtus.Pro, Vega
Now this is just an example (could switch Team Liquid with a bunch of different teams, they did win Dreamhack but there's arguments for tons of teams in that last slot), but doesn't a 16 team format with those teams listed above seem pretty damn interesting and good for their respective scenes? I think that group stage before the main invites would be a blast! Also, if one region comes to truly dominate the others, then they'll start making top 8s and top 4s in mass at previous majors which would allow for more talent from that scene to come in naturally if they can beat other regions in the qualification stage!
I've borrowed shamelessly from CS:GO for this, but that format is so good in retrospect I hope they apply it to Dota 2!
|
I don't think valve will ever run a tournament that sees teams fly around the world and then get eliminated at the hotel without ever playing on the main stage ever again.
|
I thought it went downhill from 8 direct invites that doesn't even go through groups.
Honorable mention to this: "1) 1st place from each group and second place for the 2nd place teams from EU and CN" seems really arbitrary to invite second place from eu and cn, while other regions also have 3 spots.
Cons: format is too complicated, would cost too much for only a major, it would last almost 3 weeks with breaks between each phase. Too many teams, t2~3 teams do need more space, but not on majors i would say. Leting team freedom (and so many other t2 teams) would make the tournament less important and top level.
Your first argument "pro" is more a con for me. Except boston wich was short, all major teams play way more than 3 games each. Your second one is off too, even if only infamous gets through a whole season of majors, next season they will probably shuffle making new teams, while that a lot of fierce new players have more hope of making it big, so polarization is not that big of a problem.
I personally don't see anything going for this format, except it's a longer lan (wich can be a negative because it's more expensive)
|
On January 24 2017 03:17 Pontual wrote: I thought it went downhill from 8 direct invites that doesn't even go through groups.
Honorable mention to this: "1) 1st place from each group and second place for the 2nd place teams from EU and CN" seems really arbitrary to invite second place from eu and cn, while other regions also have 3 spots.
Cons: format is too complicated, would cost too much for only a major, it would last almost 3 weeks with breaks between each phase. Too many teams, t2~3 teams do need more space, but not on majors i would say. Leting team freedom (and so many other t2 teams) would make the tournament less important and top level.
Your first argument "pro" is more a con for me. Except boston wich was short, all major teams play way more than 3 games each. Your second one is off too, even if only infamous gets through a whole season of majors, next season they will probably shuffle making new teams, while that a lot of fierce new players have more hope of making it big, so polarization is not that big of a problem.
I personally don't see anything going for this format, except it's a longer lan (wich can be a negative because it's more expensive)
I admit some of it is rather arbitrary, but you can work around that by taking out the seeding in the first group stage if you want and removing it completely, just choosing opponents randomly from the first stage.
As far as timing goes, the CS:GO events fit this format in about 10 days of game time (includes both group stages and final bracket stage), but maybe that's too long. Do you know how people view this format in CS:GO? They've been using it for a while there, and I don't see that many complaints about it. Maybe I'm just missing something.
But either way, fair enough. What would you like to see Valve do? The main reason I think we need a general format update is because we have 6 regions worth of qualifiers now. That's a bit crowded for only a 16 team field with direct invites don't you think? The only way it works with a 16 team field is either with only 4 direct invites or 10 direct invites.
|
Yeah I like your ideas in theory, but all this expansion probably won't happen and the format is confusing.
To merge theory into practice you could do 4 overarching invites and 2 qual spots from each region (EU, CIS, CN, SEA, NA, SA). Do EU and CN deserve more invites than SA? Sure. But if the goal is to grow the game then I agree, you should have 2 representatives from the region.
Also, what CS:GO tourneys do a full Swiss Group with 24 teams? I'm not seeing any on Liquipedia.
|
I don't think you need to invite a SA team just because there's a separate region there. I mean, it has been a rule to invite 1 team from each region but it's not the ideal. SEA has been lackluster lately, if they don't get good placements on dac i don't think they should get an invite either. I think the invites should be EG, OG, AF, China team (probably the best placement at DAC).
I should add that qualifiers spots could be: 2NA, 2EU, 2CN, 2 SEA, 1 CIS, 1 SA. so you can invite 6 teams. And you're done.
|
On January 24 2017 03:51 schmitty9800 wrote: Yeah I like your ideas in theory, but all this expansion probably won't happen and the format is confusing.
To merge theory into practice you could do 4 overarching invites and 2 qual spots from each region (EU, CIS, CN, SEA, NA, SA). Do EU and CN deserve more invites than SA? Sure. But if the goal is to grow the game then I agree, you should have 2 representatives from the region.
Also, what CS:GO tourneys do a full Swiss Group with 24 teams? I'm not seeing any on Liquipedia.
It's not swiss with 24, it's two swiss group stages with 16 teams in each. In effect, the top 8 from swiss group one (which includes all the qualifier teams in my scenario, and the bottom 8 from the previous major and 8 qualifier winners in CS:GO world) move on to a second swiss group with the 8 directly invited teams. The current major happening in Atlanta used this group stage format. I'll link it here for you:
http://wiki.teamliquid.net/counterstrike/ELEAGUE/2017/Major
I also don't see Valve restricting qualifier spots to favor regions at all (though I don't know how they can avoid it, even my example calls for it to some degree). They have had plenty of shots to reduce SEA qualifier spots forever and they didn't, and now SEA went from a joke region in past years to a region with plenty of good teams that can challenge all but the top 2 or so in the world at any given time.
|
On January 24 2017 03:55 Pontual wrote: I don't think you need to invite a SA team just because there's a separate region there. I mean, it has been a rule to invite 1 team from each region but it's not the ideal. SEA has been lackluster lately, if they don't get good placements on dac i don't think they should get an invite either. I think the invites should be EG, OG, AF, China team (probably the best placement at DAC).
I should add that qualifiers spots could be: 2NA, 2EU, 2CN, 2 SEA, 1 CIS, 1 SA. so you can invite 6 teams. And you're done.
You wanna leave out DC from the invites list after winning Genting? I know you just forgot them, so no big deal. But that changes your list.
The problem with 4 invites is that things get really, really messy. The idea of 6 is better though but still messy, and if Valve doesn't change the format in any tangible way then that's pretty much what they'll be forced to do until they come up with something else. I just don't see how they can keep the 16 team field without expansion as they add more qualifier regions.
Also, we only have one more big LAN to go, which is Starladder. DAC happens after the qualifiers and direct invites for Kiev have been announced. We already have 6 winners of different LANs this cycle, and that doesn't include the second place team from Boston in Ad Finem, so we're at 7 teams already unless you want to be really cruel to someone and drop them. If a team from no where comes out to win Starladder, like VG.J, then 6 invites isn't really enough anymore.
|
Very complicated and confusing group stage :\
I still hope for my single elimination format to be adopted one day. Gonna keep promoting this format. Group phase 4 groups of 4, double elimination style. Last place eliminated from tournament. Each group has its own 'grand final', rather than the winner of the winner's final automatically winning the group. Makes the group phase still important after the winner's final. First place gets seeded directly into RO8.
Main stage Games get played on main stage from here onwards. Single elimination. 2nd and 3rd place finishers face each other in a Bo5. Winners proceed to face the seeded teams. Each day will have 2 Bo5s, max 10 games. Takes 5 days to get our finalists.
12 -> 8 -> 4 -> 2
Grand finals on a fresh standalone day. Bo7 finals. Maximum 7 games. Ample time to do opening ceremonies, interviews, show matches etc.
Pros 1. Every team in the main stage are given a chance to prove themselves. If you lose 3 games to 'cheese' or a weaker team, you actually are not good enough. Probability of upsets are reduced.
2. Grand finals on a fresh day. Finalists have plenty of rest to prepare. Who cares if it is 7 games. Teams from loser bracket used to potentially play 8 games.
3. Each day on main stage has max 10 games. Plenty of time to account for delays. 4 Bo3 scheduling has always been irresponsible and reckless because days often end past midnight. Totally not accounting delays. See dotapit lol.
4. Each team plays only 1 opponent per day. Feels so weird when teams have to potentially play 2 sets in one day with certain formats/scheduling.
5. Group stage winner being top 8 and last place eliminated makes the group stage important.
Cons 1. 4 teams won't be on main stage. If a top team crashes out here, it's gonna be a pity.
2. Grand finalists probably will be hungry. The bo7 duration means breaks for meals are necessary, but how do you plan so it doesn't break momentum? Of course pre announced breaks would be good.
3. Days may end really early if it ends up with clean sweeps. I guess that's where mini events would be helpful.
|
Group stages are always too long with 16 teams, besides they're barely relevant. Making it even longer would kill it for me. Relying on valve's judgement on who to invite and how many slots they give to each region is unfortunate still the best way to go. I couldn't care less about teams making it to the grand stage or not but I understand it's still a must for fans coming to see a specific team. Even iirc all the crowds cheered well no matter the teams as long as it was some good doto.
The "region forcing" worked well for SEA to develop their scene but did not work for NA, I think valve needs to be careful to not give too many invites to certain regions. I don't care where the players come from so seeing average teams from weaker regions instead of stronger teams which did not make it through a cut throat qualifier irks me.
|
|
|
|