Made a video on the second channel discussing these types of videos: kzfaq.info/get/bejne/j7SKf7emz7e8XWg.html&ab_channel=Gbay100

Personally I agree with the guy at 2:15, he seems smart

At first something felt a bit out of place, but then he started moving his hands and then it started looking normal xD

I left feeling affirmed in my opinion that double elimination would be good but I respect your opinion and give you huge props for framing the opposing view fairly

a team upsetting a favorite and making the favorite do the losers bracket run is more enjoyable to me than them being gone and the team that upset them just getting smashed in the next series

While I can see merit to your arguments, having Worlds and TI (dota2 equivalent) at the same time this year really shifted me towards the double elimination camp. Not only does it allow for more games to watch for the viewers, it also allows for more series to be played in a day (worlds are boring as fuck with 1 series a day after groups) and more teams to advances at each stages, it also allows the meta to evolve more during the tournament and truly separates the good from the greats (TI was won by an underdog team that had to go through all the tournament favorites in the lower bracket to win it all).

To counteract your argument I feel like even with double elim, having an upset in BO5 does matter. Seeing G2 beat RNG in a bo5 matters even though RNG is not directly eliminated. It reshuffles expectations, changes narratives, brings enormous hype around a rematch, etc. Other esports showed it (just look at TI or smash bros) : an upset in winner's bracket is hype, a team or a player ranking higher than expected is hype. In fact here is another good thing about double elim : teams rankings are more accurate, which means that progressing through the bracket is important in itself, as you progress through rankings. And even though you might lose some hype for matches on winner's bracket, you get huge hype boost for loser's runs. You can see teams adapt, you can even see more upsets (for instance : MAD lose to DK but then proceed to have a good loser's run against other korean/chinese teams. I mean it is far more hype than what we got). Besides, it would enable to put 3rd seeds from groups in loser brackets, allowing for good comeback stories and even better loser runs. And obviously it increases the chance of a good grand final (you could even make it a bo7 with a game advantage for the team coming from winner's). I really feel like looking at other esports and domestic leagues proved that double elim doesn't create worthless matches or upsets, but just gives us more games, better matches and a clear cut ranking of best teams in the tournament.

I think the higher chance of a "upset" in Single-Elims does not not outweight how boring the average best of 5s in the playoffs will be. For every G2 vs RNG upset, there are 5 other 3-0 stomps to slog through. If i want to watch a competitive game where the outcome can go either team, it is much better to have two teams that are close in strength and calibre. This is especially more important for the health of the sport the closer it gets to the grand finals and the closer it gets more attention from future fans of the game. If i wanted to introduce a future fan of competitive fan of LOL to the scene, get him immersed in the energy and the passion of the clash and the conflicts, none of the Worlds Finals would probably make the list. And I think this is major flaw of Worlds. Plus there is a recent double-elim tournament of a competing MOBA that runs counter to your arguments: The recent Dota 2 Internationals where a bunch of young rookies just had a major upset against the tournament favourites in a thrilling 5 game finals. The winning team Team Spirit was filled with 4 young rookies who had just started playing competitively last year and only one other veteran. Double-elim removes any ambiguity? Well, the Lower Bracket of Double-Elim is basically a Single-Elim tournament and you don't even get the trophy if you clear the Lower Bracket. For Team Spirit a team of young rookies, their paths to the Grand Finals consisted of having to defeat the previous champtions(OG), their regional rivals who had an overwhelming win record against them before the tournament (VP), the 2nd best Chinese Dota team that had sent them to the Lower Bracket in the first place (IG) and the best EU Dota team (Secret). Each of these teams had players who had played more years than the 4 rookies combined and had won major championships before. You would have bet on any of these teams to win TI before you would have bet on Team Spirit and it would have been a logical bet. And Yet Team Spirit delivered major upset after major upset in a row in the Lower Bracket. This was G2 vs RNG repeated 4 times here. And of course the Grand Finals was a thrilling match where LGD were highly favoured to win while TS was supposed to go home proud but defeated. Yet another major upset happened and the match would be one for the history books. I will rewatch the TI 10 grand finals or even the best TI finals, TI 8 before i will even touch the Worlds Finals

That IG vs KT throwback was exactly how everyone felt during the DK vs T1 semis.

While I can definitely see the reasoning behind double elimination making things more predictable (G2 beat RNG in 2018 to surprise everyone but could they have done that again if they faced RNG in grands?) but at the same time a lot of pressure is put on one single set. The purpose of double elimination is to truly solidify who is the better team and who just got lucky or had a good breakfast the morning of. But yeah unlike traditional sports, or even other esports like smash bros, League is predicated on snowballing a lead which is why games feel so hopeless after one team gets first blood or a good teamfight.

I understand where you're coming from, but the issue I have with your arguments is that there really isn't that much hard evidence to back up your worries about double elimination. People bring up Dota's TI because not only has it used double elimination for the past 10 years, but each and every single one of your concerns simply never seems to rise up there. This past TI - TI10, had Team Spirit win, a squad of Russian Zoomers nobody has ever heard before - one which has the name TORONTOTOKYO for god's sake, 4 of which have never been to TI before, and they won decidedly after losing 0-2 to Invictus Gaming in the Upper Bracket Quarter finals, and marching their way to the Grand Finals through the lower bracket, knocking a whole slew of teams favoured to win in the process, and getting that 3-2 win against LGD - overall playing 16 games in the Main Brakcet. Their playstyle was solid. Collapse relied on an absolute godlike Magnus, and their carry player Yotaro had an immense hero pool (he played 13 different heroes in those 16 games, and he only started repeating heroes during the Grand Finals). To call this an upset would be an understatement. Each year, Dota allows players to make predictions about game stats, as well as make a fantasy bracket after group stages for how they think the bracket will play out, and only 1.6% of people predicted that Team Spirit would win. This isn't isolated to TI10 either. TI8 was won by OG, who had an extremely rocky season, which left with most of the members leaving the team 6 weeks before TI, and n0tail - the team captain, having to basically rebuild the entire team from scraps - ending up with a (at the time considered to be) Tier 2/3 Offlaner, and a midlaner who never played in a major tournament before let alone TI, and was just someone who was good on the ranked ladder. Won TI through the lower bracket. Same thing arguably in TI5, TI2, and TI1. TI6 had an unknown team win, but going as far back as the group stages they were proving themselves to be extremely dominant and were favoured to win after their performance during groups, and even then the 2nd place team was an even larger dark horse team nobody expected. In short, at least in Dota, this whole idea that Double Elimination results in predictable results and reduces the chances of major upsets simply isn't the case, and in the case of Dark Horses, often makes their story lines better since doing a massive lower bracket run requires you to go up against so many teams that are considered to be favourites that it solidifies your position as a grand finalist, rather than ending up there because you happened to get the easier half of the bracket. The point I'm trying to make is that there are many different esports out there that use double elimination brackets for their major events, and so if we want to theorize about how a certain format would play out, well the proof is in the pudding so to speak. Its pointless to theorize about possible consequences of certain formats when there is so much hard evidence that those consequences either are extremely small or non-existent. Let's do a little thought experiment - let's imagine a world where League had the exact same format as TI. To recap, Dota has Bo2 Round Robin Group Stage with 2 groups of 9, and the groups are seeded as follows: Bottom team is out of the tournament, top 4 teams get seeded into the upper bracket, and teams 5-8 get seeded into the lower bracket (for simplicity's sake, we'll pretend that Worlds will have 2 groups of 8). The playoffs are a standard Double Elimination bracket where each match is Bo3, except for the grand finals which is Bo5, and the first round of the lower bracket which is Bo1. Let's start with the Group Stages. In this scenerio no one gets eliminated after groups, and instead the group stage becomes a fight for who gets that "safety net", and who gets to fight the long way up. Also since its a Bo2 for 2 groups of 8, each team will play at least 14 matches within 4 days against 7 other teams, plus maybe some tie breaker matches if it comes down to that point. This immediately makes one massive change, since the risk of elimination is no longer present, teams now have a lot more room to experiment and try out some of their own strats - WHICH IS A HUGE DEAL. One of the biggest problems with Worlds and LolEsports in general is that when western teams come to worlds, they drop doing whatever they're good at, and start trying to replicate what the Koreans are doing. This is because with so few games at their disposal, the margin for error at worlds is so small and as such the safest option is to try and replicate what the dominant region is doing, which could result in forcing playstyles that do not suit the players. In this scenerio, each team has 14 games to try out different strategies and see what works for them, and if they see that their NA branded strats works well against the Korean meta strats, well now they can plan out a path on how they can play the tournament their own way, since they have a testing ground (which is exactly what happens at TI, the Group Stages is often a place of experimentations where the meta for the TI is discovered, and ends up setting the tone for the main bracket). This isn't to say that group stages becomes a pointless waste of time, sure no one gets eliminated, but whether or you not you make it to the Upper Bracket is still a huge deal. Not only do you get to play the rest of the tournament with a safety net, but you also don't have to deal with the Bo1 Round of Death, which as you pointed out, Bo1s are unpredictable and can lead to crazy upsets, and as a team you want to avoid that as much as possible, but still its not elimination so the room for experimentation is present. Now that Group Stages are over, we go to the double elmination bracket. The Bo1s occur and they're a bloodbath, and we're onto the Top 12. The Upper Bracket is likely filled with Korean and Chinese teams, and inevitable some good teams will likely fall to the lower bracket after the quarterfinals. Here's where the next change happens, that Korean/Chinese team is now face to face with a Top 12 team that they likely haven't went up against in Groups, and now they have to deal with a massive problem. When you're a massive Top 3 team, all eyes are on you, and everyone is studying how you play, and are carefully looking for any mistakes in your gameplay or weaknesses to exploit, meanwhile if you're a Top 12 team, nobody really knows that much about you, and as such you have a lot more room to hide your weaknesses, or to hold your cards close to the chest. As the loser bracket team, you have already played 15 matches at this tournament, so at this point you have a pretty good sense of what works and what doesn't, and this allows you to have secret strats that you can use to surprise the favoured team. All of this is precisely the secret ingredient that allows TI to have so many dark horses, and the reason why despite having double elimination, TIs consistently produce far more upsets and exciting games compared to worlds. The final thing I want to bring up is viewership stats. TI10 broke record with 2.7 million peak viewers, compared to worlds 2020 which had around 3.8 million viewers, at least in the west (China is excluded from these counts because Chinese data is extremely unreliable - Chinese streaming services are notorious for inflating their viewcounts and as such any numbers that come from that region should be taken with a grain of salt). Sure Worlds still has more viewers, but League is also a much bigger game, 10x bigger by some estimates especially in NA and Western Europe, and yet in the best case scenerio they can only get double the viewership for their world championships compared to Dota. Dota is far better at enticing its own playerbase to watch and consume esports content, and the format is likely the reason why. By having so many opportunities for upsets which results in not only more international competition but crazy games where you don't know who's going to win, people get far more invested, and are far more likely to tune in to watch the crazy action unfold, become a fan for a specific team, and overall tune in to watch TI each and every single team. Sorry for the long essay.

More group stage games, best of 16 playoffs (so maybe more overall teams going to worlds) would be the play imo. Its a fine line between enough international competition and too much. In CSGO before covid it was obviously too much. Teams would play 4 big events a month and would travel more than play. Very good teams with the luxury of skipping events would often only go to one event a month. But in League its too little imo. If MSI was comparable to worlds, witg all the top teams from every region, it would be fine. But MSI is very very limited and then all the pressure is on worlds. Worlds being at the end of the year, in a completely different meta than MSI doesnt help either. Players are already pretty tired and the meta just means too much if the entire year is decided on the very last pro patch

Your argument is based on the premise that having a more predictable set of teams reach deep stages of a tournament means reducing ambiguity. It is important that when you are making an argument you explain your position -- and while you explained your logic (and it makes sence), the premise you are basing your argument on the other hand is not elaborated upon. And I don't agree with the premise, hence I will try to explain why I believe it to be wrong. First of all, the basic reason for wanting predictable ("strongest") teams to go further in the tournament is the fact, that having stronger teams makes results less predictable. As you have correctly pointed out, viewers want the result to not be obvious, and thus we want to provide the best possible opposition to the best teams to avoid predictable stomps. Secondly, double elimination increases the volume of games, and having higher game volume increases variance in the bracket thus increasing ambiguity. While it makes it more certain that the favorites are going to go further, the path they take can be very different based on outcomes, making the tournament more exciting. Thirdly, double elimination by its nature matches teams that are close in skill in the lower bracket, and as a result matches for tournament life have higher stakes and in general are less predictable. As a result, introducing double-elimination bracket may result in more ambiguity. And if it does so (and I believe it does) it pretty much voids your entire argument. So as you can see, even though I agree with the fact that we need more ambiguity and interesting story lines, I do believe that the premise you are using for this argument is wrong and therefore I arrive to the opposite conclusion to yours. Tbh, you may disagree with my point, but the most important takeaway from my reply shouldn't be that I am right, but rather that the argument in the video is poorly constructed, since it's fully based on an arbitrary point that may be wrong.

This is the classic and eternal struggle between the two sides of competition. Fairness, and entertainment. On one side, sport is an extremely unique and amazing form of entertainment, but on the other side, the entire point is to figure out who is best. The single-double elim argument is the quintessential example. And there is no answer, as always, it's just which lesser evil you prefer. Also I found it immensly hilarious how one of your two reasons why people want double elim is the two best teams facing off in grand finals, and your main reason for why single elim is good is because that often does not happen. Doesn't that just perfectly encapsulate the entire argument? One side wants x, and the other wants all but x. imo single is fine, but one difference I would make is make group stage matches BO3. Not only would that increase international competition AT LEAST TWOFOLD, but it would also create more accurate seedings going into the bracket, making the two best teams more likely to end up at opposite sides of the bracket, but still allowing underdogs to shine due to the same amount of matches.

I'd argue that even making the tournament double elimination wouldn't take away the upsets, though making them less important in the grand scheme of the tournament would also make them less exciting.

How about a middle-ground approach for both single and double elim fans? The format being two groups of 8 teams playing bo3 double elims. And then two teams get out of each group, so from top 4 its bo5 single elim.

quite frankly, this was one of the most interesting arguments i recently heard. and comment section just makes it better. thanks for such great and high quality content. keep it up. wish you well

Given that almost all people are invested in 1-2 regions at most, I think they watch Worlds to see the best LoL there is. I will be excited if the team that the casters assure me are underdogs win but if the quality of play falls off after that it hurts the entertainment. Also, for me at least, I kinda want the "certainty" that at the end of the year I will know who the best team really was, it seems fair. If they are by far the best then they will not lose and they will not go to loser's bracket, but if they lose once then I will gladly watch the exciting and higher quality comeback run to win it all

While I see your point I am still in favor of double elims, having said that, more would need to change to make it work. I think dota has a really good structure for the internationals. Many times the winner came from a miracle run of the losers brackets. I think remove restrictions on regions and just get a set of the best teams to worlds. Put them in two large groups and seed them into upper and lower brackets. You get more games and it's more competitive.

Always nice seeing you posted

I mean, EDG came from the losing bracket (LPL use double elim), won summer LPL and Worlds 2021. I think doing double elim in regional league is important because we all want to see the best teams from each region compete, but I agree with you the current Worlds match is nice and is probably the only way we could see LCS/LEC be in the final.

Great video! Love your stuff Gbay. I will say that for part 3, the examples you give about previous Worlds' upsets, if it were double elimination, we would get to see matchups like TOP/JDG vs DRX/G2 in 2020, or KT vs RNG / C9 vs G2 in 2018 or something like this. There would still be further upset potential. Double elim doesn't guarantee the best teams make it to top 3 or top 4, just like Single elim doesn't. Sure it makes it more likely, but the same stories that develop from a Single elim bracket are equally as likely to occur in a loser's bracket due to just the percentages certain teams have to win. Double elim adds so much more for the viewer (more international best ofs) while still not taking away from crazy upsets that happen. It still would've been insane that G2 beat RNG in 2018 even if RNG ran the loser's bracket all the way to finals.

I love that you made it clear, and while I staunchly loyal towards double elimination I appreciate the fact that having it does come with some amount of devaluing of some series. I would love to see a version of the LJL Playoffs be adapted for worlds, meaning that first in groups puts you into the top bracket, first in each of the groups would play one another in a BO5, losers would drop down to play the winners of 2nd place in each of the groups. Yes that does mean a team will make it to finals only playing two BO5, but personally, if they beat the other top teams, don't they deserve that? While at the same time a 2nd place team to make it all the way to finals would have to play four BO5, its a reward for doing well in the group stage.

I absolutly love documentary Gbay. Really found your niche. Always looking forward to the next through provoking video.

Love these sit down videos

People have been complaining about how championships are crowned in almost all sports for as long as I can remember. There will always be complaining.

blessed the day is - new gbay video

Hi, Gbay. Keep the good work, buddy.

Yeah I pretty much agree with everything here. I still feel like inter-regional play is a problem but I don't think double elimination is the best way to go about it. That said, I find interesting your comments about LoL as a sport, so I'm gonna talk about that hehe. For starters, it's true that traditional sports like Football, Baseball, Soccer don't reward you for the advantages you get, and that makes it easier for the losing team to make a comeback. But I've always thought of League more similar to Boxing or MMA, where if you come to the ring prepared and you win the first round (laning phase) pretty convincingly, you have a noticeable advantage over your opponent from that moment forward in the fight, because they are hurt. What I find interesting about League is that, on top of that, there's no time limit or points limit (like in Volleyball), in League you HAVE to close the game, and no matter how far ahead you are, you HAVE to destroy their nexus. Even if your opponent is 15k behind, if they stall for long enough and get 1 good teamfight, they will take down your nexus and win, and nothing of what happened the previous 40 minutes matter. My favorite kind of League games are the ones where the underdog gets a big advantage in the first 10 minutes, because then the question is "can they close?". It's cool because, opposite to traditional sports, if the underdog gets the advantage early they actually get stronger than their opponent and they get a real chance at winning (in soccer if the underdog gets a lucky goal at the start of the game, they will most likely lose because they would've lost 0-4 anyways, now they lose 1-4) BUT at the same time the underdogs have to prove they have what it takes to close a game, and that's not easy against the favorites. Say what you want about Worlds finals but I prefer that over something like the UEFA Champions League finals between Liverpool and Tottenham a couple years ago, where Liverpool found a goal early and then they just defended for the remaining 80 minutes of the game. We all know that one SKT comeback is one of the most iconic games in League history, and this Worlds we already had something similar in game 2 between DK and MAD. The problem is people usually see those games as "throws" but man, I love those kinds of games and I wish underdogs focused more on winning early game instead of just letting better teams slowly choke them out of the game...

You make the Best videos man. Keep them coming!!

Always good to hear Gbay's perspective

It all depends on what the goal of the World Championship should be. Should we be focused on crowning the best team in the world? Should our focus be on more international play? Should there be more games? Should there be less games? The issue is that the Eastern Teams are just so much better than the Western Teams. As Western fans we want to see more opportunities for our Teams to continue to play and possibly make it out. We think that double elimination will offer that, but in reality it will stack the final days of the tournament in the Eastern Teams favor. The only reason groups seem competitive is because they are BO1 matches. If we want to make Worlds more entertaining then we would need to make it an entire BO1 single elimination tournament.

The main thing I have with MOBA double elims is the fact there isn't a chance for a bracket reset. The losers bracket team has a second life while the winners bracket team's only advantage is side selection, not even through the series. In FGC/Smash and pre-CDL what made every event and series worth watching was because you always felt like the right team was gonna win because everyone got 2 chances. Losers bracket runs are rare but immensely exciting. So much so that even now I go and see if I can find which team made the longest losers run through any given tourney.

I think you did a fantastic job of presenting both sides of the argument without any bias, great video gbay.

Single elim would be fine if we had more international events each year, but worlds and msi are the only two events we get to see all the teams from across the world compete, and so for us to only get more than one bo5 from 4 teams in the entire world feels pretty awful.

Always fun to hear an explanation from the contrarian side when it’s put together well. Great vid Gbay

Nah I think double elimination is still good regardless of random upsets Which is cool but even if they should matter they should solidify they are the best by doing it twice

I doubt anything will be more exciting than the group D day with the 4 way tie between the big 4 regions. Was such and exciting and fun day. I was pretty excited about the knockout stage but 3 out 4 3-0's is just not exciting or that fun to watch.

You're back sooner than last time. Cool.

As someone who loves international football (soccer), the excitement of the World Cup is what makes me happy. Its format is arguably one of the worst systems for showing a true deserving winner (winner only ever gets to play 7 games). The upsets of a underdog beating a giant/favorite in the bracket is what makes me excited to come back when i'm not cheering for the country I cheer for. If we removed single elimination brackets at that point with a double elimination bracket these upsets would not be as rewarding, both in league and in the World Cup. I have some issues with the play in stage, but other than that i don't have any issue with the bracket stage. Loved the video!

Here's my fix: -extend to 32 teams -remove play-ins -8 groups of 4 and a round of 16 -No teams from same region can play in the round of 16 -A coefficient based of (for example) regions performance over the past 3 years and that team's performance that year -4 seeds for group draw ( based on that coefficent) But for the MSI i feel like thats where double-elim makes sence since its only winners from spring split so removing these kinds of upsets doesnt really change much. So just by this we go from 10 international bo5's to 21 by not really changing the structure that much, where at MSI we would actually have the finals between the number 1 and 2 team, while at worlds we would have a chance for much more upsets and would be much more exciting. And one more thing for worlds, have bo5's be played more freqently, for example max 2-3 days beteween stages (qf to sf to finals), like in where else do we see QF's,SF's and Finals be played only on seperate weekends with 5-6 days between them.

tl;dw - The reason I'm against a straight double-elim bracket is I don't like the thought of sacrificing the excitement of every non-grand finals series just to have a more exciting grand finals series. I'd rather each series matter and see less competitive matches, than only have grand finals matter but get to see a few more competitive games. League already has a problem with matches being really predictable with few upsets ever occurring. We need the upsets that DO occur to matter, otherwise we'll know who's going to win before Worlds even begins - why bother watching the games? In retrospect I realize I didn't state that super clearly in the video. Hope you guys still enjoyed the discussion! New doc next month.

My favorite double elim tournament was EVO 2017 Street Fighter 5 top 8. Tokido's run from Loser's side was genuinely amazing to watch.

A possibility could be 1/8 finals instead of groups that means 16 teams has to compete in best of fives which means you will probably see more international play, there is two ways to do this, you can ditch groups all togheter and just start the 1/8 final (where you normally would start groups), or you can expand the tournament, and make it bigger, 8 groups, big event more seeds and more mixed comepetition. 8 groups, top two in the groups win proceeds to 1/8 final and the mixed competition starts, I know it will take more time and it will be harder on players, but if they make a higher price pool, and have worlds for 2-3 weeks longer i think it's doable.

The big problem of Double Elim is that loser teams are kind of benefiting from having more games than in the winners bracket.

Totally agree; bracket worlds isn't 'who is the best team now or next week', its now or never. I think the solution is to make groups bigger (2 groups of 8 or something), then 1 plays 4, etc.

I feel like, if I were just in a room somewhere and heard GBay talking, I would immediately know it was him. And I’d probably go listen to whatever he has to say. So much good information all the time

To me, doing a double elim will feel like bringing back Super Weeks from the days of EULCS and NALCS. Taxing as a whole-players, officials, and broadcast team.

Damn good points for sure, the thrill of single elim where that series you play is the most important series of the year or else you're out is mad true - good shit gbay :D

Interesting video as always Gbay! You've made some good points but now i'm only conflicted with my opinion.

so cool that Scout is now playing in the World Finals, after you made a video on him

Love you Gbay!! I'm early wooot! You're keeping me awake though.. Smh

As a fan of watching both League and Smash Bros, which uses double elimination, I think League should use a kind of modified version of a double elim, here's my argument: Watching the favorite to win make a losers bracket run is more thrilling than watching them just win every game since that was expected, but now they've shown they bleed, they can get knocked out. I think that more teams need to make it to bracket, maybe that means more teams attend worlds and there's more groups. But also, they should be best of 3's until Semi's, which would now be Winner's and Loser's final, which would be the first Bo5s. This way, being underprepared is still costly and it doesn't take all day for every series. I think that winning the winner's bracket should still give you an edge, that way teams don't lose intentionally to try to get a better bracket spot, but a full double-loser's best of 5 would take twice as long and you can't even guarantee it would happen, it would be a scheduling nightmare. Perhaps a best of 7 and start the winners off with a free game AND first side choice (you could also call this what it more would be, best of 6, with tie going to winner) so that way they still only need 3 games like they do now, but the losers need to earn back their win and need a 4th game or the other team can still win from winner's. So my hypothetical bracket would probably look 50% larger, be best of 3's until there's only 3 teams left (instead of 4, and no actual 3rd place), and comebacks and loser's runs could add more thrills. This might be a better format for something like MSI, but it still keeps a lot of what's good about the current system, the fact that it doesn't have twice as many total games, and the risk of losing at any time is a major setback (except now it's twice as much work vs complete elimination).

I think that the value we see in the actual best game that can be offered in a tournament far outweighs the excitement from an upset win eliminating a tournament favourite leaving the later rounds feeling underwhelming. A single elimination tournament basically manufactures short term underdog wins, while in a double elimination bracket a true underdog win would be more rare but would be much more meaningful and exciting.

First of all, riot should unify season format for all regions, at least the major ones. So either teams Play best of one in regular season or bo3. That would change a lot

Honestly, two groups of 8 is much much better than double elim. Other esports do double eliminations and have crazy...CRAZY runs (e.g. Team Spirit in the recent TI10, Tekken EVO Japan 2019, both Tekken World Tour) but most of the time and especially in FGC, they usually favor the best team winning the championship. Two groups of 8 (probably just single round robin is okay) make it that EVERY game matters. It's always been a problem in which that some teams lose more games than other teams but still make it out (e.g. C9 this Worlds) and honestly a 4th seed from the harder group could still surprise people.

There's a lot of good arguments that you pointed out but I think that you really didn't put enough thought about the storytelling side of the question : Storyline wise - in esport at least - there's nothing better than a loser run, nothing more thrilling than seeing a favorite dropping early into a loser bracket and needing to overcome 4-5 matches to reach the Grand finals where it needs to overcome a double difficulty against a team that has performed super well. On the other side when you have an underdog, also dropping early and everyone expects him to just lose instantly in loser bracket, they kick out opponent after opponent and perhaps win it all later on (I'm talking about the Spirit in TI run). But in order to have good storylines you need longer loser's bracket, because a 8-team double elim bracket would just make things longer and no necessarily more thrilling, we mostly need more teams at worlds, a bigger playoffs bracket (which should not be composed of China and Korea mostly, please riot), and then a loser bracket would be relevant

Banger as always

Yeah nah EDIT: Ok, there's some good points but I wanna see coinflip teams have a second chance but I also don't want to see a team's win get invalidated because they didn't really eliminate their opponent. Now I'm conflicted lmao

I compleatly agree with your points. While double elimination would probably result in a "more-skilled-winner" the format cannot match the exitement that single elim brings. Even for a team that sweeps everyone, it would be so much more boring if we knew that if an upset were to happen they were still not disqualified. Stories like the DRX-Run would probably not have happend and who knows if T1 could have beaten JDG (the likely winner of a potential loosers bracket at Worlds 23) a second time. Also, I feel that it es fine at MSI becuase the two big international Events are in a way set up to be "the competitiv one" and "the sport one"; as only the top Teams get to attend MSI and the top four (in major regions) have a seat. Thank you for making this video!

What would be cool I think would be a single 16 team bracket, where every team plays every team twice for seating. than a 16 team single elim knockout.

League is adding more comeback mechanics, for example the new Chemtech dragon dealing more damage when your HP is lower than your opponents; more items which allow for more item nuance and skill expression which could favor the losing, but more skilled teams; and objective bounties. Maybe when league improves this issue of overly predictable bo5's and game states, a double elim bracket would then add excitement and not "Take out what ambiguity is left". Great video. Fingers crossed for progress for worlds 2022.

Oh boy spicy topic

Fun thing is, exact arguments you give against double eliminations are arguments I'd use for format update (with double elim being one option). It does reduce results variance, helps dealing with League's problem of team relative strength not being transient (just because X beats Y and Y beats Z doesn't mean X beats Z). Essentially: trades away upset potential for more technical and - for a lack of better word - stable tournament that gets relative ranking much closer to what we'd expect if all possible combinations played out, and puts a lot more attention on teams that are have steady, reliable performance while pushing upsets and volatile strategies away from relevance. Except, I don't think double elim is that good at providing such outcome - much better system would be to replace semis and finals with everyone-plays-everyone group of BO5s. This way current group stage still matters (quarterfinals are still single elimination, so getting #1 in groups puts you at advantage), quarterfinal winners proceed to play against each other in total of 6 games - so less than what double elim would have, and with much less skill mismatch. The only draw possibility is 3 teams getting transient 2-1 with one getting 3-0 (A > B > C > A) that would have to be solved. Bonus: there's no chance for another boring finals, since there is no finals match.

Am I crazy for thinking DWG vs. SN was a great grand finals? First 3 games were super competitive plus the first and only penta of grand finals ever. I put it as my third favorite behind EDG vs. DK and SKT vs. SSG.

Does anyone know the music that was being played in the background?

IMO with how often a Worlds playoff series ends in 3-0 or 3-1 for the team that came in favourite to win the series, makes me lean towards double elim instead of having the possibility for an upset. I'd rather see some interesting matches we wouldn't otherwise see. Also with the current format, the best matches quite often happen in the semi final and then the final is pretty lackluster (what will happen again from the looks of it in 2021)

All I know is when it’s grand finals in a fighting game tournament, and the bracket gets reset; there is nothing more hype than what ensues..

Honestly, it really comes down to if you think a tournament should determine the best team or if it should create the most exciting viewing experience.

I would still love to see Riot TRY a Worlds with Double Elim to see what happens, ovb easier said than done as you have to book more venues etc but would still like to see an experimental Worlds year

Keep it up bro

I've seen double elimination in fighting game with people with major upsets happening all the time and sometimes watching a player from the loser brackets fighting all the way through the loser brackets is entertaining in itself because you can't help but root for that underdog when normally you wouldn't even care. For being boring to watch if its double elimination it actually has the opposite effect of boring when you think about it just in a different spot. With single elimination you cant put all your money on a third rate strategy, no-one does that. The best strategy for single elimination is to get the most consistent, safe, and simple playstyle. Minimizing risk becomes the only way to play in a tournament. A very reserved, slow playstyle is what you will see when a single mistake can cost everything. With Double elimination you may see all sort tricks or hellmarys from a team or may even witness the true power of a team before they were eliminated or something really unique. Although the downside is it can double the amount of time needed to finish the brackets.

I think something people often overlook in this conversation is that adding double elimination bracket adds more game days and therefore adds more work for the casters, hosts, producers and all the other staff members. Because Worlds usually change cities or even countries during the tournament (like Paris/London/Brussels/Berlin in 2015) those empty days are used for travelling. Having DEB would either ground the tournament to one place or make it last longer - I think there are reasonable arguments for both of these to be a bad idea. Having more than one Bo5/day is also problematic since then the length of the day can be anything between 6 to 10 games plus breaks. Imagine buying tickets for the second Quarterfinals on one day: because the first QF was a 3-2 banger your match starts two hours later than you anticipated. Then it preceeds to be a 5 game series with a big technical pause and a remake in between and you are home 2-3 hours later than you first thought. In my opinion one match/day makes scheduling easier and it helps both the production team to focus on one match up and it's narrative and presentation and fans to make plans if needed to see that exact match they want. The lack of matchups between regions is interesting because I think it's hard to base the arguments for this change solely on this. The bracket is dependant on the results of the group stage. If the group stage eliminates the majority of the western teams, well, you are left with teams from China and South Korea. For me it feels hard to argue for DEB just because "x didn't play against y" or "x played against y too soon" - it's lottery, and therefore equal for all. Making changes to it just for adding more interregional play would sound like "we want this to be harder for you". There are many fans who don't have time to watch Worlds every single day for hours - I would like to think that keeping the bracket stage simple and easy to follow with every match mattering makes it more enjoyable for the casual side of viewers: they see a match, and only winner stays in the tournament.

There are definitely issues with the current format, and I would like to see much, much more international play, but I don't think double elimination is the quick fix that many seem to think. I wouldn't say G2 reaching finals in 2019 was much of an upset though, they were among the top favourites going into the tournament and had to face off against two other favourites in DWG and SKT to get there. But, even if G2 were a favourite in 2019, watching the upset in 2018 felt much more ground breaking than the series in 2018. I don't think that series would have had nearly the same impact in a double elimination format.

Actually after world finals lpl loves double elimination system, because it reveals the real best team:EDG RNG and eliminate the fake king FPX. RNG went to loser bracket and won spring final against FPX and later won MSI. EDG also went to loser bracket and won summer final against FPX and later won worlds. DK is always the biggest enemy of lpl, glad to see double elimination choose the best to face DK and win.

In my opinion, If we are not going to get double elim's at worlds, we NEED more international events. As it is right now, between RR, MSI, and Worlds, theres a combined total of less than 20 Bo5's, we need more than at most, 100 games of international competition. Its also incredibly lame seeing an all Korea semifinals, or all lpl quarters. You already know what the end result of EDG vs LNG is going to be, or T1 Hanwha. We have seen those matchups 10000 times. I think the biggest problem is needing a new format for tournaments throughout the year.

If I can change worlds: 1. Remove playins what’s the point? 2. Instead of 4 groups why not just one big group. That we see all types of games and match ups not to mention more games since every team will play against each other twice 3. After group stage, all teams are seeded according to their rankings; top teams get seeded deeper in brackets etc. this way top teams are at least rewarded for playing well This way, we don’t need double elimination

IPL 5's Format looks insane, how did they came up with this and how was the feedback back then? Was it good to watch or nah? Bracket Phase looks huge

I think double elimination, as a debate, is good for pointing out the problems of surprise in sport that you're talking about, while also being kind of orthogonal to the problem. A double elimination that managed to have upsets and surprises would be good, a predictable single elimination would be bad. It's only about double elimination, to the degree that it's a proxy for things that people actually care about. I would argue that a huge problem of the format is that individual matches are generally not balanced. I think the solution is team autonomy. You have to find another team that wants to face you, and then it's "loser goes home." This means matches where each team thinks they have a chance, or cannot estimate their chances. In such an environment, a "one extra life" model would allow more games and comebacks without decreasing the tension much. From there, it's just setting up the details of matchmaking (can literally be done on slips of paper) and providing an incentive not to "turtle" by not playing matches (eliminate the team with the fewest wins at the end of the day, picking at random if there's a tie).

I'd like for them to try double elim at MSI and see how it goes

My problem with double elim is a bit different. If a team loses a bo5 and then wins via the losers bracket, the win is less valuable than that of a team that wins without ever being in the losers bracket. People will say, "that team would have been knocked out in a reeeal tournament". The team that beat the winners initially would be quite reasonably afforded similar prestige by having beaten them in a worlds bo5 in the same tournament with the same players and patch. In the end no team really feels like they properly won the tournament compared to the hardcore days of old (now). Winning worlds should mean being the only team to not lose in a bo5. No half-win bullshit. All or nothing.

Damnit gbay! I came into the video thinking you were straight up wrong but I end the video agreeing with you

My only problem with this whole video is that you kinda miss the point that upsets can still happen in double elim and just because a team won one best of 5, doesnt mean they will win the second, if you look at double elimination usually the upper bracket team wins about 60% of the time, meaning we usually see an upset either in the upper bracket or in grand finals. There is a reason double elimination is so loved among so many people

I wish the group stage for workds 2022 will look something more like what they are doing for Valorant Champions

What I'd like to see if them add rules in the quarterfinals that prevent the same region playing each other if possible.

Imo the Worlds format is fine, but the problem is it's so short and you have to wait an entire year to see it again. That's a huge problem for me. If they had an international tournament equivalent to Worlds 3 or 4 times a year, and we would crown a world champion every 4 months or so, that would be perfect with the existing format

I totally agree with you and the Chinese commenters. There shouldn't be any second chances when it comes to winning a title. Worlds and international competition needs to be fixed differently. They were discussing some great ideas on Bevies with the boys when Grabbz and Peter Dun were on the other day.

Hear me out here. Why don't we make the whole tournament a giant swiss format, where teams get "x" games each, without brackets or anything, just plain swiss, and all the temas get to play against as much of the other teams as possible. With this format, there's always hope for a team to comeback and get a championship even if their start was a bit rough, and at the end, the team with the most points wins or something. Adding to this idea, we should think of a way so things don't get like: oh yeah, after "y" games you can drop because you cannot win from here. Something that happens a lot in tournaments with this format, at least in my cardgame experience.

Honestly, in my opinion, they should keep it single elimination, but decrease the number of seeds from each region. There should be a MAX of three teams for the two best regions (if that is really what Riot likes, rewarding the top two regions) with the other two major regions getting 2 seeds and the rest of the spots going to the top emerging regions. This doesn't even need to change play-ins, because you can put the lowest seed of major regions (other than maybe the top region) in play-ins and give some of the top emerging regions a second seed as well. If we want to see more international competition, we need more regions participating and not just LCK, LCS, LEC, and LPL. We need to lower the number of seeds from the big regions so that we DO get international play - much like with MSI when only 1 team from each region goes. Four LCK and 4 LPL teams (or even 4 LEC when they were a top 2) is ridiculous.

I am a big fan of double elimination but this worlds basicaly showed me why format like you have in dota2 TI (which is actualy perfect fot this tournament and makes it ten times more exciting than worlds for me to watch) could actualy make worlds even worse for non korean or non chinese fans. Because the upper bracket would be korean and chinese teams only and the rest of the world would be elimininated early in lower bracket where they would end up, since seeding is based on group stage which is two big groups playing each other and teams with most wins coming to upper bracket.. It works for TI because so far you had west and east taking turns at winning it, so they are more competitive, but at league it would probably mean that biggest part of the tournament would be China and Korea only. What would help for EU and NA (and other regions) to catch up to China and Korea would be more international competition but also hard and costly to organize within standard season. Double eliminitaion is nice for people who want to watch some quality LoL being played, but for fans rooting for their individula region, it would be terrible if their region is not currently best or second best.

Not having double elimination makes League more like WWE than UFC. Now, I don't know about you, but the better team should always win. Simply put, the current format leaves a lot of questions up in the air if an upset was legit. Thus, not giving us a definitive answer if they were the better team...

I think both formats are good, you are right about it getting a bit more exciting in single elims, but I would also like to have double. 1. More games to watch, worlds is such a cool and fun time of the year and just more games would be really fun to see. 2. Like you said, there is a higher chance of the actual two best teams in the world to meet in the grand final, which I think makes sence. 2021 for example, I think the two best teams are DK and T1. And they’re meeting in the semifinal. Meanwhile the two other teams, in the other semi, are EDG and GenG who I don’t think stands a chance to either of these teams, of course there are the chance of an upset, but against DK or T1 that feels very unlikely. PS. You mentioned G2 in 2019, and this might be my EU bias talking, but I don’t think they got as far as they did because of single elim, I genuinly think they were the second best team in the world, the meta also suited them perfectly, with the Qiyanas and the Xayahs. I don’t know, maybe I’m biasaed but that’s my take

As someone who isnt that big into esports and will rarely watch all the games of a series, I think adding even more games will just make each one less exciting and also would make the worlds take alot longer.

1. If double elimination were implemented..... you can get used to watching NA getting shit on twice. 2. Implementation of a double elimination is essentially improving the grand finals at the expense of the rest of the bracket.

I want to see only 2 games of a series per day. If its a stomp the losing team has a day to look at their flaws. Most of the time it is easier to improve when you got beaten (with a possibility to rewatch the game) than it is for the winning team to improve. The viewer expierience would also go up imo, they can do more series on a day. After watching a 2-0 stomp without close games you could after that see a close series

I can understand where this is coming from, but I think we're on a point in the esport history where we can try it out and discard it if it ain't for the best. Personally a loser's bracket for those teams that manage to qualify as first in their groups, giving a reward (that's currently non existant) for pushing over just qualifying, and at the same time having a more evenly matched "quarterfinals" would which IMO give us more upsets, precisely because of the nature of the game which makes it all the more difficult for a team clearly worse than another to win 3 out of 5 games against them while a team slightly below could more reasonably do so (G2 vs RNG and imo SN vs TOP would both fall under this category), are more than enough reason to want a loser's bracket.

comment on part 2, that's why a game like CS-go can be interesting to watch, even though teams who are ahead aren't punished for loosing a round as hard as the loosing team. there's always a possibility for the team who is behind to win the next round and the next one again. mainly because it isn't the lead which determents who wins a round but more so the strategy or the mechanics used in the specific round. I believe its more similar to "normal sports" than a snowball game like league.

I think instead of a losers bracket they should do something like play-ins where 3 teams from each group get out and the 1st place team would get a bye and then the 2nd and 3rd seeds would be drafted into 4 matches and the winner of those would face the 1st seeded teams. To compensate for being drafted on same side of bracket that would only apply to the 2nd seeded teams (So for example DWG have the bye, C9 would have to be in opposite bracket but RGE would be on same side as DWG) Because the problem right now with worlds is with LCK/LPL being so far ahead most of the time you may only get 2 western teams into quarters and 6 eastern teams. To use this worlds as an example (I did my own draft into brackets assuming the rules currently in place). The first "quarterfinals" would be C9 vs LNG (Winner would play T1), next would be HLE vs 100T and winner plays GENG, next MAD vs PSG winner plays RNG and then EDG vs RGE winner plays DWG. Then if would be normal from there

7:11 - as someone from EU, I have to say, I highly doubt MAD would've won against C9 with a 3:0 score. I'm not sure who would've won, maybe I give the edge to MAD (maybe that is EU bias), but I kinda tend to a 50:50 here. Other than that: thanks for the video, gbay!

I understand your perspective, and I agree that it’s awesome to see upsets and underdog stories. However, going back to the RNG and iG example, I personally don’t like that iG was able to win the tournament, over the tournament favourites, because another team (G2) did the “dirty work” for them, in eliminating RNG. I think upsets would still happen frequently in a double elimination bracket, and that a double elimination format highlights upsets moreso than a single elimination bracket, as it pushes the winning team into a much more advantageous position than a win in single elimination, as they still have their safety net (and a potential rematch with the team they upset, which is always super exciting in other sports/esports which use double elimination). That’s just my two cents as a double elimination supporter, but I think you make some great points and I’m glad to see this being discussed.