An Update for Season 10

Users Who Are Viewing This Thread (Total: 1, Members: 0, Guests: 1)

Teeples

Member
Site Supporter
Founding Member
Did the players have a word into having a shorter league and more tournaments?

You know as much as I do about all of this, it was discussed in detail in the MKU staff chat. Both of us have only been MKU staff (well casting as well for me but that's irrelevant to what's being discussed here).
 

Shin

New member
Why don't you (don't we?) come back at a classic league with x teams and only 1 conference ? When I left MK it was this format and I found it way more competitive and interesting than conferences, which are for me only to sound "as cool as NBA"
 

Félix

Member
Founding Member
You know as much as I do about all of this, it was discussed in detail in the MKU staff chat. Both of us have only been MKU staff (well casting as well for me but that's irrelevant to what's being discussed here).

Don't worry, I just want to make sure the whole community knows we have our league ruined just because the site admins wants more small tournaments on the side, even though no one has a word. This is another major issue.
 

Jinnysoyun

Member
Founding Member
I do agree with what Félix posted. The fairness of the tournament should have priority over length or any kind of "cool ideas" like playoffs or whatever american sport uses this conference format. Since, as Félix stated, it's impossible to make exact seedings you should work on a format where every team plays each other the same amount of times first and then adjust it to length/playoffs/whatever you need.
 

Nato

Member
Staff member
Administrator
Site Supporter
Founding Member
I'm not in the MKU chat nor was I really involved in any indepth discussion on this, but I think its something we can test to see how it goes, and if its horrible, then there's no wrong in going back to the "better" format, better being whichever got better feedback of the two (or even 3 if we wanna take into account traditional division format)

In my opinion I love bracket format tournaments but its something that can't be the norm because unless your team is a top 8 or even top 16 team, you won't play much unless we do a consolation bracket, which is still something we could do.

The other option is a swiss format which wouldn't take long at all either, but again its just another suggestion.
 

Purple

Member
Site Supporter
Founding Member
Why don't you (don't we?) come back at a classic league with x teams and only 1 conference ? When I left MK it was this format and I found it way more competitive and interesting than conferences, which are for me only to sound "as cool as NBA"
They would use inches and feet if units of measurement were needed lol
 

Rookie

Member
Staff member
Administrator
Media Staff
Founding Member
Alright, so I’ve spent some time reading up everything on the thread and want to try to clear everything up and give my insight on what lead to these decisions, the logic behind it, why we thought it was a change worth making, and wrapping up I guess the bigger picture that Teeples mentioned. To start, I don’t expect anyone to read all of this, maybe just specific responses, but just to note, I’m quoting these out of order because the response is going to tie all the questions regarding a format change together. If you want a full understanding of the situation, it’s highly recommended to read through this at least the full section regarding the format, because I think it’ll tie up many loose ends / questions.

Additionally, I want to start by saying that the staff only acts in an attempt to benefit the community. Community satisfaction matters a lot to us, so I appreciate everyone for commenting and even expressing displeasure because it’s better than just not saying anything. If a format results in being incredibly unpopular, then it’s something we won’t continue. But like everyone else, staff talks about long term planning for the site and we have our own ideas that we believe in, and think will be successful. It is not our goal to hide anything or to control scheduling in a way that the community will not want, but of course there are changes we think will ultimately lead to improvement in how we run our system here on MKC. So for all the questions, I’ll give an answer on how we got to where we’re at now, and I hope you all will give it a read and understand our perspective here. I think for the sake of the future, we should start doing more feedback on each event, because people rarely do mention how much they like certain formats. Sometimes we have to take it as a lack of complaints. If you guys like a format / tournament in the future, please do speak up about it, we’ll probably try it again if/when it’s the case. I personally read as many suggestions as possible in #tournament-suggestions or even things written in general channels, so it’s worth noting. Maybe we’ll try doing more surveys soon. Anyways:

Why we want to change up the format again, how it all relates to site planning / scheduling

What would you change a format that actually works and 90% of people like

The easy answer here is that we refuse to get complacent. Obviously successes are going to be noted. But complacency is something we don’t want to fall into. The classic WCL format was a staple of the community for maybe 11 years. It was a working format, and people did like it. But last year, the MKU Staff decided to try something different that we think the community actually likes more now than the classic WCL format. If they didn’t try something different, maybe we miss out on something that the community actually does like. There is going to be hits and misses. If we think we can make improvements, we’ll try it out. If something doesn’t work out, it doesn’t mean that we can’t go back. Admittedly, the WCL format style of competition is very strange in the world of eSports. It was created in 2008 and was considered the perfect kind of thing for the MK community. As the time passes on, it's good that MKU is still popular, but we also want to modernize as well. In most eSports, the most competitive, highly regarded events don't have divisional formats. Anyone who enters a tournament could technically win it if they play well enough. It's obviously very important to have tournaments where newer players have a reason to compete, but it is very abnormal for a main event to be broken down into skill tiers. It's not a bad thing that MKU popular, and as long as it drives site participation, it's something that's necessary for the site. But we also think exposure to new kinds of ideas is really important for the community's future.

Did the players have a word into having a shorter league and more tournaments?

Like mentioned above, all suggestions are considered. It happened that in this offseason, we had a decent amount of people suggesting 8-team divisions on the basis of the skill gap being pretty large between a 1st and 12th seed of a conference. Here’s some instances of it:

g6bY4Iv.png


ag6mNdl.png


PUXlZHn.png


dE2qSdX.png


V65BZb6.png


ttoVq7J.png

So of course, this was always something that was going to be considered from our end. It also happens that when thinking about it more, the only thing that I could conclude is that perhaps a lot of good side effects would also come along from making this change. Before explaining that in detail, I want to bring up two things brought up in the same conversations back near the turn of the year.

9bc1IPA.png


0tetY9C.png


Pretty much, when Teeples says bigger agenda, the following is what he is referring to. In the first picture here, it’s mentioned that MKU Staff are saying that the rate of teams actually wanting to be moved higher has increased from 8U, perhaps due to the PR system we’ve been using. Admittedly, I don’t believe the PR has been as popular as I was hoping for on implementation, but we may get there eventually. For now, perhaps having an effect on teams craving tougher competition is justification of its existence, because it was a big goal of the PR idea to begin with. But additionally, the PR in a sense has been a way to keep a yearly flow of tournaments, divided into separate but consistent periods of time. And this is where MKU comes into the picture. We’ve ran 3 MKU Seasons of 10 Weeks the last year or two, which comes out to about 30 weeks. However, if we’re dividing our calendar years into two halves, MKU simply doesn’t work with that idea at all. Because 3/2 = 1.5, you’re going to have one MKU season in the middle of the year, in the middle of the evaluation period split. So logically, we needed to split PR seasons into three per year to accommodate properly; because if you split it 1.5-1.5, one half of the year has more tournaments / results than the other half (since results from a tournament starting in May would be going to the 2nd half of the year), and if you split it 2-1 or 1-2, then one of the MKU offseasons would be absolutely massive. It’s also a possibility to have 4 seasons of current MKU format a year, but if that were to happen, we’d essentially be getting rid of all non MKU team activities, with very little room for flexibility. I feel like the biggest fans of MKU may not even want for that to happen, as it deletes the possibility of new ideas or even returning things like a World Cup in future years. To me personally, it feels like backwards progress, like going back to the WCL era, albeit with a better format.

So under this logic, if we decide to continue with dividing events into PR seasons, we’re forced to go with 3 per year with the current MKU format. What this essentially means is that we have about 17 weeks to work with each, and MKU takes up 10 (or 11 now) weeks each. This means we generally have 6-7 weeks to run 2 additional team tournaments, and we don’t have much room for breaks in between. A lot of teams I think get exhausted with all the matches so close to each other, and I think the flow personally is not great. If we were to run an Amplify (3-4 weeks), then there is quite little time for break. Even this MKU Season will be ending in April, and we have not had one single Sunday off since the beginning of the year.

To my surprise, running an 8 team MKU really does open up the schedule / possibilities much more. Not only could we go with 6 month evaluation periods with ease now, but we could pretty easily do it without necessarily the need of cramming 2 events in between every MKU season. I think as a result, we could have a lot more flexibility in scheduling, perhaps more opportunities to try new things without worrying about confining it in a span of 2 weeks, and that perhaps events would feel more fresh. Sometimes we could have only one tournament in between MKU seasons, without the need to cram 3 per every 4 months. The longer an event, the less flexibility we have planning stuff around it pretty much. The larger the window, the more flexibility we have. We’d have shorter offseasons - if you take a look at Gumer’s message, this really stemmed off of the 3 MKUs a year scheduling. You could always decide to do away with a PR / scheduling system I suppose, but at risk of losing the perks Teeps mentioned as well as sacrificing some site organization overall. And the best part is that we’d only be losing 3 MKU weeks a year by changing, which is almost negligible from a yearlong standpoint. Despite losing a very small amount of MKU weeks a year, the proportion of matches with extra meaning (playoff matches) goes up. On top of this, this offers doubles the chances to win a division in a year, or more chances to win a title, etc. Here’s a head to head idea of what I mentioned:

NgF7KI7.png


Personally, I don’t believe overall that shorter seasons will hinder the quality of the tournaments to teams. It is still twice as long as really any other site tournament. The only true way to find out is to run the season and see how people feel I guess. Thinking about it, I would think those near the top of their divisions will still have a rewarding experience whether or not they win a division in 7 weeks or 11 weeks. Those teams who have a rough season however, will get to escape and try something new much more quickly. With the new format, those teams who have a bad Season 1 would already be done with their Season 2 before the Season 2 in the current format even begins. I think there has to be some positives overall here to consider stemming from ultimately a discussion people were having regarding lowering divisions to 8 teams each. In no way do we want to downplay MKU’s importance, which is why we wouldn’t consider this if it didn’t benefit the league itself.

Another potential positive is how it would affect the MKU Primetime streams. Disclaimer, this plays no part in any reasoning, because casts are not worth even a slight dip to the health of the league, but is just a small additional positive I felt would be neat to point out. We had a big problem with the casts this season because only half of Division 1 was able to participate at the later window (6/12). By projection, this would change to become (⅝) next season, with 5 of the current playoff teams being able to play in the later slot. We’d likely have the following comparison:

p9iL5Rk.png


Now, the thing that jumps out is that we have less castable wars, and hence less options for community entertainment. But this is actually not a bad thing, because even though there is less eligible wars, we would have more per week to choose from because the teams that are able to will play each other at a higher frequency. Additionally speaking, many eligible wars were not chosen despite existing because based on the standings, a lot of wars were deemed not good matches to cast. The decreased skill gap would also make the percentage of eligible wars that are considered ‘good’ higher. Additionally speaking, ARC would be one of the 3 teams who aren’t on the eligible list of teams, and most of their roster is able. They ended up being in a cast this season. Even if they just had 1 week of eligibility again, it would increase these numbers even further because roughly 71% of their opponents are able to as well here.

TLDR; shorter offseasons, more seasons/results, less bracket tournaments piled on top of each other.

But let’s move on and start talking actual format now.

Discussion of the Format Itself, Valid Competitions / Fairness

imo, reducing the number of teams per division is the right move.

However, I think the lenght of the tournament should be the same as now. Maybe playing two interconference matches would be the key to balance the wildcard probabilities between both conferences, and also to keep the duration of the competition.

For playoffs, I'd like to see the option 1, since the winners of the conference already are in advantage facing the wildcard teams and giving them a losers bracket might be too much. In addition, playoffs being BO3 and not having a reset option from losers bracket in GF is kinda weird.

To start, I agree wholeheartedly that Option 2 is wack because of the lack of bracket resetting. But MKU has started a trend of BO3s in playoffs, and BO3s don’t work well with resets in MK (potentially 6 matches in a row for a grand final, or two days of grand finals is kinda cringe, and even at that point playing a team 6 times in a league format is literally unheard of and has got to be super draining). This is why I think option 1 is better, even if it looks “less hype.” Besides, most of our tournaments outside of MKU run on double elimination already. In a sense this gives MKU “identity” (being the exclusive league + BO3 bracket tourney). BO3 also gives MKU some extra feeling of bulk compared to other tournaments, which I guess is fitting for the longest tournament on the site.

That being said, if inter-conference play is extended to 2 games, a new issue is created here: conferences are worthless now - because everyone would now have the same schedule. Thus, conferences would only exist as an extremely arbitrary ranking system, that would honestly be less fair than not having conferences. An 8-6 team may place 2nd in Conference B and get to host a 9-5 team who finished 3rd in Conference A, in the playoffs. Though this may already have been a case of unfortunate seeding, it becomes completely unjustified if interconference was 2 games, because the 9-5 team would’ve had a better record with the exact same schedule. So the solution at this point is if interconference is 2 games, you might as well drop conferences, and make it so it’s just double round robin. But then we just have the old MKU format on our hands. With playoffs. So pretty much the old CSL / WL format (I’m not sure if it still is) - but here is the new issue. We believe that this is a really fundamentally bad format, and it’s why MKU has never done the format or had playoffs until we introduced conference play. And here’s the explanation as to why:

It effectively screws over the team who has won the regular season, in short. Double round robin itself is a commonly used tournament format. We’d be running a double round robin tournament where the winner of the double round robin doesn’t get to then win the tournament. They would have played the EXACT same schedule as the rest of the league, and had the best record in doing so - then suddenly it doesn’t matter. You can argue that you can give them a bye I suppose, but ultimately, anything less than a title at that point is a bit of a middle finger to that team, considering they won the regular season.

What makes having conferences different is that nobody can actually be declared regular season champions at the end of the regular season. You could make the argument that this season, Requiem should be the regular season champions because they had the best record of 12-2-2 in Division 1. Except, it wouldn’t be a good / convincing argument. World Friend Star may have finished a game behind Re, but since their schedules were different, wf could maybe argue that they had a more difficult schedule. For example, HD finished high in wf’s conference, but we dropped our 2nd game to MT. Re didn’t have to play a 2nd game against MT. wf may say that the last place finishers in their conference were 6-10, while Re’s conference had a 5-11 and 0-16 team. Ultimately, there is no way to tell which conference is the superior one. They’re intentionally set up to be roughly equal. And this fact right here, gives playoffs a reason to exist: To prove who should be the champion, of a sometimes muddy picture. Double Round Robin is a FAIR tournament in itself - teams play the exact same schedule. When it is said and done, one team has won a tournament where every team had the same strength of schedule and same opportunities. This is why leagues such as the NFL / NHL have no issue crowning champions who may have not been the best team overall during the season. The schedules are never even and the same because of how divisions work. Look at the NBA right now - the season may be cancelled because of the coronavirus, and because the league is divided into conferences, fans can’t make a claim at who the champion should be. The Milwaukee Bucks have the best record in the NBA, but others are dismissing them because the Los Angeles Lakers have the best record in the conference most people consider to be better. If LA and MIL played the exact same schedule, we would have a definitive answer as to who should be considered the champion. But we don’t. Hence playoffs need to decide it!

It’s worth mentioning that playoffs probably are less fair than the old format, so I’m going to address this issue while responding to Shin.

Why don't you (don't we?) come back at a classic league with x teams and only 1 conference ? When I left MK it was this format and I found it way more competitive and interesting than conferences, which are for me only to sound "as cool as NBA"

This is an interesting comment, and it brings up the question. What do we value more? The most fair competition, or the most exciting competition? A little of both? Because you can make a valid point to say that the old WCL format (double round robin) is the most fair format we’ve had. There’s no bracket luck involved, it rewards consistency, and every team has the same schedule. Whoever wins played the same teams as everyone else and won the most among them.

Personally, I think it is not for the better. I would say it’s a bad argument to say “you can’t expect teams to be on their A game every week” to counter the WCL format. But though it may sound crazy, I think choosing the entertaining format may be for the better of the community than the possibly most fair competition. You can make an argument that as long as all the teams signed up for it and the rules don’t change, that a competition is fair. But maybe it’s true that we aren’t valuing consistency as much anymore. So why can I possibly think the less fair option may be better? Well, I think back to the best moments we’ve had in our MK8 history. I remember HD completing a 56 point comeback on the last race to win a MKU Tournament vs JBZZ. I remember wf bringing home their first MKU title on a final race vs DVP. I remember Toto getting billed off dNBC in the World Cup. Of the things you remember, did any of it happen in the MKU League format? Generally, I think it’s harder. It’s because playoff matches are definitive. They’re it. Winner takes all. The most memorable thing I remember happening during the WCL format in all of MK8 was the week that Alliance Rainbow took down Severance on Week 5 and the amount of thread views broke the site. Yet, it did nothing ultimately. AR didn’t win the division. They just had the lead one week. There wasn’t closure to it. It was just a good win, ultimately not much different than a scrimmage. You can always try to stack up the best/title deciding wars on Week 10, but it doesn’t guarantee anything. I remember Mp/xi and NF/svc as examples. When I was in Mp, we played a week 10 match vs NF (svc pretty much). NF was 9-0. We were in 2nd/3rd at like 6-3. We could’ve broken the 10-0, and maybe got 2nd. I was in the main lineup that season. It was a week 10 season finale. I chose to go to an in person MK8 tournament that was offered on my college campus instead. I won a coffee mug. I chose that over W10 of MKU between the top 2 seeds. Fast forward to 2017, xi plays svc in the grand finals of the MKU Cup I believe it was. xi had got upset by HD this tournament, a match we remember because wow, it sent xi to losers early. It had an affect. Then xi, down 30 in the last race of match 1, goes Mario Circuit and top 6’s svc to win and reset the bracket. It was madness. It was pretty much the same teams in the last match of 2 tournaments. One produced one of MK8Us most iconic final memories. The other I didn’t even bother showing up to play. It’s worth noting also that nobody can watch the rare W10 cast generally in the old format, because they’re all playing matches. In the current finals, most teams are actually out of their tournaments and free to watch also.

Besides just hype, I guess there’s more to it. There’s just more to play for each week. In the new format, a 6-3 vs 5-4 Week 5 primetime matchup may be more worth watching than a 9-0 vs 7-2 faceoff. In the WCL format, there’s no incentive caring about any matches outside of the ones affecting the division lead at this point. But in the other, there are teams near the middle who are still having impactful final regular season matches. Even for teams who don’t win a title, getting into the playoffs is an achievement on its own. I suppose that starting 0-2 in the old format is death sentence. It was demotivating. In many divisions you would have to finish 8-0 to even have a chance to win at that point. Maybe it’s more fair that losses mean more at that point. Teams may not deserve to win if they drop 3 games to begin with. But I would argue it’s more healthy for more teams having incentive to try hard late in a season because of playoffs. In Conference B of Division 1 this season, every team in the conference had a non zero shot at making the playoffs before the final week, for example. Only 2 teams had a shot to win the conference, and one team had like >95% odds (wf +1.5 games). I suppose you could argue that playoffs aren’t the most purely competitive format, but I do think the community overall likes what it adds in terms of casts, motivation, and getting to have a definitive final match guaranteed. While we’re on the topic of rough starts though:

I don't really get the point of these changes.

A year and a half ago we changed from MKU League (6 teams per division) because "if you drop 2 games, your league is over" and more reasons. And now we have a 4 teams per conference format? In my opinion, this will likely happen again, even though the skill gap gets reduced to a minimum, as you have the same amount of matches than in League.

I mean, you're having the same issue than before season 6 but the 3rd team of the conference may get into playoffs.

You’re correct in your evaluation that an 0-2 start will be more negatively impactful than it would be in S9’s format. But what I will say that the new proposed format is definitely much more similar to S9’s format than the previous Season 6 format. Let’s consider the example of having a 0-2 start and the effect it has on winning a division. The only way to win in the old format is to win a division. I took an arbitrary set of tables from WCL S7-S9 (33 results, a sufficient sample) and recorded the winning record of each:

x991edD.png


As you can see here, having a 0-2 start would lock a team out of winning in 20/33 ~61% of divisions in a 3 season span. Finishing 8-2 would not even guarantee a title also if you started 0-2, as 7 other teams won at 8-2. You would probably expect to win about 3/7 of these (since you would be tied with 1 other team, maybe 2 in these instances). So in reality, starting 0-2 you can expect to only have a 27% (1-24/33) chance of winning an old format after starting 0-2 - assuming you finish the season 8-0. Additionally speaking, it’s worth mentioning that half of the teams that won a division with a record worse than 8-2 were teams that beat a team with a better record due to bonus points (which are now gone completely).

Now even though this new proposed format is naturally more punishing on 0-2 starts than this season, it is not going to be even comparable to the old format. In the new format, while maybe if you were to take data on teams that would start 0-2, they may not make the playoffs a much higher percent of the time (due to teams who start 0-2 just naturally being weaker in general), if you want to talk about their actual elimination, then starting 0-2 on itself if not an obstacle that is insurmountable. There’s not a lot of comparable data for 8 team divisions yet, but using 4 I remember (PL S11, MKU DX S8 D6, MKU Preseason D2&D3) the average amount of wins for the 4th slot when stretching seasons to 10 games is about 6.2 wins, which rounds down to 6. I know it’s not a lot of data, but mathematically, it makes sense for a 4th place team to go about 6-4 in an 8 person 10 game season. It is much more reasonable for a team that starts 0-2 to finish 6-2 (6-4) than it is to finish 8-0 (8-2). It is also significantly more likely. There’s only one way to go 8-0 - win all your games. There’s 28 different ways you can go 6-2 over the last 8 games of a season. While some ways may be more likely than others, there is undoubtedly more flexibility when salvaging your season at this point than with the old format. So while a 0-2 start may be demotivating, it is really still not comparable to the league format, despite the equal game counts. One complaint is you may have to play the toughest teams first. This would be unlikely to say the least in this format, because that would be implying the best 2 teams are in the same conference. It’s possible due to staff error, but definitely less likely than not. Even if it were to happen, your schedule would get significantly easier afterwards. Take a look at HD in MKUDX S7: we started off 0-3, and had to play the top 2 seeds of our conference next. BRUTAL start. After Week 5, we were 6-4. The second we are in the playoff picture, that 0-3 start no longer hinders us. It still would haunt us in the WCL format.

Alright, so here is the bulk of the post, because Felix brings up lots of good points and I think I need to highlight / emphasize a lot of parts of this post separately. Let’s start with the main point:

Now, this is the main point of my post. Why do I care a lot about these mistakes seedings? Because these (inevitable) mistakes will have way more incidence on your team runs than the previous formats. It means that teams are more likely to miss playoffs (or get a free pass in them) just due to unfair seedings. In order words, it means it’s more unfair for everyone. It’s all but healthy. Why is it more unfair? Because say in a conference there is a team that is weaker than everyone else in their division. Three teams will have 20% of their season matches against that team, while the four others will have 10% of their season matches against them. The previous format had these percentages reduced (12.5% vs. 6.25%), which means you had more control on your own run.

I think this is better for most people to illustrate this with an example: the previous Division 1 would have been: 1. Prestige; 2. World Friend Star; 3. High Definition; 4. Requiem; 5. Euphorya; 6. Blizzard; 7. Arcadia Sky; 8. Rozando la Katastrofe.

Removing all matches from Star Clan, Mushroom Team, Vertex and Jean-Pierre Pernaut, the final standing would have been:

EU 7-0-3
Prestige 7-0-3
Re 6-2-2
RK 0-0-10

wf 6-2-2
HD 6-1-3
Bz 3-0-7
ARC 2-1-7

If I am a member from High Definition, I am salty to miss the playoffs while 3 other teams faced RK twice in a that short season. But what you don’t know is more than half of the staff who suggested seedings wanted Bz have a higher seed than EU. So what would have happened if Bz had seed 5 and EU had seed 6 instead? It would have make it even more harder for HD (and EU too) (based on the results). Please do not argue like it wasn’t the case anyway. As written few times already, these kind of mistakes are inevitable, and will happen again.

To start on this case, you bring up some good points here. While your example using this seasons Division 1 is fine, the conclusion you reach here is quite dishonest and misleading I believe. What you are doing here is painting a picture that HD got screwed over, because they didn’t get to play RK twice. But what you’re not mentioning is that other matches matter. You are saying that HD played at a disadvantage, but when considering all matches, you’ll see that this isn’t even the case. To measure strength of schedule without using seeds but rather results, it’s better to look at the record of your opponents. In this case, instead of just looking at how many times HD plays RK, let’s take an evaluation of strength of schedule between Prestige and HD overall, both of who placed 2nd in their conference. Let’s count the points of their opponents in this season, using ties as a half point:

Prestige: 7+7(EU)+7+7(Re)+0+0(RK)+7(wf)+6.5(HD)+3(Bz)+2.5(ARC)=47

HD: 7(EU)+7(Prestige)+7(Re)+0(RK)+7+7(wf)+3+3(Bz)+2.5+2.5(ARC)=46

So as you can see, even with HD’s lack of 2nd game vs RK, strength of schedule suggests that HD had an easier schedule than Prestige/Re did. We would not be justified to complain like you’re suggesting. How could this happen?

Well, let’s consider this also. You’re claiming that Prestige had an advantage because they got to play RK twice. Well, let’s take a look at both teams records vs RK and ARC this season, who finished 4th in your standings:

Prestige 3-0
HD 3-0

So, nothing changed. How about we look at the records vs Bz and Re?

Prestige 2-0-1
HD 2-1-0

So HD’s competition to actually finish in 2nd of their conference was likely easier. Want to know the difference? We lost to Prestige and EU, and only tied Re. That’s why we’d be left out of the playoffs in this scenario. For having a 1-1-2 record vs the other conference. And that is on US for not winning these games. Not because we didn’t get an extra game vs RK. The whole premise of this main point is based off a logical fallacy.

Take this chart to further explain the dilemma we’re discussing right now:

w2Db8bl.png


Now, we’re not actually going to ever know how the division layout will happen. Logically speaking, the top seed will have the easiest schedule because they don’t have to play themselves. The point is though that the average difficulty of a schedule based on seeding is effectively going to have the same overall average: middle of the middle seeds. As you have more and more divisions recorded, naturally it’s going to work out that 2nd seeds in both conferences are going to have slightly different roads. One will generally have a harder 3rd seed and easier 4th seed than the other in their division. There will be instances where results differ and make divisions look wack. I think this is the point of competition though. For teams to exceed expectations, not to just finish exactly as predicted.

Now, there’s going to be scenarios where this will be at times more unfair than others. In 8U, when svc, FA, xi, dominated the field, svc would appear to gain an advantage from the other 2 naturally having to play each other more. But then the winner of the FA/xi conference would actually get the easier playoff match. We’re never going to know where there’s more massive gaps (in this case, between the 3 and 4 seeds). Maybe sometimes the gap is somewhere else. Maybe you get a clean 8 team division sometimes all in the same ball park. We won’t be able to control this. Luckily now though, the top 8 is much more balanced than in 8U. Can seeding further effect how it all plays out? Sure, it can. It can make some conferences harder than others for sure, you’re right. But it’s not like the team in the opposite conference doesn’t have control of their fate. They can always just beat teams in the other conference.

Now, I don’t know how you say this ruins the league also. Because the formatting is not that much different. I understand your argument, but let’s not pretend that this can’t happen with 12 team divisions as well:

unknown.png


Additionally speaking, when you have an imbalanced win total between conferences in 12 team divisions, the win totals are much more absurd in scenarios like this. Strength of record example:

YF: 136
FE: 118

For perspective, YF on averaged played a team with one more win than FE did. YF won more games. FE hosted a playoff game. YF was not in the playoffs. I get that it's a little bit more likely to happen in 8-team division problems, but this is also a flaw of the 12-team format.

To insist more on the unfairness, take another example: say in a mid/low division, where 2 teams are way less stronger than all the other 6 teams and, by mistake, they are seeded in the same conference. This is likely to happen, and it will happen eventually if this format is kept long term.

It means that 2 of the 6 remaining teams will have 40% of their matches against 2 weaker teams (/!\ which is almost the half of the season /!\), and the remaining four teams will only have 20% of their matches against these 2 weaker teams. So two teams have a massive advantage to get a playoff spot. How can you argue again if you support this format? This format is massively unfair. It grinds my gear to see how the staff is currently running into a wall.

I understand this point, and it is valid because you are correct that this scenario is more likely to happen in the new proposed format. I would like to ask you a strange question to consider at this point, but it ties back to something earlier in the post. Do you prefer the format we’ve been running the last few seasons, or do you prefer the original MKU format? If the answer is the WCL format, then that is fair, but I would present you the same argument that I presented Shin. If the whole community preferred to go back, it would be one thing, but I think the conference format is overall preferred. But if your answer is you prefer the 12 team conference format, I’d like to ask you about why - because it is a more unfair format that doesn’t reward consistency as much. It just has other perks that justify running the format. Maybe you agree with me that the old MKU format is more fair. But I think to imply that the current format is fair and that this new format isn’t fair would be a little bit ridiculous. I would say that you have the opinion of the top part of this image, where I find the bottom to be more accurate.

u2X0LVN.png


I will give you that 12 teams probably creates a more fair tournament. But I would also give you that WCL format is the most fair tournament. But like the reasons above that 12 team format is worth changing from WCL, I think there are still reasons that exist for changing to 8 teams as well, such as overall skill gaps in divisions shrinking, and giving people more tournaments overall (including more separate MKUs). At the end of the day we have to ask where we draw the line. As long as a tournament is to the right of the valid tournament line, I think it is worth considering. An example of something to the left would be Amplify V (also admittedly, this was my fault for suggesting we could make Swiss format work with 40 teams. I’ll 100% take the blame on this and it won’t happen again) because the scheduling was unfair enough that if the tournament was repeated an infinite amount of times, the tournament would offer a systematic advantage to 5 seeds. What I think is being suggested with an 8 team tournament may be to the left of 12 team, but is still easily a valid tournament. I believe it’s valid because as you run it an infinite amount of times, the trend comes out to be overall fair (it may have some faults, here and there because of seeding or because of gaps between a 3rd and 4th seed that may disappear over time as well). I think you’re suggesting that it isn’t valid at all, and I think to suggest that the 12 team MKU can be valid but the 8 team not is honestly, incorrect. The flaws still exist in the 12-team format.

To re-emphasize, I’ll give you that 12 team MKU is a little bit more fair than the 8 team MKU. But I think you’re being a little bit disingenuous with your evaluations of unfairness in the 8 team MKU idea. To start, you say in your example of the 2 worst teams being in the same conference will let 2 teams play 40% of their matches against the worst 2 teams, and say this is almost half the matches. 40% is a lot, I’ll admit, but it’s not a 40% increase over what we have right now. It’s a 15% increase over 25%. It’s also a large possibility (maybe larger, but I don’t know for sure) that the two worst teams go to the same conference in our current format. The number 40 is large, but 25 is still significant. We could also bump divisions up to 16 teams instead of 12. This way, if the two worst teams were to be in the same conference, you would only play 4/22 matches against the bottom 2 teams, which would be even lower. But in doing this, we’re now also making teams we previously considered bottom of Division 3 have to play the top seed of D1. At some point we’ve got to draw a line on the skill gap, or to just not have divisions to begin with. Really, no matter what you pick, you’re sacrificing another point of argument.

To continue on with this point, the issues of imbalance may occur a little bit more in 8-team format than 12-team format. But to act like these problems are exclusive to 8 team by calling it unfair and not 12, it’s just not true. For instance, division balance may not always work out in 8 team divisions. Is that problem exclusive to 8 team divisions? No. A matter of fact, as seedings are messed up as you mentioned, there will naturally be really rough scenarios. Larger conferences also run the risk of being imbalanced beyond what a smaller conference could even be in terms of imbalance. In just 15 attempts (not exactly a lot) at conference format with 12 team divisions, we’ve had the following occurrence, as well as the one from earlier.

E5f0Xwc.png


It’s worth noting that in a 8-team format, as long as a conference has at least a 4-16 record against the other one, the balance would literally be better than this example. Considering it is very likely for each conference to have a single top 2-3 seed, getting just 4 wins against the other conference is not really tough at all. Here, Conference A only won 39% of the total games in the division (the absolute minimum is also 31%, not 0%). It would take only a 4-16 conference record in the new format for a conference to win 40% of the total games. In an optimal world, it should be close to 50 for each. Overall, the point is that while it’s slightly easier to have conference imbalance, the levels of imbalance have slightly more potential to be disastrous in 12 team formats.

The big worry for me here is that in every case now, if one season has a case of misinformed seeding, the new format is going to be blamed automatically, when it is not necessarily because of the new format. There always is a chance that it is the fault of the new format, but as mentioned, a lot of the problems from the 8 and 12 team formats aren’t mutually exclusive.

I hate to write this, but staff is more likely to misjudge teams that aren’t in high divisions. Sadly, it’s a fact, and once again it’s not because staff cares less about these teams. It’s just because random factors (like an unknown player popping off) are more likely to happen. This means if you are a team in a mid/low divisions, you are most likely to be victim of these unfairness.

Again, you’re right that teams can be easily misjudged, maybe we don’t know about a player for instance. It may be a tough request, but I do think we can collectively consider more data on seedings, or request the staff team takes more time. It’s honestly possible that some unknown players shouldn’t necessarily need to be unknown for instance. But yes, the odds of it happening increase the lower the division you are. But it is also worth noting that if there are few unknown players / seeds / scenarios where teams may become a victim of unfortunate unfairness, would you rather have 11 teams affected per mishap, or 7? We’ve both acknowledged that there will be some flaws in every seeding no matter how big or small the divisions are, but would this not limit the amount of teams exposed to bad seedings? Because the only true way to get out of these cases is to have one large division, which would defeat the purpose of MKU. Additionally, what is interesting is the new formats actually protect against this stuff better than the old WCL format does. The old format could be more fair from a tournament standpoint, but the old format wasn’t protected against tournaments where the team listings in each division were never balanced to begin with. At least you can play OP teams in a playoff series now. But anyway, OP teams will typically affect less teams in a 8 team format, unless there's so many mishaps that it pretty much makes up the whole league outside D1 (I don't think this has happened yet).

For those who followed Division 1, there was a team (Rozando la Katastrofe) who got a 0-16 season. If we had this new format on before, they would have been still in Division 1. Would have it been closer? Not really. However, these seedings were agreed by mostly everyone in staff once we discussed them, and no one predicted that RK would have lost every game. Moreover, Novae got the seed #24, which is the last (12th) seed in the current Division 2. With the new format, it means they would have been the last seed in the new Division 3. However, they got a very nice record of 13-0-3 this season and even got a 2nd place.

You make a valid point regarding RK and Novae, especially Novae. I honestly think that this case is a bit of an outlier. And as you mentioned, there’s going to be nothing we can do about the occasional outlier. To test if this is truly an outlier though, I went through the playoff success vs seeds for all of Season 8, and made a scatterplot.

66n8cwY.png


Only 1 12th seed made the playoffs last season, and it was a Round 1 exit. Now, aside from the crazy season that 5 seeds had (which I really doubt will happen again soon), the fact that you can see a very clear negative correlation here indicates to me that the staff, at least for Season 8, did a fairly good job of seeding the conferences. Good enough that I personally trust that mistakes like underseeding Novae are going to be quite rare. At the end of the day, they made the right call to put them in D2 as well. As for RK, perhaps we got some new information. Had we ran an 8 team division, we probably would’ve not thought any better. But this is going to happen with ANY kind of cut off whatsoever. Let me bring up an example.

Last season in Division 2, JPP1 and wf2 were the D2 wild card teams. Wf2 finished a game above JPP in the standings, but ultimately JPP won an extra game in the playoffs. JPP was seeded 12th this season, and wf2 was seeded 13th. Now wf2 is having an ok season in D2, don’t get me wrong. They are a 9-1-6 wild card team, and did win a playoff series. But up to this point, DY has been the team dominating D2 this season. JPP was not expected to do so well in D1 this season, but they won a pretty respectable 5 games. As a 12th seed, winning 5 games is generally going above expectation, and especially in Division 1. Pretty much, the staff’s 12th seed overperformed and the 13th seed underperformed (though I suppose they could still make a run, it just looks like DY is the team to beat). If it wasn’t for a single playoff match, these seeds are flipped and perhaps JPP obliterates D2 while wf2 struggles in D1. But we won’t know for sure. If we’re just running 8 team divs, we don’t have to worry about this question, and no team ends up ultimately in a D1 that they have little chance to win in at the moment since the skill gap is a bit larger. The point is, any cut off is going to introduce new potential problems that we have to worry about. You can argue that having more divisions will create more cut offs, but I’ll counter argue that this will be made up for with general skill gap balancing.

Something else you’re correct on is the scatter plot shows that it is difficult to actually nail seeds down to the exact number. While the staff tends to get the actual divisions right, conference balance is tougher. In Season 8, the conferences containing the 1 and 5 seeds definitely were a bit stronger overall. These things can happen, especially more near the mid/bottom divisions. What this scatter plot also shows though is that it happens in 12 team divisions as well. It may have a larger effect on 8 team divs at times also. Ultimately the discussion comes down to whether or not the percentage increase is worth the other benefits. I’d argue yes, because like I mentioned, I think that this is still more than a valid format. I get the feeling that you probably won’t budge, but regardless I hope this at least explains clearly why we want to try this format. I think while some of the criticisms are valid, many of them are blown out of proportion.

Additional Questions

But with less teams in each division what will happen to the teams that win their promotion, will they go to the next div or stay in the same?

In reality, promotion / relegation matches never were a very good way to determine who should be where next season. Not only do teams change significantly in offseasons with player additions and subtractions, rendering past results sometimes not very accurate for the future, but the system is a little weird at times. In division 3 this season, a relegation match would exist between a 7-9 team and a 1-15 team. If the 1-15 team wins 495-489 or something, does that suddenly change your mind on who should be relegated next season? Overall, it’s not really a consistent system and I think the staff would have better seedings if they just focused instead on results while considering player movement and offseason improvement too.

Just my opinion, but I hope this format is a joke of mkcentral staff. It's really unfair come on. Furthermore, making format so short will desmotivate people, cuz those members of teams that have less chance to play due to low level compared rest of their team, now they won't play a sh*t.

I think it was mentioned already, but the purpose of the leagues is to go all out with your best squad available. I think a lot of this response is acknowledging that a skill gap problem exists also, if anything. If teams are using their weaker players in matches intentionally, it’s likely that teams are not taking certain opponents as seriously. If anything, this is suggesting that divisions are gaining enough of a skill gap that they feel like they don’t need to worry risking losing matches even with throwing out their weaker players in the lineups. I would say most people would acknowledge that sandbagging while still expecting to win has no business in competitive leagues, especially ones where divisions exist with the intention to pit teams against rival teams close in ability.

I don’t think a shorter format should demotivate people also. The next tournament, that brings a fresh start, will be sooner as a result. I don’t believe it would ruin the impact of winning either. I’ve witnessed teams being super excited / putting banners in signatures over winning 10 match divisions for over a decade now. It doesn’t mean any less, and now there’s more chances to win, if anything. In a 72 team league, with the current format we would have 18 division champions a year. 18 different chances for people to win a MKU division every year. If we do the proposed format, this number doubles to 36 division champions a year. With more chances to win, it’s also possible that a side effect of this is less people trying to cheese the system by trying to get underseeded, since there’s now more opportunities to win to begin with.

By the way, have u though about reducing the number of teams in each conference to 5 instead of 4? I think its a posibility that might work better

This honestly is not off the table. It could even be the happy medium between slightly better skill gaps and making the playoff system closer to the balance of 12 team divisions. I’m not sure. It would be a weird season though, because this would force us to have at least 1 inter conference match a week, and the schedule would feel more random as a result. This is because if A1 plays A2, A3 plays A4, B1 plays B2, B3 plays B4, the only match left in this week is an inter conference A5 vs B5. I don’t think there’s anything inherently wrong with this idea. As discussed earlier, the format should be valid I think. If interest for this happens, it will be considered. 4 or 6 teams in the playoffs would feel quite weird too.

They would use inches and feet if units of measurement were needed lol

Absolutely false. The metric system is just better. It's base 10.
 

Cynthia.

ヒンメル
Staff member
Moderator
MKU Staff
Site Supporter
Founding Member
I haven't read everything yet but I can already say that this message is incredible
 

Purple

Member
Site Supporter
Founding Member
Alright, so I’ve spent some time reading up everything on the thread and want to try to clear everything up and give my insight on what lead to these decisions, the logic behind it, why we thought it was a change worth making, and wrapping up I guess the bigger picture that Teeples mentioned. To start, I don’t expect anyone to read all of this, maybe just specific responses, but just to note, I’m quoting these out of order because the response is going to tie all the questions regarding a format change together. If you want a full understanding of the situation, it’s highly recommended to read through this at least the full section regarding the format, because I think it’ll tie up many loose ends / questions.

Additionally, I want to start by saying that the staff only acts in an attempt to benefit the community. Community satisfaction matters a lot to us, so I appreciate everyone for commenting and even expressing displeasure because it’s better than just not saying anything. If a format results in being incredibly unpopular, then it’s something we won’t continue. But like everyone else, staff talks about long term planning for the site and we have our own ideas that we believe in, and think will be successful. It is not our goal to hide anything or to control scheduling in a way that the community will not want, but of course there are changes we think will ultimately lead to improvement in how we run our system here on MKC. So for all the questions, I’ll give an answer on how we got to where we’re at now, and I hope you all will give it a read and understand our perspective here. I think for the sake of the future, we should start doing more feedback on each event, because people rarely do mention how much they like certain formats. Sometimes we have to take it as a lack of complaints. If you guys like a format / tournament in the future, please do speak up about it, we’ll probably try it again if/when it’s the case. I personally read as many suggestions as possible in #tournament-suggestions or even things written in general channels, so it’s worth noting. Maybe we’ll try doing more surveys soon. Anyways:

Why we want to change up the format again, how it all relates to site planning / scheduling

The easy answer here is that we refuse to get complacent. Obviously successes are going to be noted. But complacency is something we don’t want to fall into. The classic WCL format was a staple of the community for maybe 11 years. It was a working format, and people did like it. But last year, the MKU Staff decided to try something different that we think the community actually likes more now than the classic WCL format. If they didn’t try something different, maybe we miss out on something that the community actually does like. There is going to be hits and misses. If we think we can make improvements, we’ll try it out. If something doesn’t work out, it doesn’t mean that we can’t go back. Admittedly, the WCL format style of competition is very strange in the world of eSports. It was created in 2008 and was considered the perfect kind of thing for the MK community. As the time passes on, it's good that MKU is still popular, but we also want to modernize as well. In most eSports, the most competitive, highly regarded events don't have divisional formats. Anyone who enters a tournament could technically win it if they play well enough. It's obviously very important to have tournaments where newer players have a reason to compete, but it is very abnormal for a main event to be broken down into skill tiers. It's not a bad thing that MKU popular, and as long as it drives site participation, it's something that's necessary for the site. But we also think exposure to new kinds of ideas is really important for the community's future.



Like mentioned above, all suggestions are considered. It happened that in this offseason, we had a decent amount of people suggesting 8-team divisions on the basis of the skill gap being pretty large between a 1st and 12th seed of a conference. Here’s some instances of it:

g6bY4Iv.png


ag6mNdl.png


PUXlZHn.png


dE2qSdX.png


V65BZb6.png


ttoVq7J.png

So of course, this was always something that was going to be considered from our end. It also happens that when thinking about it more, the only thing that I could conclude is that perhaps a lot of good side effects would also come along from making this change. Before explaining that in detail, I want to bring up two things brought up in the same conversations back near the turn of the year.

9bc1IPA.png


0tetY9C.png


Pretty much, when Teeples says bigger agenda, the following is what he is referring to. In the first picture here, it’s mentioned that MKU Staff are saying that the rate of teams actually wanting to be moved higher has increased from 8U, perhaps due to the PR system we’ve been using. Admittedly, I don’t believe the PR has been as popular as I was hoping for on implementation, but we may get there eventually. For now, perhaps having an effect on teams craving tougher competition is justification of its existence, because it was a big goal of the PR idea to begin with. But additionally, the PR in a sense has been a way to keep a yearly flow of tournaments, divided into separate but consistent periods of time. And this is where MKU comes into the picture. We’ve ran 3 MKU Seasons of 10 Weeks the last year or two, which comes out to about 30 weeks. However, if we’re dividing our calendar years into two halves, MKU simply doesn’t work with that idea at all. Because 3/2 = 1.5, you’re going to have one MKU season in the middle of the year, in the middle of the evaluation period split. So logically, we needed to split PR seasons into three per year to accommodate properly; because if you split it 1.5-1.5, one half of the year has more tournaments / results than the other half (since results from a tournament starting in May would be going to the 2nd half of the year), and if you split it 2-1 or 1-2, then one of the MKU offseasons would be absolutely massive. It’s also a possibility to have 4 seasons of current MKU format a year, but if that were to happen, we’d essentially be getting rid of all non MKU team activities, with very little room for flexibility. I feel like the biggest fans of MKU may not even want for that to happen, as it deletes the possibility of new ideas or even returning things like a World Cup in future years. To me personally, it feels like backwards progress, like going back to the WCL era, albeit with a better format.

So under this logic, if we decide to continue with dividing events into PR seasons, we’re forced to go with 3 per year with the current MKU format. What this essentially means is that we have about 17 weeks to work with each, and MKU takes up 10 (or 11 now) weeks each. This means we generally have 6-7 weeks to run 2 additional team tournaments, and we don’t have much room for breaks in between. A lot of teams I think get exhausted with all the matches so close to each other, and I think the flow personally is not great. If we were to run an Amplify (3-4 weeks), then there is quite little time for break. Even this MKU Season will be ending in April, and we have not had one single Sunday off since the beginning of the year.

To my surprise, running an 8 team MKU really does open up the schedule / possibilities much more. Not only could we go with 6 month evaluation periods with ease now, but we could pretty easily do it without necessarily the need of cramming 2 events in between every MKU season. I think as a result, we could have a lot more flexibility in scheduling, perhaps more opportunities to try new things without worrying about confining it in a span of 2 weeks, and that perhaps events would feel more fresh. Sometimes we could have only one tournament in between MKU seasons, without the need to cram 3 per every 4 months. The longer an event, the less flexibility we have planning stuff around it pretty much. The larger the window, the more flexibility we have. We’d have shorter offseasons - if you take a look at Gumer’s message, this really stemmed off of the 3 MKUs a year scheduling. You could always decide to do away with a PR / scheduling system I suppose, but at risk of losing the perks Teeps mentioned as well as sacrificing some site organization overall. And the best part is that we’d only be losing 3 MKU weeks a year by changing, which is almost negligible from a yearlong standpoint. Despite losing a very small amount of MKU weeks a year, the proportion of matches with extra meaning (playoff matches) goes up. On top of this, this offers doubles the chances to win a division in a year, or more chances to win a title, etc. Here’s a head to head idea of what I mentioned:

NgF7KI7.png


Personally, I don’t believe overall that shorter seasons will hinder the quality of the tournaments to teams. It is still twice as long as really any other site tournament. The only true way to find out is to run the season and see how people feel I guess. Thinking about it, I would think those near the top of their divisions will still have a rewarding experience whether or not they win a division in 7 weeks or 11 weeks. Those teams who have a rough season however, will get to escape and try something new much more quickly. With the new format, those teams who have a bad Season 1 would already be done with their Season 2 before the Season 2 in the current format even begins. I think there has to be some positives overall here to consider stemming from ultimately a discussion people were having regarding lowering divisions to 8 teams each. In no way do we want to downplay MKU’s importance, which is why we wouldn’t consider this if it didn’t benefit the league itself.

Another potential positive is how it would affect the MKU Primetime streams. Disclaimer, this plays no part in any reasoning, because casts are not worth even a slight dip to the health of the league, but is just a small additional positive I felt would be neat to point out. We had a big problem with the casts this season because only half of Division 1 was able to participate at the later window (6/12). By projection, this would change to become (⅝) next season, with 5 of the current playoff teams being able to play in the later slot. We’d likely have the following comparison:

p9iL5Rk.png


Now, the thing that jumps out is that we have less castable wars, and hence less options for community entertainment. But this is actually not a bad thing, because even though there is less eligible wars, we would have more per week to choose from because the teams that are able to will play each other at a higher frequency. Additionally speaking, many eligible wars were not chosen despite existing because based on the standings, a lot of wars were deemed not good matches to cast. The decreased skill gap would also make the percentage of eligible wars that are considered ‘good’ higher. Additionally speaking, ARC would be one of the 3 teams who aren’t on the eligible list of teams, and most of their roster is able. They ended up being in a cast this season. Even if they just had 1 week of eligibility again, it would increase these numbers even further because roughly 71% of their opponents are able to as well here.

TLDR; shorter offseasons, more seasons/results, less bracket tournaments piled on top of each other.

But let’s move on and start talking actual format now.

Discussion of the Format Itself, Valid Competitions / Fairness

To start, I agree wholeheartedly that Option 2 is wack because of the lack of bracket resetting. But MKU has started a trend of BO3s in playoffs, and BO3s don’t work well with resets in MK (potentially 6 matches in a row for a grand final, or two days of grand finals is kinda cringe, and even at that point playing a team 6 times in a league format is literally unheard of and has got to be super draining). This is why I think option 1 is better, even if it looks “less hype.” Besides, most of our tournaments outside of MKU run on double elimination already. In a sense this gives MKU “identity” (being the exclusive league + BO3 bracket tourney). BO3 also gives MKU some extra feeling of bulk compared to other tournaments, which I guess is fitting for the longest tournament on the site.

That being said, if inter-conference play is extended to 2 games, a new issue is created here: conferences are worthless now - because everyone would now have the same schedule. Thus, conferences would only exist as an extremely arbitrary ranking system, that would honestly be less fair than not having conferences. An 8-6 team may place 2nd in Conference B and get to host a 9-5 team who finished 3rd in Conference A, in the playoffs. Though this may already have been a case of unfortunate seeding, it becomes completely unjustified if interconference was 2 games, because the 9-5 team would’ve had a better record with the exact same schedule. So the solution at this point is if interconference is 2 games, you might as well drop conferences, and make it so it’s just double round robin. But then we just have the old MKU format on our hands. With playoffs. So pretty much the old CSL / WL format (I’m not sure if it still is) - but here is the new issue. We believe that this is a really fundamentally bad format, and it’s why MKU has never done the format or had playoffs until we introduced conference play. And here’s the explanation as to why:

It effectively screws over the team who has won the regular season, in short. Double round robin itself is a commonly used tournament format. We’d be running a double round robin tournament where the winner of the double round robin doesn’t get to then win the tournament. They would have played the EXACT same schedule as the rest of the league, and had the best record in doing so - then suddenly it doesn’t matter. You can argue that you can give them a bye I suppose, but ultimately, anything less than a title at that point is a bit of a middle finger to that team, considering they won the regular season.

What makes having conferences different is that nobody can actually be declared regular season champions at the end of the regular season. You could make the argument that this season, Requiem should be the regular season champions because they had the best record of 12-2-2 in Division 1. Except, it wouldn’t be a good / convincing argument. World Friend Star may have finished a game behind Re, but since their schedules were different, wf could maybe argue that they had a more difficult schedule. For example, HD finished high in wf’s conference, but we dropped our 2nd game to MT. Re didn’t have to play a 2nd game against MT. wf may say that the last place finishers in their conference were 6-10, while Re’s conference had a 5-11 and 0-16 team. Ultimately, there is no way to tell which conference is the superior one. They’re intentionally set up to be roughly equal. And this fact right here, gives playoffs a reason to exist: To prove who should be the champion, of a sometimes muddy picture. Double Round Robin is a FAIR tournament in itself - teams play the exact same schedule. When it is said and done, one team has won a tournament where every team had the same strength of schedule and same opportunities. This is why leagues such as the NFL / NHL have no issue crowning champions who may have not been the best team overall during the season. The schedules are never even and the same because of how divisions work. Look at the NBA right now - the season may be cancelled because of the coronavirus, and because the league is divided into conferences, fans can’t make a claim at who the champion should be. The Milwaukee Bucks have the best record in the NBA, but others are dismissing them because the Los Angeles Lakers have the best record in the conference most people consider to be better. If LA and MIL played the exact same schedule, we would have a definitive answer as to who should be considered the champion. But we don’t. Hence playoffs need to decide it!

It’s worth mentioning that playoffs probably are less fair than the old format, so I’m going to address this issue while responding to Shin.



This is an interesting comment, and it brings up the question. What do we value more? The most fair competition, or the most exciting competition? A little of both? Because you can make a valid point to say that the old WCL format (double round robin) is the most fair format we’ve had. There’s no bracket luck involved, it rewards consistency, and every team has the same schedule. Whoever wins played the same teams as everyone else and won the most among them.

Personally, I think it is not for the better. I would say it’s a bad argument to say “you can’t expect teams to be on their A game every week” to counter the WCL format. But though it may sound crazy, I think choosing the entertaining format may be for the better of the community than the possibly most fair competition. You can make an argument that as long as all the teams signed up for it and the rules don’t change, that a competition is fair. But maybe it’s true that we aren’t valuing consistency as much anymore. So why can I possibly think the less fair option may be better? Well, I think back to the best moments we’ve had in our MK8 history. I remember HD completing a 56 point comeback on the last race to win a MKU Tournament vs JBZZ. I remember wf bringing home their first MKU title on a final race vs DVP. I remember Toto getting billed off dNBC in the World Cup. Of the things you remember, did any of it happen in the MKU League format? Generally, I think it’s harder. It’s because playoff matches are definitive. They’re it. Winner takes all. The most memorable thing I remember happening during the WCL format in all of MK8 was the week that Alliance Rainbow took down Severance on Week 5 and the amount of thread views broke the site. Yet, it did nothing ultimately. AR didn’t win the division. They just had the lead one week. There wasn’t closure to it. It was just a good win, ultimately not much different than a scrimmage. You can always try to stack up the best/title deciding wars on Week 10, but it doesn’t guarantee anything. I remember Mp/xi and NF/svc as examples. When I was in Mp, we played a week 10 match vs NF (svc pretty much). NF was 9-0. We were in 2nd/3rd at like 6-3. We could’ve broken the 10-0, and maybe got 2nd. I was in the main lineup that season. It was a week 10 season finale. I chose to go to an in person MK8 tournament that was offered on my college campus instead. I won a coffee mug. I chose that over W10 of MKU between the top 2 seeds. Fast forward to 2017, xi plays svc in the grand finals of the MKU Cup I believe it was. xi had got upset by HD this tournament, a match we remember because wow, it sent xi to losers early. It had an affect. Then xi, down 30 in the last race of match 1, goes Mario Circuit and top 6’s svc to win and reset the bracket. It was madness. It was pretty much the same teams in the last match of 2 tournaments. One produced one of MK8Us most iconic final memories. The other I didn’t even bother showing up to play. It’s worth noting also that nobody can watch the rare W10 cast generally in the old format, because they’re all playing matches. In the current finals, most teams are actually out of their tournaments and free to watch also.

Besides just hype, I guess there’s more to it. There’s just more to play for each week. In the new format, a 6-3 vs 5-4 Week 5 primetime matchup may be more worth watching than a 9-0 vs 7-2 faceoff. In the WCL format, there’s no incentive caring about any matches outside of the ones affecting the division lead at this point. But in the other, there are teams near the middle who are still having impactful final regular season matches. Even for teams who don’t win a title, getting into the playoffs is an achievement on its own. I suppose that starting 0-2 in the old format is death sentence. It was demotivating. In many divisions you would have to finish 8-0 to even have a chance to win at that point. Maybe it’s more fair that losses mean more at that point. Teams may not deserve to win if they drop 3 games to begin with. But I would argue it’s more healthy for more teams having incentive to try hard late in a season because of playoffs. In Conference B of Division 1 this season, every team in the conference had a non zero shot at making the playoffs before the final week, for example. Only 2 teams had a shot to win the conference, and one team had like >95% odds (wf +1.5 games). I suppose you could argue that playoffs aren’t the most purely competitive format, but I do think the community overall likes what it adds in terms of casts, motivation, and getting to have a definitive final match guaranteed. While we’re on the topic of rough starts though:



You’re correct in your evaluation that an 0-2 start will be more negatively impactful than it would be in S9’s format. But what I will say that the new proposed format is definitely much more similar to S9’s format than the previous Season 6 format. Let’s consider the example of having a 0-2 start and the effect it has on winning a division. The only way to win in the old format is to win a division. I took an arbitrary set of tables from WCL S7-S9 (33 results, a sufficient sample) and recorded the winning record of each:

x991edD.png


As you can see here, having a 0-2 start would lock a team out of winning in 20/33 ~61% of divisions in a 3 season span. Finishing 8-2 would not even guarantee a title also if you started 0-2, as 7 other teams won at 8-2. You would probably expect to win about 3/7 of these (since you would be tied with 1 other team, maybe 2 in these instances). So in reality, starting 0-2 you can expect to only have a 27% (1-24/33) chance of winning an old format after starting 0-2 - assuming you finish the season 8-0. Additionally speaking, it’s worth mentioning that half of the teams that won a division with a record worse than 8-2 were teams that beat a team with a better record due to bonus points (which are now gone completely).

Now even though this new proposed format is naturally more punishing on 0-2 starts than this season, it is not going to be even comparable to the old format. In the new format, while maybe if you were to take data on teams that would start 0-2, they may not make the playoffs a much higher percent of the time (due to teams who start 0-2 just naturally being weaker in general), if you want to talk about their actual elimination, then starting 0-2 on itself if not an obstacle that is insurmountable. There’s not a lot of comparable data for 8 team divisions yet, but using 4 I remember (PL S11, MKU DX S8 D6, MKU Preseason D2&D3) the average amount of wins for the 4th slot when stretching seasons to 10 games is about 6.2 wins, which rounds down to 6. I know it’s not a lot of data, but mathematically, it makes sense for a 4th place team to go about 6-4 in an 8 person 10 game season. It is much more reasonable for a team that starts 0-2 to finish 6-2 (6-4) than it is to finish 8-0 (8-2). It is also significantly more likely. There’s only one way to go 8-0 - win all your games. There’s 28 different ways you can go 6-2 over the last 8 games of a season. While some ways may be more likely than others, there is undoubtedly more flexibility when salvaging your season at this point than with the old format. So while a 0-2 start may be demotivating, it is really still not comparable to the league format, despite the equal game counts. One complaint is you may have to play the toughest teams first. This would be unlikely to say the least in this format, because that would be implying the best 2 teams are in the same conference. It’s possible due to staff error, but definitely less likely than not. Even if it were to happen, your schedule would get significantly easier afterwards. Take a look at HD in MKUDX S7: we started off 0-3, and had to play the top 2 seeds of our conference next. BRUTAL start. After Week 5, we were 6-4. The second we are in the playoff picture, that 0-3 start no longer hinders us. It still would haunt us in the WCL format.

Alright, so here is the bulk of the post, because Felix brings up lots of good points and I think I need to highlight / emphasize a lot of parts of this post separately. Let’s start with the main point:



To start on this case, you bring up some good points here. While your example using this seasons Division 1 is fine, the conclusion you reach here is quite dishonest and misleading I believe. What you are doing here is painting a picture that HD got screwed over, because they didn’t get to play RK twice. But what you’re not mentioning is that other matches matter. You are saying that HD played at a disadvantage, but when considering all matches, you’ll see that this isn’t even the case. To measure strength of schedule without using seeds but rather results, it’s better to look at the record of your opponents. In this case, instead of just looking at how many times HD plays RK, let’s take an evaluation of strength of schedule between Prestige and HD overall, both of who placed 2nd in their conference. Let’s count the points of their opponents in this season, using ties as a half point:

Prestige: 7+7(EU)+7+7(Re)+0+0(RK)+7(wf)+6.5(HD)+3(Bz)+2.5(ARC)=47

HD: 7(EU)+7(Prestige)+7(Re)+0(RK)+7+7(wf)+3+3(Bz)+2.5+2.5(ARC)=46

So as you can see, even with HD’s lack of 2nd game vs RK, strength of schedule suggests that HD had an easier schedule than Prestige/Re did. We would not be justified to complain like you’re suggesting. How could this happen?

Well, let’s consider this also. You’re claiming that Prestige had an advantage because they got to play RK twice. Well, let’s take a look at both teams records vs RK and ARC this season, who finished 4th in your standings:

Prestige 3-0
HD 3-0

So, nothing changed. How about we look at the records vs Bz and Re?

Prestige 2-0-1
HD 2-1-0

So HD’s competition to actually finish in 2nd of their conference was likely easier. Want to know the difference? We lost to Prestige and EU, and only tied Re. That’s why we’d be left out of the playoffs in this scenario. For having a 1-1-2 record vs the other conference. And that is on US for not winning these games. Not because we didn’t get an extra game vs RK. The whole premise of this main point is based off a logical fallacy.

Take this chart to further explain the dilemma we’re discussing right now:

w2Db8bl.png


Now, we’re not actually going to ever know how the division layout will happen. Logically speaking, the top seed will have the easiest schedule because they don’t have to play themselves. The point is though that the average difficulty of a schedule based on seeding is effectively going to have the same overall average: middle of the middle seeds. As you have more and more divisions recorded, naturally it’s going to work out that 2nd seeds in both conferences are going to have slightly different roads. One will generally have a harder 3rd seed and easier 4th seed than the other in their division. There will be instances where results differ and make divisions look wack. I think this is the point of competition though. For teams to exceed expectations, not to just finish exactly as predicted.

Now, there’s going to be scenarios where this will be at times more unfair than others. In 8U, when svc, FA, xi, dominated the field, svc would appear to gain an advantage from the other 2 naturally having to play each other more. But then the winner of the FA/xi conference would actually get the easier playoff match. We’re never going to know where there’s more massive gaps (in this case, between the 3 and 4 seeds). Maybe sometimes the gap is somewhere else. Maybe you get a clean 8 team division sometimes all in the same ball park. We won’t be able to control this. Luckily now though, the top 8 is much more balanced than in 8U. Can seeding further effect how it all plays out? Sure, it can. It can make some conferences harder than others for sure, you’re right. But it’s not like the team in the opposite conference doesn’t have control of their fate. They can always just beat teams in the other conference.

Now, I don’t know how you say this ruins the league also. Because the formatting is not that much different. I understand your argument, but let’s not pretend that this can’t happen with 12 team divisions as well:

unknown.png


Additionally speaking, when you have an imbalanced win total between conferences in 12 team divisions, the win totals are much more absurd in scenarios like this. Strength of record example:

YF: 136
FE: 118

For perspective, YF on averaged played a team with one more win than FE did. YF won more games. FE hosted a playoff game. YF was not in the playoffs. I get that it's a little bit more likely to happen in 8-team division problems, but this is also a flaw of the 12-team format.



I understand this point, and it is valid because you are correct that this scenario is more likely to happen in the new proposed format. I would like to ask you a strange question to consider at this point, but it ties back to something earlier in the post. Do you prefer the format we’ve been running the last few seasons, or do you prefer the original MKU format? If the answer is the WCL format, then that is fair, but I would present you the same argument that I presented Shin. If the whole community preferred to go back, it would be one thing, but I think the conference format is overall preferred. But if your answer is you prefer the 12 team conference format, I’d like to ask you about why - because it is a more unfair format that doesn’t reward consistency as much. It just has other perks that justify running the format. Maybe you agree with me that the old MKU format is more fair. But I think to imply that the current format is fair and that this new format isn’t fair would be a little bit ridiculous. I would say that you have the opinion of the top part of this image, where I find the bottom to be more accurate.

u2X0LVN.png


I will give you that 12 teams probably creates a more fair tournament. But I would also give you that WCL format is the most fair tournament. But like the reasons above that 12 team format is worth changing from WCL, I think there are still reasons that exist for changing to 8 teams as well, such as overall skill gaps in divisions shrinking, and giving people more tournaments overall (including more separate MKUs). At the end of the day we have to ask where we draw the line. As long as a tournament is to the right of the valid tournament line, I think it is worth considering. An example of something to the left would be Amplify V (also admittedly, this was my fault for suggesting we could make Swiss format work with 40 teams. I’ll 100% take the blame on this and it won’t happen again) because the scheduling was unfair enough that if the tournament was repeated an infinite amount of times, the tournament would offer a systematic advantage to 5 seeds. What I think is being suggested with an 8 team tournament may be to the left of 12 team, but is still easily a valid tournament. I believe it’s valid because as you run it an infinite amount of times, the trend comes out to be overall fair (it may have some faults, here and there because of seeding or because of gaps between a 3rd and 4th seed that may disappear over time as well). I think you’re suggesting that it isn’t valid at all, and I think to suggest that the 12 team MKU can be valid but the 8 team not is honestly, incorrect. The flaws still exist in the 12-team format.

To re-emphasize, I’ll give you that 12 team MKU is a little bit more fair than the 8 team MKU. But I think you’re being a little bit disingenuous with your evaluations of unfairness in the 8 team MKU idea. To start, you say in your example of the 2 worst teams being in the same conference will let 2 teams play 40% of their matches against the worst 2 teams, and say this is almost half the matches. 40% is a lot, I’ll admit, but it’s not a 40% increase over what we have right now. It’s a 15% increase over 25%. It’s also a large possibility (maybe larger, but I don’t know for sure) that the two worst teams go to the same conference in our current format. The number 40 is large, but 25 is still significant. We could also bump divisions up to 16 teams instead of 12. This way, if the two worst teams were to be in the same conference, you would only play 4/22 matches against the bottom 2 teams, which would be even lower. But in doing this, we’re now also making teams we previously considered bottom of Division 3 have to play the top seed of D1. At some point we’ve got to draw a line on the skill gap, or to just not have divisions to begin with. Really, no matter what you pick, you’re sacrificing another point of argument.

To continue on with this point, the issues of imbalance may occur a little bit more in 8-team format than 12-team format. But to act like these problems are exclusive to 8 team by calling it unfair and not 12, it’s just not true. For instance, division balance may not always work out in 8 team divisions. Is that problem exclusive to 8 team divisions? No. A matter of fact, as seedings are messed up as you mentioned, there will naturally be really rough scenarios. Larger conferences also run the risk of being imbalanced beyond what a smaller conference could even be in terms of imbalance. In just 15 attempts (not exactly a lot) at conference format with 12 team divisions, we’ve had the following occurrence, as well as the one from earlier.

E5f0Xwc.png


It’s worth noting that in a 8-team format, as long as a conference has at least a 4-16 record against the other one, the balance would literally be better than this example. Considering it is very likely for each conference to have a single top 2-3 seed, getting just 4 wins against the other conference is not really tough at all. Here, Conference A only won 39% of the total games in the division (the absolute minimum is also 31%, not 0%). It would take only a 4-16 conference record in the new format for a conference to win 40% of the total games. In an optimal world, it should be close to 50 for each. Overall, the point is that while it’s slightly easier to have conference imbalance, the levels of imbalance have slightly more potential to be disastrous in 12 team formats.

The big worry for me here is that in every case now, if one season has a case of misinformed seeding, the new format is going to be blamed automatically, when it is not necessarily because of the new format. There always is a chance that it is the fault of the new format, but as mentioned, a lot of the problems from the 8 and 12 team formats aren’t mutually exclusive.



Again, you’re right that teams can be easily misjudged, maybe we don’t know about a player for instance. It may be a tough request, but I do think we can collectively consider more data on seedings, or request the staff team takes more time. It’s honestly possible that some unknown players shouldn’t necessarily need to be unknown for instance. But yes, the odds of it happening increase the lower the division you are. But it is also worth noting that if there are few unknown players / seeds / scenarios where teams may become a victim of unfortunate unfairness, would you rather have 11 teams affected per mishap, or 7? We’ve both acknowledged that there will be some flaws in every seeding no matter how big or small the divisions are, but would this not limit the amount of teams exposed to bad seedings? Because the only true way to get out of these cases is to have one large division, which would defeat the purpose of MKU. Additionally, what is interesting is the new formats actually protect against this stuff better than the old WCL format does. The old format could be more fair from a tournament standpoint, but the old format wasn’t protected against tournaments where the team listings in each division were never balanced to begin with. At least you can play OP teams in a playoff series now. But anyway, OP teams will typically affect less teams in a 8 team format, unless there's so many mishaps that it pretty much makes up the whole league outside D1 (I don't think this has happened yet).



You make a valid point regarding RK and Novae, especially Novae. I honestly think that this case is a bit of an outlier. And as you mentioned, there’s going to be nothing we can do about the occasional outlier. To test if this is truly an outlier though, I went through the playoff success vs seeds for all of Season 8, and made a scatterplot.

66n8cwY.png


Only 1 12th seed made the playoffs last season, and it was a Round 1 exit. Now, aside from the crazy season that 5 seeds had (which I really doubt will happen again soon), the fact that you can see a very clear negative correlation here indicates to me that the staff, at least for Season 8, did a fairly good job of seeding the conferences. Good enough that I personally trust that mistakes like underseeding Novae are going to be quite rare. At the end of the day, they made the right call to put them in D2 as well. As for RK, perhaps we got some new information. Had we ran an 8 team division, we probably would’ve not thought any better. But this is going to happen with ANY kind of cut off whatsoever. Let me bring up an example.

Last season in Division 2, JPP1 and wf2 were the D2 wild card teams. Wf2 finished a game above JPP in the standings, but ultimately JPP won an extra game in the playoffs. JPP was seeded 12th this season, and wf2 was seeded 13th. Now wf2 is having an ok season in D2, don’t get me wrong. They are a 9-1-6 wild card team, and did win a playoff series. But up to this point, DY has been the team dominating D2 this season. JPP was not expected to do so well in D1 this season, but they won a pretty respectable 5 games. As a 12th seed, winning 5 games is generally going above expectation, and especially in Division 1. Pretty much, the staff’s 12th seed overperformed and the 13th seed underperformed (though I suppose they could still make a run, it just looks like DY is the team to beat). If it wasn’t for a single playoff match, these seeds are flipped and perhaps JPP obliterates D2 while wf2 struggles in D1. But we won’t know for sure. If we’re just running 8 team divs, we don’t have to worry about this question, and no team ends up ultimately in a D1 that they have little chance to win in at the moment since the skill gap is a bit larger. The point is, any cut off is going to introduce new potential problems that we have to worry about. You can argue that having more divisions will create more cut offs, but I’ll counter argue that this will be made up for with general skill gap balancing.

Something else you’re correct on is the scatter plot shows that it is difficult to actually nail seeds down to the exact number. While the staff tends to get the actual divisions right, conference balance is tougher. In Season 8, the conferences containing the 1 and 5 seeds definitely were a bit stronger overall. These things can happen, especially more near the mid/bottom divisions. What this scatter plot also shows though is that it happens in 12 team divisions as well. It may have a larger effect on 8 team divs at times also. Ultimately the discussion comes down to whether or not the percentage increase is worth the other benefits. I’d argue yes, because like I mentioned, I think that this is still more than a valid format. I get the feeling that you probably won’t budge, but regardless I hope this at least explains clearly why we want to try this format. I think while some of the criticisms are valid, many of them are blown out of proportion.

Additional Questions

In reality, promotion / relegation matches never were a very good way to determine who should be where next season. Not only do teams change significantly in offseasons with player additions and subtractions, rendering past results sometimes not very accurate for the future, but the system is a little weird at times. In division 3 this season, a relegation match would exist between a 7-9 team and a 1-15 team. If the 1-15 team wins 495-489 or something, does that suddenly change your mind on who should be relegated next season? Overall, it’s not really a consistent system and I think the staff would have better seedings if they just focused instead on results while considering player movement and offseason improvement too.



I think it was mentioned already, but the purpose of the leagues is to go all out with your best squad available. I think a lot of this response is acknowledging that a skill gap problem exists also, if anything. If teams are using their weaker players in matches intentionally, it’s likely that teams are not taking certain opponents as seriously. If anything, this is suggesting that divisions are gaining enough of a skill gap that they feel like they don’t need to worry risking losing matches even with throwing out their weaker players in the lineups. I would say most people would acknowledge that sandbagging while still expecting to win has no business in competitive leagues, especially ones where divisions exist with the intention to pit teams against rival teams close in ability.

I don’t think a shorter format should demotivate people also. The next tournament, that brings a fresh start, will be sooner as a result. I don’t believe it would ruin the impact of winning either. I’ve witnessed teams being super excited / putting banners in signatures over winning 10 match divisions for over a decade now. It doesn’t mean any less, and now there’s more chances to win, if anything. In a 72 team league, with the current format we would have 18 division champions a year. 18 different chances for people to win a MKU division every year. If we do the proposed format, this number doubles to 36 division champions a year. With more chances to win, it’s also possible that a side effect of this is less people trying to cheese the system by trying to get underseeded, since there’s now more opportunities to win to begin with.



This honestly is not off the table. It could even be the happy medium between slightly better skill gaps and making the playoff system closer to the balance of 12 team divisions. I’m not sure. It would be a weird season though, because this would force us to have at least 1 inter conference match a week, and the schedule would feel more random as a result. This is because if A1 plays A2, A3 plays A4, B1 plays B2, B3 plays B4, the only match left in this week is an inter conference A5 vs B5. I don’t think there’s anything inherently wrong with this idea. As discussed earlier, the format should be valid I think. If interest for this happens, it will be considered. 4 or 6 teams in the playoffs would feel quite weird too.



Absolutely false. The metric system is just better. It's base 10.
Thank you Sam.
 

Troita

Member
Founding Member
Alright, so I’ve spent some time reading up everything on the thread and want to try to clear everything up and give my insight on what lead to these decisions, the logic behind it, why we thought it was a change worth making, and wrapping up I guess the bigger picture that Teeples mentioned. To start, I don’t expect anyone to read all of this, maybe just specific responses, but just to note, I’m quoting these out of order because the response is going to tie all the questions regarding a format change together. If you want a full understanding of the situation, it’s highly recommended to read through this at least the full section regarding the format, because I think it’ll tie up many loose ends / questions.

Additionally, I want to start by saying that the staff only acts in an attempt to benefit the community. Community satisfaction matters a lot to us, so I appreciate everyone for commenting and even expressing displeasure because it’s better than just not saying anything. If a format results in being incredibly unpopular, then it’s something we won’t continue. But like everyone else, staff talks about long term planning for the site and we have our own ideas that we believe in, and think will be successful. It is not our goal to hide anything or to control scheduling in a way that the community will not want, but of course there are changes we think will ultimately lead to improvement in how we run our system here on MKC. So for all the questions, I’ll give an answer on how we got to where we’re at now, and I hope you all will give it a read and understand our perspective here. I think for the sake of the future, we should start doing more feedback on each event, because people rarely do mention how much they like certain formats. Sometimes we have to take it as a lack of complaints. If you guys like a format / tournament in the future, please do speak up about it, we’ll probably try it again if/when it’s the case. I personally read as many suggestions as possible in #tournament-suggestions or even things written in general channels, so it’s worth noting. Maybe we’ll try doing more surveys soon. Anyways:

Why we want to change up the format again, how it all relates to site planning / scheduling

The easy answer here is that we refuse to get complacent. Obviously successes are going to be noted. But complacency is something we don’t want to fall into. The classic WCL format was a staple of the community for maybe 11 years. It was a working format, and people did like it. But last year, the MKU Staff decided to try something different that we think the community actually likes more now than the classic WCL format. If they didn’t try something different, maybe we miss out on something that the community actually does like. There is going to be hits and misses. If we think we can make improvements, we’ll try it out. If something doesn’t work out, it doesn’t mean that we can’t go back. Admittedly, the WCL format style of competition is very strange in the world of eSports. It was created in 2008 and was considered the perfect kind of thing for the MK community. As the time passes on, it's good that MKU is still popular, but we also want to modernize as well. In most eSports, the most competitive, highly regarded events don't have divisional formats. Anyone who enters a tournament could technically win it if they play well enough. It's obviously very important to have tournaments where newer players have a reason to compete, but it is very abnormal for a main event to be broken down into skill tiers. It's not a bad thing that MKU popular, and as long as it drives site participation, it's something that's necessary for the site. But we also think exposure to new kinds of ideas is really important for the community's future.



Like mentioned above, all suggestions are considered. It happened that in this offseason, we had a decent amount of people suggesting 8-team divisions on the basis of the skill gap being pretty large between a 1st and 12th seed of a conference. Here’s some instances of it:

g6bY4Iv.png


ag6mNdl.png


PUXlZHn.png


dE2qSdX.png


V65BZb6.png


ttoVq7J.png

So of course, this was always something that was going to be considered from our end. It also happens that when thinking about it more, the only thing that I could conclude is that perhaps a lot of good side effects would also come along from making this change. Before explaining that in detail, I want to bring up two things brought up in the same conversations back near the turn of the year.

9bc1IPA.png


0tetY9C.png


Pretty much, when Teeples says bigger agenda, the following is what he is referring to. In the first picture here, it’s mentioned that MKU Staff are saying that the rate of teams actually wanting to be moved higher has increased from 8U, perhaps due to the PR system we’ve been using. Admittedly, I don’t believe the PR has been as popular as I was hoping for on implementation, but we may get there eventually. For now, perhaps having an effect on teams craving tougher competition is justification of its existence, because it was a big goal of the PR idea to begin with. But additionally, the PR in a sense has been a way to keep a yearly flow of tournaments, divided into separate but consistent periods of time. And this is where MKU comes into the picture. We’ve ran 3 MKU Seasons of 10 Weeks the last year or two, which comes out to about 30 weeks. However, if we’re dividing our calendar years into two halves, MKU simply doesn’t work with that idea at all. Because 3/2 = 1.5, you’re going to have one MKU season in the middle of the year, in the middle of the evaluation period split. So logically, we needed to split PR seasons into three per year to accommodate properly; because if you split it 1.5-1.5, one half of the year has more tournaments / results than the other half (since results from a tournament starting in May would be going to the 2nd half of the year), and if you split it 2-1 or 1-2, then one of the MKU offseasons would be absolutely massive. It’s also a possibility to have 4 seasons of current MKU format a year, but if that were to happen, we’d essentially be getting rid of all non MKU team activities, with very little room for flexibility. I feel like the biggest fans of MKU may not even want for that to happen, as it deletes the possibility of new ideas or even returning things like a World Cup in future years. To me personally, it feels like backwards progress, like going back to the WCL era, albeit with a better format.

So under this logic, if we decide to continue with dividing events into PR seasons, we’re forced to go with 3 per year with the current MKU format. What this essentially means is that we have about 17 weeks to work with each, and MKU takes up 10 (or 11 now) weeks each. This means we generally have 6-7 weeks to run 2 additional team tournaments, and we don’t have much room for breaks in between. A lot of teams I think get exhausted with all the matches so close to each other, and I think the flow personally is not great. If we were to run an Amplify (3-4 weeks), then there is quite little time for break. Even this MKU Season will be ending in April, and we have not had one single Sunday off since the beginning of the year.

To my surprise, running an 8 team MKU really does open up the schedule / possibilities much more. Not only could we go with 6 month evaluation periods with ease now, but we could pretty easily do it without necessarily the need of cramming 2 events in between every MKU season. I think as a result, we could have a lot more flexibility in scheduling, perhaps more opportunities to try new things without worrying about confining it in a span of 2 weeks, and that perhaps events would feel more fresh. Sometimes we could have only one tournament in between MKU seasons, without the need to cram 3 per every 4 months. The longer an event, the less flexibility we have planning stuff around it pretty much. The larger the window, the more flexibility we have. We’d have shorter offseasons - if you take a look at Gumer’s message, this really stemmed off of the 3 MKUs a year scheduling. You could always decide to do away with a PR / scheduling system I suppose, but at risk of losing the perks Teeps mentioned as well as sacrificing some site organization overall. And the best part is that we’d only be losing 3 MKU weeks a year by changing, which is almost negligible from a yearlong standpoint. Despite losing a very small amount of MKU weeks a year, the proportion of matches with extra meaning (playoff matches) goes up. On top of this, this offers doubles the chances to win a division in a year, or more chances to win a title, etc. Here’s a head to head idea of what I mentioned:

NgF7KI7.png


Personally, I don’t believe overall that shorter seasons will hinder the quality of the tournaments to teams. It is still twice as long as really any other site tournament. The only true way to find out is to run the season and see how people feel I guess. Thinking about it, I would think those near the top of their divisions will still have a rewarding experience whether or not they win a division in 7 weeks or 11 weeks. Those teams who have a rough season however, will get to escape and try something new much more quickly. With the new format, those teams who have a bad Season 1 would already be done with their Season 2 before the Season 2 in the current format even begins. I think there has to be some positives overall here to consider stemming from ultimately a discussion people were having regarding lowering divisions to 8 teams each. In no way do we want to downplay MKU’s importance, which is why we wouldn’t consider this if it didn’t benefit the league itself.

Another potential positive is how it would affect the MKU Primetime streams. Disclaimer, this plays no part in any reasoning, because casts are not worth even a slight dip to the health of the league, but is just a small additional positive I felt would be neat to point out. We had a big problem with the casts this season because only half of Division 1 was able to participate at the later window (6/12). By projection, this would change to become (⅝) next season, with 5 of the current playoff teams being able to play in the later slot. We’d likely have the following comparison:

p9iL5Rk.png


Now, the thing that jumps out is that we have less castable wars, and hence less options for community entertainment. But this is actually not a bad thing, because even though there is less eligible wars, we would have more per week to choose from because the teams that are able to will play each other at a higher frequency. Additionally speaking, many eligible wars were not chosen despite existing because based on the standings, a lot of wars were deemed not good matches to cast. The decreased skill gap would also make the percentage of eligible wars that are considered ‘good’ higher. Additionally speaking, ARC would be one of the 3 teams who aren’t on the eligible list of teams, and most of their roster is able. They ended up being in a cast this season. Even if they just had 1 week of eligibility again, it would increase these numbers even further because roughly 71% of their opponents are able to as well here.

TLDR; shorter offseasons, more seasons/results, less bracket tournaments piled on top of each other.

But let’s move on and start talking actual format now.

Discussion of the Format Itself, Valid Competitions / Fairness

To start, I agree wholeheartedly that Option 2 is wack because of the lack of bracket resetting. But MKU has started a trend of BO3s in playoffs, and BO3s don’t work well with resets in MK (potentially 6 matches in a row for a grand final, or two days of grand finals is kinda cringe, and even at that point playing a team 6 times in a league format is literally unheard of and has got to be super draining). This is why I think option 1 is better, even if it looks “less hype.” Besides, most of our tournaments outside of MKU run on double elimination already. In a sense this gives MKU “identity” (being the exclusive league + BO3 bracket tourney). BO3 also gives MKU some extra feeling of bulk compared to other tournaments, which I guess is fitting for the longest tournament on the site.

That being said, if inter-conference play is extended to 2 games, a new issue is created here: conferences are worthless now - because everyone would now have the same schedule. Thus, conferences would only exist as an extremely arbitrary ranking system, that would honestly be less fair than not having conferences. An 8-6 team may place 2nd in Conference B and get to host a 9-5 team who finished 3rd in Conference A, in the playoffs. Though this may already have been a case of unfortunate seeding, it becomes completely unjustified if interconference was 2 games, because the 9-5 team would’ve had a better record with the exact same schedule. So the solution at this point is if interconference is 2 games, you might as well drop conferences, and make it so it’s just double round robin. But then we just have the old MKU format on our hands. With playoffs. So pretty much the old CSL / WL format (I’m not sure if it still is) - but here is the new issue. We believe that this is a really fundamentally bad format, and it’s why MKU has never done the format or had playoffs until we introduced conference play. And here’s the explanation as to why:

It effectively screws over the team who has won the regular season, in short. Double round robin itself is a commonly used tournament format. We’d be running a double round robin tournament where the winner of the double round robin doesn’t get to then win the tournament. They would have played the EXACT same schedule as the rest of the league, and had the best record in doing so - then suddenly it doesn’t matter. You can argue that you can give them a bye I suppose, but ultimately, anything less than a title at that point is a bit of a middle finger to that team, considering they won the regular season.

What makes having conferences different is that nobody can actually be declared regular season champions at the end of the regular season. You could make the argument that this season, Requiem should be the regular season champions because they had the best record of 12-2-2 in Division 1. Except, it wouldn’t be a good / convincing argument. World Friend Star may have finished a game behind Re, but since their schedules were different, wf could maybe argue that they had a more difficult schedule. For example, HD finished high in wf’s conference, but we dropped our 2nd game to MT. Re didn’t have to play a 2nd game against MT. wf may say that the last place finishers in their conference were 6-10, while Re’s conference had a 5-11 and 0-16 team. Ultimately, there is no way to tell which conference is the superior one. They’re intentionally set up to be roughly equal. And this fact right here, gives playoffs a reason to exist: To prove who should be the champion, of a sometimes muddy picture. Double Round Robin is a FAIR tournament in itself - teams play the exact same schedule. When it is said and done, one team has won a tournament where every team had the same strength of schedule and same opportunities. This is why leagues such as the NFL / NHL have no issue crowning champions who may have not been the best team overall during the season. The schedules are never even and the same because of how divisions work. Look at the NBA right now - the season may be cancelled because of the coronavirus, and because the league is divided into conferences, fans can’t make a claim at who the champion should be. The Milwaukee Bucks have the best record in the NBA, but others are dismissing them because the Los Angeles Lakers have the best record in the conference most people consider to be better. If LA and MIL played the exact same schedule, we would have a definitive answer as to who should be considered the champion. But we don’t. Hence playoffs need to decide it!

It’s worth mentioning that playoffs probably are less fair than the old format, so I’m going to address this issue while responding to Shin.



This is an interesting comment, and it brings up the question. What do we value more? The most fair competition, or the most exciting competition? A little of both? Because you can make a valid point to say that the old WCL format (double round robin) is the most fair format we’ve had. There’s no bracket luck involved, it rewards consistency, and every team has the same schedule. Whoever wins played the same teams as everyone else and won the most among them.

Personally, I think it is not for the better. I would say it’s a bad argument to say “you can’t expect teams to be on their A game every week” to counter the WCL format. But though it may sound crazy, I think choosing the entertaining format may be for the better of the community than the possibly most fair competition. You can make an argument that as long as all the teams signed up for it and the rules don’t change, that a competition is fair. But maybe it’s true that we aren’t valuing consistency as much anymore. So why can I possibly think the less fair option may be better? Well, I think back to the best moments we’ve had in our MK8 history. I remember HD completing a 56 point comeback on the last race to win a MKU Tournament vs JBZZ. I remember wf bringing home their first MKU title on a final race vs DVP. I remember Toto getting billed off dNBC in the World Cup. Of the things you remember, did any of it happen in the MKU League format? Generally, I think it’s harder. It’s because playoff matches are definitive. They’re it. Winner takes all. The most memorable thing I remember happening during the WCL format in all of MK8 was the week that Alliance Rainbow took down Severance on Week 5 and the amount of thread views broke the site. Yet, it did nothing ultimately. AR didn’t win the division. They just had the lead one week. There wasn’t closure to it. It was just a good win, ultimately not much different than a scrimmage. You can always try to stack up the best/title deciding wars on Week 10, but it doesn’t guarantee anything. I remember Mp/xi and NF/svc as examples. When I was in Mp, we played a week 10 match vs NF (svc pretty much). NF was 9-0. We were in 2nd/3rd at like 6-3. We could’ve broken the 10-0, and maybe got 2nd. I was in the main lineup that season. It was a week 10 season finale. I chose to go to an in person MK8 tournament that was offered on my college campus instead. I won a coffee mug. I chose that over W10 of MKU between the top 2 seeds. Fast forward to 2017, xi plays svc in the grand finals of the MKU Cup I believe it was. xi had got upset by HD this tournament, a match we remember because wow, it sent xi to losers early. It had an affect. Then xi, down 30 in the last race of match 1, goes Mario Circuit and top 6’s svc to win and reset the bracket. It was madness. It was pretty much the same teams in the last match of 2 tournaments. One produced one of MK8Us most iconic final memories. The other I didn’t even bother showing up to play. It’s worth noting also that nobody can watch the rare W10 cast generally in the old format, because they’re all playing matches. In the current finals, most teams are actually out of their tournaments and free to watch also.

Besides just hype, I guess there’s more to it. There’s just more to play for each week. In the new format, a 6-3 vs 5-4 Week 5 primetime matchup may be more worth watching than a 9-0 vs 7-2 faceoff. In the WCL format, there’s no incentive caring about any matches outside of the ones affecting the division lead at this point. But in the other, there are teams near the middle who are still having impactful final regular season matches. Even for teams who don’t win a title, getting into the playoffs is an achievement on its own. I suppose that starting 0-2 in the old format is death sentence. It was demotivating. In many divisions you would have to finish 8-0 to even have a chance to win at that point. Maybe it’s more fair that losses mean more at that point. Teams may not deserve to win if they drop 3 games to begin with. But I would argue it’s more healthy for more teams having incentive to try hard late in a season because of playoffs. In Conference B of Division 1 this season, every team in the conference had a non zero shot at making the playoffs before the final week, for example. Only 2 teams had a shot to win the conference, and one team had like >95% odds (wf +1.5 games). I suppose you could argue that playoffs aren’t the most purely competitive format, but I do think the community overall likes what it adds in terms of casts, motivation, and getting to have a definitive final match guaranteed. While we’re on the topic of rough starts though:



You’re correct in your evaluation that an 0-2 start will be more negatively impactful than it would be in S9’s format. But what I will say that the new proposed format is definitely much more similar to S9’s format than the previous Season 6 format. Let’s consider the example of having a 0-2 start and the effect it has on winning a division. The only way to win in the old format is to win a division. I took an arbitrary set of tables from WCL S7-S9 (33 results, a sufficient sample) and recorded the winning record of each:

x991edD.png


As you can see here, having a 0-2 start would lock a team out of winning in 20/33 ~61% of divisions in a 3 season span. Finishing 8-2 would not even guarantee a title also if you started 0-2, as 7 other teams won at 8-2. You would probably expect to win about 3/7 of these (since you would be tied with 1 other team, maybe 2 in these instances). So in reality, starting 0-2 you can expect to only have a 27% (1-24/33) chance of winning an old format after starting 0-2 - assuming you finish the season 8-0. Additionally speaking, it’s worth mentioning that half of the teams that won a division with a record worse than 8-2 were teams that beat a team with a better record due to bonus points (which are now gone completely).

Now even though this new proposed format is naturally more punishing on 0-2 starts than this season, it is not going to be even comparable to the old format. In the new format, while maybe if you were to take data on teams that would start 0-2, they may not make the playoffs a much higher percent of the time (due to teams who start 0-2 just naturally being weaker in general), if you want to talk about their actual elimination, then starting 0-2 on itself if not an obstacle that is insurmountable. There’s not a lot of comparable data for 8 team divisions yet, but using 4 I remember (PL S11, MKU DX S8 D6, MKU Preseason D2&D3) the average amount of wins for the 4th slot when stretching seasons to 10 games is about 6.2 wins, which rounds down to 6. I know it’s not a lot of data, but mathematically, it makes sense for a 4th place team to go about 6-4 in an 8 person 10 game season. It is much more reasonable for a team that starts 0-2 to finish 6-2 (6-4) than it is to finish 8-0 (8-2). It is also significantly more likely. There’s only one way to go 8-0 - win all your games. There’s 28 different ways you can go 6-2 over the last 8 games of a season. While some ways may be more likely than others, there is undoubtedly more flexibility when salvaging your season at this point than with the old format. So while a 0-2 start may be demotivating, it is really still not comparable to the league format, despite the equal game counts. One complaint is you may have to play the toughest teams first. This would be unlikely to say the least in this format, because that would be implying the best 2 teams are in the same conference. It’s possible due to staff error, but definitely less likely than not. Even if it were to happen, your schedule would get significantly easier afterwards. Take a look at HD in MKUDX S7: we started off 0-3, and had to play the top 2 seeds of our conference next. BRUTAL start. After Week 5, we were 6-4. The second we are in the playoff picture, that 0-3 start no longer hinders us. It still would haunt us in the WCL format.

Alright, so here is the bulk of the post, because Felix brings up lots of good points and I think I need to highlight / emphasize a lot of parts of this post separately. Let’s start with the main point:



To start on this case, you bring up some good points here. While your example using this seasons Division 1 is fine, the conclusion you reach here is quite dishonest and misleading I believe. What you are doing here is painting a picture that HD got screwed over, because they didn’t get to play RK twice. But what you’re not mentioning is that other matches matter. You are saying that HD played at a disadvantage, but when considering all matches, you’ll see that this isn’t even the case. To measure strength of schedule without using seeds but rather results, it’s better to look at the record of your opponents. In this case, instead of just looking at how many times HD plays RK, let’s take an evaluation of strength of schedule between Prestige and HD overall, both of who placed 2nd in their conference. Let’s count the points of their opponents in this season, using ties as a half point:

Prestige: 7+7(EU)+7+7(Re)+0+0(RK)+7(wf)+6.5(HD)+3(Bz)+2.5(ARC)=47

HD: 7(EU)+7(Prestige)+7(Re)+0(RK)+7+7(wf)+3+3(Bz)+2.5+2.5(ARC)=46

So as you can see, even with HD’s lack of 2nd game vs RK, strength of schedule suggests that HD had an easier schedule than Prestige/Re did. We would not be justified to complain like you’re suggesting. How could this happen?

Well, let’s consider this also. You’re claiming that Prestige had an advantage because they got to play RK twice. Well, let’s take a look at both teams records vs RK and ARC this season, who finished 4th in your standings:

Prestige 3-0
HD 3-0

So, nothing changed. How about we look at the records vs Bz and Re?

Prestige 2-0-1
HD 2-1-0

So HD’s competition to actually finish in 2nd of their conference was likely easier. Want to know the difference? We lost to Prestige and EU, and only tied Re. That’s why we’d be left out of the playoffs in this scenario. For having a 1-1-2 record vs the other conference. And that is on US for not winning these games. Not because we didn’t get an extra game vs RK. The whole premise of this main point is based off a logical fallacy.

Take this chart to further explain the dilemma we’re discussing right now:

w2Db8bl.png


Now, we’re not actually going to ever know how the division layout will happen. Logically speaking, the top seed will have the easiest schedule because they don’t have to play themselves. The point is though that the average difficulty of a schedule based on seeding is effectively going to have the same overall average: middle of the middle seeds. As you have more and more divisions recorded, naturally it’s going to work out that 2nd seeds in both conferences are going to have slightly different roads. One will generally have a harder 3rd seed and easier 4th seed than the other in their division. There will be instances where results differ and make divisions look wack. I think this is the point of competition though. For teams to exceed expectations, not to just finish exactly as predicted.

Now, there’s going to be scenarios where this will be at times more unfair than others. In 8U, when svc, FA, xi, dominated the field, svc would appear to gain an advantage from the other 2 naturally having to play each other more. But then the winner of the FA/xi conference would actually get the easier playoff match. We’re never going to know where there’s more massive gaps (in this case, between the 3 and 4 seeds). Maybe sometimes the gap is somewhere else. Maybe you get a clean 8 team division sometimes all in the same ball park. We won’t be able to control this. Luckily now though, the top 8 is much more balanced than in 8U. Can seeding further effect how it all plays out? Sure, it can. It can make some conferences harder than others for sure, you’re right. But it’s not like the team in the opposite conference doesn’t have control of their fate. They can always just beat teams in the other conference.

Now, I don’t know how you say this ruins the league also. Because the formatting is not that much different. I understand your argument, but let’s not pretend that this can’t happen with 12 team divisions as well:

unknown.png


Additionally speaking, when you have an imbalanced win total between conferences in 12 team divisions, the win totals are much more absurd in scenarios like this. Strength of record example:

YF: 136
FE: 118

For perspective, YF on averaged played a team with one more win than FE did. YF won more games. FE hosted a playoff game. YF was not in the playoffs. I get that it's a little bit more likely to happen in 8-team division problems, but this is also a flaw of the 12-team format.



I understand this point, and it is valid because you are correct that this scenario is more likely to happen in the new proposed format. I would like to ask you a strange question to consider at this point, but it ties back to something earlier in the post. Do you prefer the format we’ve been running the last few seasons, or do you prefer the original MKU format? If the answer is the WCL format, then that is fair, but I would present you the same argument that I presented Shin. If the whole community preferred to go back, it would be one thing, but I think the conference format is overall preferred. But if your answer is you prefer the 12 team conference format, I’d like to ask you about why - because it is a more unfair format that doesn’t reward consistency as much. It just has other perks that justify running the format. Maybe you agree with me that the old MKU format is more fair. But I think to imply that the current format is fair and that this new format isn’t fair would be a little bit ridiculous. I would say that you have the opinion of the top part of this image, where I find the bottom to be more accurate.

u2X0LVN.png


I will give you that 12 teams probably creates a more fair tournament. But I would also give you that WCL format is the most fair tournament. But like the reasons above that 12 team format is worth changing from WCL, I think there are still reasons that exist for changing to 8 teams as well, such as overall skill gaps in divisions shrinking, and giving people more tournaments overall (including more separate MKUs). At the end of the day we have to ask where we draw the line. As long as a tournament is to the right of the valid tournament line, I think it is worth considering. An example of something to the left would be Amplify V (also admittedly, this was my fault for suggesting we could make Swiss format work with 40 teams. I’ll 100% take the blame on this and it won’t happen again) because the scheduling was unfair enough that if the tournament was repeated an infinite amount of times, the tournament would offer a systematic advantage to 5 seeds. What I think is being suggested with an 8 team tournament may be to the left of 12 team, but is still easily a valid tournament. I believe it’s valid because as you run it an infinite amount of times, the trend comes out to be overall fair (it may have some faults, here and there because of seeding or because of gaps between a 3rd and 4th seed that may disappear over time as well). I think you’re suggesting that it isn’t valid at all, and I think to suggest that the 12 team MKU can be valid but the 8 team not is honestly, incorrect. The flaws still exist in the 12-team format.

To re-emphasize, I’ll give you that 12 team MKU is a little bit more fair than the 8 team MKU. But I think you’re being a little bit disingenuous with your evaluations of unfairness in the 8 team MKU idea. To start, you say in your example of the 2 worst teams being in the same conference will let 2 teams play 40% of their matches against the worst 2 teams, and say this is almost half the matches. 40% is a lot, I’ll admit, but it’s not a 40% increase over what we have right now. It’s a 15% increase over 25%. It’s also a large possibility (maybe larger, but I don’t know for sure) that the two worst teams go to the same conference in our current format. The number 40 is large, but 25 is still significant. We could also bump divisions up to 16 teams instead of 12. This way, if the two worst teams were to be in the same conference, you would only play 4/22 matches against the bottom 2 teams, which would be even lower. But in doing this, we’re now also making teams we previously considered bottom of Division 3 have to play the top seed of D1. At some point we’ve got to draw a line on the skill gap, or to just not have divisions to begin with. Really, no matter what you pick, you’re sacrificing another point of argument.

To continue on with this point, the issues of imbalance may occur a little bit more in 8-team format than 12-team format. But to act like these problems are exclusive to 8 team by calling it unfair and not 12, it’s just not true. For instance, division balance may not always work out in 8 team divisions. Is that problem exclusive to 8 team divisions? No. A matter of fact, as seedings are messed up as you mentioned, there will naturally be really rough scenarios. Larger conferences also run the risk of being imbalanced beyond what a smaller conference could even be in terms of imbalance. In just 15 attempts (not exactly a lot) at conference format with 12 team divisions, we’ve had the following occurrence, as well as the one from earlier.

E5f0Xwc.png


It’s worth noting that in a 8-team format, as long as a conference has at least a 4-16 record against the other one, the balance would literally be better than this example. Considering it is very likely for each conference to have a single top 2-3 seed, getting just 4 wins against the other conference is not really tough at all. Here, Conference A only won 39% of the total games in the division (the absolute minimum is also 31%, not 0%). It would take only a 4-16 conference record in the new format for a conference to win 40% of the total games. In an optimal world, it should be close to 50 for each. Overall, the point is that while it’s slightly easier to have conference imbalance, the levels of imbalance have slightly more potential to be disastrous in 12 team formats.

The big worry for me here is that in every case now, if one season has a case of misinformed seeding, the new format is going to be blamed automatically, when it is not necessarily because of the new format. There always is a chance that it is the fault of the new format, but as mentioned, a lot of the problems from the 8 and 12 team formats aren’t mutually exclusive.



Again, you’re right that teams can be easily misjudged, maybe we don’t know about a player for instance. It may be a tough request, but I do think we can collectively consider more data on seedings, or request the staff team takes more time. It’s honestly possible that some unknown players shouldn’t necessarily need to be unknown for instance. But yes, the odds of it happening increase the lower the division you are. But it is also worth noting that if there are few unknown players / seeds / scenarios where teams may become a victim of unfortunate unfairness, would you rather have 11 teams affected per mishap, or 7? We’ve both acknowledged that there will be some flaws in every seeding no matter how big or small the divisions are, but would this not limit the amount of teams exposed to bad seedings? Because the only true way to get out of these cases is to have one large division, which would defeat the purpose of MKU. Additionally, what is interesting is the new formats actually protect against this stuff better than the old WCL format does. The old format could be more fair from a tournament standpoint, but the old format wasn’t protected against tournaments where the team listings in each division were never balanced to begin with. At least you can play OP teams in a playoff series now. But anyway, OP teams will typically affect less teams in a 8 team format, unless there's so many mishaps that it pretty much makes up the whole league outside D1 (I don't think this has happened yet).



You make a valid point regarding RK and Novae, especially Novae. I honestly think that this case is a bit of an outlier. And as you mentioned, there’s going to be nothing we can do about the occasional outlier. To test if this is truly an outlier though, I went through the playoff success vs seeds for all of Season 8, and made a scatterplot.

66n8cwY.png


Only 1 12th seed made the playoffs last season, and it was a Round 1 exit. Now, aside from the crazy season that 5 seeds had (which I really doubt will happen again soon), the fact that you can see a very clear negative correlation here indicates to me that the staff, at least for Season 8, did a fairly good job of seeding the conferences. Good enough that I personally trust that mistakes like underseeding Novae are going to be quite rare. At the end of the day, they made the right call to put them in D2 as well. As for RK, perhaps we got some new information. Had we ran an 8 team division, we probably would’ve not thought any better. But this is going to happen with ANY kind of cut off whatsoever. Let me bring up an example.

Last season in Division 2, JPP1 and wf2 were the D2 wild card teams. Wf2 finished a game above JPP in the standings, but ultimately JPP won an extra game in the playoffs. JPP was seeded 12th this season, and wf2 was seeded 13th. Now wf2 is having an ok season in D2, don’t get me wrong. They are a 9-1-6 wild card team, and did win a playoff series. But up to this point, DY has been the team dominating D2 this season. JPP was not expected to do so well in D1 this season, but they won a pretty respectable 5 games. As a 12th seed, winning 5 games is generally going above expectation, and especially in Division 1. Pretty much, the staff’s 12th seed overperformed and the 13th seed underperformed (though I suppose they could still make a run, it just looks like DY is the team to beat). If it wasn’t for a single playoff match, these seeds are flipped and perhaps JPP obliterates D2 while wf2 struggles in D1. But we won’t know for sure. If we’re just running 8 team divs, we don’t have to worry about this question, and no team ends up ultimately in a D1 that they have little chance to win in at the moment since the skill gap is a bit larger. The point is, any cut off is going to introduce new potential problems that we have to worry about. You can argue that having more divisions will create more cut offs, but I’ll counter argue that this will be made up for with general skill gap balancing.

Something else you’re correct on is the scatter plot shows that it is difficult to actually nail seeds down to the exact number. While the staff tends to get the actual divisions right, conference balance is tougher. In Season 8, the conferences containing the 1 and 5 seeds definitely were a bit stronger overall. These things can happen, especially more near the mid/bottom divisions. What this scatter plot also shows though is that it happens in 12 team divisions as well. It may have a larger effect on 8 team divs at times also. Ultimately the discussion comes down to whether or not the percentage increase is worth the other benefits. I’d argue yes, because like I mentioned, I think that this is still more than a valid format. I get the feeling that you probably won’t budge, but regardless I hope this at least explains clearly why we want to try this format. I think while some of the criticisms are valid, many of them are blown out of proportion.

Additional Questions

In reality, promotion / relegation matches never were a very good way to determine who should be where next season. Not only do teams change significantly in offseasons with player additions and subtractions, rendering past results sometimes not very accurate for the future, but the system is a little weird at times. In division 3 this season, a relegation match would exist between a 7-9 team and a 1-15 team. If the 1-15 team wins 495-489 or something, does that suddenly change your mind on who should be relegated next season? Overall, it’s not really a consistent system and I think the staff would have better seedings if they just focused instead on results while considering player movement and offseason improvement too.



I think it was mentioned already, but the purpose of the leagues is to go all out with your best squad available. I think a lot of this response is acknowledging that a skill gap problem exists also, if anything. If teams are using their weaker players in matches intentionally, it’s likely that teams are not taking certain opponents as seriously. If anything, this is suggesting that divisions are gaining enough of a skill gap that they feel like they don’t need to worry risking losing matches even with throwing out their weaker players in the lineups. I would say most people would acknowledge that sandbagging while still expecting to win has no business in competitive leagues, especially ones where divisions exist with the intention to pit teams against rival teams close in ability.

I don’t think a shorter format should demotivate people also. The next tournament, that brings a fresh start, will be sooner as a result. I don’t believe it would ruin the impact of winning either. I’ve witnessed teams being super excited / putting banners in signatures over winning 10 match divisions for over a decade now. It doesn’t mean any less, and now there’s more chances to win, if anything. In a 72 team league, with the current format we would have 18 division champions a year. 18 different chances for people to win a MKU division every year. If we do the proposed format, this number doubles to 36 division champions a year. With more chances to win, it’s also possible that a side effect of this is less people trying to cheese the system by trying to get underseeded, since there’s now more opportunities to win to begin with.



This honestly is not off the table. It could even be the happy medium between slightly better skill gaps and making the playoff system closer to the balance of 12 team divisions. I’m not sure. It would be a weird season though, because this would force us to have at least 1 inter conference match a week, and the schedule would feel more random as a result. This is because if A1 plays A2, A3 plays A4, B1 plays B2, B3 plays B4, the only match left in this week is an inter conference A5 vs B5. I don’t think there’s anything inherently wrong with this idea. As discussed earlier, the format should be valid I think. If interest for this happens, it will be considered. 4 or 6 teams in the playoffs would feel quite weird too.



Absolutely false. The metric system is just better. It's base 10.

Thank you for wasting your time for answering us with graphics/statistics to make us watch everything behind this new format
 

Monfi

Member
Founding Member
Thank you for wasting your time for answering us with graphics/statistics to make us watch everything behind this new format
It wasn't a waste of time. I was skeptical if its worth trying and i couldn't see any benifits doing 8 teams at first but Rookies post made me wanna try 8 teams next season. If it fails we can always go back to old format. People really need to stop being stubborn as hell!
 

Nato

Member
Staff member
Administrator
Site Supporter
Founding Member
It wasn't a waste of time. I was skeptical if its worth trying and i couldn't see any benifits doing 8 teams at first but Rookies post made me wanna try 8 teams next season. If it fails we can always go back to old format. People really need to stop being stubborn as hell!

nah pretty sure you misunderstood like I did at first, he's saying thanks for wasting your time making all these graphics and such to save us time.

he's saying rookie sacrificed his own time for the people
 

DeltaNA

long-time anchor
SUMMIT Staff
Event Organizer
Media Staff
Site Supporter
Founding Member
that we think the community actually likes more now than the classic WCL format
Im curious as to why this was never put to a poll, despite the arguably positive points listed later in this post
 

Rookie

Member
Staff member
Administrator
Media Staff
Founding Member
Im curious as to why this was never put to a poll, despite the arguably positive points listed later in this post

We'd always be open for discussion, but I really think it's not for the best for the future of the community. In a sense, it feels like moving backwards. It's possible to run it for lower divs I would think who don't want the conference format, but the inevitable seeding imbalance (unknown players) for instance I think damage the format more in lower divs than high divs. And I feel like a playoff system is easily more captivating for the highest level, so I really can't be sure.

May make a few more clarifications soon on the thread, but not certain when
 

TromTendo

Member
Staff member
Moderator
MKU Staff
Amplify Staff
200L Staff
Founding Member
I saw somewhere above, there was a mention of how the entertainment factor could play a role in whether or not we'll follow through with the new format, and also wanting to give teams a couple of weeks off from tournaments so that not every single weekend consists of gathering lineups and playing this game. It got me thinking, if we really value entertainment, how about we incorporate an All-Stars match right before playoffs begin? Something where the best players from Conference A face off against the best players in Conference B for one match for no other reason than to give the community a cast to watch and give most players that small break that is desired.
 

Devoty

Member
Founding Member
I tried to sum up as best as I could the informations from Wikipedia (https://en.wikipedia.org/wiki/Athletic_conference) and a short article http://dearsportsfan.com/2014/08/04/conference-sports/

Conference format was born thanks to geographical restrictions (sports teams don't want to travel too much during a season to save time and money) and sports organizations history (some sports organizations were in the past separated but then merged while each keeping a conference).

In Mario Kart none of these factors exist.
But our transatlantic friends want to prove (unintentionally or not) that they have the best format on earth.
 
Top