I have proposed this in the past, but I am bringing it up again now with the drama that has happened...

I would like to explore the idea of using a series of clan ladders for the clan league.

How it would work...

1) A ladder for each template (we currently use a round robin style tournament, now replaced with clot)

2) No divisions - All players are on the ladder and compete against players from any clan in the league

3) Limit each ladder to a set amount of matchups (maybe less for 3v3 games and more for 1v1s). Similar to a seasonal ladder, but no auto ending games (we have a buffer to make sure all finish).

4) Multiple 'teams' per clan on ladders - Each clan can have multiple people/teams on a given ladder (a set limit of maybe 2-3). Only the top 'team' in a season counts towards a clan score.


That is the basics. My reasoning for this:
-multiple teams helps prevent account borrowing, subbing, etc that are used to deal with boots/inactivity. You would have a backup plan
-multiple teams also allows for a clan to use more people in a roster, even if they don't count. they can get some experience, and they MAY count if the preferred team goes inactive.
-no divisions allows for a player in a 'lesser' clan to compete against the big boys/girls, even if their clan cannot compete for the overall title (think Olympics, where a smaller nation can compete in an event, but not for some overall medal count).
-ladders allow for eventual adoption on WL page by Fizzer. This is the ideal scenario where we aren't managing this on our own, but simply doing an aggregation of ladders on the site (and maybe he adopts that too).
-this also eliminates promo drama (clans dropping off, promoting/tie breaking issues). Everything is settled on the battlefield.

Before we get too far into a debate, I would REALLY like to get some help aggregating the current league teams, imagining IF they were on a ladder. Does anyone have the time to do this?

Basically, we'd throw all clans into made up 'ladders' for each template (we could maybe just do A,B and maybe C for testing purposes). Use all results from finished games, and for the non-finished games we can simply assign a winner based on what we think would happen. It doesn't have to be exact, the point is to see how the ladders MIGHT shake out without divisions.

Then, we can apply the same scoring system we have (points per win) and see how things would shake out. But, I'd also like to have someone like Math Wolf do an aggregation on it to see how things would look using the rating from the ladder instead of pure wins/losses.

Can anybody help with this?

Seems interesting, but the only issue is having such a huge gap in skill between clans means that the ranks and stats between clans will be extremely lopsided. It will help pad some of the top clans who will have no trouble destroying weaker clans and make it harder for clans to rise up and show their potential.

I do like the concept of using the Olympic model for clan league though. It's something that is rarely used so I hope this turns out well. Perhaps this can be like its own thing in the future.

Well the matchups would be based on rating, just like any other ladder. Just because a clan isn't good on all templates doesn't mean a player can't hang with 'A' clans. Take FCC. they are getting crushed in B, but I bet Aura could compete for a ladder title. I am sure there are other clans like that. Why should he be dis-incentivized to compete just because his clan is going through a rough patch? Why not allow him to get a ladder title, even if he needs more help for a clan league title?

Interesting idea.. I personally think it is better to have seperate ladders for each division. The reason is that: It is not my idea of fun beating 10-20 noob teams just to get enough rating to play decent opponents. 6 games in CL is quite time consuming enough because it is taken more seriously than ladders, let alone 20!

So, i guess what i am saying is while your suggestion is attractive for a clan playing in qualifiers, for a clan in division A, this system might be frustrating, since early games might not be competitive.

90

Edited 7/10/2018 17:24:19

Why would you play 20? Who said that number.

My thought was the ratings of a player would be used in the following season, only for initial matchups. So you wouldn't go through that placement phase again. Or maybe you use a clan rating from the previous season and apply it to all players in the clan, only for matchup purposes.


So let's say you give the Clan finishers this:

A1 - 2500
A2 - 2400
A3 - 2300
A4 - 2200
A5 - 2100
A6 - 2000
B1 - 1900
.
.
.

And all players in those clans get that rating to start out the next year, only for matchup purposes. That way you are most likely to get matchups with people in a similar finishing clan. It would not be perfect, but it would be close enough to get the right sample of games.

Right now you can have a player in C that is maybe top 5 on 1v1 ladder, but they are playing junk opponents because their clan isn't in A. Wouldn't you rather have them playing good players? In my scenario, you could use the previous season's rating for that player (and use the clan rating as a backup if someone did not participate in that template), to try and get better matchups.

Edited 7/10/2018 17:25:37

mm yes, your suggestion would fix that problem. And perhaps division alone can determine initial rating for the first season.

How many games do you think that each team would be playing totally?

Edited 7/10/2018 17:27:16

Also, I think there would have to be two leagues. A competitive one (limited to 15-20 clans only), and a secondary one that people could promote into the competitive one (maybe 2 clans a season). That helps filter out bad matchups and also prevents someone from forming a super clan and winning immediately (they'd need to win the lower league first).

Total games? I don't know. Maybe:

1v1 - 15
2v2 - 10
3v3 - 8

Something like that. You can say that seems a lot compared to current league (6 games only). But you can simply make the season longer so nobody is burdened. Clans are getting bigger and bigger anyway. Back in the day most in CL only had 5-7 active players.

You also don't have to have 18 slots like we have now. We could cut back to 10 like we had for most seasons (only the last few have had 18).

Edited 7/10/2018 17:33:31

Imagine clan fielding two (or more) teams where only one of them will be designed to fight for top spot and others just to lower rival's rating (so, doing the best against direct rivals and losing to any low rated clans).

That is possible I suppose. Any system has a way to game it.

You could simply not count ANY games from a non-counting team. But that would lower the sample size for the counting team's rating. Not sure if that is a great outcome or not. The people would probably be mad to have wasted games playing a team that won't count that way.

I ran through some pairing ideas last night and I don't think it's be too hard to figure out a way to get 7 pairings out of 16 clans in a way that tests skill well. You'd play about half of the other clans and results would be based on BayesElo. No pairing would be based on previous matches, so stalling wouldn't be an issue. Putting two teams in the same tourney causes additional problems with pairing and creates the potential for one team to throw games. You could calculate BayesElo by excluding all games from the clan's second team, to get around the game throwing issue, but the pairing issue remains.

I have a really cool idea for a "clan battle" - which is unlike anything anybody has ever seen before. It is not a league because it is not suitable for more than 2 clans - it is a 1v1 battle of clans. The idea in a nutshell is: "risk within risk".

Let us consider the case of a "2v2 clan battle". Each clan assigns a "general" - a leader of the battle. We begin with a map, and every territory on the map is randomly distributed to the two clans. The "generals" then assign teams of size 2 to each of the territories. Every turn, a general can attack an adjacent territory from one of it's territories. When this happens, a random RT game chosen from a collection of fixed templates begins between the teams located in the attacking and defending territories. If the attackers win, then they claim the territory, whereas if the defenders win then the territory is defended. When the attackers take a territory, that same team is now present in both the territory it attacked from, as well as the territory it took. In other words, the more successful a team is, the more territories it will occupy (and can attack from in future turns). Once the whole world is conquered by one of the clans, the clan battle is over.

Pretty cool idea right? :D

90

Edited 7/10/2018 22:08:36

@90, we actually a clan battle like that many years ago. It was a you describe, except that the board was an FFA rather than just for 2 clans. Someone less lazy than me could find the forum thread for it.

This is an awesome idea! Hopefully we would stick to normal Warzone's rules and don't overrule this with super rigid "ethic" views. I like the fact that multiple teams or individuals of the same clan could participate to the ladder without affecting negatively their clan in the long run. I totally agree that it would reduce a lot accounts sharing and stuff like that without eliminating it totally though.

Edited 7/11/2018 00:28:54

Hopefully we would stick to normal Warzone's rules and don't overrule this with super rigid "ethic" views.

...

EDIT

I would like to explore the idea of using a series of clan ladders for the clan league.

=> introduce 2v2 and 3v3 to AWP
=> aggregate points per clan
=> set a date to crown CL winner once a year

Way less fun than CL as it is right now in my opinion.

Edited 7/11/2018 07:41:32

@Beren - oh that is interesting. Actually, is is a viable option to run CL in this ffa "risk-within-risk" format, but to prevent clans from deliberately ganging up to focus down certain clans, we can make the rule that every clan must attack an adjacent territory of every other clan every turn, assuming such a territory exists. That would solve our problem of clans ganging up together like people do in ffa games and makes our league a fair event. That would be a really cool league :D.

90

Edited 7/11/2018 07:42:55

wow i think this would really be...Boring.

Why remove like the most competetive and fun we have and replace it with this? we already got to many ladders as it is.

I like the format where u can go up/down in divisions and i dont want it to be luckbased on who gets the best matchup and shit like on seasonal...

To determent who is the best clan i think the clans should face the same opponents!

And also if u have like 2 teams from same clan and lets say one team is ranked 2nd and other team is ranked 4th, cant they get a matchup against eachother? if yes then its easy to manipulate the ladder aswell. so there has to be a formula that teams from same clan cant play eachother, but then again thats not really fair for the other teams that they have to play both this two good teams.

Meh I just find this a bad step for CL :)

I agree with...Kezzo?!?!?!?

In all honesty, I like the idea. But a lot of the benefits of having a clan ladder are already present in the MDL.

In addition, Seasonal CL Ladders don’t fix the problem I still have with the current format of CL, which is the one-game-decides-all aspect. Most clans in each division can beat the other in at least 1/10 games. But basing off overall superiority on just one game doesn’t necessarily determine the better clan. Seasonal matchups don’t fix this. The only option I see is a continuous ladder, but, once again, MDL does a fine job already.

There is also the problem with multiple teams on the ladder. As far as I see it, it introduces more problems and doesn’t fix the existing one. For ex., the clan’s top-scoring team is in the Top 3, and their other teams are in 30th-31st place. There’s no realistic possibility that the other two teams can compensate for the top team if one of their members goes inactive. If subbing isn’t allowed, the other clan might just simply take over the account of that player. At least with the current CL format in place, there is absolutely no reason to take over an account. But here, I wouldn’t blame players for this because there really is no legal way to counteract a player’s inactivity anymore.

Like Kryzyetifrj said previously (and I said before in other discussions), the “backup” teams have no incentive to play other than to beat other top clan teams and attempt to tank their rating. And if you continue to add exceptions to the ladder rating system (making games with backup teams not count), it just becomes too convoluted to follow and destroys any advantages in utility the ladder might have.

Finally, this idea has the same problems that all new ideas face - it’s new. There will have to be a bunch of discussion, debate and analysis before anything can be implemented. As is, the current CL format is going through a major overhaul, which I believe is currently the best format. I don’t think we need a radically different change in direction at this point in time.

There is also the Swiss-system style for tournaments which could be considered for CL.
advantages: everyone plays in the same division
pairings can be randomly at start
only ld(clans participating) rounds needed to find a clear winner
(probably 5 for CL)
mayor disadvantage: a round needs to finish before the next can start

see: https://en.wikipedia.org/wiki/Swiss-system_tournament

Me and Ag always think alike! :)