match making algorithm not running for us since latest patch 5.7.

[alegs]

Hello Nexon,

with the latest release matchmaking is not working for us anymore. I will try to describe the issue as good as I can and I would be very happy if you take my observations seriously.

Please don't copy paste parts of the ingame manual here, I did read all of that and I'm well aware of how it did work in the past and also how it ought to work by design. I know about the calculation of strength and I know, that the pool of potential opponents widens and it is more likely to get a hard opponent if the waiting time is progressing.
But this does not match, what we are observing since the latest release.


Description of observations:
In all our world war history we got more or less well-balanced matchups: We started a new world war, the search ran for some minutes up to some hours until we got our opponent assigned.
We are a good organized group with 100% participation and well-designed bases, so we could win a lot of wars although some had been very demanding. Also, we lost some wars to even better alliances. All fine so far.
Until Patch 5.7. ...
Since then, we always are matched instantly without ANY waiting time for the war search. It appears, that there is no search run at all. The result is, that we get an opponent which is far outmatching our composition 90% of the time. We are well aware that this issue is not current for all alliances and appears to be specific for us. Our main alliance e.g. still has waiting times after starting the war search although they also have issues with the match making, I think these are not similar to our problem.

Please see our war history and your logs. A great example is the current war.

My alliance is: Bad Rabbits II (15.733 current Glory)

Sincerely
Alegs

 

[Imaera]

https://forum.nexonm.com/forum/nexo...-new-matchmaking-and-glory-observations/page5

https://forum.nexonm.com/forum/nexo...aa/588936-world-war-matchmaking-changes/page5

https://forum.nexonm.com/forum/nexo...e-get-such-a-screwed-up-matchup-what-the-hell

https://forum.nexonm.com/forum/nexo...is-getting-ridiculous-way-overmatched-in-wars

https://forum.nexonm.com/forum/nexo...es/592153-ww-matchups-are-getting-out-of-hand

As you can see, everybody is complaining, nobody is listening. Hope you have more luck

 

[Imaera]

Look here
https://forum.nexonm.com/forum/nexo...-new-matchmaking-and-glory-observations/page5
and here
https://forum.nexonm.com/forum/nexo...aa/588936-world-war-matchmaking-changes/page5
And there are many other similar topics. As you can see everybody is complaining, nobody is listening. Hope you have more luck...

 

[alegs]

Look here
https://forum.nexonm.com/forum/nexo...-new-matchmaking-and-glory-observations/page5
and here
https://forum.nexonm.com/forum/nexo...aa/588936-world-war-matchmaking-changes/page5
And there are many other similar topics. As you can see everybody is complaining, nobody is listening. Hope you have more luck...

Thanks, I read those and I also posted there essentially the same description as here, before i noticed, that my post here had been approved. Sorry for the duplicate...
But... thanks very much for the approval, which gives me some more hope that you are looking into this.

I understand that there had been changes to improve waiting times which had gone nuts for some alliances.They were waiting days for getting an opponent or even didn’t get one at all.
But we had waiting times of only some minutes up to some hours which was totally OK for us because in the end we got some more less matching opponent.

Now we have - ZERO - waiting time and totally overpowered opponents. War after war, after war... Starting exactly with the latest release.

This clearly indicates, that there had been a bug introduced and the code is not doing what the developers thought it should do.

Please look at your logs and please look at the instant matches and the rationale of the assigned opponents. I’m sure you will find the issue yourself.

Also I would like to encourage to be more transparent about the algorithms. Obscurity will only protect from exploitation for a limited time. A fraction of the users will always find out and abuse weaknesses in your algorithms. If you on the other hand have a solid algorithm which is improved in iterations, there is no reason to not make it transparent.

A first step would be to display player and alliances war strength in-game. You use it for matchmaking anyhow, so why don’t let the people know and be transparent about it?

 

[alegs]

Today we got assigned the first more less balanced matchup since the new release.
Please tell me you did fix something and this isn't a lucky coincidence!

BTW: I like the newly introduced AXP. Even if you are outmatched, you still got something to fight for!

 

[TinSoldier]

Hi alegs!

I agree, it looks like your Alliance has had a difficult run of match-ups since the new War matchmaking was added. Our team is doing a deeper investigation of how the war matchmaking has done overall, but that report isn't complete yet.

From what I can see, most of these match-ups seem unfair for one of two reasons.
1.) The Alliance you're paired against looks scary (higher Ages or levels - neither which directly influence matchmaking)
2.) Alliances that are the underdogs have a much stronger offense than defense. While this may help you attack up, it hurts when the Alliance favored to win attacks your bases.

This weekend, there is the Ikko Ikki event, which has discounts on Walls, Redoubts, and Castles and Traps. These upgrades should help raise your chance at victory when you are the underdog in a World War.
https://forum.nexonm.com/forum/nexonm-games/dominations/announcements-ab/592885-ikko-ikki-event

Also glad to hear you like the AXP! Alliance Perks should also help you out in World War.

 

[Imaera]

Hi , @TinSoldier
Sorry, but a few higher walls, does not help an industrial to win against an maxed out atomic. We need a real fix, not stories for kids. Hope what you wrote is some kind of stalling untill that fix is implemented and not a definitive solution.

 

[alegs]

Hi alegs!

I agree, it looks like your Alliance has had a difficult run of match-ups since the new War matchmaking was added. Our team is doing a deeper investigation of how the war matchmaking has done overall, but that report isn't complete yet.

From what I can see, most of these match-ups seem unfair for one of two reasons.
1.) The Alliance you're paired against looks scary (higher Ages or levels - neither which directly influence matchmaking)
2.) Alliances that are the underdogs have a much stronger offense than defense. While this may help you attack up, it hurts when the Alliance favored to win attacks your bases.

This weekend, there is the Ikko Ikki event, which has discounts on Walls, Redoubts, and Castles and Traps. These upgrades should help raise your chance at victory when you are the underdog in a World War.
https://forum.nexonm.com/forum/nexon...kko-ikki-event

Also glad to hear you like the AXP! Alliance Perks should also help you out in World War.

Thanks very much for the feedback.
I’m really happy, that you are indeed looking into this.
I’m willing to stay patient while waiting for improvements.

I can understand, that some issues are making people depressed, but not everything is bad in this game and I can say: I like it very much!
I’ve played a lot of games in my life, but since some reasonable time dominations is my favourite mobile game. I also do not feel, that it was a pay to win game like many say (I have played mobile strike…! ). In dominations you have the option to invest real money to evolve faster, but you do not have to do so and anyhow can stay competitive due to the fact, that you are matched according to your current age and medals in multiplayer. If you manage to get this working for World Wars again, that would be great!

If you like to have some crowns in the pocket to speed up things every now and then, you can go for watching videos or install some of the promotion stuff (with reasonable new media awareness!) to get some and also supporting Nexon by doing so. I like this option very much and think it is a fair concept for both sides.

Also, the exponential upgrade times, which some of the players obviously dislike, are OK for me, because this way you will always have something to go for and the game never ends because you are maxed and bored...

That’s it with the positive feedback for this one.

Regarding your answer, I would like to clarify, that I am very aware on how to judge the war strength of a base. I’m not blended easily by age, XP or medals at all:
I do ignore medals completely. Instead, I look at the offensive and defensive structures very carefully. I look at what units are in the barracks and I think about, if what I see matches the players displayed XP. Next, I look at the age of the player and check how far the economic buildings are upgraded (street, caravans and in special roads! due to the fact that they boost XP overproportionally). From these observations, I can estimate what amount of research has been done in the university. Finally taking the layout into account and I have a pretty solid picture of what I’m facing. Often, I’m very happy with going against a rushed global or atomic base with shiny roads as an industrial player!
This being said, I cannot confirm that reason 1 explains the issue we had with the latest wars, because I did the above analysis for most of the opponents.

Reason 2 sadly is also not the explanation. If we had had an opponent with a much stronger offence compared to his defence, we easily would have 5-stared him from top to bottom and would have ended up rather likely in draws decided by secondary performance indicators than losing by stars.
The same is valid if you switch the underdog role the other way round...

 

[deadrock133]

Use also have seen world war matchups being 90% where our alliance (find us) is over matched. The latest matchup is a good example. They have 2 atomic, four global. We have one global and the rest industrial and enlightenment. We will lose. Have asked for updating the matchups, Nexon says matchups are working as designed. Hope Nexon fixes this soon.

 

[Belissma786]

But does anyone know what is the logic for the matchups???? How do they rank it??

 

[alegs]

Matchmaking is still broken. For an example, see our current war. (BR2 vs. Durch WarlordZ, #1 181 vs. their 269 with 4 lvl 50 Generals and an obvious Sandbag...)
Strangely this time an equal ratio of 2-win/2-loose glory points, so perhaps this is not the best example and the distribution of player strength just is unlucky for us.
But, in most cases where we got bad matchups, the glory ratio was like 700/10. So, in these cases the matchmaking algorithm "knew" for sure that he was assigning a very unbalanced matchup. Why not just allow for some waiting time in those cases? We are fighting wars for 2x24h so i have no issue waiting up to an hour or so for a match, if I get something somehow fair and demanding.

You did a great job by introducing the missile silo and the stronghold. Both have shown to be a gamechanger and I'm sure the stronghold is more and more generating income. Especially due to the fact that the free troop tactic cards have been reduced in the last couple of weeks.

But... you need to address the algorithm for matchmaking. It is not doing what it is opposed to do. So please don’t lean back and tell us everything works as designed. You have the numbers that we don’t. It should be an afternoon job for someone crawling the tables and checking the issue.

I’m still lacking the transparency in the numbers you provide.
Why don’t you just list the war weight in the war tables and make the matchmaking transparent?

 

[Tiggo]

well the more they explain how matchmaking works the more it will be exploited...its a double sided coin ...

 

[alegs]

well the more they explain how matchmaking works the more it will be exploited...its a double sided coin ...

That is a widespread misconception.
If the functionality a concept does rely on obscurity, it will probably postpone the discovery of its weakness for a certain time. Once the weakness is discovered, news will spread fast and it will be abused widely with conviction of being unnoticed.
If on the other hand you are transparent about the concept, weakness might also be discovered, but in contrast to the above, you enable a well-informed discussion that likely will show ways to improve the concept in a way that the weakness is handled properly.
If you have a product with a very limited lifespan, you might opt for the first approach.
But this game is a long runner and therefore the second approach would pay off in the long term.